UC-NRLF S3b MANUAL OF WIRELESS TELEGRAPHY FOR THE USF, OF NAVAL ELECTRICIANS 1911 r MANUAL OF WIRELESS TELEGRAPHY FOR THE USE OF NAVAL ELECTRICIANS 1 MANUAL OF WIRELESS TELEGRAPHY FOR THE USE OF NAVAL ELECTRICIANS BY COMMANDER S. S. ROBISON, U. S, NAVY 2D REVISED EDITION ANNAPOLIS, MD. THE UNITED STATES NAVAL INSTITUTE 1911 CONTENTS CHAPTER I. GENERAL REVIEW OF FACTS RELATING TO HIGH-FREQUENCY CURRENTS. Electricity; Magnetism; Electro-magnetism; Electro-magnetic induction; Methods of producing currents by electro-magnetic induction; Method of pro- ducing currents used in wireless telegraphy; Electric and magnetic fields; Electric capacity; Electric induction; Electric condensers; Discharge of con- densers; Ether waves; Reflection; Refraction; Diffraction; and Production of ether waves. CHAPTER II. UNITS. Foot-pound-minute; Centimeter-gram-second; Power; Work; Dyne; Erg; Volt; Ampere; Ohm; Watt; Coulomb; Farad; Henry. CHAPTER III. CAPACITY, SELF-INDUCTION, AND MUTUAL INDUCTION. Fundamental equation of wireless telegraphy; Difference between D. C. and A. C. due to self-induction and capacity; Skin effect of high frequency currents; Self-induction and capacity of straight wires; Mutual induction. CHAPTER IV. ELECTRIC OSCILLATIONS. Radiation of electric waves; Damped and undamped oscillations; Wave trains; Work done in producing electric waves and making dots and dashes of the telegraph code; Detection of electric waves. CHAPTER V. SENDING CIRCUITS AND APPARATUS. Generators; Transformers; Regulation of sending sets; Sending keys; Closed circuit inductances; Condensers and condenser material; Spark gaps; Energy transfer between closed and open circuits; Limitations on wave lengths; Aerials (Antenna); Anchor spark gap; Lightning switch; Open- circuit inductances; Variometers; Hot-wire ammeter; Grounds and ground connections. CHAPTER VI. RECEIVING CIRCUITS AND APPARATUS. Open circuit; Closed circuit; Inductances; Condensers; Detectors; Tele- phones; Batteries; Buzzers; Ampliphones; Recorders. 8 CONTENTS. CHAPTER VII. INSTALLATION AND OPERATION. Installation; Adjustments; Calibration; Tuning; Wave meters and their use; Measurements; Care and operation; Codes. CHAPTER VIII. MISCELLANEOUS. Wireless telephony; Production of undamped oscillations; Directive wire- less telegraphy; Portable wireless sets; Static; Standard wave lengths. APPENDICES. INDEX. MANUAL OF WIRELESS TELEGRAPHY. Chapter I. GENERAL REVIEW OF FACTS RELATING TO HIGH FREQUENCY CURRENTS. ELECTRICITY. 1. If amber is rubbed with silk a change in the condition of the amber and of the silk is produced which can be detected in various ways. This change in condition is described by saying that the amber and the silk are electrified or charged with electricity by friction. Both of these terms are derived from the Greek word " elektron," meaning amber. The silk and amber thus electrified attract each other and bodies in their -vicinity, but the silk will repel another piece of silk similarly electrified and the amber will repel another piece of amber similarly electrified. Since amber and silk have no effect on each other when not electrified, the qualities of attraction and repulsion are said to reside in the electric charges, and the fact is expressed by the statement that like charges repel, unlike charges attract each other. The silk is said to be positively, the amber negatively, electrified or charged. Positive and negative charges are indicated by plus ( + ) and minus ( ) signs. The charges are said to consist of static or frictional electricity. Bodies thus charged when not brought into contact with each other or with what are called conductors remain in an electrified condition for some time. Bringing oppositely charged bodies in contact generally removes all evidences of electrification. The charges are said to unite and, being of opposite signs, to neutralize each other, and the bodies are said to be discharged. Sparks accompanied by a sharp crackling sound are produced between highly electrified bodies when brought very near each other. After the spark has passed the bodies are found to be discharged. Charged bodies which can be discharged by sparking at greater dis- tances than others are said to be charged to a higher potential- All bodies, whatever their nature, are capable of being electrified. The presence of static charges of electricity can be shown by what are called electroscopes. One of the most sensitive, the gold-leaf electro- scope, consists of two small pieces of gold leaf, which, becoming charged 10 MANUAL OF WIRELESS TELEGRAPHY. in the same sense (i. e., positively or negatively), by touching a charged body, repel each other, and diverge, and show by their divergence the presence of electric charges. 2. Certain bodies, notably metals, have the quality of transmitting or carrying electric charges through themselves and are called conductors. Bodies lacking in this quality, or possessing it to a very limited degree, are called nonconductors, or insulators, or dielectrics, according to the purpose for which they are used. 3. When pieces of zinc and carbon are immersed in a conducting liquid (fig. l).the combination is called a primary cell. If a wire is connected to the zinc and one to the carbon and the free ends of the two wires brought near each other, these ends are found to be electrified ; the end of the wire connected to the carbon electrified like the silk ( + ) and the end of that connected to the zinc like the amber ( ) . The carbon is called the negative element or positive pole of the cell and the FIG. 1. FIG. 2. zinc the positive element or negative pole. A number of cells together is called a battery. The liquid in which the elements are immersed is called the battery solution. If the free ends of the wires are brought together an electric current is established, of which the positive direction is" said to be from the carbon to the zinc, through the wires;. from the zinc to the carbon, through the liquid. (See fig. 2, and note 1, appendix.) The current is said to be caused by a difference of potential between the carbon and the zinc. It is supposed to be made up of small electric charges transmitted through the wire in quick succession, the charges being produced by chemical or electric action between the carbon and the zinc in the liquid. The force which causes the movement of the electric charges which make up the current is called the electro-motive force and is usually written E. M. F. If the free ends of the wire in fig. 2 instead of being directly con- nected are immersed in another conducting liquid, as in fig. 3, the cur- rent will flow through this liquid. The immersed ends are called electrodes. The one at which the current enters is called the positive and the one at which it emerges the negative electrode. These are also MANUAL OF WIRELESS TELEGRAPHY. 11 called the anode and the cathode, respectively. The conducting liquid in this cell is called the electrolyte. 4. If the anode and cathode in fig. 3 are made of lead (or prepara- tions of lead) plates, and the electrolyte is a solution of sulphuric acid in water, the combination is called a secondary or storage cell or accumu- lator and a number of such cells is called a storage lattery. The FIG. 4. anode is called the positive plate and the cathode the negative plate. If, after a current has been forced through such a cell for a time, the wires from the primary cells are disconnected and the positive and negative plates connected by a wire (fig. 4) outside of the electrolyte, a current will flow, the positive direction of which will be from the positive to the negative plate in the wire, and from the negative to the positive plate in the electrolyte. 5. For convenience, a battery of primary or secondary (storage) cells is indicated as in fig. 5, the elements forming positive poles by the light FIG. 5. Cells in Series. FIG. SA. Cells in Parallel. lines and the elements forming negative poles by the shorter, heavy lines. Cells connected as in fig. 5 are said to be in series; connected as in fig. 5a, in parallel. MAGNETISM. 6. A magnet situated at a distance from other magnets and pivoted so that it is free to move, will point toward the north magnetic pole of the earth, which in some localities coincides with the north star in 2 MANUAL OF WIRELESS TELEGRAPHY. direction. That end of the magnet which points in the direction of the north star is called the north-seeking pole, or simply the north pole of the magnet. The other end is called the south pole. Similar magnetic poles, like similarly charged bodies, repel each other. Dissimilar magnetic poles, like oppositely charged bodies, attract each other i. e., two north poles or two south poles repel each other : a north and a south pole attract each other. The north pole is sometimes called the positive pole and the south pole the negative pole of the magnet. Wrought or soft iron can be magnetized but only retains its magnet- ism while under the influence of the magnetizing force; steel or hard iron once magnetized retains its magnetization permanently, and special means to demagnetize it are required. All bodies can be electrified, but all bodies can not be magnetized. 7. If a sheet of paper is held over a powerful magnet and iron filings sprinkled on the sheet, the filings will assume positions approximately FIG. 6. FIG. 7. as shown in fig. 6. Some force connected with the magnet must make the filings assume these positions, which are different from what they would be if the magnet was not under the paper; and from the way the filings are arranged, this force must act in 'the space surrounding the magnet. This space is called the field of magnetic force, or simply the field of force, and the lines in which the filings tend to arrange them- selves are called the lines of force, and we shall see in chapter II that this conception^ used as a basis for electric measurements. The direction of the lines of force at any point indicates the direction of the magnetic force at that point, and their number in any area, the strength of the field in that area. It is found that a small magnetic needle, pivoted so that it is free to move and brought near the large magnet, will lie parallel to the- direction of the lines of force at any point at which it may be placed in the field, MANUAL OF WIRELESS TELEGRAPHY. 13 and that the north pole of the needle always points along the lines of force in the direction leading to the south pole of the magnet. The direction in which the north pole of the needle points is called the positive direction of the lines of magnetic force, and the direction in which the south pole points, the negative direction of the lines of magnetic force. Lines of magnetic force are said to run from the north pole of the magnet to the south pole through the air, and back to the north pole through the steel (fig. 7). ELECTRO-MAGNETISM. 8. If the wire in fig. 1 is coiled into a spiral, as in fig. 8, with the positive direction of the electric current as shown by the arrows and FIG. 8. the battery connections, a field of magnetic force which can be explored by a small . magnetic needle, or outlined by iron filings, as in fig. 6, will be found to exist around the spiral, and the direction of the lines of force will be found the same as those around the magnet in fig. 7. If the current is reversed, the lines of force are reversed in direction. Such a spiral, when traversed by a current, is found to have all the properties of a magnet, and is called an electro-magnet or solenoid. The strength of the magnetic field around an electro-magnet rises and falls with the rise and fall of the current, and its polarity depends on the direction of the current. The positive direction of the lines of magnetic force which surround a solenoid is from the north to the south pole outside of the spiral, and from the south to the north pole inside of it, just as the positive direction 14 MANUAL OF WIRELESS TELEGRAPHY. of an electric current is from the positive pole to the negative pole out- side of a battery and from the negative to the positive pole inside of it. If the number of turns of the spiral is reduced to one it does not lose its magnetic character. The lines of force then form circles around the wire, their positive direction being shown in fig. 9, the upper side being a north pole and the under side a south pole. If the turn is straightened out, as in fig, 10, the lines of force still form circles around the wire, and the north pole of the exploring needle points in the positive direction of those lines. This direction is found to be always at right angles to the wire. FIG. 9. FIG. 10. 9. It appears from the foregoing that what is called the positive direction of motion of electric currents, or charges, is related to what is called the positive direction of the lines of magnetic force, in the manner shown by the arrows in figs. 8, 9, and 10, and, further, that the terms positive and negative as applied to electric and magnetic effects, and so largely used in connection with them, are purely conventional. (See note 2, appendix.) 10. Returning now to the statement in article 8 that the strength of the magnetic field around a solenoid rises and falls -with the strength of the current, and its polarity (i. e., the direction of the lines of mag- netic force produced) depends on the direction of the current, it can be further stated that a magnetic field exists around every wire carrying an electric current (fig. 10). The direction of the lines of force in this field depends on the direction of the current. These lines of force always enclose circles in planes at right angles to the wire. 11. Since a current is conceived to be made up of a quick succession of moving electric charges (art. 3), the above facts may be stated in another way, viz., that moving electric charges produce magnetic fields in which the lines of magnetic force enclose circles in planes at right angles to the direction of motion of the moving charges. This has been proved to be true for single static charges.* ELECTRO-MAGNETIC INDUCTION. 12. Fig. 11 represents a primary battery, with the two poles of the battery connected by a conducting wire, broken at K. A straight portion * By Professor Rowland, Johns Hopkins University. MANUAL OF WIRELESS TELEGRAPHY. 15 FIG. 11. A B of this wire is parallel to, and at a distance from another conducting wire C D. When the break at K is closed, a current flows in the circuit, and a field of force is created around the wire. Let us consider the straight portion A B in which the direction of the current is shown by the arrows, and the direction of the lines of force by the circles (shown as ellipses), at right angles to A B. Several of these lines of force are shown embracing the parallel wire C D. If gold-leaf electroscopes (art. 1) are attached to the ends C and D of the wire C D, and if the current started in A B when the break is closed is sufficiently powerful, the gold leaves will be observed to diverge, momentarily, whenever the circuit is made or broken at K. The stronger the current in A B, and consequently the stronger the magnetic field produced, the more pronounced the indications of the electroscope will be. This shows that the ends C and D of the wire C D are electrified when the current is made or broken in A B. When the current is made, the end D is negatively charged like the amber and like the wire attached to the zinc element in fig. 1, the end C positively, like the silk and like the wire attached to the carbon element in fig. 1. When the circuit is broken at K the electrification of C D is reversed, C becoming negatively and D positively electrified. A sudden increase or decrease of the current in A B produces the same effect as when the current is made or broken. It is to be noted that the electrification of C D is only momentary. As soon as the causes producing it are removed, the electric charges unite and neutralize each other through the body of the conductor. We know that when the current in A B is made, a magnetic field is created around A B which extends to and beyond C D, and that when the current in A B is broken, the magnetic field disappears, and that the only thing common to A B and C D is this magnetic field, the lines of force in which surround them both, and since we see that one kind of electrification is produced in C D when the lines of force are being created, and the opposite kind when they are being dissipated, we con- clude that the movement or creation of these lines creates the electric charges that we observe in C D. 13. In art. 11 it is stated that moving electric charges create magnetic lines of force. Now, we see the truth of the converse, viz., that moving magnetic lines of force create electric charges. These two facts are of general application and are the basis of all electro-magnetic calculations. 16 MANUAL OF WIRELESS TELEGRAPHY. 14. It is of great importance to keep clearly in mind the fact that electrification in C D only takes place when the current is made or broken or changed in A B. When there is no current in A B there are no magnetic lines of force, and consequently there is no electrification in C D. When there is a constant current in A B the magnetic field is constant and there is no electrification in C D. It is while the current in A B is rising or falling, and- the lines of force expanding from or contracting toward A B and cutting through C D as they pass, that C D is affected. A movement of the lines of force is required to electrify C D, and this movement is produced by changes in the current in A B. If the ends C and D are joined to form a complete circuit, a momen- tary current will flow when changes in the magnetic field around C D take place. We have just seen that a moving magnetic field in the vicinity of C D creates electric charges in C D. We would also find that moving C D in a magnetic field has the same effect. The change of current in A B is said to induce the current in C D, and the action is called electro- magnetic induction. The preceding facts can be stated as follows : When magnetic lines of force cut or are cut by a conductor, electric charges (i. e., a tendency to current flow) are induced in the conductor, and currents flow if the conductor forms a closed circuit, the direction of the induced currents depending on the direction of cutting. 15. When the current in A B is rising, the magnetic lines of force are expanding, and cutting C D in the direction from left to right, the direction of the momentary current in C D being as shown in fig. lla. FIG. HA. Current in A-B Rising. B P FIG. HB. Current in A-B Falling. When the current in A B is falling, the magnetic lines of force are contracting, and cutting C D in the direction from right to left, the direction of momentary current in C D being shown in fig. lib. These momentary currents or movements of electric charges in C D themselves MANUAL OF WIRELESS TELEGRAPHY. 17 produce momentary magnetic fields around C D, the direction of the lines of force of which are shown by the arrows in figs, lla and lib. It will be seen that these lines of force are opposite in direction to those which created the current in C D. ,The field of force created around C D reacts upon A B, tending to create in A B a current in the opposite direction to that already in A B, i. e., to stop it. In other words, the change of primary current in A B induces a secondary current in C D. The latter current in turn induces a tertiary current, which is in A B. This influence of two currents on each other is called their mutual induction. 16. The electric charges produced by friction (art. 1), by chemical action (art, 3), and by the movement of lines of magnetic force are all identical in their properties, and the magnetic fields produced by the movement of these charges are also identical in their properties. It is therefore evident that a very close relation exists between electricity and magnetism. 17. We have seen that the field of magnetic force around a wire carrying a current or around a magnet can be mapped out by iron filings. In a similar manner the field of electric force around charged bodies can be shown by the use of various light powders. Figs. 12 and 12b show the electric field between unlike and like charges, respectively. Figs. 12a and 12c show the magnetic field between l i2teSll/'^ FIG. 12. FIG. 12A. FIG. 12B. FIG. 12c. FIG. 12. Electric Field Charges of Opposite Sign. Attraction. FIG. 12A. Magnetic Field between Unlike Poles. Attraction. FIG. 12B. Electric Field Charges of Same Sign. Repulsion. FIG. 12c. Magnetic Field between Like Poles. Repulsion. 18 MANUAL OF WIRELESS TELEGRAPHY. unlike and like poles, respectively. The electric field between two charged bodies is found to resemble very closely the magnetic field be- tween magnet poles. In all figures it can be seen that in electric as well as magnetic fields each line of force appears to repel its neighbor, and that they have their ends on points of opposite electrification or magnet- ization. If these lines tend to shorten in the direction of their length this tendency will cause the attraction between the bodies from which they proceed. 18. It may be asked, what are these lines of force which are not visible and which can not be physically grasped? The only reply is that we believe all electric and magnetic phenomena to be the results of the disintegration of the atoms of matter or the rearrangement of their constituent parts (see note 2, appendix), the movements of which produce stresses and consequent movement or strains in what is called the ether, an almost infinitely elastic, infinitely tenuous substance which surrounds and permeates all matter and all space. The earth is immersed in an illimitable ocean of ether, just as fishes are in water. We move about in a sea of it. What we call electric and magnetic fields are places where ether move- ments and ether stresses can be detected by the phenomena, which they produce, and which are being described. An electric field is a state of strain (stretch or compression) in the ether; it can be removed between any two points by connecting them with a conductor. The release of the strain starts movements of electric charges in the conductor. Movements of these charges produce another state of strain in the ether at right angles to the first. We call this a magnetic field. We have seen that movement of either field creates the other, and that the lines of force in the two fields, when they are thus produced, are in planes at right angles to each other. When equilibrium is restored one field or the other has disappeared, though they can coexist in a transitory state. 19. It lias been proved that light and heat are forms of ether motion also, and that all movements (electric and magnetic) in the ether are propagated with the velocity of light. It has also been proved that electric movements pi'ogress along straight wires at practically the same speed that magnetic movements progress at right angles to them i. e., with the speed of light. This velocity has been measured many times and found to be 186,000 miles, or approximately 300,000,000 meters per second. We must learn therefore to think of light movements and of electric MANUAL OF WIRELESS TELEGRAPHY. 19 and magnetic actions not as being instantaneous, but as being restricted to a definite measurable speed. It takes time for electric and magnetic effects to be propagated in the ether, time for them to be propagated along a wire. The wire guides or strikes out the line of maximum disturbance. 20. Let us now return to fig. 11. Before connection at K is made, the field of magnetic force does not exist, but the wires are electrified by means of action between the zinc and carbon in the battery solution. When the break at K is closed, a magnetic field is established; when the connection at K is broken, the magnetic field disappears. The question arises, how is this magnetic field created? How is it dissipated? The reply is: It is created by movement of electric charges in A B which disturb the ether and this disturbance is propagated through the ether at right angles to A B, with the speed of light, i. e., at the rate of 186- 000 miles or 300,000,000 meters per second. This disturbance is of such a nature as to produce a state of strain in the ether which may be compared to that produced in a piece of rubber by compression or tension. The strain is relaxed as soon as its cause (i. e., the movement of the electric charges) is removed. The amount of strain (i. e., the strength of the magnetic field) decreases as the distance from the moving charges increases. It spreads in all directions, but except with very delicate instruments can not be detected at any great distance from A B. The creation and dissipation of this state of unstable equilibrium in the ether, which must be brought about by some kind of movement in it, produces electrical movement in C D, or, as it is perhaps better to say, produces electric charges in CD. CD stands in the way of and is disturbed by an advancing or receding wave of movement in the ether, originated at A B. C D is, like all other conductors, an obstacle in the path which creates an eddy, so to speak, in the ether wave and reacts, however minutely, on A B, because the movement of the electric charges produced in C D also creates an ether movement, but in the opposite direction to that proceeding from A B. 21. We have now reached a point where the electric and magnetic actions under discussion are directly applicable to wireless telegraphy, but before proceeding with this subject it is desirable to consider more fully the action of A B on C D, because the creation of electric currents by moving or varying magnetic fields, and vice versa, is the basis of industrial electric power of that used in wireless telegraphy as well as in other branches of electricity; and other facts or developments of this fundamental fact will appear which will lead to a clearer compre- hension of it. 20 MANUAL OF WIRELESS TELEGRAPHY. FIG. 11. 22. In fig. 11, C D is shown parallel to A B. If C D is slowly revolved around its own center as an axis the effect on it of making, breaking, or changing the current in A B will be found to decrease until C D is at right angles to A B, when it will disappear altogether. The lines of force are circles at right angles to A B; they do not cut C D when it is at right angles to A B because it is parallel to them, and consequently no effect is produced. The induced effects in C D will be found to increase as it is brought nearer A B and to decrease as it is removed from A B. The field near A B is stronger, and more lines of force are created there or dissipated there than at a greater distance from A B i. e., a greater disturbance in the ether takes place. 23. If the two ends of C D (fig. 11), are brought close together, but without touching, and if the current made or broken in A B is very strong, a spark will pass between the ends of C D at each make and break. If C D is separated from A B by an opaque, nonmetallic screen and the makes or breaks in A B are made to represent the characters of a code, messages sent in this code from A B can be received at C D when each is invisible from the other. By the addition of a battery to C D, similar to that producing current in A B, replies can be sent, and thus a crude wireless telegraphy produced. 24. If A B is coiled into a spiral or helix and C D. into a similar spiral or helix (fig. 13), the effect of making, breaking, or changing the current in A B is much greater than where both wires are straight; for the disturbance created in the ether that is, the number of lines of force produced by the moving charges in A B is equal, for equal lengths of the wire, and since a greater length is concentrated in the same space, the number of lines of force in that space, assuming the current in the spiral to be the same as that in the straight wire, are correspondingly greater. This stronger field would produce an increased effect on a straight wire; but the length of C D is also concentrated. Therefore the effect is increased still more. 25. We know that A B when coiled as in fig. 13 and traversed by a current forms a solenoid (art. 8, fig. 8). The space inside the coil is called the core, and it has been assumed that the surrounding substance (excluding the ether, which is present both in the interior and on the exterior of all bodies) is air. It is found, however, that if the core of the solenoid is iron, as in fig. 14, instead of air, the effect on C D is very much more powerful i. e., the numbers of lines of force created with the same current is very greatly increased. MANUAL OF WIRELESS TELEGRAPHY. This shows that it is easier to create lines of force in iron than in air; or,. to state the fact differently, lines of force permeate iron more easily than they do air. The relative ease with which magnetic lines of force are created in a substance is expressed in figures and called its magnetic permeability. The permeability of air at atmospheric pressure is called IN A-B RISING^ unity, and on that basis the permeability of the purest wrought iron is 3,000. In other words, within limits the same current will produce 3,000 times as many lines of force in iron as in a body of air of the same length and area of cross section. CURRENT iNA-lB TALLIN* FIG. 14. 26. If the iron of fig. 14 is extended to include C D, as in fig. 14a, the effect of changes in A B is increased still more, because in fig. 14 the lines of force are partly in iron and partly in air, while in fig. 14a they have an iron path throughout, and. are consequently much greater in number. C D can also be placed inside of A B or outside of it, with or without an iron core (figs. 14b and 14c).- MANUAL OF WIRELESS TELEGRAPHY. FIG. 14A. Closed-Core Transformer (current in A B falling) FIG. 14B. Open-Core, Step-Down Transformer or Induction Coil (current in A B rising). C FIG. 14c. Air-Core, Step-up Transformer (current in A B rising) A C FIG. 14o. Auto Step-Down Transformer (current in A B rising). MANUAL OF WIRELESS TELEGRAPHY. 23 27. Since the tendency to current flow in C D is produced by lines of magnetic force cutting C D, and since on making or breaking cur- rent in A B each line of force cuts C D once, for each turn in C D, if the turns in C D are decreased or increased, as in figs. 14b and 14c, the tendency to current flow i. e., the electro-motive force is raised or lowered. From this fact, and from the fact that the current in C D is opposite in direction to that in A B, the arrangements in figs. 14a, 14b, and 14c are called transformers. Fig. 14a is called a closed-core trans- former; fig. 14b an open-core transformer or induction coil; fig. 14c an air-core transformer. Transformers are called step-up or step-down with reference to whether the number of turns in the coil C D is greater or less than those in A B. Fig. 14b is a step-down; fig. 14c a step-up transformer. The coil A B is called the primary and the coil C D the secondary wind- ing, and where A B and C D have some turns common to both, as in fig. 14d, the arrangement is called an auto-transformer. FIG. 13. 28. Referring again to fig. 13 : When the break at K is closed, a current is started, which progresses upward through the coil, the mov- ing charges composing it creating a magnetic field around the wire. The lines of force, as they expand from the current in the first turn of the spiral, cut the second turn of A B in the same way that they cut C D a little later. They induce a current in the second turn opposite in direction to that in the first turn i. e., tending to stop it. The same effect is produced in the third and succeeding turns. In other words, the different parts of the coil A B react on each other just as the coil C D reacts on A B. This reactive effect of the turns on each other makes the rise in current slower than in a straight wire, and is greater when the core of the coil is of iron than when it is of air, because of the greater number of lines of force produced. 24: MANUAL OF WIRELESS TELEGEAPHY. 29. We find that a stronger current is produced by the same battery in a short wire than in a long wire of the same size and material, and in a thick wire than in a thin wire of the same length, and we say that this is due to the greater resistance of the long wire and of the thin wire as compared with the short or with the thick wire. To establish the same current in the longer or the thinner wire as in the shorter or thicker wire requires a larger battery that is, greater E. M. F. 30. N"ow, we find that when the wire is coiled into a spiral and a change in the current is taking place, the turns react on each other and resist the change of the current. This resistance does not depend on the size nor the material of the wire, but only on the amount and quick- ness of the change in the current, and is therefore of a different character from the resistance referred to above. The resistance of a wire to changes in current established in it is called its reactance, and during the change the total effect of the true resistance and the reactance is called the impedance of the wire or circuit. In circuits having reactance the production or progression of electrical effects is retarded. It takes longer to create a given current than in the same .length of straight wire. It may be said, therefore, that coiling a wire increases its electrical length i. e., increases the time it takes an electrical movement created at one end of it to reach the other. The currents in C D are said to be produced by the induction of A B on C D. The retarding effect of the coils in A B to the rise and fall of current in A B is said to be due to the self-induction of A B. It has been shown that the amount of both kinds of induction depends on the shape and arrangement of both circuits and the material (iron or air) in and around them. METHODS OF PRODUCING CURRENTS BY ELECTRO- MAGNETIC INDUCTION. 31. The currents under discussion have been illustrated as being pro- duced by batteries of primary cells, and for many purposes these are very valuable, but for the production of very powerful electrical effects advantage is taken of the fact, stated in art. 14, that when magnetic lines of force cut or are cut by a conductor, electric currents flow in the conductor, if the latter forms a closed circuit. The direction of current flow can be determined by the following rule : * (a) An increase in the number of lines of force embraced by a circuit induces a current in the opposite direction to that in which the hands of a watch move, while a decrease in the number of lines of force induces a current in the same direction as that in which the hands of a watch move, the line of sight being in both cases along the positive direction of the lines of force. (Art. 7 and fig. 7.) Or rule (b) The positive direction * From Fiske's " Electricity and Electrical Engineering." MANUAL OF WIRELESS TELEGRAPHY. 25 of the lines of force is with the hands of a watch when the current is flowing away from the observer. 'And rule (c) The currents induced by moving lines of force always tend to prevent change in the inducing current. Induced currents are, therefore, in the opposite direction when the inducing current is rising and in the same direction as the inducing current when the latter is falling. (Art. 15.) Rule (a) is illustrated by fig. 15; rule (b) by fig. 15a. From rules (b) and (c) can be deduced the following illustrated in fig. 15b, which represents a conducting wire C D below a line N" S represent- ing a field of force and its direction. When in the relative positions shown, movement of either wire or line of force toward the other creates a current to the rear, moving either one away from the other creates a current to the front. FIG. 15 It will be seen that the field N" S can be revolved through any angle around the wire C D as an axis so as to be to the right, left, or above or in any intermediate position without changing the truth of the above statement. These three rules show the relation between what we call the positive direction of the lines of magnetic force and what we call the positive direction of electric current. 32. No w let the wire C D in fig. 11 be bent until it forms a rectangle, and let it be placed in the magnetic field between the north and south poles of a powerful electro-magnet having an iron core. By bending the core into the shape shown in fig. 16, the north and south poles are oppo- site each other and a greater number of lines of force are produced, because the distance they have to travel through the air is very much shortened as compared with fig. 14. 26 MANUAL OF WIRELESS TELEGEAPHY. Exploration of the field in fig. 16 by means of iron filings or by means of a small magnetic needle will show that the lines of force extend directly from a point in the north pole to the opposite point in the south pole. In other words, that they are straight and parallel to each other, and they are so shown in fig. 16. The field is also found to be of uniform intensity, which indicates that the number of lines of force are equally distributed throughout its area. Now, if C D is moved up or down in the magnetic field, no indica- tion of a current can be perceived, and it appears that the statement in art. 14 (that when magnetic lines of force cut or are cut by a conductor electric currents flow in the conductor if the latter forms a closed circuit) is in error, but when we consider that when C D is moved upward (the FIG. 16. field being of uniform intensity) as many lines of force are cut by the bottom half as by the top half of C D, the currents induced in the two halves must therefore be equal, and since both flow to the rear we see that they neutralize each other, and the result is zero. Another way to explain this is to consider that portion of the field inclosed by C D as containing a certain number of lines of force. Those coming in when C D is moved induce a current in one direction, those going out induce a current in the opposite direction, and if as many come in as go out no effect is produced. 33. If C D were straight, electric charges would be produced on its ends and would be maintained there as long as the cutting of the lines of force continued, but bending it into a closed circuit changes conditions to the extent that cutting of lines is going on all around the circuit, some inducing charges in one direction, some in the other, and it is MANUAL OF WIRELESS TELEGRAPHY. only when there is a preponderance of cutting in one direction that a current actually flows. This would occur if C D were moved from a point where the field is weak to where it is stronger, or vice versa, but the field under discussion is supposed to be uniform. (See rule a.) If C D is rotated around one of its diameters as an axis (say the hori- zontal diameter at right angles to the lines of force) when it is hori- zontal, as in fig. 17, the lines of force included will be zero, and when vertical, as in fig. 17a, the lines of force included will be the maximum number possible in that field, so that a revolution of 90 will make an entire change in the number of lines of force passing through the rectangle. For instance, if the revolution is in the direction of the hands of a clock i. e., if the top of C D moves to the right (see fig. 17a) the upper half of C D is cutting lines of force in the direction which induces movements of electric charges to the front, while the lower half is cutting lines of force in the direction which induces movements of electric charges to the rear, so that an electric current is established in C D in the direction shown, or the number of lines of force included in C D is decreasing, and looking from N to S, the current moves with the hands of a watch. FIG. 17. 'FIG. 17A. If C D's rate of revolution is constant, a little consideration will show that when it has revolved through 90 and its plane is horizontal it is then moving at right angles to the lines of force, and consequently cutting them faster than when, its plane being vertical, it moves parallel to the lines of force for an instant and is not cutting any; also that the increase in the rate of cutting is progressive from one position to the other. It will therefore be seen that the electric current produced is a maximum when C D is horizontal, and that it is zero for an instant when C D is vertical because during that instant it moves parallel to the lines of force and therefore it cuts none. (No change in number included.) It is also evident that the increase of the current from zero to a maximum is progressive during the first 90 of revolution, that it then progressively decreases until C D has revolved through 180, and is again moving parallel to the lines of force when it falls to zero. 28 MANUAL OF WIRELESS TELEGRAPHY. As the revolution continues, that half of C D which during the first half revolution was cutting lines of force in such a manner as to induce a current to the front now cuts them in such a manner as to induce a current to the rear, its former place being taken by what was originally the lower half, so that the direction of current in C D is reversed. (Eule c.) Another maximum rate of cutting lines of force and consequent maxi- mum of current is produced when C D has revolved through 270. The current progressively increases from 180 to 270 and then decreases until when the original conditions are restored by the completion of one revolution the current has again fallen to zero. From the above and from an inspection of fig. 17a it will be seen that current is always flowing to the front in that half of C D which is going down to the right and to the rear in the half going up on the left, and that each half revolution the current changes in direction. Such a cur- rent is called an alternating current. FIG. 18. 34. This can be shown graphically in fig. 18, where the rate of cut- ting and therefore the rate of change of number of lines included in the circuit at different equidistant points in one revolution is represented by equidistant vertical lines proportional to the cutting rate, and conse- quently to- the current strength. Vertical lines above the horizontal line represent current strength in one direction and below it current strength in the opposite direction. A regular curve is produced by joining the tops of these lines. This curve is the curve of sines, because the rate of cutting and the strength of the induced current are proportional to the sine of the angle of revolution.* * Since the lines of force are horizontal, the number cut during the revolu- tion of C D through any angle is proportional to the vertical movement of the extremity of the radius of C D which generates the angles. The amount of this vertical movement is the sine of the angle, and therefore the induced current is proportional to the sine of the angle. MANUAL OF WIRELESS TELEGRAPHY. 29 35. If C D instead of forming a closed circuit entirely in the mag- netic field has its ends connected to two rings which revolve with it and touching these rings are the ends of a coiled wire (E F, fig. 19), the FIG, 19 currents induced in C D also flow through E F and make of it a sole- noid whose strength varies with the strength of the current and whose polarity reverses with the reversal of the current. If a small magnetic needle were pivoted in E F, its direction would tend to change with each reversal of the current, and it can thus be made to indicate both the direction and the amount of current flowing through the coil E F. Such an instrument is called a galvanometer. The currents in the coil E F are supplied from C D, and they are induced in C D by its movements in a magnetic field. C D has become a source of electricity like the battery in A B. E F corresponds to the coil A B in fig. 13, and the rise and fall of current in E F will produce a rise and fall of current in another coil near it, just as the make and break at K in fig. 13 induces momentary currents in C D. The currents in C D, fig. 13, were induced by interrupted current. Those induced by E F in coils near it are induced by alternate current. Interrupted current was used almost entirely in wireless telegraphy in its earlier development. It has now been replaced by alternate current. 36. It only remains now to make C D produce the magnetic field in which it revolves, and we can dispense entirely with the primary battery in A B. This can be done as follows : In fig. 20 instead of having each end of C D connected to a ring of conducting material, as in fig. 19, one ring is removed and the other split into two equal parts and an end of C D connected to each part, the ends of E F being adjusted so that as the split ring revolves with C D one end of E F is always connected FIG. 20. 30 MANUAL OF WIRELESS TELEGRAPHY. through the split ring with that half of C D in which the current is flow- ing to the front and the other end to that half in which the current is flowing to the rear. This arrangement makes the current in E F always flow in the same direction. It rises and falls with the current in C D, but does not reverse, because just as the current reverses in C D, E F changes ends, so to. speak, by breaking connection with one half of the split ring and making connection with the other. The current in E F is now said to be a pulsating instead of an alternating current, and the change can be graphically represented by transferring the part of the curve below the line in fig. 18 to a corresponding position above it, as in fig. 18a. \ FIG. ISA. The alternating current in CD is said to be rectified into a direct current in E F. The split ring by means of which it is rectified is called a commutator, and the entire apparatus (either with or without a com- mutator), a dynamo. 37. With a single coil, C D, rotating in the magnetic field the current in E F can be made to flow always in the same direction, but in order to make it constant a large number of coils, equally spaced, must be used, so that one of them is passing through the position (horizontal) in which maximum current is produced practically all the time. If there were 10 such coils, each connected to its own split ring (fig. 21), and all connected to E F, the currents in each would overlap, so that the resultant current in E F to another scale might be indicated by a line joining the highest point of each (fig. 18b). In other words the current in E F is practically constant. FIG. 18s. The revolving coils are held in place on a cylindrical drum or ring and the whole is called an armature. If this ring is made of iron the strength of the magnetic field is much increased, because the iron affords a path for the lines of force from one pole to the other and thereby lessens the distance through which they have to pass in the air. (See art. 25.) MANUAL OF WIRELESS TELEGRAPHY. 31 The tendency to current flow in C D created by cutting lines of force is called the electromotive force in C D (see art. 3), and is found to depend on the number of lines cut in a given time, so that the faster C D revolves, and the stronger the magnetic field, the greater the electro- motive force and the greater the current produced in any given circuit. Now, if the current induced in C D, instead of all flowing through E F, is divided, so that part of it flows around the core of the electro-magnet (fig. 21), this current can take the place of that produced by the battery in A B and the battery can be dispensed with. FIG. 21. 38. In art, 6 it is stated that wrought or soft iron can be magnetized, but only retains its magnetism while under the influence of the magnet- izing force. Steel or hard iron once magnetized retains its magnetiza- tion permanently and special means to demagnetize it are required. It is found that electro-magnets with soft-iron cores can be made more powerful (i. e., will give a stronger field) than if the cores are of steel, and that electro-magnets with either kind of core can be made to give much stronger fields than any permanent magnet. Also, that soft-iron cores retain a very small part of their magnetism and polarity when the current i& broken, so that, if the magnet poles between which C D revolves are made of the most efficient material (wrought iron or mild steel containing no phosphorus), when C D stops they still retain their polarity in a slight degree. . . When C D starts to revolve again the weak field generates a small cur- rent in C D, which sends this current through the wire around the poles ; this current increases the strength of the poles and consequently of the field which increases the current in C D and so on. This is called generating or building up, and continues until the limit of the power moving C D in the continually strengthening field is reached, or until the iron core is saturated, in which condition no increase of current will increase the field. 32 MANUAL OF WIRELESS TELEGRAPHY. 39. When alternating current is desired, a dynamo, in order to be self-exciting, i. e., to produce its own field, must have part of its cur- rent rectified by means of a commutator. It is more usual, however, to drive a small direct-current dynamo by means of the same power which drives the larger one, the current from the small dynamo being used to create the magnetic field in the larger one. Such a machine is called an exciter. 40. The fact that magnet poles of unlike polarity attract each other (art. 6) applies to electro-magnets, with or without iron cores, as well as to permanent magnets. Hence two electro-magnets placed as in fig. 13 will attract or repel each other according to their polarity. Each line of force apparently tends to contract in the direction of its length, and by so doing exerts a mechanical pull on the conductors which it sur- rounds. The same effect is observed between a magnet and a wire carrying a current (which, as we know, has a magnetic field around it) and between two wires, each carrying a current. They actually pull or push each other according to the quality of their magnetism, which is determined by the direction of the current. 41. If in fig. 21 the armature instead of being revolved to the right by some outside agency, is supplied with a current flowing through it in the same direction as the current it generates, it will revolve to the left. The current flowing to the front in the winding of the right half of the armature and to the rear in the winding of the left half makes of it an electro-magnet with a north pole at the bottom and a south pole at the top. The revolution is caused by the attraction of the north pole of the armature by the south pole of the field magnet, and its repulsion. by the north pole of the field magnet, This action is reversed in the south pole of the armature. The movement will be continuous, because, as the top of the arma- ture moves toward the north pole of the field magnet, the commutator acts to maintain the flow of current as before, and the consequent arma- ture poles are always at the top and bottom halfway between the field magnets. The armature thus creates a current when made to revolve, and revolves when supplied with current. In the first instance we have seen that the entire machine is called a dynamo; in the second it is called a motor. Every dynamo will run as a motor if supplied with current. Every motor will act as a generator or dynamo if made to revolve in its own field. The motor can be made to drive another armature in another field. Such a machine is called a motor-generator. It can be run with direct or alternating currents and made to generate direct or alternating cur- MANUAL OF WIRELESS TELEGRAPHY. 33 rents of a higher or lower E. M. F. For this reason it is sometimes called a rotary transformer, as distinguished from the stationary trans- formers already described. ELECTRIC AND MAGNETIC FIELDS. / 42. Electricit}^ produced by friction (art. 1) is sometimes called fric- tional electricity; by primary batteries, voltaic electricity; by electro- magnetic induction, dynamic electricity. But however produced and transformed,, all kinds of electricity are identical, and the same is true of all kinds of magnetism. Wherever there is an electric charge, stationary or moving, emanating from the charge are electric lines of force which end at other electric charges. Wherever there are moving electric charges (currents) there are magnetic lines of force also, and these magnetic lines of force are always at right angles to the direction of the motion of the charges and to the electric lines of force proceed- ing from them. And, finally, motion, or state of strain in the ether, which these lines of force represent, travels with the speed of light, and the fields of force, while more pronounced and therefore more easily detected near the moving charges, are really all pervasive. They have no limits. 43. Imagine a disturbance say an expansion of a gas to take place in the center of an immense rubber ball. A wave of tension, which be- comes less as its distance from the center increases, progresses outward to the farthest confines of the ball. When the gas contracts, a wave of contraction, also starting from the center, and decreasing with its distance from the center, progresses outward to the farthest confines of the ball. If expansion and contraction are equal the ball's former state of equilibrium is restored. In this way it can be imagined that starting a current produces a state of strain in the ether or stretches it in one direction; stopping it releases the strain. Action in both cases starts at the point where the current is produced and progresses outward with the speed of light, and a little consideration will show that it can have no limit, though it soon ceases to be perceptible except under certain conditions, to be later described. The function of wireless telegraphy is to produce these ether move- ments at will. ELECTRIC CAPACITY. 44. We can produce momentary currents in conductors, whether open or closed, by the cutting of lines of force, and the evidences of electrifi- cation are most pronounced at the ends of an open conductor, but these disappear as soon as the cutting of lines of force ceases. We find, how- ever, that electrification of amber, glass, silk, and other bodies remains 34 MANUAL OF WIRELESS TELEGRAPHY. after the rubbing ceases, and if glass plates or other nonconductors be connected to the ends of a conductor in which an E. M. F. is being generated, so that connection is made all over the surface of the glass (as it is when rubbed), the glass when separated from the conductor will be found to be electrically charged the same as when electrified by rubbing. When two plates oppositely charged (art. 1) are connected through wires leading to a galvanometer, the amount of deflection of the galvanometer needle (caused by the magnetic field of the momentary current created as the charges unite and neutralize each other) is a measure of the quantity of electricity on each plate. In testing plates of different sizes, shapes, and materials, charged to the same potential by being connected to the poles of the same source of electricity, it is found that 'different values of the throw of the gal- vanometer needle are produced. Other conditions being equal, plates having the greatest amount of surface are found to have the largest capacity. Plates of the same capacity will give a larger throw of the galvanometer when charged from a source of high than a source of low potential, so that the amount of electricity stored in an electrified body depends on its potential as well as on its capacity. 45. If two plates, oppositely charged by being connected to the poles of a battery, as in fig. 22, or to the terminals of a dynamo or transformer are discharged by being connected through a galvanometer, the throw of the galvanometer will not be as great as if the same plates, charged to the same potential by the same battery as in fig. 22a, are discharged through the same galvanometer. By being brought closer together the plates seem to have their capacity increased. It takes a greater amount of electricity to bring them to the same potential than when farther apart. If two plates charged at a distance from each other, as in fig. 22, FIG. 22. and then disconnected from the battery are brought to the position shown in fig. 22a, their potential, as measured by an electroscope, is found to be lowered. The electricity is said to be condensed by the approach of the plates, and such an arrangement is termed a condenser, a somewhat misleading term, but one generally used. . This is analogous to the increased strength of magnetic field produced by shortening the magnetic circuit while retaining the same magnetizing MANUAL OF WIRELESS TELEGRAPHY. 35 force. In both cases the field of force represents stored energy which can be made to reappear in the discharge of the condenser or the dissipation of the field. The two plates can be reduced to one if of nonconducting material, but since a nonconductor can not transmit electric charges, in order to utilize the two surfaces of the plate, each must be covered with a con- ductor which will permit the charges to distribute themselves over its area. ELECTRIC INDUCTION. 46. Electric lines of force permeate a nonconductor i. e., electric induction takes place through it, in a way analogous to that in which magnetic induction takes place through iron or air. The permeability of air for magnetic induction is taken as a standard and called unity. (See art. 25.) Its permeability for electric induction is also taken as a standard and called unity, and as we find that iron, nickel, cobalt, and oxygen have a greater magnetic permeability than air, so we find that glass, beeswax, paraffin, nearly all kinds of oil, and indeed most bodies we call insulators, have a greater electric permeability than air. The quality of a body as compared with air in this respect is called its specific inductive capacity, and bodies when considered with reference to electric induction through them are called dielectrics. (Art. 2.) It is found that the best quality of glass has nine times the specific inductive capacity of air. This means that when subjected to the same potential, the electric field, when this glass is the dielectric, is nine times as strong as that created when the medium intervening between the charges is air, it requires nine times as much work to create it, and its discharge can do nine times as much work. 47. Bodies such as iron or nickel through which magnetic induction is taking place are found to change slightly in shape, and sudden changes in the induction or lines of force permeating them produce slight sounds. The action is also accompanied by the production of heat, but as the magnetizing force (magneto-motive force) increases, the lines of force tend to reach a maximum which no increase of mag- netizing force will increase. When in this condition the magnetized body is said to be saturated. There is, however, apparently no limit to the magnetization of air. In the same way bodies (dielectrics) through which electric induction is taking place are found to change (enlarge) slightly in shape, but increase of electro-motive force (in this case potential) does not appear to tend to a maximum of electric induction. The physical strain on * the dielectric, however, continues to increase and finally reaches a point where it pierces or ruptures the dielectric, the action being accompanied 36 MANUAL OP WIRELESS TELEGRAPHY. by a sharp crackling sound and by the production of light and heat, which we call an electric spark. If the dielectric is air or a liquid, the rupture is immediately repaired by the action of the surrounding sub- stance on that heated by the passage of the spark ; but if the dielectric is a solid the rupture is permanent. Magnetization is limited by satura- tion. The limit of electrification is marked by rupture. The electric charges are found to have been dissipated after the spark has passed. The condenser is said to be discharged. If the oppositely charged plates are discharged without sparking, a slight sound is produced if the dielectric is glass. This is analogous to the minute sounds given out by magnets when magnetized or demagnetized suddenly. Magnetization or electrification seems to consist of forcing to point in the same direction the magnetic or electric polarities of the molecules of a substance. We have seen that the capacity of an electrified body depends on the area of its electrified surface, on the nearness of its charge to charges of opposite sign, and on the material of the dielectric i. e., the sub- stance intervening between the charges. ELECTRIC CONDENSERS. 48. Bodies capable of being electrified and arranged so as to present a large capacity in a small space are frequently called simply capacities, but this term is misleading, and though the term condenser is not entirely satisfactory it will be used. The total charge in a condenser depends on its potential as well as its capacity. Its potential depends on the potential of the source of electricity only, but its capacity, as stated above, depends on its size, material, and arrangement. Condenser capacities may be said to be related to each other in the same way as rubber bags inflated by gas. A large bag charged to a given pressure contains more gas than a small bag charged to the same pressure. The gas in the large bag is making no greater effort to escape per square inch (i. e., has no higher potential) than the gas in the small bag ; but it requires a longer time and more gas to charge the large bag than the small one. So when connected to the same source of electricity it requires a longer time to charge a condenser of large capacity to a given potential than it does to charge a small one to the same potential, and its power to do work is correspondingly greater. In the same way it requires a longer time to create the magnetic field of a large electro-magnet than that of a small one, and a stronger mag- netic field (within limits) is created by a large current than by a small one under the same conditions, and the energy stored in the strong field and its power to do work is correspondingly greater. MANUAL OF WIRELESS TELEGRAPHY. 37 49. It is evident that a close analogy can be drawn between the electric field in a condenser and the magnetic field around an electro-magnet. We have seen that any movement of either field creates the other; that they can exist independently only in a static condition; that, though they have no limits, the center of effort, the point of greatest activity in each, is at the body which we consider electrified or magnetized; that bodies differ in their qualities in these respects; that an actual physical change takes place in the dielectric when electrified and in the iron or nickel when magnetized, and, finally, that both electric and magnetic fields represent stored energy in an infinitely elastic medium, and we shall see that this medium, on account of its elasticity, vibrates and oscillates when either an electric or a, magnetic field is suddenly created or destroyed in it. 50. The most common and best known form of condenser is the Leyden jar, which consists of an inner and outer coating or film of tin foil or copper on a glass jar, the glass being the dielectric. Electric induction takes place through the glass and the energy is stored in the electric field, the tin foil merely serving to increase the area over which electric induction takes place, and hence the capacity of the condenser. Condensers are often made up of a large number of interlaced plates or films of conducting material, having between them for a dielectric FIXED CONDENSER VARIABLE CONDENSER HI- PIG. 23. FIG. 2 3 A. FIG. 23s. larger pieces of glass, mica, or oiled paper, alternate plates being simi- larly charged. Condensers are represented either as in fig. 23 or fig. 23a. They will be represented in this book as in fig. 23. Condensers are also made in which the relative position of the plates, and therefore the capacity, can be varied at will. These are called variable condensers, and will be represented as in fig. 23b. In variable condensers the dielectric may be glass, air, oil, mica, or paper. DISCHARGE OF CONDENSERS. 51. If, after being charged by connecting the inner coating to one pole of a source of electricity and the outer coating to the other, the two coatings are connected by means of a conducting wire the charges neutralize each other and the condenser is said to be discharged. The discharge of a condenser being a movement of electricity creates a cur- rent and consequently a magnetic field around the wire through which the discharge takes place. 38 MANUAL OP WIRELESS TELEGRAPHY. If the potential is high enough the condenser can be discharged with- out actually connecting the two coatings, for when the opposite ends of wires connected to them are brought within a certain distance of each other sparks will pass, and the condenser will be found to be discharged, the same as if the wires were actually connected. The charges unite by rupturing the air dielectric. The energy stored in the electric field appears as sound, light, heat, and other invisible ether vibrations. This spark discharge is found when analyzed to consist usually of several sparks, passing first in one direction, then in the other. Each condenser coating is charged positively and negatively in rapid succes- sion, each charge being somewhat less than the preceding until the entire energy of the original charge is dissipated. This form of con- denser discharge is oscillating. The released charge acts like a released musical string which vibrates until its energy is dissipated, and as the same string gives out the same note, whether stretched strongly or only a little, so a condenser when discharged through the same wire always vibrates or oscillates in the same period, regardless of its potential. Just as the note given out by the string depends on its material and length, so the rate of vibration of a condenser depends on its capacity, which, as we have seen, depends on its material and arrangement. 52. Another illustration of oscillatory condenser action can be given: Let fig. 24 represent two glass vessels connected by a U tube with a stopcock at the bottom of the tube. One vessel is filled with water and the other 'empty. If the U tube is large enough to permit free passage of the water, when the stopcock is opened quickly the pressure in the filled vessel will cause a sudden rush of water up the other side of the tube into the empty vessel, which will continue until it has reached nearly the same height as before (fig. 24a). It will then rush ba'dv into the first vessel, and so on, reaching a little lower level each time until equilibrium is reached at the same level in both vessels (fig. 24b) . The only action which prevents the oscillation from being continuous is friction of the water on the walls of the tube and internal friction between its molecules. Released condenser charges would also continue to. oscillate indefi- nitely if it were not for the resistance in the discharging wires and in the dielectric and the sound and light produced by the spark. These absorb the energy of the charge, and, being relatively large, a position of equilibrium is reached after a few oscillations. If the U tube in fig. 24 is very small or the stopcock only slightly opened the water will gradually rise on the other side and will finally reach a position of equilibrium without any oscillation, and it is found that if the condenser discharge takes place through a long thin wire instead of a thick one the condenser is slowly discharged through it without anv oscillation. MANUAL OF WIRELESS TELEGRAPHY. 39 53. The oscillation of the water in fig. 24 is due to its inertia. Inertia is a property of all bodies and is in amount proportional to their weight. It is represented by their resistance to change of condition, either of motion or of rest. The water in the first vessel falls by the action of gravity. Once in motion its inertia (resistance to change of condition) causes it to rise on the opposite side against the action of gravity. When gravity has overcome its inertia it falls again by gravity and is carried on by inertia. FIG. 24. FIG. 24A. FIG. 24e. It continues to overshoot the mark, so to speak, until friction, internal and external, brings it to rest. Though the electric charges on condenser coatings appear to be inde- pendent of gravity, they do possess inertia, as is shown by their resist- ance to change of direction and by their oscillatory movements. 54. Let us consider a charged condenser (fig. 25) discharged through a thick wire connecting the coatings. A break in the wire prevents the discharge until the potential is high enough to cause sparks to cross the break. One condenser coating before discharge is at a certain positive potential, the other at an equal negative potential. Both discharge through the wire in the same time, and when they have reached zero potential the electric field has been dissipated, but the moving charges 40 MANUAL OF WIRELESS TELEGRAPHY. in the wires have induced a magnetic field around the wire. The strength of this magnetic field depends on the amount of the moving charges, i. e., the strength of the current., and on the self-induction (art. 30) of the wire which, as we know, depends on its shape and the material (air or iron) in which the magnetic field is created. All the energy (except that lost by friction) which was stored in the electric field is now in the magnetic field (fig. 25a). The magnetic field, having no continuous source of magneto-motive force (current) to maintain it, collapses on the wire, producing movements of the electric charges into the condenser coatings, which now become charged in the opposite sense (fig. 25b). The electric field is again set up, containing all the remain- ing energy, and the magnetic field disappears until the charges again move toward each other. QSOLLATIWG CONDENSER DISCHARGE AT START ENERGY ALL ELECTRIC. END OF QUARTER CYCLE ENERGY , ALL MAGNETIC. FIG. 25. FIG. 25A. ENERGY ALL ELECTRIC ENERGY A' L MAGNETIC ENERGY ALL ELECTRIC REVERSED- REVERSED. LESS IN AMOUNT. FIG.- 25B. FIG. 25c. FIG. 25D. The attraction of the unlike charges for each other is analogous to the attraction of gravity for the water in fig. 24, and the magnetic field caused by the self-induction of the moving charges is analogous to the inertia of the water, which makes it rise in the second vessel, because the collapse of this magnetic field charges the condenser in the opposite sense, and for this reason self-induction is sometimes called electro- magnetic inertia. , From the foregoing illustration of what appears to take place during the oscillating discharge of a condenser we see that the energy before an oscillation begins is all electric. At the end of the first quarter of a cycle it is all magnetic. At the end of a half cycle it is all electric, but in the opposite sense. At the end of three-quarters of a cycle it is all magnetic, but with the direction of the lines of force reversed. At the end of a complete cycle or oscillation the energy is all electric again MANUAL OF WIRELESS TELEGRAPHY. 41 (figs. 25a, 25b, 25c, 25d) and in the original sense, but less in amount on account of the losses which have taken place during the transforma- tions and which are shown by the heating of the condenser and the wires (and the sound and light produced by the spark if the oscillations take place through a spark gap). At all intermediate points of a cycle the energy is partly electric and partly magnetic. 55. A complete oscillation or cycle is made up of two alternations. The highest potential reached during an oscillation is called the ampli- tude of the oscillation. The difference between the amplitude of two successive oscillations is called the damping and is a measure of the losses. The interval in time between two successive oscillations is called the period. 56. Since every body has electric capacity in proportion to its surface (art. 44), and since movements of electric charges, without which a body can not be electrified, always produce magnetic fields, every body must have self-induction, and therefore electro-magnetic oscillations can take place in it. We know that every body vibrates in its own period mechanically, and we find that every body vibrates in its own period electrically, and further that the number of vibrations or oscillations per second depends entirely on the capacity and self-induction of the body. It will be seen that while a closed circuit is necessary for the flow of a continuous or direct current, for oscillating currents a straight wire is sufficient. A circuit containing a condenser which would completely obstruct a direct current has no effect on an alternating current other than to change its sign. 57. We must be careful to distinguish between the capacity of a con- denser and the total charge in it, and between the self-induction of a wire and the total induction caused by the current in it. The capacity, it may be repeated again, depends on the material and arrangement of the charged body. The total charge that is, the total electric induction depends on the capacity and the potential. In like manner the self- induction depends on the arrangement of the conductor and the sur- rounding material (whether iron or air). The total magnetic induction depends on the self-induction and the current. 58. AVe can see in a general way that the period of an oscillating circuit depends on the capacity and self-induction of the circuit, and not on the total electric or total magnetic induction, because the capacity and self- induction are determined by the material and arrangement of the circuit, which qualities determine the mechanical period of a body. It takes longer to discharge a condenser of large capacity than one of small capacity, and it takes longer to create a given current in a circuit of large than in one of small self-induction. Increasing the potential gives^ 42 MANUAL OF WIRELESS TELEGRAPHY. more work to be done during a discharge, but also gives power to do it in the same ratio, so that increase of potential does not change the period, though it may change the amplitude of the oscillations. 59. It was stated (art. 29) that coiling a wire increases its self- induction and enables a strong magnetic field to be created around it, and that this increases the electrical length of the wire i. e., it takes an electrical disturbance started at one end of it longer to reach the other end when the wire is coiled than when the same wire is straight. Now we see that the electrical length of a wire depends on its capacity and self-induction and that its period in seconds i. e., the time of one complete oscillation (the time required for an electrical impulse started at one end to reach the other and be reflected back) must be twice its electrical length divided by the distance electricity travels in a second, which we know to be the same as light (300,000,000 meters). The capacity and inductance of a straight wire long in proportion to its thickness are so related that its electrical length is equal to its natural length. Fr.om the above the period or time of one complete electrical oscilla- tion of a straight wire one meter long is ^J-^^-Q-Q-^-^ second, and it therefore oscillates 150,000,000 times per second. ' The number of oscillations or cycles made by an alternating current per second is called its frequency. 60. We know that by coiling a wire its self-induction can be greatly increased, and its period thereby lengthened. By adding capacity to the wire in the shape of condensers its period can be lengthened still more, so that by suitable arrangements a circuit having small mechanical length, but comparatively great electrical length, can be made up in a small space.* cx o o o FIG. 26. FIG. 2 6 A. Such a circuit is shown in fig. 26. It is made up of a condenser con- nected to a coiled wire, and will be called in- this book an oscillating circuit. * It must not be forgotten that every wire possesses capacity by virtue of its surface, and self-induction by virtue of the fact that an electric current can flow in it. Even condensers have a certain amount of self-induction. MANUAL OF WIRELESS TELEGRAPHY. 43 The oscillating circuit in fig. 26 may have a break or gap in it, as in fig. 26a, Tf the potential of the condenser is sufficient to rupture the air or other dielectric in the gap, the circuit does not lose its oscillating character. The presence of the gap does, however, decrease the number of oscillations for one charge and prevents the complete discharge of the condenser, because the oscillations cease as soon as the potential falls below a certain value. The greater the loss or damping in each oscilla- tion the smaller the number of oscillations that will take place before the potential falls so low that the spark ceases. 61. As stated in art. 48, the term condenser is not satisfactory, and the word capacity is often used to mean condenser, especially in con- nection with such an oscillating circuit, the condenser being spoken of as a capacity and the coiled wire as an inductance, which means a con- ducting wire arranged so as to have large self-induction. -A FIG. 27. FIG. 27A. FIG. 27. Inductive Resistance. FIG. 27A. Noninductive Resistance. Fig. 27 represents an inductive resistance, or simply an inductance, since it is assumed that all wires have resistance. Fig. 27a represents a ncninductive resistance, or simply a resistance it represents a coil so wound that the currents in adjacent turns are in opposite directions and the coil has therefore no self-induction. 62. An oscillating circuit whose electrical length can be varied at will is represented in fig. 28. It consists of a variable condenser in connection with a fixed inductance (fig. 28), or it may consist of a fixed condenser and a variable inductance (fig. 28a), or both capacity and inductance FIG. 28. FIG. 2 8 A. may be variable, the arrow in fig. 28a being meant to show that any number of turns of the coil can be included at will. 63. Two circuits having the same electrical length are said to oscillate in resonance; their periods are equal, though the inductance and capa- city may not be the same in each. 4 44 MANUAL OF WIRELESS TELEGRAPHY. For instance, suppose the oscillating circuit (28a) is adjacent to a wire, as in fig. 28b, having the same electrical length, we know that for oscillating currents (see art. 56) a closed circuit is not necessary. We also know that by reason of their mutual induction (art. 15) the closed oscillating circuit, which we can call A B, will induce currents in the wire, which we can call C D. Since their periods are equal the induced oscillating current in C D will be suitably timed to the natural period of C D and the two circuits will oscillate in resonance. C D can be called the open circuit as distinguished from A B, the closed circuit. GROUND FIG. 28B. Oscillating circuits now used in wireless telegraphy have electrical lengths varying from 100 to 5000 meters, giving from 1,500,000 to 30,000 oscillations per second. Those first used by Marconi had electrical lengths of about 6 centimeters and oscillated approximately 2,500,000,000 times per second. ETHER WAVES. 64. As stated in art. 55, a cycle is made up of two alternations or movements in opposite directions and is represented in fig. 18. Such a curve also represents the crest, hollow, and slope of regular waves on the surface of the ocean or other body of water. The distance from crest to crest or from hollow to hollow of a water wave is called a wave length, and this distance is equal to that of two alternations. Since electro- magnetic (ether) disturbances spread in all directions with the speed of light, and when sent out by an oscillating current succeed each other at equal intervals of time, and since the magnetic and electric forces pro- duced by oscillating currents change direction during each alternation, just as the particles of water rise to the crest or fall to the hollow of a wave, their positive and negative amplitudes may represent the crests and MANUAL OF WIRELESS TELEGRAPHY. 45 hollows of waves separated by half periods or half wave lengths, an oscillating current may be called a wave producer, and the oscillations considered as movements of the ether may be called ether waves. 65. Hertz (in 1886 at Carlsruhe, Germany) was the first to show that oscillating electric currents really do produce ether waves like those of light only longer and subject to all the laws governing light waves. For this reason, wireless is sometimes called Hertzian wave telegraphy. 66. The vibrations of particles producing sound waves, as in air, con- sists of to-and-fro movements parallel to the direction of the waves, the latter consisting of alternating conditions of compression and rarefaction of the air. FIG. 18. The movement of the particles in ether waves is at right angles to the direction of propagation of the wave, and the electric and magnetic movements are also at right angles to each other at any point in the wave front. This is called transversal vibration, as distinguished from the longitudinal vibration of the particles in sound waves. When one particle 'of a substance is displaced or made to vibrate, it induces its neighbors to follow it, and starts them to vibrating in the same periods but in different phases, each particle starting to vibrate (passing the word, so to speak) at a definite interval of time after the one next to it has started. The vibrations may be longitudinal or trans- verse, as described above, or they may be circular or elliptical, but if they are regular the waves produced are regular. The amplitude of the wave (art. 55) depends on the extreme limits from its normal position of the vibration of each individual particle. The wave length depends on the time of one complete vibration of each particle and the velocity with which the displacement or vibration is propagated from one particle to another of the substance. It is found that this velocity is equal to the square root of the elasticity of the body divided by its density. 46 MANUAL OF WIRELESS TELEGRAPHY. We know that this velocity in the ether is 300,000,000 meters per second, and we conclude that the ether must have very great elasticity combined with very small density. It has been stated that electric charges or electrons are the only things which have a grip on the ether, and that when they are vibrating the ether vibrates with them. When a particle is subject to several forces at the same time, its resultant movement depends on the resultant of the forces and will vary as the forces vary, so that a body can, in effect, vibrate in more than one way at the same time, and can produce complex waves where vibrations are superimposed on each other. This is shown every day at sea by the small waves or ripples on the slopes of large ones, or the short waves from local winds superimposed and propagated in the same or different directions from the long swells due to distant storms. 67. The vibrations producing ether waves, and consequently the wave lengths and frequencies, are of an almost infinite range, for instance: Ether vibrations from 430 to 740 trillions per second (a little less than one octave) are visible to the eye and are called light. Between 870 to 1500 trillions of vibrations per second we have the ultraviolet and X-rays, and from 430 down to 300 trillions of vibrations per second what are called infrarouge rays. Below 300 and down to 20 trillions of vibrations per second we detect ether vibrations by our sense of feeling or by the thermometer, and they are called heat. Forty-five octaves lower on the same scale are the ether vibrations which we call electric waves and which are used in wireless telegraphy. The shortest of these yet measured is 0.2 of an inch in length; the longest, over 1,000,000 miles. Marconi, in his first experiments, used a pair of small spark balls which gave out waves about 12 centimeters in length. 68. Ether waves of all lengths are subject to reflection, refraction, diffraction, and absorption, and bodies, such as insulators of certain kinds, which are opaque to the short waves we call light, are transparent to the long electric waves used in wireless telegraphy. Practically all conductors are opaque to electric waves. Generally speaking, insulators are transparent to electric waves, but in transmitting the wave they absorb some of its energy. Conductors, being opaque to electric waves, partially reflect and par- tially absorb the wave energy. A simple case of wave reflection is seen when a rope hanging vertically is given a quick jerk and then held taut in the hand. A wave can be seen traveling up the rope till it reaches the top, where it is reflected, travels down the rope to the hand, is reflected there and starts up again to the top, and so continues until its energy is damped out. MANUAL OF WIRELESS TELEGRAPHY. 47 If a number of equally timed jerks are given, a succession of waves at equal intervals is sent up the rope. When reflected back they meet others coming up whose lengths are equal to those coming down. At some points the rope tends to move a certain distance in one direction with the direct wave, and the same distance in the opposite direction with the reflected wave; the result is that it does not move at all. These points are found along the rope one-half wave length apart; at all other points the rope moves or vibrates in the resultant direction of the direct and reflected wave impulse, and what are called stationary waves are set up. The points at which there is no movement are called nodes, and points at which there is maximum movement are called loops. This is shown graphically in fig. 18c. FIG. 18c. Stationary ether waves can be set up around conducting wires by suit- ably timed electrical impulses applied to the ends of the wires. 69. It will be observed that the point of support of the rope where it can not move must, in every case, be a node. So in a conducting wire, the end of the wire away from that receiving the impulses must be a current node, because no current can flow there. It can, however, and a little consideration will show that it must, be a potential loop, for while there is no movement at the point of support, the greatest pressure or tendency to move is there. Since the electrical impulses consist of variations of current and potential, which succeed each other regularly, and since at a given point we find a loop of potential and a node of current, we must, at a quarter- wave length distant, find a node of potential and a loop of current. This is shown graphically in fig. 18d, which represents the relative positions of current and potential nodes and loops in stationary electric POTENTIAL LOOP CURRENT Nooe FIG. 18D. 48 MANUAL OF WIRELESS TELEGRAPHY. waves, and illustrates the statements made in art. 54 (figs. 25a, etc.), relative to the alternations of electric and magnetic fields in oscillating condenser discharges. 70. If an oscillating current be set up in a free wire (fig. 18e) by a neighboring discharging circuit in resonance with it, the free wire will be found by measurement with a micrometer spark gap to have an alter- nating potential in it, varying from nothing at the middle point, C, to a maximum at either end somewhat similar to the full curve EOF. B PIG. 18E. If at the same time the current in the free wire could be measured, it would be found to have a maximum value at C and a minimum at the ends similar to the dotted curve A D B. If the wire A C B is not too far from the discharging resonant circuit and the wire be cut at C and an incandescent lamp L be connected to the two halves as shown in the figure, the lamp will glow. REFLECTION OF ETHER WAVES. 71. If ether waves impinge on a reflecting surface not normal to their direction, they are reflected at an angle equal to that which the reflecting surface makes with their original direction (the angle of incidence is equal to the angle of reflection), so that directed waves may be detected at points not in the line of direction by the interposition of a reflector. Air at atmospheric pressure (about 760 millimeters of mercury) is an insulator. Its density decreases with distance above the earth's sur- face, and its insulating qualities decrease with the decrease of density. At a height of approximately 45 miles above the earth's surface its pres- sure is about 1 millimeter of mercury. At the density corresponding to this pressure it is a good conductor, and though still transparent to short ether waves like those of light, it partly reflects and partly absorbs long ether waves. In the intermediate distance it is at first transparent, then partially transparent, absorbent, and reflecting, simultaneously. It is known that ether waves are guided by conducting surfaces to a certain extent (for instance, by wires), as well as reflected by them, and that otherwise they travel in straight lines. Fig. 18f shows the approxi- mate path of an ether wave started from the earth's surface and reflected MANUAL OF WIRELESS TELEGRAPHY. 49 from the upper atmosphere. It will be seen that even if the earth's surface did not guide the waves they might be detected at points below the horizon. Other causes of reflection may exist., such as large bodies of electrified air, or heavily charged clouds, which would cause interference between direct and reflected waves and make electrical shadows at certain places, i. e., points at which, owing to conditions outlined above, either the waves are so attenuated that they can not be detected or they are com- pletely neutralized. FIG. 18F. REFRACTION OF ETHER WAVES. 72. When ether waves impinge on transparent bodies at any angle other than the normal, if their velocity in the transparent body, on account of its elasticity or density, is different from that at which they were previously moving, that part of the wave first entering the body will move either faster or slower than it did before. The part outside will therefore either fall behind or gain on the first part. This action will affect each portion of the wave front as it enters the body, and the result will be that its direction of movement will be changed. The effect is to bend the wave out of its original path, and the action is called refraction. Ether waves passing through the atmosphere, whose density varies at different points, are subject to this bending action. DIFFRACTION OF ETHER WAVES. 73. When waves meet a body in their path (for instance, when the comparatively long waves used in wireless telegraphy impinge on a high island or mountain range) at the points where the wave front cuts the extreme width of the island and along the crest or summit new cen- ters of disturbance are created, which radiate some of the wave energy to points behind the island. It has the effect of bending the waves around the object, This action of waves is called diffraction. In amount it depends on the wave length. From the new centers of disturbance waves are sent out, which interfere with each other, not being propagated in the same directions. The result is that for a distance, depending on the 50 MANUAL OF WIRELESS TELEGRAPHY. width and height of the obstacle and on the wave length, a shadow exists beyond it. Partial reflection of the waves toward their source takes place on the side of the obstacle nearest the source. An attempt to show this graphi- cally is made in fig. 18g, but the best illustration is given by the motion of water around a rock on a windy day. The small back waves on the windward side are reflected to windward. The waves circling or bend- ing around the rock are diffracted. The still water in the lee of the rock is the shadow, in which no action exists. At a distance depending on the size of the rock and the wave length the zones of interference disap- FIG. 18a. pear, the regular waves from the two sides of the rock unite, and there is no evidence of its existence at points beyond, though it has decreased the total strength of the waves. For the above reasons high land between two wireless telegraph stations has the effect of decreasing the strength of signals at each station, and if close to either station may entirely prevent that station from receiving. (It may be in the shadow or be subject to interference from reflection.) The effects of reflection and diffraction on waves passing over irregular country are very pronounced. The effects of reflection, refraction, and absorption in the atmosphere are equally pronounced, the qualities of the atmosphere in all three respects varying greatly from day to day and between day and night. An ether wave traveling from one wireless-telegraph station to another MANUAL OF WIRELESS TELEGRAPHY. 51 over rough country and through an atmosphere of varying density, work- ing its way around and over mountains, being balloted from thunder clouds at one point and absorbed by semiconducting gases at another, may be said to pursue an adventurous journey. PRODUCTION OF ETHER WAVES. 74. We have now seen how to produce electric and magnetic fields, how to utilize magnetic fields for the production of electric currents in dynamos, how to increase the potential of these currents by means of step-up transformers, and ho\v by means of this high potential current to force large charges into electric accumulators or condensers and by discharging these condensers in oscillating circuits to produce what we call electric or ether waves. These operations can be represented graphi- cally or diagrammatically, as in fig. 29, which shows a separately excited A. C. dynamo in circuit with the primary winding of a step-up trans- former, whose secondary charges the condenser of an oscillating circuit containing a spark gap. GROUND FIG. 29. The secondary winding of the transformer is of many turns, in order to give a high potential. The transformer also has an iron core. The great number of turns of the secondary winding, added to the effect produced by the iron core, gives the circuit containing the secondary winding and the condenser a very large self-induction, and consequently a very long period. The circuit composed of the condenser, self-induc- tion, and spark gap has a very much shorter period, and when the spark gap is ruptured this circuit oscillates as if it were entirely disconnected from the secondary, usually completing its oscillations and coming to rest in a fraction of the period of the circuit formed by the secondary winding and condenser. MANUAL OF WIRELESS TELEGRAPHY. The oscillating circuit (condenser, spark gap, and inductance) is shown in fig. 29 near a conducting wire, having a few turns of inductance close to those of the oscillating circuit. In this circuit we can con- sider the condenser as representing the source of current, like the hattery in fig. 11, art. 12; the spark gap as the break K, the turns of inductance in the oscillating circuit as A B, and the open circuit with one end grounded as C D. The oscillating currents in A B produce like cur- rents, but in the opposite direction in C D (art. FIG. 11. 12), and C D becomes a source of ether waves. 75. The production of ether waves and their detection at a distance from the source constitutes wireless telegraphy. C D is usually called the open or radiating circuit. A B the closed or oscillating circuit. The two inductances in A B and C D form the primary and secondary, respectively, of an air-core oscillation transformer (art. 27). When arranged as in fig. 29, A B and C D are said to be inductively connected. C D may have part of its inductance common to A B. The arrange- ment in this case acts as an auto-trans- former, and the circuits are said to be direct connected (fig. 29a). If the oscillating and radiating cir- cuits have the same period, they oscillate C O(^ or vibrate in resonance. The radiating J the prac- tical unit of resistance, which is the ohm, is taken as 1,000,000,000 times the theoretical or absolute unit. Ohm's law then still remains true. I -= or amperes =- ^, MANUAL OF WIRELESS TELEGRAPHY. 57 because this equation in terms of the absolute units is ^TT (amperes) = (volts) wMch ig the game ag I= E m The size Qf 72X1,000,000,000 (ohms) ' R the units has been changed, but the proportion between them is the same as before. WATT. The practical unit of power is the watt, which is the ergs of work done per second when 1 ampere is flowing with an E. M. F. of 1 volt. This in ergs (see equation (2)) equals unit E. M. F.x 100,000,000 X unit current^ Qr 10 ^ 000 ^ 000 ergg per second. Therefore 1 watt equals 10,000,000 ergs per second. The power expended in any circuit in watts equals the product of the volts and amperes in the circuit, or P=IE (2). Ten million ergs of work is called a joule. Therefore a watt=l joule per second. We have seen that 1 H. P. = 7,460,000,000 ergs per second. There- fore 1 H. P. = 746 watts. 1 watt = approximately 0.737 foot-pounds per second. 85. After having selected the practical units, it became necessary, for the purpose of comparison and for everyday use, to represent them in practical form, because the accurate measurement of dynes and ergs is a very difficult matter practically. But it can be done in accordance with definitions given in art. 78. Also art. 81 indicates how to measure the strength of magnetic fields and how to determine and compare E. M. Fs. and currents by the ergs of work done in creating them. A volt or an ampere can thus be definitely created. The current from certain primary batteries is found to be constant when their terminals are connected by the same wire : Since current and resistance are constant, the voltage of such cells must be constant, and this voltage once determined by comparison with absolute volts as determined above, we have at once a practical concrete standard of E. M. F. It is found that the decomposition of an electrolyte (art. 1), by an electric current, always results in the separation or deposit of exactly equal quantities of the constituents of the electrolyte for equal quantities of current. The deposit in a certain time, being weighed, serves as a very accurate measurement of the amount of electricity which passes in that time, and consequently affords a very accurate means of comparing electric currents. When 1 ampere determined as above is passed through a given electrolyte, the weight of material deposited gives us at once a practical standard of current. 58 MANUAL OF WIRELESS TELEGRAPHY. XT 86. On account of the relation 7=-=-(l) between amperes, volts, and xt ohms in a circuit, if any two of them are known the other is also known, so that only two measurements of concrete units are required. The question of which two should be selected and the exact form that each should take has been the subject for discussion at a number of inter- national conferences, the latest of which has recommended that only two electrical units shall be chosen as fundamental units, viz., the inter- national ohm defined by the resistance of a column of mercury, and the international ampere defined by the deposition of silver. The volt to be defined as the E. M. F. which produces an electric cur- rent of 1 ampere in a conductor whose resistance is 1 ohm. Different methods of measurements produce slight differences in the values of the standards, but the values recognized by law in the United States are as follows : The standard (international) ohm is the resistance offered to an un- varying electric current by a column of mercury at the temperature of melting ice 14.4521 grams in mass of a constant cross-sectional area, and of a length of 106.3 centimeters. The standard (international) ampere is the unvarying current which, when passed through a solution of 'nitrate of silver in water in accordance with certain specifications, deposits silver at the rate of 0.001118 of a gram per second. As previously stated, a volt is the E. M. F. which if steadily applied to a conductor whose resistance is 1 ohm will produce a current of 1 ampere: but a concrete standard for the volt is also recognized by law, it being specified: That the electrical pressure at a temperature of 15 centigrade between the poles or electrodes of the voltaic cell known as Clark's cell, prepared in accordance with certain specifications, may be taken as not differing from a pressure of 1.434 volts by more than 1 part in 1000. The latest international conference has recommended the adoption of the Weston cadmium cell as preferable to the Clark for a standard cell. The Weston cell has an E. M. F. of 1.018 volts at 20 C. Standard resistance wires having a known ratio to the legal ohm are made, and these, with standard cells, are used for calibrating volt meters and ammeters, which are the names given to the galvanometers for indi- cating automatically the E. M. F. and current in any circuit. In this way electrical values are made uniform throughout the country. 87. In addition to the volt, the ampere, the ohm, the watt, and the joule other practical units have been adopted, the most important of which, for our purposes, are : MANUAL OF WIRELESS TELEGRAPHY. 59 COULOMB. The unit of quantity, the coulomb, which is the amount of electricity passing any point in a second when 1 ampere is flowing in the circuit. FARAD. The unit of capacity, the farad. A condenser is said to have a capacity of 1 farad when 1 coulomb of electricity will charge it to a potential of 1 volt, (Potential and E. M. F. are in some senses identical, one being the passive and the other the active state. An E. M. F. is the result of difference of potential.) If this definition were in terms of the absolute units, that for capacity would read : A condenser is said to have unit capacity when one unit of electricity will charge it to unit potential. Since by definition a condenser has a capacity of one farad when one-tenth of the absolute unit of elec- tricity charges it to a potential of 100,000,000, a farad must equal x = 10 ' 9 absolute imits of HENRY. 88. The unit of self-induction, the henry. A circuit is said to have a self-induction of 1 henry when, if the current in it is varied at the rate of 1 ampere per second, the induced E. M. F. that is, the counter or reacting E. M. F. tending to oppose the change is 1 volt. The definition of self-induction in terms of the absolute units would be : A circuit is said to have unit self-induction when, if the current in it is varied at the rate of one unit per second, the E. M. F. of self-induc- tion is unity. Since by definition a circuit has a self-induction of one henry; when, if the current is varied at the rate of one-tenth of unit current per second, the E. M. F. of self-induction is 100,000,000, such a circuit would have an E. M. F. of self-induction 10 times as great, or 1,000,000,000, if the current instead of being varied at the rate of one- tenth unit per second were varied at the rate of one unit per second. Therefore the unit of self-induction, the henry, is. equal to 1,000,000,000 = 10 9 absolute units of self-induction. By agreement among electricians self-induction is represented by the letter L; capacity, by the letter C. * When quantities are dealt with having a large number of ciphers either before or following the significant figures it is convenient to express them as multiplied by some power of ten, i e., 10 = 10 1 , 100 = 10 2 , T ^= 10- 1 , T fo = 10-*, etc. 5 60 MANUAL OF WIRELESS TELEGRAPHY. Self-induction, when expressed in terms of the fundamental units of length, mass, and time, has the dimensions of a length, and the prac- tical unit of self-induction was formerly called a quadrant on account of the fact that in the metric system, an earth quadrant i. e., the dis- tance from the equator to the north pole =1,000,000,000 centimeters, and since the henry = 1,000,000,000 absolute units of self-inductance, it was said to = 1,000,000,000 centimeters. In this notation a millihenry = 1,000,000 centimeters. (See art. 91.) 89. The units which have been considered in this chapter have been derived from the relations between electric currents and magnetic fields and are called electro-magnetic units. Another system of units, also based on the centimeter, gram, and second, called electro-static units, is in use. The relation between the absolute units of quantity in the two systems is the velocity of light in centimeters per second. This velocity is 30,000,000,000, or 3xl0 10 centimeters per second, and the electro- magnetic unit of quantity = 3 x 10 10 electro-static units. The coulomb, being one-tenth of the absolute unit, = 3 X 10 9 electro- static units. The electro-magnetic unit of potential is -gfa of the electro-static unit. In both systems a condenser is said to have unit capacity when unit quantity of electricity charges it to unit potential. From the definition of a farad, given in art. 87, we see that the quantity of electricity in a condenser equals in coulombs the potential in volts multiplied by the capacity in farads, or Q = EC, .' .C -p- . Sub- stituting for Q and E their unit values in electro-static units given 3 X 1 9 above, G j = 9xlO n , or the practical electro-magnetic, unit of JTTO capacity is 9 x 10 11 times as large as the electro-static unit. The capacity of spherical bodies is found to vary as their radii, and in the electro-static system a sphere of 1 centimeter radius has unit capacity; therefore the capacity of a sphere may be expressed by its radius in centimeters, and capacities are still expressed by some writers and manufacturers by the radius in centimeters of the equivalent sphere. A condenser having a capacity of 1 farad has a capacity equal to that of a sphere having a radius of 9 X 10 11 centimeters. A microfarad (see art. 91) =10~ 6 farads, is equal to a capacity 9x 10 11 xlO- = 9xl0 5 , or 900,000 centimeters in electro-static units. The earth's radius is approximately 65xlO T centimeters; its capacity should be approximately 7000 microfarads. 90. This difference in units is very confusing, but it exists particularly with reference to the two qualities of self-induction and capacity with which wireless telegraphy is intimately concerned. Microfarads and MANUAL OF WIRELESS TELEGRAPHY. 61 millihenrys will be used in this book, and where centimeters are found as in some catalogues and some books on electricity, the relations here given 1 millihenry = 1,000,000 centimeters electro-magnetic units; 1 microfarad = 900,000 centimeters electro-static units will enable one set of units to be converted into the other. The entire system of units used in electrical measurements is a monu- ment to the ingenuity of scientists, but productive of many difficulties to students. 91. While the volt, the ampere, and the ohm are really practical units, the farad and henry are too large for practical use. It would take a very large condenser to have a capacity of 1 farad and a coil of many turns to have a self-induction of 1 henry. Sub- divisions of the farad and henry are in practical use. Multiples and subdivisions of the other units are also frequently used, and for this purpose the prefixes kilo, meaning 1000; mega, meaning 1,000,000; milli, meaning , and micro, meaning ^ QQO Qo() , are added to the units, and such terms as kilowatt =1,000 watts, megohm =1,000,000 ohms, millivolt = milliampere = --.--AA ampere, millihenry = - henry, microfarad= i^oiooo farad ' second, are in common use. The most common practical units of capacity and self-induction (the qualities of electric circuits with which wireless telegraphy is principally concerned, because they determine the period of vibration) are the microfarad and the millihenry, but even these are too large for convenience. The terms mil, meaning i- inch; micron, meaning 1 Q00 ^ meter; circular mil, meaning area of a wire having a diameter of j^ inch, are also in general use among electricians. 92. The E. M. F. (volts) in any circuit connected with a dynamo depends only on the rate of cutting of lines of force (strength of field and rate of revolution). 62 MANUAL OF WIRELESS TELEGRAPHY. The resistance (ohms) in any circuit depends only on the material, cross section, and length of the conductor forming the circuit. The current (amperes) in any circuit depends only on the E. M. F. and the resistance in the circuit. The power (watts) in any circuit depends only on the E. M. F. and current in the circuit. The self-induction (henries) in any circuit depends only on the shape and length of the circuit, on 'the magnetic permeability (art. 25) of the material surrounding and inclosed by the circuit, and on the amount of this material. The capacity (farads) in any circuit depends only on the shape and area of its surface, on the electric permeability (art. 46) of the material surrounding the circuit, on the amount and location of this material (the dielectric), and on the position of the circuit relative to other conductors. (Straight wires are said to have distributed inductance and capacity, coiled wires have concentrated inductance, and condensers have con- centrated capacity. ) The coulombs in a charged condenser or circuit depend only on the capacity and potential of the condenser or circuit. 93. From the foregoing we can make up a table of values as follows : A volt = 100,000,000 = 10 8 absolute units of E. M. F. An ohm = 1,000,000,000 = 10 9 absolute units of resistance. An ampere -^ = 10^ absolute units of current. A watt a volt x an amp. = 10 8 x 10~ 1 = 10 7 absolute units of work per second= 1 joule per second = T J^- H. P. = 0.737 foot-pounds per second. A horse power =746 watts. A kilowatt=1000 watts. A farad= 1,000,000,000 =10 ~ 9 absolute units of capacity ' A farad in electro-static units = 9x!0 11 centimeters. A microfarad = 1 farad =1Q- 15 absolute units of capacity. A microfarad in electro-static units = 900,000 centimeters. A henry = 1,000,000,000 =10 9 absolute units of self-induction. A millihenry = - henry = 10 6 absolute units of self-induction. 1UUU A millihenry in electro-magnetic units = 1,000,000 centimeters. A standard Leyden jar or plate having a capacity of .002 microfarad has been adopted for naval use. In electro-static notation 1 standard jar has a capacity of 1800 centimeters. Chapter III. CAPACITY, SELF-INDUCTION AND MUTUAL INDUCTION IN WIRELESS TELEGRAPH CIRCUITS. FUNDAMENTAL EQUATION OF WIRELESS TELEGRAPHY. 94. It was stated in art. 56 that the period of electrical vibration of any circuit depends only on the capacity and self-induction of the circuit. When the ratio of the resistance to the self-induction of a circuit is small,, and the circuit vibrates in its own period, the period is found to be equal in seconds to 2?rV LC (3) when L is measured in henries, C is measured in farads, 71- = 3. 1416. This is called the fundamental equation of wireless telegraphy. (See table 7, appendix A.) If R is greater than 2 J ^ the circuit will not vibrate at all. For V c instance, when a condenser is discharged through a wire of great resist- ance the charge leaks out slowly without any oscillation. A nonoscillatory condenser discharge, as compared with an oscillatory discharge, is like the flow of molasses into a jar as compared with a large and sudden flow of water into a similar jar. One takes up a position of equilibrium slowly but surely, while the other vibrates and splashes and only settles down after a considerable period. Equation (3) shows that a circuit having a self-induction of 1 henry and a capacity of 1 farad would have a period of ZTT= 6.2832 seconds. Its wave length would be 1,168,000 miles. The standard wave length originally adopted for naval wireless tele- graph circuits was 320 meters; the period was approximately -g-jnnnnr second, that is, they made approximately 900,000 complete vibrations per second. The usual capacity in these circuits was 0.014 microfarad (seven 0.002 microfarad jars in parallel). Therefore the self-induction must have been 0.0022 millihenry. It will be noted that the period of a circuit varies as the square root of the product of the inductance and capacity, so that doubling either of these increases the period by V2, i. e., to 1.414 times its former value. Doubling both inductan'ce and capacity doubles the period. SELF-INDUCTION. 95. We see that all conductors must have self-induction, because we know that all currents are surrounded by magnetic fields produced by 64 MANUAL OF WIRELESS TELEGRAPHY. the currents. The production of the field creates an E. M. F. in the circuit opposite in direction to the E. M. F. causing the current and tending to stop it, so that self-induction has been defined in a qualitative manner as the inherent quality of electric currents which tends to impede the introduction, variation, or extinction of an electric current passing through an electric circuit. It has also been expressed in quantity as the number of lines of force induced in a circuit by the establishment of unit current in it. It bears the same relation to electricity as inertia does to matter; it represents its resistance to change of condition, and it might be defined as the work necessary to create unit current in a circuit. Suppose we wish to determine the work done in creating a current of value I in a circuit of self-induction L in a time T. Since L=the counter E. M. F. of self-induction when the current is varied at the rate of 1 ampere per second, the counter E. M. F. when it is varied at the rate of -- amperes per second = - . If the rise in current is uniform, the counter E. M. F. is uniform and the total work done (which equals the product of the E. M. F., current, and time) would be equal to -~ x!xT=LI 2 , were it not for the fact that j- the current rises uniformly from zero to 7 and its mean value is -=- and /v T 72 therefore the work done= =^- (4). Since the factor of time does not appear in the result it shows that it requires the same amount of work to create a given current in a circuit of given self-induction whether it is created slowly or quickly, and that this work is equal in 'joules to one-half the product of the self-induction in henries by the square of the current in amperes. Therefore in a circuit whose self-induction is 2 henries the work done in creating a steady current of 10 amperes is O NX 1 A2 equal to - - = 100 joules = 73.7 foot-pounds. values being given in centimeters. The self-induction d of two parallel wires varies as the distance between them, decreasing with the distance, so that adding straight wire to a*n aerial does not add to its self-induction in the same proportion. The relation between the inductance and capacity of a straight wire of circular section and diameter small in comparison with its length is such that its electrical length is equal to its natural length, and its wave length is therefore twice its natural length. A vertical straight wire, well grounded and of small diameter, has an apparent electrical length approximately equal to twice its natural length; its wave length is approximately four times its natural length. Pierce states that a single wire 100 feet long and -J inch diameter, when alone in space has as much capacity as an isolated flat metallic disc 16 feet in diameter.* CONDENSERS IN SERIES AND IN PARALLEL. 105. When two or more condensers are placed in parallel (fig. 28c), their total capacity C is equal to the sum of their capacities taken singly ; *j ^ c4 FIG. 28c. FIG. 28D. i. e., C = Oj + C 2 + etc. When two equal condensers are placed in series (fig. 28d), the resulting capacity is one-half of that of each taken singly, or in general * Principles of Wireless Telegraphy, by G. W. Pierce, A.M., Ph.D. (1910). MANUAL OF WIRELESS TELEGRAPHY. 71 A condenser is often placed in large aerials to shorten the natural wave length for receiving. The aerial in this case being considered as one plate of a condenser and the earth the other, when a condenser is placed between the aerial and the earth we have two condensers in series. Condensers which will be ruptured if used alone can be used in series, dividing the voltage between them. For instance, take a transformer giving 30,000 volts to be used in con- nection with condensers that will stand but 20,000, by placing 2 in series each condenser would have to stand but 15,000 volts. It will be seen that 32 jars made up into 2 condensers of 16 jars, in parallel, in each and the two condensers placed in series would only have the capacity of a single condenser of 8 jars in parallel, but the work on each jar would be four times lighter. 106. If a straight wire is broken in the middle the oscillation period of each half would be half the original period were it not for the fact that the adjacent ends of the wire and the air between them form a small condenser which has the effect of slightly increasing the capacity of each half, thus giving it a period slightly longer than half of the original period. From the above it appears that we can shorten the electrical length of an aerial (radiating circuit, art. 75, fig. 29) by putting a condenser in series with it, but we can not shorten it to less than one-half its original period. We know that by coiling the wire we can increase its self-induction and, therefore, its electrical length and wave length with very little increase in its physical length. In practice, wave lengths of the closed circuit (art. 75) are altered by changing their self-induction as above. The wave lengths of the open circuit (aerial) are increased by adding inductance. They are decreased by adding condensers in series (for receiving only). MEASUREMENT OF INDUCTANCE AND CAPACITY IN OSCILLATING CIRCUITS. 107. By comparison with standard inductances and capacities, the capacity and self-induction of circuits can be measured and their periods calculated. Measured inductances and capacities connected together so as to form an oscillating circuit are made so that the capacity or inductance (usually the capacity) or both are variable. They can be calibrated so as to show directly either the period or wave length of the circuit for any position of the variable elements. If brought near another circuit in which electrical oscillations are taking place and adjusted so as to have a maximum of current induced the two circuits are said to be in tune or resonance. (They have the 72 MANUAL OF WIRELESS TELEGRAPHY. same electrical length.) When used as above, calibrated oscillating cir- cuits are called ivave meters, ondameters or cymometers. Wave meters can be so arranged as to measure separately the induc- tance or capacity of oscillating circuits as well as their periods. If a spark gap forms part of the oscillating circuit, its period can also be directly measured by measuring the time between the successive surgings of the spark. This is done by photographing the sparks by reflection from the surface of a rapidly revolving mirror. The movement of the mirror between sparks separates their images on the photographic film, and knowing the number of revolutions of the mirror per second, the elapsed time between sparks can be calculated and hence the period of the circuit. INDUCTANCES. 108. In sending circuits the capacities (condensers) are usually fixed and the wave lengths are varied by varying the inductance (self-induc- tion). This usually consists of a bare copper wire, tube, or ribbon coiled so as to form a helix. Of course the self-induction of such a coil could be very greatly increased by providing it with an iron core, but the magnetic hysteresis loss (art, 148) would be too great. The magnetic hysteresis loss in air, like the dielectric hysteresis loss, is practically zero ; therefore, these inductances have air cores and as much of their length is included in the circuit as will, in connection with the fixed capacity (condensers), give the wave length desired. (See fig. 73 for photograph of sending helix.) MUTUAL INDUCTION AND ITS EFFECT ON OSCILLATING CIRCUITS. 109. Mutual induction between two circuits is explained in art.. 15. It is represented by the letter M and is defined as the E. M. F. generated in one circuit when the current in the other circuit is varied at the rate of one ampere per second. The two circuits are coupled together by virtue of their mutual induction and the induced current represents a transfer of energy from one circuit to the other. If their mutual induction is large, they are said to have close or tight coupling; if small, the coupling is said to be loose. It is evident that the mutual induction between two circuits depends on the self-induction of each, that is, the strength of the magnetic fields produced by varying the current. Also that it depends on the distance apart of the two circuits and the material (iron or air) intervening. It is a maximum when all the lines of force created by the current in either circuit cut the other. In this case the coupling is said to be perfect. If the two circuits in the case of perfect coupling have the same self-induction their mutual induction is equal to the self-induction of each; if different the mutual induction in such a case is equal to V LL 2 , L being the self-induction of one circuit, L 2 that of the other. MANUAL OF WIRELESS TELEGRAPHY. 73 If the two circuits are moved in relation to each other so that only part of the magnetic field created by each cuts the other, their mutual induc- tion is decreased. The ratio of the mutual induction (for any position of the circuits) to its maximum value is called the coefficient of coupling for that position, or coefficient of coupling = _ =. The mutual induction between two oscillating circuits alters the effec- tive self-induction of each (art. 92), making it apparently larger or smaller as one circuit is receiving energy from or transferring energy to the other. 110. Since the natural period of a circuit in seconds ^ir^jLG (3), if L, the effective self-induction, is varied, the period of the circuit is varied. Coupled circuits having the same or nearly the same natural periods are found to have two periods of oscillation, one faster and the other slower than the natural period of each. Therefore the open radiating circuit sends out electric waves of two lengths, one longer and one shorter than the natural wave length of the circuit. The closer the coupling the greater the difference in length of these two waves. This difference divided by the natural wave length of the circuit is called the per- centage of coupling. This can be more easily ascertained than the coefficient of coupling. For instance, if an open circuit, having a natural wave length (as determined by a wave meter) of 400 meters sends when coupled to a closed circuit of the same natural length two waves, one of 445, the other 365 meters, the percentage of couplings ^ = 20$. 111. If the circuits have loose coupling, i. e., are moved farther apart, the mutual induction is less and the difference in the wave length radiated is less. This distance can be increased until the two waves practically coincide with the natural wave length of the circuit. This is very loose coupling, but, since without mutual induction, no energy can be trans- ferred, the two can never be the same. Most of the energy is found to be in the longer wave and until recently that in the short wave was practically wasted. The method now used of generating but one wave length will be described under sending apparatus. Chapter IV. ELECTEIC OSCILLATIONS AND 3ADIATIOX OF ELECTRIC WAVES. 112. It has been stated that every oscillating circuit must contain inductance and capacity. This is true even though the circuit consists of straight wires, for these have distributed inductance and capacity. If the circuit is formed as in fig. 26a with a coil of wire and a condenser, the inductance and capacity are said to be concentrated or lumped. There is also a certain amount of distributed inductance and capacity, but in general this will be small compared with the concentrated portions. FIG. 26A. FIG. 30. FIG. 31. FIG. 26A. Non-radiating Circuit. FIG. 30. Radiating Circuit. FIG. 31. Electric Wave Leaving Oscillator. In the case of a linear oscillator (fig. 30), when the oscillations are taking place and the charges are most widely separated, we may imagine lines of electric force to be connecting each unit of positive electricity on one end to a unit of negative electricity on the other. For clearness of conception we may picture these lines of force as having a real exist- ence and exerting an elastic pull between the positive and negative units, tending to draw them together, while at the same time, provided they are running in the same direction, they tend to repel each other. These lines of force in the case of a linear oscillator, on account of their repulsion away from the oscillator, form wide loops which tend to snap off and travel away into space when the charges again rush back through MANUAL OF WIRELESS TELEGRAPHY. 75 the spark gap, thus forming electrical waves or radiation as shown in fig. 31. In the case of the circuit shown in fig. 26a, where the principal capacity lies in the condenser, the lines of force are concentrated between the condenser plates. They do not loop out to any extent, and hence such a circuit radiates very feebly. On account of these differences an open circuit oscillator (fig. 30) is often called a radiating circuit, while a closed circuit (fig. 26a) is called non-radiating, although all high frequency circuits radiate in some degree. 113, Let fig. 32 represent a closed circuit inductively connected to a vertical grounded open circuit or aerial, and suppose the spark gap to break down at the point of maximum potential of the charging current. At this instant there is no current in the closed circuit and, therefore, no current in the open circuit. The energy is all electro-static, all in the closed circuit and practically all in the electro-static field between the condenser plates, the capacity of the spark points and other parts of the circuit being very small. ? FIG. 32. FIG. 32A. As soon as discharge through the spark gap commences the field of the current in the closed circuit inductance induces movements of electric charges in the open circuit, the starting point of the disturbance being the open circuit inductance. As the charges in the open circuit separate they are connected by electro-static lines of force and surrounded by magnetic lines of force, both moving outward at the same rate that the charges move in a straight wire. The electro -static field becomes a maximum when the charge reaches the top of the wire. At this time the magnetic field is a minimum. At the expiration of a half period, when the charges meet again, the mag- netic field is a maximum, but reversed in direction. The electro-static 6 76 MANUAL OF WIRELESS TELEGRAPHY. field reverses as the charges separate again. If they can be represented as meeting in the open circuit inductance, the electro-static field just after the end of a half period can be represented as in fig. 32a, where the mutual repulsion of the electro-static lines of force outside the wire has kept them from returning as fast as the charges travel towards each other. As the charges meet, the ends of the lines of electric force unite and become closed circuits, or electric whorls shaped like smoke rings which, owing to the mutual repulsion of all their parts, expand outward, upward, and downward. It is in some such manner that we can conceive energy to be detached and sent out into space from wires forming oscillating circuits. The expanding rings touch the earth and are guided by it as by any other conductor, thus resembling near the earth expanding hemispherical shells. These may be called earthed waves to distinguish them from the free waves which exist momentarily in the vicinity of an ungrounded oscillator. 114. If the point of connection with the closed circuit is considered as at the earth, earthed waves only are generated and detached from the aerial and no free waves exist at any time. The production of earthed electric waves under these conditions is illustrated in fig. 33. FIG. 33. Earthed Electric Waves. We know that earthed waves are guided by conducting surfaces; we know that light waves are not; we do not know where the dividing line is between waves that are radiated in straight lines and those that are guided by conductors. 115. For simplicity we have described the process of radiation in terms of electro-static lines of force, but it must not be forgotten that a moving electro-static field always produces a magnetic field at right angles to itself and at right angles to the direction of movement, so that we have electro-static lines perpendicular to the surface of the earth (at least near the aerial), and magnetic lines in circles surrounding the aerial. Both the electro-static and the electro-magnetic fields reverse their directions every half wave length. MANUAL OF WIRELESS TELEGRAPHY. 77 r The process of radiation withdraws energy from the circuit just as is the case when a resistance is placed in the circuit; hence the radiation resistance is an expression often used, meaning the resistance which under the given conditions would use up the same amount of energy as that removed from the circuit by radiation. This radiation resistance depends only on the form and dimensions of the aerial and on the frequency of the oscillations,, increasing rapidly as the frequency increases. It is independent of the intensity of the oscillations and of other sources of lost energy in the circuit. Radiation resistance might be called the radiation coefficient. Accurate means of measuring it are not yet in general use. DAMPED OSCILLATIONS. 116. It has been explained (art. 54) that when a circuit consisting of a condenser, inductance, and spark gap is charged by a transformer to a potential so great that a spark passes across the gap, the electricity stored up in the condenser discharges itself through the spark gap, and by its inertia charges the condenser in the opposite sense, only at the next instant to again discharge itself, and so on. All this takes place during the time of one spark, and in fact this surging of electricity is what keeps the spark in existence after the first discharge. This surging back and forth would continue indefinitely were it not for the energy used up in the heat of the spark and in the resistance and other losses in the rest of the circuit, But as no new energy can be introduced into the circuit until the condenser is recharged, the electrical surgings decrease in intensity and finally cease. If we represent time by the horizontal axis and the amplitude of the oscillations by the vertical axis, fig. 34 will show graphically the course FIG. 34. Damped Oscillations. Energy Supplied at Beginning of Wave-train. of the phenomenon. It is exactly analogous to a light pendulum which is set swinging and which is brought to rest after a limited number of swings by the friction of the air. Gradually decreasing oscillations of this kind are called damped oscil- lations and obey the law that each succeeding amplitude is a given fraction of the one before it, that is, the amplitudes form a geometric series. 78 MANUAL OF WIRELESS TELEGRAPHY. 117. For purposes of calculation it is sometimes convenient to make use of a system of logarithms which instead of using 10 for its base, as is the case with common logarithms, uses the number 2.7183. These natural logarithms can always be obtained from the common logarithms by multiplying the latter by the factor 2.3026. If the natural logarithms of the successive amplitudes of our oscillations be taken, it will be found that the successive logarithms differ from each other by a constant num- ber. This number in the present case is known as the logarithmic decrement of the oscillations. Its chief interest to us lies in the fact that it is a measure of the energy losses in the circuit. Wave Natural logarithms Wave Natural logarithms amplitudes of the amplitudes amplitudes of the amplitudes in the ratio (constant difference^ in the ratio (constant difference) 1Q S = 0.223 10 = 0.223 1000 6.908 410 6.016 800 6.685 328 5.793 640 6.462 262 5.569 512 6.238 210 5.347 It is often expressed by the symbol 8 (delta). If we express all the losses in the circuit in terms of a resistance R which would give us the T> equivalent loss, 8===, where n is the frequency and L the induct- ance of the circuit, R and L being expressed in corresponding units, for instance, absolute units. This formula enables us at once to determine the equivalent resistance of the circuit when the damping has been measured. For the derivation of this formula and for the general mathe- matics of the damping theory, the reader must be referred to mathe- matical works on the subject. Some authors define 8 as the logarithmic decrement per half oscillation, but following the more general usage we have defined it as the decrement per whole oscillation, that is, between two oscillations in the same direction. UNDAMPED OSCILLATIONS. 118. It has been seen in the last article that the cause of the dying out of a train of oscillations in a spark circuit is the using up of energy in the circuit together with the fact that no energy can be brought in FIG. 35. Undamped Oscillations. Energy Constantly Supplied. from outside to compensate this loss. If means can be found for keeping up a constant supply of energy, our oscillations can be made to continue indefinitely and with equal amplitude (fig. 35). MANUAL OF WIRELESS TELEGRAPHY. 79 119. The electric waves produced during one set of oscillations are called a wave train. If more than one, the wave trains produced during one-half cycle of the charging current are called a group of wave trains. The duration of a wave train is the time of one oscillation multiplied by the number of oscillations in the train. It is found that the duration of a wave train is much less when the oscillating circuit (A, B, K, fig. 29) is connected to an aerial with one end free and the other earthed, like C D, than when it oscillates without any other electrical connection. The energy is radiated more rapidly, the vibrations more quickly damped. For this reason the circuit formed by the condenser, spark gap> and inductance is called the dosed or oscil- lating circuit; that formed by the aerial, inductance and ground, the open or radiating circuit. (See art. 75.) FIG. 29. 120. Considering the series of expanding hemispherical shells referred to in art. 113, and shown in fig. 33, if there is but one wave train per alternation of the condenser charging current, the thickness of one of these series is equal to the wave length multiplied by the number of oscil- lations per train. Suppose this to be 10 and the wave lengths 500 meters, then the depth of a wave train is 5000 meters, or a little more than three miles. If the frequency of the alternating current is 60 cycles, or 120 alternations per second, we have 120 wave trains per second, and since they travel at the rate of 186,000 miles per second the wave trains have intervals of 1550 miles between them, so that working at ordinary dis- tances and this frequency, each wave train has passed the receiving 'station before its successor has left the sending station. If the alternator frequency is 500, the wave trains are only 186 miles apart, or about the distance of ordinary daylight communication between ships. 80 MANUAL OF WIRELESS TELEGRAPHY. 121. Professor G-. W. Pierce, of Harvard University, has measured the period of some types of oscillating circuits used in wireless telegraphy, and it is from his published account of his experiments that the follow- ing description is derived. Suppose a spark gap set to break down at a potential of 10,000 volts, to be used in a circuit where the maximum potential reached in the condenser is 30,000 volts. Let the curve of sines in fig. 18 represent the condenser potentials of the oscillating circuit during 2 alternations, each lasting y^ of a second. The resistance of the spark gap is practically infinite before the poten- tial reaches 10,000 volts, and therefore no current passes. When the potential has risen to 10,000 volts the spark gap is ruptured. Its resist- ance decreases instantly to a fraction of an ohm, and during the first half of the oscillation the condenser is discharged to zero potential. Dur- ing the last half of the oscillation it is charged again in the opposite sense. The sparks pass first in one direction and then in the other, and the spark gap not regaining its resisting qualities, the oscillations or surgings continue until the potential (owing to losses due to the radiation of energy in the shape of electric waves, to heating the circuit, and the light and heat at the spark gap) does not rise high enough to disrupt the gap. The transformer immediately recharges the condenser, which, as soon as it again reaches a potential of 10,000 volts, breaks down the spark gap again, and a second series of oscillations begins. In the circuit under consideration the maximum charging potential is 30,000 volts, so that a condenser with a spark gap breaking .down at 10,000 volts may be charged and discharged several times during one- half cycle of the charging current. The spark acts like a trigger which suddenly releases the stored energy in the condenser, and as soon as this energy has been radiated, the trigger automatically resets itself and does not release again until the condenser is recharged. It is evident that if the spark gap in the circuit under consideration is adjusted to 30,000 volts, but one discharge of the condenser per alterna- tion will take place and but one train of waves will be sent out. Shorten- ing the gap will increase the number of discharges per alternation. The exact number for any spark-gap length will depend on the time of an alternation i. e., the frequency, and on the length of time it takes the available power to charge the condenser to the voltage required to break down the gap. Less energy per wave train will be radiated on a short gap than on a long one, because the work done varies as the square of the voltage (see art. 96) ; but the total work done may be equal, on account of the greater number of discharges. MANUAL OF WIRELESS TELEGRAPHY. 81 If the spark gap is too short, an arc is formed and no oscillations take place except those due to the frequency of the charging current. Professor Pierce has shown that the interval between wave trains may vary on account of the residual charge left in the condenser. When the spark gap's original resistance is restored, the potential of the residual charge may be opposed to the potential of the transformer and delay the charging. He has shown also that the gap sometimes partly retains its conducting character and breaks down at a lower potential than its length would indicate. This makes the sparks and oscillations irregular in strength and number and produces ragged and poor signals. In certain cases Professor Pierce notes an increase of received energy of 400 per cent when using a Cooper-Hewitt mercury interrupter in place of an ordinary spark gap. 122. With a given power the work that can be done per second is fixed. In charging a condenser W= . The number of times this is done /& per second gives the work per second, or the power expended. By increasing the frequency we can for a given power either reduce the voltage (length of gap) or the capacity of the condenser. For instance, ai a frequency of 500 cycles, for the same power, the condenser need only be 1/10 the size as for a frequency of 50 cycles. Or, keeping the capacity the same, the voltage can be reduced to I/ V 10= approximately 1/3 of that necessary for the same power at 50 cycles. A table showing the capacities necessary for given powers at different frequencies and voltages is given in table 2, appendix A. 123. When the spark gap is set to break down at the maximum charg- ing potential the condenser absorbs and stores all the energy that can be transferred by the charging transformer during the alternation. When it discharges it transfers part* of the energy to the open circuit to be radiated as electric waves. Since its period of discharge is very short as compared with that of the charging current the latter current does not appreciably change during the time the condenser is discharging. This current immediately begins to again charge the condenser, but the voltage of the latter does not rise high enough to cross the gap so that the con- denser soon begins to return energy to the charging circuit. It does this until its potential and the charging potential (and current if they are in phase) falls to zero. It then begins to absorb energy again with the reverse potential, and on reaching the maximum voltage again discharges across the' gap. Fig. 36 is an attempt to illustrate this action graphically. The area included by the curve on the left of the zig-zag line indicates the work * Experimentation has proven that from 80 to 90 per cent of the energy delivered to the transformer is transferred to the spark circuit. 82 MANUAL OF WIRELESS TELEGRAPHY. done on the condenser during the first half of an alternation; the zig- zag line indicates the number and amplitude of vibrations made by the closed circuit in transferring the energy to the radiating circuit. The area included by the curve on the right of the zig-zag line "repre- sents the work done during the second half of the alternation in recharg- ing the condenser. This work is all returned to the charging circuit. FIG. 36. MECHANICAL WORK DONE IN MAKING DOTS AND DASHES OF THE TELE- GRAPH CODE. 124. We are now in position to speak in more specific terms of the work done in sending wireless telegrams. Let us suppose that we are delivering 2 kilowatts at 60 cycles and 110 volts to a transformer, which delivers it to a condenser at a maxi- mum potential of 30,000 volts. Two kilowatts =2000 watts = 2000 joules per second = 1474 foot-pounds per second. Since 60 cycles = 120 alternations per second, the work equals approxi- mately 12.3 foot-pounds per alternation. If the work done on the condenser is in phase with the charging E. M. F., and if the spark gap is set to break down at a potential of 30,000 volts, the condenser will be discharged at the peak of the charging curve, or when one-half of the work that can be done in an alternation (12.3 foot-pounds) has been done on the condenser. The capacity of a condenser which takes 12.3 foot-pounds of work to charge it to 30,000 volts = .0372 microfarad, or approximately eighteen 0.002 microfarad jars. Suppose we are sending at the rate of 20 words per minute, that the words average 5 letters each, and that each letter is made up of 3 char- MANUAL OF WIRELESS TELEGRAPHY. 83 acters equal in length to 9 dots, then a minute can be represented as equal to 20x5x9 = 900 dots = 15 dots per second. In other words, the length of a dot is one-fifteenth of a second. Now we have 120 alter- nations per second, so that we have about 8 alternations per dot when sending at the rate of 20 words per minute; therefore a dot is made up of 8 distinct sets of discharges of the condenser and a dash of three times that number. The condenser is doing work in producing ether waves at the rate of 12.3 foot-pounds per alternation, equaling, approximately, 100 foot-pounds per dot and 300 foot-pounds per dash. It will be noted from the text that at this sending rate the frequency necessary to give 1 alternation per dot and 2 alternations per dash is only 7J cycles per second. It will be noted further that with one spark per alternation we cannot utilize 2 kilowatts continuously. We can only use it in charging the con- denser during the first half of each alternation. As soon as the discharge begins the condenser circuit oscillates in its own period as if entirely disconnected from the transformer. In this respect the charge and discharge of a condenser resembles the loading and firing of a gun. We must bear in mind, however, that though the charging may be done at any rate we desire, the discharge is very much more sudden than that of any gun. It is not necessary, therefore, except when considering methods of regulation, to devote attention to the charging of the condenser, and our minds can be concentrated on what happens during its discharge, when it forms part of an oscillating circuit. From the foregoing discussion we see that the real source of power in wireless telegraphy is the condenser, and that we can only use it inter- mittently, not more than one-fiftieth of the time, in fact, but that while working it works very energetically. DECREASE OF AMPLITUDE WITH DISTANCE FROM SOURCE. 125. From the discussion in art. 120 on the thickness of the hemi- spherical shell enclosing a train of ether waves, if we assume this thick- ness to remain constant and that part of the shell near the earth to be represented by an expanding cylinder, it is increasing in size by one dimension only, viz., circumference, and therefore the energy in any part of this shell will vary inversely as the distance, instead of inversely as the square or cube of the distance from the source, as would be the case if expansion were taking place in two or in three directions. Messrs. W. Duddell and J. W. Taylor in experiments made for the English Navy in 1905 proved that at least for distances up to 60 miles the received current as stated above varies inversely as the distance from the sending station, and the received energy varies inversely as the square 84 MANUAL OF WIRELESS TELEGRAPHY. of the distance. But additional experiments by Dr. L. W. Austin show that this law does not hold except for very short distances, and that the amplitude is lessened from other causes than those due to distance alone. We know that the energy is absorbed in the atmosphere more by day- light than by night more at high (summer) than at low (winter) tem- peratures. The amount of absorption as between one day and another probably depends also on the electric condition of the atmosphere. Long waves suffer less absorption than short ones. Irregular country produces large absorption. The absorption over some soils is for com- paratively long distances,, 30 times as great as over sea water. Trans- mission over salt water is the best. 126. As illustrating the difference in absorption between short and long waves, and (a) the greater efficiency of short waves for short distances, (b) the rapid falling off at distances above 100 miles, Dr. Austin finds : (a) Strength of received signals at 20 miles, using 300 meter waves, 5 times as great as with 1500 meter waves; at 100 miles, 300 meter waves, 4 times as great as 1500 meter waves; at 400 miles, 300 meter waves, 1.6 times as great as 1500 meter waves; at 800 miles, signals from 300 meter waves weaker than from 1500 meter waves. (b) Using 300 meter waves, strength of signals at 200 miles, 0.3 of that at 100 miles; at 400 miles, 0.053 of that at 100 miles; at 800 miles, 0.0036 of that at 100 miles. (See table 11, appendix A.) DETECTION OF ELECTRIC WAVES. 127. The direction of the magnetic lines of force at any point in a wave near the earth is parallel to the earth's surface and at right angles to a line joining the point with the source of radiation. The direction of the electro-static lines of force at any point near the earth is perpendicular to the earth's surface. An iron wire placed horizontally and parallel to the lines of magnetic force will be magnetized by a passing electric wave just as iron wires held in the magnetic meridian become magnetized ; pointed in the direction of the station the effect would be zero. It has been proposed to utilize this fact, both as a detector of electric waves and of their direction. Any conducting wire held perpendicular to the earth will be cut at right angles by the magnetic lines of force and will have electric charges induced in it which will create currents, and it is by means of the cur- rents induced in vertical conductors that electric waves are usually de- tected. A vertical wire thus situated also has a difference of potential created in its ends since it joins two points of the advancing wave whose electric potential differs. (This may also be the case in a horizontal wire if in the line joining its position with the source of radiation.) The total electric is equal to the total magnetic energy in an advancing wave. MANUAL OF WIRELESS TELEGRAPHY. 85 If two horizontal conducting plates forming a condenser are in the path of the wave, they will have electro-static charges of different poten- tials induced in them. This potential difference will vary with their vertical distance apart. If these plates are joined by a conductor, electric currents will be produced in it. We see, therefore, that there should be at least three ways of detecting electric waves: (a) By placing conductors at" right angles to the mag- netic field; (b) By placing conductors parallel to the electric field; (c) By adding to conductors at right angles to the magnetic field, conducting planes forming condensers at right angles to the electric field. It would seem that by the last method we should be able to abstract the greatest amount of energy from an electric wave and, therefore, be able to detect it at the greatest distance from its source. 128. It will readily be seen that the induction of currents in another aerial, however great the distance from the inducing aerial, is not greatly different from the inductive actions of the wires A B and C D on each other, which was discussed in the early part of this book. It was there pointed out that inductive actions caused by ether move- ments could have no limits, however small they might be at great dis- tances. In other words, every change of current sends out some non- returnable energy. Oscillating circuits of high frequency send out more non-returnable energy and radiate better than those of low frequency. Open oscillating circuits radiate faster than closed oscillating circuits. Chapter V. SENDING CIECUITS AND APPARATUS. 129. We shall now proceed to consider wireless telegraph sets in detail : GENERATORS. Induction coils (fig. 14b) with hammer breaks operated b} 7 direct cur- rent have been used to a very limited extent for naval purposes. The vibrations of the hammer were difficult to regulate and the large size necessary to handle large currents made the frequency too low for successful work. Hammer breaks were soon discarded and make and break regulated by some form of rotary motion. The most successful form was the mercury turbine interrupter. This interrupter was installed in the circuit containing the sending key and the primary winding of the induction coil. The interrupter consisted of a direct-current motor driving a centrifugal pump revolving in a chamber of mercury. The mercury was connected to one side of a break in one leg of the primary circuit. It was drawn up by the pump and delivered as a jet through a revolving nozzle. The mercury jet during a portion of each revolution struck a metallic segment connected to the other side of the break in the circuit, and, if the sending key was closed, thereby completed the circuit and built up a current in the induction coil which charged the sending condensers. When the jet passed the segments the circuit was broken. (The jet passed through grain alcohol which absorbed the spark at break.) This make and break occurred once in each revolution. The motor made approximately 1800 revolutions per minute. Assuming that the condenser was discharged only on the break, this gave but 30 dis- charges per second, or a note two octaves * lower, as compared with 120 discharges from a 60-cycle alternator. The operation of these sets was much improved by increasing the number of segments and, therefore, the number of makes and breaks per second, as many as six being used, thus giving a spark note slightly higher than that of a 60-cycle alternator. The spark in the interrupter at break always carbonized some alcohol and the latter also became mixed with mercury and formed a more or less * The octave of a note is that differing from it by 8 notes of the scale do-re-mi-fa-sol-la-si-do the octave above having twice as many vibrations per second and the octave below having one-half as many vibrations as the note referred to. Standard tuning forks vibrate 256 times per second. The pitch of a note is the number of vibrations per second producing that note. MANUAL OF WIRELESS TELEGRAPHY. 87 conducting carbon-mercury-alcohol emulsion, so that the interrupter and contents required frequent cleaning, washing, and filling. For small powers these sets, with care, gave good results, and being generally used with mechanical recording apparatus the spark note was not of marked importance. CONSIDERATIONS GOVERNING FREQUENCY OF GENERATORS. 130. Turbine interrupters were practically entirely replaced by 60- cycle alternating current generators operated by motors (on ships and at navy yards) or oil engines (isolated shore stations and light ships). These in turn are being replaced by 500-cycle sets operated by motors or engines as above. No special description of generators will be given. Sixty-cycle current was first selected because alternators of this fre- quency were commercial articles. When the use of telephones with receiv- ing sets became general it was realized that a sound of a higher note was desirable and that for the very best results the frequency (pitch) of this note should be that to which the telephone diaphragm or the operator's ear, or both, were most sensitive. A pure spark note is produced when the spark gap is so adjusted that the condenser discharges but once per alternation, thus sending out but one wave train per alternation. THE ADVANTAGES OF A HIGH SPARK FREQUENCY. 131. If two alternating currents of the same intensity but of different frequencies be sent through a telephone, it is found that the sound in the telephone produced by the current of higher frequency is much louder than that produced by the lower. This fact is due in part to the peculiarities of the human ear, which is more sensitive to high-pitched sounds than to low, also in part to the diaphragm of the telephone, which is usually of such a weight and size as to vibrate 'most readily to a sound of rather high pitch. This fact has an important bearing on wireless telegraphy, for the pitch of the sound produced in the telephone con- nected to the detector at the receiving station depends simply on the number of wave trains per second at the sending station. In order to determine exactly what is the relation between the strength of current required to produce an audible sound in the telephone and the frequency, a series of experiments was carried out on a pair of head telephones of the type ordinarily used in wireless telegraphy, the results of which are shown in the table. Frequency Volts to Frequency Volts to per produce per produce second. audible sound. second. audible sound. 60 6200 X 10-7 540 80 X 10-7 120 2900 " 660 30 180 1700 " 780 11 300 600 " 900 6 420 170 88 MANUAL OF WIRELESS TELEGRAPHY. In the first column are given the frequencies or the number of wave trains per second, and in the second the number of volts of alternating current which it would be necessary to apply to the terminals of the tele- phone to produce an audible sound. From this it is seen that it requires about a thousand times as much voltage at a frequency of 60 to produce a- sound as is required at a frequency of 900. We may assume, therefore, that if the number of wave trains at the sending station be increased from 60 to 900 per second, and the spark length be kept the same, the effect at the receiving station would be increased one thousand times. If the number of sparks be increased in this way without reducing the spark length, it is evident that the energy made use of at the sending station must be greatly increased. It will be more interesting, therefore, to calculate what the increase in sending efficiency of the station will be with increasing spark frequency if the total energy be kept constant. So if we assume that the energy is proportional to the number of wave trains, and divide the relative increase in loudness of sound in the telephone at the receiving station for any frequency by the relative increase in the number of wave trains per second, we will have -a fair comparison of the efficiencies at the two frequencies. Strength Strength Frequency. of Frequency. of signal. signal. 120 1 540 13 240 1.5 900 64 The results of such calculations are seen in the table, which shows that there would be very slight advantage in replacing a 60-cycle alternator giving 120 wave trains per second with a 120-cycle giving 240 wave trains, but that the advantage increases rapidly as the frequency is increased. The maximum sensitiveness of the telephone appears to lie in the neighborhood of 900. 132. In addition to the increase of sensitiveness of the telephone at high frequencies, there are other quite independent advantages in the use of a high-pitched spark. First, it is found in practice that a high- pitched musical signal is much more readily distinguished at the receiving station in the midst of ordinary interference and atmospheric disturb- ances; and second, at the sending station a shorter spark gap, which would generally be used with a high frequency spark, puts less strain on the insulation of the condensers and other parts of the circuit, and re- duces the losses due to brush discharges, which in many stations amount to a considerable share of the total amount of power employed. A third advantage is that with a high spark frequency larger amounts of energy can be radiated from a moderate-sized aerial without sub- jecting it to excessively high potentials. Experiments have been recently carried out in which it has been shown that in moderate frequencies with stationary spark gaps there are MANUAL OF WIRELESS TELEGRAPHY. 89 nearly always secondary discharges, irregular, but giving very high tones, so that the real advantage of the high spark frequency, from the stand- point of telephone sensitiveness, is usually less than that indicated in the table. The advantages of ease of reading, the lessening of the strain on the condensers and insulators, and the increase in effective energy capacity of the antenna,, especially when the latter is small, are very marked, so that it has been found possible to use small wireless sets of 2 K. W. capacity where formerly 5 to 10 K. W. were employed. The only difficulty involved in using very high spark frequencies lies in the cooling of the spark gap. For this purpose a rotary gap or some special refrigerating device must be used. 133. For the reasons stated above, 500 cycles has been adopted as a standard frequency for the present. An examination of table 2, appendix A,, will show how capacity for the same power decreases directly as the frequency increases, or keeping' power and capacity the same, how voltage can be decreased with increase of frequency. Condensers are more efficient when worked at a voltage below that which will give brush discharges. 134. To ensure a perfectly regular condenser discharge and thus obtain but one wave train per alternation, some generators have a disk mounted on an extension of the main shaft and revolving with it. This disk carries projecting electrodes, one for each pole of the alternator, equally spaced like the spokes of a wheel and connected to one side of the closed circuit. (See fig. 52.) In revolving they pass very close to a fixed elec- trode, or spark point, connected to the other side of the circuit, sparking taking place as the points pass one series of oscillations for each alter- nation. Generators carrying rotary spark points must of necessity be placed in or near the operating room and are to that extent objectionable on account of the noise of the spark, the additional space required, and the noise of revolution which interferes with receiving. Motor generators, or generators ' driven by engines, except as stated above, are usually arranged for being started or stopped from a distance. The controlling apparatus is mounted on a switchboard which carries voltmeters and ammeters, one each for the supply current and one each for the generator current. A frequency meter is also part of the switch- board equipment. This with the field rheostat of the motor enables the operator to adjust the speed of revolution so as to give the required frequency. TRANSFORMERS. 135. A generator designed for a certain frequency works best in con- nection with a transformer designed for the same frequency. If the size of the condenser to be used in the closed circuit is fixed, and known to the designer of the transformer, the latter can be built so that the 90 MANUAL OF WIRELESS TELEGRAPHY. secondary winding and condenser form a circuit whose natural period is that of the generator frequency ; a few such transformers have been sup- plied and are preferred to those requiring reactance regulators. Neither generator nor transformer will work without overheating at a frequency much greater than that for which they are designed on account of the increase of heating in the iron cores and frames, with increase of cycles of magnetization per second. An examination of fig. 29 will show that the generator armature wind- ing and the primary winding of the transformer form one circuit, and the secondary winding and condenser another. The reactances of these circuits should be such as to maintain the charging B. M. F. and current in phase with each other. When 60-cycle current was the standard, transformer windings were designed to give a potential of from 25,000 to 30,000 volts in the secondary when the primary was supplied with 110-volt current. With the reduction in voltage made practicable by the use of higher frequency, standard transformers now have a maximum voltage of 12,500 when supplied with 220-volt current. For small sets both induction coils and closed core transformers are satisfactory; for large sets closed core transformers are preferred. Transformers are fitted with safety spark gaps set at the maximum safe sparking potential. REGULATION OF A. C. SENDING APPARATUS. 136. Sending sets work most' efficiently when the interruptions or alternations of current are in resonance with the circuit formed by the secondary of the transformer and the sending condenser. When running on open circuit practically no work is being done by the motor or generator except that necessary to overcome friction. When the primary circuit is closed by the sending key, with the spark gap opened, so that no sparking takes place, the secondary of the trans- former charges the condenser during the first half of each alternation and receives current from the condenser during the second half of each alternation. The load thrown on the motor generator by pressing the key depends on the period in a cycle at which contact is made, but, generally speaking, it may be considered as instantaneous " full load." If -the spark gap is set so that the condenser potential breaks it down, the oscillations of the closed sending circuit practically cut out the secondary of the transformer, so that a condition of instantaneous " no load" exists as soon as the spark passes. As soon as these oscillations cease, the secondary again begins to charge the condenser, and a condi- tion of almost instantaneous full load is established. This interval is MANUAL OF WIRELESS TELEGRAPHY. 91 so short that the inertia of the moving parts of the motor generator pre- vents any change of speed or voltage, so that the instantaneous full load thrown on when the key is closed is the one affecting operation. Again, the inertia of the moving parts of the motor generator is often sufficient to keep up the voltage during the length of a dot, but not during the length of a dash. > When the key is closed the momentary current starting at that instant depends only on the reactance of the primary of the transformer and of the generator armature, since the resistance is very low. To control this sudden rush of current an adjustable choke coil, called a reactance regulator, may be placed in the primary circuit. This coil, on account of its inertia, acts as a buffer against sudden changes of cur- rent, and by means of its adjustability enables the phase relation of the E. M. F. and current in the circuit to be varied and thus the power expended to be controlled. Since the reactance regulator controls the power expended, it controls the secondary voltage and the maximum spark gap that can be used. By placing the sending key in shunt around it and having an inductive resistance in series with the key, the reactance regulator can be adjusted so that no sparking will take place, but by closing the key the current added through the shunt circuit is sufficient to cause sparking to take place. By means of this method the sudden changes from full to no load are avoided and the regulation improved, and since only a small portion of the total sending current is broken at the sending key, it is much easier to keep the contacts in good condition. A safety switch is placed in the primary lead when the method of con- trol described above is installed. This switch should only be closed when sending and should be opened at all other times when the motor genera- tor is running. A better method of control now being introduced is to have the send- ing key, by working auxiliary contacts, strengthen the fields of the motor and alternator by cutting out resistance just before the primary circuit is closed. The charge and discharge of the condenser when not sparking is indi- cated by a rustling sound, which signifies danger. This warning applies equally to induction coils and transformers, both terminals of which are dangerous when using alternating current. On account of the small penetrating effect of high-frequency currents (art. 103), it is believed that high voltages when associated with fre- quencies of above 100,000 per second are not dangerous to human life, but low frequency, high-voltage currents are very dangerous, and it must be borne in mind that a condenser being charged and discharged at the 7 92 MANUAL OF WIRELESS TELEGRAPHY. alternator frequency is very much more dangerous than when it is dis- charging across the spark gap. Other methods of generating high frequency currents are in use than by charging condensers through transformers and discharging them through spark gaps. But since other methods have not yet been practi- cally applied in the United States to any great extent, they will not be discussed in this chapter. SENDING KEYS. 137. The sending key, or the auxiliary key operated by it, is placed in one leg of the primary circuit. When placed directly in the primary circuit, sending keys, in some cases, have condensers shunted around them to absorb the spark at break. Their contacts when used to break the primary current direct, are larger than in the ordinary telegraph key on account of the larger cur- rents handled. In other respects they resemble the telegraph key. When used to operate a relay the ordinary telegraph key fills all requirements. The relay consists of a solenoid energized by the sending key, its arma- ture making and breaking the primary current in air or oil. Figs. 37, 38, 38a and 39 illustrate types of sending keys. The Slaby Arco keys shown in fig. 37 were of massive construction and very rugged. Fig. 38 shows a solenoid break, the connections for which are illustrated by fig. 38a. It will be noted that this is a positive break as well as make. Fig. 39 is practically the same as the ordinary telegraph key with large contacts. Sending keys should be adjusted to have just sufficient movement to prevent arcing and permit well defined making and breaking^ For direct breaking, though platinum contacts are largely used, com- paratively large brass or silver contacts are satisfactory. All contacts must be kept smooth and clean and their faces parallel. What is known as a " break key " is preferred. It was first used on the Stone sets, and is an ingenious and useful device for " listening in " while sending. An attachment to the sending key breaks the detector circuit just before the sending key makes contact. When the sending key is released the receiving circuit is automatically cut in, so that the receiver can " break " the sender by a call, which the latter can hear in the interval between his letters or words. For sending time signals, a Western Union relay closes a local battery having in circuit a solenoid, whose armature carries a lever which presses and releases the sending key in unison with the current impulses sent from the standard clock at the Naval Observatory. MANUAL OF WIRELESS TELEGRAPHY. 93 FIG. 37. Slaby Arco Key. FIG. 38. OF CONNECTIONS (TO TO TRANSFORMER PRIMARY TO l|0 VOLTS J).C. FIG. 38A. Solenoid Key. FIG. 39. Wireless Specialty Apparatus Co. 94 MANUAL OF WIRELESS TELEGRAPHY. CLOSED CIRCUITS (INDUCTANCE, CAPACITY, SPARK GAP) . 138. Sending circuits are illustrated by elementary diagrams in figs. 29a and 40 to 48 inclusive. The names under these figures are the names of the engineers proposing or designing sets with the connections shown. To render them capable of adjustment all wireless telegraph oscillat- ing circuits have either variable inductances or condensers or both. These condensers and inductances vary greatly in design. Those for sending circuits, especially on account of the high potentials to which they are subjected, are very different in construction and mounting from those used in receiving circuits. Fixed condensers and variable inductances are used in sending cir- cuits. The condensers may be single, two or more in series, or in parallel. Series parallel installations may be made also, just as in primary bat- teries. (See figs. 28c and 28d.) The variable inductance usually consists of a helix of comparatively large bare wire (round or flat) mounted on an insulating frame a foot or more in diameter, the turns of wire varying from about f" to 2" apart. (See fig. 73.) A large number of the sending circuits in use at present are direct con- nected, but inductively connected sets are equally efficient and have this advantage that the coupling can be varied by a movement of either the closed or the open circuit inductance as a whole without varying the wave length of either circuit. (Figs. 42, 43, and 45.) 139. In direct connected sets, three movable clips or sliders are usually provided, one for the closed and two for the open circuit (fig. 40). The closed circuit is permanently connected to one end of the helix and the circuit completed by means of the wire from the movable clip, which can be connected to any desired point. The open circuit has the ground and the aerial wire, respectively, attached to the other two clips and these are attached to such points of the helix as will give the open circuit the same natural period as the closed circuit and at the same time give the two circuits the number of turns in common necessary for the desired coupling. 140. In inductively connected sets, the closed circuit helix is the same as before, the open circuit helix is permanently attached to the ground lead and the aerial lead attached to whatever point is necessary. The mutual induction and coupling are varied by moving the open circuit helix as a whole. (Figs. 42 and 43.) Making the adjustments to particular wave lengths and couplings is called tuning and is discussed in Chapter VII. 141. In the latest Telefunken sets (fig. 46) the sending inductance in both closed and open circuits consists of flat spirally wound coils, mounted MANUAL OF WIRELESS TELEGRAPHY. 95 SHOEMAKER FIG. 40 FES3ENPEN FIG. 45 TO AERIAL S FESSETNDEN FIG. 41 TO AERIAL * S J STONE: FIG. 42 TO AER/AL II id -= n '" ^ (A o y MARCONI STONE: FIG. 43 INSULATOR LOWENSTEIN FIG. 47 TO AERIAL J iO T 3 PIERCE SLABY ARCO FIG 48 MASS1E FIG. 44 NOTE. Figures 40, 41, 42, 44, and 48 represent circuits that are not in general use to-day, and hence to the student have a historical rather than a practical value. 96 MANUAL OF WIRELESS TELEGRAPHY. parallel and close to each other in a frame. Alternate coils being con- nected to a lever by which their position relative to the others can be varied. The. coils are connected so that currents in adjacent coils oppose each other and decrease the self induction of the whole, called by the manufacturers a variometer. By means of the lever the coils can be separated and the self-induction and consequently the period of the cir- cuit regulated. This is illustrated in fig. 46 by an arrow drawn diagonally across the inductances in which it is used. Fig. 74 shows the apparatus as manufactured. 142. For older direct connected sets connections shown in figs. 40 and 44 are preferred. The S symbol indicates alternating current. In the figures referred to, the condenser is directly across the secondary terminals of the transformer, and the spark gap in one leg of the closed oscillating circuit, as contrasted with the spark gap being placed directly across the transformer terminals. (Fig. 48.) The former is considered to be a more symmetrical arrangement. 143. Attention is invited to fig. 41, which shows one leg of the trans- former directly grounded and the other leg connected direct to the aerial. All other methods of connection afford direct path to ground and path through condenser and spark gap. This method of installation affords path to ground through condenser or spark gap only and affects tuning. If the aerial is touched when current is on the transformer, the latter, having one leg grounded, is short-circuited through the body and a severe shock may be experienced. Though this method of connection is no longer used, it is referred to here to show the necessity of giving careful consideration to the relative positions of ground, spark gap, and condenser. Errors in connections are sometimes made so that the most direct path to ground is through the spark gap. This induces potentials at the gap or condenser approximately equal to those at the upper end of the -aerial and produces disagreeable inductive effects in the operating room. 144. Fig. 43 shows the preferred form of inductive connection or coup- ling, that is, one inductance above the other. This takes up less floor space and the coupling is varied by vertical instead of horizontal move- ment, as is necessary when the coils are side by side as illustrated in fig. 42. Fig. 47 indicates a method of connecting up sending sets so that the operator by moving a hand wheel or lever can change the wave length of the open and closed circuits the same amount without changing the coupling. This apparatus is just being introduced and should greatly facilitate the operator's control over his sending wave length. MANUAL OF WIRELESS TELEGRAPHY. 97 CAPACITY. 145. The condenser capacity necessary to absorb 1 K. W. at 60 cycles and a maximum potential of 30>000 volts (A" gap) is .0186 M. F. or approximately nine standard jars. (Table 2, appendix A.) At 500 cycles and 30,000 volts it is .00223 M. F., a little more than one jar. At 500 cycles and 12,500 volts (.015 gap) it is .0127 M. F., approxi- mately six jars. Older 2 K. W. sets had capacities given below and required maximum volts as indicated to absorb 2 K. W. with two sparks per half cycle. Slaby Arco, 60 cycle, capacity .014 M. F., max. volt, app. 35,000, gap .5 in. app. Fessenden, .004 65,000, 1.6 in. DeForest, . " " .02 " " 30,000, " .5 in. " Shoemaker, j Stone, * .0105 " " 40,000, " .6 in. Sets now supplied have capacities based on 500 cycles at 12,500 volts, condenser racks or tanks being arranged to hold a number of jars some- what greater than that necessary for the rated output. 146. For 2 K. W. sets standard coppered jars in air or oil are pre- ferred. Tinfoil covered jars are no longer supplied. Inside connections to Leyden jars are best made by soldering one end of a strip of copper or brass gauze to the inner copper coating and clamp- ing the other end to the charging bus bar. Outside connections are made either by supporting all jars on a con- ducting plate connected to the other charging bus bar or connecting this bar to a strap of sheet brass or copper clamped around the jar. The important point about condenser connections is that they should make a good electrical contact of comparatively large area, with the charging wire or bus bars and with the condenser jars or plates. A symmetrical arrangement of material giving as nearly as possible equal lengths of discharge paths should be made. Many kinds of springs and clips for condenser connections have been devised and are in use, but none are better than those just described. Less difficulty is experienced with connections on copper coated jars or plates than was the case when tinfoil was used exclusively for distributing the charge over the glass dielectric. 147. The condensers now in use are standard Leyden jars in air or oil. (Fig. 49.) Glass plates in air or oil (glass dielectric). Metal plates in compressed air (air dielectric) (fig. 50) and tinfoil (paper dielectric). For large powers, glass plates in oil or metal plates in compressed air are preferred. For small sets the most convenient for use installation and 98 MANUAL OF WIRELESS TELEGRAPHY. FIG. 49. Ley den Jar Battery. FIG. 49A. Moscicki Tube. FIG. 50. Compressed Air Condenser. MANUAL OF WIRELESS TELEGRAPHY. 99 inspection are the standard jars, in air or oil, or glass plates set ver- tically in oil. Fig. 49a is a special form of Leyden jar which is con- venient for some purposes. The low voltages associated with 500 cycle sending sets have made practicable the use of paper condensers as noted above for small powers. CONDENSERS AND CONDENSER MATERIAL. 148. When a piece of iron is magnetized and demagnetized i. e., goes through a cycle of magnetization a certain amount of energy is ex- pended, which appears in the shape of heat in the iron. It is supposed to be due to internal friction in the molecules of the iron and is called magnetic hysteresis. In the same way, to put a condenser through a cycle of charge and discharge requires the expenditure of a certain amount of energy, which appears as heat in the dielectric and is called dielectric hysteresis. The loss of energy due to this quality varies in different dielectrics and is a function of the frequency. In choosing condensers for the closed sending circuit, it is of great importance to find those which will absorb a minimum of energy and at the same time show no tendency to break down under the large differ- ences of potential impressed upon them. The losses of energy in condensers are of two kinds: internal losses produced by dielectric hysteresis, and external losses produced by the brush discharges at the edges of the conducting surfaces. The ideal dielectric in respect to 'the internal losses is air, as it is entirely free from internal energy absorption. When used at ordinary pressures, however, it is unable to bear any considerable difference of potential. It has been discovered that when the air pressure is increased to the neighborhood of 250 pounds, the dielectric strength becomes so great that it is suitable for use at any of the potentials ordinarily used in wireless telegraphy. Compressed air condensers are ordinarily made up in the form of a series of plates so connected that the alternate plates may be charged positively and negatively, and the whole set is enclosed in an air-tight steel tank which can be pumped up to the desired pressure. Such a condenser, while ideal in its electrical properties, is somewhat bulky, and difficulties are sometimes found in preventing leakage of the air. It is therefore common in stations where the last degree of efficiency is not demanded to make use of glass condensers, either in the form of flat plates or jars. The conducting surfaces of condensers are now generally formed of elec- trolytically deposited copper. It is generally stated that flint glass. is the glass best suited to form the dielectric. Experiments which have been made show that the internal losses of glass condensers in ordinary use amount to from 2 to 8 per cent of the total energy flowing through them. 100 MANUAL OF WIRELESS TELEGRAPHY. The losses due to the brush discharges from the edges of the conduct- ing surfaces, which sometimes amount to 30 per cent of the total energy, may be much reduced by immersing the condensers in oil or by placing several condensers in series, which reduces the individual potential differ- ence on each condenser, or by covering edges of foil (or copper) and plates with an insulating compound. 149. Practically all other insulators have a greater specific inductive capacity than air at ordinary pressure, and nearly all of them have a greater dielectric strength than air (art. 150). The Ley den jar, having long been used as a high-potential condenser, its method of manufacture being well known, and the best glass having not less than nine times the capacity of air, has been very generally used in wireless telegraph send- ing circuits. Air and oil, while requiring much larger volume to give the same capacity as glass, have the excellent property of mending themselves after puncture by a spark, while all kinds of solid or semisolid dielectrics require renewal after rupture. Mica has very great dielectric strength, as much as 5000 volts per mil, and has been used to some extent in condensers in the form of micanite. The semisolid dielectrics, such as beeswax and paraffin, have to be made up with considerable attention to the temperatures in which they are to be used, since they may melt in summer and crack in winter, but they are cheap and easily obtained. Dielectric strength of insulators per millimeter increases with decrease of thickness, except in oils, where it seems to decrease. Dielectric strength of air increases with increase of pressure. Dielectric strength of air decreases with decrease of pressure until the pressure is in the neighborhood of 1 millimeter of mercury^ when it increases. Dielectric strength of a vacuum should be infinitely great. Fleming states that with the best flint glass it is possible to store about 45 foot-pounds of energy per cubic foot of glass. The limit is set by the dielectric strength of glass. He has shown that the lengths of discharge paths of all condenser elements should be equal. Capacity varies inversely and dielectric strength directly, as the thick- ness of the dielectric, but they do not vary in the same ratio. The dielectric strength of glass condensers decreases, that of oil con- densers increases, with the frequency. 150. Tables showing the specific inductive capacity of a number of dielectrics and their dielectric strengths are given below. This data is incomplete. Data relative to the hysteresis losses of various dielectrics is almost lacking, and want of agreement is noted among different author- ities. MANUAL OF WIRELESS TELEGRAPHY. 101 Specific Material. specific inductive capacity. Dielectric strength. Volt8. ( 4,500 Air | 3,000 Hard rubber 2.29 3 40,000 India rubber 2.10 30,000 Mica 6.64 3 60,000 Micanite 8 40,000 Typewriter linen paper ... 3 45,000 Paraffin oil 2.71 8 7,000 Glass (crown) 6.r Glass (plate) 8.45 Glass ( light flint) 6.72 K ' * 20,000 Glass (extra dense flint) 9.86 1 Per millimeter for thicknesses up to 1 millimeter. 3 Per millimeter. 2 Per centimeter. * 4 Approximate. 102 MANUAL OF WIRELESS TELEGRAPHY. FIG. 51. MANUAL OF WIRELESS TELEGRAPHY. 103 DIELECTRIC STRENGTH OF AIR. 151. The dielectric strength of air is considered to be about 4500 volts per millimeter for gaps of about 1 millimeter in length, and about 3000 volts per millimeter for gaps of the length of a centimeter or more. Fig. 51 shows sparking distances in air between needle points, as deter- mined by experiment. These distances are usually greater than those obtained from equal voltage between the blunt spark points used in wire- less telegraphy. The latter probably correspond more closely to table 1, appendix A. On the other hand, this table of spark distances was determined by raising the voltage very gradually and exactly alike for each gap, while in oscillating circuits there is a convulsive rush which may produce very high potentials. This has been shown by introducing a minute spark gap elsewhere in the circuit, the effect being to greatly increase the gap, which can be ruptured by a given transformer potential. The inertia of the charge carries it forward, and just as the inertia of water in a pipe produces a great pressure if its flow is suddenly checked, so the potentials in the sending circuits may, and usually do, rise much higher than is indicated by the transformer ratio. 104 MANUAL OF WIRELESS TELEGRAPHY. * SPARK GAPS. 152. A great deal of thought and ingenuity has been expended on improving the action of spark gaps. For instance, the use of magnetic blowouts, induced and forced air drafts across the gap ; dividing them into a series of short gaps; placing gaps in parallel; enclosing them in compressed air and in nitrogen gas; making the points hollow and cool- ing them with air or water. Until recently, no method of construction for small powers was mark- edly better than the ordinary gap in air between two zinc rods, J to J inch in diameter, and these are still largely used. There, are two points in common for all good working gaps (a) The sparking surfaces must be clean and fairly smooth; (b) They must be kept from heating. The increased radiation from cooled spark electrodes as compared with heated ones is very evident. Heated surfaces give off more metallic vapor and tend to the forma- tion of a low frequency arc. There is no doubt that much of the irregularity noted in sending is due to an improperly adjusted spark gap and the effect known as " soaring " or "swinging" is probably due to the inequalities in the action of the spark gap and condensers. An open spark must be kept white and crackling and have considerable volume. If too long, it will be stringy ; if too short, an arc will be formed. All spark gaps are adjustable either in length or in number. All should be well muffled for obvious reasons. The types of spark gaps now in use are shown in figs. 52-61. The only types now supplied are fig. 52, the synchronous rotating gap, and fig. 57, quenched gap. 153. The function of the spark gap in an oscillatory circuit is'to allow the condenser to charge to the required potential, and then to break down and permit the charge to surge back and forth until its energy is dis- sipated. The ideal spark gap would be one which would insulate per- fectly while the condenser was charging and conduct perfectly while it was discharging, and the nearer these conditions can be fulfilled the more efficiently will the spark gap perform its duty. Either condition can be fulfilled alone, but the combination is somewhat difficult to obtain. The resistance of the spark gap when the discharge is passing depends upon two factors ; it increases rapidly with the spark length, and de- creases rapidly with the oscillatory current, amounting with a half-inch gap to several hundred ohms when a fraction of an ampere passes, and a small fraction of an ohm when 50 or 60 amperes are flowing. With the spark length above half an inch, the resistance with the same oscillatory current flowing may be taken as roughly proportional to the spark length. But in a condenser circuit the amount of electricity stored up in MANUAL OP WIRELESS TELEGRAPHY. 105 SPARK GAPS SHAFT or ALTERNATOR SYNCHRONOUS ROTATING SPARK GAP NON-SYNCHRONOUS ROTATING; SPARK FIG 52 FIG. 53 FIG 54 FIG. 55 X AIR BLAST QAP PARALLEL QAP uc: 1 . Elf 1 Vlf nx \ n\ m Jt/O m J r~i QUENCHED SPARK C3AP FIG. 57 FIG. 58 TELEFUNKEN (Now Obsolete) MARCONI DISC DISCHARGER FIG. 56 FIG. 59 STONE. (Now Obsolete) MASSIE QAP FOR COMPRESSEP AlR FESSENDEN-J^ FOREST FIG. 60 (Now Obsolete) FIG. 61 (Now Obsolete) 106 MANUAL OF WIRELESS TELEGRAPHY. the condenser, and hence the amount of oscillatory current, increases with the spark length. Thus we have two conditions working against each other as regards the influence of the spark length on the spark resistance; but we can increase the amount of current flowing without increasing the spark length by increasing the size of the condenser, and the most efficient form of circuit for a given power is that in which a moderate spark length, and large condensers are used. When, after the condenser is charged, the spark gap breaks down, the gap becomes filled with metallic vapor and for the time being forms a high-frequency alternating current arc. It is the presence of the metallic vapor which produces the conductivity of the spark. After the discharge ceases, however, if this metallic vapor is not removed from the gap, the insulation will evidently be poor at the time that the condenser is next being charged, hence the first condition of spark efficiency would be want- ing. It is therefore necessary to remove this vapor completely as soon as possible after the surgings of the condenser charge cease. This is done partly by cooling the electrodes of the spark gap, thus stopping the vaporization, and in some cases by blowing the vapors out of the gap. 154. In the simple gap, such as is found in sets of small power, the vapor is usuall} 7 sufficiently dissipated by the natural cooling of the electrodes and by ordinary air currents. Such a gap, however, not pro- vided with an air blast should not be enclosed. For somewhat larger powers, an air blast is ordinarily considered necessary. This carries 'away the metallic vapors and at the same time cools the electrodes. Such an arrangement is shown in fig. 54. Another form of gap for small powers which gives good satisfaction is the parallel gap (see fig. 55), in which two cylinders of zinc or brass are placed parallel to each other, and the spark runs from point- to point, never jumping twice consecutively in the same place. This wandering of the spark is facilitated by a slight roughening of the electrodes with a file. The explanation of this phenomenon of the running spark is probably as follows: The spark jumps from a slight projection on the electrode which in the course of the oscillations is burned away, so that at the next discharge an easier path is found from some other projecting point. 155. For high powers a good form of spark gap is the rotating syn- chronous gap shown in fig. 52. This consists of one or more stationary members and a rotating member made up like a wheel with projecting spokes. This in its best form is attached directly to the shaft of the alternator, and is so adjusted that a spoke comes opposite a stationary member at the exact moment that the maximum of potential is obtained in the condenser. This insures one discharge for each alternation of the current, the complete absence of conducting vapors, and gives a satisfac- tory insulation for each spark. The regularity of discharge from this MANUAL OF WIRELESS TELEGRAPHY. 107 form of gap produces a pure musical note, which is of great importance in the telephonic reception of signals. (See art, 132.) 156. Another form of rotating gap, called the non-synchronous rotat- ing gap, is shown in fig. 55. In this the wheel is rotated rapidly by an independent motor without regard to synchronism with the alternator. The face of the stationary member of the gap forms an arc of a circle long enough to a little more than cover the distance between two spokes, thus always insuring the proper sparking distance. The rotating wheel itself forms an efficient fan. 157. What is called a " quenched gap " * (shown in fig. 57) is made up of a number of copper discs accurately turned and separated by annular rings of mica about .01 inch thick. The spark is confined to the air tight space inside the mica-rings. This type of gap, if a proper number of discs are in series, also gives one discharge for each alternation of the current and produces the same pure musical note as the synchronous gap. It is almost noiseless and has the further advantage of (probably on account of its large cooling surface) quickly stopping the oscillations of the closed circuit, so that the open circuit is left free to vibrate in its own period, and it therefore radiates waves of but one length. This fact has an important bearing on the tuning of wireless telegraph sets and also on the coupling, which can without change of wave length be made that which will transfer energy from the closed circuit to the open circuit with the least loss. 158. The quenched gap can not be depended upon to operate without artificial cooling of the discs when any but very small powers are used. Like all other gaps, its action is improved by an air blast. In the case of the rotating gap the equivalent air velocity in a case of large power was about 20,000 feet per minute. Mr. J. Martin finds a very distinct gain in radiation from an air cooled gap with air pressures up to 15 pounds per square inch, which corresponds to a velocity of 82,000 feet per minute, or about 1400 feet per second. Take a single gap operating on a 1000-meter wave on the peak of the charging E. M. F. : If the coupling between the open and closed circuits is such that the closed circuit transfers all its energy to the open circuit in five complete vibrations the first group of sparks will last 1/30,000 second. To remove the conducting vapor from the gap in that time would require a minimum air velocity across the gap of 1000 feet per second if the electrodes were A inch (1 cm.) in diameter. From this point of view it would seem, therefore, that any gap will act as a quenched gap if the * Discovered by N. Wien, in the course of an investigation on electrical discharges between metal surfaces placed very close to each other, and pub- lished by him in October, 1906. 8 108 MANUAL OF WIRELESS TELEGRAPHY. air velocity across the gap is sufficiently great, and that the required air velocity varies directly as the diameter of the spark electrodes inversely as the wave length, directly as the damping and (since it is known that close coupling increases the damping) directly as the percentage of coupling. Loose coupled circuits would require a lower air velocity than close coupled ones. Fig. 56 illustrates the Marconi disc discharger, which is practically the same in principle as fig. 53 the non-synchronous rotating gap. A special motor is required to operate the discharger. It has also the dis- advantage of being as noisy as the synchronous gap. The disc discharger, like the synchronous rotating gap, is suitable for large powers. It is fitted with an auxiliary stationary gap for use in case of motor break- down. TRANSFER OF ENERGY BETWEEN COUPLED CIRCUITS. 159. The transfer of energy between coupled circuits having the same natural period is well illustrated by the mutual action of two similar pendulums connected by a flexible support. If, one being at rest, the other is pulled aside and released, the swinging pendulum gives properly timed impulses to the other through the flexible connection and starts it to swinging also, gradually decreasing its own swings while the other increases, until the first one stops ; at which time the second has reached an amplitude nearly as great as that of the first swing of the one pulled aside. In other words, all of the energy has been transferred to the second pendulum. The first one then starts again and increases its swings while the second gradually slows down and comes to rest at which time the first is again at its maximum. All the energy has been returned by the second pendulum to the first. The swings are slowly damped by air friction until the system comes to rest. If the periods of the two pendulums are not equal, or nearly so, the impulses are out of step (resonance) and no transfer of energy takes place the pendulum first started keeps on swinging and the second remains at rest. 160. If the points of support by the flexible connection are a foot or more apart (loose coupling) the second pendulum picks up the swing rather slowly and both pendulums make a large number of vibrations before the second has received all the energy from the first and the latter has come to rest. If the t points of support are close together (close coupling) the second pendulum reaches its maximum and the first comes to rest in a few vibra- tions, the transfer of energy is more rapid, and the damping greater. The ball of energy, so to speak, is tossed back and forth between them more rapidly than when they are farther apart more loosely coupled. MANUAL OF WIRELESS TELEGRAPHY. 109 Professor Pierce * has photographed the sparks in a short gap in the open circuit when oscillating in connection with the closed circuit and shows that they occur in groups. This particular circuit showed groups of four. In other words, four vibrations sufficed to transfer all the energy from one circuit to the other. Two circuits may be alike in period, but have different dampings. 161. Eeturning now to the consideration of the quenched gap and the closed circuit, what we call the closed circuit is only closed when the spark gap is conducting and its period in that condition is the one measured either when we take the time interval between sparks or deter- mine it by a wave meter. It has a different period when the spark gap is not conducting because its capacity with reference to being charged from the open circuit is less and it is therefore out of tune with the open circuit and the latter does not transfer any energy to it. The effect of the method of construction of the quenched gap seems to be to restore the nonconducting character of the gap the first time the closed circuit comes to rest, and thus leave the open circuit free to radiate. It would be inter- esting to take photographs in both circuits to determine whether this really is the case. 162. Eeferring to art. 109 on mutual induction: The open circuit is first set to oscillating in either the period longer or shorter' than its natural period and has reached its maximum when the closed circuit has stopped and opened. Thereafter the open circuit is free to vibrate in its own period, and that it changes to that period is shown by the wave meter readings, but in building up it is sending out waves of a different period. The first maximum reached in the open circuit is the highest maximum and, since no further loss by re transfer to the closed circuit takes place, the quenched gap is consequently the most efficient. It will also con- duce to efficiency to make the building up period of the aerial (when it is radiating waves of a different length) as short as possible. In other words, close coupling, but close coupling increases the induced E. M. F. in the condenser circuit. Therefore, there is a possibility with very close coupling of retransfer of energy by breaking down the gap and again closing that circuit. 163. We can, therefore, conceive of a wave train from an ordinary open circuit as made up of a series of waves whose amplitude rises and falls during the transfer and retransfer of energy from one circuit to the other; the damping depending on the coupling and being partly natural (due to heating and radiated energy in the shape of electric waves), partly artificial (due to retransfer of energy to the closed circuit). * G. W. Pierce, Principles of Wireless Telegraphy, 1910, p. 248. 110 MANUAL OF WIRELESS TELEGRAPHY. A wave train from the open circuit of a quenched gap can be repre- sented, as in fig. 62, by a building up at a certain frequency (depending on the coupling) to a maximum depending on the radiation or other losses per oscillation, and then oscillations in the natural period of the open circuit, with damping dependent on the radiation and resistance of the open circuit only. The closed circuit starting at a maximum and transferring all the energy to the open circuit in a few' oscillations as shown in the upper part of fig. 62 ; there being no retransfer of energy from the open to the closed circuit and vice versa as occurs with the pendulums discussed in arts. 159 and 160. Closed . Open SPARK FIG. 62. 164. The results of experiments indicate that the decrement of the open circuit which will give good tuning must be not greater than .2 per oscillation, and this is represented in fig. 62. This rate of decrease of amplitude gives about fifteen complete vibrations before the amplitude falls to one-tenth of the maximum. The above gives as the length of a 425-meter wave train from a 500- cycle generator about 3.3 sea miles, its duration 1/47,000 second, the distance between wave trains 186 miles, the time interval 1/1000 second, so that even with a frequency of 500 cycles we only generate electric waves about 2$ of the time. MANUAL OF WIRELESS TELEGRAPHY. Ill LIMITATIONS ON WAVE LENGTHS. 165. A certain amount of inductance is necessary in the closed circuit in order to transfer energy to the open circuit, whether the circuits are direct or inductively coupled. Since condensers of any desired capacity can readily be obtained, it is easy to make the closed circuit any electrical length we desire. There is, however, a lower limit to this, depending on the material and arrangement of the condenser and leads. Other things being equal, the larger the capacity, the longer the connecting leads; and the shortest wave length that can be obtained for a given capacity is that found when the leads from the condenser are connected in the most direct manner to those from the closed circuit and spark gap. The standard wave length for ships and shore stations was first set at 320 meters. It is now 600-1000 meters for ships. It will be noted that the increase in frequency to 500 cycles will, though the standard voltage has been lowered, permit a decrease of capacity and thus permit the radiation of larger powers on shorter wave lengths than is now practicable. Experience shows that aerials with short wave lengths radiate more efficiently than those with long ones, and that up to several hundred miles short waves travel over salt water with no great absorption; when transmission over land is necessary and for long distances over water we gain more by the reduced absorption of long waves than we lose by decreased radiation efficiency. 166. The open circuit, while it has concentrated inductance like the closed circuit, has distributed capacity which is comparatively small, and though any electrical length we desire can be obtained by adding induc- tance, it is found that concentrated inductance beyond that necessary to receive energy from the closed circuit lessens the radiation, and on that account it is necessary to increase the period of the open circuit by adding capacity in the shape of additional wires to the aerial. We have seen that, unless they are quite a distance apart, two parallel wires do not have twice the capacity of one, so that it is practically difficult to get very long wave lengths in the open circuit, especially on shipboard. The wave lengths that we can efficiently use in the open circuit are, therefore, limited by practical considerations. Since the energy in any discharge varies as the square of the voltage, and since any desired voltage can readily be obtained, the work that can be stored in a condenser of given capacity depends only on the dielectric strength of the condenser material. But in the case of the open circuit, when the first transfer of energy is completed, unless it is radiated nearly as fast as received, the maximum 112 MANUAL OF WIRELESS TELEGRAPHY. voltage in the open circuit, on account of its capacity being very much smaller, is much greater than that in the closed circuit. And we find that very high voltages, on account of difficulty of insulation, break out in sparks at all points of the circuit, that the aerial wire glows through- out its length, and the whole apparatus generally acts like a dry linen fire hose when subjected to a high water pressure i. e., it spurts electricity at all points in all directions. So practical considerations limit the wave lengths that can be efficiently used on board ship, and also limit the power that can be used with them. 167. Eef erring to the closed circuit, it is probable that the best results with any given sender are obtained when the work necessary to charge the condenser to the transformer voltage is equal to that supplied by the available power of one-half alternation. This gives but one wave train per alternation, and, if true, fixes at once the capacity of the closed sending circuits for any given power. Good results, how- ever, have been obtained by producing a condition of resonance in the secondary circuit with the primary frequency and obtaining a wave train only every two or more alternations. However, if the generator, sending circuits and aerial are properly designed, the greater power per wave train will be gained at the expense of efficiency on account of the high voltages in the aerial and the reduced spark frequency. OPEN CIRCUIT (AERIAL, INDUCTANCE, GROUND). 168. Aerials, with which the open circuit inductances of sending sets are connected, are shown diagrammatically in figs. 63 to 71 inclusive. The main principles to be remembered in connection with aerials (or antennas, as they are sometimes called) are that the higher the aerial the more efficiently the energy will be radiated in the form of electric waves and the larger the currents induced in the vertical part of the aerial the greater the amount of energy radiated. So what we need is height for efficiency; and capacity for amount (dis- tance). The former is limited by the height of mast; capacity by the amount of wire that we can conveniently support at the mast heads. Experiment indicates that distance for the same power and received on the same antenna varies as the height up to 100-200 miles. The total capacity of a ship aerial is usually 'less than one standard jar. To hold the same amount of energy as the condenser circuit the aerial is, therefore, while oscillating, charged to a higher maximum potential. The form of aerial now generally used on ships and ashore is called the flat-top or inverted L (fig. 67). The leads to the operating room are taken from one end; the other (free) end is subject to high potentials and must be well insulated. MANUAL OF WIRELESS TELEGRAPHY. 113 FIG. 63 BELLINI-TOSI FIG. 64 MARCONI GUYS (SUY9 MAST FIG. 65 UMBRELLA STONE: MASSIEL f TO SENDlNq FIG. 67 INSULATED WIRE * BARE W)f?ET FIG. 66 (Now Obsolete) TO SENPING* HELIX. FIG. 68 (Now Obsolete) TO SEMpINQ HELIX FIG. 69 (Now Obsolete) HELIX 4 TO SENDING HELIX FIG. 71 114 MANUAL OF WIRELESS TELEGRAPHY. Some T aerials are in use (fig. 70). They give greater relative capacity for the same amount of wire, but as we shall' see presently, the natural period of the flat top is not too great. T aerials sag in the center, thus decreasing their effective height and they are subject to high potentials at both ends. The other types shown are, or have been, used on shore stations, except the special receiving aerial shown in fig. 63, which is a direction aerial used on ships and which will be referred to in Chapter VIII. The umbrella aerial shown in fig. 65 has been used at some large shore stations. It is probably the best form that can be supported by a single mast. FIG. 72. Wireless Telegraph Anchor Spark Gap. LOOPED AERIALS. 169. It will be noted that the diagrams of receiving sets (figs. 82 and 83) show an aerial in the form of a loop, beyond three spark points arranged in the form of a triangle. The lower one of these points is con- nected to the sending circuit inductance so that as far as sending is concerned this aerial is the same as any other, since the high potentials used in sending easily jump the short gaps between the two sides of the loop; but for receiving it is different the weak currents can not jump the gap, which is known as an anchor spark gap, so that the circuit is only looped for receiving and not sending. MANUAL OF WIRELESS TELEGRAPHY. 115 The anchor spark gap (fig. 72) serves to cut out the sending circuit when receiving. When sending, the volume and color of the sparks in the anchor gap serve to indicate roughly whether the sending apparatus is working properly. For receiving sets not requiring a looped circuit the two sides of the loop are joined below the gap and used as, a single wire. A little consideration will show that the wave length of a loop is the same as that of half the loop on open circuit. A loop is, however, a persistent oscillator. 170. Except where they pass near conducting objects or through decks, all parts of the aerial wire are left bare on account of the lighter weight and smaller surfaces exposed to the wind as compared with insulated wire. The size of wire generally used is made up of seven strands of No. 20 B. & S. phosphor or silicon bronze wire or monnot metal having fairly high elastic strength. Stranded wire is more flexible, and the materials given above have fairly good conductivity and much greater elasticity than copper wire. The elasticity prevents permanent elongation and sagging after being hauled taut. 171. The natural wave lengths of certain aerials of the flat top type (inverted L and T aerials, figs. 67 and 70) are given in tabular form below. To the aerial is added the necessary turns on the open circuit helix to bring the natural wave length to 425 meters. It usually requires a number of turns of the helix to do this. When it is desired to greatly increase the sending wave length special loading coils are added to the open circuit. (See figs. 45, 46 and 47.) Since the closed circuit has large capacity and small self-induction a turn or more of inductance added to the closed circuit makes a large percentage addition to its self-induction and, therefore, to its wave length. But the open circuit has small capacity and relatively large self-induction, so that each additional turn does not make such a large percentage addition to its self-induction and, therefore, its increase of wave length per additional turn is much less than that of the closed circuit. (See Adjustments, Chapter VII.) 116 MANUAL OF WIRELESS TELEGRAPHY. Ship. Type. No. wires. Distance apart. Length of flat top. Vertical length of lead to operating room. Total length. Natural wave (meters). Glacier T T T T 1 1 T 1 i Inverte pyrami 10 8 6 6 8 4 4 4 4 d 2 feet 26 inches 2 feet 3 " 4 feet 4 " 5 " 4 " 170 feet 124 " 140 " 150 " 160 " 160 " 125 " 120 " 130 " 82 feet 132 " 136 " 129 " 97 " 90 " 137 " 120 " 132 " 200 " 252 feet 256 " 276 " 279 " 257 " 250 " 262 " 240 " 262 " 330 360 330 425 395 385 360 330 370 900 Mayflower Dolphin Louisiana . ... Chester Birmingham Connecticut Maine Baltimore Guantanamo a. 172. In all aerials referred to above, except the Maine and Baltimore, the long wave contained the greater amount of energy. In the case of these two aerials the greater amount of energy was radiated on the short wave. These sets (except the Shoemaker, whose closed circuit was designed to give loose direct coupling at 425 meters, and which did not require the use of the aerial loading coil for 4-25 meters) had no direct provision for changing the wave length of the aerial except in the coupling coil, and, therefore, when coupled gave a wider variation from the standard wave length than the Shoemaker sets. OPEN CIRCUIT INDUCTANCE. 173. With direct coupling the open circuit inductance forms part of the same helix as the closed circuit inductance, as has already been stated. (See fig. 40.) In inductively coupled sets the open circuit helix is movable, so that the coupling can be varied by moving the entire coil while keeping the same wave length. Provision is also made for a variable connection to the helix so that the wave length can be varied. (Figs. 42 and 43.) 174. It must not be forgotten that varying the wave length of either circuit by varying the inductance of the coupling coil or coils varies the mutual induction, as well as the self-induction, and also the coupling and damping, so that the most recent sets Fessenden (fig. 45), Tele- funken (fig. 46), Lowenstein (fig. 47) make provision for varying the wave length at some other part of the circuit than at the coupling coil, or, as in the Lowenstein sets, for automatically moving the coils so as to maintain the same coupling when the wave length is varied. These out- MANUAL OF WIRELESS TELEGRAPHY. 117 side coils are called loading coils, as distinguished from the coupling coils., by means of which energy is transferred from the closed to the open circuit (and vice versa in sets not having quenched or properly air- cooled gaps). FIG. 73. Helix and Spark Gap. The method of building the Telefunken variometer coils, shown in fig. 46, is illustrated further in fig. 74. This method of varying the self-induction of a circuit has the advantage of not having any dead ends as in the old inductance helices, shown in fig. 73. However, the vari- ometer shown in fig. 74 is not suitable for inductive coupling. None of the sending sets now in use permit the wave length of both open and closed circuits to be changed without some effort and more than one movement of the operator. Additional remarks on coupling will be found under " Adjustments," Chapter VII. For those parts of the aerial which require insulation to protect it from grounding and to protect persons, a special, heavily insulated wire called rat-tail wire, is used. 118 MANUAL OF WIRELESS TELEGRAPHY. A lightning switch (fig. 75) is installed outside the station, or where the aerial enters, by means of which it is grounded during thunder storms. The other aerial accessory the hot wire ammeter (fig. 76) r-is in- stalled in the ground lead; its uses are particularly referred to in Chapter VII. FIG. 74. GROUNDS AND GROUND CONNECTIONS. 175. As has been previously explained, wireless telegraphy makes use of earthed electric waves, as compared with the free waves discovered by Hertz and used by Marconi in his first experiments. It was soon found by Marconi that good connection to earth or to a large conducting body is essential to good working. On board ship the end of the aerial below the open circuit inductance (called the ground lead) must be well soldered, bolted, or clamped to some portion of the hull. A grounded vertical wire well earthed has a wave length not less than four times its natural length. At its free end there is a potential loop and a current node (maximum potential no current). At its earthed end there is a current loop and potential node (maximum current no potential). (See fig. 18d.) The same wire free at both ends has an electrical period equal to twice its length, and, if oscillating, has high MANUAL OF WIRELESS TELEGRAPHY. 119 potentials at both ends. If the ground connection is not good, there is a tendency to choke the current passing in and out of the earth and thus to cause a rise of potential and consequent sparking and reflection of energy at the earth connections, making the period irregular and impair- ing the sending qualities of the station. FIG. 75. Lightning Switch. FIG. 76. Hot Wire Ammeter. It should be possible to grasp the ground lead where it is soldered to the ship without injury. Inability to draw a spark there is proof of good connection. 176. At shore stations it is found that the resistance of the earth be- tween two earthed conductors a given distance apart varies widely in different localities and in the same locality with moisture and tempera- ture, and low ground resistance at a station is usually accompanied by good radiating qualities. Where and when the soil is very dry it is neces- sary to pay much greater attention to the area of the ground connections, 120 MANUAL OF WIRELESS TELEGRAPHY. and where the resistance of the earth in the vicinity of the station is high the station is a poor radiator unless an artificial ground called a " counterpoise " is installed. This can consist of any large conducting area laid on the ground or wires connected between the mast guys. The natural period of the counterpoise should be the same as that of the aerial. Generally a good ground is made by connecting the ground lead to copper plates of large area in good contact with moist earth, or to radiating lines of galvanized iron telegraph wire ending in pipes driven to moist earth, or to wire netting spread on the ground and covered with earth. At stations on tops of buildings grounds are made to the steel frames of the building and to water and gas pipes. Chapter VI. RECEIVING CIRCUITS AND APPAEATUS. 177. Receiving circuits will be considered in the following order, viz.: Open circuit, closed circuit, condensers, inductances, detectors, telephones, batteries, ampliphones, recorders. In practically all cases the same aerial wire is used for both sending and receiving. The advancing waves of electric and magnetic force from the sending aerial cut the receiving aerial and induce in it oscillating currents. If the receiving circuit has the same period as that of the passing waves, the induced oscillating currents in the aerial will increase until the energy dissipated per oscillation, by re-radiation, resistance, and transfer to other parts of the receiving circuit, is equal to that received per wave. If the receiving aerial circuit is directly or inductively connected to a closed oscillating circuit to which part of the energy received per wave is transferred during each oscillation instead of being re-radiated, this closed oscillating circuit will absorb energy, and if its period is equal to that of the arriving waves the oscillations will increase in amplitude with each half period since a closed circuit radiates slowly. If a detector is placed in either the open or closed circuit so that the oscillating currents produce differences of potential at its terminals and the maximum ampli- tude of the oscillation set up is sufficient to make it function, the passing of groups of wave trains separated into dots and dashes at the sending station can be detected at the receiving station. At the sending station the closed circuit furnishes energy to the radiating circuit, which sends it out in the shape of electric waves. At the receiving station this radiating circuit absorbs energy from the passing w r aves and transfers to the closed circuit part of what it absorbs. It is evident that no spark gap is required in the closed receiving cir- cuit and that, since no high potentials nor heavy currents need be provided for, it is not necessary that the receiving inductances and condensers should have the same dimensions or arrangement as those in the sending circuits. But in all other features receiving circuits are the exact analogue of sending circuits and the detector could occupy the place of the spark gap. However, all detectors consume energy, and placing them either directly in series with the aerial or in the closed receiving circuit is equivalent to placing a certain amount of resistance in series with these circuits, and therefore increases the resistance and 122 MANUAL OF WIRELESS TELEGRAPHY. ELEMENTARY DIAGRAMS, RECEIVING, AND DETECTOR CIRCUITS. / * I K i "s ULJ_ 150 MANUAL OF WIRELESS TELEGRAPHY. 300 3iO 32O 33O WAVE 'LENGTHS IN METERS FIG. 99. MANUAL OF WIRELESS TELEGRAPHY. 151 Dr. Austin says that if a rectifying detector and galvanometer are used for measuring the received currents the direct currents produced are also* proportional to the square of the oscillating currents ; so all these ways of measuring are directly comparable. (See curve 7.) The maximum wave meter reading is 25 at 1650 meters; at 1800 meters and 1480 meters it reads 10. By changing the setting of the wave meter either way 150 meters we reduce the strength of the received currents in the value of V25 to VlO or approximately 5 to 3. Turn now to curve III, fig. 99. We see that a change of 75 meters in the wave meter setting changes the reading from a maximum of 14 to a reading of 1. Curve III, fig. 99, is a much sharper curve than (7) of fig. 98. The same kind of wave meter being used for measuring the received currents in both cases we conclude that the sending circuit in fig. 99 is a more persistent oscillator than that in fig. 98. Compare also the shape of curve I, fig. 99, from the more rapidly damped open cir- cuit oscillating alone, with the shape of II, produced by the closed circuit alone, and III, produced by the coupled circuits. The maxima of these curves have no direct relation to each other since they are produced by different amounts of radiated energy and probably by different relative positions of the wave meter and the circuits. It is their shapes alone that are the subject of comparison and discussion. The shape of each curve will remain the same whatever the position of the wave meter (receiving circuit) . Let the portion of each curve above the heavy line X Y in figs. 98 and 99 represent the range of audibility at any distance say 100 miles from the sending set. Thus in curve II, fig. 99, a change in the receiving circuit of only 8 meters from the position of maximum loudness would render the incoming signals inaudible, while in curve 7, fig. 98, a change of 150 meters would be required to cut out sig- nals. In neither case would the lower hump audibly affect the receiving apparatus. Sharp tuning is not possible with a highly damped trans- mitter as the shape of these curves show. Neither is it possible without stiff receiving circuits. The latter should, however, be variable; that is, capable of being broadly or sharply tuned as desired (i. e., it should be possible to switch the detector from a highly damped to a rigid circuit or the circuits should be mounted so as to permit wide variations of coup- ling) . There is no more possibility of escape from a whip crack trans- mitter than from static. 208. In addition to being able to estimate damping from tuning or resonance curves we can measure it directly as follows : MEASUREMENT OF DAMPING. T> In was stated (art. 117) that the damping of any circuit 8=-^ =~ where R is the resistance, n the frequency and L the self-induction of 152 MANUAL OF WIRELESS TELEGRAPHY. the circuit. If one circuit be used to excite another and if either of the two circuits contains a variable capacity, we may plot a curve connecting the readings of the variable condenser as abscissas and the current in the second circuit as ordinates. (See fig. 100.) This is the resonance curve = 750 M. S l + B = o 038. 60 50 I 40 I o 30 o 20 z 10 48 49 50 51 52 CONDENSER SETTING IN DEGREES. FIG. 100. Resonance Curve Taken with Wave Meter. of the two circuits. The theory of coupled circuits shows that the sum of the dampings of the two circuits 8 1 + 8 2 = 7r Cm ~ C */_._*!_ ^ ? ^m T I m I where C m represents the position of the condenser in degrees for most perfect resonance, and I m the maximum current in the second circuit correspond- ing to the position of the condenser C m , and where I represents the cur- rent in the circuit corresponding to any other position C of the variable condenser. This formula becomes much simplified for practical pur- MANUAL OF WIRELESS TELEGRAPHY. 153 poses., and gives in general accurate enough results, if, instead of plot- ting a complete curve, we change the variable condenser so that for the reading G, I 2 = % P m - The quantity under the radical then becomes unity, and ^ + B 2 = Tr Cm ~ G . Two values of C should be observed ^m one on each side of (7 m , and the mean of the two values of the damping taken. If the current is measured by means of a thermo-e lenient or a perikon detector in connection with a galvanometer, the readings of the galvanometer are proportional to I 2 ; that is C is so chosen that the gal- vanometer deflection is reduced to one-half that observed with C m . If the current is read with a hot-wire instrument reading directly in am- peres, then the reading of the meter corresponding to C should be r of that corresponding to C m , since 1.41 = V2. This expression gives the true value of the dampings of the circuits only when the coupling between them is extremely loose. If the coupling is not very loose between the two circuits, the apparent value of the damping will be too large. The proper degree of coupling can be ascertained by observing the point beyond which loosening the coupling does not decrease the damping. If the damping of the wave 73 meter circuit be known or can be calculated from the formula 8 = - . T , 2nL by subtracting this from the sum of the two dampings we get at once the damping of the other circuit. If we wish to express damping in terms of wave length A instead of capacity or inductance, it may be shown mathematically that the sum of the damping 8 1 + 8 2 = 27r -^ , where as before A is the wave length, A which reduces the square of the received current to one-half of that found for resonance at \ m . 154 MANUAL OF WIRELESS TELEGRAPHY. 209. From the results of damping measurements it has been found that very sharp tuning is impracticable when a wave train contains less than 15 oscillations. This corresponds to a decrement of .2 (see fig. 62). Having measured the damping of the open circuit as coupled and found it too large it is necessary to add inductance in order to decrease it, or to weaken the coupling in order that the total resistance R may be decreased. If it is not practicable to change the wave length, the aerial must be shortened to decrease its capacity while retaining the same wave length by adding inductance. Loosening the coupling also decreases the damp- ing. Fig. 101 shows a resonance curve taken from a loose coupled sending set, showing but one maxima at 330 meters with 5 turns of inductance in the open circuit and less than -J turn in the closed circuit. This curve is steep enough to permit fairly selective receiving. MANUAL OF WIRELESS TELEGRAPHY. 155 11 156 MANUAL OF WIRELESS TELEGRAPHY. MANUAL OF WIRELESS TELEGRAPHY. 157 Fig. 102 shows much steeper curves taken from a direct connected quenched gap, 500 cycle set. The position of the two small humps in curve I, fig. 102, taken at the primary variometer indicates a coupling of - - = 22$. This is J 1 o by no means very loose coupling, but the curve shows that the aerial radiates most of its energy while oscillating in its natural period (in this case 975 meters) and that when so oscillating it is persistent enough to permit very sharp tuning. Curve II in fig. 102 taken at the aerial inductance shows but one maxi- mum which practically coincides in wave length with the maximum of curve I. ( Curve II is drawn to a different scale, so that the coincidence in maximum readings is only apparent. They are in reality smaller for the open than the closed circuit.) 210. Receiving circuits can be stiffened without changing the wave length by putting a condenser in series to decrease the capacity and then adding inductance to keep the same wave length. But the damping of sending circuits can not be conveniently changed in this way on account of the high potentials which would be induced in the series condenser. The method of measuring damping just described is applicable to receiv- ing as well as to sending circuits. Receiving circuits have in general greater resistance than sending circuits, but this is limited by specifica- tions to 4 ohms per millihenry in order not to injuriously increase the damping. . For measuring the sending current a hot-wire ammeter is installed directly in the aerial just above the ground connection. Curve 6, fig. 98, shows hot-wire ammeter readings in open circuit for various couplings and wave lengths at the G-uantanamo station. The maximum reading is for a coupling of The highest hot-wire ammeter reading shows that the circuits are in resonance and is usually taken also to indicate the best coupling; but except for circuits with quenched gaps the highest H. W. A. reading is usually obtained with a coupling which causes the radiating circuit to be too highly damped. It is therefore best to loosen the coupling until the shape of the resonance curves, or actual measurements, show sufficiently small damping; and then, by careful adjustment to resonance, attention to connections, to spark gap, and to regulator, get the highest hot-wire ammeter reading that can be obtained with that coupling and wave length. 211. In order that the performance of different sets can be compared it is necessary that all hot-wire ammeters be calibrated for reading directly and correctly in amperes. A hot-wire ammeter which reads correctly on direct current should be calibrated for high frequency as follows : 158 MANUAL OF WIRELESS TELEGRAPHY. First remove the shunt and send with reduced power so that the de- flections will approximately cover the scale. This can be done either by cutting down the actual power or by loosening the coupling between the closed circuit and aerial. Note the deflections. Then close the shunt and leaving everything else unchanged send again and note the deflection. The relation between the two deflections gives the ratio, for this wave length, of the shunted to the unshunted readings. If any other wave length is used, the shunt must be recalibrated since its effective resistance depends on the frequency. Eeports of current in aerial should always read correctly in amperes and be accompanied by report of exact frequency and input to transformer in amperes and volts. It is found that the distance of transmission varies directly as the oscillating current in the aerial, so that it is important to ascertain correctly what this current is. THE SHUNTED TELEPHONE METHOD OF MEASURING THE INTENSITY OF SIGNALS. 212. It is often desirable to make quantitative determination of the intensity of incoming signals, especially when tests are being made of either sending or receiving apparatus. This can be done if the station is provided with an electrolytic receiver, preferably of the free-wire type, and a resistance box. The connections are shown in fig. 103. I I wwvww :& < -=y=- I ^Z i T J /nnnoooooorx ^ FIG. 103. Detector Circuit with Shunted Telephone. Here L and L are wires running to the receiving circuit, K a stopping condenser, D the electrolytic, T the telephone, R a resistance box in shunt across the telephones, P the potentiometer, and C a choke coil to prevent the oscillations running around through R and P instead of passing through D when the shunt R is closed. Two 60-ohm telephones form a suitable choke. . Whatever choke coil is used, it should be tested by being placed across LL X . If the choke is perfect no oscillations will pass through it, and its presence across LL X will not diminish the loudness of the signals in the telephones. The measurement of the intensity of signal is made as follows : After the receiving circuit and detector are adjusted to give maximum loudness in the telephone, the shunt resistance R is closed and the resistance regu- MANUAL OF WIRELESS TELEGRAPHY. 159 lated until the signal just remains audible. The value of the current pulses c in the telephone, which are proportional to the energy of the incoming waves in the detector, is expressed by the following formula, where r is the value of the shunt, and t is the resistance of the telephones, and c 1 the least current audible in the telephones : r + t j f> /ii r C) c 1 is the audibility current, and the signal is often expressed as being so many times audibility. With care a series of measurements of inten- sity may be made to agree among themselves to within 5 to 10 per cent. A station is tuned, when both sending and receiving circuits are cor- rectly calibrated, coupled, and adjusted to the standard damping and wave length. 213. Inductances and capacities can be directly measured by wave meters as follows: MEASUREMENT OF INDUCTANCE AND CAPACITY. Inductance. A circuit is formed containing the unknown inductance, a known capacity (one or more standard jars), and a small spark gap. This circuit is used to excite the wave meter, and the variable condenser is varied until a maximum current in the wave meter is obtained. The two circuits being then in resonance, the product of the inductance and capacity in each is the same; that is, LG L^C^, or if L is the unknown T ini quantity, L = 7T - . C Capacity. If the spark circuit is made up with a known inductance 7" V"^ and unknown capacity, by the same process we determine that C= = . Li CARE AND OPERATION. 214. At all stations, ship and shore, the best results are invariably obtained and the most satisfactory service given by alert and careful operators who take pride in the condition of their instruments. Wireless telegraph instruments like all others depend for their efficiency on their condition and amply repay good care. An excellent operator once said that no matter how good he thought his contacts and connections were he always found that by going over them he could make them better and increase his sending and receiving efficiency. A routine, which, if followed, will ensure the proper care of a wireless set, is given in Appendix E. All sliding contacts, especially in receiver tuning coils, should be clean and bright and free from foreign matter. Sending key contacts should be kept clean and smooth and with faces parallel to each other. Detectors must be kept in their most sensitive condition and frequently tested by means of the buzzer furnished for the purpose. 160 MANUAL OF WIRELESS TELEGRAPHY. When any part -of the condenser is injured it should be immediately replaced or repaired. Any change in closed or open circuit without a corresponding change in the other throws the two circuits out of reso- nance and greatly decreases the sending radius. If the capacity in the condenser must be decreased for any cause then in order to retain the same wave length the inductance in the closed circuit must be increased. 215, The following general instructions apply to all stations: The operator shall wear the double head receiver continuously while on watch, except when necessary to communicate otherwise than by wireless. He shall satisfy himself by frequent testing with the buzzer that his detector is sensitive, and while in the vicinity of other vessels or near shore stations and using a detector that may be injured by strong sending, he shall always be alert to protect it by weakening the coupling or by opening the receiving switch. He shall familiarize himself with all sending and receiving connec- tions and adjustments and be able to tell when they are correct and to renew them when necessary ; but he shall not make any changes in any of them without the knowledge and permission of the chief electrician or operator in charge. He shall be capable of adjusting the spark gap, motor and generator rheostats and reactance regulator, so as to obtain the necessary output for the communication to be made. He shall use the shortest gap and the least power that will enable his messages to be clearly read. The spark must be kept white and crackling and have considerable volume. He shall be vigilant in noting and keeping in good condition all sending condenser connections and in keeping all articles or instruments which might be injured or cause a ground or sparking well clear of the sending apparatus at all times. He shall not, except in cases of emergency, call or send any message, when official messages are being sent or received by other vessels or stations in his vicinity. He shall be careful to file correct copies, on the official forms, of all messages sent and received by him, initialing each and filling in time and place and other information as called for on forms. He shall avoid a short and jerky style of sending. Dots and dashes and intervals must be of proper relative lengths as shown by the code in order that the sending may be clear and legible. Operators must en- deavor to attain fair speed, both in sending and receiving. Where heavy static is encountered, dots and dashes may be longer, but must preserve their relative length. The generator shall be run only during the time necessary to send messages. Where a number of tunes are ordered to be used the operators shall be MANUAL OF WIRELESS TELEGRAPHY. 161 careful to see that all circuits are correctly adjusted before attempting to send. An operator shall turn the station over to his relief clean and neat. A sending set with all connections good, closed and open circuits in resonance, no sparking from edge of condenser jars or plates, no glow from aerial and no sparking to rigging, is utilizing its power more efficiently and will be heard farther than the same set pushed to the limit but out of resonance with high resistance connections and sparking at all points. Messages shall not be sent between 11.55 A. M. and noon, 75th merid- ian time, in the Atlantic, and 120th meridian time in the Pacific. During this interval naval wireless telegraph stations send the noon time signal for the use of navigators in comparing chronometers. CODES. 216. For official use between ships of the navy and between them and naval shore wireless telegraph stations the Continental Morse code is used. Commercial shore stations in the United States, and United States coasting vessels use American Morse. All foreign stations, ship and shore, public and private, use Conti- nental Morse. American Morse is a little faster. Both codes are printed herein. The Continental Morse is a dash and dot code throughout with a maximum of four elements in any letter. The American Morse uses five elements in the letter P, four elements and a space in Y, Z and &, and a long dash for the letter L. It has a relatively less number of dashes than the Continental code and is on that account faster. It will be noted that A, B, D, E, G, H, I, K, M, N, S, T, U, V and W (fifteen out of the twenty-six. letters of the alphabet) are the same in both codes. It is to be hoped that the use of wireless telegraphy will eventually bring about an international agreement as to the elements for the re- maining eleven letters and thus provide a universal code. This will facilitate intercourse between United States ships and those of other nations and relieve operators of the necessity of learning two codes. When it is desired to communicate by the international signal book (as between two vessels whose operators do not use the same language) the " call " should be followed by the letters P R B in the Continental code. The international signal of distress is , making the letters SOS of the Continental code. The two signals given above were adopted at the International Wire- less Telegraph Conference at Berlin in 1906. The United States has not yet ratified the convention and is therefore not a party to the Inter- national Rules, but the two signals above, especially the signal of distress, 162 MANUAL OF WIRELESS TELEGRAPHY. are generally recognized. The most important rules of this convention are given in Appendix C. The rule of this convention that all stations should have three call letters is also followed in the United States. A list of these call Letters is published by the Navy Department in " Wireless Telegraph Stations of the World " and can be obtained from the Government Printing Office. Information relative to U. S. naval shore stations and shore stations in some other countries is issued in " Notices to Mariners " and shown on pilot charts published by the U. S. Naval Hydrographic Office. The rules governing communicaticn between naval shore stations and private ves- sels are published in the same manner. The substance of these will be found in Appendix B. The Act regulating apparatus and operators on steamers, which goes into effect on July 1, 1911, will be found in Appendix D. In view of the continually increasing use of wireless telegraphy it is necessary to employ concerted methods of avoiding interference other than static caused by stations, ship and shore, in the same vicinity trying to communicate at the same time, using the same wave length. These methods are not yet perfected, but are in outline as follows : (a) Standard calling wave lengths for ships and for shore stations. Say 600 meters for ships, 1000 meters for shore stations. (b) Standard communicating wave lengths or tunes different from the calling tunes, ranging from 300 to 10,000 meters, and designated by letters of the alphabet as Tune A-300 meters, Tune B-400 meters, etc. (c) Assigning specific tunes (the long ones) to shore stations which communicate only with other shore stations. Ship stations, and those which communicate with ships, call and listen for calls on 600 and 1000 meters. The station called when she acknowl- edges the call directs the calling station what tune, as C or D, to use in sending, so as to avoid interference with other tunes audible in her receiver. The above methods cannot be generally introduced until all circuits on all ships are properly calibrated and sending and receiving sets con- structed so as to permit easy, rapid, and definite changes of wave length while remaining properly coupled. 217. Whatever the speed of sending a dash is equal in length to three dots. The interval between two elements of a letter is equal in length to a dot. The interval between letters in a word is equal in length to a dash. The interval between words in a sentence is equal in length to two dashes. The length of a " space " is two dots. The long dash of the letter L in the American Morse equals two ordi- narv dashes. MANUAL OF WIRELESS TELEGRAPHY. 163 TELEGRAPH CODES. ALPHABETS. American Morse. Continental Morse. A B C D E F G H I . J K L M N O P Q R S T U V w X Y z Wait Understand Don't understand Call Finish NUMERALS. 164 MANUAL OF WIRELESS TELEGRAPHY. TELEGRAPH CODES. Continued. PUNCTUATIONS, ETC. American Morse. Period Colon Semicolon Interrogation Exclamation Fraction line Dash Hyphen Pound sterling Capitalized letter Colon followed by quotation . . . Dollar mark Decimal point Comma Paragraph Underline (begin) Underline (end) Parenthesis (begin) Parqpthesis (end) Quotation marks (begin) Quotation marks (end) Quotation within a quotation (begin) Quotation within a quotation (end) Apostrophe Spell " dot " Continental Morse. MANUAL OF WIRELESS TELEGRAPHY. 165 COMMON ABBREVIATIONS. [In use in United States telegraph services.] Abt About Af . . After Agn Again Amn American Amt Amount Anr Another Ar Answer Arv Arrive Atk Attack Atl Atlantic Awa Away Awi Awhile Ax Ask Ay Any B Be Bal Balance Bd Board Bid Bundle Bf Before Bg Being Bn Been Bot Bought Bro Brother Bk Break or back Bt But Btn Between Btr ... Better Bu Bushel Byd Beyond Bz Business Bat Battery Bbl Barrel C See Ca Came Cg Seeing Chg Charge Cr Care Ct Connect Cty City Cvl Civil Cx Capital Letter Col Collect Ck Check Da Day Dd Did Deg Degree Did Delivered Dr Doctor Drk Dark Dux Duplex DH Deadhead Ea Each Ed Editor Eng Engine Etc Et cetera Ev Ever Evn Even Exa Extra Fl Feel Fid Field Fig Feeling Flo Flow Fit Felt Fm From Fri Friday Frt Freight Gr Ground G. B. A Give better address G. A Go ahead G. S. A Give some address G. M Good morning G. E Good evening G. N Good night Gen General Ger German Gg Going Gu Guard Gv Give Gvg Giving Hb Has been Hhd Hogshead Hid Held Him Helm Hm Him Hnd Hundred Hon Honorable Hpn Happen Hqrs Headquarters Hr Here Hs His Hu House Hv Have Hw How Ify Infantry Imp Import Ix It is Ixu It is understood Kp Keep Kpg Keeping Kpt Kept 166 MANUAL OF WIRELESS TELEGRAPHY. COMMON ABBREVIATIONS. Continued. Kw Know Kwg Knowing Kws Knows Las Last Lat Latitude Lft Left Lit Little Lk Like Lt Lieutenant Lv Leave Lvg Leaving Lvs Leaves Lyg Lying Ma May Mab May be Maj Major Mar March Mas Master Mat Material Max Maximum Mch Machine Mcy Machinery Md Made Mem Member Mf d Manufactured Mgr Manager Mh Much Mil Military . Min Minute Mk Make Mkg Making Mkr Maker Mks Makes Mkt Market Ml Mail Mng Morning Mny Many Mo Month Mon Money Mrl Marshal Msg Message Msk Mistake Mst Must Mv Move Myn Million Na Name Nd Need Nee Necessary Neg Negative Ni Night No No, and New Orleans Nw None Nv Never Nun Now NX Next N. M No more Of c Officer Ofr Offer Of s Office Opr Operator Ot Out Otr Other Ov Over O. K All right PC Per cent Pd Paid Ph Perhaps Pha Philadelphia Pm Postmaster Po Post-office Pod Post-Office Department Pot President of the Potus ..... President of the United States Pr President Pra Pray Prt Part Pt Present Qk Quick Qmg Quartermaster-General Qr Quarter R Are Re Receive Red Received Reg Receiving Rcr Receiver Res Receives Ret Receipt Rek Wreck Rht Right Rlf Relief Rp Report Rpt Repeat Rr Railroad Ru Are you Ruf Rough Ry Railway Sa Senate Scotus Supreme Court of the United States Sd Should Sdu . . . Sudden MANUAL OF WIRELESS TELEGRAPHY. 167 COMMON ABBREVIATIONS. Continued. Sec Section Sed Said Sem Seem Sen Seen Sh Such Shf Sheriff Shi Shall Sig Signature Sik Sick Sis Sister Slf Self Slo Slow Sir Sailor Sm Some Sma Small Sn Soon Snc Since Snd Send Snr Sooner Snt Sent Sor Soldier Sp Ship Spfy . . Specify Spl Special Spo Suppose Ss Steamship St Street Sta State Stn Station Sto Store Str Steamer Sud Surround Sv Seven Svc Service Svd Served Sve Serve Svg Serving Svl Several Swo Swore Sx Dollar mark Sy Say S. Y. S See your service T The Tan Than Tg Thing Tgh Telegraph Tgm Telegram Tgr Together Tgy Telegraphy Th Those Thk , ..Thank Tho Though Thr Their Ti Time Tk Take Tkg Taking Tkn Taken Tkt Ticket Tlk Talk Tm Them Tn Then Tnd Thousand Tni To-night Tnk Think Tr There Tru Through Ts This Tse These Tt That Ttt That the (5) Tuf Tough Tw To-morrow Ty They U You Uc You see Un Until Uni United Upn Upon Ur Your Urg Urge Val Value Vy Very W With Wa Way Wat Water Wd Would Wea Weather Wg Wrong Wh Which Wi Will Wit Witness Wl Well Wlk Walk Wn When Wnt Want Wo Who Worn Whom Wos Whose Wr Were Ws Was Wt What Wu . . . Western Union 168 MANUAL OF WIRELESS TELEGRAPHY. COMMON ABBREVIATIONS. Continued. 13 Understand 25 .... I am busy now 30 No more 73 Accept best regards 77 Message for you 92 Deliver " Wire " Give instant possession of line for test. An addition to the foregoing XX is gradually coming into general use as a symbol for "interference" which has no counterpart in wire telegraphy. Wy Why Y Year Ya Yesterday 4 Please start me, or where 5 Have you anything for me 9 Important official mes- Chapter VIII. WIRELESS TELEPHONY. WIRELESS TELEGRAPHY USING UNDAMPED OSCILLATIONS, DIRECTION FINDERS AND DIRECTION SENDERS. 218. All wireless telephone sets thus far supplied, having proved unre- liable in action, have been withdrawn from service. The workings of these sets depended on the production of undamped oscillations in the sending circuits. The apparatus was in principle like that shown in fig. 104, using 220 volt direct current. The electrodes were copper and carbon and the arc was horizontal. A small alcohol lamp was placed immediately under the arc with its flame burning in the arc. The inductance and capacity shunted around the arc formed with it an oscillatory circuit similar to the closed circuit in an ordinary wireless transmitter. These sets were used with the ordinary ship aerial tuned to the arc circuit. The amplitude of the oscillations induced in the aerial were modified by a carbon transmitter in series with the aerial as shown in fig. 105. Talking into the carbon transmitter varied the aerial resistance. 219. Assume that the undamped oscillations had a frequency of 700,000 and the notes of the human voice varied through two octaves (say from 300 to 1200 vibrations per second). The vibrations of the telephone diaphragm by changing the resistance of the carbon modified the oscillating current in the aerial (and therefore the amplitude of the electric waves generated) in accordance with the vibrations of the voice of the person speaking. The ordinary receiving circuit having a crystal or electrolytic detector serves as well for undamped oscillations as for groups of wave trains, transforming the modified oscillations into human speech in the receiving telephone. The limit of mechanical or air vibrations recognized as sound is be- tween 30,000 and 40,000 per second. Although the undamped oscilla- tions are of a much higher frequency and therefore produce no sound in themselves, modifications of the amplitude of successive waves may be of such a nature as to produce sound by slower variations in the rise and fall of the received current. 220. The transmitting telephone may be in the arc circuit instead of the aerial as shown, or it may be inductively connected to either the open or closed circuit. There is as yet no standard practice. The telephone transmitters are specially constructed to stand the voltage and current 170 MANUAL OF WIRELESS TELEGRAPHY. induced in the aerial or that in the closed circuit. It is claimed that a type will soon be perfected that will carry 10 amperes. It has not been found practicable to vary the arc current sufficiently to produce large powers in the oscillating circuits. Arcs in parallel and in series have been tried but without marked success. 221. The arc method of producing undamped oscillations with direct current was discovered by Professor Elihu Thompson in 1892 and has been developed by many other investigators. In order to prevent the oscillations from running back to the dynamo choke coils or very high resistances must be placed in the D. C. leads. (See figs. 104 and 105.) The simple theory of the formation of the oscillations is as follows : When the shunt containing inductance and capacity is closed around the arc in a circuit like that shown in fig. 104, a part of the current flows RESISTANCE CHOKE COIL. FlG. 104. into the condenser, thus robbing the arc of a part of its current; but as the D. C. potential across an arc increases as the current decreases, this decrease in current increases the potential difference, and the condenser continues to charge. At the next instant, however, the condenser com- mences to discharge, increasing the direct arc current until it is entirely discharged; then the process repeats itself. Oscillations can be produced in this way from almost any form of arc and over a wide range of voltages, but it is found that high frequency oscillations are best produced when the direct current voltage is high (500 volts or more), and when the positive arc electrode is capable of conducting away heat rapidly. This rapid cooling of the arc plays a very important part in the production of the oscillations, as it causes the arc to die down rapidly and increases the suddenness with which the current flows into the condenser. It has also been found that when the arc is formed in an atmosphere capable of assisting in this cooling, the energy of the oscillations is vastly increased. The best gaseous conductor of heat is hydrogen, and consequently the best results are obtained in an atmosphere of hydrogen or some mixed gas or vapor containing hydrogen. Common illuminating gas gives excellent results, and recently alcohol MANUAL OP WIRELESS TELEGRAPHY. 171 introduced into the arc chamber drop by drop and vaporized by the heat of the arc has come into use. It has been suspected that these gases and vapors may have some effect on the electrical conductivity of the arc as well as on its cooling, but this point is still unsettled. 222. Another device which is made use of for increasing the energy of the oscillations which can be obtained from the arc is forming it in a magnetic field the lines of force of which are at right angles to the arc length. The action of the magnetic field is twofold; first it deflects the arc to one side, increasing its length and consequently the difference of potential between the arc electrodes, and second, it blows out of the field the conducting ions formed in the gas, thus decreasing the arc conductivity and still further increasing the difference of potential be- tween the electrodes. PILOT LAMP FIG. 105. For the successful production of oscillations a correct relation must exist between the arc current, the arc length, and the strength of the magnetic field. This relation in general can be obtained only by ex- periment. If these adjustments are not correctly made several sets of useless superposed oscillations may be produced in the condenser circuit. Therefore it is necessary in working with waves produced from the arc to examine its oscillations from time to time with the wave meter, in which, if the adjustment be correct, but one sharp and powerful maxi- mum will be found. Fleming states that the best results are obtained when the ratio of the capacity to the inductance in the oscillating circuit is small about 1 to 20 when both are measured in centimeters. 12 MANUAL OF WIRELESS TELEGRAPHY. 223. Fessenden has developed another means of producing undamped oscillations by constructing alternators giving as high as 150,000 alter- nations per second. The open circuit is connected directly to the ter- minals of these alternators and tuned to the alternator frequency. The use of these very high frequency alternators does away with all trans- formers, condensers and inductances except the aerial tuning inductance. They are, however, not yet in general use, being difficult to construct and, on account of their high speed, difficult to operate. They are suitable for either wireless telephony or telegraphy. 224. The advantage to be derived from the use of undamped oscilla- tions is considerable. We have forms of wireless detectors, like the elec- trolytic and perikon receivers, which respond in proportion to the total energy passing through them. Detectors of this kind will give the same response whether the energy is introduced in the form of an undamped continuous train of small amplitude or a damped train consisting of a few waves some of which are of large amplitude. The undamped waves offer great advantages in the way of sharp tuning, and enable the receiv- ing circuits to be so set up that they may be made comparatively free from the interference from other stations and from atmospheric dis- turbances. The advantages at the sending stations are no less important, for there, with the high potentials used in a spark circuit, a considerable portion of the energy is wasted on account of brush discharges in the condensers and in other portions of the circuit, and on account of leakage due to faulty insulation. With the undamped oscillations these difficulties prac- tically disappear, for with maximum potentials not exceeding 1000 volts in the primary circuit, an amount of energy can be transmitted to the antenna which would with the spark circuit require potentials of 30,000 or more volts. It is also claimed for the undamped oscillations that they travel over rough and broken country with much less absorption than is found in the case of the ordinary spark waves, but in regard to this and many other questions concerning the qualities of undamped oscillations we must wait for confirmation until they come into more general use. In using undamped oscillations for wireless telegraphic purposes it must be remembered that the frequency of the oscillations themselves is too high to be heard in the telephone connected with the ordinary re- ceiving circuit, and when the circuit at the sending station is closed all that would be heard is a slight click, so that there is no way of telling a dot from a dash. This makes it necessary to place a rapidly rotating circuit breaker in the circuit for the purpose of creating a buzz in the telephone at the receiving station when the circuit is closed. This circuit breaker is ordinarily placed in the aerial, while the sending key is placed either in the aerial or shunted around a few turns of the aerial induc- tance, in which case it serves merely to throw the aerial in and out of tune with the closed circuit. MANUAL OF WIRELESS TELEGRAPHY. 173 THE POULSEN TICKER RECEIVER FOR UNDAMPED OSCILLATIONS. 225. If no interrupter is used in the sending apparatus for undamped oscillations, no signals can be read at a receiving station unless the wave trains are there broken up so as to produce a buzz in the telephone. For this purpose the Poulsen ticker is sometimes used, which, at the same time does aAvay with the need of any special receiver. It consists essen- tially of a circuit breaker actuated by a small magnetic vibrator, kept in action by a dry cell. In this receiver the closed circuit is coupled very loosely to the antenna (see fig. 106), and this circuit is intermittently FIG. 106. connected to a large condenser K, of the order of a microfarad, by the ticker at A. During the time of contact the condenser K becomes charged, and when the contact is broken it discharges itself through the telephone, producing a note corresponding in tone to the frequency of the ticker. DIRECTION FINDERS. 226. The experimental installations of direction finders have been with- drawn, it not being found practicable to operate them. The principle on which they operated was that two vertical wires parallel to the plane of movement of an electric wave, if half a wave length apart, would have electric currents of opposite phase induced in them, which could be made to double the receiving effect as compared with a single wire, while if at right angles to the plane of movement of the wave, the induced currents would be in the same direction and could be made to neutralize each other. If the plane of this direction finder pointed towards the sending station, the strength of the received signals would be a maximum. If at right angles to the sending station, it would be a minimum. By swinging the ship in azimuth, the compass heading, when the strength of signal was a maximum, would indicate the line of bearing of the sending station. The practical difficulty in the way of operating this system to the best advantage is the very short waves which are necessary on account of the comparatively short distances that can be obtained be- tween wires on board ship. For instance, with masts 200 feet apart the 174 MANUAL OF WIRELESS TELEGRAPHY. wave length should be 400 feet, whereas the navy standard wave length is 1275 feet, The plane of such an aerial relative to the direction of the electric waves makes a difference in the strength of received signals whether the distance between wires is half a wave length or not. And this fact is utilized in the Bellini-Tosi apparatus where two such aerials are installed (see fig. 63) in planes at right angles to each other. The open circuit receiving coils are mounted so that they are in the same planes as their aerials. The closed circuit coil can be placed in the plane of either aerial or in any intermediate position. Its plane when the strength of signals is a maximum is approximately that of the passing waves. 227, ~No attempt to send directed waves from ships has been made. On shore, direction finders can be more successfully used than on ships. It is found by Marconi that a flat top aerial like fig. 64 sends more strongly in the direction away from the free end of the aerial and re- ceives more strongly from the direction in which it sends the best. This effect with the comparatively short horizontal part of the aerial on ships is not appreciable, but on shore where the horizontal part can be made long as compared with the vertical part it has proved to be of practical use, both as a direction sender and receiver. The trans- Atlantic wireless stations at Clifden and Glace Bay have their aerials pointing away from each other. To be used as a direction finder such an aerial would have to be revolved rapidly or the horizontal part extended in a number of directions like the spokes of a wheel, to any one of which the vertical part could be connected at will. PORTABLE SETS. 228. These, as their name indicates, are special small sets which have their own source of power, such as a foot or hand operated generator, and when used on shore have portable masts for supporting the aerial. On board ship this single wire aerial can be run up by signal halliards, and if insulated wire is used (since portable sets work usually at low voltages) no particular care need be taken to prevent the wire from touching the mast, deck or rigging. The suit-case type illustrated in fig. 107 weighs about 75 pounds com- plete. It has a motor generator for ship use, which has an output of 50 watts and can be plugged in on any lighting circuit. Small gasolene driven generators are used for some portable shore sets, the entire send- ing and receiving apparatus being mounted on wheels. The power or hand operated generator set of the suit-case type is good for about 20 miles. (See fig. 107.) A complete set is seen with condenser, inductance, and key in the left half; motor generator, quenched gap, transformer, and receiving apparatus on the right half of the case ; with the plug for connecting up with the lighting or power circuit at the upper left hand corner. MANUAL OF WIRELESS TELEGRAPHY. 175 229. To illustrate an actual wireless telegraph installation the station at Sitka, Alaska, has been selected. This station is situated on Japonski Island (see frontispiece). The masts,, rigging and rigging insulators, aerial and buildings are shown in fig. 108 ; one unit of the generating sets in fig. 109; the receiving apparatus in fig. 110. These figures repay study as illustrating a neat and workmanlike installation. The sending and receiving apparatus is after the designs of Professor Pierce. FIG. 107. N. E. S. Co.'s Portable Set Signal Corps. Figs. Ill and Ilia illustrate actual receiving sets of other types, the elementary diagrams of which are shown in figs. 83 and 86. The construction and arrangement of both sending and receiving apparatus will continue to vary, but a careful study of elementary dia- grams (figs. 40 to 48 and 77 to 88) in connection with installation dia- grams like figs. 112, 112a, 112b, 112c, which accompany each set will enable an electrician to connect up and operate any set intelligently. There are too many types of apparatus in use to warrant a detailed description or illustration of each. Such description and instructions are furnished with each set. This manual has therefore been confined to the principles common to practically all wireless sets. 176 MANUAL OF WIRELESS TELEGRAPHY. MANUAL OF WIRELESS TELEGRAPHY. 177 178 MANUAL OF WIRELESS TELEGRAPHY. FIG. 110. MANUAL OF WIRELESS TELEGRAPHY. 179 FIG. 111. Wireless Telegraph Receiver. 180 MANUAL OF WIRELESS TELEGRAPHY. TETLEFUNKEN. FIG. 112. TO AERIAL HOT WIRE AMM. SHIP MAINS CLOSED CIRCUIT OPEN CIRCUIT INDUCTANCE. INDUCTANCE FESSELNDEM . FIG. 112A. MANUAL OF WIRELESS TELEGRAPHY. 181 182 MANUAL OP WIRELESS TELEGRAPHY. MANUAL OF WIRELESS TELEGRAPHY. 183 STATIC AND PREVENTION OF INTERFERENCE. 230. Methods of preventing interference by the use of standard calling wave lengths and codified standard communicating wave lengths were referred to under codes (art. 216). The use of undamped oscillations would materially assist in the sharp tuning necessary to accomplish the above successfully ; but neither undamped nor damped oscillations can be relied upon to completely eliminate the effects of the vagrant waves and local electrification grouped under the name of " static." Every lightning discharge produces powerful electric waves which affect conductors at great distances, and since thunderstorms in warm climates, and especially in summer, are almost continuous in the sense of existing somewhere in the area in which they affect detectors, the interference caused by them is almost continuous. The waves created by lightning discharges vary greatly in length; but are highly damped and affect all aerials more or less. Again, at every wireless station the air at the top and foot of the aerial is at different potentials. The atmospheric potential gradient at any station varies with the time of day, the season of the year, and the local weather con- ditions. It is usually steeper in summer. This difference of potential tends to equalize itself through the aerial. The upper air is usually positively electrified, the earth negatively. The amount and regularity of the discharge to ground at any time depend on the difference of potential between the upper air and the ground at the time and the amount of electrified air which comes in contact with the aerial. The discharges are usually intermittent and vary in strength. Some- times they produce a continuous roar in the telephone. In this respect the note of the spark affects reception and it is possible to read a 500-cycle note through static which would render a 60-cycle note unintelligible. Whatever tends to selectivity or inertia in receiving circuits, such as large inductances, also tends to decrease static interference. Inductively coupled receiving sets afford a direct path to ground, so that static charges do not accumulate on the aerial, and the inductive coupling weakens the energy transfer of all induced currents which are out of tune. We see therefore that loose coupling, small damping and high fre- quency, which we desire for other reasons, are also desirable as tending to eliminate static interference. APPENDICES. NOTE 1. The following list of metals is arranged in such order that any one will be the positive pole of the battery when used with the metal next below it on the list as a battery element and the negative pole when used with the element next above it, the difference of potential between any two being greater the farther apart they are in the series. Carbon. Silver. Lead. Zinc. Platinum. Copper. Cadmium. Magnesium. Gold. Iron. Tin. Sodium. The amount of potential difference also depends on the battery solution, and in some instances it may be reversed. Commercial primary batteries are of copper and zinc, with an E. M. F. of approximately 1 volt, and carbon and zinc, with an E. M. F. of from 1.4 in Leclanche cells and some dry cells to 2.1 in some types of wet cells, depending on the electrolyte. NOTE 2. The relations existing between electricity and matter have been most ex- haustively investigated by Prof. J. J. Thomson, who has proved that electric- ity has an atomic structure and that it can exist separately from an atom of matter. When a current is sent through a vacuum tube, the luminous beam pro- ceeding from the cathode has been shown to consist of particles projected from the cathode. These particles are capable of turning a small wheel. The cathode beam can be deflected by either a magnetic or an electric field, and it is found to consist of particles of negative electricity or of parts of the atom negatively charged, each having about one-thousandth of the mass of an atom of hydrogen. These particles are the same, no matter what gas is used in the vacuum tube. They are usually called electrons. When an electron is broken off from an atom, the remaining part is positively charged. Currents of elec- tricity, however produced, are the result of the decomposition of atoms into positive and negative electric charges. There can be no electric current without movement of electrons. Conductors are bodies in which the break- ing up of atoms and movements of electrons take place more or less easily. Some free electrons exist in all bodies. It is by setting these into vibration and by means of this vibration making them break off similar particles from neighboring atoms, and thus propagate the disturbance throughout the mass of the conductor, that electric currents are generated. 186 APPENDICES. APPENDIX A. TABLE 1. [Extract from Fleming's Cantor lecture, Journal of Society of Arts, p. 196, January 5, 1906. Taken mostly from A. Heydweiller, " On Spark Poten- tials." Ann. der Physik, vol. 248, p. 235 (1898).] SPABK VOLTAGE BETWEEN BRASS BALLS 2 CENTIMETERS IN DIAMETER FOR VARIOUS SPARK LENGTHS. Spark length (cms.). voFtag*. S P ark len * th (cm8 ' ) ' voltage. 0.1 4,700 1 31,300 0.2 8,100 1.5 40,300 0.3 11,400 2 47,400 0.4 '. 14,500 2.5 53,000 0.5 17,500 3 57,500 0.6 20,400 3.5 61,100 0.7 23,250 4 64,200 0.8 26,100 4.5 67,200 0.9 28,800 5 69,800 TABLE 2. Condenser Capacity Required to Give Full a na -,r ,-r.n-a.ro Power for Spark Voltage of 30.000 (0.4" Gap) bp of U 000 K. W. and One Discharge Per Half Cycle. (0 15" gap). r 60 cycles. 120 cycles. "~460 cycles? 500 cycles. 1 0.019m. f. 0.009m. f. 0.002 m. f. 0.010 m. f. 2} 0.047 " 0.023 " 0.006 " 0.025 " 5 0.09a " 0.047 " 0.012 " 0.050 " 10 0.185 " 0.093 " 0.024 " 0.100 " 15 0.27& " 0.139 " 0.036 " 0.150 " 35 0.648 " 0.324 " 0.085 " 0.350 ." 1 standard jar condenser = 0.002 m. f. microfarads X spark voltage X spark voltage X spark frequency K.W. = 2,000,000,000 TABLE 3. INCREASE OF RESISTANCE OF COPPER WIRE AT A FREQUENCY OF 400,000 PEK SECOND (750 METER WAVE LENGTH). Diameter Increase of wire. in resistance. 0.2 mm. 1 per cent. 0.4 " 22 0.8 " 120 2.0 " 650 4.0 " 1000 APPENDICES. 187 TABLE 4. SPECIFIC RESISTANCE OF WATER AND SOILS. Sea water 100 Fresh water 100,000 Damp soil 10,000 100,000 Dry soil >1,000,000 TABLE 5. LOGARITHMIC DECREMENT (8) OF WAVE TRAIN AND THE APPROXIMATE NUMBER OF WAVES (N.) IN THE TRAIN BEFORE THE AMPLITUDE FALLS TO ONE-TENTH OF THE MAXIMUM. 5 N 5 N 1.0 3.5 .1 24.0 0.8 4.0 .08 30.0 .6 5.0 .06 39.0 .4 7.0 .04 58.0 .3 8.5 .03 78.0 .2 12.5 .02 116.0 Good tuning is not possible with less than fifteen waves in the train. TABLE 6. SOME COMMON UNITS EXPRESSED IN TERMS OF ABSOLUTE UNITS. 1 microfarad = 1 . 10-is c. g. s. 1 millihenry = 1 . 10 " 1 microhenry = 1 . 10 s " 1 volt = 1 . 10s 1 ohm = 1 . 10 1 ampere = 1 . 10-i " 1 watt = 1 . 107 TABLE 7. SOME COMMON HIGH-FREQUENCY EQUATIONS. The time of oscillation of a condenser circuit is T = 2 TT */ LC seconds. (L in henries, C in farads.) 1 v = n'k and T =. , n where v is the velocity, n the frequency, and X the wave length. The wave length is therefore /, m: V . &TC ^/ AJ\J) A = 1.885 */LC~ .10 9 meters. 13 188 APPENDICES. In a condenser charged N times per second the energy passing through in watts. one second is . CF 2 (C in microfarads and V in volts.) The damping of a single circuit is (R in ohms and L in henries or both in absolute units.) The damping of two circuits by the resonance method, or See art. 208. The following equation and tables are the results of experiments conducted between Brant Rock station and the cruisers Salem and Birmingham in 1909-10. See " Some Quantitative Experiments in Long Distance Radio- Telegraphy," by L. W. Austin, Reprint No. 159, from Bulletin Bu. of Stand- ards, Vol. 7, No. 3, Feb. 1, 1911. Equation: I R 4.25 x I s x * x e ~ ^ I s = Antenna current, sending, in amperes. I R = Antenna current, receiving, amperes through 25 ohms. 7i, = Height of flat-top antenna, sending station, in kilometers. A 2 Height of flat-top antenna, receiving station, in kilometers. a = .0015. d = Distance in kilometers. /. Wave length in kilometers. = 2.7183. 25 ohms = high-frequency resistance of ship aerial of 1000-meter wave length. The above equation covers the normal-day received current over salt water, through 25 ohms for two stations with flat-top aerials of any height, with any value of sending current and any wave length, provided the sending station is so coupled as to give but one wave length. The following tables (8, 9, 10 and 11, 12) illustrate the application of this equation: TABLE 8. For good communication received current should be equal to 7^ = 40 X 10- amperes through 25 ohms = 40 X 10-s watts = ^ erg per second. For auditle signals 7^ = 10 X 10- amperes through 25 ohms = 2.5 X 10- watts = V er S Per second. APPENDICES. 189 TABLE 9. Calculated Relation between Antenna Current and Distance for Two Ships with Antenna Heights of 130 Feet. A = lOOOm. Antenna Current Is- Working Distance 40.10- amp. Extreme Distance of Audibility 10. 10- 6 amp. Day. Night. (Zero Absorption) Day. Night. (Zero Absorption) 1 amp. 75 miles 90 miles 200 miles 360 miles 2 135 180 300 720 3 180 270 375 1080 5 235 450 475 1800 7 280 630 550 2520 10 345 900 630 3600 15 420 ]350 725 5400 20 475 1800 790 7200 25 525 2250 840 9000 30 565 2700 900 10800 40 630 3600 970 14400 50 685 4500 1025 18000 60 725 5400 1150 21600 TABLE 10. Good Working Distance and Sending Current for Two Stations with Flat-Top Antennas 450 Feet High. Nautical Miles. A = 1000 m. A = 2500 m. A = 3750 m. A = 6000m. 1000 15 amp. 13.5 amp. 15 amp. 17 amp. 1250 38 27 27 30 1500 91 49 44 46 1750 200 95 77 74 2000 490 155 122 105 2250 245 200 160 2500 ... 470 314 235 2750 ... 500 335 3000 ... ... 775 500 190 APPENDICES. i CO Ol o . 8 S M S 1 g I J- & fe s s a ss a s s rH CO CO CO ^ rocot2o5wr^^co^ ~ 2 S S co i-i d d wcocoo Sooo s * c CD II fcl-C CstCOlOC^ o o o ** o O S 00 -* N I 24 ^ uJ pq -^ H i^ i I 1 I CM CM 8 "f co ^ os co II * 8 || || | : s "1 H s i II , 8 8 | S | S j . CO O CO 1C CD 1 i i S 03 CC OS to OS IO CO O t- CO CM i I T-I j3 a id ^t >o o o o o o CO CD CC S S ; - S 03 >0 CO CO t- ^1 tO ^ ' -* t- -- CO l~ >O CO II - cVJ io 8 co S o 8 ' H d fa 03 05 OO t- CD O OO ?l to CO i-( i ( A II to APPENDICES. 191 APPENDIX B. 1. The facilities of the naval coastwise wireless telegraph stations (includ- ing the one on the Nantucket Shoal light-ship), for communicating with ships at sea, where not in competition with private wireless telegraph stations, are placed at the service of the public generally and of maritime interests in particular under rules which are subject to modification from time to time, for the purpose of (a) Reporting vessels and intelligence received by wireless telegraphy with regard to maritime casualties, derelicts at sea, and overdue vessels. (ft) Receiving wireless telegrams of a private or commercial nature from ships at sea, for further transmission by telegraph or telephone lines. (c) Transmitting wireless telegrams to ships at sea. 2. For the present, this service will be rendered free. All messages will, however, be subject to the tariffs of the land lines. Arrangements have been made with both the Western Union and Postal Telegraph companies for forwarding messages received from ships at sea. When a message is not pre- paid the company delivering it will collect the charges. Shipowners should arrange with companies operating the land lines as to tariffs and the settle- ment therefor. 3. The light-ship stations will report vessels and transmit messages from them if the signals are made by the international code or any other known to the officers on the light-ship. 4. When notified by the Weather Bureau of the Department of Agriculture, naval wireless telegraph stations will give storm warnings to vessels com- municating with them by wireless telegraphy. Storm warnings will be sent to the light-ships by wireless telegraphy, and storm signals furnished by the Weather Bureau will be displayed therefrom to warn passing vessels. Storm warnings and hydrographic information, such as location of derelicts and other dangerous obstructions to navigation, are sent broadcast at 8 A. M., noon (immediately after noon-time signal), 4 P. M. and 8 P. M. local standard time when received. Weather reports and other hydrographic information on file are furnished on request. 5. All vessels having the use of the n^val wireless telegraph service are requested to take daily meteorological 7 observations of the weather when within communicating range and to transmit such observations to the Weather Bureau by wireless telegraphy at least once daily, and transmit observations oftener when there is a marked change in the barometer. 6. All shipowners desiring to use any special code of signals for communi- cating with the Nantucket Shoal light-ship station or any of the shore stations, or make any other special arrangements, should communicate with the Navy Department, Washington, D. C. 7. All chambers of commerce, maritime exchanges, newspapers, news agencies, and others desiring to have vessel reports and general marine news forwarded to them regularly should communicate with the Navy Depart- ment in order that necessary arrangements for the service may be made. In no case will an operator attached to a station be allowed to act as an agent for any individual or corporation, but all vessel reports and marine news not of a private nature will be supplied to all applicants, so long as this service does not too greatly tax the personnel of the stations, when it will be necessary for those desiring information involving much time for its distribution to appoint agents, who will be allowed access to the station bulletins. 192 APPENDICES. 8. Naval wireless telegraph stations are equipped with apparatus of several systems and can communicate with all the wireless telegraph systems now in use, if tuned to the same wave length. The Department is desirous of co-operating with all shipowners wishing to avail themselves of its wireless telegraph service, and it is believed that there will be little or no difficulty in arranging for communication between its stations and ships equipped with apparatus of other systems, if the owners of the apparatus as well as the owners of the ships are desirous of establishing such communication. INSTRUCTIONS FOR COMMUNICATION BY WIRELESS TELEGRAPHY BETWEEN WIRELESS TELEGRAPH STATIONS AND SHIPS. I. A vessel wishing to communicate with a station and having ascertained by " listening in " that she is not interfering with messages being exchanged within her range should make the call letter of the station. II. The call should not be continuous, but should be at intervals of about three minutes in order to give the station a chance to answer. III. After the station answers the vessel should send her name, distance from station, weather, and number of words she wishes to send; then stop until the station makes O. K., signals the number of words she wishes to send to vessel, and signals go ahead. IV. Then the vessel begins to send her messages, stopping at the end of each 50 words and waiting until the station signals O. K. and go ahead; when all messages have been sent she will so indicate. If the sender desires to designate the Western Union or Postal Telegraph system for further transmission of his message, he should do so immediately after the address, as, for example " A. B. C., Washington, D. C., via W. U. (or P. T.)." V. When a vessel has indicated that she has finished, the station will send to the vessel such messages as she may have for her in the following order: (a) Government business, viz., telegrams from any Government Depart- ments to their agents on board. (&) Business concerning the vessel with which communication has been established, viz., telegrams from owner to master. (c) Urgent private dispatches, limited. (d) Press dispatches. (e) Other dispatches. VI. In the case of the Nantucket Shoal light-ship, it will, immediately on receiving the vessel's call, acknowledge, and (after receiving vessel's name, distance, weather report, and number of words she wishes to send) transmit the first three to Newport, and then tell the vessel to go ahead with her messages. VII. After receiving these and sending the vessel any message on file for her, the light-ship will transmit to Newport messages received from the communicating vessel in the following order: (a) Government business. (&) Urgent private dispatches, limited. (c) Press dispatches. (d) Other dispatches. VIII. A naval wireless telegraph station has the right to break in on any message being sent by a vessel at any time, and the right of way may be given at any time to a government vessel or one in distress. APPENDICES. 193 IX. When two or more vessels desire to communicate with a naval wireless telegraph station at the same time, the one whose call is first re- ceived will have right of way, and the others will be told to wait and will be taken up in turn. Vessels having been told to wait must cease calling. X. In case communication is not established with any ship for which messages are on file, the naval wireless telegraph station will notify the telegraph company from which the messages were received, giving sufficient information for them to identify the telegrams and notify the sender. XL In order to obtain the best results, both sending and receiving appara- tus should be tuned to wave length of 425 meters. (Subject to change.) XII. In order that all messages received at naval wireless telegraph sta- tions may be forwarded to ships for which they are intended, and in order that all ships equipped with wireless telegraph apparatus may receive storm warnings, they should always report when in signaling distance of a naval wireless telegraph station. XIII. The service being without charge at present, the Government accepts no responsibility for the reception or transmission of messages from or for passing vessels. Every effort will be made to transmit all messages without error and as expeditiously as possible. It must be remembered that errors are not uncommon in ordinary telegraph and cable messages, so that due allowance should be made. XIV. In order that the service may be made as good and as useful as pos- sible, complaints should be promptly reported to the Navy Department as soon as possible after the cause therefor, giving date, hour and other details, to enable the Department to investigate the case. XV. Information regarding the naval wireless telegraph service will be published in " Notices to Mariners," issued by the Hydrographic Office of the Navy Department. (Rules XV to XXXII, adopted by the International Wireless Telegraph Con- ference of Berlin, 1906, are in force at U. S. Naval coastwise wireless tele- graph stations. See Appendix C for the most important of these.) APPENDIX C. SERVICE REGULATIONS ANNEXED TO THE INTERNATIONAL WIRELESS TELEGRAPH CONVENTION. [Extracts from International Wireless Telegraph Convention, Berlin, November, 1906.] ARTICLE 2. By " coastal stations " is to be understood every wireless telegraph station established on shore or on board a permanently moored vessel used for the exchange of correspondence with ships at sea. Every wireless telegraph station established on board any vessel not per- manently moored is called a " station on shipboard." ARTICLE, 3. The coastal stations and the stations on shipboard shall be bound to ex- change wireless telegrams without distinction of the wireless telegraph system adopted by such stations. 194 APPENDICES. IV. It is understood that, in order not to impede scientific progress, the pro- visions of article 3 of the convention shall not prevent the eventual employ- ment of a wireless telegraph system incapable of communicating with other systems, provided, however, that such incapacity shall be due to the specific nature of such system and that it shall not be the result of devices adopted for the sole purpose of preventing intercommunication. 1. ORGANIZATION OF WIRELESS TELEGRAPH STATIONS. I. The choice of wireless apparatus and devices to be used by the coastal stations and stations on shipboard shall be unrestricted. The installation of such stations shall, as far as possible, keep pace with scientific and technical progress. II. Two wave lengths, one of 300 meters and the other of 600 meters, are authorized for general public service. Every coastal station opened to such service shall use one or the other of these two wave lengths. During the whole time that a station is open to service it shall be in condition to receive calls according to its wave length, and no other wave length shall be used by it for the service of general public correspondence. Each government may, however, authorize in coastal stations the employment of other wave lengths designed to insure long-range service or any service other than for general public correspondence established in conformity with the provisions of the convention, provided such wave lengths do not exceed 600 meters or that they do exceed 1600 meters. III. 1. The normal wave length for stations on shipboard shall be 300 meters. Every station on shipboard shall be installed in such manner as to be able to use this wave length. Other wave lengths may be employed by such stations provided they do not exceed 600 meters. 2. Vessels of small tonnage which are unable to have plants on board insuring a wave length of 300 meters may be authorized to use a shorter wave length. IV. 1. The International Bureau shall be charged with drawing up a list of wireless telegraph stations of the class referred to in Article 1 of the con- vention. Such list shall contain for each station the following data: (1) Name, nationality and geographical location in the case of coastal stations; name, nationality, distinguishing signal of the International Code and name of ship's home port in the case of stations on shipboard. (2) Call letters (the calls shall be distinguishable from one another and each must be formed of a group of three letters). (3) Normal range. (4) Wireless telegraph system. (5) Class of receiving apparatus (recording, acoustic or other apparatus). (6) Wave lengths used by the station (the normal wave length to be underscored). APPENDICES. 195 (7) Nature of service carried on by the station. General public correspondence. Limited public correspondence (correspondence with vessels ....). (8) Hours during which the station is open. (9) Coastal rate or shipboard rate. 2. The list shall also contain such data relating to wireless telegraph stations other than those specified in Article 1 of the convention as may be communicated to the International Bureau by the management of the Wire- less Telegraph Service ("administration") to which such stations are subject. V. The exchange of superfluous signals and words is prohibited to stations of the class referred to in Article 1 of the convention. Experiments and prac- tice will be permitted in such stations in so far as they do not interfere with the service of other stations. VI. 1. No station on shipboard shall be established or worked by private enterprise without authority from the government to which the vessel is subject. Such authority shall be in the nature of a license issued by said government. 2. Every station on shipboard that has been so authorized shall comply with the following requirements: (a) The system employed shall be a syntonized system. ( & ) The rate of transmission and reception, under normal conditions, shall not be less than twelve words a minute, words to be counted at the rate of five letters each. (c) The power transmitted to the wireless telegraph apparatus shall not, under normal conditions, exceed 1 kilowatt. Power exceeding 1 kilowatt may be employed when the vessel finds it necessary to correspond while more than 300 kilometers distant from the nearest coastal station, or when, owing to obstructions, communication can be established only by means of an increase of power. 3. The service of the station on shipboard shall be carried on by a tele- graph operator holding a certificate issued by the government to which the vessel is subject. Such certificate shall attest the professional efficiency of the operator as regards: (a.) Adjustment of the apparatus. (&) Transmission and acoustic reception at the rate of not less than 20 words a minute. (c) Knowledge of the regulations governing the exchange of wireless tele- graph correspondence. 4. The certificate shall furthermore state that the government has bound the operator to secrecy with regard to the correspondence. 2. HOURS OF SERVICE OF COASTAL STATIONS. VIII. 1. The service of coastal stations shall, as far as possible, be constant, day and night, without interruption. 196 APPENDICES. XVI. Ships in distress shall use the following signal: repeated at brief intervals. As soon as a station perceives the signal of distress it shall cease all cor- respondence and not resume it until after it has made sure that the cor- respondence to which the call for assistance has given rise is terminated. In case the ship in distress adds at the end of the series of her calls the call letters of a particular station the answer to the call shall be incumbent upon that station alone. If the call for assistance does not specify any par- ticular station, every station perceiving such call shall be bound to answer it. XVII. 1. The call letters following the letters " P R B " signify that the vessel or station making the call desires to com- municate with the station called by means of the International Signal Code. The combination of the letters " P R B " as a service signal for any other purpose than that specified above is prohibited. 2. Wireless telegrams may be framed with the aid of the International Signal Code. Those addressed to a wireless telegraph station with a view to being for* warded by it are not to be translated by such station. 3. METHOD OF CALLING WIBELESS STATIONS AND TRANSMISSION OF WIRELESS TELEGRAMS. XIX 1. As a general rule, it shall be the shipboard station that calls the coastal station. 2. The call should be made, as a general rule, only when the distance of the vessel from the coastal station is less than 75 per cent of the normal range of the latter. 3. Before proceeding to a call, the station on shipboard shall adjust its receiving apparatus to its maximum sensibility and make sure that the coastal station which it wishes to call up is not in correspondence with any other station. If it finds that any transmission is in progress, it shall wait for the first pause. 4. The shipboard station shall use for calling the normal wave of the coastal station. XX. 1. The call shall comprise the signal. the call letters of the station called repeated three times, the word " from " (" de ") followed by the call letters of the sending station repeated three times. 2. The called station shall answer by making the signal followed by the call letters of the corresponding station repeated three times, the word " from," its own call letters, and the signal APPENDICES. 197 XXIV. Before beginning the exchange of correspondence the coastal station shall advise the shipboard station whether the transmission is to be effected in the alternate order or by series (Article XVIII); it shall then begin the transmission or follow up the preliminaries with the signal (invitation to transmit). XXV. The transmission of the wireless telegram shall be preceded by the signal and terminated by the signal followed by the name of the sending station. XXVI. When a wireless telegram to be transmitted contains more than 40 words, the sending station shall interrupt the transmission after each series of about 20 words by an interrogation point and shall not resume it until after it has obtained from the receiving station a repetition of the last word duly received, followed by an interrogation point. In the case of transmission by series, acknowledgment of receipt shall be made after each wireless telegram. XXVII. 1. When the signals become doubtful every possible means shall be re- sorted to to finish the transmission. To this end the wireless telegram shall be repeated at the request of the receiving station, but not to exceed three times. If in spite of such triple repetition the signals are still unreadable the wireless telegram shall be canceled. If no acknowledgment of receipt is received the transmitting station shall again call up the receiving station. If no reply is made after three calls the transmission shall not be followed up any further. XXVIII. All stations are bound to carry on the service with as little expense of energy as may be necessary to ensure safe communication. 4. ACKNOWLEDGMENT OF RECEIPT AND CONCLUSION OF WORK. XXIX. 1. Receipt shall be acknowledged in the form prescribed by the Inter- national Telegraph Regulations, preceded by the call letters of the trans- mitting station and followed by those of the receiving station. 2. The conclusion of a correspondence between two stations shall be indi- cated by each station by means of the signal followed by its call letters. 198 APPENDICES. APPENDIX D. [PUBLIC No. 262.] [S. 7021.] An Act to require apparatus and operators for radio-communication on certain ocean steamers. Be it enacted, by the Senate and House of Representatives of the United States of America in Congress assembled, That from and after the first day of July, nineteen hundred and eleven, it shall be unlawful for any ocean- going steamer of the United States, or of any foreign country, carrying passengers and carrying fifty or more persons, including passengers and crew, to leave or attempt to leave any port of the United States unless such steamer shall be equipped with an efficient apparatus for radio-communica- tion, in good working order, in charge of a person skilled in the use of such apparatus, which apparatus shall be capable of transmitting and receiving messages over a distance of at least one hundred miles, night or day: Provided, That the provisions of this Act shall not apply to steamers plying only between ports less than two hundred miles apart. SEC. 2. That for the purpose of this Act apparatus for radio-communi- cation shall not be deemed to be efficient unless the company installing it shall contract in writing to exchange, and shall, in fact, exchange, as far as may be physically practicable, to be determined by the master of the vessel, messages with shore or ship stations using other systems of radio-communi- cation. SEC. 3. That the master or other person being in charge of any such vessel which leaves or attempts to leave any port of. the United States in violation of any of the provisions of this Act shall, upon conviction, be fined in a sum not more than five thousand dollars, and any such fine shall be a lien upon such vessel, and such vessel may be libeled therefor in any district court of the United States within the jurisdiction of which such vessel shall arrive or depart, and the leaving or attempting to leave each and every port of the United States shall constitute a separate offense. SEC. 4. That the Secretary of Commerce and Labor shall make such regulations as may be necessary to secure the proper execution of this Act by collectors of customs and other officers of the government. Approved, June 24, 1910. APPENDIX E. WIRELESS TELEGRAPH STATION ROUTINE FOR UPKEEP OF STATION OUTFIT. DAILY. Wipe off all instruments with care. Tighten contacts of receivers. Clean commutators and collector rings. Clean zinc oxide from zinc spark points, if fitted. Blow water out of air lines. Fill cylinder oil cup and lubricate governor. In winter, tend heating apparatus carefully to prevent freezing of water in cylinders, pipes, etc., and keep oil fluid if necessary. APPENDICES. 199 WEEKLY. Rub down slate panels and instrument cases, examine contacts on panels, and vaseline moving contacts lightly. Blow out armatures and fields of motor-generators, generators, and motors. Lubricate chains on engines. Clean bushings and exterior of transformers or induction coils. Wipe off glass of condenser jars in air and clean contacts if necessary. Clean jar rack. Pump up compressed air condensers, if installed. Clean and polish inductances and exposed leads of transmitter. Clean thoroughly and set up all contacts of transmitter with care. Clean and polish spark gap. Polish key. Polish wood, metal, and rubber of receiver. Vaseline lightly the contacts of receiver switch and aerial switch if fitted, after cleaning. Clean lightning switch and vaseline contacts lightly. Clean all strainers. Lubricate pistons of magnetic air valves and reducing valves. Lubricate cylinders and bearings. Lubricate working parts of valves in pipe lines and operate same. MONTHLY. Make cadmium tests of storage battery, if installed. Clean oil injection nozzles. Pack stuffing boxes of valves in pipe lines. Clean and tighten contacts of ground where accessible. SEMIANNUALLY. Change oil of motor-generators, motors, and generators. Refit and line bearings of same. Empty oil storage tank and clean gauze strainer. Dismount and clean oil tubes of lubricating system. Dismount and seat check valves. Dismount and clean tubes of feed oil distribution system. Renew asbestos packing of oil pump. Clean port openings, combustion spaces, exhaust ports, joint screws, and jackets of cylinders, and renew gaskets. Dismount Leyden jar condenser and clean thoroughly. Lower aerial, wipe off insulators, oil blocks, overhaul halliards, and renew same when necessary. Polish hard rubber of receiver, etc., using bisulphide of carbon. 200 APPENDICES. APPENDIX F. RESUSCITATION FKOM APPARENT DEATH FROM ELECTRIC SHOCK. BY ATJGUSTIN H. GOELET, M. D. The urgent necessity for prompt and persistent efforts at resuscitation of victims of accidental shocks by electricity is very well emphasized by the successful results in the instances recorded. In order that the task may not be undertaken in a half-hearted manner, it must be appreciated that accidental shocks seldom result in absolute death unless the victim is left unaided too long, or efforts at resuscitation are stopped too early. In the majority of instances the shock is only sufficient to suspend anima- tion temporarily, owing to the momentary and imperfect contact of the con- ductors, and also on account of the resistance of the body submitted to the influence of the current. It must be appreciated also that the body under the conditions of accidental shocks seldom receives the full force of the current in the circuit, but only a shunt current, which may represent a very insig- nificant part of the whole. When an accident occurs, the following rules should be promptly executed with care and deliberation: 1. Remove the body at once from the circuit by breaking contact with the conductors. This may be accomplished by using a dry stick of wood, which is a nonconductor, to roll the body over to one side, or to brush aside a wire, if that is conveying the current. When a stick is not at hand, any dry piece of clothing may be utilized to protect the hand in seizing the body of the victim, unless rubber gloves are convenient. If the body is in contact with the earth, the coat tails of the victim, or any loose or detached piece of cloth- ing may be seized with impunity to draw it away from the conductor. When this has been accomplished observe rule 2. The object to be attained is to make the subject breathe, and if this can be accomplished and continued he can be saved. 2. Turn the body upon the back, loosen the collar and clothing about the neck, roll up a coat and place it under the shoulders, so as to throw the head back, and then make efforts to establish respiration (in other words, make him breathe), just as would be done in case of drowning. To accomplfsh this, kneel at the subject's head, facing him, and seizing both arms draw them forcibly to their full length over the head, so as to bring them almost to- gether above it, and hold them there for two or three seconds only. (This is to expand the chest and favor the entrance of air into the lungs.) Then carry the arms down to the sides and front of the chest, firmly compressing the chest walls, and expel the air from the lungs. Repeat this maneuver at least sixteen times per minute. These efforts should be continued unremit- tingly for at least an hour, or until natural respiration is established. 3. At the same time that this is being done, some one should grasp the tongue of the subject with a handkerchief or piece of cloth to prevent it slip- ping, and draw it forcibly out when the arms are extended above the head, and allow it to recede when the chest is compressed. This maneuver should likewise be repeated at least sixteen times per minute. This serves the double purpose of freeing the throat so as to permit air to enter the lungs, and also, by exciting a reflex irritation from forcible contact of the under part of the tongue against the lower teeth, frequently stimulates an involuntary effort at respiration. To secure the tongue if the teeth are clenched, force the jaws apart with a stick, a piece of wood, or the handle of a pocket knife. APPENDICES. 201 4. The dashing of cold water into the face will sometimes produce a gasp and start breathing, which should then be continued as directed above. If this is not successful the spine may be rubbed vigorously with a piece of ice. Alternate applications of heat and cold over the region of the heart will accomplish the same object in some instances. It is both useless and unwise to attempt to administer stimulants to the victim in the usual manner by pouring them down his throat. While the above directions are being carried out, a physician should be summoned, who, upon his arrival, can best put into practice rules 5, 6, and 7, in addition to the foregoing, should it be necessary. FOR THE PHYSICIAN SUMMONED. 5. Forcible stretching of the sphincter muscle controlling the lower bowel excites powerful reflex irritation and stimulates a gasp (inspiration) fre- quently when other measures have failed. For this purpose, the subject should be turned on the side, the middle and index fingers inserted into the rectum, and the muscle suddenly and forcibly drawn backward toward the spine. Or, if it is desirable to continue efforts at artificial respiration at the same time, the knees should be drawn up and the thumb inserted for the same purpose, the subject retaining the position on the back. 6. Rhythmical traction of the tongue is sometimes effectual in establishing respiration when other measures have failed. The tongue is seized and drawn out quickly and forcibly to the limit, then it is permitted to recede. This is to be repeated 16 times per minute. 7. Oxygen gas, which may be readily obtained at a drug store in cities or large towns, is a powerful stimulant to the heart if it can be made to enter the lungs. A cone may be improvised from a piece of stiff paper and attached to the tube leading from the tank, and placed over the mouth and nose while the gas is turned on during the efforts at artificial respiration. INDEX. A ART. Absorption 125, 126 Accumulator 4 Adjustments 197 Aerial 113, 168-172 Air, Atmospheric pressure of 71 Dielectric strength of 151 Amber 1, 12 Ammeter 86 Calibration of 211 Hot wire 174, 19J9, 202, 210 Ampere ,. . ., 80, 93 Definition of 84 International standard 86 Ampliphone 191 Amplitude 55 Decrease of 125 Anode 3,184 Antenna. (See aerial.) Armature 37 Austin 99, 100, 125, 179, 184, 206 B Battery, Primary 3 Solution 3 Storage 4 Bellini-Tosi 226 Buzzers, Testing 189 C Calculations, Basis of 13 Calibration 197 Capacity 44, 92, 96 Concentrated 75 Distributed 75 In sending sets 1"45 Measurement of 107, 213 Notation of 89 Of straight wires 104 Relation to self-induction in long wire 59 Specific inductive 46 What it depends on 47 Care and operation 214 14 '204: INDEX. ART. Cathode 3,184 Cell, Parallel 5 Primary, illustration of v 3 Secondary 4 Series 5 Standard 86 Storage 4 Centimeter 79 Charges 1, 11 How created 13 Inertia of 53 Signs of 1 Circuits, Closed 56, 75, 138 Direct connected 75 Grounded 113 Inductively connected 75 Looped 180 Non-radiating 112 Open .75, 168 Oscillating 58, 60, 75, 121 Radiating 75, 110, 112 Stiff 182, 207 Circular mil 91 Codes .216, 217 Coherer 178, 188 Coil, Coupling 174 Loading 174 Variometer 174 Commutator 36 Condenser-s *. . 45 Connections of 146 Conventional signs for 50 Discharge of 51, 52, 53, 54, 123 Discussion of 48, 148 Fixed 50 In parallel ,. 105 In series 105 Intermittent use of , 124 Kinds of 147, 181 Leyden jars 50 , Material of 148 Non-oscillatory discharge of 94 Oscillatory discharge of 51, 52 Variable 50 Conductors, Definition of 2 Opaque to electric waves 68 Coulomb 87, 92 INDEX. 205 ART. Counterpoise 176 Coupling, Close 109 Coefficient of 109 Loose 109 Of electric circuits 109 Percentage of 110 Perfect 109 Current-s 3, 80, 92 Alternating 33, 101 Determination of direction of 31 Direct 36, 101 Direction of 9 Electric 3 Induced, direction of 14, 31 Interrupted 35 Loop 69 Received, measurement of 212 Node 69 Production of by cutting lines of force 32, 33 Pulsating 36 Rectified 36 Cycle 55 Cymometers 107 D Damping 55, 60 Formula for 117 Measurement of 208 DeForest 179, 182 Detectors ...'..! 178 Crystal 185 Electrolytic 184 Lodge-Muirhead 188 Magnetic 187 Primary cell 184 Rectifying 185 Vacuum tube 186 Dielectrics 2, 46 Rupture of 47 Strains in 47 Strength of 149 Table of 150 Direction finders -. 226 Direction senders 227 Duddell & Taylor 125 Dynamo-s 36 Building up of 38 Direct current 39 . Self -exciting 36, 39 Dyne 79 206 INDEX. E ART. Earth quadrant 88 Electric induction 46 Electricity 1 Dynamic 42 Origin of word * 1 Relation between it and magnetism 16 Source of 31 Frictional 1 Static f 1 Voltaic 42 Electrification, Duration of 44 Limit of 47 What it consists of 47 Electrode 3,184 Electrolyte 3, 85, 184 Electro-magnet 8 Illustration of 8 Electro-magnetic induction 12 Illustration of > 15 Methods of producing current by 31 Electro-magnetism 8 Electro-motive force 3, 29, 37, 80, 84, 92 Electrons 66 Electroscope 1, 12, 45 Element, Conventional sign for 5 Negative 3 Positive 3 Energy, Forms of 78 Laws of . 78 Non-returnable 128 Returnable 128 Storage of , 45 Transfer of 159 Erg 79 Ether 18 Compression of 18 Movements of 20, 43 Strain of 18 Stretch of 18 Ether waves 64 Absorption of 68, 71, 125, 126 Attenuation of 71 Detection of 71, 127 Diffraction of 68, 73 Earthed 113, 114 Formation of 66 Free 113 Guided . 71 INDEX. 207 Ether waves continued. AKT. Interference of 73 Lengths of 64, 67, 165 Measurement of 199-210 Method of changing period 106 Production of 74 Radiation of 112, 113 Reflection of 68, 71 Refraction of 68, 72 Velocity of 66 Exciter 39 F Farad 87, 93 Fessenden 99, 100, 178, 223 Interference preventer 179 Field-s, Analogy between 49 Electric 17, 18 Electro-static 113 Magnetic 17,18 Magnetic, strength of 84 Fleming 99 Foot-pound 78 Force, Definition of 78 Field of 7 Magnetising 47 Magneto-motive , 47 Frequency 59 G Galvanometer 35, 44, 199 Generator-s 41 Frequency of 130 Sending, description of 129 Gram 79 Gravity, Force of 79 Ground 175, 176 Connection and lead 175 H Heat, Definition of - 67 Velocity of 19 Henry 88, 93 Hertz 65 Hertzian waves 65 (See also ether waves.) 208 INDEX. ART. Horse-power 93 Hysteresis, Dielectric 108 Magnetic .108 Impedance 30, 101 Inductance-s 61 Fixed 62 Forms of 108, 138, 173, 174, 182 Measurement of 107, 213 Variable 62 Induction, Coil 27 Electric 46 Electro-magnetic 12, 14 Magnetic -. 47 Mutual 15, 109 Self 30, 95 Total 57 Inertia, Electro-magnetic 54 Installation 193 Insulators 2 Transparent to electric waves 68 J Joule, Definition of 84 K Keys, Break 137 Sending, types of 137 .L Length, Electrical 30, 59 Leyden jar. (See condensers.) Light, Definition of '. 19 Velocity of 19 Lines, Of force, movement of 14, 28 Of force, negative direction of 7, 9 Of force, positive direction of 7, 9 Of force, used as basis for electric measurements 7 Loading coils 179 Loops 68 INDEX. 209 M ART. Magnetic induction 14 (See also induction, magnetic.) Magnetism 6 Magnetization, Limit of 47 What it consists of 47 Marconi 158, 178, 227 Martin 158 Measurement, Of coupling. (See wave meters.) Of damping 208 Of inductance and capacity 213 Of received currents 212 Of sending currents. (See ammeter, hot wire.) Of wave lengths 199 to 205 Megohm 91 Mercury turbine interrupter 129 Microfarad 89,91,93 Micron ^ 91 Microsecond 91 Mil 91 Milliampere 91 Millihenry 88, 91, 93 Millivolt 91 Motor 41 Motor generator 41 N Nonconductors 2, 45 Nodes . 68 O Octave 129 Ohm 80, 93 Definition of 84 International standard of 86 Legal 86 Ohm's law, Deduction of 82 Ondameter 107 Operating room 193 Oscillating circuit. (See circuit, oscillating.) Oscillation-s, Damped 116 Decrement of 117 Definition of 55 Undamped 118, 218 to 225 Oscillator , . 112 210 INDEX. P ART. Permeability, Electric, of air 46 Magnetic, of air, of iron 25 Of insulators , 46 Pierce 104, 121, 178, 181 Pitch 129 Poles 3 Magnetic 6 North and south 6 Positive and negative 3 Portable wireless 228 Potential 87 Effect on work, period and amplitude 58 Illustration of 1, 3, 44 Loop 69 Node 69 Potentiometer v 178 Power 78, 92, 98 Electric, basis of .' 21 Equation of 83 Protective devices . 195 Quenched spark. (See spark, quenched.) R Reactance 30, 101 Reactance regulator 101, 135 Receiving apparatus 100, 177, 225 Receiving circuits 178 Coupling of '. . : . . . 180 Relay 137, 191 Resistance 80, 92 Illustration of 29 Inductive 61 Non-inductive 61 Radiation 115 Standard 86 Resonance 63, 107, 198 S Saturation, Of iron 38, 47 Self-induction 30, 92 (See also induction, self.) Notation of 88 Of straight wires 104 Sending apparatus 129 Direct connected 139 Efficiency of 99 Frequency of ,'. 130 INDEX. Sending apparatus continued. . ART. Inductively connected 140 Regulation of 136 Sending circuits, Types of 138 Sending helix 108 Shoemaker 179, 184 Skin effect ." 103 Solenoid, Core of 25 Illustration of 8 Solution, Battery 3 Spark, Quenched 157 Spark gap 54, 60, 74, 121 Cooling of 132 Function of 153 Resistance of 153 Rotating 134, 155 Safety 135, 195 Types of 152 Specific inductive capacity 46 Table of 150 Standard jar 93 Capacity of 94 Static discharges 184, 230 Stone 137, 178, 181 Switch, Lightning : 174 Multiple 196 T Telefunken 141, 182 Telegrams, Work done in sending 124 Telegraphy, Wireless, fundamental principle of 76 Telephones 131, 190 Time constant 102 Transformer-s 74, 135 Air core 27 Auto 27 Closed core 27 Illustration of '. 27 Open core 27 Oscillation 75 Primary winding 27, 135 Rotary 41 Secondary winding 27, 135 Stationary 41 Step down 27 212 INDEX. Transformer-s continued. ART. Step up . . . 21 Voltage of 135 Tuners, Pancake , 182 Tuning 75, 140, 182, 197 Tuning forks 129 U Units, Absolute 80 'Arbitrary 79 Electro-magnetic ." 80 Electro-static 89 Fundamental 78 Practical 80 Theoretical 84 V Variometer 141, 182 Vibrations, Electrical 56 Mechanical '. 56 Volt 80, 93 Definition of 84 Standard 86 Volt meter 86 W Watt 93 Definition of 84 Wave lengths, Standard " V. . 216 Limitations of 165 Wave meters 107, 199-210 Pierce 203 Donitz 205 Waves. (See ether waves.) Wave trains 119, 120, 121, 163 Decrement of 164 Length of 164 Whorls, Electric 113 Wien 157 Wire, rat tail 174 Wireless telegraphy, Definition of 75 Equations 98 Fundamental equation of 94 Fundamental principle of 76 Wireless telephony 218, 219, 220 Work 46, 78, 95, 96, 100, 124