RADIODYNAMICS 
 
 THE WIRELESS CONTROL OF TORPEDOES 
 AND OTHER MECHANISMS 
 
 BY 
 
 B. F. MIESSNER 
 
 Associate Member Institute of Radio Engineers, 
 Expert Radio Aide, U. S. Navy 
 
 112 ILLUSTRATIONS 
 
 NEW YORK 
 
 D. VAN NOSTRAND COMPANY 
 
 25 PARK PLACE 
 
 1916 
 
COPYRIGHT* 1916. 
 
 BY 
 D. VAN NOSTRAND COMPANY 
 
 Stanbopc flbreas 
 
 F. H.GILSON COMPANY 
 BOSTON, U.S.A. 
 
PREFACE 
 
 IN the preparation of this work the author has endeavored 
 to present in an orderly and instructive fashion the most 
 important material concerning the history, methods and 
 apparatus of Radiodynamics, the art of controlling distant 
 mechanisms without artificial connecting means. 
 
 He has aimed especially at a treatment of his subject-matter 
 that would be intelligible to the general reader without sacri- 
 ficing the technical exactitude which makes scientific work of 
 value to the trained engineer. 
 
 The chief recent developments in this new art have been 
 of a military nature, and for this reason the volume is devoted 
 for the most part to the torpedo-control applications of 
 Radiodynamics. 
 
 It is hoped that the book may prove interesting to the gen- 
 eral scientific reader, as well as to the trained engineer, and 
 to those concerned in the purely military applications and 
 possibilities of wirelessly-controlled mechanisms. 
 
 The author desires here to thank the many friends who 
 have generously assisted him in collecting his materials. He 
 desires especially to express his obligation to Professor M. H. 
 Liddell, of Purdue University for his advice and assistance in 
 the preparation of the book for press. 
 
 B. F. M. 
 
 LAFAYETTE, IND., 
 August, 1916. 
 
 in 
 
 O o n /t *t o 
 
CONTENTS 
 
 CHAPTER PAGE 
 
 I. HISTORICAL i 
 
 i IT. WIRELESS CONTROL OF MECHANISMS 6 
 
 III. PRACTICAL WIRELESS TELEGRAPHY 12 
 
 IV. ELECTROSTATIC AND COMBINED INDUCTION CONDUCTION TELE- 
 
 GRAPH SYSTEMS 19 
 
 V. ELECTROMAGNETIC WAVE SYSTEMS 27 
 
 VI. POSSIBLE CONTROL METHODS FOR RADIODYNAMICS SOUND 
 
 WAVES 33 
 
 VII. INFRA-RED OR HEAT WAVES 41 
 
 VIII. VISIBLE AND ULTRA-VIOLET WAVES 57 
 
 IX. EARTH CONDUCTION 67 
 
 X. ELECTROSTATIC AND ELECTROMAGNETIC INDUCTION HERTZIAN 
 
 WAVES 74 
 
 XI. THE ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 78 
 
 XII. SELECTORS 89 
 
 XIII. EUROPEAN CONTROL SYSTEMS . 92 
 
 XIV. WORK OF THE HAMMOND RADIO RESEARCH LABORATORY 107 
 
 XV. THE SOLUTION OF THE PROBLEMS RELATED TO BATTLE-RANGE 
 
 TORPEDO CONTROL 124 
 
 XVI. THE DIFFICULTIES ENCOUNTERED IN PROVIDING PROTECTION 
 
 FROM INTERFERENCE 137 
 
 XVII. A MEANS OF OBTAINING SELECTIVITY 145 
 
 XVIII. NATURE OF INDICATOR CURRENTS IN RADIO RECEIVERS 150 
 
 XIX. THE INTERFERENCE PREVENTER 159 
 
 XX. DETECTORS 167 
 
 XXI. METHODS OF INCREASING RECEIVED EFFECTS 175 
 
 XXII. RELAYS 180 
 
 XXIII. TORPEDO ANTENNAE 183 
 
 XXIV. RECENT DEVELOPMENTS. . . 188 
 
RADIODYNAMICS 
 
 CHAPTER I 
 HISTORICAL 
 
 From earliest times methods of conveying intelligence to a 
 distance have been universally known and utilized. Fleet- 
 footed runners, fires and torches by night, and smoke by day 
 as well as acoustic methods using both air and earth as con- 
 ducting mediums seem to have been among the first means of 
 comparatively distant signalling. We read of them in the 
 Bible (Jeremiah) and in the Greek and Latin authors; their 
 use in the far East and in Europe leaves no doubt as to their 
 wide employment amongst civilized nations. 
 
 The Indians of America from the North to Cape Horn still 
 use lighted fires and blanket-controlled smoke clouds to an- 
 nounce special tidings and convey important messages; their 
 system of optical signalling in which the arms were used, 
 furnished the basis for the semaphore, which toward the end 
 of the eighteenth century came into general use in Europe. 
 It may still be seen on any railroad. The semaphore system 
 was still further elaborated for maritime and military pur- 
 poses and today in the armies and navies of the world we 
 have semaphore and flag signalling as a very important means 
 of communicating intelligence to distances not in excess of a 
 few miles. The heliograph by day and the electric search- 
 light by day and night can both trace their evolution to the 
 primeval fire and torch. 
 
$, ; I/I y * ^ RApIODYNAMICS 
 
 The application of electricity has revolutionized all pre- 
 vious methods of signalling. The phenomenon of attraction 
 was well known to the ancients. Thales, the founder of 
 Ionic philosophy, who lived six hundred years before Christ, 
 noticed the effects of friction on amber, and Theophrastus, 
 Pliny and other writers recorded similar phenomena. 
 
 In 1727 Stephen Gray, a pensioner of the Charter House, 
 London, made an electric discharge pass over a circuit of 
 700 feet. Shortly after the discovery of the Leyden jar by 
 Muschenbroek of Leyden, in 1746, Dr. Watson, a bishop of 
 Llandaff, transmitted a charge through 2800 feet of wire. 
 In the same year he increased the distance of transmission to 
 10,600 feet through wires stretched on poles erected on 
 Shooter's Hill, London. Benjamin Franklin made similar ex- 
 periments in 1748 over the Schuylkill river at Philadelphia. 
 Le Sage of Geneva established the first telegraph system for 
 the transmission of intelligible signals in 1774*; this system 
 was based on electrostatic action. The next important law 
 was discovered by Romagnesi of Trente in 1805; but at- 
 tracted little attention until it was rediscovered in 1819 by 
 Oersted. This discovery showed that a current-carrying wire 
 is able to deflect a magnetic needle. Schweigger in 1820 
 discovered that the deflecting force was increased by winding 
 the wire several times around the needle. These very im- 
 portant discoveries paved the way for the galvanoscope and 
 galvanometer. Galvanoscopic or needle telegraphs were sub- 
 sequently evolved. 
 
 In 1832 Schilling, a Russian, devised a single-needle tele- 
 graph using reverse currents and combinations of signals for 
 an alphabet. Schilling's telegraph was developed by Gauss 
 and Weber, who built a line three miles long at Gottingen. 
 
 While Prof. A. C. Steinheil of Munich was establishing a 
 system of telegraphy in Bavaria, Gauss, the celebrated German 
 * Moigno's " T616graphie Electrique," p. 59. 
 
HISTORICAL 3 
 
 philosopher and himself a telegraph inventor, suggested to 
 him that the two rails of a railway might be used as telegraph 
 conductors. In July, 1838, Steinheil tried the experiment on 
 the Nurmberg-Furth railway, but was unable to obtain an 
 insulation of the rails sufficiently good for the current to 
 reach from one station to the other. The great conductiv- 
 ity with which he found that the earth was endowed led 
 him to presume that it would be possible to employ it instead 
 of the return wire or wires hitherto used. The trials that he 
 made in order to prove the accuracy of this conclusion were 
 followed by complete success, and he then introduced into 
 electric telegraphy one of its greatest improvements the 
 earth return circuit.-* 
 
 Following Sturgeon's invention of the electromagnet in 
 1825 and the simultaneous discovery by Faraday in England 
 and Henry in America (1831) of the laws of electromagnetic 
 induction, Morse laid the foundations in 1836 of the present 
 overland electromagnetic telegraph system. In the same 
 year in England Wheatstone with W. F. Cooke still further 
 perfected the needle telegraph and a year later put a crude 
 system of telegraphy into actual service on the London and 
 Black well Railway. In 1839 the first public line was opened 
 by Wheatstone between Paddington and Slough, England, 
 twenty miles of goose quills being used for insulation. 
 
 It was once supposed that Wheatstone was the original 
 inventor of the electric telegraph, but strictly speaking it had 
 no inventor; it is rather the result of accumulated discoveries 
 each adding its quota to advance the invention towards per- 
 fection. The greatest achievement of Wheatstone was his 
 automatic, recording telegraph, which is extensively used for 
 press and other long dispatches and which has attained 
 marvelous speeds for a mechanical recorder. 
 
 * For an account of the earth return before 1838 see Fahie's "History of 
 Electric Telegraphy to the Year 1837," pp. 343-348. 
 
4 RADIODYNAMICS 
 
 Morse constructed his electromagnetic telegraph in 1836, 
 and in the next few years he devised many important modi- 
 fications. Congress made him an appropriation of $30,000 in 
 March, 1843, an d on the 24th of March, 1844, the first tele- 
 graph line in the United States was successfully opened between 
 Washington and Baltimore, a distance of about 40 miles. 
 
 The electrostatic telegraph of Le Sage was probably the 
 first instance of the control of mechanisms from a distance 
 by the use of conducting wires. The real art of teledy- 
 namics,* however, is based on the discoveries by Romagnesi, 
 Oersted, and Schweigger of the phenomena of electromagnet- 
 ism which led up to the conception and development of the 
 electromagnetic telegraph. Since 1836, when Morse con- 
 structed his first telegraph, no very radical changes have been 
 made in the general scheme on which his system was based, 
 but it has been gradually and surely developed and brought to 
 the present stage of perfection. One very conspicuous change 
 in detail, however, is worthy of mention. The electromagnetic 
 sounder first used by Morse on the line between Washington 
 and Baltimore and exhibited in the National Museum in 
 Washington weighed one hundred eighty-five pounds. The 
 arms were three and one-half inches in diameter and eighteen 
 inches long, the same size of copper wire being used for the 
 coils as for the line wire. It was then supposed that the 
 wire of the coils and of the line should be of the same size 
 throughout, and even down to 1860 many practical telegraph- 
 ers held this view.f The sounders now used weigh about 
 one pound and require no more than about seventy-five 
 cubic inches of space. The coils are wound with wire much 
 smaller than the line wire, a great increase in sensitiveness 
 being thereby produced. 
 
 * The art of controlling mechanisms from a distance; as used here it refers 
 only to distant control, by electrical means, with or without connecting wires, 
 t London Electrical Review, Aug. 9, 1895, p. 157. 
 
HISTORICAL 5 
 
 The necessity for long-distance telegraphy brought about 
 the invention of the relay, a very sensitive form of sounder 
 which is actuated by the weak line currents and which in 
 turn controls the current for operating the sounder used in 
 receiving messages. The relay is a very important part of 
 all systems for the distant control of mechanisms as by its 
 use practically any amount of power can be controlled. The 
 mere pressure of the finger on a telegraph key through which 
 a few thousandths of an ampere flow to a distant relay is 
 sufficient to start or stop the most powerful machinery or to 
 set off explosive charges strong enough to destroy a whole 
 city. 
 
 Such mechanisms as electric bells and signals of various 
 kinds, telephone and fire alarm systems, electric clocks and 
 chimes, and time distribution systems are all developments 
 in the art of teledynamics. Present-day automatic tele- 
 phone systems, the distant control of searchlights, and the 
 wire-controlled torpedo are examples of the wonderful possi- 
 bilities along these lines. 
 
CHAPTER II 
 WIRELESS CONTROL OF MECHANISMS 
 
 Like most wonderful inventions the telegraphic transmission 
 of signals without the aid of conducting wires is in reality 
 not an invention, according to the popular conception of the 
 word, but rather the result of the combined efforts of many 
 deep-thinking scientific men extending over a period of many 
 years. After the discovery of the galvanic current and elec- 
 tromagnetism in the seventeenth and eighteenth centuries 
 the conception and development of wireless telegraphy and 
 wire telegraphy occurred at practically the same time. It 
 was in 1836 that Morse constructed his first telegraph; this 
 was not put into practical operation until 1844. I n I ^38 
 Steinheil of Munich, one of the great pioneers of electric 
 telegraphy in Europe, gave the first intelligent* suggestion 
 of a wireless telegraph. In a paper on this subject Steinheil, 
 explaining his theories and observations on earth conduction 
 telegraphy, says: 
 
 "The inquiry into the laws of dispersion according to which 
 the ground, whose mass is unlimited, is acted upon by the 
 passage of a galvanic current appeared to be a subject replete 
 with interest. The galvanic excitation cannot be confined 
 
 * Earlier but vague and impractical suggestions were made previous to this 
 time. In the Bible we find: " Canst thou send lightnings, that they may go, 
 and say unto thee, ' Here we are ? " : 
 
 In 1632 Galileo wrote of a secret art by which it would be possible to con- 
 verse across a space of several thousand miles through the attraction of a 
 magnetic needle ("Galilei Systema Cosmicum." Dial. I). The "Prolusiones 
 Academicae" of Strada, which was published in 1617, described a method of 
 communicating at a distance by means of two pivoted magnetic needles. 
 
 6 
 
WIRELESS CONTROL OF MECHANISMS 7 
 
 to the portions of earth situated between the two ends of the 
 wire; on the contrary it cannot but extend itself indefinitely 
 and it therefore only depends on the law that obtains in this 
 excitation of the ground, and the distance of the exciting ter- 
 minations of the wire, whether it is necessary or not to have 
 any metallic communication at all for carrying on telegraphic 
 intercourse. 
 
 "An apparatus can it is true be constructed in which the 
 inductor, having no other metallic connection with the 
 multiplier than the excitation transmitted through the ground, 
 shall produce galvanic currents in that multiplier sufficient 
 to cause a visible deflection of the bar. This is a hitherto 
 unobserved fact and may be classed amongst the most ex- 
 traordinary phenomena that science has revealed to us. It 
 only holds good, however, for small distances; and it must be 
 left to the future to decide whether we shall ever succeed in 
 telegraphing at great distances without any metallic con- 
 nection at all. My experiments prove that such a thing is 
 possible up to distances of fifty feet. For greater distances 
 we can only conceive it feasible by augmenting the power of 
 the galvanic induction, or by appropriate multipliers con- 
 structed for the purpose, or, in conclusion, by increasing the 
 surface of contact presented by the ends of the multipliers. 
 At all events the phenomenon merits our best attention, and 
 its influence will not perhaps be altogether overlooked in the 
 theoretic views we may form with regard to galvanism 
 itself." * 
 
 Discussing the same subject in another publication Stein- 
 heil says: "We cannot conjure up gnomes at will to convey 
 our thoughts through the earth, Nature has prevented this. 
 The spreading of the galvanic effect is proportional not to 
 the distance of the point of excitation but to the square of 
 this distance; so that at the distance of fifty feet only ex- 
 * Sturgeon's "Annals of Electricity," vol. in, p. 450. 
 
8 RADIODYNAMICS 
 
 ceedingly small effects can be produced by the most powerful 
 electrical effect at the point of excitation. Had we means 
 which could stand in the same relation to electricity as the 
 eye stands to light nothing would prevent our telegraphing 
 through the earth without conducting wires; but it is not 
 probable that we shall ever attain this end." * 
 
 Steinheil apparently received his inspiration for this method 
 of transmitting signals from his accidental discovery of the 
 conductivity of the earth in the experiments on the Nurm- 
 berg-Fiirth railroad. His explanation, which is somewhat 
 nebulous and obscured by such expressions as "multipliers," 
 
 \ \ 
 
 ^ \ 
 
 ---- " /rth* 
 
 FIG. i. 
 
 " galvanic excitation," and "galvanic induction," actually 
 amounts to this: When two earthed conducting plates are 
 connected to an electric battery, current flows through the 
 earth, but not wholly through that portion directly between 
 the plates. Instead, the current obeys Ohm's law with re- 
 gard to a circuit including conductors in parallel, i.e., the 
 current in any branch is inversely proportional to its resist- 
 ance. The number of parallel branches in the earth circuit 
 is infinite, but they obey this same law. The earth between 
 the buried plates, although having a high specific resistance, 
 has a very great cross sectional area; this accounts for the 
 relatively low resistance of earth returns. The current 
 
 * The electric eye of Hertz! "Die Anwendung des Electromagnetismus," 
 1873, p. 172. 
 
WIRELESS CONTROL OF MECHANISMS 9 
 
 density, according to Ohm's law is greatest between the 
 plates, and decreases in proportion to the distance along any 
 line at right angles to the line joining the plates. This is 
 shown in Fig. i. 
 
 SteinheiPs scheme was to so place another set of earth plates 
 connected by a wire and current indicator that the current 
 
 would traverse the earth between ^ 
 
 the sending and receiving plates, C~ "~"| 
 
 as shown in Fig. 2, and thus * y .y**' 
 
 operate the receiving instrument. I \Earth p/ates < 
 
 Steinheil's inability to signal t x' % X 
 
 over distances greater than fifty I J 
 
 feet was, no doubt, due to the '^7*^ indicator 
 
 limited capacity of his current FIG 2 
 
 supply, the insensitiveness of his 
 
 receiving indicator, and his probable ignorance of the fact that 
 
 the distance between the transmitting plates should be at least 
 
 three times the distance to be bridged, for the best results. 
 
 Another means of signalling without connecting wires was 
 disclosed by Steinheil in a classic paper on "Telegraphic 
 Communication, especially by means of Galvanism." This 
 method is particularly interesting because of its similarity 
 to the Photophone, invented by Alexander Graham Bell and 
 Sumner Tainter a half century later. Describing his idea 
 Steinheil says in part: " Another possible method of bringing 
 about transient movements at great distances, without any 
 intervening conductor, is furnished by radiant heat, when 
 directed by means of condensing mirrors upon a thermo- 
 electric pile.* A galvanic current is called into play, which in 
 its turn is employed to produce declinations of a magnetic 
 needle. The difficulties attending the construction of such 
 
 * In recent years thermo-piles have been developed to such an extent that 
 the heat radiated by stars can be detected and measured. W. W. Coblentz, of 
 the U. S. Bureau of Standards, has described, in various publications issued by 
 
10 RADIODYNAMICS 
 
 an instrument, although certainly considerable, are not in 
 themselves insuperable. Such a telegraph however would 
 only have this advantage over those semaphores based on 
 optical principles namely, that it does not require the con- 
 stant attention of the observer; but, like the optical one, it 
 would cease to act during cloudy weather, and hence partakes 
 of the intrinsic defects of all semaphoric methods." * 
 
 It is not probable that Steinheil ever worked this idea into 
 usable form, as no accounts of experiments can be found, 
 but to him is the credit really due for first (1839) suggesting 
 a means of signalling without conducting wires by the use of 
 radiant energy, and his was in all probability the first radio- 
 telegraphic system disclosed to the world. 
 
 Another way of conveying intelligence in a manner closely 
 related to those already given depends upon the sonorescent 
 property of substances. The voice-controlled transmitting 
 beam of light or heat is allowed to fall upon suitable material, 
 such, for instance, as a sheet of hard rubber. The periodic 
 expansion and contraction of this material, caused by the 
 periodic variations in the intensity of the heat imparted by 
 the beam, cause the rubber disc to reproduce the sounds made 
 near the transmitter. 
 
 Davy's Sound-relaying System 
 
 Edward Davy, in 1838, proposed a system of wireless 
 signalling, which, though not of any practical value, is worthy 
 of mention because the principle involved very closely re- 
 
 that institution, micro-radiometers which are sufficiently sensitive to detect the 
 heat of a standard candle at fifty-three miles. Edison's " tasimeter," which 
 he devised for studying the streamers of the sun during an eclipse, is reported 
 to have been so sensitive that a person at a distance of thirty feet could produce 
 a perceptible effect merely by turning his face toward the instrument. The 
 Crookes radiometer, the Duddell thermo-galvanometer, and the bolometer 
 bridge may also be used as detectors of radiant heat. 
 * Sturgeon's "Annals of Electricity," Mar., 1839. 
 
WIRELESS CONTROL OF MECHANISMS II 
 
 sembles our modern schemes of relaying, which are applied 
 in long-distance telegraphy and kindred branches of the art. 
 The energy is transmitted a short distance to a receiver 
 which responds, controls a local source of energy, and sends 
 the signal on in duplicate to the next station, this operation 
 being repeated a sufficient number of times to bridge the re- 
 quired distance. Davy, however, had in mind the conjoint 
 use of sound and electricity for accomplishing this end. His 
 plan was as follows: Stations placed about a mile apart should 
 be fitted with powerful means of producing sound waves to- 
 gether with suitable means, such as our common megaphones, 
 for directing them to the receiver and concentrating them 
 upon some delicate form of vibratory relay. This relay 
 would vibrate in resonance with the transmitted sound 
 waves, close the circuit for energizing a local means of sound 
 production similar to the first, and thus relay the signals on 
 to the next station. Obviously, such a system was impractical 
 in comparison with other ideas advanced at that time, 
 principally because of the numerous stations necessary to 
 bridge a relatively short distance, and the power required 
 at each of these to produce sound waves of sufficient ampli- 
 tude to operate the vibratory relay a mile away. John 
 Gardner of England has developed sensitive vibratory relays 
 with which he can control lights, motors, bells, etc., across 
 a large room by whistling the tone corresponding in fre- 
 quency to the natural period of the vibratory diaphragm or 
 reed.* 
 
 * For other references to this subject, see Signor Senliq d' Andres, Tele- 
 graphic Journal, vol. ix, p. 126; A. R. Sennet, Journ. Inst. Elec. Eng., No. 137, p. 
 908. See also U. S. Hydrographic Office Bulletin of May 13, 1914 on the " Fes- 
 senden Oscillator for the Detection of Icebergs," Professor Dayton C. Miller's 
 work with his '" Phonodik," described in his book on "The Science of Musical 
 Sounds " (Macmillan Co.), Tests on Fessenden Submarine Signalling Apparatus, 
 Journal U. S. Art. War, Apr. 1915; see also Sci. Am., July 4, 1914 and the 
 American Magazine, April, 1915. 
 
CHAPTER III 
 PRACTICAL WIRELESS TELEGRAPHY 
 
 The first experiments of importance with the new earth 
 conduction telegraphy appear to have been made by Professor 
 Morse, who, in 1842, actually transmitted signals a distance 
 of nearly a mile across the Susquehanna river.* 
 
 In a letter to the Secretary of the Treasury which was sub- 
 sequently laid before the House of Representatives, Morse 
 says: . , 
 
 "In the autumn of 1842, at the request of the American 
 Institute, I undertook to give to the public in New York a 
 demonstration of my telegraph, by connecting Governor's 
 Island with Castle Garden, a distance of a- mile; and for this 
 purpose I laid my wires properly insulated beneath the 
 water. I had scarcely begun to operate, and had received 
 but two or three characters when my intentions were frus- 
 trated by the accidental destruction of a part of my con- 
 ductors by a vessel, which drew them up on her anchor, and 
 cut them off. In the moments of mortification I immedi- 
 ately devised a plan for avoiding such an accident in the 
 future, by so arranging the wires along the banks of the river 
 as to cause the water itself to conduct the electricity across. 
 The experiment, however, was deferred until I arrived in 
 Washington; and on Dec. 16, 1842, I tested my arrangement 
 across the canal and with success. The simple fact was 
 then ascertained -that electricity could be made to cross a 
 river without other conductors than the river itself; but it 
 was not until the last autumn that I had the leisure to make 
 
 * From this we learn that Morse actually operated a wireless telegraph 
 before his Washington-Baltimore wire system was opened for service. 
 
 12 
 
PRACTICAL WIRELESS TELEGRAPHY 
 
 a series of experiments to ascertain the law of its passage. 
 The following diagram (Fig. 3) will serve to explain the 
 experiment : 
 
 "A, B, C, D are the banks of the river; N, P is the battery; 
 G is the galvanometer; ww are the wires along the banks 
 connected with copper plates, f, g, h, i, which are placed in 
 the water. When this arrangement is complete, the elec- 
 tricity, generated by the battery, passes from the positive 
 pole P to the plate h, across the river through the water to 
 
 w ..d^r- 
 
 plate i, and thence around the coil of the galvanometer to 
 plate f, across the river again to plate g, and thence to the 
 other pole of the battery. The distance across the canal is 
 eighty feet; on August 24 the following were the results of 
 the experiment* . . . showing that electricity crosses the 
 river and in quantity in proportion to the size of the plates 
 in the water. The distance of the plates on the same side of 
 the river from each other also affect the result. Having 
 ascertained the general fact I was desirous of discovering the 
 best practical distance at which to place my copper plates, 
 and not having the leisure myself, I requested my friend, 
 Professor Gale, to make the experiments for me." . . . 
 
 The experiments made by Professor Gale indicate that the 
 distance between the plates along the shores should be approxi- 
 mately three times greater than that from shore to shore 
 
 * The table containing information only of general interest is omitted. 
 
14 RADIODYNAMICS 
 
 across the stream, since four times the distance did not give 
 any increase in power and less than three times the distance 
 diminished the deflections of the galvanometer considerably. 
 Between 1854 and 1860 James Bowman Lindsay made similar 
 attempts at wireless telegraphy by utilizing water as the 
 conducting medium. 
 
 With an apparatus like that of Morse, Lindsay finally suc- 
 ceeded in signalling across the river Tay, where it is more 
 than a mile wide.* 
 
 J. W. Wilkins of the Cooke and Wheatstone Telegraph Co. 
 also experimented with earth conduction telegraphy in 1845, 
 and published the results of his investigations in the Mining 
 Journal, March 31, 1849, under the heading " Telegraph 
 Communication between England and France." f 
 
 Invention of the Telephone 
 
 After the invention of the telephone, in 1876, wireless 
 telegraphy went forward with leaps and bounds. The 
 marvelous sensitiveness of this instrument, which will give 
 audible responses under the application of less than one- 
 millionth of a volt of electromotive force, is "largely responsible 
 for the great progress made along these lines. Even wireless 
 telephony was introduced. 
 
 Its use in a telegraph line running parallel to another line 
 through which telephone conversation and singing was being 
 carried, led to the accidental discovery of its extraordinary 
 sensitiveness to induction currents in 1877, by Mr. Charles 
 Rathbone of Albany, N. Y.J 
 
 * See "Electrical Engineer," vol. xxiii, pp. 21-51; Kerr, "Wireless Teleg- 
 raphy," 1898, p. 40. 
 
 t For detailed accounts of his work see Fahie's "History of Wireless 
 Telegraphy," pp. 32-38. 
 
 t Journal of the Telegraph, Oct. i and 16; and Nov. i, 1877. For simi- 
 lar observations see Telegraphic Journal, Mar. i, 1788, p. 96; Journal of the 
 Telegraph, Mar. 16, 1878, Dec. i, 1877; The Electrician, vol. vi, pp. 207-303. 
 
PRACTICAL WIRELESS TELEGRAPHY 15 
 
 These observations on inductive effects in telephone cir- 
 cuits began to be investigated; in 1877 Prof. E. Sacher of 
 Vienna found that signals from three Smee cells sent through 
 one wire 120 m. long could be distinctly heard in the telephone 
 in another and parallel wire 20 m. distant.* 
 
 Prof. John Trowbridge of Harvard University was the first 
 to systematically study the problem of electromagnetic in- 
 duction signalling. His attention is concentrated chiefly on 
 the use of interrupted or alternating currents at the trans- 
 mitter and a telephonic receiver; in other respects his circuit 
 was practically the same as Morse's (Fig. 3).t 
 
 In 1884 Trowbridge described another plan using a com- 
 bination of both electromagnetic induction and earth con- 
 duction; later he discussed the phenomena of induction, 
 electromagnetic and electrostatic, as distinguished from leak- 
 age or earth currents, and with reference to their employment 
 in wireless telegraphy, t 
 
 Experiments of Alexander Graham Bell 
 
 About 1882 Alexander Graham Bell made some successful 
 experiments along this line suggested by Trowbridge. In his 
 paper read before The American Association for the Ad- 
 vancement of Science, in 1884, he says: 
 
 "A few years ago I made a communication on the use of 
 the telephone in tracing equipotential lines and surfaces. I 
 will briefly give the chief points of my experiment, which was 
 based on experiments made by Professor Adams of King's 
 College, London. Professor Adams used a galvanometer in- 
 stead of a telephone. 
 
 "In a vessel of water I placed a sheet of paper. At two 
 
 * Electrician, vol. i, p. 194. 
 
 t His investigations are discussed in detail in "The Earth as a Conductor 
 of Electricity," Am. Acad. Arts and Sc., 1880; see also "Silliman's Am. Journ. 
 Sc., 1880. 
 
 J Sc. Am. Supp., Feb. 21, 1891. 
 
16 RADIODYNAMICS 
 
 points on that paper were fastened two ordinary sewing 
 needles, which were also connected with an interrupter that 
 interrupted the circuit about one hundred times a second. 
 
 "Then I had two needles connected with a telephone; one 
 needle I fastened on the paper in the water, and the moment 
 I placed the other needle in the water I heard a musical sound 
 in the telephone. By moving this needle around in the 
 water, I would strike a place where there would be no sound 
 heard. This would be where the electric tension was the 
 same as in the needle; and by experimenting in the water 
 you could trace out with perfect ease an equipotential line 
 around one of the poles in the water. 
 
 "It struck me afterwards that this method, which is true 
 on the small, is also true on the large scale, and that it might 
 afford a solution of a method of communicating electric signals 
 between vessels at sea. 
 
 "I made some preliminary experiments in England, and 
 succeeded in sending signals across the river Thames in this 
 way. On one side were two metal plates placed at a distance 
 from each other, and on the other two terminals connected 
 with the telephone. A current was established in the tele- 
 phone each time a current was established through the 
 galvanic circuit on the opposite side, and if that current was 
 rapidly interrupted you would get a musical tone. 
 
 "Urged by Professor Trowbridge, I made some experiments 
 which are of very great value and suggestiveness. The first 
 was made on the Potomac river. I had two boats. In one 
 boat we had a Leclanche battery of six elements, and an in- 
 terrupter for interrupting the current very rapidly. Over the 
 bow of the boat we made water connection by a metallic plate, 
 and behind the boat we trailed an insulated wire, with a 
 float at the end carrying a metallic plate, so as to bring these 
 two elements about one hundred feet apart. I then took 
 another boat and sailed off. In this boat we had the same 
 
P&1CTICAL WIRELESS TELEGRAPHY 17 
 
 arrangement, but with a telephone in the circuit. In the 
 first boat, which was moored, I kept a man making signals; 
 and when my boat was near his I would hear those signals 
 very well a musical tone, something of this kind: turn, 
 turn, turn. I then rowed my boat down the river, and at a 
 distance of a mile and a quarter, which was the farthest I 
 tried, I could still (distinguish those signals. 
 
 "It is therefore perfectly practicable fof steam vessels with 
 dynamo machines to know of one another's presence in a fog 
 when they come, say, within a couple of miles of one another, 
 or, perhaps at a still greater distance. I tried the experiment 
 a short time ago in salt water of about twenty fathoms in 
 depth; I used then two sailing boats, and did not get so 
 great a distance as on the Potomac. The distance, which we 
 estimated by the eye, seemed to be about half a mile; but 
 on the Potomac we took the distance accurately on the shore.' 7 
 
 In 1886, convinced of the practicability of his method, Bell 
 says further: 
 
 "Most of the passenger steamships have dynamo engines 
 and are electrically lighted. Suppose, for instance, one of 
 them should trail a wire a mile long, or any length, which is 
 connected with the dynamo engine and electrically charged. 
 The wire would practically have a ground connection by 
 trailing in the water. Suppose you attach a telephone to 
 the end on board. Then your dynamo or telephone end 
 would be positive, and the other end of the wire trailing be- 
 hind would be negative. All of the water about the ship 
 will be positive within a circle whose radius is one-half the 
 length of the wire. All of the water about the trailing end 
 will be negative within a circle whose radius is the other half 
 of the wire. If your wire is one mile long there is then a 
 large area of water about the ship which is affected either 
 positively or negatively by the dynamo engine and -the 
 electrically charged wire. It will be impossible for any ship 
 
l8 RADIODYNAMICS 
 
 or object to approach within the water so charged in relation 
 to your ship without your telephone telling the whole story 
 to the listening ear. Now if a ship coming in this area has a 
 similar apparatus, the two vessels can communicate with 
 each other by their telephones. If they are enveloped in 
 fog, they can keep out of each other's way. The ship having 
 the telephone can detect other ships in its track, and keep 
 out of the way in a fog or storm. The matter is so simple 
 that I hope our ocean steamships will experiment with it." : 
 This method of signalling, attempted later by Messrs. 
 Rathenau, Rubens, and Strecker, was finally carried to a 
 distance of nearly nine miles, but the advent of the work of 
 Maxwell and Hertz followed by the practical application of 
 their theories and discoveries by Marconi and others, proved 
 such an advance in method, and the futility of trying to 
 make earth conduction systems duplicate the records of the 
 new Hertzian-wave telegraphy was so evident that work along 
 that line was practically discontinued. 
 
 * Public Opinion, Jan. 31, 1886. 
 
CHAPTER IV 
 
 ELECTROSTATIC AND COMBINED INDUCTION- 
 CONDUCTION TELEGRAPH SYSTEMS 
 
 Professor Dolbear's Electrostatic Telegraph 
 
 In 1882, at about the same time as A. G. Bell, Professor 
 Dolbear of Tufts College, Boston, was also engaged on the 
 problem of wireless telegraphy. His apparatus was some- 
 what more suggestive than any hitherto proposed and was 
 awarded a United States patent in March, 1882. The fol- 
 lowing is an extract from his patent specification: 
 
 b 
 Hi 
 
 OR 
 
 FIG. 4. 
 
 "In the diagram (Fig. 4), A represents one place and B 
 a distant place. C is a wire leading into the ground at A, 
 and D a wire leading into the ground at B ; G is an induction 
 coil having in the primary circuit a microphone transmit- 
 ter, T, and a battery, F, which has a number of cells suf- 
 ficient to establish in the wire C, which is connected with one 
 terminal of the secondary coil, an electromotive force of, say, 
 
 19 
 
20 RADIODYNAMICS 
 
 ioo volts. The battery is so connected that it not only 
 furnishes the current for the primary circuit, but also charges 
 or electrifies the secondary coil and its terminals C, and Hi. 
 
 "Now if words be spoken in proximity to transmitter T, 
 the vibration of its diaphragm will disturb the electrical con- 
 dition of the coil G, and thereby vary the potential of the 
 ground at B, and the receiver will reproduce the words spoken 
 in proximity to the transmitter, as if the wires CD were in 
 contact, or connected by a third wire. 
 
 "There are various well-known ways of electrifying the 
 wire C to a positive potential far in excess of ioo volts, and 
 the wire D to a negative potential far in excess of ioo volts. 
 
 "In the diagram, H, Hi, Ha represent condensers, the con- 
 denser Hi being properly charged to give the desired effect. 
 The condensers H and H2 are not essential, but are of some 
 benefit; nor is the condenser Hi essential when the secondary 
 coil is otherwise charged. I prefer to charge all these con- 
 densers, as it is of prime importance to keep the grounds of 
 wires C and D oppositely electrified, and while, as is obvious, 
 this may be done by either the batteries or the condensers, I 
 prefer to use both." 
 
 In the Scientific American Supplement, Dec. n, 1886, 
 Professor Dolbear gives some additional particulars: 
 
 "My first results were obtained with a large magneto 
 electric machine with one terminal grounded through a Morse 
 key, the other terminal out in free air and only a foot or two 
 long; the receiver having one terminal grounded, the other 
 held in the hand while the body was insulated, the distance 
 between grounds being about sixty feet. Afterward -much 
 louder and better effects were obtained by using an induction 
 coil having an automatic break and with a Morse key in the 
 primary circuit, one terminal of the secondary grounded the 
 other free in air, or in a condenser of considerable capacity, 
 the latter having an air discharge of fine points at its opposite 
 
TELEGRAPH SYSTEMS 21 
 
 terminal. At times I have employed a gilt kite carrying a 
 fine wire from the secondary coil. The discharges then are 
 apparently nearly as strong as if there was an ordinary 
 circuit. 
 
 "The idea is to cause a series of electrical discharges into 
 the earth without discharging into the earth the other termi- 
 nal of the battery or induction coil a feat which I have been 
 told so many many times was impossible, but which certainly 
 can be done. An induction coil isn't amenable to Ohm's 
 law always! Suppose at one place there be apparatus for 
 discharging the positive pole of the induction coil into the 
 ground, say, 100 times a second, then the ground will be 
 raised to a certain potential 100 times a second. At another 
 point let a similar apparatus discharge the negative pole 100 
 times a second; then between these two places there will be 
 a greater difference of potential than in other directions, and 
 a series of earth currents, 100 per second, will flow from one 
 to the other. Any sensitive electrical device, a galvanometer 
 or a telephone, will be disturbed at this latter station by 
 these currents, and any intermittence of them, as can be 
 brought about by a Morse key in the first place, will be seen 
 or heard in the second place. The stronger the discharges 
 that can be thus produced, the stronger will the earth currents 
 be of course, and an insulated tin roof is an excellent terminal 
 for such a purpose. I have generally used my static telephone 
 in my experiments, though the magneto will answer. 
 
 "I am still at work on this method of communication, to 
 perfect it. I shall soon know better its limits on both land 
 and water than I do now. It is adapted to telegraphing be- 
 tween vessels at sea. 
 
 "Some very interesting results were obtained when the 
 static receiver with one terminal was used. A person stand- 
 ing on the ground a distance from the discharging point could 
 hear nothing; but very little standing on ordinary stones, as 
 
22 RADIODYNAMICS 
 
 granite blocks or steps; but standing on asphalt concrete, the 
 sounds were loud enough to hear with the telephone at some 
 distance from the ear. By grounding one terminal of the 
 induction coil to the gas or water pipes and leaving the other 
 end free, telegraph signals can be heard in any part of a big 
 building and its neighborhood without any connection what- 
 ever, provided the person be well insulated." 
 
 Explanation of Dolbear's System 
 
 Although Professor Dolbear's circuit arrangements re- 
 semble somewhat those of Marconi, his system lacked the 
 essential features which, later, were applied so successfully, 
 namely, electrical oscillations of high frequency at the trans- 
 mitter and suitable detecting apparatus at the receiver. 
 
 Dolbear's results were clearly those of electrostatic induction 
 and not, as he believed, due to conducting effects through 
 the earth; the earth connections merely served to furnish 
 one side of an electrostatic condenser, the other sides of 
 which were supplied by the elevated conductors; the same 
 results can be secured by using insulated metallic capacity 
 areas, now known as counterpoises, instead of the earth as 
 the lower halves of the radiating and receiving aerial systems. 
 This is made plain by a study of the drawings taken from his 
 patent specification. 
 
 The functions of the elevated condensers, H, Hi, and H2, 
 and of the battery b (Fig. 4), are not evident, since the under- 
 lying principle upon which the whole system is based does 
 not explain their necessity. This principle is nothing more 
 than a statement of the laws of electrostatic induction; it 
 can best be understood by a study of the properties and 
 action of the circuits with the unnecessary apparatus omitted. 
 At the transmitter we then have a voice-controlled source of 
 high potential, one end of which is earthed and the other con- 
 nected to an insulated elevated conductor. At the receiver 
 
TELEGRAPH SYSTEMS 
 
 Stress Lines of 
 Electrostatic Field 
 
 lew fed Charged 
 Conductor 
 
 we have a similar elevated conductor earthed through an 
 electrostatic telephone. When sound waves impinge on the 
 microphone of the transmitter, fluctuating currents are set 
 up in the primary of the induction coil; these produce fluctu- 
 ating potentials at the terminals of the secondary winding, 
 which are conducted to the elevated capacity area; the 
 latter with the earth forms an electrostatic condenser with 
 the intervening air as the dielectric. The electrostatic field 
 of force of this condenser extends radially out and down- 
 wards from the aerial wire 
 in curved lines, as is graphi- 
 cally shown in the accom- 
 panying diagram. (Fig. 5.) 
 Now .if an insulated body, 
 such as the elevated wire 
 
 i- ,-i T '.,1 -'-i~'' -' <-.:- . -Earth* '"' " 
 
 of the receiver, lies within 
 
 this field of force, potentials FlG> 5> 
 
 will be induced on it, the amplitude of which varies in unison 
 
 with the variations of potential on the transmitting aerial 
 
 wire. 
 
 Since the earthed plate of the electrostatic telephone in 
 the receiver remains constant at the earth's potential and 
 since the other plate is connected to the elevated wire and 
 subject to the inductive action of the transmitter, a varying 
 difference of potential is therefore set up between the plates, 
 with a consequent variation of attraction between them. 
 One of the plates, which is a diaphragm of flexible metal or 
 of some such material as thin sheet mica covered with a tin 
 foil conducting area, is therefore made to vibrate and repro- 
 duce sounds produced at the transmitter. 
 
 Lowenstein's Electrostatic Telegraph 
 
 Mr. Fritz Lowenstein, a consulting and research engineer 
 of New York, engaged in radio research work, suggested a 
 
24 RADIODYNAMICS 
 
 similar method' of signalling to short distances in iqi2. 
 This was based principally upon the marvellous sensitiveness 
 of his potential operated receiving device. It could be used 
 advantageously with the (magneto) telephone and was 
 therefore adapted for both telegraphy and telephony; the 
 telegraphic system, however, gave the best results for dis- 
 tance; telegraphic signals were sent, with his apparatus, from 
 his laboratory at 115 Nassau Street to the Liberty Build- 
 ing at the corner of Nassau and Liberty streets, about a 
 half mile distant. The transmitter consisted of a 2o,ooo-volt 
 transformer, the primary of which was energized by a 500- 
 cycle alternating current. One terminal of the secondary 
 was grounded to the water pipe system; the other was con- 
 nected to a single, nearly vertical conductor (No. 8 stranded 
 copper), the upper well-insulated end of which extended to a 
 drop wire from the top of a three hundred-foot office building 
 nearby. 
 
 The receiving station near the top of the Liberty Building 
 consisted of a one hundred-foot length of bell wire suspended 
 from a pole out one of the windows and connected to Mr. 
 Lowenstein's ion controller detector,* the other terminal of 
 which was grounded to the water pipes. The sensitive 
 telephone connected to the instrument clearly indicated the 
 Morse signals sent out at the transmitter.- These were made 
 by opening and closing the primary circuit of the transformer 
 with a Morse key. 
 
 Passing over the work of Thomas A. Edison, W. F. Melruish, 
 C. A. Stevenson, Professor Erich Rathenau, and others, we 
 come to another serious attack on the problem of wireless teleg- 
 raphy, which was executed in a masterly way by Sir William 
 Preece, engineer-in-chief of the postal telegraph system in 
 England. 
 
 * A potential-operated, ionized gas-detector and amplifier for radio- 
 telegraphy, radiotelephony, and wire telephony. 
 
TELEGRAPH SYSTEMS 
 
 Preece's Induction-Conduction System 
 
 Preece's system was a combination of three previously 
 existing systems, namely, earth conduction, electrostatic 
 induction, and electromagnetic induction. Although it is 
 certain that each of these three phenomena played a part in 
 the transmission of the signals, their relative importance has 
 not been definitely determined. A brief explanation will 
 serve to make his method clear. 
 
 The signals were transmitted between two long, horizontal 
 wires, one at the transmitter and one at the receiver. These 
 
 Condenser 
 
 Listening-//! tfey 
 
 'Earth 
 
 Receiving Telephone 
 
 FIG. 6. 
 
 Earth 
 
 wires were supported parallel to one another on telegraph 
 poles and were connected to earth plates of considerable area 
 at their two ends. The diagram, Fig. 6, shows the con- 
 nections at each station, which is a combined transmitter and 
 receiver.* 
 
 The pulsating currents through the sending wire and the 
 earth produce a variable electromagnetic and electrostatic 
 field, which induces a fluctuating E.M.F. in the receiving cir- 
 cuit. This is indicated by sounds in the telephone. The in- 
 duced currents are also augmented by the currents conducted 
 through the earth itself. 
 
 * The use of a " breaking-in " key in this circuit will be found very inter- 
 esting to practical operators since a number of inventors, within the last few 
 years, have brought forward this principle as novel for use in radio systems. 
 
26 RADIODYNAMICS 
 
 It has been shown that the hemispheroidal mass, repre- 
 sented by the lines of current-flow from one plate to the other, 
 can be replaced electrically by a resultant conductor of defi- 
 nite form and position. This is illustrated in Fig. 7, where 
 L is the line wire, PP the earthed plates, POP, PAP, etc., 
 the equipotential lines of current-flow, and R the resultant 
 conductor. The induction effects occurring between two such 
 circuits are therefore the same as if they were composed 
 entirely of metallic conductors of the same physical and 
 electrical characteristics as the line wires with their result- 
 ant earth conductors. At Loch Ness, where the parallel 
 wires were about three miles long, the calculated depth of the 
 
 L 
 
 resultant earth conductor was about nine hundred feet. This 
 arrangement of parallel line wires with earthed ends therefore 
 gave all the advantages of signalling between huge, single-turn 
 coils, with the increased effect due to earth conduction, and 
 without the almost insuperable difficulties involved in con- 
 structing such coils above the earth. 
 
 In March, 1898, this system was permanently established 
 for signalling between Lavernock Point, on the mainland, 
 and Flat-Holm in the Bristol Channel, a distance of over 
 three miles. Fifty Leclanche cells and an interrupter fre- 
 quency of four hundred makes and breaks per second were 
 used for transmitting, and a telephone served as the receiving 
 indicator. The signals were very distinct, and it is said a speed 
 of forty words a minute has been attained without difficulty. 
 
CHAPTER V 
 ELECTROMAGNETIC WAVE SYSTEMS 
 
 The profound speculations and mathematical researches of 
 Maxwell on the electromagnetic nature of light, followed by 
 the brilliant work of Hertz and his successors, are so familiar 
 to the scientific public that a brief resume of the evolution of 
 the art is here sufficient. 
 
 Again we see that radio signalling, like most wonders of 
 science, has not been an invention, in the popularly accepted 
 meaning of the word, but rather a gradual, step-by-step de- 
 velopment in which many prominent men of science have 
 
 M 
 
 FIG. 8. 
 m EE is the glass tube. P, P, the connectors, and M, the filings. 
 
 played a part. Maxwell's theories, published in 1865, laid 
 the foundation, and Hertz, by a long and carefully executed 
 series of experiments, paved the way; Hertz's successors, 
 men who foresaw the practical value of these discoveries, 
 utilized the material he laid bare for the production of a 
 serviceable means of communication. 
 
 The greatest need in the extension of Hertz's work to 
 greater distances, was a receiving wave detector of high 
 sensitiveness. A crude form of such a detector had, as early 
 as 1866 been used by S. A. Varley as a lightning arrester. 
 In 1890 Prof. E. Branly of the Catholic University of Paris 
 rediscovered the effects, already utilized by Varley, of Hertzian 
 
 27 
 
28 
 
 RADIODYNAMTCS 
 
 waves on the conductivity of metallic filings. He also ob- 
 served the restoring or decohering effect of light tapping on 
 the filings tube. In 1893 Sir Oliver Lodge repeated Hertz's 
 experiments, using the "Branly tube," or " coherer," as he 
 
 Capacity Ana 
 
 FIG. 9. 
 
 called it, in place of the micrometer spark gap in the Hertz 
 resonator. Branly 's coherer is shown in Fig. 8. Lodge's 
 apparatus, connected for reception of signals, is shown dia- 
 grammatically in Fig. 9. With this apparatus he was able to 
 observe Hertzian waves at distances up to about 150 feet. 
 
 Early Work of Nikola Tesla 
 
 Nikola Tesla, after completing the application of his dis- 
 covery of the rotating magnetic field to electric motors in 
 1888, turned his attention to the problem of transmitting 
 electrical energy to a distance without wires. His earliest 
 plans were to transmit energy not only in small amounts, for 
 purposes of communication, but also in amounts sufficient for 
 industrial purposes. 
 
 The first public announcements of these plans were made in 
 February and March, 1893. He delivered lectures before the 
 Franklin Institute in Philadelphia, and the National Electric 
 Light Association in St. Louis. However, in 1891, he had 
 already described and shown, in a lecture before a scientific 
 society, a method of lighting an electric lamp at a short dis- 
 tance without connecting wires. High-frequency oscillations 
 were used in these experiments, but the power of the apparatus 
 was small in comparison with that of his later lectures and 
 experiments. 
 
ELECTROMAGNETIC WAVE SYSTEMS 29 
 
 In these lectures he expressed the conviction that: "It 
 certainly is possible to produce some electrical disturbance 
 sufficiently powerful to be perceptible by suitable instruments 
 at any point on the earth's surface." 
 
 Describing his plan in detail he says: 
 
 "Assume that a source of alternating currents be connected 
 as shown in the accompanying diagram (Fig. 10) with one of its 
 terminals connected to earth (convenient to the water mains) 
 and with the other to a body of large surface, P. When the 
 electric oscillation is set up, there will be a movement of electri- 
 city in and out of P, and alternating currents will pass through 
 the earth, converging to or diverging from the point C, where 
 
 FIG. 10. 
 
 the ground connection is made. In this manner neighboring 
 points on the earth's surface within a certain radius will be 
 disturbed. But the disturbance will diminish with the dis- 
 tance, and the distance at which this effect will still be per- 
 ceptible will depend on the quantity of electricity set in motion. 
 Since the body P is insulated, in order to displace a considerable 
 quantity the potential of the source must be excessive, since 
 there would be limitations as to the surface of P. The condi- 
 tions might be adjusted so that the generator or source, S, will 
 set up the same electrical movement as though its circuit were 
 closed. Thus it is certainly practicable to impress an electric 
 vibration, at least of a certain low period, upon the earth. 
 Theoretically it should not require a great amount of energy 
 to produce a disturbance perceptible at great distance, or even 
 
30 RADIODYNAMICS 
 
 all over the surface of the globe. Now, it is quite certain that 
 at any point within a certain radius of the source, S, a properly 
 adjusted self-induction and capacity device can be set in action 
 by resonance. Not only can this be done, but another source 
 Si, similar to S or any number of such sources, may be set to 
 work in synchronism with the latter, and the vibration thus 
 intensified and spread over a large area; or a flow of electricity 
 produced to or from the source Si, if the same be of opposite 
 phase to the source S. Proper apparatus must first be pro- 
 duced, by means of which the problem can be attacked, and I 
 have devoted much thought to this subject." 
 
 Tesla continued his investigations along these lines and in 
 1898 had already developed apparatus of great power giving a 
 pressure of four million volts and discharges extending through 
 sixteen feet. At that time and even today this is considered 
 remarkable. From 1899 to I 9 ne continued his investiga- 
 tions and in 1900 he published, in the Century Magazine, a 
 long article of absorbing interest and of great suggestiveness on 
 "The Problem of Increasing Human Energy." Therein he 
 described and illustrated with actual photographs his appara- 
 tus for producing pressures of over twelve million volts and 
 capable of delivering energy at the rate of seventy-five thou- 
 sand horse-power. 
 
 Professor PopofFs Receiver 
 
 Professor Popoff, in a communication to the Physico- 
 Chemical Society of St. Petersburg, in 1895, described a 
 form of receiving apparatus designed by him for the study 
 of atmospheric electricity. His circuit arrangement which is 
 shown in Fig. n is different from that of Lodge in that one 
 terminal of the coherer is grounded, and the other is connected 
 to a vertical conductor extending above the housetop. Here 
 is introduced the well-known method of utilizing the electric 
 signal bell for an automatic decoherer. Professor Popoff also 
 
ELECTROMAGNETIC WAVE SYSTEMS 31 
 
 used a form of tape recorder which automatically recorded 
 the duration of the electrical disturbances in the atmosphere. 
 This apparatus and circuit arrangement is precisely the same 
 as that used by Marconi in his early experiments. That 
 Popoff foresaw the possibilities of his receiver for Hertzian 
 wave telegraphy is clearly evidenced by the concluding para- 
 
 FIG. ii. 
 
 graph of a paper read before the Institute of Forestry of 
 St. Petersburg. "In conclusion," he says, "I may express 
 the hope that my apparatus, with further improvements, 
 may be adapted to the transmission of signals to a distance 
 by the aid of quick electric vibrations (high-frequency oscilla- 
 tions) as soon as a means of producing such vibrations possess- 
 ing sufficient energy is found." 
 
 Marconi's Early Work 
 
 With Hertz's oscillator and PopofFs receiver Marconi began 
 his experiments on his father's estate near Bologna, Italy, in 
 1895. Although only 22 years of age he had already acquired 
 much knowledge of Hertzian waves, having studied under 
 
32 RADIODYNAMICS 
 
 Professor Rosa of the Leghorn Technical School, and ac- 
 quainted himself with the published writings of Professor 
 Righi of the University of Bologna. After a year of experi- 
 menting Signor Marconi went to England and filed, in the 
 Patent Office of Great Britain, an application for a patent, 
 which was duly granted. 
 
 Later Improvements 
 
 Improvements in both the more efficient generation and 
 reception of the electromagnetic waves have, since 1895, 
 chiefly engaged the attention of radio investigators. Among 
 the more important advances may be mentioned the intro- 
 duction of the Tesla high-frequency transformer for coupled 
 circuits by Lodge and Braun, instead of the direct spark- 
 excited antenna of Marconi; the discovery and adoption of 
 detectors suitable for use with the telephone ; the introduction 
 of alternating current and high spark frequencies for trans- 
 mission; the utilization of Wien's discovery of the quenched 
 spark gap; and the more recent attacks on the problem of 
 selectivity. The recent work of Fessenden, Alexanderson, 
 and Goldschmidt on the direct production of high-frequency 
 alternating current of continuous amplitude for electric wave 
 telegraphy and telephony, is worthy of mention. 
 
CHAPTER VI 
 
 POSSIBLE CONTROL METHODS FOR RADIO- 
 DYNAMICS SOUND WAVES 
 
 Every teledynamic system has two principal parts, namely, 
 (i) the apparatus for the transmission and reception of the 
 controlling energy, and (2) the apparatus or mechanisms to 
 be controlled. This broad subdivision applies to such simple 
 forms as the telegraph, where the energy-transmitting medium 
 is a metallic conductor, and the receiver a relay controlling 
 a sound-producing mechanism, as well as to the very com- 
 plicated systems utilizing the ether as the connecting link. 
 
 Of these two divisions the first is to us by far the most 
 important, if for no other reason because of the difficulty it has 
 presented in the practical solution of such representative 
 problems as torpedo control. It therefore demands careful 
 consideration, especially with reference to a proper selection 
 of the kind of radiant or other energy to be used. 
 
 The following table gives some of the most important forms 
 of radiant energy in ether and air, their vibration frequencies, 
 and detecting means capable of actuating mechanisms: 
 
 Waves 
 
 Frequency per sec. 
 
 Detector 
 
 Acoustic 
 
 1 6 to 35,000 
 
 Vibratory relay 
 
 Hertzian 
 
 50,000 to 2 billions 
 
 Hertzian wave detector 
 
 Infra-red, or heat 
 Visible 
 
 2 tO 40OO " 
 
 4000 to 8000 ' ' 
 
 Thermoelectric cells 
 Selenium cells 
 
 Ultra-violet 
 
 8000 to ? 
 
 Trigger vacuum tube. 
 
 Besides these radiant energy means we may mention 
 earth conduction, electrostatic induction, and electromag- 
 netic induction. 
 
 33 
 
34 RADIODYNAMICS 
 
 Choice of Control Energy 
 
 A number of important factors must be taken into con- 
 sideration in order to make the best choice of these several 
 control methods. Although the Hertzian wave system is 
 employed in nearly all of the suggested applications of radio- 
 dynamics, and is to all appearances the most reliable and 
 best, who can say that any one of these other possible methods, 
 if it received the proper attention, would not be much simpler, 
 and at the same time still more reliable? 
 
 Reliability is the factor of prime importance in the abso- 
 lute and accurate control of a dangerous weapon like a torpedo, 
 travelling, as it does, at a speed of between thirty and forty 
 miles per hour and carrying large quantities of highly ex- 
 plosive material. Simplicity, freedom from accidental or 
 intentional interference, and cost are other points which 
 demand careful thought. The maximum range at which 
 control is necessary, and indeed possible, is limited by vision. 
 This, in clear weather, does not exceed eight miles, for even 
 with a good binocular the torpedo cannot be seen beyond 
 that distance. In cloudy or stormy weather the operations 
 may be limited to two or three miles. This does not mean 
 that the usefulness of the wirelessly directed torpedo is 
 limited to calm, clear weather, for any attacking fleet or ship 
 would be subject to the same conditions, inasmuch as the 
 distance and accuracy of their fire is greatly affected by the 
 condition of the sea and weather. 
 
 Difficulties to be Overcome 
 
 The reader may think of the four-thousand mile accom- 
 plishments of modern radiotelegraphy and immediately con- 
 clude that the problem of getting a sufficient amount of 
 energy to the vessel is one of comparative simplicity. On 
 the contrary this is one of the chief difficulties, and it has only 
 lately begun to be surmounted. 
 
SOUND WAVES 35 
 
 The following table will serve in a rough way to show the 
 comparison between transmitted and received energies in 
 various types of electrical energy- transmitting systems: 
 
 Watts transmitted Watts received Ratio 
 
 Power line io 6 io 6 i 
 
 Cable telegraph i icf" 3 lo" 3 
 
 Telephone icr 2 icr 6 icr 4 
 
 Radiotelegraphy io 5 icr 8 icr 13 
 
 From this table it may be seen that of the one hundred kilo- 
 watts used at a high-power radiotelegraphic transmitting 
 station but one ten- trillion th part is received at a distance 
 corresponding to the maximum working range, i.e., the 
 range at which the received power is measured in hundred- 
 million ths of a watt. This range in daylight is usually in the 
 neighborhood of three thousand miles, but is subject to con- 
 siderable variation from day to day and from season to season; 
 the night range is also very much greater than the day range 
 during those parts of the year when atmospheric disturbances 
 cause the least amount of interference. 
 
 In long-distance radiotelegraphic sets the transmitter is of 
 such power (25 to 100 kw.), as would be excessive for torpedo 
 control in coast defence. But far more important than this 
 is the fact that the telephone, which is used as the receiving 
 indicator in wireless telegraph sets, will give readable signals 
 under an impressed e.m.f., of less than one-millionth of a volt, 
 while to trip the most sensitive relay under ideal conditions 
 requires about one-thousandth of a volt e.m.f. applied to 
 its terminals. Under the conditions of shock and vibration 
 aboard a small vessel in a rough sea the restoring spring of 
 such an instrument must be set under sufficient tension to 
 prevent the making of false contacts; the sensitiveness is 
 thereby reduced to from one-fifth to one-tenth of its highest 
 value. From these values we can readily see that a radio- 
 telegraphic receiver may easily be as much as 5000 times as 
 
RADIODYNAMICS 
 
 sensitive as the type necessary for the absolutely reliable 
 control of mechanisms. Practically all systems of wireless 
 signalling depend for their long-distance operation on this 
 
 FIG. 12. 
 
 Prof. Fesseriden's submarine sound signalling apparatus used to detect the 
 presence of submarines. (Published by permission.) 
 
 extraordinary sensibility of the telephone; when used with a 
 relay the distance over which they are operative likewise de- 
 creases tremendously. 
 
 Sound Waves in Radiodynamics 
 
 The employment of sound waves in air for radiodynamics has 
 not been productive of any noteworthy results. Submarine 
 
SOUND WAVES 37 
 
 signalling, however, has been developed to the point where 
 the transmitting bell signals have been received at distances 
 
 FIG. 13. 
 Operator sending submarine sound signals with the Fessenden apparatus. 
 
 (Courtesy of the American Magazine.) 
 
 up to about 25 miles. Prof. R. A. Fessenden, one of the 
 pioneers and authorities on radiotelegraphy in the United 
 
38 RADIODYNAMICS 
 
 States, signalled across Massachusetts Bay during the spring 
 of 1914, with a submarine sound wave apparatus which he 
 invented. Figs. 12, 13, 14 and 15 show Professor Fessenden 
 and various parts of his submarine signalling system. These 
 
 FIG. 14. 
 
 Vibrating steel diaphragm used as both transmitter and receiver in the Fessen- 
 den submarine signalling system. (Courtesy of the American Magazine.} 
 
 photographs are reproduced through the courtesy of the 
 American Magazine.* 
 
 Although little has been done with this signalling system 
 in adapting it to the severe requirements of torpedo control, 
 its possibilities are not unworthy of consideration. 
 
 The fact that most steamships, war vessels, and submarine 
 boats are now equipped with submarine signalling apparatus 
 is ample proof of the practicability of this system for fog and 
 
 * For further details of submarine signalling apparatus see: Jour. Am. 
 Soc. Nav. Engrs., Aug., 1914; Mar. Engr. and Nav. Archt., May, 1914; Proc. 
 Am. Inst. Elec. Engrs., July, 1912. 
 
SOUND WAVES 
 
 39 
 
 warning signalling. As practiced, a submerged bell, electri- 
 cally operated, is used as the transmitter, while a submerged 
 microphone transforms the received sound waves into elec- 
 trical effects observable upon a telephone receiver; this re- 
 ceiving apparatus is in all respects the same in principle as 
 the ordinary telephone which we have in our offices and 
 
 FIG, 15. 
 
 Professor Reginald A. Fessenden taking observations on the sound waves sent 
 out by submarines. (Courtesy of the American Magazine.) 
 
 homes. An electric ear of this kind is usually installed on 
 each side of the vessel, and two telephones provided in the 
 pilot house for the observer. By switching from one to the 
 other of these the general direction of the transmitter can 
 usually be determined, since the receiving microphone on the 
 side of the boat nearest the bell will give the stronger signal. 
 When the signals are of equal strength in both telephones the 
 direction of the bell at the dangerous reef can be determined 
 
40 RADIODYNAMICS 
 
 by swinging the ship. The practicability of apparatus based 
 on such an energy transfer method although not assured is 
 not wholly uncertain. One advantage of no mean impor- 
 tance is that the torpedo could be entirely submerged, offering 
 no target for the enemy's gun fire. Every other system ex- 
 cept earth conduction in practice would require a portion of 
 the receiving apparatus to project above the water. 
 
 By utilizing sound waves of frequencies below the audible 
 limit (16 per second), the control impulses could not be 
 detected by the enemy unless they were provided with special 
 apparatus for that purpose. If such a transmitter be used 
 with tuned mechanical elements in connection with current 
 amplifying devices at the receiver, it is possible that an 
 extremely simple and effective system of control could be 
 developed.* The torpedo, although invisible, could be ac- 
 curately located by means of two submerged microphones, 
 which would respond to signals sent out by the torpedo 
 itself. This scheme has been used in the European War 
 to detect the presence of hostile submarine boats. The 
 principal difficulties to be met in the use of submarine 
 sound waves for torpedo control are the interfering signals, 
 which the enemy might easily send out, and the very weak 
 electrical effects produced by the transmitter at battle-range 
 distances. The former is an extremely difficult problem. 
 The latter might be overcome by using a simple form of 
 amplifier, such as De Forest's. 
 
 * Such a scheme was described by the author in a lecture on The Wirelessly 
 Directed Torpedo, before the Indianapolis-Lafayette section of the American 
 Institute of Electrical Engineers in October, 1913. 
 
CHAPTER VII 
 INFRA-RED OR HEAT WAVES 
 
 Omitting Hertzian waves for the present we come to the 
 infra-red rays as a possible means of effecting mechanism 
 operation at a distance. The great sensitiveness of the 
 bolometer, thermo-pile, and other thermal and thermo- 
 electric detectors suggests the use of these rays as a form of 
 wave energy capable of serving our needs. 
 
 No mention of the use of radiant heat for operating dis- 
 tant switches has been found in scientific literature. As a 
 means of telephoning to short distances, however, it was 
 among the first to be suggested, as previously stated. 
 
 Stimulated by the accounts of the extreme sensitivity of 
 radiant heat detectors and of their use in the measurement of 
 stellar radiations, the writer has given some thought to the 
 possibility of using heat waves as a control agency in a system 
 for the wireless direction of torpedoes. 
 
 Let us consider first the general advantages and disadvan- 
 tages of such a control energy, assuming that we have generat- 
 ing means of such power and receiving detectors of such 
 sensitiveness that we are able to control switches at useful 
 distances. 
 
 One of the first advantages, and perhaps the greatest, lies 
 in our ability to direct this energy at will. By means of the 
 highly-polished, parabolic surfaces of such metals as silver 
 and zinc, we can direct practically the whole of our generated 
 energy into a beam of parallel rays. Surfaces of silver and 
 zinc, when well polished, will absorb no more than two or three 
 
 41 
 
42 RADIODYNAMICS 
 
 per cent of the incident radiant energy, the remaining ninety- 
 eight or ninety-seven per cent being reflected. 
 
 In order to secure the advantages of direction by the use of 
 parabolic reflectors, we must confine our source of heat to a 
 comparatively small area. But if the area be small the rate 
 at which the energy is radiated per unit of area must be 
 correspondingly large. A high radiation rate per unit of area 
 can only be obtained with a high temperature. In order then 
 that we may be able efficiently to utilize heat radiations we 
 must have first, a source of easily controlled energy which can 
 readily be converted into the energy of radiant heat; second, 
 a means of developing an extremely high emission rate per 
 unit area; third, a means of limiting the radiation to a small 
 area; and fourth, a properly shaped reflecting surface of 
 material suitable for directing the heat energy developed into 
 a beam of parallel rays. Disregarding our primary assump- 
 tion, we must in addition be able to project these rays upon 
 a swiftly moving receptor at five miles distance with sufficient 
 effect to produce definite, mechanical movements at will. 
 
 These requirements are admirably met in our present high- 
 power searchlights. Electricity as a prime source of energy 
 lends itself easily to our needs because of its extreme flexi- 
 bility; the electric arc as a means of transforming this energy 
 into heat is not only extremely efficient, but fulfills the require- 
 ments of small area and very high temperature as well. The 
 energy in the visible portion of the electric arc spectrum does 
 not exceed ten per cent of the input energy; but with this we 
 are not particularly concerned, since a " black body" receiving 
 surface will enable us to convert practically all of the radiation 
 incident upon it, including, besides all of the infra-red, the 
 visible, and most of the ultra-violet also. 
 
 The energy emission rate per unit of area, which is a 
 function of the energy density per unit of crater surface, is 
 exceedingly high; the energy density may reach twenty-one 
 
INFRA-RED OR HEAT WAVES 
 
 43 
 
 and one half watts per square millimeter, and the temperature 
 may rise to the vicinity of three thousand eight hundred 
 degrees Centigrade. Moreover the energy of the high-tem- 
 perature portion is limited to a comparatively small value 
 by the low coefficient of thermal conductivity of the electrode 
 
 material. This allows a suf- 
 
 fluently close approach to 
 the. "point source" ideal de- 
 sired with parabolic re- 
 flectors, for practical utility. 
 Electric searchlights, or 
 "projectors," as they are fre- 
 quently called, have been 
 built with parabolic reflecting 
 mirrors sixty inches in di- 
 ameter. Such a projector of 
 the type used in the United 
 States Navy is shown in Fig. 
 16. The power of these 
 projectors can easily be 
 raised to fifteen or twenty 
 kilowatts. Were it necessary, 
 heat-wave generators of this 
 kind could be constructed 
 having a capacity for trans- 
 
 FIG. 16. 
 
 Sixty-inch projector used with radio- 
 dynamic torpedoes. (Published by per- 
 mission of General Electric Co.) 
 
 forming an electrical energy 
 
 of one hundred kilowatts 
 
 into the energy of radiant heat. The infra-red radiations of 
 
 the electric arc may be increased by the addition of barium 
 
 chloride to the arc electrodes. 
 
 Invisibility 
 
 A searchlight transmitter can be installed directly on a 
 harbor or coast line and so masked as to be completely in- 
 
44 RADIODYNAMICS 
 
 visible to ships several miles at sea. The electric power would 
 preferably be generated at a central power plant and trans- 
 mitted over a hidden high-tension transmission line to a 
 number of these hidden control stations. By means of step- 
 down transformers (and rotary-converters if it is necessary to 
 use direct-current arcs), the high tension line currents fur- 
 nished by the central station would be transformed to currents 
 of proper potential and power for the operation of the high- 
 power, electric-arc, heat-wave generators. 
 
 Control operators at each hidden transmitter would be in 
 constant communication with each other and with the military 
 head of the harbor defenses in order that the control operations 
 might be constantly in the hands of the operator in the most 
 advantageous position with respect to the attacking war vessels. 
 
 Selective Operation 
 
 Since heat waves as a control agency, unlike Hertzian waves, 
 sound waves, electromagnetic and electrostatic induction, or 
 earth conduction, can be directed at will, their use demands 
 no consideration of the selectivity problem, the solution of 
 which has ever been practically unattainable under the con- 
 ditions imposed in torpedo control. Although it is possible 
 to produce Hertzian waves with fronts perpendicular to the 
 direction of propagation, the difficulties involved in construct- 
 ing reflectors of sufficient size for the wave-lengths necessary 
 are very great from a practical point of view. Other means 
 have been developed for directing Hertzian waves, among 
 which may be mentioned the radio-goniometer of Bellini and 
 Tosi, but in practice it has been found that the power of such 
 transmitters is limited. 
 
 In order to prevent the enemy from projecting the beams of 
 their own searchlights onto the receiver of our torpedo, the 
 latter is provided with a gyroscope which serves to keep the 
 receiving heat detector always facing toward our own trans- 
 
INFRA-RED OR HEAT WAVES 45 
 
 mitters on shore and away from the enemy at sea, a screen of 
 opaque material on the side toward the sea providing means 
 cf intercepting the rays from the enemy's lights. This same 
 gyroscope at night serves to keep from the view of the enemy 
 at sea, the screened signal lights on the torpedo, which at all 
 times are plainly visible from shore, and which are automati- 
 cally operated by the control apparatus within the torpedo. 
 Their purpose is to permit the control operator on shore to 
 follow the direction of the torpedo without keeping his trans- 
 mitting searchlight directed upon it, and thus in continuous 
 view of the attacking ships, and at the same time automatically 
 to signal back the operations occurring on the torpedo. 
 
 The rays of the searchlight are invisible in bright daylight 
 unless an observer be directly in their path ; this is desirable, 
 inasmuch as it prevents the enemy from locating the screened 
 control stations. At night the powerful light is a distinct 
 advantage in locating any attacking ships, and, when neces- 
 sary, in following the torpedo itself. Should it become 
 necessary to have the control station invisible during the night 
 as well as by day, suitable ray niters would be necessary. 
 Substances which will screen off or absorb the visible radia- 
 tions and allow the longer infra-red waves to be transmitted, 
 that is, substances which are said to be "diathermanous," are: 
 black fluorite, smoky quartz, black glass, and a strong solution 
 of iodine in carbon disulphide; gases not near the point of 
 condensation are also highly diathermanous. 
 
 Dispersion and Atmospheric Absorption 
 
 The best of our present-day searchlights are not capable of 
 producing strictly parallel rays. The non-parallelism usually 
 amounts to at least three or four degrees. Because of this 
 dispersion the beam of a searchlight which at the mirror is 
 sixty inches in diameter, may be five hundred or a thousand 
 feet in diameter at a distance of five miles. It is obvious, 
 
46 RADIODYNAMICS 
 
 therefore, that the illumination intensity directly in front of 
 the searchlight will bear to the illumination intensity at five 
 miles a ratio equal to the ratio of the respective areas of the 
 beam at these points; this equals the ratio of the squares of 
 the radii of the beam at these points. In the case of the sixty- 
 inch searchlight, assuming that at five miles the beam has a 
 diameter of one thousand feet, this ratio would roughly equal 
 twenty-five thousand to one. It is possible, however, that 
 the dispersion could be reduced by a more careful attention to 
 this useless waste of energy. No necessity has yet arisen for 
 such a reduction in searchlights as now used, since it is desir- 
 able to illuminate the entire length of modern, five-hundred- 
 foot battleships at such distances. 
 
 Atmospheric Absorption 
 
 Some of the energy of the rays is absorbed in the atmosphere. 
 If the vibrating rates of the atmospheric gases are equal to 
 any of the vibration rates in the projected waves, part of the 
 energy of those particular waves will be absorbed. In this 
 connection it may be possible so to choose the electrode 
 materials for the arc that vibration rates produced in the arc 
 will not be equal to those of the atmospheric gases, thereby 
 evading the energy losses due to this cause. 
 
 In foggy or rainy weather the atmospfoeri% absorption would 
 be materially increased because water is not very diather- 
 manous. It is also true, however, that in such weathefbattle 
 ranges are materially decreased because of the decrease in the 
 limit of vision, which, in turn, is brought about by mist, rain, 
 or fog. It is difficult to foretell whether or not the two would 
 decrease at the same rate. 
 
 Receiving Radiant Heat Detectors 
 
 The development of sensitive radiant heat detectors has 
 followed several distinct lines corresponding to the varying 
 
INFRA-RED OR HEAT WAVES 47 
 
 phenomena of radiant energy in the form of heat waves whose 
 lengths are longer than 0.77 ju- The length of the visible waves 
 lies between 0.77 //, and 0.39 /*, those above 0.39 ju being in the 
 ultra-violet. 
 
 Those effects of radiant heat which have been used in the 
 production of sensitive detecting instruments may arbitrarily 
 be classified as follows: 
 
 i Volumetric expansion (chiefly of gases). 
 
 2. Thermoelectric currents. 
 
 3. Resistance change in electrical conductors. 
 
 4. Stresses in rarefied gases. 
 
 5. Linear expansion of solids. 
 
 As an example of the first may be mentioned the micro- 
 radiometer of Weber.* This instrument is a combination of 
 a differential air thermometer and a Wheatstone bridge. A 
 thin glass tube which contains at its center a drop of mercury 
 surrounded on both sides by a solution of zinc sulphate, con- 
 stitutes two arms of the bridge. Platinum electrodes sealed 
 in the bulbs at each end of the tube dip into the zinc sulphate 
 solution. One of the bulbs, which is made of an opaque 
 non-conducting material, and coated inside with lampblack, 
 is fitted with a fluorite window. When radiant energy enters 
 through the non-absorbing, fluorite window it is absorbed by 
 the contained gas and by the lampblack. Thus heated, the 
 gas expands and pushes the liquid toward the opposite bulb. 
 This changes the relative lengths of the mercury column and 
 of the solution between the platinum terminals; the balance 
 of the bridge being upset, a deflection of the galvanometer 
 consequently occurs. This instrument was stated to be 
 sensitive to a temperature change of one millionth of one 
 degree. 
 
 * Weber, Archiv. Sci. phys. et Nat. (3) 18, p. 347; 1887. 
 
48 RADIODYNAMICS 
 
 Thermoelectric Detectors 
 
 These radiant heat detectors may be divided into two groups, 
 namely, those in which the detector and the sensitive gal- 
 vanometer with which it is used are two separate and distinct 
 instruments, and those in which the two are combined into 
 a single instrument. The thermopile is representative of the 
 first group, and for the second we have the radiomicrom- 
 eter. 
 
 Let us first consider the thermopile. From the very be- 
 ginning of radiant energy measurements, the power of this 
 form of wave energy in the ether for developing electric 
 currents in circuits containing junctions of dissimilar metals, 
 has found wide application. Tyndall, Rubens, and other 
 pioneers in this domain secured very satisfactory results with 
 the thermopile, in spite of its great heat capacity. Rubens 
 has described* a linear thermopile consisting of twenty junc- 
 tions of iron and constantin wires about o.i mm. to 9.15 mm. 
 in diameter (resistance 3.5 ohms). When used with a gal- 
 vanometer having a figure of merit of i = i .4 X io~ 10 amperes 
 (resistance = 3 ohms, period = 14 seconds), a deflection of 
 one scale division indicated a temperature change of i.i X 
 icT 6 . A candle at five meters gave a deflection of 10 cm. or 
 250 cm. at one meter. The area of the exposed face of the pile 
 is about 1.6 cm. 2 . The heat capacity was such that its 
 stationary temperature was reached in less than seven 
 seconds. 
 
 If p = the thermoelectric power in microvolts per degree 
 ( = 53 X iQ" 6 volts for iron and constantin), n = the number 
 of junctions exposed, and r = the internal resistance, of the 
 thermopile; and if we combine the pile with a galvanometer, 
 which, with an internal resistance of w ohms, gives a deflection 
 of m millimeters per microampere, then a deflection of i mm. 
 
 * Rubens, Zs. fur Instrumentenkunde, 18, p. 65; 1898. 
 
INFRA-RED OR HEAT WAVES 49 
 
 indicates a change in temperature at the junctions of A/ degrees 
 when 
 
 npm 
 
 The highest efficiency is obtained when the resistance of the 
 thermocouple is equal to the combined resistance of the 
 connecting wires and of the auxiliary galvanometer. 
 
 Coblentz has described* a linear thermopile of bismuth- 
 silver junctions which had a heat capacity low enough to attain 
 ninety- two per cent of its maximum temperature in two 
 seconds. It has a completely opaque surface, this novelty 
 being secured by a series of overlapping receivers; it has a high 
 sensitivity; the materials are sufficiently strong to withstand 
 rough usage; it is quick acting, and yet sufficiently massive to 
 permit operation in the open without being disturbed by the 
 cooling effect of air currents. 
 
 The efficiency of the thermocouple is such that one micro- 
 watt of radiant power produces about 0.02 microvolt per 
 thermojunction in the thermopiles of bismuth-silver, or in 
 larger units i watt = 0.02 volt. 
 
 At present we have no exact knowledge of the mechanical 
 equivalent of the radiations of large searchlights, but for 
 sunlight we have accurate data. Upon the reasonable assump- 
 tion that we can develop searchlights which, with the aid of 
 collecting and concentrating means at the receiver, will produce 
 received effects at five miles equal to those produced by sun- 
 light without such concentrating means, we may proceed to 
 make calculations on a sunlight-source basis. Mr. W. W. 
 Coblentz has kindly made for the author the following cal- 
 culations on the current developed in a thermocouple with 
 sunlight as a source: 
 
 * Various Modifications of Bismuth-Silver Thermopiles Having a Con- 
 tinuous Absorbing Surface, Scientific Papers of the Bureau of Standards, No. 
 229, p. 132 
 
50 RADIODYNAMICS 
 
 The solar radiations reaching the earth's surface are about 
 i.o to 1.2 gr. cal. cmT 2 per minute = gV gr. cal. cm? sec." 1 , 
 or about -^ watt per cm. 2 per second. For a quick-acting 
 thermopile the receiver has an area of about 0.04 cmT 2 , so that 
 when exposed to sunlight the amount of radiant power inter- 
 cepted is 
 
 0.04 
 
 - = 0.003 watt. 
 
 This would produce 0.016 X 0.003 = 4& X io~ 6 volt, or a rise 
 in temperature of about one-half degree centigrade. 
 
 By increasing the number of couples to 100 and placing the 
 whole in vacuo, the sensitivity could be increased 200 times. 
 The e.m.f. developed would then be 200 X 48 X io~ 6 , or very 
 nearly o.oi volt. Within recent years relays have been 
 perfected which will operate with impressed voltages of 
 approximately 0.003 vo ^- A factor of safety is therefore, 
 apparent, since the received current is three times that re- 
 quired for operation. These rough calculations indicate that 
 heat-wave control systems are quite within the range of 
 possibility. 
 
 Our calculations are based on the assumption that we can 
 produce, at five miles, thermoelectric effects equal in magni- 
 tude to the effects produced at the earth's surface by the solar 
 radiations. It is probable that this assumption can be realized 
 in practice. Even if this were not possible, we have means of 
 increasing the received effects so that a much smaller heat 
 intensity at the receiver would produce the desired results. 
 The De Forest three-stage amplifier is capable of amplifying 
 minute received currents to from five hundred to a thousand 
 times their original strength. These amplifiers operate best 
 with pulsating or alternating received currents. It would be 
 a simple matter to use a current interrupter of either the 
 motor-driven or vibrating-buzzer type for breaking up the 
 direct current produced in the thermopile. This would intro- 
 
INFRA-RED OR HEAT WAVES 5 1 
 
 duce some complications. The thermopile and galvanometer 
 relay, with an auxiliary relay capable of handling larger 
 currents would, however, form a very simple and reliable 
 receiver. 
 
 - Radiomicrometers, Bolometers, and Radiometers 
 
 Three other well-known types of radiation detectors are the 
 radiomicrometer, the bolometer, and the radiometer. The 
 radiomicrometer, which was invented independently by 
 d'Arsonval and Boys, consists essentially of a moving-coil 
 galvanometer of a single-loop with a thermo-junction at one 
 of its ends. It is, as previously stated, a combined thermo- 
 couple and galvanometer. 
 
 The bolometer is simply a Wheatstone bridge, two arms of 
 which are made of very thin, blackened, metal strips of high 
 electrical resistance and high temperature coefficient, one or 
 both of which are exposed to radiation. 
 
 The radiomicrometer, because of its great delicacy, is not so 
 suitable for radiodynamics as the separate thermopiles and 
 galvanometer relays. The bolometer is an extremely sensitive 
 radiation detector, but careful precautions must be observed 
 in keeping at a constant temperature the air in which it is 
 contained so as to avoid the drifting of the zero position of the 
 auxiliary galvanometer. 
 
 The radiometer of Crookes, a scientific toy which may be 
 seen in many jeweller's windows, has been modified for radiant 
 energy measurement. Nichols* has described a radiometer 
 consisting of two blackened vanes of platinum attached to a 
 horizontal arm and suspended in a vacuum by a quartz fiber. 
 Although instruments of this type will detect a change in 
 temperature of one one-nr'llionth of one degree, their extreme 
 delicacy and sluggishness make them less suitable than ther- 
 mopiles for radiodynamics. 
 
 * Phys. Rev., 4, p. 297; 1897. 
 
RADIODYNAMICS 
 
 The Tasimeter 
 
 Edison's tasimeter consists essentially of a vulcanite rod and 
 a microphonic contact. The vulcanite rod, which has a high 
 coefficient of linear expansion, is made to exert a pressure on 
 the microphonic contact by means of a screw press. A slight 
 expansion of the rod, brought about by a slight increase in 
 temperature, causes a change in pressure on the microphonic 
 contact, and consequently a change in its resistance. When 
 the microphone forms one arm of a Wheatstone bridge the 
 apparatus becomes a very sensitive radiation detector. 
 
 Edison used a solid rod of hard rubber in compression 
 against two blocks of carbon, as shown in Fig. 17. Owing to 
 
 Compression Adjustment 
 
 Carbon Blocks 
 
 ff'IWI'U 
 
 ' Hard 'Rubber 
 Rod 
 
 FIG. 17. 
 Simple form of Edison's tasimeter. 
 
 the large mass of rubber in the rod and to the low coefficient of 
 conductivity of hard rubber, this form is sluggish in its action. 
 The author has modified this instrument in order to increase 
 its sensitiveness and to decrease the period required to attain 
 its maximum temperature under the action of a given intensity 
 of received radiation. The modification consists substantially 
 in substituting a sensitive telephone microphone of the carbon 
 granule type for the blocks of carbon, and in replacing the 
 hard-rubber rod with a thin strip of hard rubber. This strip 
 is maintained in tension by an adjustable spring. The arrange- 
 ment of microphone, hard-rubber strip, and adjusting screws 
 is shown in Fig. 18. 
 
INFRA-RED OR HEAT WAVES 
 
 53 
 
 This instrument, when connected in circuit with a battery 
 and ammeter, will readily indicate a change in current sufficient 
 to operate a relay, if influenced by the heat of a bunsen burner 
 at a distance of one meter. Although not lacking in sensitive- 
 ness, heat detectors of this type are subject to vibration, jars, 
 and sounds; a weakness which disqualifies them for radio- 
 dynamics, particularly in the radiodynamics of torpedo control. 
 
 Thermostats 
 
 In an endeavor to provide a sensitive, quick-acting, heat- 
 detecting instrument which will close a circuit directly without 
 
 FIG. i 8. 
 
 the aid of a sensitive relay, the author has experimented with 
 various types of thermostats. After experimenting with 
 mercury-in-glass thermostats designed especially for quick 
 action, with composite-strip thermostats of both the straight 
 and spiral types, and with alcohol, mercury, and gas ther- 
 mometers, the conclusion was reached that a modification of 
 the differential gas thermometer would be far more suitable 
 than the other types, because of its high sensitiveness and 
 rapidity of action. 
 
 The most satisfactory form thus far produced is shown in 
 Fig. 19. The general scheme of this type of instrument was 
 suggested to the author by Prof. E. S. Ferry, of Purdue 
 
54 
 
 RADIODYNAMICS 
 
 University. In the drawing A is the heat absorber of thin, 
 lampblacked platinum, B and Z>, two glass gas-chambers, 
 connected by a glass tube of small bore; B is lampblacked 
 inside. M is a thread of mercury; W-W are water or alcohol 
 columns whose function is to prevent the mercury from moving 
 by jumps under the action of the expanding gas in B ; and C-C 
 are contact wires of platinum sealed into the glass tube so as 
 to make contact with the mercury thread. 
 
 If heat rays fall upon the platinum disc A, they are absorbed 
 and their energy appears as a rise in the temperature of A. 
 
 FIG. 19. 
 
 A has a very small heat capacity because of its low specific 
 heat and its thinness, and it therefore requires but a small 
 amount of heat energy to raise its temperature. Since 
 platinum has a high coefficient of thermal conductivity, the 
 heat is rapidly conducted to the gas in the closed chamber B. 
 This chamber is so designed that the distance from the plati- 
 num to any part of the enclosed gas is small, in order that the 
 conduction-time-lag through the gas may be a minimum. 
 The inside of B is lampblacked in order to prevent escape of 
 heat by radiation through its walls. The temperature of the 
 
INFRA-RED OR HEAT WAVES 55 
 
 enclosed gas therefore rises rapidly to the temperature of the 
 absorber. This gas, which is especially chosen for its maxi- 
 mum coefficient of volumetric expansion and its minimum 
 specific heat, expands and pushes back the liquids in the tube. 
 The mercury M will then short-circuit the two Connecting 
 wires C-C, thus closing the external circuit. Jf the source of 
 heat be removed the order of actions is reversed. The^ ab- 
 sorber A then becomes a rapid and efficient radiator, and the 
 heat of the gas in B is dissipated through conduction to and 
 radiation from the platinum disc A. Any change in normal 
 temperature, i.e., the temperature of the air in which both A 
 and D are contained, will not ca'use any appreciable change 
 in position of the mercury thread because pressures will be 
 produced in the two gas chambers which are equal and opposite. 
 
 W-W consist of some non-conducting liquid of low specific 
 gravity. Without some such steadying means the mercury 
 thread will have a tendency to move in jumps. The specific 
 gravity should be low in order that a slight difference in the 
 levels of the two columns will not require a great difference in 
 pressure between A and D. If mercury were used a relatively 
 large difference in pressure between A and D would be neces- 
 sary in order to produce a motion of M sufficient to bridge 
 contact wires just above the normal position of the mercury 
 surface. 
 
 The author has constructed thermostats of this type which 
 will operate satisfactorily in strong sunlight, an exposure of 
 from one to five seconds being sufficient to produce the maxi- 
 mum deflection of the mercury thread. The complete periods 
 were considerably less than twice these values. 
 
 These results can probably be improved upon. The author 
 has experimented with gases containing vapors of alcohol, 
 ether, carbon tetrachloride, and similar liquids whose satu- 
 rated vapors have a high coefficient of volumetric expansion. 
 The results of these experiments are promising. 
 
56 RADIODYNAMICS 
 
 A differential gas thermostat of this type, if developed to the 
 proper sensitiveness, would be as near the ideal of a heat-wave 
 receiver as we can hope to reach. Its extreme simplicity and 
 ruggedness, and the absence of the usual sensitive relay are 
 its chief advantages; all other types of receiving apparatus, 
 whatever the nature of the control energy, require, besides 
 various other apparatus, a sensitive relay, usually of the 
 galvanometer type; this, in turn, requires a more rugged relay 
 for handling the electrical energy used in performing the 
 various operations aboard torpedoes. 
 
 Heat-wave control systems, we may therefore state, are 
 not only within the range of possibility but of probability as 
 well. The extreme simplicity and ruggedness of both trans- 
 mitter and receiver, the absence of masts and other aerial 
 targets, the satisfactory solution of the interference problem, 
 the near approach to complete invisibility of both transmitter 
 and torpedo, and the almost indiscernible form of the control 
 energy are factors that commend heat waves as a connecting 
 link between the shore and the wirelessly directed torpedo. 
 
CHAPTER VIII 
 VISIBLE AND ULTRA-VIOLET WAVES 
 
 As shown by the table, the visible waves vary in frequency 
 from 8000 billions down to 4000 billions per second, repre- 
 senting the various colors from violet, down through blue, 
 green, and yellow to red, and including all the thousands of 
 intermediate shades. Quite apart from the various optical and 
 chemical effects these waves are capable of producing, their 
 chief interest to us lies in their ability to effect changes in 
 the electrical characteristics of various substances. These 
 changes can be utilized for the operation of delicate indi- 
 cating or relaying instruments. Selenium, described more 
 fully in a subsequent chapter, is the most important of these 
 substances affected by light. Systems of telegraphy and 
 telephony based on its peculiar property of changing its 
 electrical resistance under the influence of light, have occupied 
 the attention of numerous scientific men since Willoughby 
 Smith's discovery of that property in 1875. But no ac- 
 count of its application to torpedo control can be found. 
 The writer ventures here to present some experimental data 
 and observations made by him, especially in view of the 
 possible adoption of a light-wave selenium control system for 
 the Hammond dirigible torpedo. 
 
 From a number of selenium cells of varying types and re- 
 sistances two, made by Dr. Korn of Vienna, were chosen. 
 The sunlight and dark resistances of one of these were 2000 
 and 5200 ohms respectively; of the other, 1300 and 3000 ohms. 
 These two cells were to be used in selective light telegraphy 
 tests and it was decided first to learn the applied e.m.f. 
 
 57 
 
58 RADIODYNAMICS 
 
 range in which the cells operated with the greatest sensitive- 
 ness and smallest inertia. The 2000-5200 ohm cell was first 
 tried. It was connected in a series circuit with a battery and 
 microammeter, and with a potentiometer for accurate regu- 
 lation of potentials. The operation was much better with 
 
 l( 1 
 
 
 
 
 
 
 / 
 
 
 'I/ rHl 
 ^L--20crn-->{r l 
 
 ~~ /&~| | T 
 
 Yotf- Sensitiveness Cut 
 Horn Selenium Celt No. 
 
 I* 
 
 
 2 
 
 
 
 j 
 
 've 
 
 / 
 
 / 
 
 
 / 
 
 1 
 
 / 
 
 
 / 
 
 
 
 
 
 
 
 / 
 
 
 A 
 
 f 
 
 
 
 
 
 
 
 / 
 
 / 
 
 / 
 
 
 
 
 
 
 i 
 
 ,/ 
 
 $ 
 
 / 
 
 
 
 
 
 
 
 3 
 
 / 
 
 f 
 
 
 
 
 
 
 
 
 / 
 
 Z 
 
 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 
 
 
 
 / 
 
 {/ 
 
 P 
 
 
 
 
 
 
 
 
 i2 
 
 / 
 
 
 
 
 
 
 
 
 / 
 
 * / 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 1.4 VOLT DRY CELLS 
 FIG. 20. 
 
 the highest current density permissible with the micro- 
 ammeter, so it was decided to use a milliammeter and higher 
 potentials. Potentials as high as 25 volts were used and the 
 results are given in the curve of Fig. 20. This shows graphi- 
 cally the relation between applied voltage, current, resist- 
 ance, and the current change between light and darkness. 
 
VISIBLE AND ULTRA-VIOLET WAVES 59 
 
 The current change is the factor of principal importance. It 
 is noted that this value increases directly as the voltage, and 
 that the highest value corresponds to the highest value of cur- 
 rent density permissible. It was learned that if this exceeded 
 five or six milliamperes for an hour or more, the cell would 
 get out of order, and a telephone inserted in the circuit in- 
 dicated that the current was varying at a rate of several 
 thousand per second. The sound was an irregular hissing 
 or frying noise, resembling closely the sounds in a radio re- 
 ceiver due to heavy atmospherics, or that heard in an ordi- 
 nary telephone during the progress of a thunder storm in 
 the immediate vicinity. The tests were made with a i6-c.p. 
 carbon filament electric light, at a distance of 50 cm. 
 
 The 1300-3000 ohm cell, which we shall designate No. 2, 
 was given a similar test under slightly different conditions. 
 The i6-c.p. light was placed at a distance of 10 feet, and a 
 five-inch condensing lens was used to increase the intensity 
 of illumination on the active surface of the cell. The curves 
 of Fig. 21 show the relation between the e.m.f., current, 
 resistance, and current change with the No. 2 cell. 
 
 By a comparison of Figs. 20 and 21, it is readily observed 
 that the combination of the No. 2 cell with the condensing 
 lens was more sensitive at a distance of 10 feet than the 
 No. i cell at about 20 inches. The indications of the milli- 
 ammeter showed that when the cells were brought from 
 darkness to light the resistance dropped very quickly to 
 about two-thirds the resistance change value, and then to 
 the lowest value in about five or ten seconds. From this 
 observation it is obvious that the cell would operate much 
 more efficiently for slow variations in the light intensity than 
 for rapid, such as are used in light telephony, since the lagging 
 part of the current change represented by the change occur- 
 ring, say one-fiftieth of a second, after a given change in illumi- 
 nation, would be of no value where the variation frequency 
 
6o 
 
 RADIODYNAMICS 
 
 exceeded 50 per second. It would seem, therefore, that the 
 higher characteristics of the human voice, which may reach 
 vibration rates of five thousand per second, would be re- 
 produced with less distinctness than the lower characteristics. 
 This is actually the case with light telephony as well as with 
 
 I 
 
 
 
 
 
 
 
 f" 
 
 
 
 % 
 
 ^-" 
 v 
 
 
 
 --/o> 
 
 ~t 
 
 
 . ) 
 
 p 
 
 s 
 
 
 p. 
 
 //- Se 
 Corn i 
 
 nsiti 
 Se/en 
 
 tenes 
 'um (. 
 
 s Cur 
 
 ;<?// A 
 
 
 
 
 / 
 
 
 Vo 
 / 
 
 ye 
 
 O.I 
 
 
 / 
 
 
 
 
 
 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 / 
 
 ^ 
 
 
 
 
 
 
 
 
 y 
 
 / 
 
 
 
 X 
 
 
 
 
 
 & 
 
 7 
 
 
 x 
 
 X 
 
 
 
 
 
 
 / 
 
 
 ^ 
 
 x 
 
 
 
 
 
 
 / 
 
 
 ^ 
 
 
 
 
 
 
 
 / 
 
 
 X 
 
 
 
 
 
 
 
 / 
 
 ^ 
 
 / 
 
 
 
 
 
 
 
 A 
 
 // 
 
 ? 
 
 
 
 
 
 
 
 
 /s 
 
 * 
 
 
 
 
 
 
 
 
 
 2 4 6 8 1C 
 
 1.4 VOLT CELLS 
 FlG. 21. 
 
 wire telephony, the inertia being too great to permit the re- 
 sistance to follow the rapid variations in the light's intensity; 
 in wire telephony, however, the inertia is mechanical rather 
 than electrical, except in very long distance lines where 
 the capacity comes into play. It was also found that the 
 cells were more sensitive in some parts of the active surface 
 
VISIBLE AND ULTRA-VIOLET WAVES 6 1 
 
 than in others. The five-inch condensing lens was used to 
 furnish a small spot of intense light with which to explore 
 the surface of the cell. When the diameter of this spot was 
 made about one-sixteenth of an inch the best results were 
 secured. The surface of the selenium, to secure this con- 
 dition, was placed very near the focal point of the lens. The 
 sensitive spots could then be accurately located, and in some 
 places the cell was several times as sensitive as in others. Ex- 
 perimental investigations of the cause of this spot sensitive- 
 ness were not made. 
 
 A test was made with the No. 2 cell with a view to de- 
 termine the possible value of a selective light signalling and 
 radiodynamic control system devised by the writer. In these 
 tests a light interrupter was used to effect the periodic illumi- 
 nation and darkening of the selenium cell at rates up to 300 
 per second.* 
 
 Fig. 22 illustrates the general arrangement of the apparatus 
 as well as the results of the test. The light used was a 4-c.p. 
 carbon filament electric lamp, and the interrupter was, as 
 shown, similar to those used on motion picture machines; it, 
 however, had a much larger number of blades, and, as shown 
 in the sketch, was attached to the shaft of a fan motor. A 
 rheostat and tachometer permitted variation and exact 
 knowledge of the speed, so that any desired interruption fre- 
 quency could be obtained. 
 
 At the receiver a three-inch condensing lens was used with 
 the selenium cell. The latter was connected in circuit with 
 a variable battery, a milliammeter, and the primary of a 
 ferric transformer. The secondary of the transformer was 
 connected in a series oscillatory circuit with another lumped 
 inductance and variable capacity of the Korda type. By 
 varying the capacity or inductance of this circuit, it could be 
 
 * In later tests an arc light supplied with alternating current of from 60 to 
 600 cycles was used to furnish the periodically fluctuating light. 
 
62 
 
 RADIODYNAMICS 
 
 brought into resonance with the periodicity of the light in- 
 terruptions. The oscillatory currents set up in this circuit 
 by the action of the pulsating currents in the selenium cell 
 circuit, when the two were in resonant operation, were less 
 than one ten-millionth of an ampere. Since that value did 
 not give a satisfactory telephone signal, an amplifying device, 
 
 60 
 
 Motor ,/ Light Interrupter 
 
 Amplifier 
 
 \ ' 
 Forced 
 
 I I 
 Resonance 
 
 45 
 
 15 
 
 50 100 ISO 200 250 300 350 400 450 
 
 FREQUENCY 
 
 FIG. 22. 
 
 Resonance curve of the author's selenium, selective, light-wave, radiodynamic 
 
 control system. 
 
 due to F. Lowenstein, was used, which, in turn, effected the 
 operation of a telephone of the high resistance type, and a 
 microammeter. With this arrangement the telephone signals 
 received when the circuit and light interrupter were in res- 
 onant operation, were very loud, the interruptions of the 
 light being heard as a clear musical tone, equal in frequency 
 to the frequency calculated from the speed of the motor, and 
 
VISIBLE AND ULTRA-VIOLET WAVES 63 
 
 the number of blades on the shutter disc. The microam- 
 meter indicated a signal current of about 50 microamperes 
 as the maximum value for resonance adjustment. 
 
 The selectivity was so good that it was difficult to keep 
 the apparatus in tune, owing to slight variations in the speed 
 of the shutter motor caused by changes in the line voltage. 
 The resonance curve shows very clearly the degree of selec- 
 tivity obtained. It is observed that when the frequency of 
 the interruptions was between zero and about one hundred, 
 the selective circuit was forced to vibrate in its own period, 
 and the effect on the receiving instrument was practically as 
 great as when the purely resonant operation occurred. The 
 amplitude of the superimposed fluctuating current, on the 
 normal dark current in the selenium cell circuit, which con- 
 tained a considerable ferric inductance, was probably 50 times 
 greater for frequencies, say, below 100, than for frequencies 
 in the neighborhood of 800, the natural frequency of the 
 circuit during the test. This accounts for the fact that the 
 indicator currents were so high with the forced operation in 
 comparison with those of resonant operation. The funda- 
 mental vibration rate of the circuit at this particular setting 
 was approximately 345 complete periods per second. The 
 selectivity no doubt could have been improved upon by a 
 better design of light interrupter, which would produce 
 sinusoidal variation in the selenium cell's resistance, and by 
 using inductances of less resistance than were used in this 
 experiment. 
 
 Several electric lights in the room, including a mercury-arc 
 light and a gas light placed directly between the condens- 
 ing lens and the four candle-power light, did not mate- 
 rially affect the strength or quality of the received signals. 
 It is believed that the distance could have been considerably 
 increased or the intensity of the signal light reduced without 
 greatly reducing the strength of the received signals. The 
 
64 RA DIOD YNA MICS 
 
 telegraphic signals were sent with an ordinary Morse key 
 connected in series with the signal light and operated according 
 to the Morse or International codes. 
 
 Selective mechanically tuned elements were also tried in 
 place of the tuned electrical circuit and an even greater 
 degree of selectivity was obtained. These tuned elements 
 were of the different types due to Ruhmer, Lowenstein, and 
 Pickard, and are known as monotone amplifones or selective 
 reed relays. A complete description is not permissible. 
 The electrical circuit is, however, to be preferred for practical 
 use on account of its nice adjustability and ease of handling, 
 and because it is not, like the amplifone, subject to jars or 
 sounds. 
 
 Although these experiments were rather encouraging, later 
 tests made with a 24-inch searchlight during both day and 
 night furnished conclusive evidence that, with the illumina- 
 tion furnished by such a source of light and the best selenium 
 cells procurable, the operative range of such a system would 
 not exceed a distance of one mile. The tests were therefore 
 discontinued. 
 
 Ultra-violet Radiations 
 
 Ultra-violet light has a powerful effect in facilitating the 
 discharge of electrons from negatively charged conductors and 
 entirely overcoming the hindrance ordinarily experienced. 
 The light waves may be conceived as shaking up the neutral 
 atoms condensed around the electrons and setting the latter 
 free. 
 
 Suppose we have two conductors in a vacuum at a difference 
 of potential of, say, 300 volts, and that ultra-violet rays are 
 made to fall on the negative wire. The negative electrons 
 are set free to enter the vacuum and are repelled by the 
 negative conductor; at the same time they are attracted by 
 the positive conductor, and, since nothing prevents them, 
 
VISIBLE AND ULTRA-VIOLET WAVES 65 
 
 they pass from one to the other, and with a velocity of ap- 
 prpximately 6600 miles per second. (See Fig. 23.) 
 
 Although we have but little data oh distances at which these 
 effects are observable, we know that they can be brought 
 about, and a method of applying this property of ultra-violet 
 light at once suggests itself. 
 
 In the "Electrische Zeitung," July, 1898, Prof. E. Zickler 
 proposed to use this property of these radiations for teleg- 
 
 JOOVb/fs- 
 
 Vacuum Tube 
 
 < 
 
 Ultra Violet ftays 
 
 FIG. 23. 
 
 raphy. He succeeded on a small scale and believed that with 
 a 25-ampere lamp and suitable reflectors, good results were 
 possible over several kilometers. His proposal was based on 
 an at first inexplicable phenomenon observed by Hertz. In 
 the course of his experiments on resonance it was observed that 
 the intensity of the sparks at the detector was greatly increased 
 by placing a screen between it and the exciting spark. Later 
 the curious effect was attributed solely to the ultra-violet 
 rays emitted by the exciting discharger. 
 
 For such simple apparatus as that required for exploding 
 mines where but one energy impulse is required, the problem 
 is quite simple, requiring only an electric explosive cap in 
 series with the battery and vacuum tube. When the rays 
 are directed upon the negative electrode of the vacuum tube 
 the discharge occurs immediately, and the current passing 
 
66 RA DIOD YNA MICS 
 
 through the electric cap effects the ignition or detonation of 
 the explosive material. 
 
 For the control of complicated mechanisms, however, a 
 relay and some form of electric switching device are required. 
 
 Although ultra-violet rays have all the advantages of in- 
 audibility, invisibility, and simplicity of generation and 
 reception, in view of the known fact that the earth's atmo- 
 sphere very strongly absorbs their power it is very probable 
 that operation of such a system at useful distances would 
 involve considerable difficulty. 
 
CHAPTER IX 
 EARTH CONDUCTION 
 
 The early suggestions of Steinheil and their subsequent 
 application, as previously outlined, have been quite seriously 
 considered as a very simple substitute for the comparatively 
 complicated Hertzian wave apparatus for torpedo control. 
 In view of the simplicity of the apparatus, and of the all- 
 important problem of selective control offered by earth con- 
 duction telegraphy, experiments were conducted by the writer 
 in Gloucester harbor (1912) to determine roughly its value for 
 torpedo control. 
 
 The general plan of the proposed system is shown graphi- 
 cally in the accompanying drawing (Fig. 24). 
 
 Heavy insulated wires lead from the submerged plates to 
 the station, where, by means of a switchboard, heavy currents 
 can be sent between any two of the plates. The six different 
 current fields thus capable of production are shown in the 
 drawing by the curved lines. 
 
 The receiver comprises two conducting plates, one at the 
 bow of the torpedo and one at or trailing behind its stern, with 
 insulated conducting wires extending inside to the terminals 
 of a sensitive relay. The necessity of so many current fields 
 is obvious when we consider that we must be able to operate 
 the sensitive relay regardless of the torpedo's direction of 
 motion, i.e., the direction of the current field must correspond 
 to the direction of extension of the receiving plates. If means 
 for accomplishing this are not provided it is possible to lose 
 control of the vessel altogether. 
 
 The first thing to determine in these experiments was 
 
 67 
 
68 
 
 RADIODYNAMICS 
 
 whether or not it would be possible to transmit enough energy 
 to the receiver, at distances up to the limit of vision, with 
 
 Submerged 
 
 Submerged 
 P/ofe '. 
 
 Submerged 
 .' Plate 
 
 Connecting 
 WiretoP/at 
 
 'horeLirre 
 
 Connecting Wires 
 to PlaJes 
 
 Submerg 
 Plate 
 
 Control Station 
 FIG. 24. 
 
 apparatus of such power as is consistent with practical con- 
 siderations. With this end in view, apparatus was arranged 
 as shown in the sketch (Fig. 25), which also shows the received 
 
 20 A mp. 
 
 ^> 
 
 200 
 
 200 
 
 Distance between Sending Pis. 200 Ft. 
 - Receiving 8 Ft. 
 
 , Area each - - tOSq.Ft. 
 
 * " " Sending " 25 " - 
 
 10 Received current values snown arelO' f Ampt 
 
 FIG. 25. 
 
 currents indicated by the Weston microammeter, for different 
 positions of the receiving boat in the current field. This in- 
 strument was connected directly between the receiving plates 
 
EARTH CONDUCTION 69 
 
 which were fastened at the bow and stern of a ten-foot row 
 boat. 
 
 The transmitter consisted of a 25-volt, 1 2o-ampere-hour 
 storage battery connected to two copper plates having an 
 effective area of about 25 square feet each. One of these 
 plates was the regular ground for the radio set aboard an 8-ton 
 house boat; the other was fixed to the bottom of a row boat, 
 moored about 200 feet distant. 
 
 A No. 1 6 bell wire, extending from the mast of the house 
 boat down to the row boat, served to complete the circuit. 
 The results are clearly shown in the drawing. 
 
 A very curious, unexplained phenomenon was observed dur- 
 ing these tests. The readings were taken on transmitted im- 
 pulses of about two seconds duration, with intervening periods 
 of rest, and were made in response to signals, by an assistant 
 on the house boat, who opened and closed the circuit between 
 the battery and the overhead line wire extending to the distant 
 sending plate. In this way one battery terminal was con- 
 tinually connected to the earth plate beneath the boat. 
 
 In addition to the usual earth current these currents may 
 be found almost anywhere on the surface of this earth and 
 were in this case of course independent of the sending battery 
 readings were obtained which varied up to 50 microamperes 
 according to the position and direction of extension of the 
 receiving boat; and these readings gradually increased as the re- 
 ceiving boat neared the house boat. When the house boat plate 
 was disconnected from the storage battery the received current 
 dropped to the normal value. Signals could thus be sent, up 
 to distances of about 50 feet, simply by making and breaking 
 the connection between the battery and plate, with absolutely 
 no current flowing from the battery. The arrangement of 
 apparatus and results are shown in Fig. 26. 
 
 The following day (Sept. i, 1912) further tests were made 
 with increased power and distance. The transmitting current 
 
70 RADIODYNAMICS 
 
 was obtained from a bank of four no- volt, 5o-ampere mer- 
 cury-arc rectifiers, regulated by a series resistance, and meas- 
 ured by an ammeter. The respective areas of the transmitting 
 
 (< sort 
 
 Copper Ground 
 Plate 
 
 FlG. 26. 
 
 plates as well as the distances between them are given in the 
 drawing (Fig. 27). The leads to the earth plates of the 
 transmitter were composed of twenty-foot sections of No. 20 
 copper strips, one and a half inches wide, soldered together. 
 
 //Ov. D.C. 
 
 50 Amp. 
 
 400 
 
 100 
 
 400 
 
 '35 
 
 10 
 
 Distance between Sending Plates 400 Ft. 
 Receiving " 8 Ft. 
 
 A rea each " " 8 Sa. Ft. 
 
 Sending " . SO & 15 Sq.Ft. 
 Received current values shown are 10' 6 Amp. 
 
 FlG. 27. 
 
 The maximum current obtainable was 50 amperes, and the 
 received current values shown were secured with this trans- 
 mitting current. 
 
 On Oct. 3, this distance was again increased, this time to 
 approximately 1000 yards between the sending plates. Fig. 
 28 shows the results of this test. 
 
EARTH CONDUCTION 
 
 The received current values resulting from these different 
 tests show that, in a line between the two transmitting plates, 
 the position of the receiving plates for weakest signals is 
 midway, and for strongest, nearest either of the plates. With 
 
 Distance between Sending Pis. 3000 Ft 
 Receiving ~ 30 Ft. 
 
 Area each " - BSaFt. 
 
 " . " Sending - SO " "' 
 
 Received current values shown are 10 Amp 
 
 FIG. 28. 
 
 a transmitting power of over five kilowatts, the received 
 currents at the position of minimum signal strength (which is 
 the distance determining factor), that is, at a distance of less 
 than one-half mile, were no more than sufficient to operate a 
 
 FIG. 29. 
 
 sensitive relay of the type necessary for torpedo control. 
 These results were far from encouraging and the tests were 
 discontinued. 
 
 A somewhat different scheme due to Mr. H. Christian 
 Berger, an electrical engineer of New York, was next tried. 
 The transmitting energy was a high-frequency oscillatory cur- 
 
72 RADIODYNAMICS 
 
 rent, and the receiver was of the regular radio type, with two 
 earth connections instead of one earth and one aerial. 
 
 The earth plates were of copper, 100 and 25 square feet 
 area, respectively, separated 400 feet, and connected, as shown, 
 in series with the oscillating condenser circuit. A hot wire 
 ammeter in this circuit indicated a current of four amperes. 
 The primary energy was delivered to the condenser circuit by 
 a 3-kw., 1 10 to 20,000- volt, 6o-cycle transformer. (See Fig. 29.) 
 
 The receiver connections, illustrated in the small drawing 
 at the right, are those of a common radiotelegraphic receiver 
 with the exception previously noted. The distance between 
 the receiver grounds was about 250 feet, and the distance 
 between the two sets of grounds of transmitter and receiver 
 was estimated at about 500 feet. 
 
 Dr. L. W. Austin, head of the U. S. Radio Laboratory in 
 Washington, D. C., has shown that a radiotelegraphic sender 
 may exert a very large amount of power for the brief periods 
 of time during which the condenser discharges. 
 
 Considering a condenser of 0.04 mfd., charged to a potential 
 
 of 10,000 volts, Q is equal to ^^ X 10,000, or 0.0004 
 
 1,000,000 
 
 coulomb. 
 
 The work done in charging such a condenser to that poten- 
 
 ,. , . V 2 C io,ooo 2 X 0.00000004 i 
 
 tial is equal to , or - , or 2 joules, or 
 
 2 2 
 
 0.737 X 2 = 1.47 foot-pounds. It can furnish that amount 
 of energy in one discharge. 
 
 If the condenser is discharged through a circuit of such self- 
 inductance as will give a wave length of 1000 meters, the 
 oscillation frequency will be 300,000, and the alternations 
 600,000 per second. 0.0004 coulomb will create an average 
 current of 0.0004 X 600,000 = 240 amperes. Were the wave 
 length much shorter the current would be correspondingly 
 greater, as is shown by the following example. 
 
EARTH CONDUCTION 73 
 
 The above condenser is discharged through an inductance 
 which will give a wave length of 500 meters. The alternation 
 frequency of this circuit would then be 1,200,000 per second. 
 0.0004 coulomb will create an average current of 0.0004 X 
 1,000,000 = 480 amperes. By this we see that the current 
 in an oscillatory circuit is inversely proportional to the wave 
 length. 
 
 If the energy of 2 joules stored in the condenser is radiated 
 in five complete oscillations, the rate of doing work, if the effi- 
 ciency of conversion is unity, is 2 joules in - - second = 
 
 600,000 
 
 240,000 per second = 240 kw. This shows very clearly that 
 although the available energy is very small, the rate of doing 
 work, i.e., the power of a wireless telegraph sender, may be 
 very great for a short period of time. 
 
 This peculiarity of a condenser discharge is, no doubt, the 
 basis for Mr. Berger's suggestion. The scheme is distinctly 
 novel, utilizing, as it does, oscillating currents of high fre- 
 quency, but with earth conduction, and not etheric radiation, 
 as the means of transferring the energy. 
 
 The tests given this system were very severe in that the 
 conditions imposed were far from ideal; but at that time it 
 was believed that if no telephone signals could be received 
 under those conditions, the system would be valueless for relay 
 operation. No signals were received during this test and the 
 experiments were discontinued. 
 
CHAPTER X 
 
 ELECTROSTATIC AND ELECTROMAGNETIC INDUC- 
 TIONHERTZIAN WAVES 
 
 Following the scheme of Dolbear, the author experimented 
 with electrostatic induction as a possible means of torpedo 
 control at the Hammond Radio Research Laboratory in 1912. 
 
 The transmitter consisted of a ioo,ooo-volt transformer, 
 especially built for the purpose, energized by alternating 
 current of from 60 to 1000 cycles. One terminal of the 
 
 ill 
 
 Detectoi 
 
 ffeceiver 
 
 FlG. 30. 
 
 secondary was grounded, and the other connected to the 
 station antenna,* which was insulated for 1,000,000 volts with 
 Electrose strain insulators. 
 
 The receiving apparatus consisted of an antenna,f on the 
 house boat Pioneer, connected to a very sensitive form of 
 potential operated radio detector which will be more fully 
 described in a subsequent chapter. 
 
 Fig. 30 shows schematically the circuit arrangements. 
 
 * 300 ft. high, 400 ft. flat top. t 30 ft. high, 20 ft. flat top. 
 
 74 
 
HERTZIAN WAVES 
 
 7S 
 
 The curve, Fig. 31, was made by taking readings of the re- 
 ceived currents, as indicated by a Weston microammeter, 
 connected to the receiving detector. The Pioneer was 
 started seaward within about a hundred feet of the trans- 
 mitting station, and readings taken every minute near the 
 shore; after the steeper part of the curve had been passed 
 the readings were taken at longer intervals. 
 
 1000 2000 
 
 Dl STANCE IN FEET 
 
 FIG. 31. 
 
 An attempt at tuning the receiver to the frequency of the 
 alternating current used at the transmitter was made by in- 
 troducing a variable inductance of large value and a tuning 
 condenser in series with the antenna, and connecting the de- 
 tector to a point of maximum potential in this circuit. Fig. 
 32 shows this circuit. 
 
 The transmitter was then changed to the regular radio 
 type, the wave length being pushed far beyond the natural 
 period of the antenna by means of loading inductances in 
 both open and closed circuits. This was done to increase 
 the potential to the highest possible value, in order to in- 
 
76 RADIODYNAM1CS 
 
 crease the distance of operation. With an emitted wave 
 length of about 3000 meters, the potential was slightly in- 
 creased over that previously obtained with the transformer, 
 but no material increase in received results was noted. 
 The group frequency was 120 per second. In order to meas- 
 ure the electrostatic effects alone, no tuning to the high fre- 
 quency oscillations was attempted, the receiving antenna being 
 
 111 
 
 Antenna 
 
 Iron Core Inductance 
 
 Poteritio 
 Detector 
 
 -*= Earth 
 
 FIG. 32. 
 
 connected only to the detector. The low-frequency tuned 
 antenna circuit was then substituted for this receiver as be- 
 fore. The curve obtained in the best series of tests (Fig. 31) 
 shows that with our very sensitive relay operating under 
 working conditions, the maximum range would be only 
 up to about 1000 feet, a distance far too short for torpedo 
 operation. This method of control was also abandoned. 
 
 Electromagnetic Induction 
 
 Beyond the work of Preece, Trowbridge, Edison, and 
 others, already mentioned, very little has been done in the 
 field of electromagnetic induction for radiodynamics. 
 
 The fact that the transmitting and receiving coils or line 
 wires must be in parallel planes is one of the chief objections 
 to this system for transmitting energy impulses to a movable 
 boat. 
 
 The writer has not been able to find any accounts of work 
 
HERTZIAN WAVES 77 
 
 along this line. Although the difficulties are not in them- 
 selves insuperable, from a practical point of view they have 
 been considered too great in comparison with other systems. 
 No effort therefore has been made to utilize electromagnetic 
 induction as a means of controlling dirigible, self-propelled 
 vessels. 
 
 Hertzian Waves 
 
 Hertzian waves, as every one knows today, are by far the 
 most important means of wirelessly transmitting energy, 
 either for the communication of intelligence or for the con- 
 trol of self-acting apparatus of whatever nature. We are 
 not so much interested in presenting historical matter per- 
 taining to the very large amount of work done since Marconi's 
 first experiments; nor do we wish to burden the reader with 
 detailed theoretical or practical considerations of the many 
 phases of the radio signalling art, it being assumed that he is 
 sufficiently acquainted with the art as it now stands to under- 
 stand the accounts of its special application in the compara- 
 tively new field of radiodynamics; these special applications 
 will be hereinafter described in sufficient detail to be readily 
 understood by those possessing some knowledge of electricity. 
 
CHAPTER XI 
 
 THE ADVENT OF WIRELESSLY CONTROLLED 
 TORPEDOES 
 
 Although the subject of this chapter does not include 
 torpedoes in general, it is nevertheless important to have 
 some knowledge of the ordinary torpedo, and some facts 
 pertaining to its advantages and disadvantages, if we wish to 
 obtain a clear conception of the wirelessly controlled weapon 
 now being perfected for modern naval warfare. 
 
 The torpedo is claimed to be an American invention, being 
 said to have sprung from the fertile brain of Benjamin Franklin, 
 who, during the Revolution, experimented with this then un- 
 heard of method of marine attack. The first attempt in war 
 of which we know was made in the harbor of Brest, on the west 
 coast of France, in 1801, under the orders of Napoleon. This 
 first test under actual war conditions was made by an Ameri- 
 can, Robert Fulton, the father of steam navigation. Fulton 
 used a submarine boat, the drawings and designs of which 
 have never been published. He is said to have obtained con- 
 siderable success in his experiments, but he failed in an 
 attempt to blow up an English man-of-war, whereupon 
 Napoleon withdrew his support, and the scheme was not 
 carried into practical operation. 
 
 We next hear of torpedoes in the Russian war of 1854, 
 when one of prodigious power was exploded in the harbor of 
 Cronstadt, through copper wires connecting with a galvanic 
 battery on shore. 
 
 Again, during our own Civil War, the torpedo made its 
 appearance in improved form. It was employed for harbor 
 
 78 
 
ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 79 
 
 defense chiefly in and around Charleston. It was also used 
 with deadly effect during the Spanish-American War, and, in 
 the hands of the Japanese, inflicted great damage to the 
 Russian fleet in the battle of Fuishima Straits. 
 
 Every modern battleship is equipped with from two to 
 four torpedo tubes. The United States alone has over 60 
 
 FIG. 33. 
 The first Holland submersible. 
 
 torpedo boat destroyers, 30 torpedo boats, and 50 subma- 
 rines, representing a cost of at least fifty million dollars and 
 manned by over three thousand officers and men. 
 
 The modern torpedo, for the handling of which all these 
 vessels have been built, is about eight feet long and nearly 
 two feet in diameter at the largest part. It is propelled by a 
 compressed-air motor fed from tanks containing air under 
 about seventy atmospheres pressure, and is kept laterally 
 
8o 
 
 RADIODYNAMICS 
 
 stable, and on its intended course by a gyroscope. It has a 
 speed of from twenty-five to forty knots, a range of from one 
 
 FIG. 34. 
 Types of torpedoes. 
 
 thousand to four thousand yards, and carries from two to 
 three hundred pounds of highly explosive material, usually 
 gun-cotton. It is launched from a "torpedo tube," a form 
 
 *";. 35- 
 Modern U. S. submarines. 
 
 of compressed-air gun, which on battleships and submarines 
 is submerged, and on torpedo boats and destroyers is so 
 
ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 8 1 
 
 mounted on deck that the missile can be fired in the desired 
 direction without swinging the ship. Figs. 33 to 39 will aid 
 
 FIG. 36. 
 View of a battleship in a dry dock showing submerged torpedo tube. 
 
 in making clearer the explanations relating to torpedoes and 
 to the different types of ships upon which they are used. 
 
 FIG. 37. 
 Deck type of torpedo tube used in launching torpedoes from torpedo boats. 
 
 Before actually firing a torpedo allowances must be care- 
 fully made for such variable factors as speed and direction of 
 
82 
 
 RADIODYNAMICS 
 
 both the target and the firing ship, the direction and velocity 
 of the wind, and the condition of the sea. 
 
 
 FIG. 38. 
 U. S. S. South Carolina equipped with two submerged torpedo tubes. 
 
 The percentage of hits at the extreme range of four thou- 
 sand yards is not greater than twenty-five; and when the 
 sea is disturbed, even at much shorter ranges the accuracy is 
 still less. 
 
 The torpedo boats of both surface and subsurface types are 
 chiefly relied upon to do the torpedoing, and, because of the 
 
 FIG. 39. 
 Torpedo boat destroyer entering Norfolk Navy Yard. 
 
 fact that in order to do accurate firing the distances must 
 not be great, these vessels are subject to the very hot fire 
 of the enemy's torpedo defense battery of three- and six-inch 
 guns. The torpedo attacks are, however, usually made under 
 
ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 83 
 
 the cover of darkness or fog; this fact, coupled with their 
 great speed and small size, is their only protection. If they 
 can approach near enough to discharge accurately a torpedo 
 before being discovered and illuminated by the searchlights 
 of the enemy, all is well ; but if not, it is probable that their 
 thin hulls will be riddled with three-inch shells before they 
 can escape. 
 
 The principal advantage of the ordinary torpedo is that 
 usually with one well-directed shot, a small, comparatively 
 inexpensive craft carrying from ten to fifty men can totally, 
 or at least seriously, disable a huge fighting machine like a 
 modern dreadnought, carrying a thousand men and costing 
 from five to fifteen million dollars. Its disadvantages are 
 principally the great risk to human life accompanying its use, 
 the comparatively poor accuracy of the firing, and the fact 
 that if a shot is a failure, the five thousand dollar torpedo 
 cannot be recalled. 
 
 In the year 1897, when wireless telegraphy was still in its 
 infancy, Ernest Wilson, an Englishman, was granted a British 
 patent on a system for the wireless control of dirigible, self- 
 propelled vessels. The primary object of this invention was 
 to provide a weapon for use in naval warfare, which, if in the 
 form of a dirigible torpedo, controlled from a shore or ship 
 wireless installation, would be most deadly in its effect on a 
 hostile fleet. No mention has been found of actual apparatus 
 constructed according to Wilson's plans. 
 
 To Nikola Tesla, probably more than to any other investi- 
 gator, belongs the credit of first constructing a dirigible vessel 
 which could be controlled from a distance without connecting 
 wires. His experiments were begun in 1892 and from that time 
 on he exhibited a number of wirelessly-directed contrivances in 
 his laboratory at 35 South Fifth Avenue, New York City. In 
 1897 he constructed a complete automaton in the form of a 
 boat (Figs. 40, 41 and 42), which would steer itself in obedience 
 
RADIODYNAMICS 
 
 to guiding impulses of Hertzian waves sent out from shore. 
 On Nov. 8, 1898, he was granted a United States patent on 
 this invention. In this patent he mentions the use of all 
 
 FIG. 40. 
 
 Nikola Tesla's telautomaton, controlled by Hertzian waves, which is the first 
 radiodynamic boat. 
 
 forms of control energy including electromagnetic induction, 
 electrostatic induction, conduction through earth, water, and 
 the upper atmosphere, and all forms of purely radiant energy. 
 
ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 85 
 
 The drawings, of which there are ten, illustrate in detail the 
 nature and arrangement of the apparatus. These drawings 
 were made to scale from the completed model, which he had 
 in operation at that time. 
 
 FIG. 41. 
 Side view of Tesla's boat. 
 
 Wilson's was the pioneer patent in that branch of radio- 
 telegraphy now known as radiodynamics. Since then a large 
 number of patents in this field have been taken out by various 
 inventors, and several of those who have been so fortunate 
 
 FIG. 42. 
 Interior view of Tesla's model radiodynamic torpedo. 
 
 as to secure the means, have developed their respective systems 
 in the effort to realize their possibilities. 
 
 Gardner of England, Wirth, Beck and Knauss of Ger- 
 many, Gabet and Deveaux of France, Roberts of Australia, 
 
86 
 
 RADIODYNAMICS 
 
 and Tesla, Sims, and Edison of the United States have 
 during the last fifteen years attempted to solve the problem in 
 a practical way. All of these investigators save Roberts, 
 Simms, and Edison have applied their systems on boats 
 intended primarily for torpedoes, which they control by 
 Hertzian waves. Sims and Edison, with the cooperation 
 of the United States Government, developed a system for 
 
 FIG. 43. 
 Roberts's (Australia) wirelessly directed airship exhibited in 1912. 
 
 controlling a dirigible torpedo through a trailing conductor, 
 and Roberts has applied his system to dirigible balloons. Fig. 
 43 shows A. J. Roberts and his wirelessly-controlled airship as 
 it appeared on the lecture platform. The twelve-inch induc- 
 tion-coil transmitter may be seen at the right on the table. 
 At A is the coherer, tapper, relay, and coherer battery; at B is 
 a rotary switch of the Tesla type; at C are several cells of a 
 storage battery and two signal lights; at D are two propelling 
 and steering motors which are mounted at the ends of a 
 
ADVENT OF WIRELESSLY CONTROLLED TORPEDOES 87 
 
 centrally-pivotted, horizontal frame about two feet long. 
 When both are rotating the airship moves directly ahead. 
 Steering is accomplished by stopping one of the motors. A 
 single wire about 4 feet long serves as the antenna. The length 
 of the airship is 15 feet and the weight is approximately 16 
 pounds. The gas bag consists of four layers of pig intestine. 
 The intestines. of over 4000 pigs were used in the construction 
 of this bag. The maximum control distance is about 500 feet. 
 
 These inventors have had various degrees of success in 
 their endeavors to perfect their inventions, but apparently 
 none have reached the goal. It is true that they have con- 
 trolled the movements of vessels from a distance without the 
 aid of conducting wires, but at best the apparatus has worked 
 spasmotically, unsatisfactorily, and the greatest distance at 
 which their vessels have been controlled has not exceeded 
 one-half mile. But why, we may ask, have these able experi- 
 menters failed to secure the desired results when wireless 
 telegraphy, the mother of radiodynamics, has made such 
 wonderful progress? 
 
 OA analyzing the situation we find that early in the art 
 potential-operated receiving devices, such as the coherer, were 
 used, which permitted the use of recording mechanisms. As 
 the art progressed new receptive devices were discovered 
 which, in connection with the marvellously sensitive telephone, 
 proved the coherer comparatively insensitive and unreliable. 
 Coherers were then discarded and replaced by the detectors 
 and telephones, which provided a means of signalling over 
 vastly greater distances with the same transmitting power as 
 before. The new detectors, while forming a very desirable 
 combination with the telephone, were entirely unsuitable with 
 relays, and, therefore, those interested in the control of 
 mechanisms were compelled to retain the coherer as the 
 receiving detector. This is the reason for the poor success 
 attained in the field of radiodynamics. The coherer, being 
 
88 RADIODYNAMICS 
 
 capricious in its action, sometimes operates when the trans- 
 mitting key is closed, and sometimes when no signals are sent, 
 possibly steering the boat to starboard when the signal should 
 have turned her to port, or stopping the engine when full speed 
 was desired. The coherer, because of its unreliability has 
 heretofore been the barrier to the full realization of the inven- 
 tion's possibilities. 
 
CHAPTER XII 
 SELECTORS 
 
 We have already discussed the possible control methods for 
 use in radiodynamics. The function of these, as has been 
 pointed out, is to operate an electromagnetic switch, or relay, 
 as it is called, from a distance. We have also shown how selec- 
 tivity in the operation of a relay can be secured by an applica- 
 tion of the familiar principles of resonance; for example, the 
 methods of tuning in radiotelegraphy to periodically recurring 
 characteristics of the emitted wave energy. 
 
 Since there are a number of mechanisms to be controlled, 
 each with a distinct operation to perform aboard the torpedo, 
 it is evident that we must have either a separate receiver and 
 relay for each mechanism, or else some kind of selector appara- 
 tus controlled by a single receiver and relay. As pointed out 
 by Hammond, we can, therefore, make two broad classifica- 
 tions of the control systems, namely, (i) monopulse, those in 
 which a single kind of impulse controls a single relay, which, 
 in turn, controls a means of selecting the desired circuit, and 
 (2) polypulse, those in which a different kind of impulse and a 
 separate receiver and relay are used for each circuit to be 
 controlled. 
 
 A further classification, depending on the type of relay- 
 controlled selector used by various experimenters, follows; 
 this classification also has two main heads, namely, (a) those 
 systems involving the time factor in impulse emission, and 
 (b) those independent of the time factor. 
 
 Under a we have: 
 
 (i) BlondePs, Gray's, and Mercadier's methods of con- 
 
 8 9 
 
90 RADIODYNAM1CS 
 
 trolling separate mechanisms by the use of tuned mechanical 
 or electrical elements at the receiver, and a transmitter capable 
 of transmitting impulses of varying group frequency. 
 
 (2) The author's system which utilizes a method of oper- 
 ating separate mechanisms by impulses of different length. 
 
 (3) The author's method of using a transmitter of variable 
 impulse frequency and a receiver with a solenoid of high self- 
 inductance in which the current is made to vary by change of 
 impulse emission rate so that its core can be made to assume 
 any one of a number of positions. 
 
 (4) Gardner's system in which the ratio between the length 
 of impulses and of the intervening periods of rest is varied, and 
 in which a solenoid is used at the receiver, its core assuming a 
 definite position for each value of that ratio. 
 
 Under b we may place the following: 
 
 (1) Tesla's method which uses one type of impulse to con- 
 trol a number of different mechanisms, as a clock fitted with 
 a hand, which, operated from a distance, could be made to stop 
 on any five-minute point, the hand needing to pause say two 
 seconds before the energy would be exerted. 
 
 (2) Walter's method, which depends on synchronous me- 
 chanical rotation; as two clocks, one at the transmitter and 
 another at the receiver, each fitted with a hand, which moved 
 synchronously over their respective dials, and so arranged that 
 when the transmitting hand is depressed and stopped, say, at 
 the figure six, an impulse will be sent out that effects the 
 pulling down and stopping of the receiving hand which is also 
 at six, thus closing the circuit; when the transmitting clock 
 hand is raised, the impulse ceases and both then resume their 
 synchronous rotation. 
 
 (3) There is still another method upon which modern auto- 
 matic telephone systems are based, namely, the method of 
 closing any one of a plurality of circuits by sending the correct 
 number of impulses; for example, we have a square figure 
 
SELECTORS Ql 
 
 divided into 100 equal squares; beginning at the lower left- 
 hand square let us number it o, the next above it i, the next 2, 
 and so on up to 9; then let the square at the right of o be 
 called i, the next beside it 2, the next 3, and so on to 9; the 
 squares above these are numbered in exactly the same way, 
 that is, the columns are all of the same figures. We have a 
 checker normally at the o position that can only be moved 
 twice to place it in any square, and only in two directions, i.e., 
 up, and to the side. Now in order to get our checker in space 
 67, for instance, we move it six blocks up and seven to the 
 side. In the case of the loo-circuit selector illustrated by this 
 checkerboard, two different sets of impulses are necessary, one 
 which effects the raising of the contact arm to the desired row, 
 and another to move it laterally to the position desired. 
 
 Among the earliest devices we find that of Tesla, used also 
 by Orling and Braunerhjelm, Jamieson and Trotter, Roberts, 
 and Varicas, employing a form of rotating commutator or 
 its equivalent. 
 
 Of these, Tesla's, Varicas', and Roberts' only have been 
 actually put in practice. Varicas' boat carried out in 1901 
 the simple steering evolutions required but the apparatus 
 was quite unprepared to cope with intentional interference. 
 Roberts, as before stated, applied his system in 1912 to a 
 small dirigible gas balloon for theatrical exhibitions. None of 
 these operations were carried on at any considerable distance, 
 probably not farther than a few hundred feet, but no authentic 
 data is available. All employed the primitive filings coherer. 
 
CHAPTER XIII 
 EUROPEAN CONTROL SYSTEMS 
 
 The following extract from an article on " Telemechanical 
 Problems in the Wireless World," by L. H. Walter, M.A., 
 taken with the kind permission of the author, describes some 
 of the systems experimented with in Europe during the past 
 15 years. 
 
 Walter's Selector System 
 
 "As an example of codal selectors the system devised by the 
 writer in 1898 may be taken, not because it is considered the 
 best for the writer is prepared to admit that in its early 
 form it left something to be desired but because it is 
 practically the earliest comprehensive method, and also be- 
 cause it has served as a model upon which 'numerous later 
 systems have been founded, such as those due to Chimkevitch, 
 to Hiilsmeyer, Branley and others. The system was worked 
 out as a selecting device suitable for any telemechanical, as 
 opposed to telegraphic purpose, although Righi and Dessau 
 in their 'Telegraphic ohne Draht/ and also Mazzoto in his 
 book, describe the writer 's arrangement as though it were to 
 be used for selective wireless telegraphy, like the later system 
 of Anders Bull, which has many points of similarity. The 
 original idea involved the use of synchronous rotating discs at 
 the sender and receiver, both released by the act of sending a 
 preliminary signal. One complete revolution of the disc then 
 resulted, if no further impulses were received, and the arrange- 
 ment was then in its initial receptive state. The receiver disc 
 comprises a number of contact studs placed on the periphery 
 
 92 
 
EUROPEAN CONTROL SYSTEMS 93 
 
 of the disc, corresponding to the code signal selected for one 
 circuit; these studs are all connected with a safety device. If 
 during the disc's rotation impulses are received when the 
 contact brush is exactly on one of the studs so connected, the 
 safety device has its circuit closer advanced one step for each 
 such correctly timed impulse, and finally makes an operative 
 contact when the desired evolution (steering, firing, etc.) is 
 carried out. Should, however, any impulses arrive when the 
 brush is not on a codal stud, the safety device flies back to its 
 initial position, thus preventing the actuation of the apparatus 
 by unauthorized or interfering impulses. It is well under- 
 stood that the transmitter has as many codal discs as there are 
 circuits to be controlled, and there are a corresponding number 
 of safety devices. Special relays are also used for the purpose 
 of stopping an evolution when the next evolution is one which 
 cannot be carried out without conflicting with the first (e.g., 
 'helm to port' and 'helm to starboard'). 
 
 " Although apparatus of this type was kept at work in the 
 author's laboratory for several years it has never been fitted 
 on an actual boat, owing to the fact that the idea appeared to 
 be before its time, as people at that date were not inclined to 
 take even wireless telegraphy very seriously. Hiilsmeyer's 
 system, however, which dates from 1906, and is practically 
 identical with that of the writer, was tried in Germany on a 
 practical scale, and is said to have proved satisfactory, al- 
 though it has not been possible to obtain any further particu- 
 lars. The much-discussed system due to Professor Branley is 
 also carried out on almost identical lines; his earlier arrange- 
 ment of 1906 being later completed, in 1907, by the addition 
 of a safety device like that of the writer." 
 
 Gardner's Torpedo Control 
 
 "The highly ingenious system devised by J. Gardner ap- 
 pears to have been the first comprehensive arrangement to be 
 
94 RADIODYNAMICS 
 
 put into operation on a practical scale, and has proved to be 
 one of the most thoroughly reliable methods of controlling 
 vessels by means of wireless impulses. 
 
 "At the first glance the apparatus, which is based upon 
 an application of Watt's centrifugal governor, appears to be 
 unlike any of the other systems; but on looking at it from 
 the point of view of a selecting system it is clear that the 
 device combines the properties peculiar to both the classes 
 as already defined. The governor, with its hinged balls 
 maintained near the axis by means of a spring, is normally 
 stationary, in which condition the sliding collar on the spindle 
 is in the rest position, and all circuits are open. When im- 
 pulses are received from a suitable transmitter, a step-by- 
 step arrangement causes the governor spindle to rotate; the 
 governor balls tend to fly out against the action of the spring, 
 and the collar moves along the spindle, carrying with it a 
 contact brush which is able to pass over a series of contacts 
 connected to the various circuits to be controlled. When the 
 periods of impulses and no impulses are equal, the governor 
 maintains a constant speed, and it is thus possible, by vary- 
 ing the relative durations of the impulse periods and the 
 periods of etheric silence, to make the contact brush pass on 
 to any required contact, and to maintain it there. At the 
 highest speed of rotation the firing contact is made. 
 
 "One of the chief advantages of this system is that, should 
 anything go wrong, the 'off' position is reverted to auto- 
 matically, and the torpedo comes to rest. 
 
 "By the kindness of Mr. J. Gardner the writer is able to 
 give a photograph of this interesting dirigible torpedo, which 
 is the only British wirelessly controlled craft that has stood 
 the strain of actual tests. Although these trials were carried 
 out when there was quite a lot of shipping about, there has 
 never been any mishap and this the inventor attributes to 
 the simple property possessed by the system of causing the 
 
EUROPEAN CONTROL SYSTEMS 
 
 95 
 
9 6 
 
 RADIODYNAMICS 
 
 vessel to rest when the impulses cease. The short funnel 
 which will be noticed in the photograph, Fig. 44, is for the 
 escape of battery gases; the aerial being supported from a 
 pole which fits in the socket just forward of the funnel." 
 
 Deveaux's Dirigible Torpedo Boat 
 
 "The method adopted by Lalande and Deveaux is exceed- 
 ingly simple, but the boat represents the most ambitious 
 attempt in the history of wirelessly controlled apparatus. 
 The selector system comprises (i),' a circular distribut- 
 
 FIG. 45- 
 
 ing switch, having on it the studs pertaining to the cir- 
 cuits to be controlled; and (2), a circuit closer which only 
 allows the current to pass when the distributing switch arm 
 has reached the desired contact stud. In the actual ap- 
 paratus the distributing switch has twelve studs, of which 
 nine lead to the nine operating circuits employed; the re- 
 maining three are distributed among the others and con- 
 stitute rest positions with a view to saving the switch arm 
 from having to execute a complete circle each time. Fig. 45 
 is a diagram of the connections. 
 
 "In order to carry out the double function mentioned, an 
 
EUROPEAN CONTROL SYSTEMS 97 
 
 electromagnet M is provided which moves forward by one 
 tooth at each Hertzian impulse, a twelve-toothed step-by-step 
 arrangement connected to the distributor arm D. During 
 the period when this arm is being advanced, no closing of the 
 operating circuits is possible owing to the circuit closer being 
 opened by means of a projection on the end of the armature of 
 the electromagnet; this is shown at P. Thus if twelve im- 
 pulses are received, the distributor would describe a complete 
 revolution. On the other hand, if the impulses cease after the 
 distributor has been carried from a rest stud to one of the 
 operative studs, the circuit closer will complete the circuit after 
 a brief interval of time, which is caused to elapse owing to the 
 intervention of a retarding device. This latter consists of a 
 train of clockwork, which, by virtue of its inertia, does not 
 allow the circuit closer to operate at once; a delay of twice 
 the time required for the distributor to be moved forward by 
 one tooth has been found sufficient. M. Deveaux's paper, 
 which was published in the Bulletin of the Societe Inter- 
 nationale des Electriciens, in 1906, will be found to give full 
 information as to the circuit arrangements, but no illustra- 
 tions of the boat itself. 
 
 "By the courtesy of M. Montpellier, the editor of 1'Elec- 
 tricien, the writer is able to make good this deficiency, and 
 to give two photographs of this craft; Fig. 46 shows the vessel 
 when hoisted out of the water, and Fig. 47 gives a general 
 idea as to its appearance and the visibility of its antenna 
 when afloat; the French cruiser Saint Louis, shown in the 
 background, was watching the trials which took place off 
 the port of Antibes in the early part of 1906. 
 
 11 The boat itself, which weighs 6700 kg., consists of two 
 cigar-shaped bodies formed of steel plate, one above the 
 other on the principle of the Sims-Edison dirigible torpedo. 
 The upper cylinder, 9 meters long by 45 cm. in diameter, acts 
 as a float; it is provided with two small masts, which serve 
 
9 8 
 
 RADIODYNAMICS 
 
 i 
 
 I 
 
 I 
 
 .s 
 "s 
 
 
EUROPEAN CONTROL SYSTEMS 
 
 99 
 
 to support the wireless antenna, consisting of five wires kept 
 at a height of about three meters; and these masts have 
 lamps about halfway up, for the purpose of facilitating 
 steering operations. The lower cylinder is n meters in 
 
 FIG. 47. 
 
 length, and i meter in diameter, and contains the torpedo- 
 ejecting tube and a Whitehead torpedo of 450 mm. diameter; 
 the accumulator battery and propelling motors are also con- 
 tained therein. The control apparatus is intended to be 
 placed in the lower cylinder, where it would be protected 
 
100 RADIODYNAMICS 
 
 from the enemy's gunfire by two meters of water, but in the 
 trials the apparatus was placed in a sheet metal box on the 
 top of the upper cylinder in order to be available should any 
 adjustments be required. The trials were carried out over 
 a comparatively short radius, 400 to 1800 meters, but it is 
 stated that these distances could easily have been exceeded 
 though to what extent is not said. 
 
 "The transmitting station from which the boat's evolutions 
 were controlled was on land, and had a five- wire antenna 15 
 meters in height; but no information is available as to the 
 actual wave length employed, although, from the size of the 
 receiving antenna, it was probably very short, of the order 
 of 80 to 100 meters." 
 
 Wirth, Beck and Knauss 
 
 Remarkable achievements in the line of torpedo control 
 have been accomplished in Germany, where two unmanned 
 motor boats 33 and 50 feet in length have been steered, 
 stopped, started, and controlled in every way by electric 
 waves transmitted from the shore without the use of wires. 
 The system employed is the invention of C. Wirth of Nurem- 
 berg. It has been brought to its present stage of develop- 
 ment by several years of experiment, conducted by Wirth 
 and his cooperators, the manufacturer Beck and a merchant 
 named Knauss; it is protected by numerous German and 
 foreign patents. 
 
 The first success was attained in 1910 with an electric 
 launch on a lake near Nuremberg. The vessel was 33 feet 
 long, and was propelled by a 5 horse power motor, and an 
 accumulator battery of 80 volts and 300 ampere-hours. 
 The first public demonstration was given with this boat in 
 1911 before the German Fleet Club; in this demonstration 
 the unmanned boat fired a signal shot and then set itself in 
 motion. Travelling at a speed of about 10 miles it was made 
 
EUROPEAN 
 
 101 
 
 to turn right or left or to stop completely and start again by 
 the controlling operator in obedience to the requests of mem- 
 bers of the Fleet Club. Each order was obeyed within from 
 one to five seconds, and signal lights flashed back the receipt 
 of the impulses. The manuevers were continued for several 
 hours. 
 
 A boat 50 feet in length was later exhibited in Berlin, at 
 the invitation of the German Fleet Club. An antenna of 
 
 Explosive Head 
 
 Receiving and Steering Motor 
 
 Control Apparatus 
 
 FIG. 48. 
 Proposed form of Wirth radiodynamic torpedo. 
 
 four wires was stretched between the cupola of the Kaiser 
 Pavillion and the restaurant on the shore of Lake Wann. 
 The transmitting apparatus which was installed at the 
 restaurant was of the induction coil type, and was of about 
 100 watts capacity. The various operations performed on 
 the boat were accomplished by sending impulses by means 
 of a Morse key. The boat was equipped with an antenna 
 of four wires about 15 feet high, a radio receiver capable of 
 adjustment to different wave lengths from the transmitter, a 
 distributor or selector, electric steering apparatus, signal 
 guns, lights, and fireworks apparatus. The tuning of the 
 
102 ; L RA'DtOVYNAMICS 
 
 apparatus could be altered by sending a long signal; this was 
 for the purpose of evading interference. 
 
 The general scheme of the Wirth torpedo is shown in 
 Fig. 48. The diagram which is here presented in Fig. 49 
 shows the essential parts of the control system, and the 
 circuit arrangements. The coherer 38, of the usual filings 
 
 FIG. 49. 
 
 type, is connected in the circuit of the battery 37 and a 
 sensitive relay 39. The armature of this relay, 36, serves to 
 close a second circuit including the battery 13, by which the 
 electromagnet 14 is operated. By means of the latter, there 
 is operated the lever 8 which serves to rotate a ratchet wheel 
 7 by means of a pawl in the usual way, each time an impulse 
 is received ; at 40 is a tapper for the coherer. Arriving impulses 
 cause the ratchet wheel to advance by one, two, three, etc.. 
 
EUROPEAN CONTROL SYSTEMS 
 
 I0 3 
 
 teeth, and as the wheel is mounted integral with a contact 
 disc, the latter is rotated at the same time. The fixed brush 
 thus comes over a metal contactor or otherwise over the in- 
 sulated part between the contacts according to the position of 
 the ratchet wheel. Should there be a contact piece under the 
 brush, the circuit of the battery 4 is closed and one of the 
 six electric motors, I-VI is set in motion. By the rotation 
 of the motor there is set working a spring contact device 
 which will be further mentioned, and such contacts act to 
 
 FIG. 50. 
 
 close the circuit of the apparatus which is to be finally worked, 
 such as the movement of the rudder of the boat, etc. A 
 second motor of the series serves to work another apparatus, 
 and there is used one motor of small size for each operation 
 to be carried out. The purpose of the motors is to furnish 
 a time element device, which allows distinction between long 
 and short impulses. 
 
 Fig. 50 shows the apparatus which is used for two dis- 
 tinct movements, namely, for steering to right or left. At I 
 is a relay which is worked by the coherer, and at II the 
 contact disc before mentioned. At Ilia and Illb are two 
 small electric motors for making the contacts, this latter 
 
104 
 
 RADIODYNAMICS 
 
 being carried out by the spring contact devices IVa and 
 IVb, one for each motor. 
 
 The coherer action sends current impulses by means of the 
 relay I into the electromagnet of the contact disc. According 
 to the number of impulses which are sent, we have the brush 
 placed on a metal contact or in the insulated interval. When 
 the brush is on one of the uneven-numbered contacts, the 
 motor Ilia is set working, and it acts on the spring contact 
 device IVa so as to operate the small contact switch noticed 
 
 at the front. Such contact 
 thus gives current for op- 
 erating the movement of 
 the rudder to the left by a 
 suitable electromagnetic de- 
 vice. When the brush is 
 on one of the even-num- 
 bered contacts the motor 
 Illb is set running, and it 
 works the corresponding 
 spring switch IVb so as to 
 give current for a second 
 magnetic device, for bring- 
 ing the rudder to the right- 
 hand side. The mechanism 
 of the spring contact device 
 
 FIG. 51. 
 
 is arranged on the retarding plan, so that it first sends out a 
 wave signal which is received at the sending station; two 
 seconds later it closes the switch. 
 
 Should the brush remain but a short time on one of the 
 contacts, this will give no effect, as the motor takes a certain 
 time to start up, and thus the motor gives a method of work- 
 ing by means of long contacts, but not by short ones. 
 
 When the brush is on an insulated part of the disc the de- 
 vice is inactive, and the rudder comes automatically to the 
 
EUROPEAN CONTROL SYSTEMS 
 
 105 
 
 zero or central position. The signal which is sent back to 
 the shore station is seen on the paper strip of the receiver, 
 and the operator thus has a check on the working of the ap- 
 paratus, and can correct any wrong working by subsequent 
 signals. Wirth's transmitting antenna is shown in Fig. 51. 
 
 Dr. E. Branley's Control System 
 
 Dr. Branley, of Paris, in addition to various other kinds of 
 distant control apparatus, devised an instrument with the 
 
 
 , 
 
 m 
 
 FIG. 52. 
 
 purpose of protecting the receiver against a continuous stream 
 of sparks such as the enemy might send out in time of war. 
 
106 RADIODYNAMICS 
 
 This, like a previous system of Fessenden's, utilizes breaks in 
 a continuous emission of energy as the signal or controlling 
 impulses, instead of " makes," with periods of rest intervening. 
 If interfering signals are sent continuously the apparatus can- 
 not be operated by any other signals, even from the controlling 
 station, but should the interfering signals cease for a short 
 time, the controlling operator can perform the desired opera- 
 tion by making the required number of breaks in his own 
 continuous stream of signals. 
 
 Dr. Branley's protective device consists of a horizontal 
 disc moved by clockwork, and is kept constantly in rotation 
 first to the right, then to the left, by electromagnets which 
 are acted on by distant waves. The rotation of the disc 
 causes a series of contacts for closing different circuits cor- 
 responding to the different operations to be performed. The 
 whole is so arranged that when a continuous stream of energy 
 is received the disc rotates forward and back. If the dis- 
 turbing signals cease for a brief period of time the control 
 operator sends a code signal, which acts upon the disc and its 
 contacts in such a way that the operation is performed. 
 
 In the present type of apparatus the waves are received 
 upon a new type of coherer which is shown in Fig. 52. It is a 
 modified form of Dr. Branley's tripod coherer, and is made up 
 of a polished steel cylinder at the lowest part. It is fitted on 
 an upright support, and from this three arms hang down by 
 means of pivots. The arms carry well-rounded steel projec- 
 tions which bear lightly on the cylinder so as to make the 
 coherer contact. The whole is enclosed in a vacuum chamber 
 in order to protect the coherer from the action of the air. 
 Such a coherer is useless when subject to vibration. 
 
CHAPTER XIV 
 
 WORK OF THE HAMMOND RADIO RESEARCH 
 LABORATORY 
 
 Following the rotary switch scheme of Tesla, John Hays 
 Hammond, Jr., head of the Hammond Radio Research Labo- 
 ratory at Gloucester, Mass., began his experiments in the 
 summer of 1910. No detailed accounts of these first experi- 
 ments are available, as no systematic method of keeping 
 records of the work had then been inaugurated, but it is 
 known that mechanisms designed to steer a small boat were 
 operated at a distance of three or four hundred feet. This 
 apparatus, however, was never actually used in a boat for 
 steering purposes. 
 
 During the following winter an entirely new set of control 
 apparatus was designed in New York from Mr. Hammond's 
 plans. The object in view was to build a control apparatus, 
 which could be attached to existing automobile torpedoes. 
 The coherer receiving set, relays, rotary switch, cut-off and 
 center-stop mechanism, and batteries, were all contained in a 
 brass tube about one foot in diameter and six feet long. 
 
 This apparatus was set up on a float landing about a thou- 
 sand feet from the transmitting station. An antenna 15 feet 
 high and 20 feet long was improvised, and after much careful 
 adjustment, signals were received which were capable of start- 
 ing, stopping or reversing the steering motor. The transmit- 
 ting antenna was of the inverted L-type, about 80 feet high 
 and 200 feet long. An antenna current of about 2 amperes was 
 registered. The transmitting set was of the Clapp-Eastham 
 type, 60 cycle, 3 kw. 
 
 107 
 
108 RADIODYNAMICS 
 
 After these preliminary tests the apparatus was set up in 
 a i2-foot gasoline launch, with a 1 5-foot antenna supported 
 by bamboo poles. Considerable trouble was experienced in 
 these tests. Due to the engine vibration, the sensitiveness of 
 the Seimmans and Halske relay, as well as the Marconi co- 
 herers had to be greatly reduced. The Hertzian and inductive 
 effects from the gas engine caused considerable trouble until 
 the engine pit was entirely encased in sheet iron; this, however, 
 did not eliminate the coherer trouble although it decreased 
 it. The instruments were almost inaccessible for adjustment; 
 the moist, salt air made matters still worse by corroding the 
 multitude of contacts. Finally a determined attempt at rudder 
 operation was made even though the action of the apparatus 
 was far from what had been expected, and indeed necessary. 
 The motor in the tube was accordingly connected to the boat's 
 steering post by chain and sprockets, but when the current 
 was switched on the motor was found to possess less than 
 half the power required in turning the rudder hard over when 
 the boat was under way. Three weeks were spent in futile 
 attempts to eliminate the difficulties; then the tube and most 
 of its contents were relegated to what was then the scrap heap, 
 and now the historical collection. 
 
 Simplified Apparatus 
 
 Plans were at once formulated for the construction of much 
 simplified apparatus, which could be thoroughly tested under 
 conditions in which it could be protected from the weather, 
 and observed and adjusted while in operation. An old house 
 boat of about eight tons displacement fulfilled all the require- 
 ments for a floating laboratory splendidly. She was fitted 
 with a gasoline engine capable of driving her four knots an 
 hour, and forty-foot masts for supporting the antenna. This 
 boat is shown in Fig. 53. 
 
 Coherers and relays of highest sensitiveness combined with 
 
HAMMOND RADIO RESEARCH LABORATORY 
 
 109 
 
 the necessary ruggedness were secured from the electrical 
 instrument makers in America and Europe, A steering motor 
 of increased size was procured and mechanically connected 
 to the steering wheel on the house boat by a worm-wheel 
 reduction gear; a hand-operated clutch permitted either radio 
 or manual control of the 
 wheel. 
 
 With this new appa- 
 ratus installed in the 
 Pioneer (as the house boat 
 was afterwards named), 
 where it was protected 
 from the weather in an 
 atmosphere that could be 
 kept dry and warm by a 
 coal stove, and arranged 
 for continual observation 
 and adjustment while in 
 operation, the results were 
 more satisfactory. The 
 filings coherer, however, 
 continued to be the chief 
 source of our difficulties. 
 Every known remedy for 
 the trouble was applied to 
 increase the sensitiveness and reliability, but despite all these, 
 the sheet iron protection from stray Hertzian and inductive 
 effects, the protective resistances and capacities for preventing 
 sudden rise of potential at current-breaking points, despite 
 the care exercised in the selection and adjustment of jiggers, 
 relays, decoherers, etc., the results were so discouraging that 
 it was decided to discontinue the use of the filings coherer, 
 and adopt the Lodge-Muirhead mercury-steel-disc coherer. 
 Several complete receivers of this type, which had been dis- 
 
 FIG. 53- 
 
no 
 
 RADIODYNAMICS 
 
 carded from actual service, were purchased from the United 
 States Navy. These had become obsolete and useless be- 
 cause of the advent of the telephone receiving sets. The 
 best of these was installed on the Pioneer and was found to 
 be more sensitive and reliable than the filings coherer. Fig. 54 
 shows this receiver as installed aboard the house boat and 
 Fig. 55 is a detail drawing of the Lodge-Muirhead coherer. 
 
 After this change had been made the boat could be kept 
 under fairly good control at distances up to and over a mile. 
 It was steered over a prearranged course during both day 
 
 FIG. 54. 
 
 and night, and in all conditions of sea and weather. The 
 course was by no means simple, covering, as it did, circles 
 around several buoys, and a complete circle around the harbor. 
 Fishing and other vessels were continually moving about the 
 harbor but no great difficulty was experienced in avoid- 
 ing them, and, at the same time, keeping on the course. 
 It was found possible to steer the boat against either of 
 several upright spar buoys a mile from the point of control. 
 At night lights, automatically controlled by the steering 
 mechanism, kept the "helmsman" at the transmitting key on 
 shore informed of the boat's action. A white light would 
 shine each time an impulse took effect; in this way the con- 
 
HAMMOND RADIO RESEARCH LABORATORY 
 
 III 
 
 trol operator on shore was immediately informed if the receiv- 
 ing apparatus or part of the control apparatus had gotten out 
 of order. As long as the rudder was in the central position no 
 lights save the required running lights were burning. As soon, 
 however, as the rudder moved to one side or the other a red 
 or green light on the yard arm would be connected, depending 
 on the resultant direction of 
 the boat, and this would con- 
 tinue to burn so long as the 
 direction of the steering mo- 
 tor's rotation was not reversed. 
 When the rudder reached the 
 extreme hard-over position an 
 additional red or green light 
 would flash, the two of the 
 same color remaining illumi- 
 nated while the boat was turn- 
 ing in her circle of shortest 
 diameter. If the direction of 
 motion was left then the two 
 lights would be green in color; 
 if right the color would be red. 
 As soon as the rudder was 
 again started back to the cen- 
 ter the two lights would go 
 out and a single light of the 
 opposite color would come on; when the rudder was stopped 
 at the mid-position by the automatic center-stop mechanism, 
 the white light would again flash for an instant, signifying 
 that fact. 
 
 The steering of the boat was accomplished by sending 
 Hertzian wave impulses, which, affecting suitable receiving 
 and switching devices, controlled the one-fourth horse-power 
 electric motor mechanically connected to the steering wheel. 
 
 FIG. 55. 
 
 A is the steel disc with a polished 
 knife-edge; B is the small cistern of 
 mercury covered with a film of oil; 
 K is a leather wiper; H and E are 
 the terminals. 
 
112 
 
 RADIODYNAMICS 
 
 The rudder, by this means, could be made to move to port or 
 starboard at will, or set at any intermediate position from the 
 transmitting station. 
 
 During the next year some valuable additions were made 
 for carrying on the experimental and research work; the size 
 
 FIG. 56. 
 
 of the station was increased, two 33O-foot towers (see Figs." 56 
 and 57) were erected for supporting the antenna, a battery of 
 mercury-arc rectifiers was installed to furnish direct current 
 for the operation of two 5-kw. 5oo-cycle motor generator sets, 
 two ioo,ooo-cycle alternators, a 24-inch searchlight, and vari- 
 ous other apparatus. A 4o-foot gasoline launch of 150 horse- 
 power and over 25 knots was built for use as a torpedo, and 
 
HAMMOND RADIO RESEARCH LABORATORY 113 
 
 FIG. 57. 
 Installing the antenna system at the Hammond Station. 
 
 FIG. 58. 
 
RADIODYNAMICS 
 
 FIG. 59. 
 
 other valuable additions were made to the control system, 
 which permitted a greater range and more reliable operation. 
 
 The battery of four General 
 Electric 5o-ampere recti- 
 fiers is shown in Fig. 58. 
 Fig. 59 shows the 5-kw. 
 Lowenstein Transmitter. 
 
 Steering Apparatus 
 
 A brief description of the 
 control apparatus is here 
 necessary in order to form 
 a clear conception of some 
 of the important details. 
 It has been previously mentioned that a control system is 
 composed of two main parts: (i) the transmitter and receiver, 
 and (2) the mechanism to be controlled. The principal parts 
 of the mechanism, which is the rudder control apparatus, are 
 the electromagnetically operated reversing switch, the steering 
 motor, and the source of pov. er. 
 
 The rotary switch, shown in Fig. 60, is essentially an 
 insulating drum fitted with contact pieces; it can be revolved, 
 step by step, through successive contact positions with a set 
 of brushes by means of an electrorragnet and pawl and 
 ratchet. The contact positions and blank or "neutral" 
 positions alternate; moreover, the contact positions are of 
 two kinds, one for clockwise rotation of the motor, the other 
 for counterclockwise rotation. The sequence of positions, 
 then, as the electromagnet is impulsively operated, is port, 
 neutral, starboard, neutral, port, neutral, and so on in the 
 same order. 
 
 This is easily understood when the rotary switch is looked 
 upon as a simple, double-pole, double-throw reversing switch 
 connected to the armature of the steering motor, the shunt 
 
HAMMOND RADIO RESEARCH LABORATORY 115 
 
 field of which is continuously excited. A diagram of this con- 
 nection is shown in Fig. 61. 
 
 Consider the switch in the upper or neutral position, where 
 the armature is disconnected from the source of power. There 
 
 FIG. 60. 
 
 Electrically-operated rotary switch designed by the author and used in the 
 Hammond System of control. Relay at right, drum switch in the center, 
 and operating magnet at left. S 
 
 are two possible ways of closing the switch, corresponding to 
 the two possible directions of motor rotation. One of these 
 will, by swinging the rudder to port, cause the boat to steer 
 around to port; the other will effect starboard motion. The 
 only difference between these 
 two reversing switches is that 
 with the hand type, the motor 
 after being stopped, can be 
 made to run in the same direc- 
 tion again without the necessity of passing over the position 
 for opposite rotation. With the rotary switch this cannot 
 be accomplished unless some auxiliary instrument be used to 
 prevent the motor's rotation while passing over the undesired 
 position. 
 
 D.C. 
 
 Field 
 
n6 
 
 RADIODYNAMICS 
 
 The steering motor should preferably be of the shunt type 
 with the field winding continuously energized. This is 
 important to secure quick action. Motors larger than one- 
 fourth horse-power cannot well be used with such a controlling 
 switch because the unregulated starting current becomes 
 excessively large. Where greater power is required for rudder 
 operation a pneumatic control apparatus is more effective. 
 The one-fourth horse-power motor was found large enough for 
 the 33-mile " Radio," which had a displacement of about four 
 tons. 
 
 To Rotary -Switch Magnet 
 
 <?/ A ,vcv 'Travelling ''Guide Rod fytf^top 
 ^-Starboard Stop Block ,'Adjustmen 
 
 nag A djustment 
 
 ~ 6ia * e Base 
 FIG. 62. 
 
 The source of power is preferably a storage battery. It 
 should be of the most rugged type. An Edison 30- volt, 120- 
 ampere hour battery gave excellent service in the Gloucester 
 experiments. 
 
 A crude, but nevertheless operative, control system can be 
 made up of these essential parts, since with them the steering 
 motor, which is mechanically connected to the steering wheel, 
 can be started, stopped, and reversed, the worm-wheel reduc- 
 tion gear serving to lock the wheel in any desired position after 
 the power has been cut off from the motor. The value and 
 reliability of such a crude controller, however, is much increased 
 
HAMMOND RADIO RESEARCH LABORATORY 117 
 
 by the use of auxiliary instruments, which make it possible to 
 greatly reduce the skill required of the controlling operator. 
 The cut-off and center-stop mechanism is one of these. It 
 consists essentially of a threaded shaft bearing a small traveling 
 block, fitted with two fingers as shown in Fig. 62. This shaft 
 is connected directly, or by means of a flexible shaft, to the 
 reduced-speed shaft of the steering motor; near each end of 
 the worm shaft is a platinum-tipped contact spring. At the 
 center is a short stiff spring provided with a contact screw on 
 each side. 
 
 The operation is as follows: The apparatus is so adjusted 
 that the central position of the traveling block corresponds to 
 the central position of the rudder, and the end contact springs 
 are so placed that the operative leg of the motor circuit is 
 broken by the motion of the traveling block when the rudder 
 reaches the extreme right and left positions. The center-stop 
 contacts are engaged by the fingers on the traveling block a 
 short time before the rudder reaches the central position. 
 By closing the circuit of the electromagnet which operates the 
 rotary switch, these contacts make electrical connections which 
 effect the turning of the drum to the neutral position, and 
 thus stop the motor. The adjustment must be very care- 
 fully made in order to make the rudder stop at the exact 
 position corresponding to straight-ahead motion of the 
 boat. 
 
 Suppose the boat is steering ahead. An operator at the 
 control station desiring to take control depresses his key once, 
 the necessary length of the impulse transmitted being from 
 one-half to one second. This impulse will move the drum to 
 either a port or starboard position, and is sent in order to get 
 his bearings. Let us assume that the boat immediately begins 
 to swing to starboard. The steering motor has been energized 
 and is swinging the rudder to the right. The time required 
 for the rudder to move from the center to either extreme posi- 
 
Il8 RADIODYNAMICS 
 
 tion is from five to fifteen seconds, depending on the speed of 
 the steering motor. This time could be fixed at any value 
 within these limits by adjustment of the field weakening 
 rheostat connected in series with the motor's shunt field. 
 
 The operator, then, could allow the rudder to continue its 
 motion to the extreme position, where it would be automati- 
 cally stopped by the cut-off mechanism, or it could be stopped 
 at any intermediate position, by sending another correctly- 
 timed impulse after the first. This second impulse would 
 simply turn the drum to the next neutral position and thus 
 stop the motor. If he allowed the rudder to go to the extreme 
 position the boat would travel in its circle of minimum 
 diameter until the correct number of impulses were sent to 
 change the rudder's position. 
 
 Since the drum was revolved to a starboard position by the 
 first impulse, it remains there, the motor circuit for starboard 
 rotation being opened by the cut-off mechanism. Obviously, 
 then, two impulses are required to rotate the drum to the next 
 port position. If these are sent, the motor will rotate in the 
 opposite direction, and, if no more impulses are sent within 
 the fixed time limit, the rudder will move to the central posi- 
 tion and automatically stop. This central stop, as before 
 explained, is effected by the automatic rotation of the drum 
 to a neutral position. Remembering, then, that in this case 
 the neutral position is followed by a starboard position, one 
 impulse will send the boat to starboard again and the minimum 
 number of impulses for port steering is three. If these are 
 transmitted, the rudder will turn to the extreme port, unless 
 stopped at some intermediate position, by sending an additional 
 impulse. 
 
 It is thus seen that the nucleus of the apparatus is simply a 
 wirelessly-operated switch, supplemented by auxiliary ap- 
 paratus, which automatically perform operations that would 
 be very difficult for the distant operator. 
 
HAMMOND RADIO RESEARCH LABORATORY IIQ 
 
 In practice a difficulty developed which was only overcome 
 after considerable experimentation. It was found that, when 
 the rudder and traveling block were in their central positions, 
 and three impulses were sent at the maximum speed permitted 
 by the inertia of the various parts, the drum remained on the 
 undesired contact long enough to allow a considerable number 
 of revolutions of the steering motor in the direction opposite to 
 that desired. The following would be the result : 
 
 The first impulse would effect rotation in the wrong direc- 
 tion, the motion of the traveling block being great enough 
 to allow the contact finger to pass under the spring. The 
 second impulse would bring the next blank space into posi- 
 tion, stopping the block's motion; the third would effect the 
 desired rudder motion and the block would immediately begin 
 to move with the rudder in the desired direction. After a few 
 revolutions, however, the center-stop finger would engage its 
 contacts, and as a result the motor would stop. 
 
 In this way it was found impossible to get the rudder 
 through the central position from one side to the other. 
 Tests without the automatic stop, however, clearly indicated 
 the futility of trying to dispose of it, it being found im- 
 possible to steer a straight course because of the difficulty of 
 stopping the rudder at the exact point necessary for steering 
 directly ahead. 
 
 Attempts were made to surmount this difficulty by the 
 introduction of considerable inertia into the rotating parts 
 connected to the motor, so that in the brief interval of time 
 necessary in passing over undesired operative positions of the 
 rotary switch, the motor could not develop enough speed to 
 cause the above-mentioned difficulty. This was partially 
 satisfactory, but did not completely solve the problem. 
 
 The solution was finally found in a "time relay," especially 
 built for this purpose after our plans, in New York. This 
 instrument consists principally of an iron-clad solenoid with 
 
120 
 
 RADIODYNAMICS 
 
 a movable core, and an adjustable air dash pot. A plan view 
 of this relay is shown in Fig. 63. 
 
 When energized, the core is drawn up in a period of time that 
 can be easily varied, by adjustment of a thumbscrew, from a 
 fifth of a second to ten seconds. A set of heavy platinum con- 
 tacts is fixed on the core and an adjustable screw so that a local 
 circuit, carrying ten or fifteen amperes can be satisfactorily 
 opened and closed at the end of the "in" stroke. A light 
 spring serves quickly to bring the core back to the normal 
 position after the solenoid is de-energized. This is facilitated 
 
 Field Frame 
 
 Back Spring Co// 
 
 Solenoid 
 
 .... to"--- 
 
 FIG. 63. 
 
 by a one-way valve which permits the expulsion of. the air in 
 the dash pot with but little retardation on the return stroke. 
 The solenoid windings were connected in parallel with the 
 electromagnet of the rotary switch, and the contacts were 
 connected in one leg of the battery circuit to the motor 
 armature. By this means the motor armature current could 
 not be exerted until a definite length of time had elapsed 
 after the impulse had been sent, and thus the undesired oper- 
 ative positions on the drum switch could be passed over 
 without difficulty. The time limit was usually set at about 
 one second. Another time relay of the same kind was used 
 for engine control. In this way short impulses of about one- 
 
HAMMOND RADIO RESEARCH LABORATORY 
 
 121 
 
 half second duration were used for steering, and long im- 
 pulses of about ten seconds were used for starting or stopping 
 the engine. The engine control in these experiments was 
 effected by opening or closing the ignition circuit, the engine 
 being so adjusted that it would stop at the required point at 
 which the explosions occur. It was found possible to start 
 the engine in this way as long as an hour after it had been 
 stopped. This second time relay was also used for firing ex- 
 
 LU 
 
 Rotary Switch Magnet 
 
 Potenfio Sensitive Relay 
 Detector ' 
 
 I 1 ! 
 
 '"- oadinc/ Inductance 
 Tuning Capacity 
 
 Time Relay 
 for Engine Control ' 
 
 Rotary Switch 
 / Connections 
 
 io cnqi 
 Clutcfi- 
 
 FlG. 64. 
 
 plosive charges of powder placed on top of the boat's cabin. 
 The complete circuit is shown in Fig. 64. 
 
 During the summer and autumn of 1912, with this im- 
 proved apparatus, the results were very encouraging. The 
 25-knot Radio of which Fig. 65 is a reproduction was con- 
 trolled with reliability and precision at ranges of over three 
 miles, a distance far in excess of that attained by European 
 investigators. Demonstrations made before the U. S. War 
 Department authorities proved beyond a doubt that the 
 dirigible torpedo would be of great use in naval warfare, 
 especially for coast defense. 
 
122 RADIOD YNA MICS 
 
 The operations can be carried on at night almost as well as 
 by day. With a special reflector of the triple-mirror type, 
 and a searchlight, the maneuvers of the boat can be followed 
 with ease from the control point, although from any other 
 position the reflected light, and the boat itself, are practically 
 invisible. 
 
 The broad idea of the Hammond torpedo control system 
 for coast defense is this: 
 
 One powerful transmitting station is employed and suitably 
 placed in some protected situation with respect to the gunfire 
 
 FIG. 65. 
 
 of the enemy. A number of operators are placed along the 
 shore; these different operators have wire connection with 
 the wireless station and telephone connection with one 
 another. A number of torpedoes may be used, each of which is 
 controlled by an impulse of specific characteristics. These 
 are moored at a central protected torpedo base, and one or 
 more can be controlled at the same time by either of the 
 operators at the hidden control stations. The torpedoes are 
 started at the base and passed on to the control of the oper- 
 ators in the most advantageous positions, who operate under 
 orders transmitted over the telephone lines by the military 
 
HAMMOND RADIO RESEARCH LABORATORY 123 
 
 head of the fortifications. The wireless station is equipped 
 with a wave generator for each torpedo, and by means of the 
 wire connections, which extend from the control point to 
 the central station, impulses can be sent from either of these 
 generators by either of the control operators. 
 
CHAPTER XV 
 
 THE SOLUTION OF THE PROBLEMS RELATED TO 
 BATTLE-RANGE TORPEDO CONTROL 
 
 The problem of performing a number of operations aboard 
 a moving vessel at distances of over a mile is by no means a 
 
 120 
 
 Galena, 
 Y 
 
 RECEJ.VER 
 
 1300m. 
 
 \SOO-~-~. 
 
 ^ TRANSMITTER 
 
 U i i I 
 
 _L 'HWA.-IOAmps. 
 
 Distance- intensity received current 
 
 on "D/RIGIA" with Solid- Rectifier' Detectors 
 
 ZQ 30 
 
 MINUTES 
 3 
 
 MILE.S. 
 
 FIG. 66. 
 
 simple one, even from a theoretical point of view. It might 
 be mentioned that controlling a number of mechanisms with 
 but a single relay is easy in comparison with wirelessly con- 
 
 124 
 
BATTLE-RANGE TORPEDO CONTROL 125 
 
 trolling a relay five miles away, which at the same time is 
 immune from accidental or intentional interference from 
 other sources of energy. Then, too, when we reflect that 
 probably less than one-billionth of the energy radiated into 
 space by the transmitter reaches the receiver, the quanti- 
 tative aspects of the problem begin to come into evidence. 
 
 Fig. 66 is a graphical representation of the current re- 
 ceived on board the Pioneer, which was first used as a torpedo 
 in the Gloucester experiments. The transmitting energy in 
 the antenna * was about two kilowatts at a wave length of 
 approximately 1300 meters. The group frequency was 1000 
 and, with the air cooled quenched spark gap, and loose in- 
 ductive coupling between the open and closed oscillatory 
 circuits, the decrement was 
 fairly low. 
 
 The receiving antenna as 
 shown in the illustration 
 of the Pioneer was of the 
 inverted L type, 30 feet 
 high and with a 2o-foot 
 flat top, the detector was F IG> 57. 
 
 one of the solid rectifier 
 
 type, essentially a crystal of galena with a light spring 
 contact. It is shown in Fig. 67 and was designed by the 
 author. The current values given were read directly upon 
 a Weston microammeter. The readings were taken at five- 
 minute intervals, except in close proximity to the station, 
 and the distances corresponding to these intervals were 
 computed from the boat's speed and log readings. This 
 curve shows how quickly the received current drops down 
 within a mile, and how it remains almost constant after this 
 distance is well passed. The high value of the received cur- 
 rent within a short distance of the transmitter may be 
 
 * 300 ft. high, 400 ft. flat top. 
 
126 RADIODYNAMICS 
 
 due to the augmentation of the Hertzian effects by the purely 
 electrostatic effects, as evidenced by the curve made in the 
 experiments with an electrostatic telegraph (Fig. 20). The 
 received current at three miles was only about 3-icr 6 amperes. 
 So far as we were able to learn there was no relay, possessing the 
 necessary mechanical and electrical stability, which was sensi- 
 tive enough to operate reliably on such a small amount of 
 energy. The most sensitive relay we could procure in the 
 United States and Europe, which was rugged enough to 
 operate reliably under the conditions of shock and vibration 
 aboard a small high-speed boat in a rough sea, required about 
 300. icr 6 amperes. We had as high a transmitting antenna 
 as was practicable, the most efficient transmitting apparatus, 
 and the most sensitive receiving set obtainable, and yet the 
 
 breach between the available 
 and the required received cur- 
 rent, was so wide that it ap- 
 peared almost impossible to 
 bridge it. With a sender that 
 could deliver only one mi- 
 croampere at four miles, and a 
 receiving relay that required 
 300 microamperes for opera- 
 tion, the problem was a serious 
 and discouraging one. 
 
 FKJ 68 The first step in the solu- 
 
 tion was in the improvement 
 
 of the sensitive relay. This was of the Weston pivoted galva- 
 nometer type, and is reproduced in Fig. 68 by the courtesy of 
 the Weston Electrical Instrument Company. The permanent 
 magnet was replaced by an electromagnet, which, by increasing 
 the field intensity, more than doubled the sensitiveness. Fig. 
 69 shows graphically the effect of variation in the field energiz- 
 ing current. Later the author replaced the delicate platinum 
 
BATTLE-RANGE TORPEDO CONTROL 
 
 127 
 
 contacts by a single platinum point on the movable arm, 
 and an adjustable globule of mercury. This increased the 
 operating sensitiveness from twenty to thirty times, for only 
 an extremely small contact pressure was required to keep the 
 circuit closed under considerable vibration. These relay im- 
 provements therefore increased the receiver's sensitiveness 
 about fifty times. 
 
 100 
 
 Sensitiveness Curve of 
 Remodelled Wesfon Galvanometer 
 
 r-^- 
 
 2 4 e 
 
 FIELD CURRENT AMPS. 
 
 FIG. 69. 
 
 Fig. 70 is a plan view of this improved sensitive relay. 
 T and Ti are terminals of electromagnet windings, W and 
 Wi, surrounding soft iron cores. When in operation T and 
 Ti are connected to a source of direct current. T2 and T3 
 are terminals of movable coil C, the pivots and mountings of 
 which are not shown. B is a light arm fixed to C. Terminal 
 T4 is connected to arm B. L is a non-oxidizable contact 
 piece fixed to B. M is the top of a column of mercury ex- 
 tending into and above the block H. The size of the globule 
 
128 
 
 RADIODYNAMICS 
 
 M above H is adjustable by screw S. The distance of M 
 from L in its normal inoperative position may be varied by 
 the adjusting screw Si. 
 
 Ordinary instruments of this kind have permanent magnets, 
 but by the use of electromagnets and suitably shaped pole 
 pieces, a much more intense field, and consequently a greater 
 sensitiveness, could be secured. 
 
 FIG. 70. 
 
 When C is energized by current flowing in the right direc- 
 tion, arm B carrying contact L will move toward M. L will 
 make contact with and move into M and establish a good 
 low-resistance connection in the local circuit connected to T4 
 and T5. When C is de-energized, the spiral spring Q causes 
 B to return to the normal position. With a relay of this 
 description currents of a few microamperes could be relayed 
 under conditions of vibration which necessitated a current of 
 a hundred microamperes with the best of ordinary sensitive 
 relays of the solid contact type. 
 
BATTLE-RANGE TORPEDO CONTROL 
 
 129 
 
 The next step in the solution of the control problem was 
 to discard the Lodge-Muirhead coherer, and to adopt the 
 vacuum-tube rectifier, which was perfected in this country 
 by DeForest. This is about twice as sensitive as the best 
 solid or electrolytic rectifiers, and has the additional advantage 
 of being more stable, both electrically and mechanically. 
 
 2400 
 
 2000 
 
 1600 
 
 1200 
 
 soo 
 
 400 
 
 Distance- Intensity Received Current 
 on "D/RiG/A"'wifh Pofentio" Detector. 
 
 In attempting to improve this detector, the writer dis- 
 covered a connection arrangement which made the detector 
 a true potential-operated device. The other existing forms 
 of vacuum detectors as well as the many forms of solid recti- 
 fiers, electrolytic, thermal, thermoelectric and other detectors 
 are practically all conceded to be current operated, and be- 
 cause of this fact they not only consume energy, but also 
 
130 RADIODYNAMICS 
 
 decrease the receiver's selectivity by increasing the damping 
 of the receiving circuits. This change in the receiving circuit 
 made the instrument approximately twenty-five times as sensi- 
 tive for relay or indicating instrument operation; this can 
 be readily observed from a comparison of Fig. 7 1 with Fig. 66. 
 Both curves were made on the same trip, one going out to 
 sea and the other returning, in order to insure the greatest 
 possible similarity in the conditions. The transmitting energy 
 was also kept constant, the only variable factor being the 
 distance. 
 
 To prove that this circuit arrangement made the detector 
 a potential operated device, four of these detector circuits, each 
 with its separate indicating instrument, were arranged so that 
 they could be simultaneously in connection with an antenna 
 circuit tuned to distant signals. It was found that in no case 
 was the signal intensity in the first set decreased by connect- 
 ing on one or all of the other three. Moreover the signals 
 in all four receivers were approximately equal. The slight 
 inequalities were due to difference in sensitiveness of the 
 separate detectors. The effect on a single indicator could be 
 proportionally increased by connecting it to the secondary 
 winding of a transformer, having separate primaries which 
 were connected to the separate detectors. Theoretically this 
 circuit furnishes a means of securing any received current 
 desired, simply by connecting a sufficient number of units. 
 
 With these circuits the vacuum detector can be adjusted so 
 that the paralyzing effect of strong signals is not encountered. 
 This makes the detector electrically stable and is a very im- 
 portant feature. The detector can also be adjusted so that the 
 local battery current, which we shall call the field current, 
 increases or decreases as desired, when the signals arrive. We 
 will not attempt to give a theoretical consideration of the 
 detector's action, but simply explain the various circuits 
 employed. 
 
BATTLE-RANGE TORPEDO CONTROL 
 
 Til 
 
 Fig. 72 represents a vacuum tube detector, comprising ex- 
 hausted glass bulb H, in which are fixed filament W, grid G, 
 and plate F, the terminals of which are 
 T, Ti, T2, and T$. These terminals lead 
 through H in the usual manner. Fi, Gi, 
 and Wi, show more clearly the shapes and 
 relative sizes of the plate, grid, and filament, 
 respectively. 
 
 Fig. 73 is a diagrammatic representation of 
 the author's circuit arrangement for use with 
 the instrument shown in Fig. 72. When used in a circuit, such 
 as that shown in Fig. 72, this vacuum tube is called a Poten- 
 tio detector. That part of the diagram included in the circle 
 J is the Potentio detector circuit, and the remainder is 
 
 simply one of the large number 
 of ways in which it may be ap- 
 plied in a radio receiving set. 
 
 W is the hot wire filament, 
 which is maintained at an in- 
 candescent temperature by the 
 battery Bi, the degree of incan- 
 descence being varied by resist- 
 ance R. G is the grid, and F 
 the plate or cold electrode, which 
 
 FIG. 73- 
 
 is connected through the indicating instrument I, such as a 
 telephone, to the positive pole of the battery B. This bat- 
 tery in practice consists of about thirty cells; it is connected 
 to W through the variable connecting means K. The grid 
 terminal T is connected to some point in the receiving circuit 
 where the highest potentials are developed by the incoming 
 waves. 
 
 In this receiving circuit A is the antenna, L the open cir- 
 cuit tuning and coupling inductance, C a variable tuning 
 condenser, and E the earth connection. This open receiving 
 
132 
 
 RADIODYNAMICS 
 
 circuit is coupled to the closed resonant circuit comprising 
 inductance Li and condenser Ci. 
 
 Fig. 74 illustrates another type of receiving circuit employ- 
 ing the well-known Oudin resonator principle for increasing 
 the potentials. 
 
 Fig. 75 illustrates a circuit arrangement and apparatus 
 used in producing an indicated effect greater than can be 
 obtained with one detector. With ordinary current-operated 
 detectors only one can be used advantageously in a receiv- 
 ing circuit, since with a plurality of detectors requiring current 
 energy for their operation, the energy of the incoming waves is 
 
 LLJ 
 
 Lz 
 
 
 I U 7T 
 
 FIG. 74. 
 
 FIG. 75. 
 
 divided between the detectors, and consequently no increase 
 in effect is obtainable. It is possible and advantageous, how- 
 ever, to connect a plurality of Potentio detectors to one 
 antenna circuit, or circuits coupled to a single antenna, in 
 order to obtain an indicated effect much greater than is 
 possible with, one Potentio or with other detectors. 
 
 In Fig. 75 the antenna A2, condenser 4, inductance 1/3, 
 and earth E form the open receiving circuit. To a point of 
 maximum potential, such as T5, is made a connection which 
 leads to the grid terminals T of a plurality of Potentio circuits 
 represented by 0, Oi, and 62. O, Oi, and 02 are similar to 
 that part of Fig. 73 included in the line J, with the exception 
 
BATTLE- RANGE TORPEDO CONTROL 
 
 that the primary windings P, Pi, and P2 are used with O, 
 Oi, and O2 respectively instead of the indicator represented 
 by I. D is the core of a transformer in which P, Pi, and ?2 
 are the primaries acting conjointly on secondary S. The 
 latter as shown is connected to indicating instrument Ii. 
 This obviously receives the effects of O, Oi, and 62 combined 
 when the receiving circuit A2, L5, C4, E2 is energized by 
 received currents. 
 
 Fig. 76 is the cascade circuit for amplification. In this cir- 
 cuit arrangement the received energy develops potentials in L6 
 
 LL|A 5 
 
 L6 
 
 /C5" 
 
 FIG. 76. 
 
 FJG. 77. 
 
 which so influence the Potentio detector 03 as to cause varia- 
 tions in the battery current flowing through ?3. The varia- 
 tions of current in P3 induce corresponding variations in the 
 secondary winding Si. These are of higher potential and in 
 turn effect the induction of currents of still higher potential 
 in 82. The final effects at the indicator 1 2 are thus increased 
 considerably over the initial effects. 
 
 Fig. 77 represents an adaptation of the Potentio circuit 
 which has been found especially valuable for the operation 
 of indicators of the galvanometer type. Experimental results 
 prove it to give indicated effects on galvanometers 25 times 
 as great as those obtainable under similar conditions with 
 solid rectifiers and other well-known detectors of equal 
 
134 RADIODYNAMICS 
 
 sensitiveness. No improvement, however, is noted in tele- 
 phone operation. 
 
 The antenna circuit comprising antenna A4, inductance 
 coil L8, condenser C6, and earth 4, is coupled to the coil Lg 
 by means of L8. The condenser C; must be connected 
 in the circuit as shown in order to secure the greatly in- 
 creased effect not noticeable with other detectors or circuit 
 arrangements. 
 
 In practice 63 and Ri are adjusted until the desired indi- 
 cation is secured at 13, care being taken that the applied 
 voltage at 63 is not too high. By experiment the adjust- 
 ment should be made so that a decrease in the normal current 
 occurs when the signal arrives. Then by varying the capacity 
 of C; while signal impulses are arriving the indications at 13 
 may be made to remain during and after the actuating signal 
 has ceased. The length of this time of indication after the 
 signal has ceased can be increased or diminished by variation 
 of the capacity Cy. With C; short circuited or with the 
 connection to C.J broken the effects at I are very much less, 
 so that it can be readily understood that the presence of Cy 
 between the cold electrode and the end of Lg not connected 
 to Gi fulfills the condition necessary for obtaining the de- 
 sired operation. 
 
 There is no danger of burning out 13 by the action of ex- 
 cessively strong signals, since with signals above a certain 
 intensity value, the current in 13 will remain at zero. The 
 normal field current is diminished by the incoming signals in 
 proportion to the strength of these signals. This is another 
 important feature. 
 
 The Potentio detector stood up very well under the very 
 severe conditions. It must be remembered, as is shown by 
 the received current curve (Fig. 71), that the detector must 
 be rugged enough to remain in perfect adjustment under the 
 strongest signals received within a few hundred feet of the 
 
BATTLE-RANGE TORPEDO CONTROL 135 
 
 transmitter, and at the same time it must be sensitive enough 
 to operate the relay at the extreme range of eight miles. It 
 must be capable of performing these functions for hours at a 
 time, perfectly, without a single hitch, or the necessity of 
 adjustment. The strongest signals in our experiments in- 
 variably caused the familiar "blue arc" between the plate 
 and filament, with the usual forms of connection with this 
 type of detector. This necessitates opening and reclosing of 
 the field battery circuit to restore the normal condition of 
 sensitiveness. 
 
 With the Potentio this is impossible. The blue arc is 
 caused by too great a density in the ionic field current. There 
 is a critical value which varies for different bulbs depending 
 on the degree of exhaustion and the distances between the 
 plate and filament. It is only necessary to bring the field 
 current to the critical value either by increase in the field 
 battery voltage, or by superimposing the currents arising 
 from incoming signals upon the normal battery current 
 through the ionic field, to start the arc, 
 
 It is possible to obviate this trouble, but only at the ex- 
 pense of sensitiveness. The most sensitive adjustment is 
 obtained when the field current is just below the critical 
 point. If the incandescence of the filament or the voltage 
 of the field battery is reduced sufficiently, the normal field 
 current can be made so low that the strongest incoming 
 oscillations will not cause a sufficient superimposition of cur- 
 rent to bring its value up to the critical point. But the cost 
 of this freedom from arcing is a great reduction in sensitive- 
 ness. 
 
 The adjustment of the Potentio is such that the normal 
 field current is safely enough below the critical value to allow 
 for increases due to occasional vibrations of the plate and 
 filament, which reduce the distance between them, and yet 
 high enough to insure good sensitiveness. Instead of in- 
 
136 RADIODYNAMICS 
 
 creasing the field current the received oscillations decrease 
 its value. The strongest signals, instead of causing the blue 
 arc, can only bring the field current down to zero from its 
 normal value. Thus it is seen that the signal effect of the 
 Potentio is a change in the normal current, a change which 
 decreases its value away from instead of towards the critical 
 point; that instead of producing excessively strong indicator 
 operating currents with excessively strong signals, the 
 Potentio automatically prevents such an effect by ceasing to 
 furnish an increase in field current change when signals in- 
 crease above a certain critical value. 
 
CHAPTER XVI 
 
 THE DIFFICULTIES ENCOUNTERED IN PROVIDING 
 PROTECTION FROM INTERFERENCE 
 
 The selectivity problem, which is by far the greatest of all 
 difficulties to be overcome i-n the successful evolution of a wire- 
 lessly controlled torpedo, is one of comparative difficulty, 
 depending upon the degree of non-interferability desired. A 
 selective receiver is like a safety vault. The operator at the 
 transmitter may be likened to the owner of a safe who alone 
 possesses the combination. No safes are absolutely burglar 
 proof, and their value depends principally on the length of 
 time required for a skilled cracksman to reach the inside. 
 Likewise no receivers are absolutely selective for the simple 
 reason that an operator bent on interfering can take observa- 
 tions and measurements on the signals sent out to the selec- 
 tive receiver (just as the burglar may watch the opening of 
 a safe by the owner), and adjust his own apparatus to emit 
 waves of exactly the same characteristics as those of the 
 transmitter designed or adjusted for the selective receiver's 
 operation. The burglar, instead of trying to learn the com- 
 bination, may use sheer force in reaching the inside of the 
 safe. In the same manner a hostile transmitting station can 
 perform the desired effect in the selective receiver, i.e., operate 
 the receiving indicator or relay, by the emission of very strong 
 signals so that forced oscillations are set up. This is known 
 as the "whip crack" effect, and it is believed that very few 
 receivers are immune from it. 
 
 The best wireless signaling sets, such as those used by the 
 U. S. Navy are considered very selective, and yet the inter- 
 
138 
 
 RADIODYNAMICS 
 
 ference existent between them is very serious. Take, for ex- 
 ample: Washington (Fig. 78) is receiving a message from the 
 Eiffel tower station in Paris, which is sending at 3000 meters 
 wave length. It is easily possible for a station in California, 
 
 t with an equal amount of 
 
 power, and at practically the 
 same distance, to make the 
 Paris message unintelligible 
 at Washington, simply by 
 sending signals at or within, 
 say, two or three per cent of 
 3000 meters wave length. 
 
 Again, some insignificant 
 station with little power 
 within a few miles of Wash- 
 ington, could make it practi- 
 cally impossible for them to 
 receive any messages at all 
 from distant stations by send- 
 ing out broadly tuned sig- 
 nals of high damping. These 
 highly damped, or untuned 
 signals, by the previously 
 
 mentioned whip crack effect, cause the receiving antennae 
 to vibrate in their own periods, and thereby produce a great 
 deal of interference. 
 
 To illustrate: Sing a clear, steady tone into a piano. The 
 string, which when struck emits that tone, will audibly 
 vibrate in resonance. Then shout loud and gruffly into the 
 same piano; practically every string will be set in vibration, 
 producing a dull roar. The first tone corresponds to the 
 signals sent out by a tuned transmitter; the second (noise) 
 to the whip crack signals emitted by an untuned trans- 
 mitter. 
 
 FIG. 78. 
 
 Towers of the U. S. Naval Station at 
 Arlington, Va. 
 
PROTECTION FROM INTERFERENCE 139 
 
 Another comparison might serve to illustrate the degree of 
 selectivity necessary for torpedo control. We have three 
 persons A, B, and C. A and B are together at one place, and 
 C is, say, a mile away. The problem is to allow A to hear 
 what C says while B is shouting in his ear. Impossible, you 
 say? Yet the torpedo problem is practically an exact analogy. 
 We must be able to make the electromagnetic ear of the 
 torpedo hear our control impulses eight miles away while it 
 
 FIG. 79. 
 Transmitter of Telefunken 5 kw. set aboard U. S. S. South Carolina. 
 
 is at the very side of a battleship capable of almost deafening 
 it with the force of it's own powerful signals. 
 
 It is true that we could sidetrack the real problem, and 
 simply use such a large amount of energy at our shore station, 
 that more energy could be delivered to the torpedo at eight 
 miles than the hostile battleship could deliver at a hundred 
 feet. This is possible because the amount of energy that can be 
 efficiently used aboard a modern battleship does not greatly 
 exceed 5 kw. Such a 5 kw. set is shown in Figs. 79 and 80. 
 This is due to the limited size of the antenna. A shore station, 
 with practically no limits on the size and height of its antenna, 
 
140 
 
 RADIODYNAMICS 
 
 can easily use 100 kw. The shore station also has the advan- 
 tage of being able to direct its energy somewhat in the general 
 direction of operations. This can be accomplished either by 
 the use of an inverted L-type antenna, with a flat top long in 
 proportion to its height, as suggested by Marconi, or by 
 means of the Bellini-Tosi Radiogoniometer. 
 
 FIG. 80. 
 Receiving set aboard U. S. S. South Carolina. 
 
 Sidetracking the real problem by using a land station of 
 tremendous power is, however, not the practical and efficient 
 solution. In the first place, although desirable, absolute 
 selectivity is not necessary. A receiver that will require, say, 
 fifteen minutes time for the enemy to learn its combination, 
 will, in all probablity, satisfy the requirements. 
 
 We have described the systems used by the principal in- 
 vestigators abroad; those who have observed closely will 
 have seen at once that not one of these systems is immune to 
 intentional interference for the simple reason that no pro- 
 vision is made for avoiding the broadly tuned or whipcrack 
 signals, that may be sent out by any transmitter. 
 
PROTECTION FROM INTERFERENCE 141 
 
 What use, we may then ask, and well, are the various types 
 of codal selectors and protective devices, when any hostile 
 battleship can absolutely lock the receiving apparatus so 
 that not even the controlling operator can get in a signal, 
 by the simple process known to operators as " sitting on the 
 key." 
 
 Systems like those of Anders Bull, Walter, Branley, and 
 others, providing complicated apparatus, aside from not being 
 able to cope with interference, actually defeat their own end 
 by their very presence; designed to increase the reliability of 
 operation, they diminish it by increasing the number of 
 mechanisms, especially those electrically operated, which are 
 likely to get out of order. 
 
 The keynote of success in developing mechanisms that must 
 operate without adjustment is simplicity. The simplest form 
 of distributor that will accomplish the end in view, namely, 
 to close any one of a number of circuits, is all that is neces- 
 sary and indeed is to be greatly preferred. Wirth's ap- 
 paratus which provides means of changing the receiver's 
 wave length, is somewhat nearer the solution, but it, like 
 the others, provides no means of getting away from forced 
 oscillation effects; any system that does not do this is useless 
 for torpedo control. 
 
 The selectivity problem cannot be solved by any form of 
 codal selector or protective device inserted in the receiving 
 circuits after the relay; they must be placed before the relay, 
 that is, they must protect it from interference if they are to 
 serve their purpose. Instruments like the resonance relays 
 and monotelephone amplifiers have this protection inherently 
 by virtue of their vibratory elements tuned to the spark 
 frequency of the transmitter, but these, as pointed out else- 
 where, are subject to vibration, shocks, and sounds. 
 
 In order to reach the solution we must devise systems that 
 not only have a high degree of selectivity for tuned signals, 
 
142 RADIODYNAMICS 
 
 but also provide means of evading the whip crack effects of 
 broadly tuned or plain-aerial transmitters. 
 
 Interference preventers have been invented, which, to a 
 large extent, prevent disturbances from untuned transmitters 
 of whip crack signals. At Washington, in 1910, the writer 
 witnessed government tests of such a receiver, worked out by 
 Fessenden. Reception of signals from 
 a station about 400 miles away, at 
 Brant Rock, Mass., was carried on 
 while a five-kilowatt station less than 
 a mile away was sending interfering 
 signals. The wave lengths of the two 
 transmitters were different by only a 
 few per cent. Fig. 81 is a diagram of 
 JLW this receiver. The relation between 
 the two sets of coils is such that when 
 the same current passes through the 
 E two primaries, no current is induced 
 
 FlG " gi in the secondary circuit, the two 
 
 secondary inductances, Li and L2 
 
 being wound in opposition. By tuning each circuit separately 
 to the incoming signals, and then throwing one of them slightly 
 out of tune, the broadly tuned and whip crack signals will 
 divide equally between the two primaries, while the tuned 
 signals will be received by one side alone, their strength un- 
 diminished by the presence of the other circuit to ground. 
 
 When such a receiver is used in conjunction with a trans- 
 mitter of the undamped, continuous wave type, like the high 
 frequency alternator, or the Duddell-Thompson arc, a very 
 considerable degree of freedom from interference is possible 
 for acoustic signaling. 
 
 Whether such a system is selective enough for torpedo 
 control depends upon the effectiveness of two possible methods 
 of producing interference. One is, to listen for the controlling 
 
PROTECTION FROM INTERFERENCE 143 
 
 impulses, measure their wave length, and then adjust their 
 own transmitter to send out similar signals. Whether or not 
 they can do this in the time necessary for the torpedo to reach 
 them (probably ten minutes), is not known. Since probably 
 not more than twenty-five short course-correcting impulses are 
 necessary to guide the torpedo to a target to a distance of, say, 
 six miles^ it is a matter of conjecture. The other interference 
 method is to use one of the siren interference machines in- 
 vented in Germany. It consists essentially of a set of rotat- 
 ing switches, which automatically, and in rapid succession, 
 cause a series of sharply tuned waves of gradually increasing 
 length to be emitted. This, however, as an interference de- 
 vice, has the disadvantage that the available power is divided 
 among the different wave lengths used. 
 
 Assume that the energy is 5 kw., and that we use 20 sepa- 
 rate wave lengths. The energy on each wave length (not 
 taking into consideration the difference in efficiency with 
 change in wave length) is one-twentieth of the total or one- 
 fourth kilowatt. For telephone operation this would not 
 apply, since that instrument would give the full indication 
 during the short time that the particular wave length affect- 
 ing it was being emitted, and thus make reception of other 
 signals impossible; but for relay operation unless the separate 
 wave lengths were each emitted for a time equal to the me- 
 chanical vibration period of the relay's movable element, or 
 longer, the effect would be equivalent to the effect of a one- 
 fourth kilowatt transmitter in continuous operation on the 
 correct wave length. Even though the rotating switches 
 were rotated at a speed slow enough to cause a closure of the 
 relay each time the correct wave length was emitted, the 
 fact that nineteen other similar wave lengths must be sent 
 out in succession, each for the same length of time, makes 
 the interfering impulses so far between that corrections can 
 be made from the control station. 
 
144 RADIODYNAMICS 
 
 The second method, even with the limitations explained, 
 is probably the better of the two methods, as the shore station 
 could send out confusing or blind signals differing in wave 
 length from the steering impulses, so that the enemy afloat 
 would have no means of determining which was actually the 
 correct one. 
 
 In order to increase the selectivity of torpedo operation to 
 such a point where interference is much more impractical, 
 Mr. Hammond and the writer worked out a number of Selec- 
 tive control systems. Of these only a few will be described. 
 Since the work done along this line at the Hammond Radio 
 Research Laboratory is practically the only work of this 
 kind in the United States, and since nothing new is forthcom- 
 ing from European inventors, these selective systems repre- 
 sent the latest improvement along this line. 
 
CHAPTER XVII 
 
 A MEANS OF OBTAINING SELECTIVITY 
 
 Selective Transmitter-Receiver Unit. Fig. 82 illustrates a 
 type of transmitter-receiver unit suggested by the writer in 
 1911. H.F.A. is a high-frequency alternator, or other high- 
 frequency current producer, which supplies energy to Li 
 through interrupters Ii, 12, and key K. 
 
 The principle applied in this selectivity scheme is the same 
 as that brought forward by Blondel in his spark-tuned re- 
 
 41 ^ 
 
 Ai Az 
 
 H.F, 
 
 L3 I 
 
 
 FIG. 82. 
 
 ceiver some time ago, but it has the additional selectivity of 
 one or more other circuits, besides the high-frequency and 
 spark-tuned circuits. Moreover, this other circuit, which is 
 tuned to an intermediate frequency between the two men- 
 tioned at the transmitter, has an inaudible periodicity, so 
 that the signals cannot be heard at all by an ordinary re- 
 ceiver. The wave length of this inaudible frequency is so 
 far above the wave lengths used in signaling that the ordi- 
 nary receiving circuits will not respond to it, and so far below 
 that of the spark-tuned circuits that they would give no indi- 
 cation even if the frequency were audible. 
 
 145 
 
146 
 
 RADIODYNAMICS 
 
 Supposing the stiffness of the receiving circuits is such as 
 to require twenty impressed oscillations to swing them up to 
 full amplitude, then the ratio of the different frequencies at 
 the transmitter should be 20 to i, that is, the transmitter 
 emits several frequencies, all in the same wave, the values 
 of which drop down in steps according to the ratio 20 to i or 
 the ratio found most suitable. 
 
 Consider the wave length of the emitted waves to be 1000 
 meters. The oscillation frequency corresponding to that 
 value is 300,000 per second. The oscillation frequency at 
 H.F.A. would then be 300,000. Allowing 20 oscillations to 
 
 141 
 
 L35 
 
 D.C. 
 
 FIG. 83. 
 
 FIG. 84. 
 
 the wave train, the interrupter Ii would have a. frequency of 
 15,000 per second. That is, at each contact of Ii, 20 com- 
 plete oscillations from H.F.A. would occur in Li and be 
 radiated from A. Dividing 15,000 by 20 we have 750, the 
 frequency of interrupter 12. When key K is closed antenna 
 A then radiates electric waves of 1000 meters length, which 
 are broken up into an inaudible group frequency by Ii. This 
 resultant signal is then broken up into a frequency of a lower 
 order, determined by the speed of 12. 
 
 Figs. 83 and 84 show two other ways of obtaining the same 
 results as with Fig. 82. In Fig. 83 G is an alternating-current 
 generator having a periodicity of about 7500 cycles, which 
 excites the field windings of high-frequency alternator H.F.A., 
 
A MEANS OF OBTAINING SELECTIVITY 147 
 
 the latter being rotated at such, speed as to give a 300,000- 
 cycle current. The current delivered by H.F.A. will then 
 have a periodic amplitude variation of a frequency correspond- 
 ing to G, namely 15,000. H.F.A. delivers this periodically 
 varying 3oo,ooo-cycle current through interrupter I and key 
 K, to antenna A by means of the inductively coupled coils, 
 Li and L2. 
 
 When interrupter I is stationary or short circuited, an- 
 tenna A radiates electric waves of 1000 meters length at an 
 amplitude frequency of 15,000, which being above the audible 
 limit, will not be heard by a spark receiving set. If I is 
 rotating so as to give 750 contacts per second, A will radiate 
 this wave of two periodic characteristics at a rate of 750 per 
 second, which of course is audible. Thus in Fig. 83 G takes 
 the place of Ii, in Fig. 82, for producing the 15,000 per second 
 group frequency. 
 
 Fig. 84 shows another method of producing a high-fre- 
 quency current having two or more group frequencies within 
 or out of the range of audibility. B is an arc oscillatory-cur- 
 rent generator of the Duddell-Thompson type, fed from a 
 source of direct current. In shunt around the arc is a con- 
 denser, Ci, inductance, L2, and interrupter, I. 
 
 Consider I as being short circuited and circuit B-C-Li 
 open, then as is well known in the art, when B, Ci, and L2 
 are properly adjusted, oscillatory currents will be generated 
 in the circuit B-Ci-L2, the frequency of the alternating 
 currents developed being dependent principally on the values 
 of Ci and L2. Now if circuit B-C-Li be closed oscillations 
 will be generated in it of, say, 7500 cycles. This has the 
 effect of producing a 7500 cycle amplitude variation of the 
 current of the circuit B-Ci-L2 and antenna A being in 
 resonance with B-Ci-L2, will radiate electric waves of 
 300,000 oscillatory frequency, and at a group frequency of 
 15,000 impulses per second. 
 
148 
 
 RADIODYNAMICS 
 
 Now if interrupter I, giving 750 contacts per second, be 
 included in the circuit B-Ci-L2, the radiated waves will be 
 broken up into the audible frequency of 750 per second. 
 
 The receiving circuits, as shown in Fig. 82, are tuned to the 
 oscillatory current frequency, the inaudible amplitude fre- 
 quency, and the audible group frequency. Antenna circuit 
 A2-L3-C and circuit L4~Ci are both tuned to the trans- 
 mitter oscillation frequency. By the action of rectifier Ri, 
 L5 receives unidirectional currents from L4~Ci, thereby 
 energizing circuit L6-L7-C2, which is in resonance with the 
 frequency produced by the interrupter giving 15,000 contacts 
 
 *UJ 
 
 FIG. 85, 
 
 FIG. 86. 
 
 per second. If 12 of Fig. 82 is in operation, circuit L8-C3-P 
 will then be energized, and telegraphic signals may be pro- 
 duced at P by the transmitting key K. 
 
 Figs. 85 and 86 show two other methods of producing the 
 ultra audible group frequencies of the high-frequency currents. 
 In Fig. 85, I is a high-frequency alternator supplying current 
 to antenna 6 through interrupter 3 and key 2, and coupling 
 coils 4 and 5. Motor 7 is mechanically connected to coil 5 
 and rotates it in such a way as to produce a periodic ampli- 
 tude variation, the frequency of which is ultra-audible. 
 
 In Fig. 86, antenna 8 is inductively connected to high- 
 frequency alternator 12 and ultra-audible frequency alter- 
 
A MEANS OF OBTAINING SELECTIVITY 149 
 
 nator 16 by means of coupling coils 9 and 10, and 14 and 15 
 respectively. An interrupter n and key 13 are in circuit 
 with 12. 
 
 When the transmitter is in operation, n interrupts the 
 current from 12 at an audible rate, and by the action of 16, 
 the amplitude of the antenna current is varied periodically, 
 the periodicity being dependent upon the frequency of 16. 
 
CHAPTER XVIII 
 
 NATURE OF INDICATOR CURRENTS IN RADIO 
 RECEIVERS 
 
 The absolute necessity for a simple and effective inter- 
 ference preventer for our torpedo control system led the 
 author, in the fall of 1912, to investigate the nature of the 
 phenomenon accompanying the reception and indication of 
 alternating-current waves of radio lengths. 
 
 Prof. G. W. Pierce, of Harvard University, has already 
 done considerable work along this line, and in his book on 
 I I the " Principles of Wireless Telegraphy " are found the 
 results of his extensive researches on detectors and 
 rectification phenomena, together with hypotheses 
 : r ~l based on them. Although Professor Pierce's work has 
 been mainly along the line of determining the cause of 
 rectification in radio detectors, he also presents brief 
 
 but concise discussions on the nature of the received 
 FIG. 87. 
 
 currents in the indicator circuit. 
 
 On pages 173-174, explaining the action of solid rectifiers 
 in a receiving circuit like that shown in Fig. 87, he says: 
 
 "A train of incoming waves produces an alternating e.m.f. 
 in the antenna circuit. This e.m.f., when in one direction, 
 produces a large current through the detector, charging the 
 antenna. When the e.m.f. reverses the current from the 
 antenna to the ground through the carborundum is smaller, 
 thus leaving the antenna charged with a small quantity of 
 electricity. The effect of the whole train of waves is addi- 
 tive, so that this charge on the antenna is cumulative. The 
 accumulated charge on the antenna escapes through the 
 
NATURE OF INDICATOR CURRENTS 151 
 
 telephone shunted about the carborundum, causing the dia- 
 phragm to move. Each subsequent train of waves causes a 
 similar motion of the diaphragm, which is evidenced as a note 
 in the telephone, equal in pitch with the train frequency of 
 the waves. 
 
 "It is immaterial whether the detector permits the larger 
 current to flow upward, charging the antenna positively, or 
 permits the larger current to flow in the downward direction, 
 charging the antenna negatively. The explanation is the 
 same in both cases. 
 
 "With very slight change this explanation can be made to 
 apply also to those cases in which the detector is in a con- 
 denser circuit coupled inductively or directly with the antenna 
 circuit." 
 
 Such a modification consists essentially in substituting the 
 stopping condenser for the capacity, and the coupling coil 
 for the inductance of the antenna. That the two 
 circuits are of the same type is seen by an inspec- 
 tion of Figs. 87 and 88. In both cases the detector . ,. 
 has the alternating e.m.f . impressed upon it and, 1 1 
 as Dr. Pierce told the author personally, the 
 charge accumulates, in the stopping condenser, 
 and discharges through the indicator at a rate FlQ 88 
 equal to the transmitter's group frequency, and 
 in exactly the same manner as in the previously mentioned 
 circuit. 
 
 The best value for the stopping capacity, if this be true, 
 would be such that with signals of medium intensity, it would 
 be completely charged by one wave train. The resistance of 
 the receiving telephone also influences their best value, and 
 must be taken into consideration. 
 
 The application of such an explanation to relay operation 
 is, however, of principal interest to us. There is no reason 
 why the theory applied to arriving wave trains of 1000 per 
 
152 RADIODYNAMICS 
 
 second frequency should not hold equally well for trains of 
 much greater length, provided the stopping condenser is 
 large enough to absorb all the energy delivered by the rectifier 
 during that longer train. Neither is there any reason ap- 
 parent why wave trains composed of equal amplitude oscilla- 
 tions should not act in the same way as do the damped trains. 
 
 Granting these suppositions, we could use undamped oscil- 
 lations in trains of any length, and a stopping condenser 
 sufficiently large to absorb all the energy supplied to it during 
 that time by the detector. At the end of the wave train then 
 the accumulated charge in the condenser would discharge 
 through the relay with much greater effect than could be 
 obtained with short successive wave trains. Experiments 
 based on these suppositions were performed, but no increased 
 relay deflections could be obtained. 
 
 In a sketch of the action of wireless telephonic apparatus, 
 on page 305 of his book, Dr. Pierce says: 
 
 "The receiving apparatus is identical with that employed 
 in wireless telegraphy, and makes use of a receiving antenna 
 coupled with a circuit containing some type of rectifying 
 detector; e.g., an electrolytic detector, a crystal contact de- 
 tector, or a vacuum tube rectifier. About the detector is 
 shunted a sensitive telephone receiver. 
 
 "The action is as follows: If an unmodified train of electric 
 waves having a frequency higher than the limit of human 
 audibility (35,000 vibrations per second) arrives at the re- 
 ceiving station, the receiving circuit, if properly tuned, will 
 sustain electric oscillations which, passing through the de- 
 tector, will be rectified and will give a series of rectified im- 
 pulses to the receiving telephone circuit. 
 
 "These impulses, being all in one direction, will act as a 
 continuous pull on the diaphragm, a continuous pull for 
 the reason that the diaphragm cannot follow the rapid suc- 
 cessive impulses, and because also, on account of the inductance 
 
NATURE OF INDICATOR CURRENTS 153 
 
 of the telephone circuit these impulses are modified electrically 
 into practically continuous current through the receiver.''' 
 
 An application of this explanation for a receiver which 
 discriminates between spark and undamped wave signals 
 for relay operation at once suggests itself. If the inductance 
 of the telephone or relay is high enough to smooth the high- 
 frequency direct-current impulses into a practically con- 
 tinuous current, the indicator current, then, with unbroken, 
 undamped oscillations, is practically unvarying and uni- 
 directional. For spark signals or chopped continuous wave 
 signals of audible frequencies, the high-frequency direct- 
 current impulses are modified into one impulse the length of 
 the train, but the self -inductance of the indicator is insufficient 
 to appreciably affect these longer train-length impulses, so 
 that they pass through unmodified. However, by inserting 
 a choke coil of large value, the broken signals may be greatly 
 reduced in intensity, while the unbroken signals remain 
 practically the same as before, except for a decrease in ampli- 
 tude due to the added ohmic resistance of the choke coil. A 
 selective receiver based on this principle will be described in 
 a subsequent chapter. 
 
 That these two explanations do not agree is evident, but 
 it is difficult to understand why the action for continuous 
 waves should be other than the action for damped trains of 
 waves. 
 
 In November of 1912 the writer performed some experi- 
 ments in the effort to verify either of Pierce's theories, or to 
 unearth the true explanation of the nature of the received 
 current in the indicator circuit. These experiments, although 
 crude and incomplete seem to shed new light upon this little 
 investigated phenomenon. 
 
 The writer presents the data and conclusions derived from 
 these tests in the hope that they may incite further investi- 
 gation. 
 
RADIODYNAMICS 
 
 Experimental Determination of the Nature of the Indicator- 
 operating Currents in a Radio Receiver 
 
 Fig. 89 shows diagrammatically the connections and 
 arrangement of apparatus, and the following table gives data 
 relative to the apparatus used. 
 
 FIG. 89. 
 
 B Storage battery. 
 
 E Ericcson test buzzer. 
 
 K Key. 
 
 V Murdock variable condenser (max. cap. 0.002 mfd.) set at 100 
 
 Vi Murdock variable condenser (max. cap. 0.002 mfd.) set at 120 
 
 V2 Murdock variable condenser (max. cap. 0.002 mfd.) set at 45 
 
 V3 Murdock variable condenser (max. cap. 0.002 mfd.) set at 60 
 
 4 Murdock variable condenser (max. cap. 0.002 mfd.) set at 180 
 
 P Primary Murdock inductive tuner 69 turns. 
 
 Pi Primary Murdock inductive tuner 72 turns. 
 
 S Secondary Murdock inductive tuner, contact stud No. 2. 
 
 Si Secondary Murdock inductive tuner, contact stud No. 3. 
 
 D Iron pyrite detector. 
 
 Di Iron pyrite detector. 
 
 F 3ooo-ohm Schmidt- Wilkes telephone receiver. 
 
 R Weston relay (microammeter movement). 
 
 Fi 300-ohm Marconi wavemeter phone. 
 
 L 2500-ohm, 8-c. p., carbon filament lamp. ' 
 
 1 25oo-ohm (d. c.) No. 34 copper wire coils (2) on a laminated wire 
 
 core. 
 H Distance of apparatus in circle from remainder, 10 feet. 
 
 Experiments and Observations 
 
 With the apparatus arranged and adjusted as shown and 
 in operation, the following experiments and observations were 
 made: 
 
NATURE OF INDICATOR CURRENTS 155 
 
 1. Test for tuning: With the coupling between the coils of 
 the two tuners moderately weak (secondaries about three- 
 fourths the way out of primaries), the tuning was fairly 
 sharp, the point of maximum signal intensity being capable 
 of determination to within i or 2 degrees on either of the 
 variable tuning condensers, V, Vi, or 3. 
 
 2. To prove that tertiary circuit, Si-V3, does not receive 
 its energy direct from primary exciting circuit, V-P, instead 
 of from the rectifier D, as intended: Signals in F were di- 
 minished to inaudibility by variation of Vi. 
 
 3. Test for difference in energy between secondary (S-Vi), 
 and tertiary (Si-V3) circuits: F connected across D, with 
 Pi disconnected, indicated a signal that was only slightly 
 greater than at Di. 
 
 4. With D elements out of contact, it was found necessary 
 to change Vi to 60 degrees for resonance, but signals at F 
 were very sharply tuned and much stronger. 
 
 5. R, Fi, L, and I were separately thrown in circuit with 
 D and Pi with the following results at F, the signal inten- 
 sities being in the order given. 
 
 i O. 
 
 2 R. 
 
 3 ....Fi 
 
 No great difference between the intensities; 
 signals moderately strong. 
 
 4 L. 
 
 5 1. Signals very weak. 
 
 6. All apparatus included by circle H was replaced by a 
 closely coupled set of coils on a laminated core, a variable 
 condenser, and a telephone, all compris- 
 ing a low-frequency oscillatory circuit of 
 such wave length as to respond to the 
 group frequency of the buzzer signals. 
 (See Fig. 90.) The signals at F2 were considerably reduced 
 with this arrangement, but fairly good spark tuning could be 
 secured. 
 
156 RADIODYNAMICS 
 
 7. Change of detectors at D: Detectors of various types, 
 such as the different forms of the vacuum tube rectifier, 
 carborundum, and electrolytic, were used at D with practi- 
 cally no change in results, save that in some instances the 
 signals through the whole series of tests were stronger or 
 weaker due to differences in detector sensitiveness. 
 
 Conclusions Drawn from Tests 
 
 The very fact that energy in the form of tuned high-fre- 
 quency oscillatory currents is developed in the tertiary 
 circuit is a proof that the effect of a train of waves is not ad- 
 ditive; that the whole train of waves does not pass through 
 the indicator as a single pulse in one direction, but that each 
 separate impulse in the train passes through the indicator 
 without losing its distinctness, and is not smoothed down into 
 one impulse with the others in the train. The fact that this 
 tertiary circuit was 10 feet distant from the other circuits, 
 and the fact that when the audions were used as detectors, 
 the signals at F could be cut out completely by de-energizing 
 the filament, show conclusively that the currents in the 
 tertiary circuit were not set up by direct induction from the 
 primary exciting circuit, and that therefore they were due to 
 the "blow excitation" of the distinct high-frequency direct- 
 current impulses arriving from D, the rectifier. The fact 
 that there was tuning, and that the selectivity was so good 
 that the signals at F could be rendered inaudible by varia- 
 tion of Vi, also strongly .support this proof. 
 
 The fact that the signals at F were very nearly as strong 
 as when F was connected across D, indicates that the rectified 
 high-frequency impulses were not flattened greatly, due to 
 inductive resistance, as otherwise the energy delivered by 
 Di to F could not have been so great, and the tuning would 
 not have been so good. 
 
 With D out of contact the result was simply to connect V2 
 
NATURE OF INDICATOR CURRENTS 157 
 
 in parallel with Vi, Pi being in this connecting line, the wave 
 length being thereby increased so that it, was necessary to 
 decrease Vi to 60 degrees in order to restore resonance con- 
 ditions; the current in Pi was therefore alternating instead 
 of pulsating direct, and maximum value instead of half value, 
 due to the chopping off action of the rectifier. The signals 
 at F were, therefore, much stronger. 
 
 The fact that there was no very noticeable decrease in 
 signal intensity when N was changed successively from O to 
 R, F, and L, apparently indicates two things, namely: (i) the 
 resistance of the detector D was very high, for additional 
 resistances up to 2500 ohms in its circuit with P did not 
 greatly change the total resistance of the circuit, since by 
 their addition the signals were not greatly diminished; and 
 (2), coils of wire wound on magnetized or permanent magnet 
 cores present little inductive resistance to high-frequency 
 pulsating currents. (The coil of the Weston relay R sur- 
 rounds a core magnetized by the permanent magnet of the 
 instrument; the coils of the Marconi phone, Fi, are wound 
 on permanent magnet poles.) 
 
 When I, the coils of copper wire inductively wound on a 
 laminated core, having a resistance equal to that of the lamp, 
 L, was connected, the signals became very nearly inaudible. 
 This shows that inductances with soft, laminated iron cores 
 present a relatively high inductive resistance to high-frequency 
 pulsating currents. Since these coils had considerable dis- 
 tributed capacity it is believed the weak signals that were 
 heard were due to the conductive effects of this capacity. 
 
 The intensity of the signals was decreased to this great 
 extent, because the high-inductive resistance of I obliterated 
 the separate pulses in the train, and lumped them into one 
 unidirectional-current impulse the length of the train, which, 
 having a very low frequency, could not swing up circuit 
 Si-V into resonant operation; the detector Di and tele- 
 
158 RADIODYNAMICS 
 
 phone F, therefore received no energy, and so no signals were 
 heard. 
 
 That this lumping action does occur was shown with 
 the low-frequency tuned circuit, which responded to the 
 group frequency of these impulses. The reduction in in- 
 tensity of the received current at F2 in this low-frequency 
 circuit was due to the fact that with the apparatus at hand, 
 a high resistance was inevitable in order to secure the in- 
 ductance necessary for obtaining the long wave length re- 
 quired in that circuit. 
 
 The tests with different detectors show that all give prac- 
 tically the same effects. 
 
CHAPTER XIX 
 THE INTERFERENCE PREVENTER 
 
 The writer devised this receiver in order to utilize fun- 
 damental principles relating to the flow of direct and 
 alternating currents for the production of a highly selective 
 radio system. 
 
 These properties have been applied in wire telephony, and 
 kindred branches of the electrical arts, and, more specifically, 
 deal with electrical circuits and their properties. These 
 properties may be so varied as to make the circuit conductive 
 to currents of constant value, while greatly resisting the flow 
 of varying currents, and vice versa. In other words, by in- 
 serting an electrostatic condenser in a metallic circuit con- 
 nected to a source of alternating potentials, an alternating 
 current would flow, but the same circuit would present an 
 infinitely high resistance to the flow of a direct current. 
 Also by inserting a coil of high self-inductance in a metallic 
 circuit connected to a source of direct unvarying currents, 
 the resistance to direct currents would be low, while the same 
 circuit would greatly impede the flow of an alternating or 
 varying current, the extent of the impedance depending upon 
 the inductance of the coil, the limits between which the 
 amplitude of the current varies, and the frequency of the 
 alternations or variations. 
 
 In radio signaling systems of today two principal kinds of 
 electric wave producers are in general use. The older of 
 these is the spark system, with which electromagnetic waves 
 are generated by the spark discharge of a condenser. The 
 waves are sent out in groups, the group frequency being de- 
 
 159 
 
160 RADIODYNAMICS 
 
 pendent upon the frequency of the alternating current charg- 
 ing the condenser, and the spark-gap setting, and the length 
 of the waves upon the electrical sizes of the condenser, and 
 the inductance through which it discharges. 
 
 The number of waves in a train is governed by the damping 
 of the circuit, which, in turn, depends on the various sources 
 of energy loss, such as heating, and radiation of electro- 
 magnetic waves. 
 
 For example, take a 5oo-cycle transmitter emitting a 1000- 
 meter wave. The 5oo-cycle alternator is connected to the 
 condenser circuit through a step-up transformer. The con- 
 denser is charged during the first half of each alternation of 
 the primary current and discharges across the spark gap 
 when its potential reaches the necessary value. This dis- 
 charge occurs at the peak of the alternating-current wave in 
 the primary circuit, is oscillatory in nature, and the number 
 of oscillations in the train is dependent upon the damping. 
 If the damping is small 15 oscillations may occur in the de- 
 cadent train before the potential drops too low to overcome 
 the resistance of the spark gap. 
 
 The time during which the discharge takes place, therefore, 
 with a looo-meter wave (300,000 frequency), and 15 com- 
 plete cycles to the train, would be - ? of a second, or one 
 
 150,000 
 
 ten-thousandth of a second. Thus with the 5oo-cycle, 1000- 
 meter wave transmitter, once in each one- thousandth of a 
 second the condenser discharges for one ten-thousandth of a 
 second, producing a decadent train of, say, 15 oscillations. 
 
 The other type of transmitter is the continuous wave 
 generator, which either by the high-frequency alternator or 
 the Duddell-Thompson arc, produces continuous undamped 
 waves. The waves, instead of being produced in decadent 
 trains during only one-tenth of the time, as with the spark 
 sets, are generated continuously and with constant amplitude. 
 
THE IN TERFERENCE PRE VEN TER 1 6 1 
 
 At the receiving station the effects produced by the two 
 types of wave generators is somewhat different. 
 
 With the spark set the oscillatory currents developed in 
 the receiving antennae by the transmitter, and built up by 
 resonance are rectified, and flow through the indicating in- 
 strument. The telephone, which is used as the indicating 
 instrument, therefore receives a direct-current impulse for 
 each discharge of the transmitting condenser. These im- 
 pulses, although consisting of a number of separate impulses, 
 act as one pull on the telephone diaphragm, which vibrates 
 at a rate of 1000 times per second, corresponding to the trans- 
 mitting group frequency. This periodic motion produces an 
 acoustic note of high pitch. Dots and dashes are distinguished 
 by the length of time during which the note is produced, 
 dots, say, one-tenth second, and dashes one-fifth to one-half 
 second. This note is produced in the receiving telephone 
 only when the transmitting key is closed. 
 
 With the undamped wave transmitter there is no audible 
 variation in the received rectified current, the effect of which 
 is practically the same as that of a continuous direct current 
 in the receiving telephone. Therefore a continuous pull on 
 the diaphragm results so long as the sending key is depressed. 
 
 This, then, is the essential difference, from a low-frequency 
 consideration, between the effects of spark transmitters and 
 undamped wave transmitters. The former produces a 
 periodic received current, while the latter produces a constant 
 received current. 
 
 At present radio stations using these different systems pro- 
 duce considerable mutual interference. By the use of suitable 
 apparatus we may differentiate between the two kinds of 
 effects at the receiving station, and thereby secure a greater 
 degree of selectivity. 
 
 The following description covers methods for accomplishing 
 the desired results, with particular reference to the circuit 
 
162 
 
 RADIODYNAMICS 
 
 arrangements illustrated by the drawings. Figs. 91, 92 and 
 93 show graphically three common types of radio trans- 
 mitters. 
 
 Fig. 91 illustrates a spark transmitting set. The alter- 
 nating-current generator G supplies current to primary P 
 of step-up transformer T, through the controlling key K. 
 Secondary S supplies high potential current for charging 
 condenser C, to break down the spark gap SG. When the 
 potential reaches the sparking value, C discharges across SG, 
 and through inductance L, and by electromagnetic induction 
 and resonance, an oscillatory current is set up in the radiating 
 system composed of antenna A, inductance Li, and earth E. 
 
 $ 
 
 JT 
 
 FIG. 91. 
 
 FIG. 92. 
 
 In Fig. 92 a continuous wave transmitter is shown. High- 
 frequency alternator, H.F.A., sets antenna Ai into electrical 
 vibration by means of coupling coils L2 and L3, as is common 
 in the art. The antenna is earthed at E, and signals are sent 
 by making a condition of resonance or dissonance between 
 H.F.A. and the tuned radiating system Ai-I/j-Ei. This is 
 accomplished by short-circuiting or otherwise changing the in- 
 ductance or capacity of the radiating system with the trans- 
 mitter key in such a manner as to produce the desired signal. 
 In this way the signals may consist either of periods of work 
 or of rest of the radiator. 
 
 Fig. 93 illustrates an undamped wave transmitter, based 
 
THE INTERFERENCE PREVENTER 
 
 on the principles of the Duddell-Thompson oscillatory arc. 
 The direct-current generator Gi, which preferably gives a 
 potential of about 500 volts, supplies direct current to the 
 electrodes of the arc AR, through the choke coils L4 and LS. 
 When properly adjusted, electric oscillations are set up in the 
 closed oscillatory circuit, comprising arc AR, condenser Ci, 
 and inductance coil L6, the frequency of which is determined 
 principally by the values of Ci and L6. By electromagnetic 
 induction and resonance oscillatory currents are produced in 
 the radiating system composed of antenna A2, inductance 
 Ly, and earth E2. In order to send signals, the key K2, 
 which establishes resonance, is used. 
 
 FIG. 93. 
 
 FIG. 94. 
 
 Figs. 94, 95 and 96 are schematic representations of re- 
 ceivers based on the idea of reducing interference between 
 spark and undamped wave systems. 
 
 Fig. 94 illustrates the circuit arrangements of a receiver 
 for use with an undamped wave transmitter, such as Fig. 92 
 or 93. 
 
 The action is as follows: By the phenomenon of electric 
 wave propagation and reception, when the transmitting key 
 is closed, alternating currents of high frequency are de- 
 veloped in the resonant receiving antenna system, which in- 
 cludes antenna A$, inductance L8, condenser C2, and earth 
 3. Circuit L9-C4-D is energized, and current energy is sup- 
 
1 64 RA DIOD YNA MICS 
 
 plied to detector D, through the stopping condenser 4. D 
 is a detector such as a thermal or thermoelectric; Lio is a 
 choke coil, and I an indicating instrument for translating the 
 received currents into effects observable with one or more of 
 the physical senses. D produces a unidirectional current in 
 the circuit D-Lio-I, for each wave train impressed upon it. 
 That is, for signals from a 5oo-cycle spark transmitter, D 
 produces a pulsating, unidirectional current in the indicating 
 circuit, the frequency of which is 1000 per second, equal to 
 the transmitting group frequency, and for signals from an 
 undamped-wave transmitter. D produces a unidirectional, 
 unvarying current in the indicator circuit. Therefore it is 
 obvious that the distinguishing difference between spark and 
 undamped-wave signals is that one produces a periodic re- 
 ceived current, while the other produces a constant received 
 current. The detector D, it must be understood, has too 
 much inertia to follow the high-frequency impulses of the 
 oscillatory current, which are of the order of 500,000 per 
 second, but it can and does follow the impulses correspond- 
 ing to the group frequency, which need not be greater than 
 1000 per second. This inertia or lagging action is due to the 
 fact that detectors of this class which are operated by the 
 heat developed by the incoming oscillations, cannot heat and 
 cool with sufficient rapidity to follow the enormously high 
 number of periodic variations in the heat-producing current. 
 
 Referring now to Fig. 94, choke coil Lio is of such value 
 as to greatly impede the flow of the periodically varying cur- 
 rents produced by spark transmitters, while direct currents 
 set up by continuous wave transmitters flow unimpeded. 
 For this reason the interfering effect of a near-by spark station 
 on a continuous wave receiving station is greatly reduced. 
 
 Fig. 95 illustrates another method of securing the same 
 freedom from disturbance. The receiving antenna system, 
 composed of A4, Ln, C5, and 4 is coupled to the closed 
 
THE INTERFERENCE PREVENTER 
 
 oscillatory circuit, comprising Li2 and C6, with which it 
 
 is in resonance. Circuit Li2-C6 supplies oscillatory-current 
 
 energy to detector Di, which furnishes unidirectional current 
 
 to winding W of indicating the instrument and to primary 
 
 Pi of transformer Ti. Secondary 
 
 Si is connected to winding Wi A, 
 
 through stopping condenser C8, and 
 
 rectifier T>2 rectifies the alternating Ln 
 
 current supplied by Si for use at 
 
 Wi. 
 
 The indicating instrument is here 
 represented as a relay in which M 
 is the moving element, but any other 
 
 Cs 
 
 FIG. 95. 
 
 form of indicating instrument may be used, or W and Wi 
 may be independent primaries of an induction coil, both of 
 which, when in operation, produce equal and opposite effects 
 upon a secondary coil, connected to the indicating instru- 
 ment, while one, operating alone, produces the signal effect. 
 
 The operation is as follows: When continuous wave signals 
 are received, Di supplies unidirectional currents to W and 
 Pi. There is no induction of current into Si, because the 
 currents in Pi do not vary, and therefore only W of Ii is 
 energized, and M is attracted, i.e., the relay is operated. 
 
 If periodic currents are delivered by Di, such as are set 
 up by spark transmitters, currents are induced in Si and 
 therefore Wi receives direct-current impulses by the action 
 of D2 and C8. Now Ti, W, and Wi are so proportioned that 
 with spark signals of the common frequencies, the magnetic 
 effects of W and Wi are equal and opposite. M will, there- 
 fore, be unaffected when group-frequency signals are received, 
 but will operate without difficulty with continuous wave 
 signals. 
 
 Fig. 96 represents the circuit arrangements and apparatus 
 necessary to prevent interference from continuous wave 
 
i66 
 
 RADIODYNAMICS 
 
 transmitters to receivers of spark signals, such as are pro- 
 duced by the transmitters of Fig. 91. Antenna A6, induc- 
 tance Li6, condenser Ci2, and earth E6, form the receiving 
 antenna circuit. Coupled to this is the closed circuit com- 
 posed of inductance Li7 and ca- 
 pacity Ci3. The two circuits are 
 tuned to resonance with each other 
 and with the transmitter. When 
 energized, Li7~Ci3 supplies energy 
 to detector D4, through stopping 
 condenser Ci4. D4 delivers uni- 
 directional currents to primary P2 
 of transformer T2. Secondary 82 
 
 Lie 
 
 Ll7 
 
 PQ* 
 
 FIG. 96. 
 
 is connected to telephone F2. The continuous currents 
 produced in the circuit D4-P2 by continuous wave trans- 
 mitters produce no induced currents in 82. Therefore F2 
 does not operate when continuous wave signals are re- 
 ceived. Spark signals, however, produce periodic direct 
 currents in P2, which by induction produce alternating cur- 
 rents in 82 and F2. F2 therefore receives spark signals with- 
 out difficulty, but remains inoperative 
 for continuous wave signals. 
 
 Fig. 97 is a schematic representation 
 of a thermal detector circuit. D is the IG< 97 ' 
 
 fine wire of the thermal detector, which is connected in series 
 with choke coil Li8, indicating instrument 12, and a source 
 of direct current Z, which is a battery and potentiometer. 
 
 This circuit is suitable for use with the antenna circuit 
 shown in Fig, 94 for the continuous wave receiver. 
 
CHAPTER XX 
 DETECTORS 
 
 According to the definition adopted by the standardization 
 committee of the Institute of Radio Engineers, a radio de- 
 tector is "that portion of the receiving apparatus which, con- 
 nected to a circuit carrying currents of radio frequency, and 
 in conjunction with a self-contained or separate indicator, 
 translates the radio-frequency energy into a form suitable for 
 the operation of the indicator. This translation may be 
 effected either by the conversion of the radio-frequency 
 energy, or by means of the control of local energy by the 
 energy received." 
 
 A wrong impression relative to the exact function of a de- 
 tector in a wireless receiver has been prevalent among those 
 engaged in radio work. This misconception, as pointed out 
 by Professor Pierce, is that detectors are more sensitive to 
 electrical energy than the telephone, galvanometer, or relay is. 
 
 Detectors are necessary only because the energy of the 
 high-frequency received current is in an unsuitable form for 
 use with the indicating instruments employed. This is 
 obvious when we consider such instruments as the Hetrodyne 
 receiver of Fessenden's, which is an indicator so arranged 
 that the high-frequency currents themselves operate it no 
 detector or translating device of any kind being required. 
 
 Because the frequency of the oscillations is so high (of the 
 order of a million per second), the moving coils of galva- 
 nometers, the diaphragms of telephones, or even the light 
 fiber of the Einthoven string galvanometer cannot follow them. 
 No motion, and consequently no indication therefore results. 
 
 167 
 
l68 RADIODYNAMICS 
 
 The energy must be applied either in such a form that it 
 acts in one direction on the indicator, as required in the tele- 
 phone and galvanometer, or, if alternately in opposite direc- 
 tions, the frequency of the alternations must be so low that 
 the inertia of the moving parts of the indicator does not come 
 greatly into play. In the case of the telephone this frequency 
 should not exceed 2000 per second, about 1000 per second 
 being the best value; with the Einthoven string galvanometer 
 the best frequency is still lower, in the neighborhood of 100 
 per second; coil galvanometers have such a slow period, about 
 i to 10 seconds, that they, for all practical purposes, are be- 
 yond consideration in this respect. 
 
 True, alternating-current instruments depending on the 
 Thompson effect have been constructed, which give uni- 
 directional deflections for alternating currents of radio fre- 
 quency, and, like the Hetrodyne receiver, do not require a 
 detector, but they are so insensitive that they can be used 
 only where comparatively large energies are received, such 
 as in the wave-meter application by Dr. Seibt.* 
 
 For the operation, then, of our common and most sensitive 
 indicators, we require some form of translating device; this is 
 not, as has been supposed, for the reason that the detector is 
 a wonderfully sensitive instrument, but because it furnishes 
 a means of utilizing the marvelous sensitiveness of these 
 indicators. 
 
 Taken singly the detector is perhaps the most important 
 part of a radiodynamic system. It is to the torpedo what 
 the ear is to a telephone operator; all orders are received 
 through it; without it wirelessly directed torpedoes would be 
 impossible, just as the telephone would be impossible with- 
 out human ears. It is delicate, necessarily, because of the 
 slight effects it must respond to; like the human ear it must be 
 
 * Elihu Thompson, Eke. World, May 28, 1887; see also Proceedings Inst. 
 Radio Engineers, Vol. i, Part 3, 1913; and Phys. Review, Vol. 20, p. 226, 1905. 
 
DETECTORS 169 
 
 able to stand up under heavy cannonading as well as to hear 
 weak signals from a distance; it must be rugged to withstand 
 the severe conditions imposed; rugged, because subject to 
 strong effects, both mechanical and electrical, which tend to 
 break down its original sensitive adjustment; rugged for the 
 reason that the possibility of readjustment in a dirigible 
 torpedo is excluded. 
 
 An ideal detector is one that is extremely sensitive, and at 
 the same time immune to disturbances which make readjust- 
 ment a necessity; one that will operate with the faintest 
 signals, as well as stand up under the strongest electrical and 
 mechanical shocks. 
 
 Although close approaches have been made to this ideal, 
 the perfect detector has not yet been produced. Those in 
 use purely for signaling, i.e., radiotelegraphy and telephony, 
 where an operator is constantly in attendance, are near 
 enough for all practical purposes, but for such work as torpedo 
 control, they are not yet what they should be. Even though 
 the best, namely, those designed or modified especially for. 
 this purpose, do operate perfectly for hours at a time' under 
 the conditions of torpedo control, yet they cannot be de- 
 pended upon absolutely, and absolute dependence, absolute 
 reliability in the detector are pre-requisites for a really suc- 
 cessful dirigible torpedo. 
 
 Since the first electric oscillator of Hertz, which consisted 
 of a bent wire with the ends very near together, a number of 
 different types of detector have been brought out. These 
 new types and modifications have been steadily improved in 
 sensitiveness and reliability. 
 
 Detectors may be classified under the following titles: 
 Coherers. Crystal rectifiers. 
 
 Magnetic detectors. Electrolytic detectors. 
 
 Thermal detectors. Electrometer detectors. 
 
 Thermoelectric detectors. Vacuum detectors. 
 
1 70 RADIODYNAMICS 
 
 A further classification may also be made which places 
 detectors under one of two general heads, namely, potential 
 operated detectors and current operated detectors. 
 
 The following table gives this classification according to the 
 present theories of operation for these detectors: 
 
 Potential Operated Current Operated 
 
 1. Loose contact coherers. Magnetic. 
 (Filings, Lodge-Muirhead, mi- Thermal. 
 
 crophonic contacts, etc.) Thermoelectric. 
 
 2. Capillary electrometer. Crystal rectifiers. 
 
 3. Potentio vacuum detector. Electrolytic detectors. 
 
 Vacuum detectors. 
 
 The potential group operate like a trigger in that they 
 control local sources of energy which effect indicator oper- 
 ation, and depend on the potential of the received currents. 
 
 The current operated group depend upon the current 
 effects of the received energy. In some there is a local source 
 of energy which is called somewhat into play by the action of 
 the received current. The bolometer, which comes under the 
 thermal class, is one of these. In others the oscillatory energy 
 alone affects the indicator operation. Among these is the 
 crystal rectifier, which, chopping off the even or odd alter- 
 nations in a received wave train, leaves only impulses of one 
 sign, positive or negative. In others still, both the incoming 
 energy and local energy called into play by it act upon the 
 indicator. The crystal rectifiers with a local battery are 
 examples of these. 
 
 For torpedo control a detector must be able to withstand 
 the heavy electrical shocks at the shortest ranges, and at the 
 same time be sufficiently sensitive to operate the relay at 
 distances up to eight or ten miles. In addition to this it 
 must not be affected by the mechanical vibration and shocks 
 met with aboard a small self-propelled craft in a rough sea, 
 
DETECTORS 171 
 
 and remain in operative adjustment for at least one hour 
 under such conditions. 
 
 Coherers, as before stated, are sufficiently sensitive, but 
 their action is erratic; heavy received currents cause detri- 
 mental effects; as a whole, they are far from the solution of 
 the detector problem. 
 
 Magnetic detectors are very stable, both electrically and 
 mechanically; they will not burn out with the strongest 
 signals, nor lose their adjustment when subject to severe 
 mechanical shocks, such for instance as those arising from 
 heavy gun fire. Their failing, however, is insensitiveness, in 
 which they are below most detectors in use. 
 
 Thermal detectors, such as Fessenden's barreter and the 
 bolometer, are mechanically stable, but they are subject to 
 burnouts from strong signals, and are insensitive. The fine 
 platinum wire, the resistance changes in which arise from tem- 
 perature variations produced by the oscillatory currents flow- 
 ing through it, can be fused by received currents of excessive 
 intensity. Immunity from these burnouts can only be secured 
 by increasing the thickness of the fine wire, but this again 
 reduces the sensitiveness, which at the best is not even equal 
 to that of the magnetic detector. 
 
 Thermoelectric detectors employing a junction of two dis- 
 similar metals, such as bismuth and antimony, which, when 
 heated by the passage through it of oscillatory currents, 
 produce direct thermoelectromotive forces, have, like the 
 magnetic detector, the necessary stability, but they,, too, are 
 insensitive. They are also somewhat handicapped in having 
 a comparatively large heat capacity, so that a signal several 
 seconds long must be sent before the temperature of the 
 junction rises to the maximum value for a given signal in- 
 tensity; likewise it requires a similar length of time for cool- 
 ing. DuddelPs thermogalvanometer, which is probably the 
 most sensitive of the combined thermoelectric detector and 
 
172 RADIODYNAMICS 
 
 galvanometer, and of thermoelectric detectors in general, 
 though valuable for the purposes of measurements, is not 
 sufficiently rapid or sensitive for use in a radiodynamic system. 
 Crystal rectifiers, sometime called also solid rectifiers, 
 though used in about 90 per cent of radio stations, and though 
 more sensitive than any of those hitherto described, are still 
 too low in sensitiveness for use in torpedo control work. 
 This has been previously pointed out in connection with the 
 received current curve. These also can be burned out by 
 excessively strong signals, so that readjustment is necessary, 
 and they can be thrown out of adjustment by vibration or 
 gun fire. 
 
 Electrolytic detectors are about equal in sensitiveness to 
 the crystal rectifiers, but are not so much used as they were 
 before the advent of the crystal rectifiers. They, too, are 
 subject to burnouts, and the most sensitive types, the free 
 point electrolytics, are not mechanically stable. The glass 
 point electrolytic, in which the fine wire anode is sealed in 
 glass and immersed in the acid electrolyte, though not possess- 
 ing this latter defect to so great a degree is less sensitive and 
 is also subject to burnouts. 
 
 The capillary electrometer detector (see Fig. 98), as in- 
 vented by Armstrong and Orling of England, consists of a 
 minute capillary glass tube filled with mercury. 
 The small end of this tube is immersed in an 
 acid solution. Under the action of a current the 
 electrolytic polarization of the contact causes a 
 change of the surface tension of the mercury. 
 Under this influence the mercury rises or falls in 
 the capillary tube. A low-power microscope is 
 used to observe the minute motion of the mer- 
 cury column. It is said a delicate capillary elec- 
 trometer will give a readable deflection with an applied e.m.f. 
 of one ten-thousandth of a volt. In order, however, to pro- 
 
DETECTORS 1 73 
 
 duce a motion sufficiently to act as a relay (one-sixteenth inch), 
 the e.m.f. must be increased to such an extent that the sensi- 
 tiveness is too much reduced to make the instrument of value 
 for mechanism control. 
 
 Vacuum detectors * have previously been discussed in detail. 
 Some are potential operated, others are current operated, 
 according to the circuit arrangements employed. It has been 
 shown that with a suitable form of circuit, the vacuum de- 
 tector approaches nearer the ideal by far than any other 
 
 FIG. 99. 
 
 detector. Its sensitiveness is as much- as 20 times as great 
 as the best of other detectors, and it is not subject to burn- 
 outs or severe mechanical shocks. It is this detector and 
 the circuit which makes it potential operated that has made 
 possible the extraordinary success attained by Mr. Hammond 
 in the Gloucester torpedo control experiments. It is pictured 
 in Fig. 99 as used in the DeForest System. 
 
 The hetrodyne receiver of Fessenden, depending on the 
 principle of beats for its operation, cannot be used for relay 
 operation, but the beats principle can be applied for ampli- 
 fication purposes. This will be described in another chapter. 
 
 * For detailed accounts of the very important work recently carried out by 
 Dr. Irving Langmuir, Dr. Lee DeForest, and others, see General Electric Review, 
 March 1915, May 1915, and Proceedings Inst. Radio Engrs., Sept. 1915. 
 
174 RADIODYNAMICS 
 
 The frequency transformer, or tone wheel, of Dr. Gold- 
 schmidt is another application of the beats principle. Al- 
 though a very efficient form of detector and very satisfactory 
 for telephones, it is unsuitable for the operation of our most 
 sensitive relays, which require direct current, because it, like 
 the hetrodyne, produces an alternating current for indicator 
 operation.* 
 
 * For complete description of the hetrodyne receiver and U. S. Navy test 
 data, see Proc. Institute Radio Engineers, Vol. i, part 3, 1913; a complete 
 description of the Goldschmidt frequency transformer is contained in Proc. 
 Inst. Radio Engrs., Vol. 2, No. i, 1914. 
 
CHAPTER XXI 
 METHODS OF INCREASING RECEIVED EFFECTS 
 
 Various means for increasing the intensity of received 
 signals have been proposed and utilized within the past ten 
 years. These are called amplifiers, amplifones, variable re- 
 lays, intensifiers, etc., but the generally accepted term is 
 amplifier. It may be defined as a relay which modifies the 
 effect of a local source of energy in accordance with variations 
 in received signals and, in general, produces a larger indica- 
 tion than could be had from the incoming energy alone. 
 
 If a really satisfactory amplifier were available the serious- 
 ness of the detector problem in radiodynamics would be 
 greatly reduced, for then a receiving detector possessing the 
 necessary stability, though lacking in sensitiveness, could be 
 employed. 
 
 To fulfill this requirement, an amplifier must, first of all, 
 be capable of amplifying with a high ratio; and, next in im- 
 portance to this, it must neither be subject to burnouts nor 
 mechanical disturbances; this presupposes no necessity for 
 readjustment for at least several hours; simplicity is also a 
 very desirable element. 
 
 Amplifiers may arbitrarily be classified as follows: 
 
 1. Microphonic contact amplifiers. 
 
 2. Generator amplifiers. 
 
 3. Vacuum tube amplifiers. 
 
 4. Hetrodyne amplifiers. 
 
 Of these the microphonic contact amplifiers were the first 
 to be developed, and they are most used. They consist 
 
 175 
 
176 RADIODYNAMICS 
 
 essentially of a combined telephone receiver and transmitter, 
 the same diaphragm serving both. The rectified received 
 currents flow through the telephone electromagnet on one 
 side of the diaphragm the consequent vibratory motion of 
 which alters the resistance of the adjustable microphonic con- 
 tact on the opposite side. Those in use for radiotelegraphy 
 usually are so made that they will give a maximum response 
 
 FIG. ioo. 
 
 only for impulses of the correct group frequency. These are 
 called spark-tuned or monotelephone relays. The common 
 types employ a diaphragm as the mechanically tuned element. 
 The Pickard, Ruhmer, Brown, and Telefunken amplifiers are 
 examples of this type. Others have a steel reed with a very 
 pronounced period of vibration, to increase the selectivity. 
 Lowenstein has constructed a very sensitive instrument of 
 this type. The instrument devised by F. C. Brown is shown 
 in Figs, ioo and 101. 
 
 This type of amplifier has the disadvantage of being subject 
 to vibration, jars, and sounds; it also requires frequent ad- 
 
METHODS OF INCREASING RECEIVED EFFECTS 177 
 
 justment. Although exploited commercially by several lead- 
 ing radio companies it has never been extensively adopted for 
 commercial use, even for radiotelegraphy. 
 
 The generator amplifier consists of a small generator through 
 the field coils of which the rectified received currents are 
 made to flow. The armature currents, with all the charac- 
 teristics of the field currents, but much amplified, are used 
 for indicator operation. Alexanderson has built such an 
 
 FIG. 101. 
 Connection diagram of Brown relay. 
 
 amplifier, which he designed especially for telephony, and 
 succeeded in securing amplification ratios as high as 20 to i. 
 
 It is believed that amplifiers based on this generator prin- 
 ciple present the most satisfactory solution of amplification 
 problems. They are not subject to mechanical disturbances; 
 they cannot be burned out, and they can be constructed for 
 high amplification ratios. Driven continuously by a small 
 electric motor, a generator amplifier would require no atten- 
 tion or adjustment. They also lend themselves easily to spark 
 tuning when a variable condenser is connected across the 
 field coils. 
 
 Vacuum tube amplifiers have been brought out independ- 
 ently by Lowenstein and DeForest. With three vacuum tube 
 
178 RADIODYNAMICS 
 
 detectors arranged in cascade it is claimed amplification ratios 
 as high as 120 to i have been obtained. Such an arrange- 
 ment is shown in Fig. 102. 
 
 These instruments though possessing a high amplification 
 ratio, and not greatly affected by jars or vibration, have a 
 multiplicity of adjustments and sometimes are thrown out 
 of operation by very strong signals, which produce the familiar 
 "blue arc." Although not so desirable 
 as the generator amplifier, they are 
 much more satisfactory than the micro- 
 phone amplifiers, and may yet be 
 brought to the desired state of perfec- 
 tion.* 
 
 The beats principle has been applied 
 by Fessenden for amplification purposes 
 in radiotelegraphy. 
 
 In the latest form of this receiver, 
 which, as before stated in the chapter on 
 detectors, is called the Hetrodyne re- 
 ceiver, a local source of undamped and 
 variable high-frequency oscillations is arranged so as to act 
 on the receiving antenna circuit, and so adjusted that the fre- 
 quency of its alternations is very nearly equal to the frequency 
 of the incoming waves . The effect of these two very nearly 
 equal frequencies, as in acoustics, is to form electrical beats, 
 or alternate additions and subtractions of the two independent 
 forces, of a periodicity equal to the difference between the two 
 original frequencies. 
 
 The incoming frequency is fixed, but the local frequency 
 can be altered at will, and any beat frequency desired can be 
 produced to suit the acoustic conditions. 
 
 When no beats are produced the two frequencies are equal; 
 
 * For complete description of the DeForest Audion Amplifier, see Proc, 
 Inst. Radio Engrs., Vol. 2, No. i, 1914, page 24. 
 
METHODS OF INCREASING RECEIVED EFFECTS 179 
 
 obviously by calibrating the local source of oscillations, a 
 very useful means of measuring the exact wave length of a 
 distant transmitter is furnished. 
 
 It is said amplification ratios as high as 20 to i have been 
 secured with such an arrangement. The principal difficulty 
 with this system, however, is a reliable generator for the local 
 oscillatory currents at the receiver. Arcs are troublesome 
 and require constant attention; high-frequency alternators 
 are very cumbersome (existing types weighing at least 1000 
 pounds, and possessiig a rotor which makes 20,000 r.p.m.) 
 and at the same time expensive. For this reason it would 
 be next to impossible to utilize this amplifying principle for 
 torpedo control, unless some simple and reliable wave gen- 
 erator be developed.* 
 
 * See Proc. Inst. Radio Engrs., Vol. i, Part 3, 1913. The author in 1911 
 under the direction of Mr. Fritz Lowenstein experimented successfully with 
 vacuum tube rectifiers as a means of producing sustained high-frequency os- 
 cillations for use in beat amplifying and selective systems and also as a wave- 
 generator for radio telephony. (For a very complete consideration of micro- 
 phonic contact amplifiers, see extracts from a paper presented before the I. of 
 E.E., London, which appeared in the Elec. Rev. and Western Elect., Vol. 56, 
 Nos. 23 and 24, "A Telephone Relay," I and II.) 
 
CHAPTER XXII 
 RELAYS 
 
 The importance of a relay in a radiodynamic system is 
 second only to that of the detector, and its requirements are 
 just as exact. That is, it must have great sensitiveness, 
 ruggedness, stability, and small inertia. 
 
 The sensitiveness necessary in the relay to bridge a given 
 distance depends upon a number of factors, namely, the 
 height and power in the transmitting antenna, and the 
 efficiency of the receiving detector, or detector and amplifier. 
 Obviously these factors must be taken into consideration 
 for the reason that they are interdependent. For torpedo 
 control it is of little consequence what the sensitiveness of any 
 single one of the receiving instruments is so long as the final 
 result, namely, the opening and closing of the relay contact, 
 can be reliably effected from the transmitter at the required 
 distance, and so long as the combination is immune to dis- 
 turbances of whatever nature which must be encountered. 
 The sole function of the detector, amplifier, and relay in 
 mechanism control is to open and close an electric circuit 
 at the will of the control operator. Any combination of the 
 above-named instruments that will accomplish this result 
 with absolute reliability is a satisfactory solution of the prob- 
 lems involving each of the three elements separately. That 
 combination, however, which is most simple, least cumber- 
 some, and least expensive is to be preferred. 
 
 In radiodynamic work where the distance is not limited 
 by vision, as it is with torpedoes, each of the elements should 
 have the maximum sensitiveness in order that the distance 
 
 180 
 
RELAYS 181 
 
 of operation may be as great as possible. The desirability 
 of this is evident for such use as call-bell operation in radio 
 signaling. 
 
 Relays are commonly classified as polarized and non- 
 polarized. The motion of the movable element in the former 
 reverses with a reverse in the direction of the current energiz- 
 ing it, while in the latter the motion is always unidirectional. 
 
 In the polarized relays either an armature consisting of a 
 permanent magnet, or a coil through which the current flows, 
 is the movable element. 
 
 No. 554. 
 
 FIG. 103. 
 Polarized relay of the high-resistance type. (Courtesy J. H. Brunnell & Co.) 
 
 The non-polarized types usually have a soft iron armature 
 or core which is always attracted in one direction regardless 
 of the direction of the current in the electromagnet or solenoid 
 influencing it. 
 
 The most sensitive non-polarized relays, such as those used 
 in telegraph offices, require a current of three or four milli- 
 amperes to trip them. The most sensitive of the polarized 
 type, as developed for use with coherer receivers by the 
 Marconi, Slaby-Arco, Ducretet, Telefunken, and other com- 
 panies, require about 400 microamperes under operating con- 
 
182 RADIODYNAMICS 
 
 ditions. Such a relay is shown in Fig. 103. Movable coil 
 relays, with permanent magnet fields and solid local circuit 
 contacts as previously described are more sensitive than 
 the above ferric armature types, requiring in the neighbor- 
 hood of 200 microamperes for operation. When fitted with 
 strong electromagnetic fields and a mercury-platinum contact 
 arrangement, the movable coil relays can be made to operate 
 reliably on from about 30 to 5 microamperes, depending on 
 the mechanical disturbances encountered. 
 
 A very sensitive galvanometer of ordinary construction 
 and about 1000 ohms resistance will give a visible deflection 
 with less than one ten-millionth of a volt, but such an instru- 
 ment requires a very solid support, such as a heavy masonry 
 pillar, and the slightest vibration or current of air will cause 
 the delicately suspended coil to move. Suspension coil gal- 
 vanometers, though possessing very high sensitiveness, can- 
 not be used for relays because of their extreme delicacy. 
 Even uni-pivot galvanometers, such as the portable Paul 
 instruments, which will give a go-degree deflection for 10 
 microamperes, though at least ten times as sensitive as the 
 author's modification of the Weston dual-pivot relay, cannot 
 be used for relay purposes except under ideal conditions in 
 the laboratory. They require leveling screws, and though 
 not to quite so great a degree as the suspension coil galvanom- 
 eter, are still much too delicate for use aboard a torpedo. 
 
 Likewise galvanometers of the vibration type like Eintho- 
 ven's, which are capable of use in radio receiving stations for 
 recording messages photographically over great distances, are 
 not rugged enough for torpedo control work. 
 
 The capillary electrometer can be used as a relay, but, as 
 before stated, its sensitiveness is not sufficiently high. 
 
 It is believed that the remodeled Weston relay, as used by 
 Hammond, is the most satisfactory instrument for this kind 
 of work. 
 
CHAPTER XXIII 
 TORPEDO ANTENNA 
 
 It is not the purpose here to discuss the many details in 
 connection with the ordinary types of antenna used for radio 
 work and means for supporting them. I wish merely to 
 make a few remarks on antennae for special use in torpedo 
 control work, and briefly to describe the most recent pro- 
 posals for improvement of this essential part of the receiving 
 apparatus. 
 
 Obviously, as shown long ago by Marconi, the receiving 
 antenna should be as high as possible, since the received cur- 
 rent increases with the height. Marconi enunciated at one 
 time an empirical law that, for simple vertical sending and 
 receiving antennas of equal height, the maximum working 
 telegraphic distance varied as the square of the height of the 
 antennae. The experiments of the General Electric Co., of 
 Berlin, also roughly agree with Marconi's law. Dr. L. W. 
 Austin has worked out a formula, which, taking into account 
 the antenna heights as well as the transmitting power and 
 atmospheric absorption, gives the approximate signaling 
 range of any transmitter and receiver.* 
 
 The length of the horizontal portion of an antenna is also 
 of some importance. 
 
 We see then that for our torpedo we require an antenna of 
 the greatest possible height and length. It is very doubtful 
 whether, with the type of craft used for torpedoes, this height 
 
 * For a discussion of this equation, see Austin, L. W., Bulletin Bureau 
 Standards, 1911, Vol. VII, No. 3, pp. 315-363, "Some Quantitative Experi- 
 ments in Long Distance Radio Telegraphy." 
 
 183 
 
184 
 
 RADIODYNAMICS 
 
 could be made to exceed the length of the vessel. The best 
 practice, as shown in the antennae, in use on the submarine 
 boats of the navies of the world, substantiate this statement. 
 
 I , The 40-foot Radio in the Gloucester 
 
 experiments had a three-wire inverted 
 L-type antenna, with 6-foot spreaders 
 of light bamboo; it was about 20 feet 
 above the water and about thirty feet 
 long, supported by two 3 -section masts 
 made of the lightest steel tubing con- 
 sistent with strength. These weighed 
 about 15 pounds each. The antenna 
 wire was of the usual phosphor-bronze 
 variety having 7 strands of No. 22 
 wire. A single 1,000,000- volt strain 
 insulator between each spreader and 
 mast-head blocks furnished the neces- 
 sary overhead insulation. For the 
 leading-in insulation a 5oo,ooo-volt roof 
 type leading-in insulator was used. 
 This was protected from the flying 
 spray by an improvised hood. These 
 insulation precautions were taken as a 
 result of experiments which proved 
 their necessity with the potential- 
 operated vacuum detectors. 
 
 Experiments were made with this 
 antenna in the effort to increase its 
 effective length. By increasing its 
 length it would be possible to increase 
 its natural wave length and thus di- 
 minish the value of the energy absorbing loading inductances 
 necessary for tuning to the transmitted waves. 
 In this connection a field worthy of experimentation is one 
 
 FIG. 104. 
 
 Common type of radio 
 tower. 
 
TORPEDO ANTENNAE 185 
 
 which covers the possibilities relative to variation of trans- 
 mitting wave length between two antennae of widely different 
 natural periods. , 
 
 It is well known that a transmitting antenna will operate 
 most efficiently only at that wave length corresponding to 
 the natural period of the antenna with just sufficient induc- 
 tance in series for coupling to the closed circuit. If the wave 
 length be increased energy-absorbing loading inductances are 
 necessary; if decreased, an energy-absorbing series capacity 
 must be used. 
 
 The receiving oscillatory system likewise has a definite 
 wave length for which it will operate most efficiently, and for 
 the same reasons. It is known, however, that the current in 
 an oscillatory circuit is inversely proportional to the wave 
 length, so that although the receiving antenna is operating 
 inefficiently at a wave length below its natural wave length, 
 it is possible that the receiver, as a whole, works at an in- 
 crease in efficiency. Again, while the receiver works best 
 with short wave lengths, the power that can be handled by a 
 transmitting antenna decreases with its natural wave length, 
 and so it is possible that the large transmitting powers made 
 possible by high, large capacity, long wave length antennae 
 will entirely overbalance the detrimental effects due to in- 
 efficiency in the reception of the waves. This presents an 
 interesting field for experimentation.* 
 
 At the suggestion of Dr. Lee DeForest, the "Radio's" 
 antenna was fitted with an extension in the form of two long 
 wires attached to the after spreader and reaching down to a 
 light wooden float 30 or 40 feet astern; the swift motion of the 
 boat kept the wires taut. Long pennant-like pieces of cloth 
 
 * See "Optimum Wave-length in Wireless Telegraphy," by A. H. Taylor, 
 Physical Review, Vol. i, No. 4, Apr. 1913, pp. 321-325. Also, "Determination 
 of Wave-length in Radio Telegraphy," A. S. Blatterman, Electrical World, 
 Vol. 64, No. 7, Aug. 15, 1914, pp. 326-329. 
 
l86 RADIODYNAMICS 
 
 through which light wires connected to the rear end of the 
 antenna were woven, and which stood out almost horizontally 
 from the mast head when the boat was in motion, were also 
 tried. Neither method, however, was found of any material 
 benefit. 
 
 Water Antenna. A very novel form of antenna was in- 
 vented several years ago by Fessenden. It consists of a 
 stream of water thrown vertically upward through a coil of 
 copper tubing by a centrifugal force pump. Although possi- 
 bly inoperative in fresh water a torpedo so equipped might be 
 practicable in salt water, which has a higher conductivity. 
 The hollow coil serves as a means of coupling the water 
 antenna to the receiving apparatus. 
 
 The apparent advantage of such an aerial conductor is that 
 it cannot be shot away by the enemy. No data relating to 
 actual use, either experimental or practical, of an antenna 
 of this type can be found. 
 
 The U. S. Navy has experimented with submerged receiving 
 antennas, for use in signaling to submarine boats equipped 
 with radio apparatus. The antenna consisted of a type of 
 conductor, very heavily insulated with rubber and other in- 
 sulating compounds, known as "rat- tail." The antenna wire 
 was thus completely insulated from the water, although be- 
 neath its surface. The author assisted in these tests which 
 were made at Washington, in 1909. Audible signals were 
 received with such an antenna in the Potomac river at Alex- 
 andria, Va., about seven miles from the two-kilowatt trans- 
 mitter at the Washington Navy Yard. 
 
 These tests after considerable experimenting at Charleston 
 and Boston with submarine boats were finally discontinued. 
 
 Another type of antenna, which has a marked directive 
 effect, and experimented with by Dr. Franz Kiebitz, of the 
 General Telegraph Department, of Berlin, has aroused con- 
 siderable interest within the past two years. A straight wire 
 
TORPEDO ANTENNA iSj 
 
 is stretched horizontally a few feet above the earth, and the 
 receiving apparatus connected in the middle. The best 
 directions of reception are those to which the free ends of the 
 wire point. In other forms the two ends are grounded; in 
 still others only one end is grounded, the receiving apparatus 
 being connected near that end.* 
 
 * See Proc. Inst. Radio Engrs., Dec. 1915: " The Effectiveness of the Ground 
 Antenna in Long Distance Reception." 
 
CHAPTER XXIV 
 
 RECENT DEVELOPMENTS 
 
 Pneumatic Steering Apparatus. The latest development 
 in torpedo-control apparatus has been to discard electric 
 steering gear, and to adopt apparatus designed for use with 
 
 FIG. 105. 
 
 The head telephones enable the operator to listen to the control impulses; the 
 instrument in front of the operator is a searchlight control apparatus. 
 
 compressed air. In Figs. 105 and 106 may be seen a control 
 operator at the Hammond Laboratory. Fig. 107 is a view 
 of this Laboratory, and Fig. 108 is a view of Hammond's 
 latest boat. 
 
 188 
 
RECENT DEVELOPMENTS 
 
 189 
 
 This change has made possible a great simplification of 
 apparatus, and a corresponding increase in reliability; inci- 
 dentally it has also increased the accuracy of control because 
 of the swiftness with which the operations are performed. 
 
 FIG. 106. 
 
 The gratifying results now being secured with pneumatic 
 apparatus are ample evidence of the truth of the afore- 
 mentioned statement that simplicity is a highly important 
 factor in apparatus where adjustment is not possible; and 
 that even very simple electric devices are uncertain in their 
 action. 
 There are only three operations necessary for the control 
 
igo 
 
 RADIODYNAMICS 
 
 of a dirigible torpedo, namely: (i) rudder to port, (2) rudder 
 to starboard, and (3) engine control. The following is one 
 
 FIG. 107. 
 
 of a number of pneumatic systems devised by the author 
 with this " simplicity" idea in mind. In addition to the 
 
 FIG. 108. 
 
 simple and rugged nature of the apparatus it also possesses 
 the advantage that, unlike other systems, it does not require 
 
RECENT DEVELOPMENTS 
 
 191 
 
 an especially trained operator; even with the simplest of the 
 old systems, such as Gardner's and Hammond's, a very con- 
 siderable amount of practice is necessary in order to attain 
 expertness in the boat's control. 
 
 This system was designed in 1912 for use with selector 
 systems in which a gradual back and forth rectilinear motion 
 of the movable selector element, as distinguished from the 
 
 Wire/ess Controlled 
 
 Rudder 
 
 FIG. 109. 
 
 step by step circular motion of some, can be secured. Three 
 of the author's methods as well as that of Gardner's, are 
 adaptable for this purpose. 
 
 As shown in Fig. 109 the function of the movable selector 
 element is to control air valves, which in turn control the 
 energy used in performing the desired operations. A brief 
 description will serve to explain the action of the apparatus. 
 
 Normally the transmitting impulses are of such character 
 
IQ2 RADIODYNAMICS 
 
 that the valve V is partially open, thus allowing compressed 
 air to escape from the tank to the cylinder. At this normal 
 position the pressure inside the cylinder reaches a certain 
 value and then remains constant, due to the pull of the 
 large spring S, and to the action of the adjustable escape 
 valve E. If the valve V is opened wide the piston moves 
 quickly, and with great power to the left, due to the fact 
 that the escape valve cannot take care of this increased 
 inrush of air; if the valve is closed farther than the normal 
 position, the piston will be moved in the opposite direction 
 by the spring S. By this means, the rudder, as shown, can 
 be made to move to either side as swiftly or as slowly as de- 
 sired, and maintained in any position, simply by altering the 
 position of the steering wheel at the transmitting station. 
 This alters the speed, or the ratio of on to off periods, of the 
 impulses. Thus any inexperienced operator can steer the 
 boat. 
 
 For engine control a rotary valve (not shown) operated by 
 the solenoid, is used. This has but two kinds of positions 
 corresponding to start and stop. When it is desired to start 
 the engine, one turn of the steering wheel to the extreme 
 right (farther than for the hard over position) is made; the 
 operation for stopping is exactly the same. This can be 
 done quickly so that no interference with the steering evolu- 
 tioned need be experienced. The required air pressure is 
 maintained in the tank by a compressor actuated by the 
 propelling motor. 
 
 Another system of the writer's depends on the selecting 
 action of a dash-pot retarded solenoid apparatus. By send- 
 ing an impulse of one second, then allowing a short break, 
 and then holding the impulse again, number one circuit can 
 be operated. For the other circuits the first impulse need 
 only be changed to 2, 3, 4, etc., seconds, according to the 
 number of the circuit to be operated. The circuit continues 
 
RECENT DEVELOPMENTS IQ3 
 
 to be closed as long as the last impulse is held; when it is 
 stopped the selector arm returns to the normal position. 
 
 Not mentioning the work now being carried on both in the 
 United States and in Europe on the control of trains by 
 control systems based on electromagnetic induction at dis- 
 tances of a few feet, the latest development along the lines 
 of distant control has been reported from France and Italy 
 in connection with the "F ray" naval experiments made in 
 the Solent. It may be worth recalling that Signor Ulivi 
 made a number of experiments in the presence of the French 
 authorities at Villers-sur-Mer in August of 1913. 
 
 "The 'F rays' were originally discovered by a professor at 
 the University of Nancy, and there has been considerable 
 controversy from time to time as to their potency, and some 
 have even doubted their existence. On the other hand, ac- 
 cording to certain reports, the effects obtained by Signor 
 Ulivi were wonderful, and amazed the French authorities. 
 No less a personage than General Joffre is said to have been 
 impressed by them, and to such an extent that he asked the 
 inventor to prepare a plan by means of which an enemy's 
 magazines and powder supplies might be blown up from a 
 distance. 
 
 "What Signor Ulivi has since done in France has remained 
 a profound secret; in fact it is not known whether he has done 
 anything at all. Immediately after the first articles had 
 appeared in the papers, in August of 1913, it is understood 
 he was asked to go to England to submit some tests to the 
 British Admiralty. His experiments in France were chiefly 
 carried out at Havre and Villers-sur-Mer. They were wit- 
 nessed by General Joffre, General Curieres, de Castelnau, 
 Major Ferrie, and a delegate of the Minister of War, Captain 
 Cloitre. The first tests consisted of a series of submarine 
 mines of which there were ten, placed at intervals of 600 
 meters. Signor Ulivi, at i the appointed moment, touched 
 
IQ4 RADIODYNAMICS 
 
 a lever, and one by one the mines exploded without any 
 visible agent. He declared that he had done it by a con- 
 centration of the power of F rays. He was next asked to 
 blow up some powder magazines in an old hulk, which he 
 also did successfully. 
 
 " The technical officers who had witnessed the tests next 
 wanted to prepare mines in their own way and defied him to 
 explode them. This he is alleged to have refused to do at 
 one moment, and a discussion arose. Were the experiments 
 sincere or not? The question was asked and sides were taken 
 at the time; but the dispute was suddenly hushed up or 
 dropped. The fact is that every subsequent move of Signor 
 Ulivi has been shrouded in mystery."* 
 
 Self-Directing Torpedoes. 
 
 The latest tendencies along torpedo-control lines have been 
 towards the development of apparatus which will give a tor- 
 pedo the power of self -direction. 
 
 In 1912 the author, in collaboration with John Hays Ham- 
 mond, Jr., developed such an apparatus, which was called 
 an " orientation mechanism." It is more generally known 
 now as the "electric dog." It is shown in Figs, no, in 
 and 112. 
 
 "This orientation mechanism in its present form, consists 
 of a rectangular box about three feet long, one and a half 
 feet wide, and one foot high. This box contains all the in- 
 struments and mechanism, and is mounted on three wheels, 
 two of which are geared to a driving motor, and the third, 
 on the rear end, is so mounted that its bearings can be turned 
 by electromagnets in a horizontal plane. Two five inch 
 condensing lenses on the forward end appear very much like 
 large eyes. 
 
 * Extract from an article in the "London Times." 
 
RECENT DEVELOPMENTS 
 
 195 
 
 "If a portable electric light be turned on in front of the 
 machine it will immediately begin to move toward the light, 
 and, moreover, will follow that light all around the room in 
 many complex manoeuvers at a speed of about three feet per 
 
 Pony Relay 
 
 Wiring Diagram- Electric Dog 
 FlG. I 10. 
 
 second. The smallest circle in which it will turn is of about 
 ten feet diameter; this is due to the limiting motion of the 
 steering wheel. 
 
 Upon shading or switching off the light the dog can be 
 stopped immediately but it will resume its course behind the 
 
RADIODYNAMICS 
 
 moving light so long as the light reaches the condensing 
 lenses in sufficient intensity. 
 
 " The explanation is very similar to that given by Jaques 
 Loeb, the biologist, of reasons responsible for the flight of 
 moths into a flame. According to Mr. Loeb's conclusion, 
 which is based on his researches, the moth possesses two 
 minute cells, one on each side of the body. These cells are 
 sensitive to light, and when one alone is illuminated a sensa- 
 
 FlG. III. 
 
 Interior of Electric Dog. 
 
 tion similar to our sensation of pain is experienced by the 
 moth; when both are equally illuminated, no unpleasant 
 sensation is felt. The insect therefore keeps its body in 
 such a position, by some manner of reflex action, as will in- 
 sure no pains, and in this position the forward flying motion 
 will carry it directly toward the source of light. 
 
 "The orientation mechanism possesses two selenium cells, 
 corresponding to the two light sensitive organs of the moth, 
 which, when influenced by light effect the control of sensitive 
 relays, instead of controlling nervous apparatus for pain pro- 
 
RECENT DEVELOPMENTS 197 
 
 duction, as is done in the moth. The two relays controlled 
 by the selenium cells in turn control electromagnetic switches 
 which effect the following operations; when one cell or both 
 are illuminated the current is switched onto the driving 
 motor; when one cell alone is illuminated, an electromagnet 
 is energized and effects the turning of the rear steering wheel. 
 The resultant turning of the machine will be such as to bring 
 the shaded cell into the light. As soon and as long as both 
 
 FIG. 112. 
 Electric Dog in Action. 
 
 cells are equally illuminated in sufficient intensity, the ma- 
 chine moves in a straight line toward the light source. By 
 throwing a switch, which reverses the driving motors con- 
 nections, the machine can be made to back away from the 
 light in a most surprising manner. When the intensity of 
 the illumination is so decreased by the increasing distance 
 from the light source, that the resistances of the cells approach 
 their dark resistances, the sensitive relays break their respec- 
 tive circuits, and the machine stops. 
 "The principle of this orientation mechanism has been 
 
198 RADIODYNAMICS 
 
 applied to the Hammond dirigible torpedo for demonstrat- 
 ing what is known as attraction by interference. That is, 
 if the enemy tries to interfere with the guiding station's con- 
 trol, the torpedo will be attracted to it. The torpedo is fitted 
 with, apparatus similar to that of the electric dog, so that if the 
 enemy turns their search light on it, it will immediately be 
 guided toward that enemy automatically. 
 
 "In order that the search light used by the control opera- 
 tor may not have this same effect, use is made of a gyroscope 
 to keep the turn table upon which the cells are mounted, 
 in a fixed position relative to the earth. In this way no mat- 
 ter how much the torpedo turns, or in what direction it is 
 traveling the selenium cells will always face from the shore 
 and toward the attacking battleship in the open sea. 
 
 "By means of two directive antennae, instead of two sele- 
 nium cells the same principle may be applied for attraction 
 by interference when Hertzian, instead of light waves are 
 used. Sound waves might also be utilized in a similar man- 
 ner so that the sound reaching the torpedo (which would be 
 equipped with two submerged microphones made sensitive 
 and directive by megaphone attachments) from the pound- 
 ing of the battleships engines and other machinery, would 
 effect its attraction in a way analogous to the attraction of a 
 source of light for the orientation mechanism. It is just 
 possible, too, that similar apparatus could be used for the 
 detection of submarines, or for defense against them.' 7 
 
 The electric dog operates in a single plane, the horizontal; 
 the author has developed plans for extending its operations 
 to both horizontal and vertical planes, by using two sets of 
 the orientation apparatus operating at right angles to one 
 another. These plans include the use of all forms of radiant 
 energy. 
 
 * Extract from a paper on Torpedo Control by the author in the Purdue 
 Engineering Review, 1914. 
 
RECENT DEVELOPMENTS 199 
 
 With such a double orientator a new defense against the 
 submarine becomes possible. Captain K. 0. Leon of the 
 Swedish navy has already applied the electric dog principle 
 to the automatic direction of torpedoes, the soun4 waves 
 sent out through the water from the hull of a ship acting as 
 the attracting stimulus; it, is but a step to apply a double 
 "orientator of this type to torpedoes that will seek out and 
 destroy any submarines within its range of hearing. This 
 same type of automatic director is suitable for use with 
 aerial torpedoes, explosive-laden mechanical moths, which 
 will sweep down upon the ships of the air with a sting that 
 will blow them into a thousand pieces. The electric dog 
 which now is but an uncanny scientific curiosity may within 
 the very near future become in truth a real "dog of war," 
 without fear, without h^art, without the human element so 
 often susceptible to trickery, with but one purpose; to over- 
 take and slay whatever comes within range of its senses at 
 the will of its master. 
 
INDEX 
 
 Adams, Prof., experiments of, in 
 electromagnetic induction signal- 
 ling, 15. 
 
 Amplifiers, classification of, 175. 
 De Forest's, 50. 
 Generator, 177. 
 Hetrodyne, 178. 
 Lowenstein's, 62. 
 Microphonic, 175. 
 Monotelephonic, 141. 
 Vacuum tube, 177. 
 Antennae, Austin's law for height of, 
 
 183. 
 
 Circuit adjustment of, 134. 
 Marconi's law for height of, 183. 
 Of Kiebitz, 186. 
 On "Pioneer" 125. 
 On "Radio" 185. 
 Submerged receiving, 186. 
 Torpedo, 183. 
 Water (Fessenden), 186. 
 Armstrong and Orling, capillary elec- 
 trometer, 172. 
 
 Austin, Dr. L. W., experiments by, 
 
 with radiotelegraphic sender, 72. 
 
 Formula of, for antennae height, 
 
 183. 
 Automatic recording telegraph, 3. 
 
 Balloon, dirigible, of Roberts, 86. 
 Battle-range torpedo control, 124. 
 Bell, Alexander Graham, experiments 
 of, in electromagnetic induction 
 signalling, 15. 
 Photophone of, 9. 
 Plan of, for marine signalling, 17. 
 Beck, experiments of, in torpedo con- 
 trol, 100. 
 
 Berger, H. Christian, apparatus of, 
 
 in earth conduction, 71. 
 Bolometer, 10, 41, 51. 
 Boys, radiomicrometer of, 51. 
 Branley, codal selector of, 141. 
 
 Control system of, 105. 
 
 Protective device of, 106. 
 Branly, experiments with Hertzian 
 
 waves, 27. 
 
 "Branly tube," or coherer, 28. 
 Bull, Anders, codal selector of, 141. 
 
 Capillary electrometer (Armstrong & 
 Orling), 172. 
 
 As relay, 182. 
 Cells, silenium, 57. 
 
 Selectivity of, 63. 
 Codal selector, of Anders Bull, 141. 
 
 Branley, 141. 
 
 Walter, 141. 
 
 Wirth, 141. 
 Coherers, 171. 
 
 Branly's, 28. 
 
 Control energy, choice of, 34. 
 Control systems, classification of, 89. 
 
 Beck's, 100. 
 
 Branley's, 105. 
 
 Deveaux's, 96. 
 
 Hammond's, of torpedoes, 122. 
 
 Knauss's, TOO. 
 
 Wirth's, 100. 
 
 Cooke, W. F., needle telegraph of, 3. 
 Crooke's radiometer, 10, 51. 
 Crystal rectifiers, 172. 
 Current, density of, 8, 9. 
 
 d'Arsonval, radiomicrometer of, 51. 
 Davy's sound-relaying system, 10, n. 
 
 201 
 
2O2 
 
 INDEX 
 
 De Forest, amplifier, 50. 
 
 Vacuum-tube rectifier, 129. 
 Density of current between earth- 
 plates, 8, 9. 
 Detectors, 167. 
 
 Capillary electrometer, 172. 
 
 Electrolytic, 172. 
 
 Ion controller (Lowenstein), 24. 
 
 Magnetic, 171. 
 
 Potentio, 134. 
 
 Radiant heat, 46. 
 
 Thermal, 171. 
 
 Thermoelectric, 48, 171. 
 
 Vacuum, 173. 
 
 Deveaux's dirigible torpedo boat, 96. 
 Diathermanous materials, 45. 
 Dirigible torpedo boat (Deveaux), 
 
 96. 
 Dolbear, Prof., electrostatic system 
 
 of, 19. 
 
 Double orientation mechanism, 194. 
 Duddell's thermogalvanometer, 10, 
 
 171. 
 
 Duddell-Thompson arc, 142, 147, 
 160, 163. 
 
 Earth conduction, 3, 67. 
 
 Experiments in selective control, 
 offered by, 67. 
 
 Low resistance of, 8. 
 
 Plan of Berger, 71. 
 
 Plates (Steinheil), 9. 
 Edison, dirigible torpedo c/, 86. 
 
 "Tasimeter" of, 10, 52. 
 "Electric Dog," 194. 
 Electric wave producers, 159. 
 Electrolytic detector, 172. 
 Electromagnet, invention of, 3. 
 Electromagnetic induction, 76. 
 
 Laws of, 3. 
 
 Telegraph, 3. 
 
 First overland system of, 3. 
 Development of, 4. 
 
 Signalling, 15. 
 
 Electromagnetic sounder, 4. 
 Wave systems, 27. 
 
 Later improvements in, 32. 
 Marconi, early experiments of, 
 
 Tesla, experiments of, 28. 
 Electrometer, capillary, 172. 
 
 As relay, 182. 
 Electrons, effect of ultra-violet rays 
 
 on, 64. 
 
 Electrostatic telegraph systems, 10. 
 Of Le Sage, 4. 
 Of Lowenstein, 23. 
 
 Electrostatic and electromagnetic in- 
 duction, 74. 
 
 "F Ray," discovery of, 193. 
 
 Experiments of Uiivi in, 193. 
 Faraday, discovery of laws of elec- 
 tromagnetic induction by, 3. 
 Fessenden, Prof., hetrodyne receiver 
 
 of, 167, 173, 178. 
 Interference preventer of, 142. 
 Submarine signalling system of, 36. 
 Water antennas of, 186. 
 Franklin, Benjamin, experiments of, 2. 
 
 Invention of torpedo by, 78. 
 Frequency transformer (Golcl- 
 
 schmidt), 174. 
 
 Fulton, Robert, experiments by, with 
 torpedoes, 78. 
 
 Gale, Prof., experiments of, in water 
 conductivity, 13. 
 
 Galileo, early theory of, 6. 
 
 "Galvanic excitation" of Steinheil, 8. 
 
 "Galvanic induction of" Steinheil, 8. 
 
 Galvanometers, 2. 
 As relays, 182. 
 
 Gardner, John, sensitive vibratory re- 
 lays of, ii. 
 Torpedo control, system of, 93. 
 
 Galvanoscope, 2. 
 
 Gauss, experiments of, 2. 
 
INDEX 
 
 20 3 
 
 Generator amplifier, 175. 
 Goldschmidt, Dr., frequency trans- 
 former of, 174. 
 
 Goose quills, use of, for insulation, 3. 
 Gray, Stephen, early discovery by, 2. 
 
 Hammond, John Hays, Jr., experi- 
 ments by, 107. 
 Steering apparatus of, 114. 
 Torpedo control, system for coast 
 
 defence, 122. 
 Hammond radio research laboratory, 
 
 work of; 107. 
 Heat, detectors of, 10. 
 Heliograph, i. 
 Henry, invention by, 3. 
 Hertzian waves, 77. 
 
 Branly's experiments with, 28. 
 Lodge's experiments with, 28. 
 Hetrodyne receiver (Fessenden), 
 167, 173, 178. 
 
 Indians, signalling by, i. 
 
 Indicator currents in radio receivers, 
 experiments of G. W. Pierce in, 
 150. 
 Nature of, 150. 
 
 Induction-conduction telegraph sys- 
 tems (Preece's), 25. 
 
 Inductive effects in telephone circuits, 
 
 IS- 
 Infra-red or heat waves, selectivity 
 
 of, 44. 
 
 Use in torpedo control, 41. 
 Interference preventer, 159. 
 
 Of Fessenden, 142. 
 
 Ion controller detector (Lowenstein), 
 24. 
 
 Knauss, experiments of, in torpedo 
 control, 100. 
 
 Leon, Capt. K. O., experiments with 
 torpedoes, 199. 
 
 Le Sage of Geneva, 2. 
 
 Electrostatic telegraph of, 4. 
 Leyden jar, discovery of, 2. 
 Light telephony, 60. 
 Lindsay, James Bowman, experiments 
 
 of, in water conductivity, 14. 
 Lodge, Sir Oliver, experiments with 
 
 Hertzian waves, 28. 
 Lowenstein, amplifier of, 62. 
 
 Electrostatic telegraph of, 23. 
 Ion controller detector, 24. 
 
 Magnetic detectors, 171. 
 Marconi, early experiments of, 31. 
 
 Law of, for height of antennae, 
 
 183- 
 Marine signalling, Bells plan for, 17. 
 
 Method of Rathman, 18. 
 
 Method of Rubens-, 18. 
 
 Method of Strecker, 18. 
 Microphonic amplifier, 175. 
 Micro-radiometer (Weber), 47. 
 Monotelephone amplifier, 141. 
 Morse, foundation of overland system, 
 
 3- 
 Development of overland system of, 
 
 4- 
 
 Sounder of, 4. 
 
 Experiments of, with earth con- 
 duction, 12. 
 
 Report of, to government, 12. 
 "Multipliers" of Steinheil, 8. 
 Muschenbroek of Leyden, 2. 
 
 Nichol's radiometer, 51. 
 
 Orientation mechanism (" Electric 
 
 Dog"), 194- 
 
 Applied to Hammond dirigible tor- 
 pedo, 197. 
 Double, 198. 
 Experiments of Captain Leon with, 
 
 199- 
 Oersted, discovery by, 2, 4. 
 
2O4 
 
 INDEX 
 
 Parabolic reflectors, 42. 
 
 Photophone, of Bell and Tainter, 9. 
 
 Pierce, Prof., G. W., experiments of, 
 in indicator currents in radio re- 
 ceivers, 150. 
 
 Pliny, early discovery by, 2. 
 
 Pneumatic steering apparatus, 188. 
 
 Potentio detector, 134. 
 Adjustment of, 135. 
 
 Poppoff s receiver, description of, 30. 
 
 Preece's induction-conduction system, 
 description of, 25. 
 
 Radiant energy in ether and air, 
 
 vibration frequencies of, 33. 
 Radiant heat detectors, 46. 
 
 Classification of, 47. 
 Radio control, experiments in Europe, 
 
 193- 
 
 Recent developments in, 188. 
 Radiodynamic torpedo (Tesla), 85. 
 Radiodynamics, sound waves in, 36. 
 Radio-Goniometer (Bellini and Tosi), 
 
 44, 140. 
 
 Radiometer, 51. 
 Radiomicrometer of d'Arsonval and 
 
 Boys, 51. 
 
 Radio receivers, indicator actions of, 
 163- 
 
 Indicator currents in, 150. 
 Radiotelegraph, experiments with, 
 
 (Austin), 72. , 
 First, 10. 
 
 Power of transmitter, 35. 
 Range of received power, 35. 
 Radio tower, common type of, 184. 
 Rathbone, Charles, discovery by, 14. 
 Receivers, hetrodyne (Fessenden), 
 
 167, 173, 178. 
 Poppoff s, 30. 
 Selective, 137. 
 
 "Whip-crack" effect in, 137. 
 Receiving wave detector, Varley's use 
 of, 27. 
 
 Rectifiers, crystal, 172. 
 
 Vacuum tube of De Forest, 129. 
 Reflectors, parabolic, 42. 
 Relay, capillary electrometer as, 182. 
 
 Galvanometer as, 182. 
 
 Importance of, 180. 
 
 Improvements of, 1 26. 
 
 Invention of, 5. 
 
 Non-polarized, 181. 
 
 Polarized, 181. 
 
 Resonance, 141. 
 
 Sensitive vibratory, of Gardner, n. 
 Resonance relay, 141. 
 Roberts, dirigible balloon of, 86. 
 Romagnesi of Trent, discovery by, 2, 
 4- 
 
 Sacher, Prof., E., experiments of, in 
 induction, 15. 
 
 Schilling, telegraph of, 2. 
 
 Schweigger, discovery by, 2, 4. 
 
 Searchlights, electric atmospheric ab- 
 sorption and dispersion of rays, 
 
 45- 
 
 Invisibility of rays, 43. 
 Selective receivers, 137. 
 Selective transmitter-receiver unit, 
 
 145- 
 
 Selectivity, means of obtaining, 145. 
 Selectors, 89. 
 Branley's, 92. 
 Codal, 141. 
 Hiilsmeyer's, 93. 
 Walter's, 92. 
 Self-directing torpedoes, 194. 
 
 Experiments of Capt. Leon with, 
 
 199. 
 Orientation mechanism applied to, 
 
 199. 
 
 Semaphore system, i. 
 Signalling, early methods, i. 
 Flag, i. 
 
 Submarine (Fessenden's apparatus 
 for), 36. 
 
'INDEX 
 
 205 
 
 SUenium, 57. 
 
 Cells, 57. 
 
 Selectivity of, 63. 
 Sims, dirigible torpedo of, 86. 
 Siren interference machines, 143. 
 Sonorescent property of substances, 
 
 10. 
 
 Sound waves in radiodynamics, 36. 
 Sound-relaying system, of Davy, 10. 
 Sounder, electromagnetic, of Morse, 4. 
 Steering apparatus (Hammond), 114. 
 
 Pneumatic, 188. 
 Steinheil, system of telegraphy of, 2. 
 
 Radio-telegraphic system of, 10. 
 
 Scheme of earth-plates of, 9. 
 
 Use of railway by, 3. 
 
 Wireless telegraph of, 6, 9. 
 Sturgeon, invention by, 3. 
 Submarine signalling, Prof. Fessen- 
 den's apparatus for, 36, 40. 
 
 Tainter, Sumner, photophone of, 9. 
 "Tasimeter," Edison's, 10, 52. 
 Telautomaton (Tesla), 84. 
 Teledynamics, development of, 4, 5. 
 Teledynamic system, principal parts 
 
 of, 33- 
 
 Telefuncken, transmitter of, 139. 
 Telegraph, automatic recording, 3. 
 
 Electrostatic, 4-23. 
 
 Electromagnetic, 4. 
 
 First invented, 2. 
 
 First overland, 3. 
 
 First in U. S., 4. 
 
 Needle, 3. 
 
 Wireless, 6, 12. 
 Telephone, inductive effects in, 15. 
 
 Invention of, 14. 
 
 Wireless, 14. 
 Telephony, light, 60. 
 Tesla, Nikola, early experiments, 28. 
 
 Invention of wirelessly controlled 
 vessel, 83. 
 
 Radiodynamic torpedo ot, 85. 
 
 Tesla, Nikola, Telautomaton of, 84. 
 Thales, early discovery by, 2. 
 Theophrastus, early discovery by, 2. 
 Thermal detectors, 166, 171. 
 Thermocouple, 49. 
 Thermoelectric detectors, 48, 171. 
 Thermogalvanometer, Duddell's, 171. 
 Thermo-electric pile, 9. 
 Thermopile, 41, 48. 7 
 
 Of Coblentz, 49. 
 
 And galvanometer relay, 51. 
 Thermostat, 53. 
 
 Torpedo, advantages and disadvan- 
 tages, 83. 
 
 Antennae of, 183. 
 
 Battle range, control of, 124. 
 
 Coast defense (Hammond), 122. 
 
 Control systems for, 93, 96, 100. 
 
 Demonstrations of, before U. S. 
 War dept., 121. 
 
 Description of, 79. 
 
 Dirigible, 86, 102. 
 
 Experiments of Leon with, 199. 
 
 First test of, 78. 
 
 Invention of, 78. 
 
 Methods of launching, 80. 
 
 Radiodynamic (Tesla), 85. 
 
 Self-directing, 194. 
 Transmitter of Telefuncken, 139. 
 Trowbridge, Prof., John, study of 
 electromagnetic induction signal- 
 ling, 15- 
 
 Ultra-violet radiations, 64-66. 
 Ulivi, experiments of, in F Rays, 
 193- 
 
 Vacuum-detector, 173. 
 
 rectifier, of De Forest, 129. 
 Vacuum-tube, amplifier, 177. 
 Varley's use of receiving wave detec- 
 tor, 27. 
 Vibration frequencies of radiant 
 
 energy in ether and air, 33. 
 
206 
 
 INDEX 
 
 Waiter, codal selector of, 141. 
 Watson of Llandaff, early discovery. 
 
 by, 2. 
 Wave systems, electromagnetic, 27, 
 
 28,31,32. 
 
 Waves, infra-red, 41, 44. 
 Weber, micro-radiometer of, 47. 
 Wheatstone, needle telegraph of, 3. 
 Automatic recording telegraph of, 
 
 3- 
 
 "Whip-crack" effect in receivers, 137. 
 
 Wilkins, J. W., experiments of, with 
 earth conduction, 14. 
 
 Willoughby Smith, discovers proper- 
 ties of silenium, 57. 
 
 Wilson, Ernest, invention by, of wire- 
 less control of vessels, 83. 
 
 Wireless telegraphy, 6. 
 First, of Steinheil, 6. 
 Later, of Steinheil, 9. 
 Practical, 12. 
 
 Wireless transmission of energy, ex- 
 periments of Nikola Tesla, 28. 
 Wireiessly controlled vessels, system 
 
 for (Wilson), 83. 
 Invention of (Tesla), 83. 
 Description of, 84. 
 Wirth, codal selector of, 141. 
 
 Experiments of, in torpedo control, 
 100, 102. 
 
 Zickler, Prof. E., use of ultra-violet 
 rays in telegraphy, 65, 66. 
 
LIST OF WORKS 
 
 ON 
 
 ELECTRICAL SCIENCE 
 
 PUBLISHED AND FOR SALE BY 
 
 ' D. VAN NOSTRAND COMPANY, 
 
 25 Park Place, New York. 
 
 ABBOTT, A. V. The Electrical Transmission of Energy. A Manual for 
 the Design of Electrical Circuits. Fifth Edition, enlarged and rewritten. 
 With many Diagrams, Engravings and Folding Plates. 8vo., cloth, 
 675 pp Net, $5.00 
 
 ARNOLD, E. Armature Windings of Direct-Current Dynamos. Exten- 
 sion and Application of a general Winding Rule. Translated from 
 the original German by Francis B. DeCress. Illustrated. 8vo. 
 cloth, 124 pp $2.00 
 
 ASHE, S. W. Electricity Experimentally and Practically Applied. 
 Second Edition. 422 illustrations. 12mo., cloth, 375 pp. .Net, $2.00 
 
 ASHE, S. W., and KEILEY, J. D. Electric Railways Theoretically and 
 Practically Treated. Illustrated. 12mo., cloth. 
 
 Vol. I. Rolling Stock. Second Edition. 285 pp Net, $2 . 50 
 
 Vol. II. Substations and Distributing Systems. 296 pp Net, $2 . 50 
 
 ATKINSON, A. A. Electrical and Magnetic Calculations. For the use 
 of Electrical Engineers and others interested in the Theory and 
 Application of Electricity and Magnetism. Fourth Edition, revised 
 Illustrated. 12mo , cloth, 310 pp Net, $1 .50 
 
ATKINSON, PHILIP. The Elements of Dynamic Electricity and Mag- 
 netism. Fourth Edition. Illustrated. 12mo., cloth, 405 pp. .$2.00 
 
 Elements of Electric Lighting, including Electric Generation, Measure- 
 ment, Storage, and Distribution. Tenth Edition, fully revised and new 
 matter added. Illustrated. 12mo., cloth, 280 pp $1 .50 
 
 Power Transmitted by Electricity and Applied by the Electric Motor, 
 
 including Electric Railway Construction. Illustrated . Fourth Edition, 
 fully revised and new matter added. 12mo., cloth, 241 pp. . .$2.00 
 
 AUSTIN, E. Single-Phase Electric Railways. 346 illustrations 
 8x11, cloth, 308 pp Net, $5.00 
 
 AYRTON, HERTHA. The Electric Arc. Illustrated. 8vo., cloth, 479 
 pp Net, $5.00 
 
 BAKER, J. T. The Telegraphic Transmission of Photographs. 63 
 illustrations. 12mo., cloth, 155 pp Net, $1.25 
 
 BARHAM, G. B. Development of the Incandescent Electric Lamp. 
 25 illustrations, 2 plates,. 5 J x8| cloth. 206 pp $2 . 00 
 
 BEDELL, FREDERICK. Direct and Alternating Current Manual. 
 With directions for testing and a discussion of the theory of electri- 
 cal apparatus. Second Edition, greatly enlarged. Assisted by C. A. 
 Pierce. Illustrated. 6x9, cloth, 373 pp Net, $2 .00 
 
 ELAINE, ROBERT G. The Calculus and Its Applications. A practical 
 treatise for beginners especially engineering students. 79 illustra- 
 tions. 12mo., cloth, 330 pp Net, $1 .50 
 
 BONNEY, G. E. The Electro-Plater's Hand Book. A Manual for Ama- 
 teurs and Young Students of Electro-Metallurgy. Fifth Edition, 
 enlarged. 61 Illustrations. 12mo., cloth, 208 pp $1 .20 
 
 BOTTONE, S. R. Magnetos For Automobilists; How Made and How Used 
 A handbook of practical instruction on the manufacture and adapta- 
 tion of the magneto to the needs of the motorist. Second Edition, 
 enlarged. 52 illustrations. 12mo., cloth, 118 pp Net, $1 .00 
 
 BR9ADFOOT, S. K. Motors, Secondary Batteries and Accessory Appara- 
 tus. 16 illustrations. 16mo., cloth. 100pp. (Installation Manuals 
 Series) Net, 0.75 
 
 BROUGHTON, H. H. Electric Canes. Their design, construction and 
 application. 600 illustrations and plates. 120 tables, 6x9, cloth, 
 848 pp Net, $9.00 
 
CHILD, C. D. Electric Arcs. Experiments upon arcs between different 
 electrodes in various environments and their explanation. 58 illus^ 
 trations. 5x7, cloth, 203 pp Net, $2 .00 
 
 COLLIS, A. G. High and Low Tension Switchgear Design. 94 illus- 
 trations and folding plates, 6x9, cloth, 233 pp Net, $3.50 
 
 Switchgear and The Control of Electric Circuits. 47 illustrations. 
 
 4ix6J, cloth, 85 pp. (Installation Manuals Series) Net, 0.50 
 
 COMSTOCK, D. F. , and TROLAND, L. T. The Nature of Matter and 
 Electricity. An outline of modem views. Illustrated. 5x7, 
 cloth, 200 pp Net, $2.00 
 
 COOPER, W. R. Primary Batteries: their Theory, Construction, and Use 
 
 New Edition in Press 
 CRAIG, J. W., and WOODWARD, W. P. Questions and Answers About 
 
 Electrical Apparatus. Third edition revised and enlarged. Illustrated. 
 
 4^x6% Leather, 256 pp Net, $1.50 
 
 CRAMP, W. Continuous Current Machine Design. 137 illustration. 
 8vo., cloth, 240 pp Net, $2.50 
 
 CREEDY, F. Single-Phase Commutator Motors. 98 illustrations, 6x9, 
 cloth, 120 pp Net, $2.00 
 
 CROCKER, F. B. Electric Lighting. A Practical Exposition of the Art 
 for the use of Electricians, Students, and others interested in the 
 Installation or Operation of Electric- Lighting Plants. 
 
 Vol. L The Generating Plant. Seventh Edition, entirely revised. Illus- 
 trated. 8vo., cloth, 482 pp $3.00' 
 
 Vol. II. Distributing System and Lamps. Sixth Edition. Illustrated 
 8vo., cloth, 505 pp. 
 
 CROCKER, F. B., and ARENDT, M. Electric Motors : Their Action, 
 Control, and Application. Second Edition, Revised and Enlarged. 
 169 illustrations. 6Jx9J, cloth, 315 pp Net, $2.50 
 
 CROCKER, F. B., and WHEELER, S. S. The Management of Electrical 
 Machinery. -Being a thoroughly revised and rewritten edition of the 
 authors' "Practical Management of Dynamos and Motors." 
 Eighth Edition. Illustrated. 16mo., cloth, 232 pp Net, $1 .00 
 
 CUSHING, H. C., Jr. Standard Wiring for Electric Light and Power 
 
 Illustrated. 16mo., leather, 156 pp $1 .00 
 
 CUSHING, H. C., Jr. and HARRISON, N. Central Station Management. 
 5^x7^, cloth, 417 pp Net, $2.00 
 
 DAVIES, F. H. Electric Power and Traction. Illustrated. 8vo., cloth 
 293 pp. (Van Nostrand's Westminster Series.) Net, $2 .00 
 
 Foundations and Machinery Fixing. 52 illustrations. 16mo., cloth, 
 
 146 pp. (Installations Manuals Series) Net, $1 .00 
 
DEL MAR, W. A. Electric Power Conductors. Seeond edition reused. 
 69 illustrations. 5|x8, cloth, 330 pp Net, $2 . 00 
 
 DEVEY, R. G. Mill and Factory Wiring. 126 illustrations. 16mo., 
 cloth, 209 pp. (Installation Manuals Series) Net, $1 .00 
 
 DINGER, Lieut. H. C. Handbook for the Care and Operation of Naval 
 Machinery. Second Edition. 124 Illustrations. 16mo., cloth, 
 302 pp Net, $2 .00 
 
 DUNCAN, W. G. and PENMAN, D. Electrical Equipment of Collieries. 
 157 illustrations, 6ix8|, cloth, 329 pp Net, &4.00 
 
 DWIGHT, H. B. Transmission Line Formulas for t Electrical Engineers 
 and Engineering Students. 27 illustrations. 2 folding plates, 12mo , 
 cloth, 143 pp Net, $2 .00 
 
 DYNAMIC ELECTRICITY: Its Modern Use and Measurement, chiefly 
 in its application to Electric Lighting and Telegraphy, including: 
 1. Some Points in Electric Lighting, by Dr. John Hopkinson. 2. On 
 the Treatment of Electricity for Commercial Purposes, by J. N. Shool- 
 bred. 3. Electric-Light Arithmetic, by R. E. Day, M.E. Fourth 
 Edition. Illustrated. 16mo., boards, 166 pp. (No. 71 Van Nos- 
 trand's Science Series.) 50 cents 
 
 UCK, J. Light, Radiation and Illumination. Translated from the German 
 of Paul Hogner. 59 illustrations. 5: Jx8, cloth, 100 pp. Net, $2.50 
 
 ECCLES, W. H. Wireless Telegraphy and Telephony. Illustrated. 5x8*4 
 half leather, 418 pp Net, $4.50 
 
 JEDDY, L. C. Laboratory Manual cf Alternating Currents. 21 illustra- 
 tions, 5J^xS, cloth, 26 pp Net, 50 cents 
 
 EDLER, R. Switches and Switchgear. Translated by Ph. Laubach. 
 335 illustrations. 6x9, cloth, 412 pp Net, $4 . 00 
 
 EDGCUMBE, K. Industrial Electrical Measuring Instruments. Illus- 
 trated. 8vo., cloth, 227 pp. 
 
 ENGINEERING AS A CAREER. By Prominent Engineers. Edited by 
 C. E. Drayer and F. H. Newell. Stftf^, 226 pp. Paper 75 cents, 
 cloth -.$1 .00 
 
 EWING, J. A. Magnetic Induction in Iron and other Metals. Third 
 Edition, revised. Illustrated. 8vo.. cloth, 393 pp Net, $4.00 
 
FISHER, H. K. C., and DARBY, W. C. Students' Guide to Submarine Cable 
 Testing. Fourth Edition, enlarged. Illustrated. 8vo., cloth, 
 252 pp Net, $3.50 
 
 FLEMING, J. A., Prof. The Alternate-Current Transformer in Theory 
 
 and Practice. 
 Vol. I. : The Induction of Electric Currents. Fifth Issue. Illustrated. 
 
 8vo., cloth, 641 pp Net, $5.00 
 
 Vol. II. : The Utilization of Induced Currents. Third Issue. Illus- 
 trated. 8vo., cloth, 587 pp Net, $5.00 
 
 Propagation of Electric Currents in Telephone and Telegraph Con- 
 ductors. Illustrated. 6 }$x9}j, cloth, 323 pp Net, $3 .00 
 
 Handbook for the Electrical Laboratory and Testing Room. Two Vol- 
 umes. Illustrated. 8vo. ; cloth, 1160 pp. Each vol Net, $5.00 
 
 FOSTER, H. A. With the Collaboration of Eminent Specialists. Electri- 
 cal Engineers' Pocket Book. A handbook of useful data for Elec- 
 tricians and Electrical Engineers. With innumerable Tables, Dia- 
 grams, and Figures. The most complete book of its kind ever pub- 
 lished, treating of the latest and best Practice in Electrical Engineer- 
 ing. Seventh Edition, completely revised and enlarged. Fully Illustrated. 
 Pocket Size. Leather. Thumb indexed. 1636 pp $5.00 
 
 -Engineering Valuation of Public Utilities and Factories. 50 blank 
 
 forms. 6x9, cloth, 361 pp Net, $3-00 
 
 Electrical Cost Data, Estimates and Working Tables. In Press 
 
 FOWLE, F. F. The Protection of Railroads from Overhead Trans- 
 mission Line Crossings. 35 illustrations. 12mo v cloth, 76 pp 
 
 Net, $1.50. 
 
 FREUDEMACHER, P. W. Electrical Mining Installations. 36 illustra- 
 tions. 16mo., cloth, 192 pp. (Installation Manuals Series) .Net, $1 .00 
 
 FRITH, J. Alternating Current Design. 27 illustrations. 6x9, cloth, 
 131 pp Net, $2.00 
 
 GARRARD, C. C. Electric Switch and Controlling Gear. 423 illustra- 
 tions, 5HxS, cloth, 656 pp Net, $6.00 
 
 GANT, L. W. Elements of Electric Traction for Motonnen and Others. 
 Illustrated with Diagrams. 8vo., cloth, 217 pp . . Net, $2 . 00 
 
 GEAR, H. B. and WILLIAMS, P. F. Electric Central Station Distribution 
 Systems. Their Design and Construction. Second Edition, revised 
 and enlarged. 187 illustrations. 5&x?H, cloth, 455 pp. . .Net, 3.50 
 
GERHARDI, C. H. W. Electricity Meters; their Construction and Man- 
 agement. A practical manual for engineers and students. Illus- 
 trated. 8vo., cloth, 337 pp ........................... Net, $4 .00 
 
 GOLDSMITH, RUDOLF. The Alternating Current Commutator Motor. 
 The Leakage of Induction Motors. In one volume. 247 illustrations. 
 5ix8|, cloth .................... ..................... Net, $3.00 
 
 GORE, GEORGE. Tne Art of Electrolytic Separation of Metals (Theoret- 
 ical and Practical). Illustrated. 8vo., cloth, 295 pp ____ Net $3.50 
 
 GROTH, L. A. Welding and Cutting Metals by Aid of Gases or 
 Electricity. 124 illustrations. 8vo., cloth, 280 pp. (Van Nostrand's 
 Westminister Series) ................................ Net, $2 . 00 
 
 HALLER, G. F. and CUNNINGHAM, E. T. The Tesla High Frequency 
 Coil; its construction and uses. 12mo., cloth, 56 illustrations, 130 
 pp ........................ ...... . .................. Net, $1.25 
 
 HASKINS, C. H. The Galvanometer and its Uses. A Manual for Elec- 
 tricians and Students. Fifth Edition, revised. Illustrated. 16mo., 
 morocco, 75 pp .......................................... $1 . 50 
 
 HAUSMANN, E. Telegrr.ph Engineering. A Manual for Practicing 
 Telegraph, Engineers and Engineering Students. 192 Illustrations. 
 , cloth, 416 pp ................................ Net, $3 .00 
 
 HAY, ALFRED. An Introductory Course of Continuous-Current En- 
 gineering. Illustrated. 8vo., cloth, 327 pp ............ Net, $2 . 50 
 
 HAYES, H. V. Public Utilities, Their Cost New and Depreciation. 5fx8,- 
 cloth, 275 pp .................... .................... Net, $2.00 
 
 - Public Utilities; Their Fair Present Value and Return. 5fx8t, cloth, 
 200 pp ............................................. Net, $2.00 
 
 HEATHER, H. J. Electrical Engineering for Mechanical and Mining 
 
 Engineers. 183 illustrations. 5fx8|, cloth, 344 pp ..... Net, $3.50 
 
 /f 
 
 HEAVISIDE, 0. Electromagnetic Theory. Two Volumes with Many 
 Diagrams. 8vo., cloth, 1006 pp. Each Vol ............ Net, $5 .0*0 
 
 Vol. IH. 529 pp ................................. ---- .Net, $7.50, 
 
 HOBART, H. M. Heavy Electrical Engineering. Illustrated. 8vo., > 
 cloth, 338 pp ............... - ---- ...... ...... .'.... ..... Net, $4.50 " 
 
 Design of Static Transformers. 101 illustrations. 6x9, cloth, 189 
 pp ....................... .:;. . . .......... ; ..... :... .Net, $2.00 
 
Electricity. A text-book designed in particular for engineering 
 
 students. 115 illustrations. 43 tables. 8vo., cloth, 266 pp. f Net,$2.00 
 
 Electric Trains. 88 illustrations. 8vo., cloth, 220 pp. .. Net, $2 . 50 
 
 Electric Propulsion of Ships. 44 illustrations. 8qo., cloth, 167 pp. 
 
 Net, $2.00 
 
 HOBBS, W. R. P. The Arithmetic of Electrical Measurements. . With 
 numerous examples, fully worked. S'xteenth Edition, revised and 
 edited with six additional chapters by A. R. Palmer. 12mo. cloth, 
 126 pp 50 cents. 
 
 HOUSTOUN, R. A. Studies in Light Production. 22 illustrations. 
 5%x8%, cloth, 120 pp Net, $2. CO 
 
 HUTCHINSON, R. W., Jr. Long-Distance Electric Power Transmission: 
 Being a Treatise on the Hydro-Electric Generation of Energy; Its 
 Transformation, Transmission, and Distribution. Second Edition. 
 Illustrated. 12mo., cloth, 350 pp Net, $3 .00 
 
 - and THOMAS, W. D. Electricity in Mining. Being a theoretical 
 and practical treatise on the construction, operation and main- 
 tenance of electrical mining machinery. 12mo., cloth In Press 
 
 INCANDESCENT ELECTRIC LIGHTING. A Practical Description of 
 the Edison System, by H. Latimer. To which is added: The Design 
 and Operation of Incandescent Stations, by C. J. Field; A Descrip- 
 tion of the Edison Electrolyte Meter, by A. E. Kennelly; and a 
 Paper on the Maximum Efficiency of Incandescent Lamps, by T. W. 
 Howell. Fifth Edition. Illustrated. 16mo., cloth, 140 pp. (No. 
 57 Van Nostrand's Science Series.) 50 cents 
 
 INDUCTION COILS: How Made and How Used. Eleventh Edition. 
 Illustrated. 16mo., cloth, 123 pp. (No. 53 Van Nostrand's Science 
 Series.) 50 cents 
 
 JEHL, FRANCIS. The Manufacture of Carbons for Electric Lighting 
 and other purposes. Illustrated with numerous Diagrams, Tables, 
 and Folding Plates. 8vo., cloth, 232 pp Net, $4.00 
 
 JOHNSON, J. H. Arc Lamps and Accessory Apparatus. 20 illustrations. 
 16mo., cloth, 135 pp. (Installation Manuals Series). .Net, 75 cents. 
 
 JOHNSON, T. M. Ship Wiring and Fitting. 47 illustrations. 16mo., 
 cloth, 92 pp. (Installation Manuals Series) Net, 75 cents. 
 
 JONES, HARRY C. The Electrical Nature of Matter and Radioactivity. 
 Third Edition, revised. 5*x<S*, cloth, 218 pp Net, $2.00 
 
 KAPP, GISBERT. Alternate-Current Machinery. Illustrated. 16mo., 
 cloth, 190 pp. (No. 96 Van Nostrand's Science.) Series. . .50 cents 
 
KAPPER, F. . Overhead Transmission Lines and Distributing Circuits. 
 297 illustrations. 2 plates, 7MxlO, cloth, 310 pp Net, $t.OO 
 
 KENNEDY, R. Electrical Installations of Electric Light, Power, and 
 Traction Machinery. Illustrated. 8vo., cloth, 5 vols. The set $15.00. 
 Each $3 . 50 
 
 KENNELLY, A. E. Theoretical Elements of Electro-Dynamic Machinery. 
 Vol. I. Illustrated. 8vo., cloth, 90 pp $1 . 50 
 
 KERSHAW, J. B. C. The Electric Furnace in Iron and Steel Production. 
 
 Illustrated. 8vo., cloth, 74 pp Net, $1 . 50 
 
 Electrometallurgy. Illustrated. 8vo. s cloth. 303 pp. (Van Nos- 
 
 trand's Westminster Series.) Net, $2 .00 
 
 KINZBRUNNER, C. Continuous-Current Armatures; their Winding and 
 
 Construction. 79 Illustrations. 8vo., cloth, 80 pp Net, $1 .50 
 
 Alternate-Current Windings; their Theory and Construction. 89 Illus- 
 trations. 8vo., cloth, 80 pp Net, $1 . 50 
 
 The Testing of Alternating Current Machines in Laboratories and Test 
 Rooms. A practical work for students and engineers. Vol I. Gen- 
 eral Tests; Transformers, Alternators. 141 illustrations. 5\x8f, 
 cloth, 164 pp Net, $2.00 
 
 KLINGENBERG, G. Large Electric Power Stations. Their design and 
 construction, with examples of existing stations. 180 illustrations, 
 7 plates, 7x 10, cloth, 260 pp Net, $5 .00 
 
 KOESTER, F. Hydroelectric Developments and Engineering. A practi- 
 cal and theoretical treatise on the development, design, construction, 
 equipment and operation of hydroelectric transmission plants. 
 Second Edition. 500 illustrations. 4to., cloth, 475 pp. . .Net, $5.00 
 
 Steam-Electric Power Plants. A practical treatise on the design of 
 
 central light and power stations and their economical construction 
 
 and operation. Second Edition. Fully Illustrated. 4to., cloth 
 
 . 455 pp Net, $5.00 
 
 LANCASTER, M. Electric Cooking, Heating and Cleaning. Edited by 
 W. E. Lancaster, American Edition by S. L. Coles, 305 illustrations, 
 6x8*, cloth 340 pp Net, $1 .00 
 
 LARNER, E. T. The Principles of Alternating Currents for Students of 
 Electrical Engineering. Illustrated with Diagrams. 12mo., cloth, 
 144 pp Net, $1 .25 
 
LEMSTROM, S. Electricity in Agriculture and Horticulture. Illustrated. 
 8vo., cloth Net, $1 .50 
 
 LIVERMORE, V. P., and WILLIAMS, J. How to Become a Competent 
 Motorman : Being a practical treatise on the proper method of oper- 
 ating a street-railway motor-car; also giving details how to over- 
 come certain defects. Second Edition. Illustrated. 16mo., cloth, 
 247 pp Net, $1 .00 
 
 LIVINGSTONE, R. Mechanical Design and Construction of Generators. 
 122 illustrations. 5|x8|, cloth, 228 pp Net, $3 . 50 
 
 Mechanical Design and Construction of Commutators. 62 illustra- 
 tions, 5ix8f, cloth, 93 pp Net, $2.25 
 
 LOCKWOOD, T. D. Electricity, Magnetism, and Electro-Telegraphy. A 
 Practical Guide and Handbook of General Information for Electri- 
 cal Students, Operators, and Inspectors. Fourth Edition. Illus- 
 trated. 8vo., cloth, 374 pp $2.50 
 
 LODGE, OLIVER J. Signalling Across Space Without Wires: Being a 
 description of the work of Hertz and his successors. Third Edition 
 Illustrated. 8vo., cloth Net, $2 .00 
 
 LORING, A. E. A Handbook of the Electro-Magnetic Telegraph 
 Fourth Edition, revised. Illustrated. 16mo., cloth, 116 pp. (No. 
 39 Van Nostrand's Science Series.) 50 cents 
 
 LUCKIESH, M. Color and Its Application. 126 illustrations, 4 color 
 plates. 6x9, cloth, 350 pp $3 .00 
 
 Light and Shade and Their Applications. 135 illustrations, 10 
 
 tables, 6Mx9M, cloth, 278 pp ! Net, $2.50 
 
 MALCOLM, W. H. Theory of the Submarine Telegraph Cable. In Press 
 
 MANSFIELD, A. N. Electromagnets : Their'Design and Construction. 
 
 Second Edition. Illustrated. 16mo., cloth, 155 pp. (No. 64 Van 
 
 Nostrand's Science Series.) 50 cents 
 
 Manufacture of Electric Light Carbons. Illustrated. 5ix8, cloth. 
 
 Net, $1.00 
 MASSIE, W. W., and UNDERBILL, C. R. Wireless Telegraphy and 
 
 Telephony Popularly Explained. With a chapter by Nikola Tesla. 
 
 Illustrated. 12mo., cloth, 82 pp Net, $1 . 00 
 
 MAURICE, W. Electrical Blasting Apparatus and Explosives, with 
 
 special reference to colliery practice. Illustrated. 8vo., cloth. 
 
 167 pp Net, 3.50 
 
 The Shot Firer's Guide. A practical manual on blasting and the 
 
 prevention of blasting accidents. 78 illustrations. 8vo., cloth, 
 
 212 pp Net, $1.50 
 
MIESSNER, B. F. Radiodynamies. The wireless control of torpedoes 
 and other mechanisms. 112 illustrations, 5^x814, cloth, 211 pp. 
 
 Net, $2.00 
 
 MONCKTON, C. C. F. Radio Telegraphy. 173 Illustrations. 8vo 
 cloth, 272 pp. (Van Nostrand's Westminster Series.) Net, $2 .00 
 
 MONTGOMERY, J. W. Electric Wiring Specifications 4x6^, cloth, 
 107 pp Net, $1 50 
 
 MORECROFT, J. H. and HEHRE, F. W. A Short Course in Testing of 
 Electrical Machinery. Third Edition. Illustrated. 8vo., cloth, 
 160 pp ' Net, 1 .50 
 
 MORGAN, ALFRED P. Wireless Telegraph Construction for Amateurs, 
 Third Edition. 167 illustrations. 5ix7*, cloth, 236 pp.. .Net, $1 .50 
 
 NERZ, F. Searchlights, Their Theory, Construction and Application. 
 Translated by C. Rodgers. 47 illustrations. 6x8,, cloth, 145 pp. 
 
 Net, $3.00 
 
 NIPHER, FRANCIS E. Theory of Magnetic Measurements. With an 
 Appendix on the Method of Least Squares. Illustrated. 12mo., 
 cloth, 94 pp $1 .00 
 
 OHM, G. S. The Galvanic Circuit Investigated Mathematically. Berlin, 
 1827. Translated by William Francis. With Preface and Notes 
 by Thos. D. Lockwood. Second Edition. Illustrated. 16mo., cloth, 
 269 pp. (No. 102 Van Nostrand's Science Series.) 50 cents 
 
 OLSSON, ANDREW. Motor Control as used in Connection with Turret 
 Turning and Gun Elevating. (The Ward Leonard System.) 13 
 illustrations. 12mo., paper, 27 pp. (U. S. Navy Electrical Series 
 No. 1.) Net, .50 
 
 OUDIN, MAURICE A. Standard Polyphase Apparatus and Systems. 
 Sixth Edition, revised. Illustrated with many Photo-reproductions, 
 Diagrams, and Tables. 8vo., cloth, 369 pp Net, $3 .00 
 
 PALAZ, A. Treatise on Industrial Photometry. Specially applied to 
 Electric Lighting. Translated from the French by G. W. Patterson, 
 Jr., and M. R. Patterson. Second Edition. Fully Illustrated. 
 8vo., oloth, 324 pp .$4.00 
 
 PARR, G. D. A. Electrical Engineering Measuring Instruments for Com- 
 mercial and Laboratory Purposes. With 370 Diagrams and Engrav- 
 ings. 8vo., cloth, 328 pp... , Net, $3.50 
 
PARSHALL, H. F., and HOBART, H. M. Armature Windings of Electric 
 Machines. Third Edition. With 140 full-page Plates, 65 Tables, 
 and 165 pages of descriptive letter-press. 4to., cloth, 300 pp. .$7.50 
 
 Electric Railway Engineering. With 437 Figures and Diagrams 
 
 and many Tables. 4to., cloth, 475 pp Ne-t, $7.50 
 
 Electric Machine Design. Being a revised and enlarged edition of 
 
 "Electric Generators." 648 Illustrations. 4to., half morocco, 601 
 pp Net, $12.50 
 
 PATCHELL W. H. Application of Electric Power to Mines and heavy 
 
 Industries. 91 illustrations 61^x914, cloth, 203 pp Net, 4.00 
 
 
 
 PATTEESON, G. W. L. Wiring Calculations for Light and Power In- 
 stallations. 139 illustrations, 5^x7%, cloth, 203 pp Net, $2.00 
 
 Electdc Mine Signalling and Installations. Illustrated. 5Mx7K, 
 
 110 pp Net, $1 . 50 
 
 PERRINE, F. A. C. Conductors for Electrical Distribution : Their Manu- 
 facture and Materials, the Calculation of Circuits, Pole Line Con- 
 struction, Underground Working, and other Uses. Second Edition. 
 Illustrated. 8vo., cloth, 287 pp Net, $3 .50 
 
 POPE, F. L. Modern Practice of the Electric Telegraph. A Handbook 
 for Electricians and Operators. Seventeenth Edition. Illustrated. 
 8vo., cloth, 234 pp $1 .50 
 
 RAPHAEL, F. C. Localization of Faults in Electric Light Mains. Third 
 Edition In Press 
 
 RASCH, E. Electric Arc Phenomena. Translated by K. Tornberg. 
 52 illustrations. 5x7, cloth, 187pp Net, $2 . 00 
 
 RAYMOND, E. B. Alternating-Current Engineering, Practically Treated. 
 Third Edition, revised. With many Figures and Diagrams. 8vo., 
 cloth, 244 pp Net, $2 .50 
 
 REDFERN, T. B. and SAVIN,- J. Bells, Indicators, Telephones, Fire and 
 Burglar Alarms. 85 illustrations. 4ix6|, cloth, 123 pp. (Installa- 
 tion Manuals Series) 0.50 
 
 RICHA1DSON, S. S. Magnetism and Electricity and the Principles of Elec- 
 trical Measurement. Illustrated. 12mo., cloth, 596 pp-. .Net, $2.00 
 
 ROBERTS, J. Laboratory Work in Electrical Engineering Preliminary 
 Grade. A series of laboratory experiments for first and second- 
 year students in electrical engineering. Illustrated with many 
 
 v Diagrams. 8vo., cloth, 218 pp Net, $2.00 
 
ROLLINS, W. Notes on X-Light. Printed on deckle edge Japan paper 
 400 pp. of text, 152 full-page plates. 8vo., cloth Net, $5.00 
 
 RUHMER, ERNST. Wireless Telephony in Theory and Practice. Trans- 
 lated from the German by James Erskine-Murray. Illustrated. 
 8vo., cloth, 224 pp Net, $3.50 
 
 RUSSELL, A. The Theory of Electric Cables and Networks. 71 Illus- 
 trations. 8vo., cloth, 275 pp Net, $3 . 00 
 
 SAYERS, H. M. Brakes for Tramway Cars. 6x9, cloth, 76 pp. Net, 1 .25 
 
 SEVER, G. F: Electrical Engineering Experiments and Tests on Direct- 
 Current Machinery. Second Edition, enlarged. With Diagrams and 
 Figures. 8vo., pamphlet, 75 pp Net, $1 . 00 
 
 SEVER, G. F., and TOWNSEND, F. Laboratory and Factory Tests in 
 Electrical Engineering. Second Edition, revised and enlarged. Illus- 
 trated. 8vo., cloth, 269 pp Net, $2.50 
 
 SEWALL, C. H. Wireless Telegraphy. With Diagrams and Figures. 
 
 Second Edition, corrected. Illustrated. 8vo., cloth, 229 pp. . Net, $2 . 00 
 
 Lessons in Telegraphy. Illustrated. 12mo., cloth, 104 pp .Net, $1 .00 
 
 SEWELL, T The Construction of Dynamos (Alternating and Direct 
 Current). A Text-book for students, engineering contractors, and 
 electricians-in-charge. Illustrated. 8vo., cloth, 316 pp $3.00 
 
 SHELDON, S., and HAUSMANN, E Dynamo-Electric Machinery: Its 
 
 Construction, Design, and Operation. 
 Vol. I.: Direct-Current Machines. Ninth Edition, completely re-written. 
 
 Illustrated. 12mo., cloth, 281 pp Net, $2.50 
 
 Vol. II.: Alternating-Current Machines : Tenth Edition, rewritten. 
 
 cloth, 353 pp Net, $2.50 
 
 Electric Traction and Transmission Engineering 127 illustration. 
 
 12mo., cloth, 317 pp Net, $2.50 
 
 SLOANE, T. O'CONOR. Elementary Electrical Calculations. A Manual 
 of Simple Engineering Mathematics, covering the whole field of 
 Direct Current Calculations, the basis of 'Alternating Current 
 Mathematics, Networks, and typical cases of Circuits, with Appen- 
 dices on special subjects. 8vo., cloth. Illustrated. 304 pp 
 
 Net, $2.00 
 
SMITH, C. F. Practical Alternating Currents, and Alternating Current 
 Testing. Third Edition. 236 illustrations. 5|x8|, cloth, 476 pp. 
 
 Net. $2.60 
 
 SMITH, C. F. Practical Testing of Dynamos and Motors. Third Edition, 
 108 illustrations. SJxSf, cloth, 322 pp Net, $2.00 
 
 SNELL, ALBION T. Electric Motive Power. The Transmission and Dis- 
 tribution of Electric Power by Continuous and Alternating Currents. 
 With a Section on the Applications of Electricity to Mining Work. 
 Second Edition. Illustrated. 8vo., cloth, 411 pp Net, $4.00 
 
 SODDY, F. Radio-Activity; an Elementary Treatise from the Stand- 
 point of the Disintegration Theory. Fully Illustrated. 8vo., cloth, 
 214 pp Net, $3.00 
 
 SOLOMON, MAURICE. Electric Lamps. Illustrated. 8 vo., cloth. (Van 
 Nostrand's Westminster Series.) Net, $2 .00 
 
 SWINBURNE, JAS., and WORDINGHAM, C. H. The Measurement of 
 Electric Currents. Electrical Measuring Instruments. Meters for 
 Electrical Energy. Edited, with Preface, by T. Commerford Martin. 
 Folding Plate and Numerous Illustrations. 16mo., cloth, 241 pp. 
 (No. 109 Van Nostrand's Science Series.) 50 cents 
 
 SWOOPE, C. WALTON. Lessons in Practical Electricity: Principle 
 Experiments, and Arithmetical Problems. An Elementary Text- 
 book. With numerous Tables, Formulae, and two large Instruction 
 Plates. Fifteenth Edition, revised. Illustrated. 8vo., cloth, 462 pp 
 
 Net, $2.00 
 
 THIESS, J. B. and JOY, G. A. Toll Telephone Practice. 273 illustrations. 
 8vo., cloth, 433 pp Net, $3.50 
 
 THOM, C., and JONES, W. H. Telegraphic Connections, embracing recent 
 methods in Quadruple* Telegraphy. 20 Colored Plates. 8vo., 
 cloth, 59 pp $1.50 
 
 THOMPSON, S. P., Prof. Dynamo-Electric Machinery. With an In- 
 troduction and Notes by Frank L. Pope and H. R. Butler. Illus- 
 trated 16mo., cloth, 214 pp. (No. 66 Van Nostrand's Science 
 Series.) 50 cents 
 
 Recent Progress in Dynamo-Electric Machines. Being a Supple- 
 ment to " Dynamo-Electric Machinery." Illustrated. 16mo., cloth, 
 113 pp. (No. 75 Van Nostrand's Science Series.) 50 cents 
 
 TOWNSEND, FITZHUGH. Alternating Current Engineering. Illustrated. 
 8vo , paper, 32 pp Net, 75 cents 
 
UNDERBILL, C. R. Solenoids, Electromagnets and Electromagnetic 
 Windings. Second Edition. 218 Illustrations. 12mo., cloth, 345 
 pp Net, $2 . 00 
 
 URQUHART, J. W. Electroplating. Fith Edition. Illustrated. 12mo., 
 cloth, 230 pp $2.00 
 
 Electrotyping. Illustrated. 1 2mo. , cloth, 228 pp $2 . 00 
 
 VOSMAER, A. Ozone. Its Manufacture and Uses. 76 illustrations. 
 6x9, cloth, 210 pp Net, $2 . 50 
 
 WADE, E. J. Secondary Batteries: Their Theory, Construction, and Use. 
 Second Edition^ corrected 265 illustrations. 8vo., cloth, 501 pp. 
 
 Net, $4.00 
 
 WADSWORTH, C. Primary Battery Ignition. A simple practical pocket 
 guide on the construction, operation, maintenance, and testing of 
 primary batteries for automobile, motorboat, and stationary en- 
 gine ignition service. 26 illustrations. 5x7, cloth, 79 pp Net, .0.50 
 
 WALKER, FREDERICK. Practical Dynamo Building for Amateurs. 
 How to Wind for any Output. Third Edition. Illustrated. 16mo., 
 cloth, 104 pp. (No. 68 Van Nostrand's Science Series.). .50 cents. 
 
 Electricity in Mining. Illustrated. 8vo., cloth, 35 pp $.50 
 
 WATT, ALEXANDER. Electroplating and Refining of Metals New 
 Edition, rewritten by Arnold Philip. Illustrated 8vo., cloth, 704 
 pp Net, $4.50 
 
 Electro -metallurgy Fifteenth Edition. Illustrated 12mo., cloth, 
 
 225 pp S1.00 
 
 WEBB, H. L. A Practical Guide to the Testing of Insulated Wires and 
 Cables Sixth Edition. Illustrated. 12mo., cloth., 118 pp .. .$1 .00 
 
 WILKINSON, H. D. Submarine Cable-Laying, Repairing, and Testing. 
 Second Edition, completely revised. 313 illustrations. 8vo., cloth, 
 580 pp Net, $ 3 .00 
 
 WILSON, J. F. Essentials of Electrical Engineering. 300 illustrations. 
 6x9, cloth, 355 pp Net, $2.50 
 
 WRIGHT, J. Testing, Fault Localization and General Hints for Linemen. 
 19 Illustrations. 16mo., cloth, 88 pp. (Installation Serious Manuals.) 
 
 Net, 50 cents. 
 
YOUNG, J. ELTON. Electrical Testing for Telegraph Engineers. Illus- 
 trated. 8vo., cloth, 264 pp Net, $4 .00 
 
 ZEIDLER, J., and LU3TGARTEN, J. Electric Arc Lamps: Their Princi- 
 ples, Construction and Working. 160 Illustrations. 8vo., cloth, 
 188 pp Net, $2.00 
 
 A 96=page Catalog of Books on Electricity, classified by 
 subjects, will be furnished gratis, postage prepaid, 
 on application. 
 
 3,500 M-l-17 
 
D. VAN NOSTRAND COMPANY 
 
 are prepared to supply, either from 
 
 their complete stock or at 
 
 short notice, 
 
 Any Technical or 
 
 Scientific Book 
 
 In addition to publishing a very large 
 and varied number of SCIENTIFIC AND 
 ENGINEERING BOOKS, D. Van Nostrand 
 Company have on hand the largest 
 assortment in the United States of such 
 books issued by American and foreign 
 publishers. 
 
 All inquiries are cheerfully and care- 
 fully answered and complete catalogs 
 sent free on request. 
 
 25 PARK PLACE NEW YORK 
 
14 DAY USE 
 
 RETURN TO DESK FROM WHICH BORROWED 
 
 LOAN DEPT. 
 
 This book is due on the last date stamped below, or 
 on the date to which renewed. 
 Renewed books are subject to immediate recall. 
 
 
 
 REC'D LD 
 
 
 r ..29 19S7 
 
 
 
 
 &$* 
 
 
 
 
 ^ i O 
 
 
 F^C'13 
 
 
 ^m 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 LD 21-100m-6,'56 
 (B9311slO)476 
 
 General Library 
 
 University of California 
 
 Berkeley 
 
389478 
 
 UNIVERSITY OF CALIFORNIA LIBRARY