GIFT OF ELECTRIC LIGHTING. ELECTRIC LIGHTING TRANSLATED FROM THE FRENCH OF LE COMTE TH. DU MONCEL BY ROBERT ROUTLEDGE, B.Sc. (LoND.), F.C.S. Author of "DISCOVERIES AND INVENTIONS OF THE NINETEENTH CENTURY," "A POPULAR HISTORY OF SCIENCE," ^T. •'} , - i"M£ WITH SEFENTY-SIX ILLUSTRATIONS SECOND EDITION LONDON GEORGE ROUTLEDGE AND SONS BROADWAY, LUDGATE HILL NEW YORK: 9 LAFAYETTE PLACE 1883 BY THE SAME AUTHOR. In Crown Si'o, cloth, 10s. 6d. DISCOVERIES AND INVENTIONS OF THE NINETEENTH CENTURY. BY ROBERT ROUTLEDGE, B.SC. (LOND.), F.C.S. With 400 Illustrations, Portraits, &c. TRANSLATOR'S PREFACE. THE book in the reader's hand is a translation of the second edition of the Comte du Moncel's "E Eclair- age Electrique" published at Paris, by the Messrs. Hachette, in 1880. Though the original work is one of a series in- tended for popular use, it leaves nothing to be desired on the score of scientific treatment. An able and thorough handling of the subject that occupies the present volume is, indeed, precisely what would be expected by those who are acquainted with the very eminent scientific position of the Comte du Moncel as an investigator and as an author. While the general reader will find in the following pages all that he can possibly require to know about electric lighting, and find it, too, laid before him in full and lucid explana- tions, divested of all perplexing technicalities, it is believed that the book may prove of more extended and important utility, by supplying serviceable guidance to the daily in- creasing number of persons who are called upon to consider how far this latest gift of science can be applied to their purposes. Such are more especially engineers, architects, managers of industrial and commercial establishments of all kinds, municipal officers, members of local boards, directors of railway and steamship companies, lessees of theatres, cura- tors of museums, and of picture galleries, etc., besides other classes of persons too numerous to mention. A remarkable feature of the present work — because it is a merit by no means common, either in scientific or in popular books— is, that 461554 VI TRANSLATOR'S PREFACE. from the first page to the last, the practical point of view is never for an instant abandoned. No special knowledge of electrical science is needed for the intelligent perusal of this work ; nevertheless,! a reader altogether ne\v to the subject would do well to acquire clear ideas of a few elementary facts. Almost any of the cheap manuals which now abound would suffice to convey all the information required; but it could not, perhaps, -be attained more pleasantly and profitably than by a reference to Professor TyndalFs " Notes of a Course of Seven Lectures on Electrical Phenomena and Theories," which may be had for a few pence. The present translation contains all the author's notes, references, and appendices. To these, two other appendices are now added by the translator. The first gives the English equivalents of the French weights, measures, etc., which, on account of their now almost universal employment for scientific purposes, it has been deemed expedient to retain in the text. The second is a brief notice of some forms of incandescent lamps, and of another recent invention, which have come into prominence since the publication of the origi- nal work. Several lamps of this kind may now be said to have proved completely successful in practice, and the forms described in the appendix are those which have attracted the largest share of public attention, as offering the simplest and most effective solution of nearly all the difficulties that have hitherto attended the application of electric lighting to domestic and certain other purposes. The translator, having been unable to refer to some of the English and American newspapers, journals, and reports quoted by the author, has usually contented himself with a faithful rendering of the French version; but it is not, of course, to be expected that the ipsissima verba of the TRANSLA TOR'S PREFACE. Vli originals will appear by this process of re-translation. The inverted commas, used to mark passages cited verbatim, have generally been suppressed in such cases. With but two or three unimportant exceptions, the illus- trations in this volume are identical with, or equivalent to, those of the original work. The translator and the pub- lishers desire gratefully to acknowledge the kindness of Messrs. Blackie, in allowing the use of several cuts from their excellent edition of Deschaners Natural Philosophy, namely, those which appear in this volume as Figs, i, 2, 4, 5, 10, n, 12, 13, 15, 1 8, 38, and 72. A like acknowledgment is also due to M. Werdermann for his permission to use the illustra- tions forming Figs. 37 and 57. R. R. CONTENTS, PART I.-GENERAL CONSIDERATIONS. HISTORICAL SKETCH IMPORTANT DEFINITIONS ......... 4 Electric current ............... ^ Electro-motive force ............... - Tension of a current ...... ... ...... 5 Electrical potential .., ............ 6 Electrical intensity ......... ... 6 Electrical resistance ... ... ... ... 7 Electrical conductivity ...... 8 Electrodes ... ......"... g Polarization ......... ......... g Electric Units ......... ...... ... ]0 WHAT THE ELECTRIC LIGHT IS ... ... '^/.^ "s ,.. 12 THE VOLTAIC ARC PART II.-GENERATORS OF ELECTRIC LIGHT. VOLTAIC GENERATORS ... ..,......, ... 2$ Groves and Sunset? s batteries '.,•, .,v ... 25 CONTENTS. Voltaic constants Grouping of elements for tension or for quantity ... 30 Conditions for maximum current ... ... ... 32 Derived circuits 32 Mode of Propagation of Electricity 35 THERMO-ELECTRIC GENERATORS 3cS History ... ... ... , 38 Clamontfs piles ... ... 39 MAGNETO-ELECTRIC GENERATORS • ..>. .. 46 Induced current ... .., 46 History of the Question 46 THE DIFFERENT MODES OF GENERATING INDUCED CUR- RENTS, AND THE LAWS WHICH GOVERN THEM... 52 Laws of Induced Currents ... 54 VARIOUS MAGNETO-ELECTRIC GENERATORS OF ELECTRIC LIGHT 6l The Alliance machine ••• 62 Wilde's Machine ... .... -* -^ — 68 Ladd's machine ... ....... 71 Gramme's Machine <• ... 74 Siemens' Machine ... ... ••• ;.sv*^'.-. ••• 82 Be Meritens' machine ... ••• . . 85 Wallace Farmer's Machine 90 Brush's Machine ... ... ... 9- It ii ruin's Machine 95 Trouve's Machine ... 97 Loiitin's Machine 98 ;J.l l'['iT!7'.-TJ^-'! • ; . • -4 /• •*•!• Ji "> • •• ' MACHINES WITH ALTERNATE REVERSIONS OF THE CUR- RENTS, ESPECIALLY APPLICABLE TO THE DIVI- SION OF THE ELECTRIC LIGHT . .X, ICX> CONTENTS. xi PAGE Lout ill's System . . i oo Gramme's System 104 Siemens' System 107 Jablochkoff 's System ...no COMPARATIVE EXPERIMENTS ON THE EFFECTS PRO- DUCED BY THE DIFFERENT ELECTRO-MAGNETIC MACHINES • Ill EFFECTS OF THE RESISTANCE OF EXTERNAL CIRCUITS... I2O RESULTS PRODUCED BY COUPLING TWO MACHINES ... 122 TABLE OF THE AMERICAN EXPERIMENTS... 123 COMPARATIVE TABLE OF THE ENGLISH EXPERIMENTS 124 — 5 TABLE OF NEW EXPERIMENTS (SHOOLBRED) ... 126 — / APPARATUS FOR EVOLVING THE ELECTRIC LIGHT ... 128 Light produced between carbon electrodes... 128 "Jacquelairis Method ... 131 Carre's Method f,.. 132 Gauduiffs System ... 1 34 Effects produced by the addition of Metallic Salts to the Carbons prepared for the Electric Light ... /.?-*'.-' 135 Metallized Carbons ... ... ... ... 137 Effect of heat on the conductivity of the carbons ,.. 139 Light produced by means of conductors of indifferent conductivity 140 Jnblochkoff's System ... ... \ ,'&•> ••• ••• J40 Lodyguine and Kosloff's System < .... 144 Edison's System .... ... ... 145 Systems nf E. Reymer, Wtrdermajm, and others ... 146 Light produced by means of an inductive action ... .., * v.i r ••• !5° xn CONTENTS. PART III.— ELECTRIC LAMPS. PAGE VOLTAIC ARC LAMPS 152 Foil ran It and Duboscq's Lamps 154 Serrin's Lamp ... ... 161 Siemens' Lamp .".." ... ... 165 Lout hfs Lamp ... "" ... 169 De Mersanne's Lamp 170 Biirgin's Lamp ... 172 fiaiffe's Lamp '. ... ... ... 174 Carre's Lamp 177 Brush's Lamp 178 Jaspar's Lamp 180 RapiefTs Lamp ... ... 1-81 Baro's Lamp ... ,'.. 185 Wallace Farmer's Lamp ... 186 Houston and Thompson's Lamp 186 Molera and Cebrian's Lamp 187 Lamps of various kinds ... ... 188 Regulators with hydrostatic actions ... 19° INCANDESCENT LAMPS ... 195 King's and Lodyguine's Lamps. ]95 Komi's Lamp • '96 It o ii I i mime's Lamp ... •- • • J9tS Sawyer-Man's Lamp ... , ... J99 E. Reynier's Lamp ... ... 199 Werdermann's Lamp . ... 205 Trouve's form of Reynier's Lamp • • 209 Ducretet's form of Reynier's Lamp t.; .211 Tommasi's Lamp ... ••• -12 CONTENTS. xiii PAGE Edison's Lamp , 212 Automatic lighter of Reynier's Electric Lamps 216 ELECTRIC CANDLES 217 Condensers 227 Other Systems of Electric Candles 229 jamirfs Candle 231 Soligniacs Candle ... ... ... ... .. 236 La^ suits. Of these various machines, it is evidently those which call a motor into play that are applicable for electric lighting,. for it is in this form that electricity can be obtained under the most economical conditions. These machines will there- fore exclusively occupy our attention ; but before this we think some indications should be given of the laws regulating induced currents, and of the different causes which concur for their production. THE DIFFERENT MODES OF GENERATING INDUCED CURRENTS, AND THE LAWS WHICH GOVERN THEM. Besides the induction effect which we have already ex- plained, there are many other causes capable of developing induced currents. Every action, the effect of which is to diminish or to increase the power of a magnet already acting on an induction coil may give rise to induced currents which will be direct when there is diminution, and inverse whea there is increase. This increase may result from the action on the magnetic poles of an armature of soft iron, and dimi- nution will result from the withdrawal of that armature. Again, if the pole of a permanent magnet be passed before an iron core surrounded by a coil, a double action will be produced : i°, a current which may be termed an electro- dynamic induction current, which will result from the successive passage of the spires of the induced coil before the pole of the inducing magnet, and which will be the more energeti- cally developed the more suitable precautions are taken to avoid the hurtful induction to which the parts of the spires behind those directly affected by the magnet are liable ; 2% DIFFERENT MODES OF GENERATING CURRENTS. 53 a current to which I have given the name of polar inversion current, and which results from the inversions of magnetic polarities to which the core is successively subject, by reason of the movement of the inducer.* These two currents * In order that the origin of the currents which now play so important a part in the new electric generators may be thoroughly understood, we shall •examine what takes place when a bar of soft iron surrounded by a magne- tizing coil is brought near one of the poles of a permanent magnet, for instance, near the north pole. At the instant of approach there will be pro. duced a first current, a current of magnetization, whose direction will depend upon which end of the bar is acted upon by the magnetic pole. This current is due to the transformation of the bar into a magnet. If the magnet is with- drawn, a current will again be induced in the direction opposite to the former, and which will correspond with the demagnetization of the bar. But if the pole of the magnet be brought near the middle of the bar, there will be no current produced, because the bar itself will then form a magnet with a consequent point, and the effect of the induction on the one side of this point will be counteracted by that on the other side. The same effects will ensue if the coil is deprived of the iron core. Now, it follows from this principle that, if the iron bar surrounded by its coil be bent so as to form a ri ig, neither currents of magetization nor of demagnetization can be obtained by the approach or recession of the magnet, to what point soever it may be directed, for the parts to the right and to the left of the point affected are then polarized in the same manner. Nevertheless, if the magnet he moved parallel to the axis of the bar, that is to say, circularly about the ring, it will no longer be the same thing, and a current may be produced, due neither to magnetization nor to demagnetization, but which will, under certain con- ditions, last during the whole time the magnet is revolving in the same direc- tion. This current may be the result of two different and simultaneous actions, but in order that it may be manifested it will be necessary for the coil to have a certain arrangement, for even if the induction be produced by a single magnetic pole, the two opposite parts of the ring affected would always be polarized in a different direction, and would give rise to contrary currents. One of these actions is the result of the magnetic disturbance produced in the core itself by the successive inversion of its polarities— a disturbance which must, as I have proved, give rise to a reaction analogous to that observed when an effect of demagnetization is made to follow an effect of magnetization under opposite conditions ; and as under such con- ditions the resulting currents have the same direction, and in consequence of the progressive motion of the inducer they follow each other without in- terruption, a continuous current is the upshot, and this changes in direction only when the direction of the movement of the inducing pole is changed. The other action results from the motion itself of the magnetic inducing system before the spires of the induced coil, or what amounts to the same thing, f; om the motion of these coils before the inducing system. (Sice Note A. ) 54 ELECTRIC LIGHTING. which are continuous, are manifested during the whole period of the magnet's movement, and their direction depends upon that of this movement, but it always corresponds with a de- magnetization current, that is to say, with a current in the same direction as that of the magnetic current of the core affected.* It is these currents which act in the Gramme machines. The other reactions which were first discussed have given birth to magneto-electric machines, of which the best-known forms are those of Dujardin, Breton, Duchenne, Wheatstone, Breguet, &c. ; but as these machines have not produced effects sufficiently powerful to form generators of electric light, we shall say no more about them. There are yet other sources of induction which have pro- mised to give origin to magneto-electric machines, and of their number is that which has given rise to the peripolar in- duction machine of Le Roux; but these sources are still too feeble to be applied advantageously. Laws of Induced Currents., — Many experiments made on induced currents have proved : — i°. That the quantity of electricity put in action in a circuit is proportional, other things being equal, to the intensity of the inducing current and to the length of the induced circuit. 2°. That it is independent of the duration of the inducing action, and varies only with the magnitude of the initial cause of induction. 3°. That the tension of the induced current varies on the contrary with the duration of the inducing action, and in- creases with the rapidity of the variation of the inducing cause; a circumstance which goes to show that the tension of an induced current is proportional to the algebraic deriva- tive of the function of the time expressing the law of succession of the values of the intensity in the induced current. * For the explanation of these effects see Note A at the end of the book. DIFFERENT MODES OF GENERATING CURRENTS. $5 4°. That the direct currents have a shorter duration than the inverse currents. 5°. That it follows from this last property that the inverse and direct currents, though formed by equal quantities of electricity, are able to act differently ; for the direct currents having a shorter duration than the inverse currents, have a greater tension, and may therefore act on a circuit of greater resistance. 6°. That the duration of the direct current is independent of the resistance of the induced circuit, whilst that of the inverse current increases with that resistance and with the number of spires in the coil; whence it follows that in ordinary cases the electro-motive force of the direct current must be greater than that of the inverse current, and that the ratio of these forces increases with the length of the induced wire. Induced currents may, like battery currents, be changed into tension currents or into quantity currents, not only ac- cording to the mode in which the induced coils are con- nected, but also according to the insulation of the wire, its thickness, its length, and the composition of the magnetic core causing the induction. With a wire very rine, very long and insulated with all the precautions adopted for the elec- tricity of the glass-plate machines, sparks of more than a metre in length have been obtained ; and with a magneto- electric machine having a wire of large diameter a current may be obtained possessing sufficient quantity to produce the effects of the voltaic battery. The laws of induced currents with regard to the effects they produce through the external circuit are the same as those of voltaic currents, but it must be admitted that in this case the resistance of the generator is represented by a quantity much greater than that which can be deduced from its direct measurement if voltaic currents were employed. Thus the calorific work supplied by an induction machine is expressed by Joule's formula, in which the value of the resistance of 56 ELECTRIC LIGHTING. the induced wire would be expressed by a quantity variable no doubt with the machines, but which may, with the Alliance machines, be six times greater than its true value. It follows then, that, taking into account this increase ot resistance, this work is proportional to the square of the electro-motive force developed, and to the resistance of the external circuit, and is inversely proportional to the square of the total re- sistance of the circuit, and this leads to the conclusion that the maximum effect is obtained when the resistance of the external circuit exceeds that of the induced circuit by the quantity with which this last must be increased in order that Ohm's formula may be applicable to these machines. We shall not speak of the laws relating to induced currents of different orders, for we shall have to concern ourselves but little about these effects in the machines used for electric lighting. But we must carefully study the influence on the induced currents of the shape, the dimensions, and the com- position of the magnetized core, as well as that which results from the speed ot the alternate magnetizations and demag- netizations. If it is considered that the power of electro-magnets is proportional to the diameters of the magnetic cores, and to the square root of their length, it may be supposed that it would be of advantage to make the iron cores of induction apparatus the thickest and the longest possible. But as the alternations of magnetization and demagnetization are pro- duced much more slowly with large iron cores than with small ones, and as again a cylindrical metallic surface allows of the formation of local induced currents which are developed to the prejudice of the induced currents themselves, some middle course had to be sought for, and recourse was had to bundles of iron wires or of juxtaposed thin plates of sheet- iron, the magnetic adherence of which is not sufficiently perfect to be equivalent to a continuous mass of that metal. This means has yielded excellent results, as we shall farther on have occasion to observe. DIFFERENT MODES OF GENERATING CURRENTS. 57 Experiments on the length of the magnetic cores have not yet been sufficiently multiplied and sufficiently conclusive to enable a very definite law to be formulated with regard to them. Poggendorff, Muller, and several otner physicists have, however, observed in a simple coil with a straight core that the inducing action is stronger at the middle of the core than in any other point, and therefore they have advised that the greatest possible number of spires should be accumulated on this part of the coil, which implies giving to the coils the form of a spindle. This result will be understood if it be considered that the middle of a coil corresponds with the re- sultant of all the dynamical effects of the individual currents of the magnetic core. As to the length of the coils them- selves, it seems that there is an advantage in making them FIG. 13. rather long to obtain tension, and rather short to obtain quantity. Siemens, however, has contrived a form of induc- tion coil entirely different from the ordinary forms, which has given excellent effects. It is represented in Fig. 13. The magnetic core is made of a cylinder of iron, in which a wide groove is formed longitudinally surrounding the cylinder, and in this the wire of the coil is wound parallel to the axis cf the cylinder, and is bound with bands which prevent it from yielding to the influence of centrifugal force when the bobbin rotates. The parts of the iron cylinder left uncovered form the polar ends of the coil. The bobbin of course revolves on the axis of the cylinder, and in a semi-cylindrical cavity formed in the iron armature fitted to the two poles of the inducing magnet, as shown in Fig. 14. The two ends of the wire of the coil terminate in a commutator shown at the end 58 ELECTRIC LIGHTING. of the axis of the bobbin, on which press two springs con- nected with the circuit. When it is wished to use cylindrical cores of iron there is. an advantage, in respect to induction, to form them of iron tubes, slit longitudinally, and to put two or three of them one within the other. In the large machines of the Alliance Company this plan has been made use of. FIG. As to the more or less advantageous effects of quicker or slower alternations of magnetization and demagnetization, the question is complicated on account of the magnetic inertia of iron. According to the proportionality of the tension of the currents to the derivative of the function of the time, one would suppose that these alternations ought to be as rapid as possible ; but the tardiness with which iron is magnetized and demagnetized much complicates the effect produced, and it is observed that whilst slow interruptions of DIFFERENT MODES OF GENERATING CURRENTS. 59 the inducer are favourable to the development of the tension of the induced current in RuhmkorfFs coils, a high speed of rotation, and, therefore, very rapid alternations of magnetiza- tion and demagnetization, are required to produce the maxi- mum effect with magneto-electric machines. It may, there- fore, be said that in general a rapid succession of magnetiza- tions and demagnetizations increases the tension of currents, and allows the magnetic cores to acquire their maximum of magnetization. It will be understood by this that the number of successive inducing effects which a machine will furnish, to be under the best conditions, will depend upon its construction, and that the more readily its magnetic organs are able to undergo magnetization and demagnetiza- tion, the greater should be its speed. There are also certain conditions in the relative arrange- ment of the induction organs which are more or less favour- able to a high speed ; for example, in an ordinary magneto- electric machine formed of several induction coils the mean intensity of the sum of all the transmitted currents increases with the velocity of rotation, but in a less ratio than the in- crease of the velocity. This increase depends on two cir- cumstances, that is to say, on the intensity of the current itself, and on the greater or less number of coils connected for tension. The greater the number of coils arranged for tension the less rapidly does the electro-motive force in- crease, and this slowness of increase is the greater as the intensity of the current is greater. " It follows from all this," says Le Roux, " that the increments of intensity cost more and more when it is sought to obtain them by increase of velocity ; for, if theoretically each turn con- sumes a quantity of work proportional only to the useful effect it will finally produce, practically such turn causes the loss of a certain quantity of work expended on passive resistance of every kind. Nevertheless, in the calorific applications of electricity, it is of importance to obtain these high intensities, since the useful effect is proportional to their square in a given time ; but in the 60 ELECTRIC LIGHTING. chemical applications, for which the effect is proportional to the first power of the intensity, it is advantageous for the economy of motive force not to drive the machine at high speed." If the increase of speed in magneto-electric machines augments the intensity of the resulting induced currents, on the other hand the increase in the intensity of the currents opposes the rotatory movement. The more frequent sepa- rations of parts attracting each other must give the motive power a greater sum of mechanical resistances to overcome ; but there is also in the very circumstance of the development of electrical work effected a mechanical reaction, which is the consequence of the transformation of the physical forces. Everybody knows the beautiful experiment of Foucault's, which consists in making a copper disc revolve between the poles of a powerful electro-magnet. While the magnet is not in action the disc may be turned with any required velocity, but as soon as the electro-magnet comes into action the disc becomes more and more heated, the resistance offered to its rotation increases considerably, and soon becomes so great that the speed of the machine cannot be further increased. The calorific effect produced has there- fore necessitated an expenditure of force by giving rise to an increase of resistance, and the measure of that additional mechanical resistance must be the equivalent of the physical action which has caused it. Now an action of this kind must evidenly be produced in magneto-electric machines in rapid rotation, and especially in dynamo-electric machines. There is yet an interesting question to study, namely, to ascertain the duration of induced currents, the period which elapses between the closing or opening of an inducing cir- cuit and the appearance of the induced current, and how the intensity of the current behaves at the different phases of its appearance. Blaserna and Mouton have made some very interesting researches on this subject, of which we shall indicate the principal results. According to Blaserna the time which elapses between the VARIOUS GENERATORS OF ELECTRIC LIGHT. 6 1 inducing action and the appearance of the current is less than the fifty thousandth part of a second, and the current, feeble at the commencement, gradually increases, then diminishes, and ceases in a space of time varying with the intensity of the induced current, but which is on the average the two hundredth of a second. Mouton has shown that this dimi- nution of the induced current takes place by oscillations suc- cessively decreasing. To conclude, we shall add that if several induction effects can arise under the influence of the same inducer, the total resulting current cannot have a greater intensity than that which would result from a single one of these effects, if the action causing this last were acting under its maximum condition ; for the inducing force is divided like the attractive force. Hence certain machines in which several kinds of inductions are combined do not yield more than others in which only one kind of induction is in action. VARIOUS MAGNETO-ELECTRIC GENERATORS OF ELECTRIC LIGHT. In tomes IT. and V. of my Expose des applications de- relectridtc I have given a nearly complete history of the various induction generators that have been invented ; but as we have now to consider only machines for light, only those generators which have yielded the most important results will occupy our attention, and these are : — 1°, the machines of the Alliance Co.; 2°, the dynamo-electric machines of Wilde, Ladd, Wallace Farmer, Siemens, Lontin, &c. ; 3°, the elec- tro-magnetic machines with rings of Gramme, De Meritens1, Brush, and Biirgin; 4°, the divided current machines of 62 ELECTRIC LIGHTING. Lontin, Gramme, Siemens, &c. We shall of course begin with the Alliance machines, the earliest of all The Alliance machines. — The comparatively powerful effects produced by the magneto-electric machines of Pixii and of Clarke at once suggested the idea of using them as economical generators of electricity, and it was supposed that by making them of a large size and driving them with a steam-engine, not only might the cost of electricity, which is so great with batteries, be considerably reduced, but currents of much greater regularity and constancy might be obtained. By the year 1849 Nollet, professor of physics in the military school at Brussels, proposed in fact to construct a Clarke's machine on a large scale, and by introducing into it the im- provements which the progress of science and his own ex- periments had suggested, he invented the machine which is now known as the Alliance Machine. Unfortunately his labours were arrested by death, when he was about to see his scheme, not indeed successful, for its success was not to be immediate, but to see it submitted to practical tests. Bold speculators, helped by rich and powerful persons, succeeded indeed in floating a scheme, founded on Nollet's machines, for extracting gas from water. This was a mad notion, but for that very reason it found adherents, and it was the Company then formed which, under the name of The Alliance, set up by the year 1855, in buildings belonging to the Hotel des Invaltdes, the first large magneto-electric machines known in Europe. Of course the results obtained were miserable, to say the least, and in 1856 the Company was compelled to go into liquidation. It was reorganized some time afterwards, and having ap- pointed Berlioz as its manager, it endeavoured to profit by the considerable materials it had formed. I was then con- sulted, and indicated to Berlioz the application which he could make of them to the electric light and to electro- plating. But for that purpose many improvements had to VARIOUS GENERATORS OF ELECTRIC LIGHT. 63 be introduced into these machines, and all the electricians of the time contributed each a stone to the edifice. Among these improvements there is one by which double the effect is at once obtained. It was suggested by Masson, it that lime professor of natural philosophy at the Ecole Centrale, whom I took to see these machines, and consisted in the suppression of the commutator formerly used in all such machines. All these improvements, which" were ad- mirably carried out by Van Malderen, the Company's engi- neer, made it possible, in conjunction with the improvements he himself introduced, to carry the machines to a degree of perfection which the most sanguine had scarcely hoped for, and which it was thought could not be surpassed. And then, for the first time, attention was turned to the availability of the electric light for the illumination of ships and lighthouses. Jn conjunctfon with Reynaud and Degrand, experiments were thereupon instituted, and shortly afterwards, in 1863, it was found possible to illuminate the lighthouses of La Heve in this way. About the same time experiments were made regarding the use of the electric light for ships. These experiments did not at first prove entirely successful, owing chiefly to the opposition of the navy. But after a while the importance of the results obtained came to be understood. France thus led the van in this double path of progress, as she still does in the matter of public illumination ; but we must point out that the civilized world is indebted to the Alliance Company, .and to the enterprise of its intelligent manager Berlioz, for these beautiful applications of electricity. Now that we have acknowledged our obligations to a Com- pany whose efforts fortune failed to reward, we are going to study the constructive details of their machines. The principle of the Alliance Company's machine much resembles that on which Clarke's machine is based, but the mechanical arrangement is so contrived that the induction -coils and magnets may be multiplied without inconvenience. 64 FJ.ECTRIC LIGHTING. Sixteen coils are placed on a bronze wheel provided with suitable receptacles and fastening-collars by which they are strongly clasped. This arrangement, which for brevity we shall call the disc, revolves between two rows of horseshoe magnets, fixed parallel to the plane of the disc on a particu- lar kind of frame constructed of wood only, at least in the vicinity of the magnets. Each row contains eight magnets, and thus presents sixteen poles at uniform distances .apart. Thus there are as many poles as coils, and when any one of the latter is opposite a pole the remaining fifteen are in a like position. One machine may have several discs mounted on the same axle, with several rows of magnets fixed on the same frame. The number of discs does not usually exceed six, as the machines would be too long, and it would be difficult to avoid flexure of the axle and the frame. It should be borne in mind that the coils must revolve as close as possible to the magnets, but without touching them. At the present time nearly all these machines have only four discs. Fig. 15 shows a general view of a six-disc machine, and Fig. 1 6 indicates the manner in which each coil E is pre- sented before the poles A, B ; B", A"; a, b ; //, a', &c., of the different magnets, which are so placed in the contiguous rows that a north pole may be opposite to a south pole in order to polarize in opposite directions the magnetic core of each coil. The ends of the wires of the coils are attached to flat pieces of wood fixed on the bronze wheel, and on these the coils may be arranged either for tension or for quantity, like the elements of an hydro- electric battery. One pole of the total current is connected with the axle' c, which com- municates with the frame through the bearings ; the other pole ends in a metallic collar, D c, concentric with the axle, and insulated from it by wood or ebonite, and these two poles are connected with the external circuit by the cast-iron frame of the apparatus, and by a friction-spring which, by means of a wire, is in communication with a binding screw. VARIOUS GENERATORS OF ELECTRIC LIGHT. 65 We shall now enter into some details of the parts of these achines. The magnets are, as we have said, of a horseshoe form. and come from the workshops of Allevard ; they weigh about 20 kilogrammes each, and are composed of five or six plates of tempered steel, fitted to each other by grinding, and held together by screws; the thickness of each plate is about 5 <36 ELECTRIC LIGHTING. i centimetre. For the sake of perfect uniformity of thick- ness at their polar extremities, and also for readiness of ad- justment, they are provided with soft iron poles. Each plate is magnetized by the ordinary process, and the bundle should support three times its own weight. This magnetization is, moreover, improved rather than otherwise by the action of the machine. FIG 16. The bobbins have undergone many alterations; at the present time they are formed of iron tubes split longitudinally, and have brass caps also divided in order to diminish disad- vantageous induction. They are 10 centimetres long and 4 centimetres in diameter, and there are two tubes— one within the other. The length and action of the wires of these bobbins should depend, as will be readily understood, on the resistance of the external circuit, on the number of bobbins, and on the VARIOUS GENERATORS OF ELECTRIC LIGHT. 67 kind of work required. Nevertheless, experience has shown that in the use of the machine for the electric light, the best results have been yielded with bundles containing eight insu- lated wires, each i millimetre in diameter, 30 metres long, and joined eight together. These wires are covered with a layer of cotton impregnated with a solution of Judea bitu- men in spirits of turpentine, or in benzine. This kind of varnish has the advantage of .drying very quickly, and not easily cracking ; it is besides very thin, so that it does not materially increase the thickness of the different layers of spires. According to the experiments of Jamin and Roger, the electro-motive force of the current sent from a six-disc Alliance machine, with the coils arranged for tension, and with a velocity of rotation of 200 turns per minute, is equi- valent to 226 Bunsen cells; but when the coils are arranged for quantity the electro-motive force is equivalent to only 38 Bunsen cells. The resistance of the generator was equivalent to that of 655 Bunsen cells in the first case, and to 1 8 in the second case. The light yielded may be nearly represented by that of 230 Carcel lamps, and the cost of the current pro- ducing it would be, according to the calculations of Reynaud, the inspector of lighthouses, i fr. 10 c. per hour. But since these experiments were made great improvements have been effected in the machines, and we shall presently give the figures representing their performance and their cost ; here we shall merely state as the net result that the ordinary Alliance machines with four discs are able, with a speed of 400 turns per minute, to maintain 3 Jablochkoff electric candles. Under exceptional circumstances this yield has even been very much increased, and machines have been spoken of capable of supplying 6 candles, but the ordinary machines are far from giving such a result. Some time after the definite establishment of the machines we are speaking of, Holmes, a former employe of the Alliance Company, made in England machines of the same kind, 5 - 2 68 ELECTRIC LIGHTING. but of larger dimensions, which however yielded results in- ferior to those just mentioned. Nevertheless, they were fitted up for the illumination of lighthouses on the English coast. It will be seen farther on by the trials made in Eng- land itself that of all the machines tried these gave the least advantageous results, and we shall, therefore, say nothing more about them. Wilde's Machine. — Soft iron being, by reason of its greater magnetic conductivity, capable of giving a much greater maximum magnetization than the tempered steel of which permanent magnets are formed, Wilde thought that there might be an advantage in employing as inducing organs electro-magnets instead of magnets, and that a small supple- mentary magneto-electric machine, actuated by the same motion as the induction machine proper, might be used to magnetize them. Reasoning in this way, he was led to the machine represented in Fig. 1 7, which was the starting-point for all the machines subsequently designated dynamo-electric machines. We shall, in fact, soon see that this small supplementary or priming machine was unnecessary, provided the whole or a part of the induced current were made to traverse the wire of the electro-magnets. Nevertheless, Wilde's machine pro- duced very good results, and it was the first machine of small dimensions which was able to generate the electric light ; but unfortunately it required a great speed of rotation, As shown in Fig. 17, this machine is composed of two parts, of which the upper one is in a manner a miniature copy of the other. The former is a Siemens' magneto- electric machine, with a magnet M of 16 plates, between whose poles, m ttt turns a Siemens' bobbin size, No. 58 S IOO 2-16 2-32 0-83 420 3 '5 85< SIEMENS, medium •> size, No. 68 ) too 2-16 2-32 0-83 420 3'3 85' 2 HOLMES, coupled 1,100 9-84 4'3y. 5 '24 11,480 6-5 2 GRAMMES, coupled 640 5-i6 2*58 4'o6 5,706 I0'5 42( 2 SIEMENS, medium -> size, Nos. 58 & 68 [• coupled ) 200 4'32 2 '32 i o'83 840 6-6 85< 124 HE ENGLISH EXPERIMENTS. Total Luminous Intensities. Luminous Intensities for each Horse-power of Force absorbed. Classification of the Machines in order of Merit. Condensed Light. Diffused Light. Condensed Light. Di L Can- dles. Tused ight. Carcel Lamps. Can- dles. Carcel Lamps. Can- dles. Carcel Lamps. Can- dles. 476 Carcel Lamps. 1.523 217*57 L523 217*57 68-00 476 68-00 6 j 1.953 279-00 1.953 279-00 543 77*57 543 77*57 5 6,663 951-86 4,016 57371 1,257 179-57 758 108-29 4 6,663 915-86 4,016 57371 1.257 179*57 758 108-29 4 4,818 2116-86 8,932 1276 "co 1,512 : 216 oo 9TI 130-14 3 5-539 791-29 3-339 477-00 1,582 226 "oo 954 136-29 2 6,864 980-57 4.138 59i*i4 2,080 297-14 '-254 433 179-14 I 2,811 40I-57 2,811 401-57 432 6171 6171 — 1.396 1628 -oo 6,869 981 -29 1,085 iSS-oo 654 93*29 — 4.134 2019-14 8,520 1217-14 2,141 305-86 1,291 i84*43 — 12 TABLE OF NEW EXPERIMEN Number of Candles s Engines. measured horizontally. a o V a c Cylin< p, Lamps ij i used. •s 1 ,S I'd a | a £ V 5 & §"§ o ^ 3 ^ I H 1** K 1 V Q" * K M Holmes . , Holmes Steam jq | j ' Alliance T O"Q C 1 O q QOO 667 Serrin Meritens large 8,600 I 228 Holmes Gas o T ,, small 1, 860 I 24.0 880 Duboscq Tr6 q i Siemens, medium 4,100 1.254 850 Siemens Steam — 13 I Gas . . . g 91 Gramme, for a single ) light, (A) 3 960 320 Steam ,, 1,824 730 920 •^'^ — — - 2 8t;o i 018 802 „ 2,520 740 830 Gas... 143 6| ] g 800 Brush Steam 6 Large Brush 1,230 077 i 340 6 ; Small QOO 2qn 6 i Large Wallace 821 800 9 *' Small A AC> 6 i Gramme Division Ma-O chine, 16 lights j Gramme Division Mao chine, 20 lights j 250 — 640 650 Jablochkoff... •- 142 10 r Lontin Division Ma- > chine, 6 lights $ 570 — 350 Serrin-Lontin •• no 9i i." 126 tRAWN UP BY SHOOLBRED. Engines. Remarks. NOMINAL. I 5 >ower per Light. EFFECTIVE. Localities. Observers. Observations. "rt § 1 1 '«£ 3'2 3'6 4 '5 7-0 »/S 3 '3 5'4 47 27 2'5 2-3 3'4 1-84 3*26 376 to to 8 8 0 8 8 2 3 '5 P p 9 io-9 4-2 8-9 8'2 3 5'i South Foreland Douglass N j Alternating I Magneto- ( Electric I Currents. V .. Dynamo- P^lectric > Machines with 1 Continuous Currents. / ^ Dynamo- [ Electric f Machines with \ Alternating ) Currents. Paris Allard Royal Institution Douglass South Foreland C London Stereo- 7 (. scopic Co. 3 La Chapcl'e, Paris ... Rouen Amos..... (Chemin de fcr du '(. Nord Powell Paris C Edmundson, West- ) ( minster j Philadelphia .,. Trcsca Shoolbrcd f Committee of the I Franklin Institute Allard — — 3-89 1-25 y Avenue de TOpera, ) (. Paris .J" Thames Embankment c St. Lazare Station, > t Paris 3 f Chemin de fer de i 1'Ouest 127 128 ELECTRIC LIGHTING. APPARATUS FOR EVOLVING THE ELECTRIC LIGHT. Light produced between carbon electrodes. — We saw at the beginning of this work that the electric light was produced by the passage of an electric discharge or current through a gaseous or solid body having a conductivity suffi- ciently small to become incandescent from the enormous heat developed by the passage of the electricity. We saw also that in order to obtain this light under the most favour able conditions it is necessary to use electrodes made of sub- stances capable of disintegration, and of ready combustibility, and that of all substances carbon is that which yielded the best results. It was Davy who first conceived the idea of using carbons as electrodes for producing the voltaic arc, but these carbons were sticks of charcoal quenched in water. These were, however, so quickly consumed that other physicists endea- voured to substitute a more durable form of carbon for the wood charcoal. Foucault was the first to make use of that product of coal which is deposited in the inside of the retorts employed in gas-making. He thus obtained a voltaic arc of much greater durability. There was, however, much that was objectionable in retort- coke, from its want of uniformity and from its admixture with earthy matters, which caused the light produced to be far from steady. The carbons were destroyed also in conse- quence of the fusion of the siliceous matter contained in them, and generally emitted vapours which, being better conductors than the arc, carried off a portion of the current as a non-luminous discharge. It is true that by suitably choosing these carbons, and by cutting them from the uniform portions of the deposits, good ones might be obtained ; but this material for the voltaic arc APPARATUS FOR THE ELECTRIC LIGHT. 129 has long been objected to, and has often proved an obstacle to the applications of the electric light. Yet even at the present time, and in spite of the progress which has recently been made in the preparation of carbons for the electric light, there are some who prefer the retort-coke. The above mentioned inconvenience of retort-carbons, as may readily be supposed, soon caused physicists and manufacturers to seek a method of preparing them of a purer chemical composition, and with a more uniform physical constitution. By certain processes of which we shall pre- sently speak, this has gradually been achieved. Into the composition of these artificial carbons certain metallic salts have sometimes been made to enter, in order that, under the threefold influence of the electrolytic, the calorific, and the reducing actions, the metals of the salts might be deposited on the negative pole, where their combustion in the air would increase the light of the arc itself. Carre, Gauduin, and Archereau have made experiments of this kind, which we shall describe farther on ; but the results have not been entirely satisfactory, as the light has generally been rendered flickering and unsteady. The simple carbon has therefore commonly been preferred. The resistance of the carbons is very variable. When they are of uniform section and quality the resistance is, of course, dependent on their length or thickness. Experi- ments made at Silvertown by Hospitalier and Robert Gray have shown a resistance of 3^25 ohms per lineal metre for Carre's carbon of 4 millimetres diameter. A copper con- ductor of the same dimensions would have a resistance of only 0*001315 ohm, and therefore the copper conducts 2,471 times better than the carbon. These figures may occasion surprise, but it should be remembered that they relate to carbons of small diameter, and that, if referred to carbon of i centimetre in diameter, the resistance is not more than o'52 ohm, or about 50 metres of telegraph wire. As early as the year 1846 Staite and Edwards had patented 9 130 ELECTRIC LIGHTING. electrodes for the electric light, made of a mixture of pounded coke and sugar, which, after having been moulded and powerfully compressed, was baked, then steeped in a con- centrated solution of sugar, and heated again to whiteness. Three years afterwards, in 1849, Leonolt patented carbons for the same purpose, made of two parts of retort-coke, two parts of wood charcoal, and one part of liquid tar. These substances were mixed into a paste, then subjected to power- ful compression, afterwards covered with a coating of sugar- syrup, and left for 20 to 30 hours at a high temperature. They were afterwards purified by successive immersions in acids. Lacassagne and Thiers conceived the idea of purify- ing sticks of carbon by steeping them in fused caustic, potash, or soda. The object of this operation was, accord- ing to the author, to change the silica contained in the carbon into soluble silicates. They were then steeped a few moments in boiling water, and afterwards exposed to a current of chlorine in a porcelain tube heated to redness, in order to convert the earths left unattacked by the potash or soda into volatile chlorides of silicium, calcium, potassium, iron, &c. A short time after these experiments Curmer proposed to form carbons by the calcination of a mixture of lamp black, benzine, and turpentine, the whole moulded into a cylindrical form. The decomposition of these substances left a porous coke, which was saturated with resin or syrup, and again calcined. These had little density and a low conducting power, but they were very uniform and free from all im- purities. The greatest success was, at the period we are speaking of, obtained by Jacquelain, formerly the chemist of the Ecole Centrale. Experiments on the electric light made with his carbons for the French lighthouse authorities were so con- clusive that the problem was supposed to be solved. In connection with these experiments I cannot refrain from giving here a letter written to me in 1858 by Berlioz, at that time manager of the Alliance Company, a man of intelli- APPARATUS FOR THE ELECTRIC LIGHT. 131 ;gence, who was then most enthusiastic as to the results ob- tained by his machines : — . . . I ought to have written sooner to tell you about the machine: it works perfectly well and with increasing power, and we have an admirable light. I wanted to tell you of this, which I consider of consequence, for, as usual, we have to thank you for this important matter. You induced me to see Jacquelain about his pure carbon coke, and this coke gives a steady light without flame, and of remarkable brilliancy. I am sorry you are not here, as you have been kind enough to bestow a fatherly care upon our machine. This evening I am showing the light to my superintending committee. We are able at a distance of 60 metres to read very small writing, and no doubt we could read at nearly a kilometre if space permitted. We have also magnificently illuminated the dome of Les Invalides, which is about 300 metres from our apparatus ; but I hope we shall soon perform the experiment on the Seine in a steamboat, as you advised. Thus with our machine taking the place of the voltaic pile, with the Jacquelain carbons, a regulating mechanism suited for alternately contrary currents, and a good reflector, the problem of < electric lighting on sea-going ships will be completely solved. Jacquelairis Method. — To obtain a pure carbon Jacque- lain makes use of the carbides of hydrogen, either those obtained in the distillation of coals, shales, turf, &c., or the ; numerous products formed during the carbonization of these combustibles in close vessels, or those represented by the heavy coal, shale, or turf oils, or by volatizable organic sub- rstance. " These organic materials," says Jacquelain, " contained in a cast-iron receptacle, are introduced into a cast-iron boiler on a lower level by a conducting pipe furnished with a cock. The boiler is also provided with a cock for emptying it. The vapours are carried by a cast-iron pipe into a horizontal retort made of refractory earthenware, and fitted with a screen for retarding the passage of the gaseous products. This communicates with -two cast-irons receivers, forming an inverted U, where the lamp 9—2 132 ELECTRIC LIGHTING. black is collected. Any obstruction in the retort is removed by means of a scraper. From the last receiver which forms the second branch of the inverted u, a curved pipe leads the hydro- gen gas and the undecomposed volatile products under a grating." Unfortunately this process was very incomplete, and offered no security for the quality of the products. Along with ex- cellent carbons giving very satisfactory results there were others very bad, and sometimes even worse than those ob- tained from retort-coke. Therefore Carre determined to study the problem afresh, and he thus speaks of it in the Comptes rendus de r Academic des Sciences of the igth February, 1877, P- 346:— Carres Method. — "The superiority for various experi- ments of artificial carbons, and the possibility of purifying by alkalis, acids, aqua-regia, &c., the carbonaceous powders that enter into their composition, then led me to seek for some means of producing them economically. By moisten- ing the powders either with syrups of gum, gelatine, &c., or with fixed oils thickened with resins, I succeeded in form- ing pastes sufficiently plastic and consistent to be forced into cylindrical rods through a draw-plate placed at the bottom of a powerful compression apparatus under the pressure of about 100 atmospheres. Carbons are now manufactured by this process, and 1 have at various times presented some of them to the Academic des Sciences and to the Societe a"* encouragement. " These carbons have three or four times the tenacity and are much more rigid than retort-coke carbons, and cylinders of 10 millimetres diameter may be used of a length of 50 centimetres without any danger of their bending or crossing during the breaks of the circuit, which often happens with others. They may be as easily obtained of the slenderest diameters (2 millimetres) as of the largest. " Their chemical and physical homogeneity g;ives great steadi- ness to the luminous arc ; their cylindrical form, combined with the regularity of their composition and structure, causes their cones to continue as perfectly shaped as if they had been turned APPARATUS FOR THE ELECTRIC LIGHT. 133 in a lathe, and therefore there are no occultations of the point of maximum light like those produced by the projecting and com- paratively cold corners of the retort-coke carbons. They are not liable to the inconvenience of flying in pieces when first lighted, as the others are, in consequence of the great and sudden ex- pansion of the gas contained in their closed cellular spaces, which are sometimes I cubic millimetre in capacity. By giving them one and the same uniform density they are consumed by the same amount for an equal section ; they are much better conductors, and, without the addition of any substances other than carbon, they are even more luminous in the proportion of i 25 to i." The preparation preferred by Carre is a mixture of pow- dered coke, calcined lamp black, and a syrup made of 30 parts of sugar and 12 parts of gum. The following formula is given in his patent of the 151!! January, 1876 : — Very pure coke, finely powdered ... ... 15 parts. Calcined lamp black 5 ., Syrup of sugar ... J to 8 „ The whole is well pounded together, and from i to 3 parts of water are added to make up for the loss by evapora- tion, and to give the required degree of consistence to the paste. The coke is to be made with the best samples ground -and purified by washing. The paste is then compressed and passed through a draw-hole, and the carbons are afterwards piled in crucibles and exposed for a certain length of time to a high temperature. Details of the operations for prepar- ing these carbons will be found at page 54 of H. Fontaine's •work. We may add that Carre, being desirous of imparting to the electrical light the hues most suitable for theatrical pur- poses, has succeeded in so treating his carbons that they impart to the light a tint which, instead of being bluish- white, has a rosy-yellow hue, that is very advantageous for bringing out the complexions of the actresses. Carre's manufacture has, since the great experiments made 134 ELECTRIC LIGHTING. by the Jablochkoff Company, been much extended, and,, thanks to the large scale on which his process is being carried out, he can now send out carbons 50 per cent, better than at first. These carbons are now at a price which allows of their practical use, and this circumstance greatly conduces to the development of the applications of the electric light. Gauduin' s System. — The carbons manufactured byGauduin are of pure carbon, their base being lamp black ; but as the price of this substance is comparatively high, and its management is difficult, Gauduin was obliged to seek for a better source of carbon, and he obtained it by heating in a closed vessel common rosin, pitch, tars, resins, bitumens, natural and artificial oils, essences, or organic matters, which, after decomposition by heat, leave sufficiently pure carbon. These products are put into crucibles and heated to bright redness, the volatile matters being conducted into a con- densing chamber, whence they are carried by a copper worm- tube with the condensed liquids, such as tars, oils, spirits, and carbides of hydrogen, into another worm-tube, where they are collected for use in the manufacture of the carbons. There remains in the retort some more or less compact car- bon, which is pulverized as finely as possible, and collected either alone or mixed with a certain quantity of lamp black by means of the carbides of hydrogen obtained as secondary products. Thus prepared, these carbons are completely free from iron, and are superior to those met with in commerce. For moulding his carbons the inventor uses steel moulds capable of withstanding the highest pressure of a powerful hydraulic press. The moulds are arranged like draw-plates, and their arrangement has been much improved by Gauduin, for in his process the sticks are supported throughout their whole length, so that they do not break by their own weight, as often happens with ordinary draw-plates. Quite recently Gauduin has further improved his process. APPARATUS FOR THE ELECTRIC LIGHT. 135 Instead of carbonizing wood and reducing the charcoal to powder, he selects a suitable piece of wood, which he cuts to the shape required in the carbon ; then he converts it into hard charcoal, and finally soaks it, as in the manufacture we have described. The heating of the wood is conducted slowly so as to drive off the volatile matters, and the final heat- ing takes place in a reducing atmosphere at a very high tem- perature. All impurities are removed from the wood by a preliminary washing in acids and in alkalis. Gauduin also points out a method of stopping the pores of the wood by heating it to redness and exposing it to the action of chloride of carbon and various carbides of hydrogen. He expects thus to produce electric carbons which will be very slowly consumed and will give a light absolutely steady. These carbons have not, however, done what was expected of them, therefore they have not become general, and are little used except by the firm of Sautter and Lemonnier. At the present time it is difficult to obtain them, and it is doubt- ful if they can compete with those of Carre, who employs as his raw material a very cheap substance, namely coke. Effects produced by the addition of metallic Salts to the Carbons prepared for the Electric Light. — As we have already seen, page 129, attempts have been made to obtain some advantage by associating the carbons with metallic salts, which might supply, independently of the voltaic arc, a light due to the combustion of the metal carried to the negative electrode. The following experiments have been undertaken by Gauduin, Carre, and Archereau : — Gauduin mixed the following substances with the pure carbon : — phosphate of lime from bones, chloride of calcium, borate of lime, silicate cf lime, pure precipitated silica, mag- nesia, borate of magnesia, alumina, silicate of alumina. The proportions were so calculated as to leave 5 per cent, of oxide after the carbons were dried. They were submitted to the action of an electric current always in the same direction, supplied by a Gramme machine sufficiently powerful to main- *36 ELECTRIC LIGHTING. tain an arc of 10 to 15 millimetres long. The following results were obtained : — i°. The phosphate of lime was decomposed, and the re- duced calcium burnt in the air with a reddish flame ; the light measured by the photometer was double that given by retort-coke carbons of the same section. The lime and the phosphoric acid were, however, diffused in the air, producing some smoke. 2°. The chloride of calcium, the borate, and the silicate, were also decomposed. But the boric and silicic acids seemed to escape the action of the electricity by volatilizing. The light was less than with phosphate of lime. 3°. Silica made the carbon worse conductors, and dimi- nished the light; it also fused and was volatilized without being decomposed. 4°. Magnesia, borate and phosphate of magnesia were de- composed, and the magnesia vapour passing to the negative pole burnt in the air with a white flame, which much in- creased the light — 'less so, however, than the lime salts. Alagnesia, boric, and phosphoric acid were diffused in the air as smoke. 5°. Alumina and silicate of alumina were difficult to de- compose ; a very strong current and a considerable arc were required. Under these conditions the aluminum was seen to issue in vapour from the negative pole like a. jet of gas, and to burn with a bluish little — luminous flarne. Archereau found that the introduction of magnesia into the carbons increased their illuminating power in the pro- portion of i '3 4 to i. According to Carre it would seem— i°. That potash and soda doubled at least the length of the arc, rendered it silent, and, by combining with the silica always present in retort-coke, they eliminated it in the state of transparent, vitreous, and often colourless, globules at 6 or 7 millimetres from the points. The light was increased in the proportion of 1*25 to i ; .. APPARATUS FOR THE ELECTRIC LIGHT. 137 2°. That lime, magnesia, and strontia increased the light in the proportion of 1*30 or 1-50 to i, giving it different colours ; 3°. That iron and antimony brought up the increase to i*6o and 170; 4°. That boric acid increased the duration of the carbons by covering them with a glassy coating which protected them from the air, but without increasing the light ; 5°. That the impregnation of pure and regularly porous carbons by solutions of various substances is a convenient and economical means of producing their spectra, but it is preferable to mix the elementary substances with compound -carbons. Metallized Carbons. —According to E. Reynier, the carbons being consumed a little by combustion on their lateral surfaces, which glow for a length of 7 or 8 millimetres above and below the luminous point, there is a complete loss as regards the light, and it would be an advantage to cover them with a metallic covering in order to avoid this lateral combustion. It follows, in fact, from his experiments in the workshops of Sautter and Lemonnier with a Gramme machine of the 1876 form, that the metallized carbons are consumed less than the ordinary carbons, as the following results show:— easures imps. Dimensions of the Carbons. State of the surface of the Carbons. Length of Carbons used in one Hour. «J •2-5 I'9 d — 7 mm. J S=o'3846sq. cm. | Uncovered Covered with copper Covered with nickel At the + pole. 1 66 mm. 146 , 106 , At the — pole. 68 mm. 40 , 38 , • Total. 234mm. 186 144 Jete 947 917 d = 9 mm. I S =0*6358 sq. cm. | Uncovered ... ... Covered with copper Covered with nickel 104 , 98 , 68 , 50 , 132 104 528 553 5*6 I38 ELECTRIC LIGHTING. " These experiments," he says, " were made with Carre* car- bons and a Serrin lamp. It was observed that with naked carbons those of the smallest diameter had the longest points, as might have been expected ; but with the metallized carbons the reverse was the case, a circumstance difficult to explain. " We may, however, conclude from these experiments : — " i°. That independently of the improvement in the shape of the positive carbon, covering it with nickel lengthens the duration by 50 per cent, of the carbons of 9 millimetres, and by 62 per cent, that of the 7 millimetre carbons. Covering with copper also increases the duration by an amount intermediate between that of the naked and of the nickelized carbons ; " 2°. That with equal sections the metallization of the carbons does not seem to modify the amount of light yielded by them ; " 3°. That the luminous power of the carbons of small diameter is much superior for the same electric intensity to that cf car- bons of large diameter, which is explained by the fact that con- ducting bodies placed in a circuit composed of conductors of large section or of high conductivity become the more heated as their diameter is smaller ; and also from the fact of the polariza- tion being the more energetic, the smaller are the carbons. This the more concentrates the resulting calorific effects of which we have spoken above. It depends also upon the circumstance, that to obtain the maximum of light it is necessary that the re- sistance of the circuit of the voltaic arc should be as nearly as possible that of the generator ; "4°. That the metallization by allowing carbons of small instead of large section to be used for the same time of action, gave advantageous results. This metallization is effected gal- vanically." According to A. Ikelmer, this system of metallization can have no advantage except in so far as it may lessen the re- sistance of the carbons, and as the thin layer of metal is oxidized by the influence of the high temperature and disin- tegrated for a distance that may reach as much as 10 centi- metres, an improvement of the conductivity cannot in this way be obtained, any more than the prevention of lateral combustion. Consequently, he thinks that the problem APPARATUS FOR THE ELECTRIC LIGHT. 139 would be solved much better, at least so far as increased con- ductivity is concerned, by connecting with metallic rods. These rods may be placed either within the carbons themselves or in the electric candles on both sides of the insulator, and thus they may become heated without danger of oxidation. Jablochkoff, who had patented this system of rods as early as November, 1878, finds them advantageous only so far as they lessen the consumption of the electrodes; and if instead of a covering of carbon use is made of a covering of mag- nesia and an iron rod, as in the preparation of these candles^ the consumption might be reduced to one-eighth. But this result is obtained at the expense of the brilliancy of the light produced, which is then reduced to that of six or eight gas- jets. Nevertheless, as there are cases where there is an ad- vantage in increasing the duration of the candle at the cost of the luminous intensity, he has patented this system with the idea of using it in Russia for lighting carriages. Effect of heat on the conductivity of the carhons* — Heat is known to modify the electric conductivity of bodies, diminishing that of metallic conductors, and usually increasing that of substances of mediocre conductivity, whether liquid or solid;* and carbon is precisely one of these. According to the researches ot Borgman, the tempera- ture of a carbonaceous substance heated to orange-red de- creases its resistance in the following ratio : — For wood charcoal 0x30370 — between 26° and 260* For Dormez anthracite ... 0-00265— „ 20° „ 260* For Alibert plumbago ... 0-00082— „ 25° „ 250° For coke 0-00026— „ 26° „ 275* It would seem that even a feeble calorific radiation causes a diminution of resistance in plates of wood charcoal, and that between 100 and 125 degrees the resistance of pine- wood, elm, and ebony charcoals notably varies. * See my Paper on the conductivity of mediocre conductors, p. 27. 140 ELECTRIC LIGHTING. Light produced by means of conductors of in- different conductivity. — We have said at the commence- ment of this chapter, that one of the means of producing the electric light is the heating which takes place when a power- ful current traverses a body of indifferent conductivity inter- posed between two electrodes of good conductivity. We have also seen that rods of carbon and of refractory sub- stances constitute these bodies of indifferent conductivity, and that Jablochkoff on one hand, and Lodyguine and Kosloff on the other, had made some very interesting ex- periments on this subject. It is this new .method of pro- ducing the electric light that we are now about to consider, and we shall begin by Jablochkoff's system, which is the most curious. Jablochkoff ^ s System. — In this new system the induction currents from a Ruhmkoft's coil of moderate dimensions are used, and a piece of slightly baked kaolin, two milli- metres thick and one centimetre wide, which forms the semi- conducting substance, is required to supply the incandescent point, or rather the luminous source, for the whole appears to be illuminated. With a single coil two sources of light can easily be obtained in one circuit, but by increasing the number of induction coils and the power of the generator, the number of these luminous sources may be indefinitely increased— a circumstance which may in some degree solve the difficult problem of the division of the electric light. We shall, however, consider this question farther on. The arrangement of the system is very simple ; the small piece of kaolin is introduced between two little iron nippers which form the polar electrodes, and which are themselves attached to two clips capable of moving horizontally by means of a screw. These nippers seize the piece of kaolin by its' upper slightly thinned edge, and even a little beyond this edge, in order that the apparatus may be more easily lighted ; for this apparatus must be lighted, and it will readily be understood that the substance is not of itself sufficiently APPARATUS FOR THE ELECTRIC LIGHT. 141 conductive, even of induced currents, to be capable of passing a current able to produce the electric light. To remedy this defect of conductivity, the plate of kaolin must be warmed in the neighbourhood of the electrodes, and this is done in a very simple manner, by connecting the two iron clips mentioned above by a rod of retort carbon. By first taking a spark from one of the clips the carbon glows and transmits its heat to the neighbouring part of the kaolin, which fuses and affords a passage to the electric influence, at first through a very short space (i or 2 millimetres), then over one progressively longer in proportion as the carbon is moved along the kaolin, and which at length occupies the whole length of the latter when the carbon has reached the second iron clip. The current then follows the track of fused matter which is progressively formed, and reveals itself to the eye as a band of dazzling light, appearing much wider than it really is on account' of irradiation. Care must be taken to concentrate the heat developed by the carbon by means of a reflector of refractory matter, which may consist of a plate of kaolin. The light thus supplied is, as I have already said, very steady, very brilliant, and much softer than the arc light. Its power depends, of course, on the resistance of the circuit and on the number of luminous points interposed, but with a weak electric force it is equal to one or two gas-jets. Kaolin appears to be the best substance, because, being prepared as a paste, it can be made very homogeneous; but other substances are capable of producing the same effects; magnesia and lime have indeed given very good results.* * The conductivity of this kaolin, studied by means of the process used in my researches on substances of mediocre conductivity, showed traces of the passage of the voltaic current produced by 12 Leclanche elements only when the specimen had remained in a cellar for more than a day. When it was kept in an inhabited apartment it gave no deviation, and when it was heated to redness in a spirit-lamp it gave a deviation of only i degree. It is therefore necessary in order to obtain the important effects that have been mentioned, that the electricity of tension should, in consequence of the '-esist- 142 ELECTRIC LIGHTING. A very interesting circumstance was discovered in the ex- periments made by Jablochkoff, and that is, that the currents supplied by the induction apparatus intended for the light are most advantageously exerted by a magneto-electric gene- rator giving alternately reversed currents, such as those supplied by the Alliance Company or the Lontin Company. With such a generator, the induction apparatus does not re- quire either a condenser or an interrupter, and the intensity of the current is increased by the suppression of these. On the other hand the tension is notably diminished, for in ex- periments I have witnessed, the spark was hardly 2 milli- metres long ; but in order to obtain calorific effects, tension is specially necessary, and we have seen that irrthis respect the results have left nothing to be desired. Thanks to this plan, a Ruhmkoff induction machine can supply the electric light, and this result is of the more importance that a regu- lator of the light is not required to render it steady, and that the consumption of kaolin is almost insignificant (i milli- metre per hour). The magneto-electric generator itself need not be powerful, and it may, as we have already said, be pro- portioned to the number of luminous jets desired, care being taken to connect a suitable number of induction coils having their induced wires not too thin. When the illumination ot a long strip of kaolin is required under the influence of a very powerful current, it becomes necessary in order to light it to mark a line of lead on the upper edge of the bad conductor from one electrode to the other. The current is at first conducted by this streak, but is not long in heating the kaolin and in producing the effects we have described. This arrangement enables a large quan- tity of light to be obtained in a small space, for in order to increase the effect it suffices to fold the strip several times on itself like an electric multiplier. ance it meets in its passage, be accumulated within the very substance of the bad conductor, and should be transformed into heat by not being able to flow away fast enough as an electric charge. APPARATUS FOR THE ELECTRIC LIGHT. 143 According to Jablochkoff, the luminous intensity of these different sources varies according to the arrangement and dimensions of the coil, and the number of lights interposed in the circuit of each coil. They may therefore be so arranged as to supply light of different intensities from their minimum light of i or 2 gas-jets up to a light equivalent to 15 jets. u In this system," says Jablocbkoff, " the method of distribution of the circuits is in fact reduced to a central artery represented by a series of anterior wires corresponding with the inducing helices of the different coils, and with as many partial circuits as there are coils ; these last circuits corresponding with the induced wires of the coils, and ending separately at the different luminous ioci which are to be maintained. Each of these foci is therefore perfectly independent, and can be extinguished or lighted sepa- rately. Under these conditions the distribution of the electricity becomes very similar to that of gas, and I have been able to have 50 foci simultaneously illuminated with different intensities.3' Jablochkoff has lately rendered more practical the system we have just described by causing the current supplied by the small pattern of the magneto-electric Alliance machine to act directly. In order to impart more tension to the "currents he adopts a condenser of rather large surface to one of the wires going from the machine to each apparatus. This •condenser is composed of sheets of tinfoil, india-rubber, and varnished silk, alternated and folded as in the English con- densers for submarine cables. In this way, with a total con- densing surface of 200 square metres, seven points of light may be obtained instead of two, and what is more curious, the increased effect is produced even with currents alter- nately reversed. The arrangement of this system is besides very simple ; one of the armatures of each condenser is con- nected with one of the two wires of the machine, and the second wire of this machine is connected with one of the •clips of each light apparatus, while the other clip corresponds with the second armature of each condenser. There is then produced within each condenser successive fluxes of each 144 ELECTRIC LIGHTING. kind of the contrary electricities, which, by charging the- condensers, cause the illumination of the plates of kaolin, oa which streaks of black lead extend from one clip to the other- Lodygiiine and Kosloff 's System. — Of the various methods used for obtaining luminous effects by the narrowing of the section, :pf a good conductor, that contrived by Lody- guine and Koslqff had given some very interesting results These results even made some noise in 1874, for the effects were nearly similar to those we have just mentioned; but in order to produce them a. .much greater electrical force was required, and the portions of the apparatus brought to a white heat, being made of retort-coke of Very narrow section, did not present the desired conditions of solidity and stability. In this system these little needles of carbon were cut out of carbon prisms of at least i centimetre across, and were fixed between two insulated clips connected with the two- branches Of the circuit, as in JablochkofPs system.* In order to prevent their combustion they were enclosed in vessels free from air, or simply hermetically sealed, so that the oxygen of the enclosed air should not be renewed. With a powerful Alliance machine four luminous foci could,, it is said, be obtained in this way, and the lighting power was very satisfactory. Unfortunately these carbons were fre- quently broken, and it was quite a task to replace them. To obviate this inconvenience several ingenious arrangements were invented, of which we shall speak farther on; but all these systems have scarcely yielded anything very satisfactory from a practical point of view. The slender rod of carbon was in fact consumed parti- * It -seems that the connection of the carbons intended to become red hot with tlic wires of the circuit was one of the difficulties which checked 'Lodyguine and Kosloff. In fact, by making the wire penetrate into the carjiop, the latter was broken, on account of the difference of the ex- pansion pf the metal and that of the carbon, and again the metal by touch- ing the parts'qf the carbon heated to whiteness was melted. Kosloff, after many experiments, has, it appears, avoided these difficulties by using a special'metal to form the supports of the carbon rods. APPARATUS FOR THE ELECTRIC LIGHT. 1 45 cularly in the middle, and it was observed that the residue showed a notable part of the carbon to have been ignited, hence a considerable loss in the strip of carbon. Edisorfs System. — In the year 1845 Draper had endea- voured in America to take advantage of the incandescence of a platinum wire, rolled in a spiral form, as a focus of electric light, and in 1858 there was much said of a system invented by De Changy, which 'was merely the same thing.* Latterly Edison has taken up the notion, and has made noise about it in the newspapers great enough to lower very considerably the shares of the gas companies. But this system, besides lacking novelty, only very imperfectly solves the problem, and in the chapter on incandescent electric lamps will be seen the means he used, not to produce the electric light, for everybody knew that, but to prevent the platinum spirit from melting when the intensity of the heat exceeded the melting point of platinum. t Hospitalier has also invented a kind of regulator for the same purpose, but it is more complicated, and makes the apparatus a sort of incandescent electro-magnetic lamp. We think that all these methods of electric lighting founded solely on the effects of incandescence leave much to be desired, and we * In the Compies rendus de r Academic dcs Sciences de Paris, 2/th Feb., 1858, it will be seen that Jobard, of Brussels, had announced to the company that De Changy had just solved the problem cf the divisibility of the electric light by help of the incandescence of platinum spirals. (See my Expose" des applications de I' Electricity t. IV., p. 501, ame Edition.) f It appears, according to some American newspapers, that the pretended marvellous discovery of Edison is not serious. Here is what we read in the New York Herald of 5th Jan., 1879: — Mr. Edison has received from the Electric Light Company 100,000 livres in order to continue his experiments. He has already spent, it is believed, 76,000 livres up to the present time, and he has yet produced nothing but promises ; but it may be taken as certain that none of these will be fulfilled, and that no important revolution as regards the electric light will come from Menlo Park for the next 50 years, at least if one may judge by the present rate of p'rogress. What will be done there will depend in a great measure on the conclusions arrived at after all the researches and experiments conducted in the whole world, the results of which are continually sent to Mr. Edison. 10 146 ELECTRIC LIGHTING. believe those to be preferable in which combustion and the voltaic arc are united to incandescence. We shall see that in the systems of Reynier, Werdermann, &c., the results are in fact much more satisfactory. Systems of E. Reynier, Werdermann, and others. — By the beginning of the year 1878 Emile Reynier, struck by the ad- vantages which incandescent effects offered for the ready production of the electric light, and especially for its sub- division, bethought himself of combining these advantageous effects with those of the voltaic arc, and for that purpose he arranged the carbons of the King or Lodyguine system in such a manner that they might burn and furnish at the point of contact a small voltaic arc resulting from the repulsions produced by contiguous elements of the same current, as in the case of Fernet's and Van Malderen's regulators. He therefore arranged above a large fixed carbon a very slender rod (about 2 millimetres diameter), which was supported vertically by means of a heavy holder, and he connected this carbon with the current at a suitable height above the fixed carbon, so as to give a bright incandescence to the thin car- bon. And by this arrangement as the thin rod of carbon was consumed at the point of contact with the large carbon, it was renewed by a progressive advance produced by the weight of the holder. From this combustion, however, ashes were produced, which accumulated round the point of con- tact, and therefore he arranged the apparatus so that the large carbon might by a rotatory movement cause the ashes to fall off. Under these conditions Reynier was able to light 5 lamps with the current from a Bunsen battery of 30 cells, and he was even able to keep one of the lamps lighted for more than a quarter of an hour with the current from a Plante polarization battery of 3 elements. Some time after- wards the same idea was taken up by Werdermann, who used an arrangement the reverse of Reynier's. The stick ot carbon was pushed upwards by a counterpoise, and thus the large carbon did not require to be moved. According to APPARATUS FOR THE ELECTRIC LIGHT. 14? him this plan gave very good results, and with an electro- plating Gramme machine arranged for quantity he was able to light in 10 derived circuits 10 lamps of this kind, each giving a light equal to 40 candles.* The experiment he made on this occasion on the influence exerted by electrodes of carbons of different diameters, being very interesting as regards the present question, we here give them from a paper presented to the Academic des Sciences on the i8th November, 1878:— " When the electric arc is produced between two carbons of the same section," says Werdermann, " the changes in the polar extremities take place thus : the positive electrode heated to whiteness takes the shape of a mushroom, is hollowed into a crater form, and is consumed twice as fast as the negative electrode. The latter, which is only heated to redness by the current, is then slowly shaped into a point, and the length of the arc is in pro- portion to the tension of the current. " Things do not pass thus if a different section is given to each electrode. When the section of the positive electrode is gradually lessened, and that of the negative electrode increased, the red heat seen at the point of the latter diminishes more and more, whilst the heat of the positive electrode increases in proportion to the reduction of the section. The electric current no longer passes over the space between the electrodes with the same facility, and in order to maintain the voltaic arc the electrodes must be brought nearer together. " A strange phenomenon then appears ; the end of the positive electrode enlarges considerably, and the current shows a ten- dency to equalize the two surfaces, that is to say, to give to the positive electrode as much as possible the same section as the negative. The greater the difference between the sections of the carbon, the more must the distance between them be lessened ; and to avoid too great a swelling of the positive electrode the tension of the current must be a little reduced, which is easily done by using a Gramme machine, with which the tension of the * According to Werdermann, the machine required only 2 horse-power to yield this light. But I am informed that this is incorrect, and that a much greater force must be taken into account. IO — 2 148 ELECTRIC LIGHTING. current is proportional to the speed, the resistance of the coil being constant. " We thus reach a limit where the distance between the elec- trodes becomes infinitely small, that is to say when the electrodes are in contact. This is when their sections are nearly as i to 64 ; then the negative electrode scarcely gets heated, and it is therefore not consumed. Under these conditions it is only the positive electrode which is consumed, while it produces a beauti- ful absolutely fixed white light so long as the intimate contact between it and the negative electrode is maintained. It is in reality a light produced by an infinitely small voltaic arc. When the reverse method is adopted, that is to say, when, instead of lessening the section of the positive electrode, the section of the negative is gradually reduced, and the section of the positive increased, the light of the latter is seen to gradually diminish, and the heat of the negative electrode to increase. "When the sections of the electrodes are nearly as i to 64, and they have been placed in contact, no light is any longer given off by the positive electrode ; the negative one alone pro- duces the light. It is curious that when a voltaic arc is set up between the two carbons, the smaller electrode is always shaped into a point, whether it be positive or negative." Fig. 37 represents the series of changes in the form of the electrodes when their respective dimensions are made to vary. The electrodes in the centre represent the ordinary electrodes of equal section. In the three systems on the left are shown the effects produced as the lower electrode, which is the positive one, increases, and the three on the right show the effects of a successive increase of the upper and negative electrode. As we have seen, Werdermann has been able by derivation to obtain with an electroplating Gramme machine the light- ing of 10 electric lamps. The resistance of the coil of the machine was o'ooS ohm, and the electro-motive force was equal to that of 4 Daniell cells, with a speed of 800 turns per minute. At this speed the current corresponded with 66*06 webers; but with a speed of 900 turns it corresponded APPARATUS FOR THE ELECTRIC LIGHT. 149 with 88 '49 webers. Nevertheless, with the 10 lamps arranged by Werdermann the velocity of 800 turns was sufficient fc According to Werdermann, the resistances of the circuit were : — For one lamp 0*392 ohm For five lamps 0^076 „ For six lamps ,, 0-037 „ 150 ELECTRIC LIGHTING. It is surprising that with a current having so little tension such results could be obtained, and some sceptical persons would at first have denied the fact, contending that in order to obtain a focus of electric light an electro-motive force equal to that of at least 30 Bunsen cells was required ; but such persons did not observe that with incandescent lamps there is no appreciable solution of continuity in the metallic circuit, and that a source of electricity of quantity suffices to produce incandescence in such a circuit. If from the resistance of a light circuit, four or five thousand metres of telegraph wire be taken away, together with the nearly equivalent resistance of the battery, and that of the electro- magnetic apparatus of the regulator, it may be understood how an electro-motive force equal to that of only 4 Daniell cells can produce effects of incandescence in a circuit of extremely small resistance, and even produce several points of light, by derivations from the current, since the total resistance of the circuit is then in a manner diminished proportionably to the number of derivations. We are not sufficiently familiarized with effects of this kind, and mistakes are often made by confounding phenomena that are produced under very different electrical conditions. Light produced by means of an inductive action* — A short time before his death, Fuller, who had been one of Edison's fellow labourers, invented a system of electric lighting, on which we think it right to say a few words, although it appears to us scarcely practical. Here, however, is the description of it given in the Telegraphic Journal ' : — In this system the principal current does not produce the light, but it engenders another current in a series of induction coils, and each lamp is lighted by the current of one of the coils. An alternating current must be used. The induction coils were constructed as follows: — Two magnetic cores, placed parallel to each other, were magnetically connected at one end. Round the centre of each of these cores was a APPARATUS FOR THE ELECTRIC LIGHT. 151 head of soft iron, and at a suitable distance a head ot in- sulating substance. The external extremities of the cores were covered with insulated copper wire, and were connected together, as well as to the principal wire (or to the generator), in such a manner as to produce two opposite magnetic poles. Between the soft iron heads and the coils spirals of finer wire were wound, the fineness depending on the tension re- quired. To one of the iron heads an iron arm was jointed capable of turning and being applied to the other head, and thus magnetically connecting the N. and S. poles. In a system thus arranged, if a current is sent through the principal wire, and quickly reversed in the opposite direction, the polarities of the magnetic cores will change, and these changes will produce an induced current of high tension in the smaller coils ; this second current, if it is carried to a lamp by suit- able wires, will maintain the carbon or platinum in a high state of incandescence. Two or four induction coils may be connected together for lights of the highest power. PART III.— ELECTRIC LAMPS. IN order to obtain a continuous action from the carbons used for the electric light, these carbons must, in proportion to their consumption, be brought near each other, so that the intensity of the current may be kept as constant as possible. Now in order to obtain this result, arrangements have been contrived which effect it automatically, and these form what are called regulators of the electric light, or simply electric lamps. Of course the construction of these apparatus varies according as the light is produced by the voltaic arc or by incandescence. VOLTAIC ARC LAMPS. Electric lamps are of an earlier date than is generally supposed. In 1840 they consisted of a kind of Lannes exciter, the balls of which were replaced by sticks of carbon, which were pushed forward by hand as they were consumed, and the apparatus had the form represented in Fig. i. A little later an attempt was made to render the forward move- ment of the carbons automatic by placing the holders under the control of clockwork, or of electro-magnetic effects, which could act as a balance, that is to say, on the least variation in the intensity of the current. Then the idea was conceived of arranging the, carbons so as to burn like a candle, and it is to this last plan that recourse has been had in the attempts at electric lighting of the streets, which so astonished visitors during the whole period of the Exhibition of 1878. VOLTAIC ARC LAMPS. 153 The first automatic electric lamp appears to have been in- vented in 1845 by Thomas Wright; but it was not until 1848, when Stake and Petrie, in England, and Foucault, in France, invented their regulators that attention was attracted to the matter ; and before these apparatus could be regarded as capable of any practical application, Archereau on the one hand, and J. Duboscq on the other, had used them for numerous experiments in projection. From that time, and especially after the curious results obtained by the machines of the Alliance Company, people began everywhere to apply* themselves to the improvement of these apparatus, and a multitude of systems were invented, the most important being those of Serrin, Duboscq, Siemens, Carre, Lontin, Rapieff, Brush, &c. Having in the 5th vol. of my Expose des appli- cations de relectricite described most of these numerous systems, I shall here concern myself only with those that have become common in practice. Regulators of the electric light may be divided into six classes, viz. : i°. regulators founded on the attraction of sole- noids, and to this class belong the regulators of Archereau, Loiseau, Gaiffe, Jaspar, Carre', and Brush; 2°. regulators founded on the approximation of the carbons by successive movements produced electro-magnetically? and among these we may mention the regulators of Foucault, Duboscq, Deleuil, Serrin, Siemens, Girouard, Lontin, Mersanne, Wallace Farmer, Rapieff; 3°. regulators with circular car- bons, the most important types being the regulators of Thomas Wright, Lemolt, Harisson, and Reynier; 4°. regu- lators with a hydrostatic action, amongst which we may name those of Lacassagne and Thiers, Pascal, Marcais and Du- boscq, Way, Molera, and Cebrian; 5°. reaction regulators, such as those of Fernet, Van Malderen, and Bailhache; 6°. electric candles on the Jablochkoff plan, and others. The incandescent carbon regulators forming a class by them- selves, we shall discuss farther on. Of all these apparatus, those of Foucault and Duboscq, T54 ELECTRIC LIGHTING. Serrin, Gaiffe, Siemens, Carre, Lontin, RapiefT, Brush, and Biirgin are the only ones which are practically used, and therefore we shall describe the details of these only. Fo lira ii It and Duboscq's Lamps.— Foucault was, as we have seen, one of the first to invent the regulator with a fixed luminous point working by successive electro-magnetic movements. He thus describes his apparatus to the Academic des Sciences : — "The two carbon holders tend to move towards each other by springs, but they can only move by setting in motion a train of wheels, the last of which is controlled by a catch. It is here that electro-magnetism comes into play; the current that lights the carbons passes through the spires of an electro-magnet with an energy varying with the intensity of the currents. This electro-magnet acts on a piece of soft iron drawn away on the other hand by an antagonist spring. On this piece of soft iron is mounted the detent which en- gages the wheel and allows it to pass when necessary, and the direction of the movement is such that it presses on the wheel when the current becomes stronger, and liberates it when the current becomes weaker. Now, as the current becomes stronger or weaker precisely when the interpolar distance decreases or increases, it will be understood how the carbons have the power of approaching at the instant the distance between has just been increased, and that this approach cannot proceed to actual contact because the in- creasing magnetization resulting from it opposes an insur- mountable obstacle, which rises of itself as soon as the inter- polar distance is again lessened. " The approach of the carbons is therefore intermittent; but when the apparatus is well adjusted, the periods of rest and of advance succeed each other so rapidly that they are equivalent to a continuous progressive movement." Foucault does not explain how he adjusted the greater or less approach of the carbons ; it was probably by giving the VOLTAIC ARC LAMPS. pulleys over which passed the cords that moved the carbons an un- equal diameter proportional to the amount of the con- sumption. Nor does he describe the way in which the detent acted; but it seems from his description that it was merely by a simpte pressure against a drum fixed on the axle of the two pulleys, on which were wound in opposite directions the cords attached to the carbon -hold- ers. Be that as it may, this appara- tus has been the starting point for all those we are about to mention, and which require a definite place for each pole of the battery. Some years after the apparatus just 156 ELECTRIC LIGHTING. described had been invented, and after Duboscq had pointed • out the defects he had found in most of the regulators then in use, and even in those he had made himself, Foucault invented a new form which we represent in Fig. 38, and which has been hitherto generally resorted to for all experi- ments of projection. This is the model which has been con- structed by J. Duboscq. In this new system the carbon-holders B and D end below in racks on which a clockwork movement acts by means of a double wheel, which is arranged in such a manner that the two carbons advance towards each other, and the lower one D moves through twice the distance that the upper one passes through. This arrangement was governed by the un- equal consumption of the two carbons, which, as shown on page 20, is double for the positive carbon. To obtain this result, the two racks engage with the two wheels of which we have spoken, and these wheels, fixed on the same axis, have the number of their teeth in the proportion of 2 to i. To cause this arrangement to act upon the carbons, it is only necessary so to contrive matters that when the intensity of the current becomes a little too feeble an electro-magnet shall act on the clockwork, and that its action shall be stopped when, by the approach of the carbons, the voltaic arc offers less resistance. It is upon this principle that nearly all the regulators of this class have been based ; but in order to obtain a perfectly regular action, the problem to be solved was much more complicated, and the special arrangements we have now to study were necessary. The defect of the regulators founded on the principle we have explained above, was that the electro-magnet armature intended to release or to arrest the clockwork was in a state of unstable equilibrium, and therefore liable to be driven against one or the other of the two stops which limited its play without ever being able to remain in an intermediate position. In order to remedy this, the antagonistic spring R of the electro-magnet E was made, not to act directly on the VOLTAIC ARC LAMPS. 157 armature, but on the extremity P of a piece jointed to a fixed point x, and having its edge shaped to a particular curve, and in rolling pressing on a projection, which thus represents a lever of variable length, as in Robert Houdirts electric dis- tributor. The armature must then always remain thus fluc- tuating between the two limiting positions, for at each moment the antagonistic force opposed by the spring to the attraction of the electro-magnet is compensated by the effect of the lever thus arranged. Or in other words, the position of the armature every instant expresses the intensity of the current. Whilst the intensity preserves its desired value, which is co- relative to the distance maintained between the carbons, the armature is balanced in such a manner as to prevent any movement of approach or recession ; the moment the current becomes too strong or too weak, there is a recession or approach, for the lever T attached to the branch of the arma- ture lever F produces these effects by the oscillation of an escapement anchor t fixed at the end of the lever T, which engages and disengages a double clockwork mechanism, the action of which we shall now study. In this, the fly vanes o and o' act as detents. This mechanism represented on a larger scale, in Fig. 39, is set in motion by two spring barrels L L', each of which controls its own trainwork of wheels, the last piece bearing the fly vane mentioned above. The system controlled by the barrel L tends to separate, and the other to bring them together; but in order that these mechanisms working in oppo- site directions may act on the movers of the carbons, a special mechanical contrivance was necessary, and Foucault had re- course to an arrangement invented by Huyghens, consisting of two sun and planet wheels f and e, fitted to a wheel s, which rotates on the axle gh. It is this wheel which governs the motion of the double wheel acting on the racks H and D ; but it can come into action only when the wheel work in con- nection with the flies o o' is liberated by the electro-mag- netic effect and the movement of the detent /. When in con- ELECTRIC LIGHTING. sequence of the liberation of the fly o', the wheelwork c d is allowed to act, the barrel i! makes the wheel S turn in the direction of the arrow, and the two carbons are brought FIG. 39. nearer together. At the same time, the two sun and planet wheels are turned, but without producing any effect, for they revolve round the wheels h and d; but when the second fly o is liberated the planet wheel e is set in motion by the wheel a and the pignon h, and by acting through the planet VOLTAIC ARC LAMPS. 159 wheel / on the wheel d it compels the wheel S to move a little in the direction opposite to the former. But to effect this it is necessary that the barrel L should have as much power as L'. In any case the motion can only be very small. The wheels a and b must be of one piece, but they must move with gentle friction on the axle gh in order that the wheel S may turn with making them share its motion. If this much has been understood, it only remains to examine the mode of action of the apparatus. The arc being established between the two carbons, the attractive action of the electro-magnet is counterbalanced by the antagonistic spring, so that only one of the arms of the escapement anchor engages the fly o'. As the carbons con- sume, the armature F is the less attracted the more the arc is lengthened, but no sudden movement occurs ; the armature drawn away by the spring rolls on the jointed curve x, Fig. 38, and at last the hammer liberates the fly o', and engages the fly o ; the carbons are then approximated until the in- tensity of the current is sufficient to re-establish the power of the electro-magnet. If the carbons are too near, the arma- ture F is attracted the more, and the escapement anchor will set free the fly o, which will cause a separation of the carbons. Let us add, that the carbons are capable of being moved in two ways by hand in order to fix the position of the luminous point in the first instance. Thus, the upper carbon may be moved independently of the lower one, the whole system may be raised or lowered, the separation of the car- bons continuing unchanged; and this is required for pro- perly centring the light in projections. An important improvement has lately been made in this apparatus. It had been found that changes in the intensity of the current modified the condition of the magnetic core, and that the magnetic power persisted more or less. The antagonistic spring regulated by a screw, seen on the right in Fig. 38, was therefore unable for a given electric intensity to l6o ELECTRIC LIGHTING. maintain the equilibrium, and if too great a tension were given to it the movement of the apparatus became too jerky. This inconvenience has been got rid of by so arranging the armature, to which a curved form is given, as to vary its distance from the poles of the electro-magnet. This motion is produced by the friction of an eccentric lever. This slight modi- fication is shown in Fig. 40, which represents the external appearance of the apparatus. As the least variation sensibly alters the effec- tive power of the attraction, the action of the electricity may then be easily and very exactly regu- lated, according to the power of the electrical generator at a given instant. In this new model the positive pole may be placed above or below as required. The -j- pole should correspond with the upper carbon for lighting purposes, and with the lower carbon for optical experi- ments, such as the combustion of nnetals, &c. "This new regulator," says Du- FIG. 40. boscq, " fulfils all the conditions re- quired for the application of the electric light to scientific experiments, and to the illumination of lighthouses, vessels, workshops, theatres, &c. " In the present state of science the electric light is as com- monly produced with the magneto-electric machine as with the battery ; it may even be affirmed that the industrial generator of the electric light is magnetic action : witness the electric lighting of lighthouses, ships, yards, &c. It was therefore VOL TAIC ARC LAMPS. 1 0 r requisite to make the regulator suitable for these two sources of electricity. When the arc that springs between the carbons is derived from, the battery, they are consumed in the ratio of i to 2 ; if, on the other hand, it is derived from the magneto-electric machine, the consumption of both carbons is alike, since the currents are alternating. In the former case it is necessary to arrange the advance of the carbons in the ratio of I to 2, and in the latter to make it equal. An addition to the mechanism enables the change of the relative velocities of the carbons to be made in an instant, according as one or the other source of elec- tricity is made use of. " Thus improved, the new regulator is perfectly adapted to all the applications of the electric light." Serrin's Lamp. — Of all the regulators yet invented, that of Serrin is the one most used when a prolonged illumination is required. We have had the pleasure of following the various phases through which this apparatus has passed since its invention, and we were the first to give a complete account of it in tome IV. of the second edition of our Expose des applications de Felectritite, published in 1859. At a later period, Pouillet, in a report made to the Academic des Sciences, explained its ingenious arrangements. Finally, the experi- ments made with the machines of the Alliance Company showed that it was then the only apparatus that could work with currents alternately reversed. Since that period this re- gulator has been constantly used in the various experimenls that have been made with the electric light, and it is the one adopted in the illumination of lighthouses. We must there- fore dwell a little on this ingenious apparatus, which is so sensitive that an india-rubber ring placed between the two carbons is sufficient to arrest their progress without the ring losing its shape. This apparatus, which is able, like that of Duboscq, to keep the luminous point stationary, is essentially formed ol two mechanisms mutually connected, but each producing its own effect on the movement of the carbons. One of these ii 1 62 ELECTRIC LIGHTING. mechanisms, in direct communication with the electro-rnag- netic system, forms an oscillating system, composed of a kind of double jointed parallelogram, to which are attached the tube and accessories belonging to the lower carbon. It is formed of four parallel horizontal arms, turning on the tube of the upper carbon-holder and connected together by two vertical cross pieces. The other mechanism, which we shall call the advancing mechanism, is connected with the carbon -holders, and is formed of wheelwork, together with a rack and draw chain. The first mechanism, while acting, as we shall see, directly on the lower carbon-holder, controls the action of the second mechanism ; and this last, by rendering regular the ap- proach of the carbons towards each other, carries out the mechanical effect begun by the first. With this object, the tube belonging to the lower carbon and forming part of the oscillating system, carries a stop of triangular shape which acts on the wings of a fly, forming the last moving piece of the advancing mechanism. The oscillating system connected by the two vertical cross pieces carries a cylindrical arma- ture, which being placed within the influence of an electro- magnet acting tangentially on it, can more or less lower it according to the intensity of the current traversing the system. There are two opposing springs fixed to the lower arms of the oscillating system, and these springs are attached to the supports of the wheels, and raise the system when the current does not act with energy sufficient to counteract their effect. It follows therefore from this arrangement that, with an adequate electrical intensity, the oscillating system is suffi- ciently lowered to arrest the advancing system, and with an inadequate intensity this last system, being set at liberty, allows the carbons to approach each other merely by the effect of gravity on the upper carbon-holder, which is heavy enough to produce the movement of the system. Let us now see how the advancing system is arranged. In the first place, it is formed of a set of four wheels, the VOLTAIC ARC LAMPS. 163 first of which engages a rack, making part of the upper carbon- holder, and has its axle provided with a drum, round which is wound a Vaucanson chain ; this chain, after having passed over a pulley, is attached to a piece connected with the lower carbon-holder. It follows that when the oscillating system, in consequence of the inactivity of the electro-magnet, has liberated the wheel, the upper carbon-holder is free to de- scend, and in descending not only causes all the wheels to turn, but also raises, by means of the Vaucanson chain, the lower carbon-holder. This action goes on until the current, by becoming stronger, produces a more powerful action on the electro-magnet, which thereupon engages, by means of the oscillating system, the fly in the train of wheels. The ap- paratus is then stopped until the energy of the electro-magnet again diminishes. A second opposing spring, acted on by an adjusting screw, allows the sensitiveness of the instrument to be increased or diminished at pleasure. Lastly, a chain hanging down is so arranged as to act as a counterpoise to compensate, in the action of the oscillating system, for the loss of weight occasioned by the consumption of the lower carbon. The current is transmitted to the lower carbon through a bent flexible plate which follows the carbon in its move- ments, and it is transmitted to the upper carbon through the mass of the apparatus and the electro-magnet which has the free extremity of its coil connected with an external binding screw. There is nothing special about the lower carbon-holder: it is merely a socket provided with a pressure screw for hold- ing the carbon ; but the upper carbon-holder is more com- plicated, in order that it may have two rectangular move- ments so as exactly to fix the two carbons in the desired relative position. The positive carbon is maintained above the negative by means of a tube, supported by two horizontal jointed arms controlled by two screws. One of the screw? allows the carbon-holder to be displaced in one vertical II — 2 1 64 ELECTRIC LIGHTING. plane, and the other, by means of an eccentric, displaces the carbon in the vertical plane perpendicular to the former. Serrin has made several patterns of his regulator to suit the greater or less electric intensities that are to act upon it. His largest pattern is arranged to burn carbons of 15 milli- metres broad, or 225 square millimetres in section; and, in spite of these large dimensions, it is as sensitive as the smallest patterns. In the pattern made for lighthouses, the inventor has made several important modifications. Thus, by means of a little arrangement adapted to the chains of the carbon- holders, he has been able to shift the luminous point to a certain extent without putting out the light. This is very important in the application of these apparatus to lighthouses, for it gives the means of correctly centring the light with regard to the lenses. Again, as these regulators have to act with extremely powerful currents, the heat in the circuit would be sufficient to burn the .insulating covering of the coil in the electro- magnet, and thus destroy its effect. Serrin has made the electro-magnetic spirals with metallic helices unprovided with insulating covering, and so arranged that the spires cannot touch each other. In order that these helices might be adapted to the magnetic cores and to the discs of the electro- magnet with a sufficient insulation, he has covered these cores with a somewhat thick layer of vitreous enamel, as well as the inner parts of the discs ; and in order to obtain as many spires as possible with the maximum of section, he has cut these helices from a copper cylinder of a thickness equal to that of the coils. In this way the electro-magnetic helices are represented by a kind of close screw thread of a projection, equal to that of the discs, and having its central part repre- sented by the magnetic cores and the covering of enamel. It will easily be understood that with this arrangement the helices may be carried to a very high temperature without the spires ceasing to be insulated from each other, for they are not in contact, and they are separated from the body of the VOLTAIC ARC LAMPS. 165 electro-magnet by a substance which cannot be changed by the most elevated temperature, Besides, the large section of the spire,> thus formed makes the heating more difficult than with ordinary arrangements, and that is not dne of the least advantages of this kind of electro-magnet I ought, in justice, to state that previously to Serrin an electro-magnet of this kind had been invented by Duboscq for his regulator, but he had riot taken care to enamel the parts in contact with the helices, considering that precaution as useless, on account of the great section of the spires of the helix, which prevented them from being raised to redness. Nor did he form his spires in the same way: they were simply a strip of copper hammered into a spiral. Siemens' Lamp. — The last of Siemens' lamps, which is much used in England and in Germany, is represented in Fig. 41. Like that of Serrin, it can be lighted automatically, and the two opposite actions required for the separation and approach of the carbons are determined by the weight of the upper carbon-holder and by the electro-magnetic vibration of a rocking lever which acts on a clockwork mechanism driven in the opposite direction by the weight of the car- bon-holder. This mechanism, composed of four wheels, is arranged nearly as in the regulators we have just described, and it is on the last wheel i, furnished with a ratchet and a fly with wings, that the vibrating electro-magnetic acts. This last is formed of a bent lever L, jointed at Y, and carrying at M the armature of the electro-magnet E. This is the principal organ of the apparatus, for on one side it carries a contact piece which forms with the stud x the vibrating circuit- breaker, and in the second place the antagonistic spring of the system, the tension of which is regulated by means of the screw R, and finally the driving and stopping catch Q, which acts on the clockwork mechanism by means of the ratchet wheel i. A fixed piece s supports the end of this catch, in order to liberate the wheel i, at a suitable inclination of the 1 66 ELECTRIC LIGHTING. lever L. Finally, a screw K, which passes through the case of the lamp, allows the distance of the armature to be pro- perly adjusted to the current employed, and a small addition N, which also projects from the case, shows whether the electro-magnetic system is properly vibrating. The wire of the electro-magnet E is moreover connected with the mass of the apparatus, in order that the current which passes through and illuminates the carbons may be derived through the circuit-breaker at each attractive movement of the armature, and produce the vibration of the lever L by closing a short circuit. The action of this apparatus is very simple. When a current passes through the electro-magnet E, the catch Q is withdrawn from the wheel i, and the upper carbon-holder, by weighing on the wheels of the clockwork, sets them going until the carbons guided by racks engaging these wheels are brought into contact. But if under these conditions the generator is put into communication with the lamp by the binding screws z and c, the current traverses the electro-magnet E, the mass of the apparatus, the upper carbon-holder, the lower carbon- holder, and returns to the generator by the communication connecting this with the binding screw z; the carbons then glow at their point of contact, the electro-magnet becomes active, and the catch Q, by acting on the ratchet wheel I, causes it to advance one tooth, by which the carbons are separated. But in this movement a contact is set up at x between the lever and the screw c, and the current, finding less resistance in passing by that path than through the electro-magnet E, in a great measure leaves the latter ; then the armature, being no longer sufficiently attracted, causes a backward movement of the lever L, which again withdraws the catch Q, destroys the contact at x, and brings about a new attraction of the armature involving a new movement of the wheel i; and as these alternating motions are more rapidly effected than that which results from the setting in motion of the wheels by the action of the upper carbon-holder's VOLTAIC ARC LAMPS. 167 weight, the carbons are soon sufficiently apart to produce a voltaic arc of suit- able size, which increases in length in consequence of the consumption of the carbons ; but when their dis- tance apart be- comes too great, the intensity of the current becoming too weak is unable to cause the at- traction necessary for the action on the wheel i, and then the wheels can turn freely, causing thereby the approach of the carbons, which goes on until the current has re- gained an inten sity sufficient to again produce the effects we have already studied. By a suitable ad- justment of the screws R, K, and z, the double in- FlG- 4*. verse action we have just examined may be made very re- gular. But this adjustment is very delicate, and that is perhaps an inconvenience of the system. 1 68 ELECTRIC LIGHTING. The apparatus is moreover provided with two other systems of regulating screws, one of which moves the two carbons and the luminous point without extinguishing it, and the other allows one of the carbons to be displaced. Finally, screws attached to the upper carbon-holder give a means of easily fixing the carbons with regard to each other so as to supply a diffused or a condensed light. Two small bull's-eyes in the side of the lamp allow the action of the delicate parts of the mechanism to be observed, and the effects of the regu- lator to be noted. In order that the generator of light may work always under the same conditions, whatever may be the variations in the circuit external to the lamp, Siemens has interposed in the circuit a regulator of resistance which we are going to describe, as it has more importance than at first sight would be supposed ; for these variations, by changing the conditions of velocity in the motor, oa one hand, and on the other by causing many sparks, may injure the machine and the col- lector. By 1856 Lacassagne and Thiers saw the necessity of a regulator of this kind, and had invented one which I have described in my Expose des applications de V electricity t. V., p. 506, and which was an accessory of their electric lamp ; but these systems had been little used before Siemens. Siemens' arrangement consists of an electro-magnet with a thick wire, the armature of which acts in the manner of\ relay on a contact which, when the armature is not attracted, has the effect of introducing into the circuit a derivation with a resistance nearly equal to that of the voltaic arc. With a suitable regulation of the antagonistic spring, the derivation is therefore substituted for the voltaic arc whenever the re- sistance of the latter becomes so great that the armature is no longer retained. This is what happens not only when the lamp is put out or withdrawn from the circuit, but also when very great variations occur in the working of the lamp. The helix forming the derivation is placed in a tin vessel filled with water to prevent the wire from becoming too hot VOLTAIC ARC LAMPS. 169 during long interruptions of the current, such as those re- quired by replacing the carbons. The lamp just described is not, however, the only one made by Siemens. He has already patented eight forms. Lout ill's Lamp. — From a published account of the Lontin machines we extract the following description of this lamp : — " The first and principal advantage of this regulator is that its moving and regulating parts are such that it can work in any position, upright, horizontal, or upside-down, and without being stopped or changed in its action by even the strongest shocks or oscillations. " In these regulators a quite new application has been made of the heating produced by the current in a metallic wire to cause the separation of the carbons and keep it perfectly constant. Thus the use of electro-magnets is dispensed with, and the con- sequent cost of the additional electric power required to over- come their resistance in the circuit, while the length of the arc remains absolutely fixed, so that a more regular light may be obtained. " The approach of the carbons in proportion to the combustion is obtained by another not less happy application of a derived current taken from the light current itself, and acting as follows: — " There is a solenoid in the apparatus formed of a coil of fine wire in quantity sufficient to offer a very great resistance to the current. This coil encloses a movable iron rod, which when at rest keeps back the moving power that brings the carbons nearer together. So long as the carbons are at a distance adjusted to the amount required for a good light, the whole current passes through them, on account of the great resistance of the coil; but when the separation increases, a small portion of the current passes through the coil and excites it so that the movable iron rod is attractive, and the moving power released from its stop brings the carbons nearer by the amount required for maintain- ing the length of the arc : at this moment the solenoid ceases to act, and the irod rod again stops the motor, which has merely to bring the carbons nearer together and is extremely simple." This employment of a derivation from the light current 1 70 ELECTRIC LIGHTING. may also be advantageously applied to all the regulators which automatically separate the carbons, and it renders their action more certain and regular, whatever may be the variations in the intensity of the current. This system has been successfully applied to the Serrin regulators, used at the Chemin defer de r Quest (St. Lazare station). De Mersaime's Lamp. — De Mersanne's regulator was invented to enable straight carbons to supply an electric light for at least sixteen hours consecutively. The system, represented in Fig. 42, is essentially formed of two slide boxes B B', fixed on a strong upright stand of cast iron. Through these slide the two cylindrical carbons c and c', each 75 centimetres or more in length, moved by a re- gulated action. In order that the carbons may be adjusted to have their points in the same vertical line, the boxes are capable of turning a little vertically, and the upper one can also be turned horizontally. The sliding system of both boxes consists of four grooved rollers, two of which are con- nected with the two ends of a lever, and are pressed against the carbons by a spiral spring v. These rollers serve as guides, and the other two, of larger diameter and with a roughened surface, act as the movers of the carbons. For this purpose these are set in motion by wheels fixed on an axle, connected by mitre wheels with a vertical shaft A A. This shaft, being capable of turning in either direction, ac- cording to the action of the regulating apparatus, can make the carbons approach to, or recede from, each other. The carbons are supported and protected outside of the boxes by enclosing tubes. The regulating apparatus is placed in a case below the slide box B' of the lower carbon. It consists in the first place of clockwork driven by a spring barrel,* and by an. electro-magnet E forming part of a derived circuit from the * This barrel, when wound up, will act for 36 hours without attention, VOLTAIC ARC LAMPS. 17* two carbons, as in Lontin's system ; and in the second place, of another electro-magnet M interposed in the same deriva- tion, and acting on the box B' of the lower carbon in such a manner as to separate the car- bons when they come into con- tact. When the apparatus is not in action, the carbons are generally separated by a greater or less interval; but as soon as the circuit is closed through the apparatus, the two electro- magnets are excited, for the current then wholly passes through the derivation, and the clockwork is liberated, while the lower holder is so inclined as exactly to bring the two carbons one over the other. The advance of the carbons takes place slowly, and when they 'come into contact, the current, finding a more direct path, abandons the derivation and the electro-magnets, and passes almost entirely through the circuit of the carbons, which then glow, and imme- diately supply the voltaic arc. The electro-magnet M having become inactive, the box B' of the lower carbon-holder is slightly inclined forward, and by this causes not only the disjunction of the carbons, but also a sufficient separation of their points in consequence of the action produced on the FIG. 42. 1 7 2 ELECTRIC LIGHTING. wheels by the movement of the box. The lamp is thus lighted; but in proportion to the consumption of the carbons, the resistance of the light circuit increases, and the current, passing with more intensity into the derivation, soon becomes sufficiently powerful to release the clockwork that brings the carbons together, until the current has regained its full inten- sity in the light circuit. Things go on in this way until the carbons are entirely consumed. It will easily be understood that with this arrangement there is no limit to the length of the carbons, since it may exceed that of the apparatus, and without their requiring any particular position. Their forward movement takes place as if they glided between the fingers of two hands, pushed towards each other by two thumbs guiding their progress. This lamp, like those of Serrin and Siemens, may be lighted from a distance, which is not one of the least of its advantages. A medal was awarded to it at the Exposition Universelle of 1878. Biirgiu's Lamp. — This lamp is already rather old. and we are surprised that it has not been described anywhere, for, according to what Soret has communicated to me, it works in the most satisfactory manner. It was shown at the Exposition of 1878, but, its inventor being absent when the jury went round, it was not examined. It is used con- tinuously at Geneva for the theatre and for the illumination of a public clock. Biirgin has made two patterns of this lamp, one of which is for industrial purposes, and is represented in Fig. 43 ; the other, more complicated, is intended for scientific experi- ments. The principle of the lamp is very simple : the two carbon-holders tend to approach each other continuously by the influence of a spring barrel or a counterpoise; but they can obey this tendency only when a check controlled by an electro-magnetic action permits the passage of the chain or chains which support the carbon-holders, so that, according VOLTAIC ARC LAMPS, 173 as the current is more or less energetic, there is motion or rest in these carbon-holders. In the pattern shown in Fig. 43, this result is obtained by means of a large wheel R, which carries on its axle the pulley c, on which is wound the chain supporting the lower carbon -holder. The axle of the wheel rests on a piece of iron A A fixed to a jointed paral- lelogram, and serving as the armature of an electro-magnet E con- nected with the light circuit. A spring break F presses on the circumference of this wheel, and is suffi- ciently bent to prevent the wheel from turn- ing when the latter is at the proper height, that is to say, when the armature A is at its greatest approximation to the electro-magnet E ; but when, in conse- quence of the weaken- ing of the current, this armature is at a greater distance, the wheel, by dropping with the ar- mature, withdraws from the break and is then able to turn by the effect of the weight of the lower carbon (or of a spring barrel attached to this carbon-holder), acting through the chain which is wound about the pulley c. Thereupon the lower carbon-holder rises, and the current, resuming its FIG. 43. 1 7 4 RLE C 7R1C LIGHTING. energy, quickly produces a fresh engagement of the wheel, which checks the rise of the carbon at the proper instant. The slight movement of the attracted armature A, when the carbons are in contact, suffices to allow the passage of enough chain to bring about the separation of the carbons when the circuit is closed. In this pattern the upper carbon is fixed, and therefore the luminous point changes its position, a thing of no conse- quence for ordinary illumination; but for experiments of projection, the two carbons must be so arranged as to move simultaneously in the proportion of 2 to i, and for this pur- pose Biirgin fixes the two carbon-holders to two chains, which are wound upon two pulleys of unequal diameter, mounted on the axis of the large regulating wheel, so that each move- ment of that wheel causes a double displacement of the car- bons. An adjusting screw attached to the break allows the apparatus to be made more or less sensitive. In this pattern it is the weight of the upper carbon-holder which, as in Serrin's regulator, brings the carbons together, and the attrac- tive action of the electro-magnet first determines their sepa- ration in order that an arc may be formed, and afterwards .stops them so as to maintain their due interpolar distance. Gaiffe's Lamp. — In 1850, Archereau, reflecting upon the considerable space that a rod of soft iron will move within a coil under the influence of the magnetic attractions, invented A regulator based upon that principle. He formed one of the carbon-holders of a rod, half of copper and half of iron, placed within a long coil, and to properly balance the attrac- tive force he used a counterpoise. This was, therefore, one of the simplest of regulators, and it had the advantage of being capable of lighting at a distance. In the hands of .skilful experimenters, it could work well, but the movements being too abrupt and the oscillation too large, it often went •out, and was not in fact a practical lamp. Jaspar and Loiseau succeeded in lessening these defects, but it was not VOLTAIC ARC LAMPS. 17$ FIG. 44. until Gaiffe invented the regulators represented in Fig, 44 that the advantage of this plan became obvious. 1 7 6 ELECTRIC LIGHTING. In Gaiffe's Lamp the two carbon-holders H, H' are movable, as in Foucault's and in Serrin's arrangements, and they are so arranged as to keep the luminous point stationary. For this purpose their motion is controlled by two racks K, u, which engage two wheels M', o, of unequal diameter, and moved by a simple spring barrel on the axle of which they are fixed. This barrel is wound up by the mere separation of the carbon-holders, which are, however, perfectly balanced and turn between sets of rollers. The lower carbon-holder is terminated by an iron rod K, to which is attached a rack, and this rod is placed within an electro-magnetic bobbin L, with a coil increasing in diameter from its upper to its middle part in order to compensate for the unequal action of the spring barrel through the whole range of this movement. Finally, a small wheel R connected by the wheels M', o, and another to the wheel M gives the means of simultaneously act- ing on the two racks so as to raise or lower the luminous point. In the normal state the carbons touch each other, and when the current traversing them excites the coil, the lower carbon-holder is lowered at the same time that the upper one HVI is raised, and this effect is continued until the attractive force of the coil balances the resistance of the spring barrel, and thus a voltaic arc is produced. Of course the length of the arc depends upon the tension of the spring in the barrel, and this can be regulated by a screw. So long as the arc remains under the same conditions of resistance, the effect is maintained ; but the moment the resistance increases on account of the consumption of the carbons, the power of the spring prevails over the electro-magnetic action, and the car- bons are brought nearer together, until a condition of equili- brium is again attained, and the same thing goes on until the carbons are entirely consumed. The small mechanism fitted to the wheels of the racks allows the luminous point to be raised or lowered, by means of a key, without the lamp going out, and this is necessary in optical experiments in order to properly centre the light. VOLTAIC ARC LA MIS. 177 Carre's Lamp. — Carre's Lamp, which received a gold medal at the Exposition of 1878, and is represented in Fig. 45, is only an ingenious improvement on Ar- chereau's and Gaiffe's regulators. As in those, the electro-magnetic ac- tion is founded on the attractive effects of sole noids, but these effects are by an ingenious ar- rangement very much magnified, and the me- chanical action is pro duced, as in the regula- tors of Serrin, Foucault, £c., by clockwork acting on two racks D, E, fixed to the carbon-holders, and controlled by a detent brought in action by the electro-magnetic system. This system is formed of two coils B, B', having their axes slightly curved, and into these pass the ends of a soft iron core A A bent into an S shape, and turning at its middle part about the centre c. A double set of antago- nistic springs ;-, ;-', regu- FlG ^ lated by a system de- pending upon the adjusting screw v, allows the force op- posing the attraction of the coils to be suitably adjusted, and 12 r/8 ELECTRIC LIGHTING. a rod / fixed to the magnetic core acts on the detent of the clockwork. The motion of the wheels causes a forward movement of the two racks in the proper proportion for keeping the point of light stationary. The light current passes through the two coils, and according as its intensity is greater or less, the iron core is more or less attracted within the coils, a sufficient enfeeblement causing a release of the detent, whereupon the approach of the carbons ensues. In this system, as in those of Archereau, Gaiffe, Jaspar, Loiseau, £c., the increase of the current has the effect of separating the carbons from each other, and the clockwork brings them nearer together ; but as under these conditions the path of the movable piece in the electro-magnetic system is considerable, and as the attractive effect is much less sudden than in the case of electro-magnets with turning armatures, the separations of the carbons are produced freely and without oscillations, which is an advantage.* These are the regulators which, during the whole duration of the Exposition, have worked with the machines of the Alliance Company. With alternately reversed currents, they offer real advantages, for J. Van Malderen has shown that with these currents an iron rod is almost as strongly drawn into a coil as with direct currents, only there is a much greater heating of the electro-magnetic system ; but to com- pensate for this there is much less residual magnetism, and therefore greater sensitiveness. Brush's Lamp. — The report of the American Commis- sion appointed to examine magneto-electric machines having bestowed great praise on this lamp, we have considered it our duty to give a description of it here, although it appears to us inferior to those we have in France. This lamp, which is represented in Fig. 46, is, like the pre- ceding, based on the attraction of solenoids. The electro- * See the laws of the attractions of solenoids in tome II. of my Expose des applications de rtlectriciU, p. 132. VOLTAIC ARC LAMPS. I79 magnetic coil is at A, above the upper carbon-holder, and is supported by the arm b fixed to a vertical piece c sliding within a column so as to adapt the apparatus to different lengths of carbons. Inside of the coil a magnetic core d moves freely. This core is hollow, and through it passes the copper rod // of the upper carbon- holder, sliding freely for its whole length. A sort of collar h grasps it, however, a little below the magnetic core, and is so arranged that when it presses on a cross piece //, forming .part of the fixed system of the arm b> it leaves the rod // free, which then de- scends by its own weight. The collar therefore maintains the rod only when it is itself raised up, and this takes place almost constantly during the working of the appa- ratus, for a catch e fixed on the magnetic core supports it above; but in its rise it cannot go beyond a certain limit which can be adjusted, for a screw .v has a head large enough to hold it by its upper edge. The magnetic core itself is kept up by a cross piece on which act two spiral springs, the rods of which serve at the same time as guides. The carbon-holders have no particular feature ; one is fixed to rod // the other to an 12 — 2 FIG. 46. l8o ELECTRIC LIGHTING. arm fitted to the supporting column, and this last is arranged in such a manner as to be capable of moving when the con- sumption of the carbon requires ; for in this system it is only the course of the upper carbon that is regulated electro- magnetically, as in Archereau's original regulator. We must add that the coil A is formed of two helices which may, by means of a commutator shown at the top of the coil, be arranged either for tension or for quantity, according to the conditions of the experiment. The action of this apparatus will be readily understood: when it is not working, the two carbons k k are in contact, and the current can pass through them when the carbon- holders are connected with the electric generator. Under the influence of the current, then at its maximum intensity, the magnetic core d is raised up, carrying with it, by its catch,, the collar h ; the rod //is then raised, and the two carbons separated ; the voltaic arc is produced, and, so long as the electric action is kept within proper limits, the apparatus re- mains in the condition brought about by the rise of the core d i but when the consumption of the carbons becomes sufficiently great to notably weaken the current, the core d drops down again, and with it the iron rod // and the collar h. If this descent is not complete, the carbons ap- proach each other only by the distance the core has dropped ; but should it be so great that the collar h presses on the cross piece //, the rod // is liberated and falls by its own weight until the approach of the carbons is sufficiently great to bring about a fresh ascent of the core //. Ja spar's Lamp. — Jaspar, of Liege, was one of the first to turn his attention to electric lamps, and in the second edition of our Expose des applications de r electricity published in 1856, we have described the first system he sent to the Exposition of 1855. Since that time he has paid little atten- tion to the subject, and it was not until the Exposition of 1878 that he again entered the lists with other inventors of VOLTAIC ARC LAMPS. l8l lamps and obtained a gold medal. This honour compels us to give some details of this new apparatus, which, it appears, acts remarkably well. The apparatus is, like the preceding, based upon the attraction of solenoids acting directly on the carbon-holders by means of chains passing over pulleys of unequal diameter, and this action can be regulated by a counterpoise which acts on the antagonistic force.' But the improvement in the original apparatus consists in this : the lower carbon acts on a rod with a piston which moves freely in a tube filled with mercury, and produces two useful effects. In the first place, it prevents any abrupt movement of the carbon holders, for the mercury being able to pass but slowly through the narrow space between the piston and the interior surface of the tube, resists a too rapid rise or descent ; in the second place, it affords an excellent contact for the negative rod. The small counterpoise serving as antagonistic force has also a happy arrangement. It slides on a small rod jointed horizontally, with one end connected with the free extremity of the chain of the lower carbon-holder ; so that, by moving backward or forward an externally projecting stud, its effect may be increased or diminished. Experiments made with this lamp have proved very satis- factory. It is now applied at Sautter and Lemonnier's, in the workshops of Cockerill and Seraing, in Denayer's paper manufactory at Willebroek, at the Gare du Midi at Brus- sells, &c. Rapieff's Lamp. — This system is only an extension of Bailhache's, in which the carbons are kept always at the same relative distance in spite of their consumption, by means of a spring pressing on them like the candle springs uf carriage lamps. In Bailhache's system the carbons were kept at a distance suitable for the formation of the arc, by two hollow cones of calcined magnesia in which their points were placed, and which formed a kind of stop-collar. As 1 82 ELECTRIC LIGHTING. their extremities burned away, they were pushed forward by the springs, and as they became thin by burning at the ex- tremities only, the cold part always remained within the re- fractory cone. In Rapieff's system the same effect is pro- duced, but there are no refractory cones, these being replaced by four sticks of carbon joined at the ends two by two at an acute angle, and so arranged one above the other as to form two angular systems in planes perpendicular to each other. The points of contact which form the electrodes are sepa- rated one from the other by the distance desired for forming the voltaic arc. Counterpoises with pulleys are arranged to push the carbons of each system one against the other, and then, as these carbons are consumed, they constantly advance towards the common point of intersection, which remains always at the same place. Fig. 47 represents this regulating system, the carbons of Avhich, a a', b b\ seem to form an X, with this difference, that the two lower carbons are placed in a plane perpendicular to that in which the upper ones are placed. The voltaic arc is produced at c, between the upper and the lower pairs of points. As the carbons are consumed, they slowly move nearer to each other in each couple, under the influence of a counterpoise w, which, through the cords and the pulleys VJ f h d a! a e g d' b' b, pushes the sticks of carbon against each other. This counterpoise is guided in its course by two columns s s', which at the same time serve as conductors for transmitting the current to the two systems of electrodes a a', b b', supported also by the two metallic arms d h, d' g. As the upper carbons are to be connected with the positive pole, they fire of course longer than the lower. This arrange- ment is completed by an electro-magnetic system enclosed in the base of the apparatus ; and its office is, when the current passes and the four sticks of carbon are in contact, to sepa- rate the two systems to the distance necessary for the for- mation of the arc. This effect is produced by means of a cord attached to the electro-magnetic armature, which cord, VOLTAIC ARC LAMPS. 183 FIG. 47- passing within the column s, acts on the arm cF g. A re- flector of a cup shape, either of silvered copper or of porce- lain, is fixed a little above the point of contact of the carbons, 184 ELECTRIC LIGHTING. . and adjusting screws enable the luminous beam to be turned in any required direction. With carbons 20 inches long and 5 millimetres in diameter, the light supplied by the lamp lasts, according to Rapieff, for seven or eight hours ; but with a diameter of 6 millimetres these carbons may last two hours longer. This light may be reckoned at IOD or 120 gas-jets, or at 1,000 candles; but with Rapieff's smaller patterns one may be obtained not ex- ceeding 5 gas-jets. The inventor has also constructed patterns in which the preceding arrangement is reversed, in order that they may be hung from the ceiling. According to him, the resistance of the arc does not exceed 3 ohms, or 300 metres. In another arrangement Rapieff combines the action of the voltaic arc with the luminous effect of a piece of kaolin placed above the arc. The four carbons are then arranged so as to form the four edges of a pyramid broken off at its summit, where a kind of bell of kaolin is placed, like the extinguisher of a lamp, and this when glowing increases the illuminating power, according to Rapieff, by 40 per cent. The carbons are Carre's, and a Gramme machine is the generator. With this machine 10 lamps of the first described pattern may be lighted by placing them in the same circuit, but only 6 are used in the Times office at London, where this mode of lighting has been in use for some time. According to the Telegraphic Journal of the ist Novem- ber, 1878, Rapieff's system of electric lighting, introduced into England by E. J. Reed, under the direction of Apple- garth, has given excellent results in the trials which have been made at Smithfield and at the establishment of the London Times. The compositors' room is now lighted by 1 8 Rapieff lamps, and 6 are appropriated to lighting the offices. "The great advantage of this lamp," he says, "is that it will go a whole night without a renewal of the car- bons ; its intensity is always constant, even when the carbons are burned very low, and in this respect the lamp is prefer- VOLTAIC ARC LAMPS. 185 able to the Jablochkoff candle, for in this the current in- creases in energy as the candle burns on account of the decrease in the length of carbon the current has to traverse, whilst in the Rapieff lamp that length is always the same. In order that the extinction of one lamp may not cause that of other lamps, Rapieff arranges the electro-magnet that separates the carbons in such a manner that it acts as a com- mutator. When the current • passes through the electro- magnet, the commutator is not brought into action, and the circuit is completed through the lamp ; but- when this is extin- guished or withdrawn from the circuit, the electro-magnet in question becoming inactive causes the current to pass through a derivation on which a resistance equal to that of the lamp circuit is introduced, and the circuit of the other lamps is not interrupted thereby. This effect is obtained by means of a second armature, which, being attracted when the current passes, acts as a keeper, thus increasing the electro-mag- netic action exercised on the lamp, and sets the commutator in action, when the current no longer passing, the armature yields to the antagonistic action. In a new pattern, Rapieff has replaced the upper carbons of the regulator we have described by a large piece of carbon, which, as in Werdermann's lamp, does not burn. This arrangement had, however, been indicated by Rapieff in his patent of 1877, so that he cannot be accused of having imitated Werdermann in his new arrangement. It is this form that is now most employed. Baro's Lamp. — This apparatus is composed simply of two metallic tubes placed vertically one beside the other, and in these slide freely two sticks of carbon, which rest on a block of magnesia. These tubes are separated by an insu- lating substance, but a screw regulates the distance between the carbons at the end in contact with the magnesia, so that the arc may be produced at this point of contact under the desired conditions, and be kept there in spite of the con- £86 ELECTRIC LIGHTING. sumption of the carbons, for these continually tend to ap- proach the block of magnesia by their own weight. A lamp of this kind had previously been invented by Staite. Wallace Farmer's Lamp. — This system is so primitive that we are surprised that it has received any attention. It was invented with a view to obtaining a longer duration of the carbons, and with this object thin plates of carbon are used instead of carbon pencils. These plates are placed so as to form a very small angle with each other, so that the arc may be progressively displaced as the carbons are hollowed out. An electro-magnetic mechanism analagous to those of the other lamps brings the plates nearer together as the action proceeds, and this mechanism acts by variations in the inten- sity of the current. It is certain that a lamp on this system will remain lighted for a long time ; but we very much doubt whether the electric intensity it produces can be compared with that of other lamps, for the incandescent part of the carbons is then considerably reduced in extent and calorific intensity. Houston and Thomson's Lamp.— As regards arrange- ment, Houston and Thomson's Lamp resembles the ordinary apparatus ; only the lower carbon, instead of being connected with a clockwork mechanism, is supported by an arm fixed to a strong flat spring. This arm carries the armature of an electro-magnet placed below it, and is insulated metallically from the upper carbon-holder, which is placed in communi- cation with the positive pole of the electric generator. The electro-magnet is in metallic communication with the lower carbon and the negative pole of the generator, so that the circuit is completed through the elastic arm of the lower carbon-holder and the two carbons. The electro-magnetic armature forms therefore a trembler, as in an electric bell, and the effect is a very rapid series of interruptions of the current between the two carbons; and this occasions repeated sparks, VOLTAIC ARC LAMPS. I7 and thence a continuous light by their superposition on the organ of sight. In order to avoid a continuance of the action on the carbon-holders when the carbons are consumed, the upper carbon-holder has a head, which by meeting a circuit- breaker stops the current through the apparatus. Under these conditions, the spark of the extra current of the electro- magnetic system is joined to that of the generator, and increases its brilliancy. Houston and Thomson do not, of course, announce this regulator as suitable for feeble currents and lights. Maiche sought the same object, and Lemolt also used this means to produce a voltaic arc with powerful generators, only it was a clockwork mechanism which gave the vibrating movements to the carbons. Hlolera and Cebriaii's Lamp. — This system of lamp was invented for distributing the light in several directions by optical means. The apparatus consists, therefore, of a regulator, with the luminous point enclosed in a polyhedric cage, with faces provided with Fresnel's lenses. With this object, the two carbons are supported by two systems of regulating carbon-holders entering at a certain angle within the cage, and directed towards its centre. These carbon- holders are fitted to the two ends of a tube filled with a liquid, in which they form a kind of pistons which may be pushed to a greater or less distance according to the pres- sure exercised by the liquid in the tube. At one part of this tube there is a kind of bellows, the movable part of which is provided with a plate of iron, which serves as the armature of an electro-magnet placed in the arc circuit. This arma- ture being more or less energetically attracted, according to the greater or less intensity of the light current, exercises a greater or less pressure which causes the carbons to advance proportionally to their consumption, and therefore maintains them in the centre of the cage. This pressure may, how- ever, be regulated by means of an antagonistic screw. When J88 ELECTRIC LIGHTING. the current is interrupted or weakened, the spring acts so as to bring the carbons nearer together; when, on the contrary, it increases, the inverse effect is produced and the carbons are separated. Lacassagne and Thiers had, as will be pre- sently seen, used a system of the same kind. On each face of the polyhedron, formed by the reflexion cage, abut tubes which are arranged so as to carry the rays projected by the lenses in certain directions ; and according to the Scientific American, these tubes may be placed in the streets, along buildings, or even under the floors, in order to distribute the light in a house. For that purpose it would suffice to fix a stout branch pipe with a properly inclined mirror at each change of direction. By means of this system, says the American journal, each street in a town may be provided with one or more pipes carrying a certain quantity of light, which may always be governed by simply changing the position of the reflectors. The problem may thus be reduced to the conditions of a water supply. Without stopping to discuss these rather fantastic dreams of American inventors, we think we ought to give here some figures they have published relative to the system of lamp we have described. According to these, 195 lamps on this system can be supplied by i-horse power, and a light obtained equivalent to 1,958 candles, for which it follows that this illumination would cost only one-twentieth the price of gas. It is needless to say that we cannot accept such figures, and if we give them, it is only to show to what length the prolific imaginations of Americans will go. (See the Tele- graphic Journal of i5th July, 1879, p. 231.) Lamps of various kinds. — In order to complete our monograph of the voltaic arc lamp, it remains only to speak of various plans which, although they have not given very important results in practice, nevertheless have an originality of character which deserves attention. Of this number is VOLTAIC ARC LAMPS. 189 Girouard's lamp, which is a clockwork regulator somewhat re- sembling that of Foucault or Serrin,but acting under the power of a kind of relay-regulator, and therefore capable of being controlled from a distance. The system consists, therefore, of two apparatus: i°. An electro-magnetic coil with thick wire, through which the current of the generator passes, and which acts on a double contact ; 2°. A lamp with a double clock- work motion, on which act two electro-magnets with fine wires, excited by the current of a battery of very feeble intensity. The current of the generator, after having traversed the electro-magnet of the relay-regulator, passes directly, there- fore, through the carbons of the lamp, and their distance apart is regulated by an independent system which works by the influence of the two contacts of the relay-regulator. When the current has its full intensity, the mechanism con- trolling the separation of the carbons is set in motion, and when on the contrary the current is too much enfeebled, the .second clockwork mechanism is liberated, and this brings the carbons nearer together. In order to obtain this double action in opposite directions, Girouard used a spring barrel with a double movement. The complete description of this system may be found in vol. V. of my Expose des applica- tions de V clectridte, p. 495. In order to increase the duration of the carbons, circular carbons have been used. Indeed, the first electric lamp was made by Thomas Wright on this system. Subsequently, in 1849, Lemolt, taking up Wright's idea, constructed a better apparatus, in which the two discs of carbon, supported by two curved and jointed levers, were set in motion by a double system of pulleys driven by the clockwork mechanism. A spiral spring connecting the two curved levers, pressed the two carbon discs against each other, and these were separated at very short intervals by the action of an eccen- tric put in motion by clockwork. The result was a series of sparks succeeding each other rapidly enough to impress the sight as a continuous light. After this system came that of 190 ELECTRIC LIGHTING. Harisson, in which one of the carbons was replaced by a carbon cylinder, movable on its axis, so as to make the con- sumption slower. The upper carbon was moved by an electro-magnetic system, which determined the formation of the arc and kept it con- stant by means anala- gous to those employed in other regulators. Ducretet has improv- ed this lamp by arrang- ing it so as to give simultaneously with the arc powerful incandes- cent effects. Fig. 48 represents this new ar- rangement, which, I am told, has given very good results. Lastly came Reynier's system, the most com- plete of all, in which each of the carbon discs was set in motion sepa- rately by a special clock- work mechanism, and the separation necessary for the production of the arc was obtained by an electro-magnetic system acting on one of the carbon-holders, and producing effects like those in the other regulators of this kind. (See my Expose, t. V.,p. 502.) Besides the apparatus of which we have just spoken, there exists a class of electric lamps to which, in my work on the applications of electricity, I have given the name of regu- lators with hydrostatic actions, and which are at least very interesting, though not very practical. For the regulating FIG. 48. VOLTAIC ARC LAMPS. 191 organs, they have liquids which act either by communicating vessels, or by serving as the medium of discharge under certain conditions, or by producing an effect like that of the moderator lamps. The principal forms in this class are those of Lacassagne and Thiers, Pascal of Lyons, Margais and Duboscq, and Way. In Lacassagne and Thiers' apparatus, invented in 1856, only the lower carbon is movable, and it is directed by a float fitted in a long cylinder filled with mercury, which is placed in communication, by a tube, with a reservoir of the same liquid, and this last is fixed on the column that supports the upper carbon-holder. The tube that establishes the communication between the two vessels is bent across one of the poles of a powerful electro-magnet so as to present its curvature beneath the armature, and from this arrange- ment it results that the armature/pressing on the part of the tube when the current has its whole intensity, plays the part of a stopper, So long, therefore, as the current keeps all its power, the level remains the same in the two vessels, and the lower carbon continues motionless ; but when the current becomes feeble by the increase of the length of the arc, the tube is then liberated a little, and a small quantity of the liquid passes from the reservoir into the tube, and this causes the carbon to rise until the current, having resumed its original intensity, has again occasioned the obstruction of the tube. The action of the spring antagonistic to the electro-magnet is regulated by a second electro-magnet placed in a very resisting derivation of the circuit, and which acts in the same direction as it would on the armature of the large electro-magnet, which is for this purpose prolonged beyond its pivot. Margais and Duboscq's regulator is a kind of moderator lamp, the rack of which acts on the two carbon-holders by means of a double pinion, and in which the movement of the piston depends on the more or less rapid flow of the oil below it. For this purpose, the upper and lower parts of the 192 ELECTRIC LIGHTING. body of the lamp are connected by a tube, at one part of which is a circular opening stopped by an extensible mem- brane, and above this membrane a kind of plug is applied controlled by the armature of an electro-magnet, which, as in the foregoing system, stops or allows the flow of the liquid according to the greater or less intensity of the current. In Way's lamp, which was much talked about in 1856, and which moreover caused its inventor's death, the carbons between which the voltaic arc is generally produced were replaced by a tine stream of mercury issuing from a small funnel and received into an iron basin also containing mer- cury. The two poles of the generator being connected, one with the funnel and the other with the basin, a series of voltaic arcs were produced between the successive drops of the discontinuous stream, and the combination of these arcs formed a source of light tolerably brilliant and uniform. The luminous vein was enclosed by a glass tube sufficiently narrow to become so hot that the mercury should not condense, on its surface ; and as the action took place out of contact with oxygen, the mercury was not oxidized. Way modified this first arrangement by using two jets of mercury instead of a single one, and these jets were so arranged that they met together at one point, from which they flowed on in drops. He also closed and interrupted the currents continually by means of a small electro-motor set in motion by the battery, and which drove the mercury pump that supplied the jets. But in spite of these improvements, it was necessary to abandon this apparatus on account of the mercurial vapours that escaped from it, and which at length killed the inventor. Moreover, the light attained to little more than a third of that produced by the same current between two carbon points. As a special form of arc lamp, I must say a few words about J. Van Malderen's regulator, which is based upon the repulsions between the contiguous elements of one and the same current. It is a kind of suspended compass, the VOLTAIC ARC LAMPS. 193 jointed arms of which carry at their ends the carbon-holders, which are therefore horizontally opposite to each other. These two arms of the compass are insulated from each other and are very readily movable. They are connected with the two branches of the circuit, and when the carbons come into contact by the tendency of the carbon-holders to place themselves vertically, the passage of the current which then takes place produces a repulsion that separates the carbons and produces the arc. There is then established a state of equilibrium between this repulsive force and that due to gravity. This state of equilibrium is sufficiently stable to render the arc comparatively steady. But this system can be used only with currents of small intensity, and it cannot be considered of any practical use. The same may. be said of that of Fernet, which is arranged according to the same principle. In conclusion, we shall mention a lamp invented by Dubos, which is rather interesting from its arrangement. In this lamp the luminous point remains fixed without any clockwork, in consequence of the shape of the two carbons, which are semicircular. These carbons are supported by two arms jointed to a pivot in the centre of the circumfer- ence formed by the two carbons. As in the ordinary regu- lators, an electro-magnetic mechanism fixed to the carbon- holders brings them together as they are consumed. As these displacements take place circularly, the point of contact of the two remains always at the same height, and consequently the same is the case with the luminous point. There are yet a few other lamps which were shown at the Exhibition at the Albert Hall, and which have been described in various works, in those of Fontaine and Higs, among the rest ; but these lamps are merely more or less complicated modifications of those we have described, and therefore we shall here merely make a note of them. Of these, we shall i94 ELECTRIC LIGHTING. mention the lamps of Regnard, Hiram Maxim, H. Fontaine, Marcus, Crompton, Hackley, Krupp or rather Dornfeld, Chertemps, &c. This last is little more than Archereau's regulator improved, but it has the advantage of being cheap and of working in a tolerably steady manner. It was shown at the Exposition of 1878, and we give a representation of it in Fig. 49. Dornfeld's system resembles Biirgin's, that of Hackley is like Van Malderen's. Crompton's lamp is only a Serrin regulator, in which the carbons may be placed above or below the case containing the mechanism, and thanks to this arrangement the lamp can simultaneously give two luminous points. Finally, H. Fontaine's and Hiram Maxim's lamps much resemble Serrin's lamp, but have their electro-mag- netic systems a little different. These last two apparatus are re- presented and described in the new edition of Fontaine's book, as is also that of Marcus, which FlG 49 is very like Gaiffe's. Regnard's lamp is only a complication of De Baillache's, and makes use of very slender carbons, and is therefore of little practical application. INCANDESCENT LAMPS. 195 INCANDESCENT LAMPS. The interesting experiments undertaken by Lodyguine and Kosloff prompted several inventors to contrive lamps for obtaining the electric light 'by the incandescence of the carbons. It seems, however, according to Fontaine, that King, as early as 1845, nac^ invented the first lamp of that kind.* King's and Lodyguine's Lamps. — King's lamp consists of a slender rod of retort carbon fixed at its ends into two carbon cubes, and supported by a stand with two porcelain arms. The whole is enclosed in an exhausted tube, and the rigid conductors traversing this tube interpose the little car- bon rod in the circuit of the electric generator, by which it is made to glow sufficiently to give a brilliant light. This is, it will be seen, a system somewhat resembling that of Lodyguine and Kosloff, mentioned on page 144. This idea was taken up in 1846 by Greener and Staite, and in 1849 by Petrie. "Illumination by incandescence," says Fontaine, " and the principle of its production, had long been forgotten, when, in 1873, a Russian physicist, Lody- guine, resuscitated both, and made a small lamp, which was subsequently improved by Konn and Bouliguine. In his lamp Lodyguine used carbons in one piece, diminish- ing the section at the luminous centre, and in the same apparatus he placed another carbon, through which, by means of a commutator, the current could be sent when the first carbon was consumed. Kosloff, who came to France ex- pecting to work Lodyguine's patent, effected some little im- provement in this lamp, without, however, getting at anything * It is stated that this lamp was invented by J. W. Starr. (Sec the Telegraphic Journal <£ the ist January, 1879, p. 7 and 15.) 11 2 196 ELECTRIC LIGHTING. passable. A relative of True, the Paris lampmaker, at whose house Kosloff's experiments were made, also worked at this lamp with much enthusiasm without effecting any material improvements ; and it was not until Konn had invented his lamp in 1875 tnat ^ was possible to enter upon experiments which held out any promise of practical advantages from lamps of this kind. It was Duboscq who first made this lamp in France. It o icn's Lamp. — In this apparatus, which is represented in Fig. 50, each luminous centre, instead of having only one carbon, was provided with four or five, and all these carbons A B, arranged vertically and circularly, were terminated by small carbon cylinders in which were incorporated, in the upper parts, rods of copper A B of a successively decreasing length. Their lower part was connected with one of the branches of the circuit, but the upper part was connected with the other branch only by means of a kind of jointed metallic cap which rested on them by its own weight. Their height, however, being different, this cap could only touch the longer one. Now, it would follow from this arrange- ment that if this last were to break or be completely con- sumed, the cap would fall on to the next longer one, and would thus send the current through a fresh carbon, which would be instantly illuminated. The latter again breaking, the cap would transmit the current into a third carbon, and so on until the last. Experience had shown that five carbons thus arranged wera ample for an evening's illumination, and during some experiments at which I was present I have twice seen the lamp working at the moment of the rupture taking place. Each of these quintuple systems of carbons was of course enclosed in a hermetically sealed vessel w, from which the air had been exhausted, and their difference of height was calculated so that the curvature produced by the great heat should not cause a division of the current. INCANDESCENT LAMPS. 197 When all the carbons of one lamp were consumed, the cap, by meeting with a copper rod, continued the circuit ; so FIG. 50. that, if there were several lamps included in the same circuit, the extinction of one of them did not involve that of the others. According to the experiments made at Florent's, at St. J98 ELECTRIC LIGHTING. Petersburg, where three of these lamps are in operation, each of them gives a light equal to 20 Carcel lamps, and they are worked by the currents of one of the Alliance Company's machines. Bouliguine's Lamp. — This lamp accomplishes nearly the same object as the preceding, but by using only one car- bon. It is, like the former, composed of a copper base, two vertical rods, two bars for conveying the current, and a valve for exhausting. One of the rods is perforated by a small opening from top to bottom, and nearly all its length there is a slit which allows two small lateral projections to pass through. The carbon is introduced into this rod as intoaporte-crayon, and it tends to rise, by the action of counterpoises attached by very small cords, to the lateral projectors on which the carbon rests. The part of the carbon to be made incandescent is held between the lips of two conical blocks of retort carbon. A screw placed under the base enables the length of the rod which carries the upper conical block to be increased or diminished, and thus a greater or less length can be given to the luminous part. The closing of the globe is accomplished, as in the preceding apparatus, by the lateral pressure of several caoutchouc discs. When the lamp is placed in a circuit, the carbon rod glows and becomes luminous, until at length it breaks. At this instant a little mechanism controlled by an electro- magnet opens the lips of the carbon-holders : the counter- poise of the upper one pushes out of the groove any frag- ments which may remain, and the counterpoise of the lower one raises the carbon rod, which goes into the upper block and re-establishes the current. The mechanism controlled by the electro-magnet again acts, but in the reverse direction of its former movement, the porte-crayons are tightened, and the light reappears. This system, according to Fontaine, does not always yield INCANDESCENT ' LAMPS. 1 99 good results, on account of the multiplicity of its parts, but \vhen by chance it does work regularly, the light it gives is more intense than that of the Konn lamp. Sawyer-lHan's Lamp. — This, lamp is nothing more than King's or Lodyguine's in all its simplicity, except some in- significant arrangements for lessening the calorific radiation, and we are surprised at the long accounts we read in the English journals about this lamp, which is probably not better than the older ones. In this system, the vessel con- taining the incandescent carbon is filled with nitrogen to avoid combustion and the deposition of volatile products on the surface of the vessel. The carbon itself is rather short ; its resistance does not exceed 0*95 ohm, and each light is supplied with a derivation, in order that the conditions of the distribution of the current may not be changed when any one of the lamps is extinguished. The commutator used for light- ing and extinguishing the lamp is, moreover, arranged in such a manner that the current gains its full intensity only by degrees, and after having passed through successively weaker resistances. It is, according to Sawyer- Man, from neglect of this precaution that the carbons of incandescent lamps are so soon spoiled. Finally, an electro-magnetic register is in- cluded in the circuit, and has the same effect as a gas-meter. We shall not speak of the system of distributing the current among all these lamps, for it is founded on the principle of derivation, and is identical with that of Werdermann, which we shall presently consider. A detailed description of this lamp may, however, be seen in the Telegraphic Journal Q{ ist January, 1879. We think it unnecessary to say more about it, than to express our surprise that English and American inventors trouble themselves so little about the earlier in- ventions. E. Reynier's Lamp. — We have seen, on page 146, the principle on which this lamp is based. The lamp is the most 2OO ELECTRIC LIGHTING. important of all those that we are now examining ; it perhaps may some day, under one form or other, solve the problem FIG. 51. of the divisibility of the electric light. The best known form of it is represented in Fig. 51. It is composed, as will be INCANDESCENT LAMPS. 2OI seen, of a long and slender rod of carbon c c 2 millimetres in diameter, supported on a heavy carbon-holder A, which slides on a hollow column D between four rollers. This rod rests on a cylinder of carbon R, turning on a horizontal arm ^ fixed to the column. A guide, fitted with a brake F, en- closes the rod of carbon to a short distance (6 millimetres about) from the carbon cylinder, and at the same time con- ducts to it the positive current, which returns to the gene- rator by the carbon cylinder and its support. A second brake F, placed behind the apparatus and resting on the support A of the movable carbon through an opening made in the tube D, moderates the action of the weight of the support A, according to the degree of pressure exerted on carbon disc R. For this purpose, the support of the axle of this disc forms a rocker, and on the side of the rocker opposite the disc R is fixed the second brake, of which we are speaking. The point of contact of the rod of carbon with the cylinder is placed a little eccentrically, in relation to the vertical pass- ing through the axis of the cylinder, so that at each lowering of the system from the consumption of the rod, a small tan- gential impulse may be given to the cylinder, which causes it to make a slight movement that suffices to throw down the ash accumulated at the point of contact. Without this pre- caution, this ash might interfere with the brilliancy of the light, at least with the impure carbons which are at present in use. The problem which Reynier undertook to solve was, as may be seen, to make a long slender rod of carbon incan- descent towards its extremity, and while wasting at the end, to be moved forward continuously and regularly. Fig. 52, taken from the French patent of the iQth February, 1878, dearly explains the data of this problem. " A cylindrical or prismatic rod of carbon c," says Reynier,* *' is traversed between / and j by a continuous or alternative •- See le compte rendu des stances de la Socittt dc Physique for April- July, 1878, p 96. 2O2 ELECTRIC LIGHTING. current, intense enough to make this portion incandescent. The current enters or leaves through the contact /, and it leaves or enters through the contact B. The contact /, which is elastic, presses the rod laterally; the contact E touches it endways. Under these conditions, the carbon wastes at its extremity j quicker than at any other place, and tends to become shorter. Consequently, if the carbon C is continually pushed in the direc- tion of the arrow so as to always press on the end contact B, it will advance gradually as it is consumed by sliding in the lateral contact /. The heat developed by the passage of the current in the rod is greatly increased by the combustion of the carbon. FIG. 52. FIG. 53. " In practice, I replace the fixed contact by an end contact P., Fig. 53, which removes the ashes of the carbon. The rotation of the end contact is concomitant with the progressive move- ment of the carbon rod, so that the position of the latter on the end contact acts as a brake on the moving mechanism. " The principle of this new system of lamp having been esta- blished, it was easy to invent simpler arrangements for carrying it out. The samples which I have the honour of laying before the Society will explain themselves at a glance. The progression of the carbon c and the rotation of the end contact B are obtained by the descent of the heavy rod of the carbon-holder.* To wind up the lamp, this rod has merely to be lifted up. The rod of * Comparative experiments made by Reynier prove that the renewal of the end contact is indispensable for obtaining a somewhat prolonged action with the ordinary carbons of commerce. INCANDESCENT LAMPS. 203 carbon is put in its place without any adjustment. There is no apparatus for regulating." Previous to constructing the form of lamp we have described, Reynier had contrived a more complicated one, FIG. 54. in which the carbon cylinder was moved by a clockwork mechanism, checked by the pressure of the carbon rod, the end of which formed a brake, so that the clockwork acted only as the carbon consumed. But he soon saw that the problem admitted of a simpler solution. Quite recently Reynier has given his lamp the form shown in Fig. 54, which yields the best results. In this new pattern 204 ELECTRIC LIGHTING. the lateral contact has a less complicated arrangement, and the weight P which acts on the movable carbon to produce the end contact, is placed above the upper end of the carbon. It may therefore be selected as required. The end contact itself is fixed instead of turning round. The arrangement given to the apparatus enables it to be suspended from a ceiling. It can also be placed horizontally : in that case a spring propels the movable carbon. Experiments made at Sautter and Lemonnier's, with ten Reynier lamps and a Gramme machine with a speed of 930 turns, gave the following results in a circuit represented by 100 metres of copper wire of 3 millemetres, or 30 metres of telegraph wire : Number of lamps Indications of the Luminous intensity of Total yield intension. galvanomstre. each lamp. of light. 5 25° . 15 jets 75 Jets. 6 22° • [3 - 78 7 20° 10 — 70 - 10 15 5 ~ So - Serrin's regulator gave under the same circumstances, with a deviation of 21°, a luminous intensity of 320 jets. As regards total luminous intensity, the use of incandescence lamps is therefore less advantageous than that of arc lamps; but with the former there is the advantage of the division of the light, and the possibility of obtaining it with comparatively small electrical power. They consume about 10 centimetres of carbon per hour; but by taking larger carbons and arrang ing the generators for quantity, this consumption may be much reduced. In the experiments made at the Societc d* encouragement, six lamps could be lighted with the current from 25 double Bunsen cells. Their light through ground glass globes was soft, and appeared about equal to that of two or three gas-jets. They could be lighted and extin- guished at will. In the year 1876, S. A. Varley had patented a lamp on a similar plan which we show in Fig. 55, and which is de- scribed as follows : — INCANDESCENT LAMPS. 205 " A rod of carbon T rests gently by its weight and that of its carrier on the periphery of a roller of retort carbon N, and the imperfect contact thus produced gives rise to the incandescence and combustion of the stick of carbon at its extremity." We may remark, however, that with this arrangement, in which the transmission of the cur- rent to the movable carbon is not produced by a sliding contact, placed at a short distance from the point, that the light is only produced at the point of contact, and it can be but very feeble, whilst with Reynier's system, in which this sliding contact exists, the portion of the carbon be tvveen it and the roller is heated to incandescence. It appears also, on FIG. 55. an attentive examination of the pa- tent, that the carbon roller N was made of alternately in- sulating and conducting portions ; and as it was subjected to a rapid rotatory movement, the light was the result of a series of sparks which were produced at the moment of the passage of the carbon T on each of the insulating portions. Werdermann's Lamp. — Werdermann's lamp is in prin- ciple merely Reynier's lamp reversed ; but this arrangement is more practical for public illumination, for which it was specially intended, and it makes use of the voltaic arc action. It is represented in Fig. 56. This system consists essentially of a slender carbon b moving within a metallic tube T, which serves at once as guide and conductor of the current A collar fixed on the lower part connects it — by means of two cords which leave the tubes by two grooves and pass over two pulleys — to a counterpoise P that tends constantly to raise the carbon, and to keep it lightly pressed against a large carbon disc c of 2 206 ELECTRIC LIGHTING. indies diameter, which is kept in a fixed position by a ver- tical support D. This support is fixed to a kind of funnel- shaped covering s, which receives the ashes from the com- bustion, and permits a glass globe to be fitted to the lamp. The upper carbon disc is connected with the negative pole of the generator, and the metallic guide T of the carbon pencil corresponds with the positive pole, so that only the portion of the carbon between the tube and the upper carbon is brought to incan- descence. This incandescence is in- creased by the action of the small voltaic arc which, as we have said, is formed at the point of contact of the two carbons, and it is increased also by the combustion of the at- tenuated carbon. The upper carbon, by virtue of its greater mass, neither burns nor changes. The action of the counterpoise is regulated by means of a spring K provided with an adjusting screw, which, by press- ing more or less on the part of the tube surrounding' the carbon, acts as a brake. The greater or less pres- sure of this brake depends on the formation of the voltaic arc between the two carbons, or merely on the incandescent effects, and it is in this circumstance that the difference between Werdermann's and Reynier's system specially resides. In the latter the movable carbon is too contracted to take part in the formation of an arc, whilst in the other this carbon is sufficiently free for this effect to be produced ; and thus the diameter of the movable carbon may without inconvenience be increased for the purpose of making it last longer. In the new forms of the Werdermann lamp the upper disc of INCANDESCENT LAMPS. 207 carbon has been advantageously replaced by a disc of copper. Experiments made in England with a Gramme machine arranged for electroplating and work- ed by a steam engine of (it is stated) two horse- power, * gave, according to Werdermann, the fol- lowing results : — " i°. When the current of the machine was distri- buted between two lamps, the brightness of the light was equal to that of 360 candles. This light was white, and apparently free from the blue and red ray so often perceived in the light of the voltaic arc. It was, besides, perfectly steady. "2°. By placing in the circuit 10 derivations, each corresponding with one lamp, as shown in Fig. 57, 10 luminous centres could be obtained, each representing about 40 can- dles. In order to render the action steady, coils a a a of small resistance are interposed in each de- rivation. Under these conditions the resistance of each lamp was 0-392 ohm, and therefore the total resistance of the circuit was only 0x337 ohm. * According to the particulars sent to me, this power should be greater. 20« ELECTRIC LIGHTING. " 3°. The consumption of the carbons in the smallest pattern of the lamps did not exceed 2 inches an hour, and for the large pattern 3 inches. A length of one metre might be used. They were Carre carbons." With this system, and for that matter with Reynier's system also, all the lamps could be lighted or extinguished at once or successively ; and as their brilliancy cannot be great, trans- parent instead of ground glass globes may be used. Fig. 58 shows the commutators in each derivation, for throwing into it the resistance a', for the interruption of the circuit, and for direcf transmission ; they are metallic rings, FIG. 58. divided into 4 segments, within which is placed a plug half of metal and half of an insulating substance. The two upper parts of each ring are connected, one with the lower carbon of the lamp, the other with the upper carbon ; but the con- nection in the last instance is effected through a resistance equal to that of the lamp, or 0*392 ohm. The lower part on the left is connected with the positive wire, and the lower part on the right is made of ebonite. When the plug is in the first position shown in the figure, the current traverses the lamp directly, because the communication is established by the metallic part between the two metallic sectors at the left ; when it is in the second position, the current no longer passes through the carbons but through the resistance coil; and as this is equivalent to that of the lamp, there is no change in the distribution of electricity. Finally, when the plug is INCANDESCENT LAMPS. 209 in the third position, the circuit is interrupted, and the other lamps benefit by the portion of the current that was pre- viously passing through the lamp. Besides the resistances just spoken of, there are others placed in the course of the positive wire of each derivation, which are indicated in Fig. 57, and serve to change the brilliancy of any particular lamp. This system of electric lamp has lately been used in London for lighting the South Kensington Museum, where it is said to have produced a well-spread light of a fine quality, an agreeable softness, and a remarkable steadiness. (See the journal La Lumiere eleclrique of the i5th July, 1879, p. 72.) Some very successful experiments were also made at Paris, in which six to eight lamps were lighted by the current of an Alliance machine. It is asserted that it will be possible to light a much greater number with lamps of a smaller pattern and with a particular arrangement of the machine. We shall wait and see the result for ourselves before pro- nouncing on this matter. We have just learnt that with a Gramme machine, driven by a gas engine of 6 horse-power, it has been found possible to light 1 5 Werdermann lamps, the carbons of which were 4-5 millimetres in diameter, each lamp giving a light of 25 Carcel lamps. It was, however, a new form of this kind of lamp which gave this result, and this form is not sufficiently known for us to describe it here. (See ^Tote I> at the end of this volume.) Trouve's form of Reynier's Lamp.— Trouve' has just given a new arrangement to Reynier's lamp, which is some- what that of Werdermann, and at once rather practical and economical. Fig. 59 will give a precise idea of it. The small carbon E, which gives the incandescence, is inclosed in a long tube with a slit all its length, which allows the movable portion D to push the fine carbon against the mas- sive cylinder D under the influence of a counterpoise P pro- 210 ELECTRIC LIGHTING. FIG. 59. vided with rollers. This fine carbon is guided between two spring rollers g, which also carry the current to it. The cylinder c is of copper, and presents itself to the carbon in INCANDESCENT LAMPS. 211 such a manner as to be able to turn by the tangential pres- sure against it, in proportion as the illuminated rod B B is consumed. This lamp works well, and can henceforth be ap- plied to domestic illumination under the influence of a current from 6 Bunsen cells. The rod of carbon is of a more slender diameter than those used by Reynier. E. Arnould has lately invented a lamp the arrangement of which is exactly that of the preceding apparatus, but under much less excellent conditions, and yet certain newspapers have not hesitated to announce it as something new. Ducretet's form of Reynier's Lamp.— An arrange- ment of the kind just al- luded to has also been given to Reynier's lamp by Ducretet. In this form, shown in Fig. 60, the rod of carbon c M, instead of being pushed by a counter- poise, is plunged in a column of mercury, filling a long iron tube T, which forms the body of* the lamp. A cap and an elastic collar B guide the rod, and, as well as the mercury, impart to it the positive polarity. A carbon cylinder H, sup- ported by a metallic arm s, movable in a socket with a regulating screw, allows any required length to be given to the incandescent portion. The current reaches the apparatus by the conductor /, and leaves FIG. 60. 14—2 2 1 2 ELECTRIC LIGHTING. it by the conductor /' and the interrupter M v. In order to avoid the poisonous vapours emitted by the mercury in con- sequence of becoming heated, the cap B is insulated from the tube T by means of a substance not a conductor of heat, and it communicates with the base of the lamp through the copper conductor /, which, being a good conductor of heat, diffuses it and withdraws it from the mercury. This system, by its simplicity, permits the lamp to be very cheaply made. It had. however, been already devised by Reynier, who had pointed it out amongst the arrangements that might be given to his apparatus. Tommasi's Lamp.— In order to obtain a longer duration of the working of the preceding lamps, Tommasi arranged them like a revolver. For this purpose, the lamp is formed of an iron tube about 3 centimetres in diameter, turning on a pivot and supporting five carbons of 30 centimetres length, and these, one after another, are by a rotatory movement brought into contact with the carbon disc of the negative pole, and thus the light is produced for eight or ten hours. The forward movement of these carbons is effected by mer- cury contained in the tube, as in the preceding system, and, according to Moigno, the light thus supplied is very uniform and steady. This lamp has the shape and size of an ordi- nary Carcel lamp. Edison's Lamp.— The reputation that Edison acquired by the invention of the phonograph was the cause of con- siderable financial disasters, when some time ago he an- nounced that he had finally found the long-sought solution of the problem of the division of the electric light. He was taken at his word, and in America, as well as in England and in France, the shares of the gas companies fell enormously. It was forgotten that the American newspapers were often the propagators of false news; and since the phonograph triumphantly refuted the denials with which its announce- INCANDESCENT LAMPS. 2 1 3 ment was greeted, people ceased to recollect the proverbial tanards Americains. Be that as it may, this pompous an- nouncement caused many losses, and that at three different times. People have now recovered from their fears, and, as always happens, have passed from one extreme opinion to its opposite. The truth is, that if Edison's invention were at first but a very small matter, it might some day gain import- ance ; and according to the information I have received from ocular witnesses, the last patents of Edison's, which require not less than 200 pages for their description, might contain discoveries of the greatest importance. It seems that even Edison's failures have not discouraged American capitalists, and that they have put at his disposal not only all the necessary amounts, but also engineers of all kinds, elec- tricians, mechanicians, chemists, &c. According to the persons who gave me this information, Edison has con- structed a dynamo-electric generator superior to all those we have mentioned, and capable of utilizing 90 per cent, of the motive power. Without attaching too much importance to these accounts, it must be admitted that a man so ingenious as Edison has not for so long a time studied a question without making something of it ; and while waiting until the new patents are published, we must meanwhile give some account of the lamp which has fluttered the financial world. It is repre- sented in Fig. 6 1. According to Edison, the loss which is involved in the division of the electric light is not peculiar to this kind of light, but also occurs with gas when the elements of its flow are placed under the like conditions. Thus, if a great number of gas-jets are opened and the total amount of light measured, a smaller luminous intensity will be found than that resulting from a small number of these jets under the same pressure of gas. But if that pressure be suitably regu- lated, things may be so arranged that this loss would not -exist. " Now," says Edison, " properly arrange the condi- 214 ELECTRIC LIGHTING. tions of the flow of the electric current, and you can do for electricity that which has been done for gas." This reason - FIG. 61. ing is correct only up to a certain point, for the conditions of maximum intensity of a current with or without derived INCANDESCED T LAMPS. 2 1 5 currents are well known, and the arrangement of the gene- rator in any given case is indicated perfectly by Ohm's formulae. As regards this matter, we do not think that Edison can discover anything more than we know already : and we believe that it is rather by the material arrangements of his lamp that he will have any chance of arriving at the solution he is seeking. Now Edison's lamp, which has many very different forms, consists of the following : The principle of the lamp, which is represented in Fig. 61, is the incandescence of a spiral of platinum wire a alloyed with iridium. And in order to prevent the spiral from burn- ing when its temperature exceeds a certain point, Edison places within the spiral a metallic rod G, which by dilating causes a contact to be made at z, by means of the lever s, precisely at the moment when the heat is about to reach that degree. The current is then derived through this con- tact, and immediately lowers the temperature of the spiral, which causes a disjunction of the derivation and arrests the cooling. The spiral then begins again to be heated, and is thus maintained at a temperature which can vary only be- tween very narrow limits, which may be regulated by the greater or less separation of the contact piece i of the deri- vation, by the resistance of the derivation, or by a resistance regulator variable with the pressure, based on the principle which Edison has applied to telephonic transmitters. Can a platinum spiral by this means acquire a temperature sufficiently high to produce light ? This appears the more doubtful from the fact that De Changy did construct a very ingenious regulator under similar conditions, and that, never- theless, his spirals were volatilized when they were heated to such a point as to become luminous. It is true that, according to the American journals, Edison has discovered a new metallic alloy, having its fusing point much higher than that of any known metal.* * According to Edison's last patent, the incandescent spiral is formed of platinum and iridium, and is covered with a metallic oxide, either of cerium, 216 ELEC'lRIC LIGHTING. Automatic lighter of Reynier's E-lectric Lamps.— To avoid the inconveniences which may result from the extinction of an electric lamp — an extinction which may involve that of all the rest of the lamps in the same circuit — Reynier arranges an automatic system of permutator, the effect of which is to light other duplicate lamps placed in the neighbourhood. This system consists of a kind of electro-magnet relay which sends the current of one circuit into another, when the armature of this relay, being no longer attracted, touches its limiting stop. If for this pur- pose two duplicate lamps are used, as Reynier proposes, these lamps are introduced into two derivations of equal resistance arranged between the two conductors of the generator, and the relays which are to act in these lamps are interposed in the circuit of the first lamp : the contacts of these relays are also arranged so as to establish the com- munication of the derivations with the circuit at the separation of each relay. When the first lamp is lit, the electro-mag- nets of the relays being active, the communications of the derivations with the circuit are severed, and the lamp works alone ; but if the latter should be extinguished, either by the consumption of the carbons or by accident, the current is immediately sent into the derivations ; but as there is .then but one relay which is active, there is only one of the duplicate lamps at work, and it is only when this one is in its turn extinguished that the third is lighted. This system may readily be arranged to suit the various ways in which the luminous central arc is connected with the principal circuit. zirconium, calcium, magnesium, or of some other metal — i.e., an oxide not altered by a high temperature. The effect of this covering is designated by Edison by the name of pyro-insulat>on, and in order to cover the wire it suffices to dip it into a solution of one of these oxides, then into an acid, and to pass the spiral through the flame so as to evaporate the aqueous portions and leave only the oxide on the wire. (See the Telegraphic Journal of i5th July, 1879.) ELECTRIC CANDLES. 217 ELECTRIC CANDLES. JablochkofFs electric candles, which have lately been so much spoken of, are certainly not the ideal of the electric light ; but on account of the a'bsence of all mechanism, and the comparative regularity of their action, they can be applied to public illumination, and assuredly this could not have />een done with any electric lamp invented up to that time. It is they, and the influential company working the invention, that have enabled those beautiful experiments to be under- taken, and those splendid illuminations to be made of the Avenue de f Opera, of the Arc de I'Etoile, of the Chambre des Deputes, of the shops of the Louvre, of the Theatre du Chdtelet, &c., which astonished all the visitors to Paris during the Exhibition of 1878, and proved that electric lighting was not the chimera that those interested in gas companies wished to think it. It was the Jablochkoff lamps that excited that rage for electric lighting in every country, which will, as sure as fate, shortly lead to the substitution — or at least the partial substitution — of gas illumination by electric illumina- tion. These candles are, moreover, much used, being now daily lighted in more than 1,500 lamps. We must there- fore devote a long chapter to them. If two perfectly straight carbons are placed side by side parallel to each other, and separated by an insulating layer capable of being fused or volatilized by the passage of £ie electric current between the two carbons, a lamp without machinery is obtained, which gives light like a candle, that is to say, it burns progressively until the two carbons are en- tirely consumed. This is the principle of the Jablochkoff candle represented in Fig. 62. Numerous experiments have been made to ascertain what is the best insulating material to place between the carbons, 2l8 ELECTRIC LIGHTING. and what are the best dimensions for them. After many trials, preference has been given to plaster of Paris as the insulator, and to Carre's carbons 25 centimetres long and 4 millimetres in diameter. When giving a light equal to from 25 to 40 gas-jets, these candles will last for one hour and a half; but we shall presently see that by a very simple arrangement the light founded on this system may be made to last as long as may be de- sired. A Jablochkoff candle is, then, composed of two insulated Carre carbons <:, d, 25 centimetres long, slightly pointed at their upper extremities, and separated by the insulator already mentioned, provided at its upper end with a thin layer of lamp black for conducting the current when the candle is first lighted. This layer is made of black lead mixed with gum, and the candle is charged by merely dipping its end in the mixture. At first the two carbons were connected by a plumbago point kept inks place by a piece of asbestos paper; the preceding plan is now found to be much simpler. At the lower end the carbons are fitted into copper tubes, by which they are connected with the circuit when the candle is placed in the chandelier ; these tubes are inserted in the piece M (made of a silicate or other suitable substance) so as to prevent the carbons from separating from their insulating partition. Before the manufacture (° Carriage and packing .-. 5° Various IOQ Total ... 2,300 and the working expenses per hour would be : Fr. c. Retort carbon ... °'21 Coke for motive power o' 1 5 Working expenses and management ... o'io Sinking of 2,300 fr., divided over 500 hours of lighting yearly 0-49 Total 0-92 If lighted throughout every night, that is to say, 4,000 hours in a year, the cost falls to 53 centimes. With an hydraulic motor the expenses are 77 centimes for an illumination of 500 hour?, and 38 centimes for an illumi- nation of 4,000 hours. A workshop of 60 metres by 20 will require two lamps, which will entail a cost of i fr. 77 c. per hour instead of 3fr. 50 c., the expense of 100 gas-lights that would have to l>e used in such a case, and which would give only one-fifth or one-sixth of the total intensity of light. If the workshop is open all the night, the cost per hour becomes o fr. 97 c. with the electric light, and 3fr. 07 c. for the gas. It will be interesting to compare these indications with the results that were published in England after the Trinity House •experiments, and we give an abstract of these from a paper xead before the Society of Civil Engineers of London, by Higgs and Brittle, entitled " On some recent Improvements in Dynamo-electric Apparatus. " 246 ELECTRIC LIGHTING. "Although under certain circumstances these two agents un- doubtedly come into competition, they have two separate fields. Hitherto gas been generally employed for lighting spaces of both large and small dimensions, because a better source of light for large spaces has not been procurable with economy. But for lighting large spaces that are not subdivided by opaque objects or screens, it is a want of economy to employ gas. If, in fact, a gasworks were to be constructed simply for lighting large spaces, as does occur in some extensive works, the disbursement neces- sary to establish even a small gasworks would, compared with that necessary to establish the electric light, be a considerable multiple of the latter. Assuming light-power proportional to- horse-power expended (although it increases at a greater rate),. 100 horse-power would give 150,000 candles light ; if this be dis- tributed from three points, the cost of each lamp per hour would not be more than js. 6d., or £i 2s. 6d. per hour for the three, each light centre giving an illumination which would enable small print to be read at a distance of a i mile from the light. A burner giving the light of 20 cafrdles consumes 6 cubic feet of gas per hour, which may be manufactured at a cost of 2s. per 1,000 cubic feet. This gives 7,500 burners' light only, and 45,000 cubic feet of gas at a cost of £4. 105-. per hour, a ratio of 4 to i in favour of electric lighting. Electric lighting, where adopted, has been found to be generally more economical than gas light- ing, but the economical ratios differ greatly, and are dependent chiefly upon the price of gas and the motor power employed. For large spaces the cost of electric lighting is about one-fourth, or even one-fifth of that of gas lighting, when steam has been used as a power, and wear and tear are reckoned. With a gas engine as motor, the ratio has only been as I to 3, the greatest economy having been with a turbine as motor. At 'Dieu's work- shops at Davours the cost per hour for gas is 2s. o 632^. against i s. j.2d. for electric lighting. Ducommun . finds, taking into account wear and tear and interest, that gas costs 2*25 times more than the electric light, which ratio increases to 7'i5 when wear and tear and interest are left out of consideration. At Siemens Brothers' Telegraph Works the cable shops are im- perfectly lighted with 120 gas-burners. Each of these burners consumes 6 cubic feet per hour, at a cost of $s. <)d. per 1,000 cubic feet The cost of fixing gas-pipes, including cost of pipes, COST OF ELECTRIC LIGHTING. 247 burners, cocks, &c., for the 120 burners is ^60. Taking interest at 5 per cent., to include wear and tear and renewals, there results for i. ooo hours' consumption per annum — £ J. d. Interest 900 Cost of gas consumed .,. ... 135 o o ,£144 o o "At the utmost the 120 burners cannot give more than 2,400 candles' light, and naturally but a small percentage of this is reached. Further, when steam or other vapour or fog arises the gas-jets are obscured. The space being subdivided, it is neces- sary to employ three machines. These three machines, with lamps, conducting wires, mounting, &c., cost ^250.* " Thus the economy is 2 to i in favour of electric lighting. But there is the further advantage that the lighting is perfect, and that steam or vapour or fog does not cause inconvenience. If, however, the ratio of light-intensities were adapted to the ratio of efficiency, the advantage would be considerably higher (20 to i) in favour of electric lighting. It may be laid down, as proved by experiment, that for lighting large spaces not too much subdivided, the advantage is greatly in favour of the electric light ; but that when numerous light-centres of sihall intensity are required, or where the space is much subdivided, the advantage is in favour of gas. This advantage will cease when a practical method of subdividing the electric light has been obtained. In places where opaque objects or screens occur that only throw- shadow, but are not of sufficient size to completely block out the light from the space they inclose, reflectors can be utilized to overcome the difficulty of shadows. When the electric light is capable of minute subdivision, it will undoubtedly compete with gas on terms of the highest advantage, since the cost of establish- ing a gasworks will be many times in excess of that necessary to supply the electric light to a district. * Interest 15 per cent, upon ^'250 37 10 o Carbons, coals, attendance, oils, &c., per 1,000 hours 35 4 o £72 14 o 248 ELECTRIC LIGHTING. " The objection that the glare of the electric light is trying to the eyes of the workpeople, has been overcome by inclosing the light in an opaque reflector, the rays being projected on to a screen, or on to the ceiling or roof of the building, whence they are diffused, giving to the space lighted the appearance of illumi- nation by daylight." PART V.— APPLICATIONS OF THE ELECTRIC LIGHT. THE comparative cheapness of the electric light and its concentrated power have long ago given rise to the idea of applying it in a large number of special cases, and latterly the hope has been entertained of even using it as a means of public illumination. But without speaking of this appli- cation, which, as we shall presently see, has not yet been completely successful, there are a multitude of cases in which this method of illumination can now be employed under favourable conditions; as, for example, in the lighting of large workshops, great retail establishments, works carried on at night, goods stations of railways, drifts in mines, &c. These are cases in which no* other system of lighting can yield such adequate and advantageous results. Such are the applications of the electric light to lighthouses, to military operations, to navigation, to submarine work, to projecting on a screen certain scientific experiments, to theatrical effects, to public rejoicings, to maritime signals, &c. These applications we are now about to discuss, and we shall begin with the most general of all, namely, public lighting. Application to Public Illumination. — Since Davy discovered the wonderful illuminating power of the electric discharge between two carbons, many attempts have been made to apply it to public illumination. These attempts have not yielded very satisfactory results, and this must necessarily have been the case, for besides' the cost of this mode of lighting, which was very high, the thing desired was not an intense and concentrated light : such a light is, in fact, insupportable to the eye when near, and it is inca- pable of illuminating a sufficiently wide area to give a real 249 250 ELECTRIC LIGHTING. advantage over lights dispersed in a number of different points. This truth was established a score of years ago by the experiments in the Place du Carrousel, not indeed with the electric light, but with a light equally intense, which formed the centre of a splendidly luminous sphere. It was finally admitted that this single point was Tar from supplying the same advan- tages as ordinary gas-jets, which could be placed as required. Now, if we remember that the special characteristic of the electric light is its concentrated power, we may conclude th^t if this light had to remain subjected to the same condi- tions under which it existed a few years ago, it would never have been regarded as a mode of public illumination. Never- theless, the considerable reduction of the cost of this light,, and the methods of dividing it under favourable conditions,, which have recently been discovered, have changed the aspect of the question. The important experiments undertaken in 1878 by the JablochkofT Company gave rise to new ideas, which have now been taken up by all civilized countries, and promise to lead to some important results. It is therefore not surprising that the gas companies should have been affected, and their shares depreciated. Nevertheless, we think the depreciation has been unduly exaggerated, for, as we remarked at the beginning, it is difficult to believe that uses will not be found for gas, considering the important pur- poses to which it can be applied in very many branches of industry. We have already explained the manner in which the electric light can be divided sufficiently for the requirements of public illumination, by means of induction machines with fixed coils, Jablochkoff candles, the lamps of Lontin, Rey- nier, Werdermann, &c. We must not, however, suppose that this idea is new. The division of the electric light has long been sought after, and several plans have been proposed, such as those of Wartmann, Quirini, Liais, Deleuil, Ronalds, Lacassagne and Thiers, and Martin de Brettes, which I have described in my Expose des applications de APPLICATIONS OF THE ELECTRIC LIGHT. 251 /. K, p. 550. But these systems, which were founded either on derivations of the current, or on rapid successive permu- tations of the current through a certain number of different circuits, or on the revolving projection of a beam of light, were destitute of arrangements sufficiently energetic and well con- trived to solve the problem. It was not until Lontin, Lodyguine, and Jablochkoff had made their earlier experi- ments that the possibility of dividing the electric light began to gain some degree of credence. There was still, however, great doubt as to the possibility of illuminating a space of con- siderable length. It was supposed the loss of electric intensity from the length of the conductors could absorb all the power of the generator, and that sufficient electric energy would not be left to light several lamps. But Jablochkoff's experi- ments, performed on the whole length of the Avenue de rOp'era, with only one machrne for each side of the street, completely removed all doubts on this head, and from that time, as we have already said, the problem of electric light- ing has come to the front in every country. We must there- fore devote a few lines to these remarkable experiments, which are still carried on at the present moment, and are being extended to other important thoroughfares in Paris. In the Avenue de r Opera and the Place du Theatre- Fran$ai$ there are 32 lamp-posts carrying the electric light, 16 on each side. The lamp-posts are the ordinary ones of the city of Paris, supported by circular oaken pedestals i '5 metres high, and surmounted by lamps like that shown in Fig. 65, page 225. Each lamp holds six candles, and seven wires enter it from the interior of the lamp-post after having passed through a commutator of six contacts placed within the pedestal. The machines which work these different lamps are set up in two cellars under the middle of the street, and therefore the wires are divided into two bundles for each machine ; two of these bundles go up the street on each side, and the other two down. The wires are buried in the earth below the pave- ments, and besides their insulating covering of gutta-percha. 252 ELECTRIC LIGHTING. and tarred canvas, they are protected by well-fitted drain- pipes. There are openings in front of each lamp-post, and in these are made the connections between the wires of each lamp and those of the circuit. Of course the seven wires are only between the commutator and the candles ; everywhere else there are only two for each set of four lamps. It appears that the two machines require 36 horse-power in order to light 32 lamps on each side of the street, that is, every lamp requires 1*12 H.P. ; but it should be noticed that in this estimate the power needed to overcome the resistance of the conductors is included, and that this must be conside- rable is obvious from the fact that for the most distant con- ductors it is nearly 1,000 metres of wire. Jablochkoff asserts that each of the lamps represents from 25 to 30 gas-jets; but F. Leblanc declares that it does not exceed 12 jets, and it is on this estimate that the last contract between the Jablochkoff Company and the city of Paris is based. We must, however, bear in mind that nearly 45 per cent, of the light is absorbed by the enamelled glass globes; so that it is possible that each light represents 50 or 60 Carcel lamps according to Jablochkoff, or 22 to 24 such lamps ac- cording to Leblanc. According to the original agreement between the city of Paris and the Jablochkoff Company, the cost to the city was six times that of gas. It is now much reduced, and unless the company is a loser by the new agree- ment it has entered into with the city of Paris, the cost may be said to be nearly that of gas, for it is only one-fourth as much as under the first agreement. The debates in the Municipal Council on the renewal of their agreement may illustrate the cost of maintaining this system of lighting. The illumination of the Avenue del1 Opera, together with that of the Place deV Opera, the Place du Theatre Fran^ais, and the fa£ade of the Corps Legislalif, originally cost i franc 25 centimes for each lamp 'per hour. No\v accor- ding to Mallet, 68 gas-jets may be had for this sum, and thus electric lighting would be at a disadvantage in tiie propor- APPLICATIONS OF 7^HE ELECTRIC LIGHT. 253 tionof i to 6. But the amount of 1-25 francs was in the last discussion in the Council reduced to 30 centimes, which it seems the Jablochkoff Company have accepted, as the lighting is continued. We think, however, that under these conditions the company must be out of pocket, for accord- ing to the calculations of the Reporter to the Municipal Council the cost must amount to 75 centimes. As the subject is interesting on account of the difficulty of obtaining exact rfafa, it may be useful to give a summary of what was said on this subject in the Municipal Council. According to the Report of Cernesson, the cost of main- tenance for each electric lamp per hour may be deduced from the following list of the expenses of lighting 62 lamps for the space of one hour : — Fr. c. Motive power (sundry expenses) 3-20 Fuel (for the engines) ... 664 Oil for lubrication 1*23 Wages of attendant ... 3'2o 62 candles at 50 centimes each 3i'oo Total ... 4527 This gives for each candle o fr. 75 cent. Yet the Com- pany were content to demand only o fr. 60 cent, and wished the Council to enter into an agreement on those terms : the Report did not, however, recommend a compliance with this, but established the principle of paying the Company in. proportion to the amount of light supplied. " Each of the electric lamps," says the Report, "having been supposed to give as much light as u gas-jets consuming 140 litres per hour, and costing each about 2\ centimes, the Com- mission considers that a sum of 30 centimes per lamp per hour should be allotted to the Company. According to the preceding scale the Company would be entitled to only 27 centimes." As to the Electric Lighting Company's project of under- taking for three years the illumination of the principal 254 ELECTRIC LIGHTING. thoroughfares of Paris, the Commission distinctly refused to bind itself in any particular way, but decided that in the present condition of affairs, electricity, as represented by the Jablochkoff candle, could not be considered as having reached such a degree of perfection that it could supersede 1 gas ; that, nevertheless, the progress that had been effected was sufficiently real and important to justify a continuance of the experiments on a larger scale. The Commission therefore considers that the Avenue de T Opera should con- tinue to be lighted as before for one year, beginning on the 1 5th January, and that the Company might apply its system in two new positions in the city of Paris, viz., the Place de la Bastille &&& one of the wings of the central markets. (See the journal LElectririte of the 5th and 20th January.) The illumination of the Avenue de r Opera and of the Place du Thedtre-Francais allows but one candle to each lamp, and there are in all forty-eight luminous points. On the Place de r Opera there are only eight lamps, but two candles burn together in each of them; the two triple lamps placed on each side of the facade of the Opera-House have simple candles only. These two lamps are supplied by two Alliance machines, but those in the Place de r Opera are worked by a Gramme dividing machine, and this requires for the electric lighting of this part of Paris the employment of four machines, each using from sixteen to eighteen horse-power. The Place de la Bastille and the central market are each lighted by sixteen lamps, In spite of the fierce hostility with which the gas com- panies attacked the newspapers in these trials of electric lighting, every civilized country is now turning its attention to this subject. The cities of Stockholm, St. Petersburg, Amsterdam, and San Francisco have already taken steps to establish the electric light. The City of London is at present engaged in realizing this application of electricity, and the Thames Embankment is now lighted in this way. Even America has made repeated attempts to reach an immediate APPLICATIONS OF THE ELECTRIC LIGHT. 255 solution of the problem. Shall we have merely the poor courage to adopt this beautiful application of science only .after it has been carried into effect by every other nation, .as in the cases of the electric telegraph, railways, &c. ? This would be hard, after having made the first experiments. We have hitherto discussed only the light given by the Jablochkoff candles; but this system is not the only one which may be employed for 'public illumination, and it is, perhaps, not even the most economical for the light pro duced. The globes of enamelled glass are, for one thing^ a bad arrangement, as they prevent the utilization of the total light, and the Baccarat frosted glass, which allows 70 per cent, of the light to pass instead of 55, have been tried, as we have seen, 'with much advantage. Again, there is now being tried Paris's new kind of enamelled glass, which absorbs only 35 per cent, of the light, and Clemandot's globes, formed of concentric spheres of transparent glass, with a packing of glass between them. This arrangement, when tried at shops -of the Louvre, absorbed, it is said, only 24 per cent, of the light instead of 45. With the Reynier and Wedermann systems applied to the existing lamps, a greater division of the light, and probably a smaller consumption of the carbons, -would be obtained, with equal electric intensity. If the car- bons were placed within the column of the lamp, which could easily be done in the Wedermann arrangement, they would burn a whole night without requiring attention. Again, with new machines it would be possible to reduce in such a degree the expenditure of motive power that the high price of the electric light, which under the present conditions is the strong point of the partizans of gas, may be sufficiently lowered to compete successfully with gas itself. For lighting on the spot, the problem has, as we have seen, been long ago completely solved ; and it is not affirmed that by increasing the section of the conductors, and expending upon them as much as the cost of gas-pipes, public illumination would not .be placed under conditions of economy as favourable as if the -'5 6 ELECTRIC LI G PI TING. light were near the machine. We therefore see no reason why the cost of electric lighting should not come to be lower than that of gas lighting. It is only a question of time, and the important point was to prove that public illumination by electricity was not physically impossible — a matter upon which we now know what opinion to hold. It has been asserted that the electric light was hurtful to the eye, disagreeable in its effect, and alarming to horses. The experiments of the last six months have not given me any such impression; and if we stand in the evening at the corner of the Rue de la Paix and the Avenue de r Opera, and compare the ilumination of the two thoroughfares, especially as regards the houses on either side, we should suppose that one of the streets is dark. Certainly, when the electric lighting of the Avenue de r Opera is given up, the public will find a vast difference, and will not easily become accustomed to the gas-lamps, which serva only to make the darkness visible ; and yet these lamps contain three gas-jets where a little while ago there was only one. As to the bluish hue of the electric light, it looks cold only because we are accustomed to reddish lights ; but if a red tint were required, it would not be difficult to impart it, by in- troducing certain colouring salts into the composition of the carbons. It is, however, difficult to suppose that this objec- tion; is seriously advanced ; for it seems to me that a white light, resembling that of the sun and showing colours in their proper hues, is preferable to one which casts a false tint on all objects. Those who find fault with the electric light on this ground must equally object to those beautiful moonlight effects that are so much praised by artists and poets. In the pamphlet published by the Jablochkoff Company the advantages of electricity for public lighting are thus set forth :— "Besides the economy of the light it has the following qualities : — " i°. It does not alter colours, but allows the closest tints to APPLICATIONS OF THE ELECTRIC LIGHT. 257 be distinguished, which would be impossible with gaslight. In every industry where the qualities of objects have to be judged of from their colour, in those where there are sortings or selec- tions according to tints, in assorting stuffs or thread of various shades, the electric light is of unquestioned utility. " 2°. The heat given off by electric lighting is extremely small. It is well known at what trouble and cost a very imperfect ven- tilation of apartments in which a, number of gas-jets are burnt can be obtained. "3°. In industrial establishments the electric light supplies a general illumination, which facilitates superintendence at the same time that it simplifies all work of transport, management, &c., &c. It enables, therefore, the number of workmen employed in night shifts to be diminished, and consequently the extent of the premises in which the work is carried on to be reduced. There is thus, together with less cost, an economy of labour and an economy of the prime cost of the establishment. " 4°. It removes the danger of using gas, which results either from negligence or from an escape from the pipes, as when the turning of a tap is forgotten or the pipes are fused in a fire otherwise originating. " Nothing of this kind happens when the electric light is used. Instead of the fusible lead-piping, from which the gas can so readily be allowed to escape, and which can so easily be maliciously injured, the electric communication is by a cord or copper wire covered with an insulating substance. If the circuit were cut, which could only happen intentionally, the reverse would take place from what occurs in the case of gas, for the fluid would no longer flow, and the light would be extinguished. So that, while in the one case the gas from a leakage very easily occasioned, either by accident or by malice, spreads rapidly and forms with the air a mixture that a spark would cause to explode ; in the other case, if the circuit is interrupted there is merely a simple extinction of the lights. " Fires may also be occasioned by the direct action of the gas- jets on combustible materials. Now, an electric lamp having a llaine of scarcely any size takes the place of a great number of gas-jets, and the chances of a fire are proportionately reduced. It is, moreover, important to observe that as the combustion of gas gives off much heat, the inflammation of combustible mate- 17 25 8 ELECTRIC LIGHTING. rials is facilitated by the high temperature of places lighted by gas. " The electric l;ght, on the contrary, gives off extremely little heat, since a lamp yielding a light equal to that of several hun- dred stearine candles does not give more heat than a single candle. "5°. All the lamps supplied by the same dynamo-electric machine are lighted up instantaneously." When only a portion of the space has to be lighted in a given direction, and under an angle not exceeding 180 degrees, the diffusion projectors invented by J. Van Mal- deren may be used with advantage. These consist of a kind of parabolic mirror, of which the electric light occupies the centre, and with the front part closed at a short distance from the arc by a ground glass which receives the beam of parallel rays sent out by the mirror, spreads them out so that all the space in front is illuminated. It is stated that the intensity of the illumination is greatly increased in this way. I think it will be interesting here to reproduce a letter by Lontin relating to the divisibility of the electric light, which has lately been so much spoken of, and has been put forward as a new discovery, the anterior labours of Wartmann, Quirini, and others being ignored. The letter was addressed to the journal LJ Eledricite, and appeared in the number for 5th November, 1878 : — " Allow me to remind you that two years ago / took out a patent for photo-electric regulators, which divide perfectly the current supplied to them. " The Alliance electro-magnetic machine which was working in the Exhibition, supplied, it is true, 4 Jablochkoff candles, but this same machine, without any change, has worked 12 of my regu- lators. Here, I think, we truly have the divisibility of the electric light, a divisibility the more real since each of my regulators gave a luminous intensity of 19 Carcel lamps. This luminous power may, moreover, be yet further reduced, for I have obtained in- tensities of only 4 Carcels. I think I may conclude from these experiments that the Alliance machine, or one of my dynamo- APPLICATIONS OF THE ELECTRIC LIGHT. 259 electric machines suitably arranged for that purpose, can supply 50 regulators. " The illumination of the Lyons railway station last year was supplied by one of my electric generators yielding 12 currents, each current maintaining 2 or 3 lamps. The new apparatus I am now fitting up will allow of 4 lamps being inserted in each current. "At the Saint-Lazare station each current supplies 2 or 4 lamps, the intensity of which is 'regulated according to the re- quirements of the service." Application to the Illumination of Lighthouses. — We are now no longer in the region of hypotheses, the application of the electric light to lighthouses being a fait accompli for more than fifteen years (1864), and I am not aware that any serious accident has interrupted the experi- ments. Most of the important lighthouses on the coasts of France, Russia, and England are thus lighted ; and it is to the bold enterprise of the Alliance Company and its intelli- gent director, Berlioz, that the civilized world owes this beautiful application, which has certainly prevented many maritime disasters. It is true that Berlioz has been ener- getically assisted in his experiments by the Lighthouse Ad- ministration, and among others by Reynard and Degrand, who, after many intelligent experiments, arranged the light- houses of La Heve on this new system towards the end of 1863. Some time afterwards England imitated us, and em- ployed the electro-magnetic machines of Holmes, which were merely an inferior copy of those of the Alliance. Le Roux has published in the Bulletin de la Societe d^ Encouragement a very interesting paper on this kind of application, and we should have had great pleasure in reproducing it here had space permitted, but we shall give merely a summary of it, referring the reader to tome XIV. of the Bulletin de la Societe, p. 762. At the present time, dynamo-electric macnlnes appear to be preferred, and the Telegraphic Journal in the number of 17— 2 260 ELECTRIC LIGHTING. the ist December, 1877, gives many details of the manner in which the apparatus is fitted up in the lighthouses at Lizard Point. This would be very interesting to describe, but for want of space we shall at present content ourselves with considering the way in which the electric light is arranged at the top of the lighthouses. The illuminating part of a lighthouse is constructed, as shown in Fig. 68, of a glass lantern formed of a certain num- ber of Fresnel's lentilles-a-echelons (lenses in steps), in the centre of which the luminous focus is situated. This lantern is turned by powerful clockwork, and it is the passage of the separating zones between the different lentricular parts that produces those eclipses by which the light of a lighthouse is distinguished from that of an ordinary fire. The smaller the luminous point the more the effect is magnified by the lenses, and in order that the light shall be visible from a great dis- tance it is essential that the luminous focus should be as bright and as small as possible. Now, the electric light solves this double problem, and therefore it appears expressly made for lighthouses. Nevertheless, as the electric light regulators are occasionally liable to extinction, and a prolonged ex- tinction might cause serious disasters, the regulators (gene- rally Serrin's or Siemens') are arranged in duplicate. They slide into the lantern on small rails placed on the surface of a cast-iron table, a stop arrests them when at the focus of the lenses, where they instantly light themselves, and this is one of the great advantages of the electric light, espe- cially with the regulators to which we refer. - The electric communication is established on one hand by means of the cast-iron table, and on the other by a metallic spring which presses against the upper part of the lamp at a convenient point. The substitution of one lamp for another does not require more than two seconds, the one which is withdrawn pas:>mg out by one of the railways whilst its substitute enters by the other. The light may also be made to pass instantly from one apparatus to the other by means of a APPLICATIONS OF THE ELECTRIC LIGHT. 261 FIG. 68. 262 ELECTRIC LIGHTING. commutator, but to centralize the two points of light is a more difficult matter. The carbons used in lighthouses are 7 millimetres wide and 27 centimetres long, and their rate of consumption is estimated at 5 centimetres per hour at each pole, at least with machines giving alternating currents. In spite of this uniform consumption there is, however, a slight difference., and the upper carbon is consumed a little faster than the lower, in the ratio of 108 to 100. The regulators are there- fore very well adjusted, but as it is important that the varia- tion of the luminous point shall be under 8 millimetres, without which the rays will not be sent out to the horizon, it is necessary to carefully attend to this light In order that the keepers may be able easily to observe the progress of the carbons, an image of them is projected on the wall by means of a lens of short focus, a horizontal line is marked on the wall, and the points must be equally distant from that line. As a deviation of i millimetre is represented by a deviation of 22 millimetres on the wall, defects in the adjust- ment are readily seen. This apparatus began to work at the lighthouse on the south cape of La Heve on the 26th December, 1863, and after fifteen months of experiments it was decided -to apply the same plan of lighting to the second lighthouse. From that time electric lighting was definitely established. As to the machines, which, like regulators, are provided in duplicate, they are generally placed in the base of the lighthouse tower, with the steam-engines for driving them, and well-insulated cables of a large diameter convey the current to the regulators, as we have already said. According to Le Roux, it seems that even with the Alliance machines of 4 discs, the cost of the light is, on the average, one-seventh of that of oil. In the natural state of the atmosphere the Alliance machines with 4 discs give a light visible at 38 kilometres, and those with 6 discs have a range of 50 kilometres ; but it is curious APPLICATIONS OF THE ELECTRIC LIGHT. 263 that in foggy weather the electric light does not carry farther than that of oil lamps. There are no.w several electric lighthouses in France, Eng- land, Russia, Austria, Sweden, and even in Egypt. Their working is everywhere satisfactory. Application to the Lighting of Ships.— One of the most important applications of the electric light is for lighting ships in their course so as to avoid fouling, and show the en- trances to ports in the night. The earliest attempts were made with the magneto-electric machines of the Alliance Company, and although the results were not entirely satisfactory, they were sufficiently successful to show that a solution of the problem would be effected in the immediate future.* The inconve- niences for which the plan was blamed may thus be summed up: — The electric light produces round it a whitish cloud, fatiguing to the eye and injurious to observations; the fixed electric light, by its great intensity, obliterates the regulation green and red lights, which is a source of real danger; near the shore, ships might mistake the electric lantern for a light- house, and take a false course; finally, the apparatus is cum- bersome, and the cost of fitting it up too great for the service it renders. The greater part of these objections have lately been re- moved, by raising the luminous lantern to a certain height, by making the light intermittent, and by using the Gramme * The first attempts of the Alliance Company, at that time directed by Berlioz, were made as early as 1855, on board the .Jerome-Napoleon, whose commander, M. Georgette Dubuisson, was a strong supporter of the system. They were afterwards repeated on board the Saint-Laurent, the Forfait, the d ' Estrte, the Heroine, the Coligny, the France ; and it may be seen by the Reports in Les Mondes, tome XVI 1 1., pp. 51, 325, 458, 593, 637; tome X VI. t pp. 488, 594 / tome XIII., pp. 177, 405, 423; tome VIII., p. 592, that if the naval service in general attached little importance to this application, several distinguished officers fully appreciated its value. At that period, it is true, the electric lighthouses that have given such good results on board the Amtrique had not been fitted up on ships, but an electric light lantern, very ingeniously arranged, was fixed on the mizzen-mast, and thus removed one of the principal objections that had been made. 204 ELECTRIC LIGHTING. machines, which occupy a small space and are not costly. It was on board the transatlantic steamer EAmkrique, and under the direction of Captain Pouzolz, that this new system was first established ; and it appears to have succeeded per- fectly. Fontaine gives the following details on this subject :— " The lantern is placed in the upper part of a turret ascended by internal steps, so it is not necessary to go on the bridge, for this turret rises above one of the companion-ladders. This arrangement is very advantageous, especially during heavy weather, when the bow is with difficulty accessible from the bridge. The turret was at first 7 metres high, but Pouzolz had it lowered by 2 metres to increase its stability, and to lower the level of the luminous beam ; so that this turret is now 5 metres above the bridge. Its diameter is i metre, and it is placed in the fore part of the steamer at 1 5 metres from the stern. " The lantern properly so called has prismatic glasses ; it is able to illuminate an arc of 225° while leaving the steamer almost entirely in shade. The regulator, which is on Serrin's plan, is hung to the cardan. A small seat in the upper part of the tower allows the attendant to regulate the lamp. The luminous beam is about 8 diameters wide. " The Gramme machine which supplies the luminous arc has a power of 200 Carcel lamps, and is driven by a motor on the Brotherhood system, which reduces the space occupied by the two to i '20 metres by o'6o. These two machines are placed on a false floor in the engine-room, 40 metres from the lantern. " All the wires pass through the captain's cabin, who has under his control commutators by which he can at will turn the light on or off in each of the two lamps without stopping the Gramme machine. " The novelty of the arrangements on L'Ame'rique consists in the automatic intermittance of the light in the lantern. This intermittance is given by a very simple commutator attached to the end of the axle of the Gramme machine, which has the effect of alternately sending the current into the lamp, and into a closed metallic bundle of the same resistance as the voltaic arc, which bundle is heated and cooled alternately. This arrangement was adopted in order to keep the Gramme machine, which makes APPLICATIONS OF THE ELECTRIC LIGHT. 265 850 revolutions per minute under always the same conditions as regards the external circuit. According to Pouzolz's calcula- tions, the best relation between the eclipses and appearances of the light would be a light of 20 seconds and an eclipse of 100 seconds. "The height of the luminous focus is 10 metres above the water, and the possible range of the light, in consequence of the depression of the horizon, is 10 marine miles (18,520 metres) for an observer with his eye at 6 metres above the water. " In order to light up the topsails and the gallant-sails while the low sails are left in obscurity, Pouzolz had made a frustum of a cone in tin-plate, and placed it in the movable lamp, with the large opening outwards. In this way the Ameriqut was visible at a great distance from ships and signal-stations when the captain allowed the electric light to continue in action during the whole night." It will be seen by this description that all the objections offered to the use of the electric light on ship-board have been removed by this new arrangement, and Pouzolz answers those which have been made to the use of an intermittent light by stating that the light produced by short flashes has never incommoded the sight of any officer of the watch or look-out man at the cathead, and that the brilliancy of the green and red side-lights is not at all diminished by the use of the lighthouse in front. Since the dreadful collisions which have taken place within the last four years, there is now more inclination to resort to the electric light for ships ; and we see that, according to Fontaine's book, in 1877, a certain number of Gramme machines have been set up on board several French, Danish, Russian, English, and Spanish ships of war, among which we may mention the Livadia and the Peter the Great of the Russian navy, the Richelieu and the Suffren of the French navy, and the Rumancia and the Victoria of the Spanish navy. It remains to describe the projecting apparatus, which, on account of the small space over which the light is to be thrown, is different from the lenticular apparatus of 266 ELECTRIC LIGHTING. lighthouses. This apparatus is not essentially different from that which was put up on board of the Jerome Napoleon. This FIG. 60. consisted of a parabolic reflector, in the focus of which was the voltaic arc produced by a Serrin regulator. This reflector, APPLICATIONS OF THE ELECTRIC LIGHT. 267 ,a little prolonged in front, was closed by a Fresnel lens, in order to transform the divergent beam into a parallel one. Finally, behind the regulator and the reflector was a small spherical reflector. The whole was mounted in a chamber movable on a pivot which, by a lever and rotating platform, per- mitted the beam to be sent in any direc- tion. Further, a ma- rine telescope fitted to the apparatus enabled the points of the horizon lighted by the beam to be observed. By placing in front of this beam coloured glasses, the light could be colour- ed green or red, and thus made suitable for marine signalling. In Sautter and Le- monnier's projector, represented in Fig. 69, the spherical and parabolic reflectors do not exist, and the whole consists of a Fresnel lens composed of 3 droptric and 6 catadroptric elements. This lens is enclosed in a wide cylindrical tube, which, being supported on a pivot with the whole electrical arrangement, can be turned in any desired direc- tion. In Siemens' projector, represented in Fig. 70, the para- bolic reflector is placed behind the lamp ; and the latter is FIG. 70. 268 ELECTRIC LIGHTING. also provided with two lenticular arrangements for projecting the image of the carbons on a screen, thus facilitating their adjustment. Application to Nautical Signals at long range.— Nocturnal signals exchanged between the various ships of a squadron are often inefficient on account of the feebleness of the light, and it would be desirable to make them clearer and visible at a greater distance. To accomplish this, De Mersanne arranged a particular system of regulator, which could not only be controlled at a distance, but was also re- gulated without requiring the presence of an attendant near the instrument. This regulator has its carbon-holders mounted on two vertical rods provided with a screw movement, and capable of turning round on their own axes by an electro-magnetic mechanism controlled by a commutator. The apparatus is enclosed in a large lantern provided in its central part with a cylindrical system of " lenses-in-steps," at the focus of which the luminous arc is placed, and which is so arranged as to direct the light according to the height which the beam has to reach. Now it is in order to always keep the luminous point in the right place that the above-mentioned electro- mechanism is applied. This is composed of two straight and two horse-shoe electro-magnets, arranged in two perpendicular lines in a vertical plane. In the centre of these four electro- magnets there is on a forked armature a lever provided with a steel tooth, which passes between two parallel ratchets inversely disposed to the lower end of the two rods of the carbon-holders. When no current is passing in the electro- magnets the tooth is exactly between the two ratchet-wheels; but if, by means of the commutator, the current is passed through one of the straight electro-magnets— the upper one, for example — the lever already mentioned is raised, and the tooth at the end enters between two teeth of the upper ratchet-wheels, without, however, producing any effect ; and APPLICATIONS OF THE ELECTRIC LIGHT. 269 it is only when the current is made to pass through the right electro-magnet that the latter causes the lever to turn, and pushes the tooth by one notch. The screw rod of the regu- lator then turns, by an amount proportionate to the escape- ment of this tooth, and lowers the corresponding, carbon- holder. If now the lower straight electro-magnet be excited, the tooth of the lever engages the lower ratchet-wheel of the rod, and when the current is' sent through the left electro- magnet the rod in question turns by the space of a tooth of the ratchet-wheel, but in the direction opposite to the former movement; and this causes the carbon to be raised that before was lowered. As the other carbon is capable of being similarly acted upon in the same way, the luminous point can thus be placed at any desired elevation, at what- ever distance the operator may be from the regulator ; and he can separately or simultaneously move both carbons as occasion requires. As to the signals, there are two methods of proceeding. Either the light of the signalling apparatus maybe extinguished by a commutator, or the lights may be hid by a screen elec- trically made to descend in front of the arcs. In the latter case the apparatus are fitted up with a special system of electro- magnets, by which the movement is easily effected. De Mer- sanne has fitted up several patterns, which may be applied to any other kind of regulator; the problem is not one of any difficulty. The signalling apparatus just described has been made to work by hand ; but the regulation of the light may obviously be automatically produced in a very simple way, by causing a mechanism connected with the light-producing current to act on the commutator already mentioned. One small detail in the construction of the commutator is of some importance. It is a platinum wire which glows when the lamp itself is lighted, and is extinguished with the latter. The person sending the signals is therefore aware, although- he may not see the lamp, when this latter is lighted. 270 ELECTRIC LIGHTING. Application to the Arts of War.— The extreme in- tensity of the electric light, and the ease with which it can at will be made to instantly appear and disappear at a dis- tance, render it capable of important applications in military operations, either for signals, or as a means of illuminating at night a distant point to be observed, or to light up the work of the assailants in sieges. Martin de Brettes published, twenty years ago, an interesting paper on this subject, and this we have reproduced in the second edition of our Expose des applications deT Electricity tome III., page 258. We have here space for only a few extracts : " Signals in the field or at a siege," says Martin de Brettes, " are chiefly intended for the transmission of orders or urgent despatches. It is therefore clear that the best system of lumi- nous signals is that in which the light is most simply produced, is seen from the greatest distance, and has the greatest regula- rity in the appearance of the lights combined to produce the signs required in a telegraphic correspondence. "As to the property possessed by the electric light of being seen at a considerable distance, its superiority for a good system of signals cannot be disputed. Nevertheless, rockets may in general, under ordinary circumstances, be advantageously em- ployed on account of their simplicity, the ease with which they can be carried about and used. But when a powerful permanent luminous signal is required, the electric light will be of great help, and may prevent the use of a captive balloon in the field. " Again, circumstances occur in war in which an illumination of a longer or shorter duration is required ; for instance : " To reconnoitre a fortification, the besieger requires a mo- mentary light sufficient for his purpose, and not so prolonged as to attract the attention of the besieged. " To direct the fire of a battery on a given point, that point must be lighted up long enough to allow a good aim to be taken. "In order not to be taken by surprise by the opening of trenches, the besieged should continuously light up the ground where that operation is likely to be executed. " The lighting up of a battle-field or of a breach a': the time of the assault, requires also an illumination of an indefinite duration. APPLICATIONS OP THE ELECTRIC LIGHT. 27 1 " Thus there may be required in war, either a momentary illu- mination, or one prolonged for perhaps the whole night. We have already seen that these two illuminations can be readily produced at will by the electric light, by closing or interrupting the voltaic circuit." Martin afterwards explains the conditions for applying the electric light so as to obtain these results. At the time when his paper was written, however, the light could not be produced by electro-magnetic machines, and it would have been necessary to work with the cumbersome materials of a battery, which made the problem one of much greater difficulty. Now, thanks to the small dimensions of the magneto-electric machines, very considerable luminous inten- sities may be obtained, and this kind of application of the light becomes an easy matter. The magneto-electric — or, best of all for this purpose, the Gramme — machine, may be mounted to a portable engine, which can be as easily moved about to any required position as the cannons. The system used in France is driven by a Brotherhood three- cylinder machine. The electro- magnets of the Gramme machine are thin, flat, and very wide ; the bobbin has two current collectors, and a commutator mounted on the arma- tures allows the machine to be joined up for tension or for quantity. This plan, as shown in Fig. 71, has been adopted by France, Russia, and Norway. According to Fontaine, it was found, by experiments made at Mount Vale'rien with a machine thus arranged, that an observer, placed beside the apparatus, is able to see objects 6,600 metres distant, and to distinguish clearly details of construction at a distance of 5,200 metres. To obtain these results, the Gramme machine must have a power of 2,500 lamps, and the projector must concentrate the light by re- flection and refraction, as in those projectors we have de- scribed for lighting ships at sea. When the electro-magnets of the machine are joined up for quantity, it turns with a speed of 600 revolutions per 272 ELECTRIC LIGHTING. minute, and expends 4 H.-P. • the light produced varies from 1,000 to 2,000 lamps. In the second case it makes 1,200 turns, employs 8 H.-P., and gives light equal to from 2,000 to 2,500 lamps. When the weather is clear, the machine is joined up for quantity, and the expenditure of steam is then FIG. 71. small, the working easy, and the carbons are consumed slowly. When the weather is foggy and thick, the machine is arranged for tension, the expenditure of steam increases, the working requires rather more care, and the carbons are consumed quickly. With the Brotherhood motor the change of power is effected instantaneously. For war signalling, Gramme has fitted up a small machine which can be worked by a man's arm. This machine. APPLICATIONS OF THE ELECTRIC LIGHT. 273 worked by four men, gives a light equal to 50 Carcel lamps. The French Government has lately had it tried. Experiments with machines arranged in nearly the same way were made in Berlin in 1875. The light produced was intense enough at a mile's distance to allow ordinary writing to be read. When a mirror placed in front of the regulator was inclined so as to reflect the rays upwards, a luminous track was thrown upon the clouds, and from a distance appeared like the tail of a comet, in which the signals made before the^ mirror showed themselves. Mangin's projector is the one adopted in France for mili- tary operations, and it is in arrangement somewhat similar to that of Siemens', which we have shown in Fig. 70. It is mounted on a low light truck, by which it can easily be taken wherever required. This apparatus, described in detail in the Memorial de I'officier du genie (Engineer Officer's note- book), is composed essentially of a concavo-convex glass mirror with spherical surfaces of different radii. The con- vex face is covered with silver, and reflects. This mirror, go centimetres in diameter, has the property of being free from spherical aberration, notwithstanding its diameter and its focal length being nearly equal. Between the mirror and the luminous arc is a concavo- convex lens with its concavity towards the light. Its use is to collect upon the mirror a greater number of the rays, and thus increase the amplitude of the field of illumination. The beam leaving the apparatus when the lamp is at the focus of the lens is exactly bounded by a circumference almost free from penumbra, and without other divergence than that due to the dimensions of the source of light, which divergence is about 2\ degrees. The light is uniformly dis- tributed over the whole surface. This apparatus also possesses the property of being made at will, either to light up a considerable space or to concen- trate its intensity on one point, and this renders it extremely suitable for certain military operations. A simple displace- 18 ELECTRIC LIGHTING. ment of the luminous point produced by a screw effects the change. In some other experiments made with these apparatus at Toulon and at Cherbourg an unexpected fact was esta- blished, namely, that when the concentrated beam is pro- jected upon a ship, the pilot has much difficulty to enter a port. This is a new means of defence. It has been proposed to send signals from captive balloons. In this case the signal regulator of De Mersanne may be advantageously employed. Lighting of Railway Trains. — The intense brilliance of the electric light, and the easy method of throwing it in any direction, have suggested in its employment for lighting railway trains running at night, and for announcing them at a greater distance were it only by the illumination of the sky at the place they are passing. Experiments made on the Chemin de fer du Nord have been perfectly successful, and seem to indicate that this plan of lighting will one day become general. In the meanwhile Girouard has invented the following system. The electric generator, or Gramme machine, is fitted up on the tender, and is driven by a toothed wheel moved by an independent piston fixed on the lower part of the frame. One of Watt's governors controls the admission of the steam. A copper tube connected with a cock fixed on the engine is coupled to a pipe leading to the slide valve of the motor cylinder. In order to protect the magneto-electric appa- ratus from rain and dust, it is enclosed in a wooden casing, and only the cylinder remains outside. It will easily be seen that this arrangement is very solid, although indepen- dent of the engine. Its parts can be attended to by the person who usually cleans the engine. In front of the locomotive there is firmly fixed a lantern containing an electric lamp provided with a powerful re- flector, and in front of the lantern is placed, at an angle ot APPLICATIONS OF THE ELECTRIC LIGHT. 275 45 degrees, a semi-transparent platinized plate of glass. This glass is mounted in a frame so arranged that it can be turned a little to the right or left, while still inclined at the same angle. Further, a frame containing three coloured glasses, red, white, and green, is placed in front of the reflector, and serves at the same time to protect the lantern from wind and rain. A jointed rod proceeds from the frame of the inclined glass, and another from the frame carrying the coloured glasses, and these rods are connected with two small levers within reach of the engine-driver. The lamp is connected with the magneto-electric machine by two cables, and when the current passes in the lamp the luminous rays are thrown forward by the reflector, but, as the glass is slightly platinized, only a portion of the beam proceeds straight forwards, whilst the other is projected upwards towards the sky in the form of a cone. By means of the first lever this cone can be turned obliquely either to the right or to the left, while the forward illumination still continues, and by means of the second lever the rays can be coloured either red or green. Now, by giving a certain meaning to each combination, a considerable number of signals may be formed. Besides this, although a train may be passing through a deep cutting or be hidden from sight by curves and inclines, or although the direct view of it may be intercepted by bridges or other objects, the beam projected vertically indicates its position at a great •distance. Application of the Electric Light to the Lighting of the Drifts in Mines, &c.— Several men of science, and among the rest De la Rive, Boussingault, and Louyet, have laid claim to having thrown out the first suggestion of the use of the electric light in mines. The idea originated, as seems to me to have been proved, with Louyet; but the application of it was certainly not made until 1845 by Boussingault. 1 8— 2 276 ELECTRIC LIGHTING. Everybody knows of the dangers miners are exposed to when gas, issuing from the beds ofxoal, comes into contact with the flame of a lamp. A dreadful explosion takes place, and fires the whole drift. These sad accidents are called explosions of fire-damp. Now as the electric light is inde- pendent of a supply of air, for it can be produced in a vacuum, the danger of fire-damp will obviously be avoided, by enclosing each lamp in hermetically-sealed globes, placed in the drifts where the miners are working. It will, however, "be necessary to exhaust these globes, lest the dilation by heat of the enclosed air should break them. There is then not the least danger to be feared, for the luminous points are thus completely separated from the external air. In order to avoid the considerable cost of setting up the electric light, Dumas and Benoit conceived the idea of sub- stituting the light of the inductive spark in a vacuum; they therefore arranged the vacuum tube spirally, and placed it in an outer tube, provided with copper fittings for suspending. The exhaustion is made on Morren's gas, in order to obtain a fine white light. I have spoken at length about this kind of illuminating tubes in my account of Ruhmkorff's induction apparatus (5th edition), and to this I refer the reader. The electric light produced by the Alliance machines was successfully used in 1863 by Bazin for lighting the slate quarries of Angers. A machine with 4 discs was capable of lighting a gallery 60 metres long, by 50 metres wide and 40 metres high. The machine was near the opening of the shaft, and the electric current was transmitted by wires 150 metres long. In spite of the diminution of intensity oc- casioned by this great length of wire, the illumination proved so satisfactory that the workmen expressed their delight by loud applause. These advantageous results were confirmed on several different occasions, and it was found that the effective labour of the workmen was increased by a fifth or a sixth — a net advantage of 15 or 20 per cent, to add to the comfort of the workmen, which it was desirable to secure APPLICATIONS OF THE ELECTRIC LIGHT. 277 even at a high cost. There were, however, only two points of light. (See Les Mondes, tome /., page 691, and tome //, pp. 221 and 278). The electric lighting of the slate quarries at Angers has lately been provided for in a permanent manner by Lorain. For this purpose a Gramme machine of the pattern described on page 75, and two Serrin regulators, have been used. The large subterranean gallery thus lighted up is no less than 100 metres in length, with a width varying from 15 to 50 metres, and a height of 60 metres. The whole of it — walls, vaulted roof, and floor — is black, yet in spite of the absence of reflection from its surfaces, it is illuminated almost as well as it would be in broad daylight under a clear sky, and with great satisfaction to the workmen and advantage to the Company. A speed of 800 turns per minute would appear to be sufficient for obtaining a splendid light with the Gramme used, if the regulator had been placed near it ; but on account of the depth of the quarry the regulator was about 350 metres from the source of electricity. In order to obtain a good light and a regular working of the apparatus, it was requisite to give the machines a mean and nearly constant velocity of I>135 revolutions per minute. Conducting wires of a greater diameter are about to be used, and this will allow of such a reduction of the velocity that the electro magnets will not become heated beyond 50°. The machines have been running simultaneously and con- tinuously day and night since they were put up, eight months ago. With regard to duration and continuity, this experiment is the most conclusive that has hitherto been made with the Gramme machine, which has victoriously stood this severe test. (See the journal La Lumiere ekctrique for i5th May, 1879.) The employment of the electric light for the illumination of works carried on at night was one of the first useful appli- cations of this method of lighting, and, dating from the works 278 EL ECTRIC LIGHTING. of the Pont Notre-Dame, where it was brought into use for the first time, it has always been employed whenever any important work had to be expeditiously performed. It has, for instance, been made use of in the works for the Docks Napoleon, in the rebuilding of the Louvre, &c., &c. In such applications the lantern is commonly set up at the top of a wooden post, and is furnished with a reflector for throwing the light downwards. It has also been proposed to apply the electric light in field labours in order to expedite harvest operations. Albaret, the head of a great firm of agricultural implement manufacturers at Liancourt, has lately made at Mornant and at Petit-Bourg, experiments that have proved successful. The apparatus (described in the journal L'&lec- tricite of the 5th Sept., 1878) is composed of (r°) a portable steam-engine ; (2°) a dynamo-electric machine of some kind or other ; (3°) a post made of iron bars, serving to carry the lantern and the lamp, and fitted to the portable engine. The engine may, if it has sufficient power, be utilized to drive a thrashing machine. A windlass in front of the chimney enables the post to be raised or lowered. Application to Lighting Railway Stations, Work- shops, &c.— The electric lighting of large workshops and railway stations is now un fait accompli. Following Her- mann-Lachapelle, who was one of the first to enter upon this course, a multitude of other manufacturers now use it with very great satisfaction. Fontaine's book tells us that Gramme machines now illuminate the establishments of Ducommun, at Mulhouse; of Sautter and Lemonnier, at Paris ; ofMenier, at Crenelle, Noisiel, and Roye; the spinning-mills of Dieu- Obry, at Daours; of Ricard fits, at Mauresa (Spain); of Buxeda, at Sabadell (Spain) ; of David, Trouillet, and Ad- hemar, at Epinal ; of Bourcard (Doubs) ; of Horrocks and Miller, at Preston; the weaving-shops of Gr^goire, at Creve- Coeur-le-Grand ; of Manchon, at Rouen; of Brindle, at Pres- ton; of Mottet and Baillard, at Rouen: of Isaac Holden, APPLICATIONS OF THE ELECTRIC LIGHT. 279 at Rheims; the workshops of Coron and Vignat, at Saint- Etienne; of Maes, at Clichy; of Descat-Leleu, at Lille; of Pulher and Sons, at Pesth; of Carel, at Ghent; the yards of Jeanne Deslandes, at Havre; the workshops of Mignon, Ronart, and Delinieres at Montluc,on; the quay of the canal between the Marne and the Rhine, at Sermaize; the goods station at La Chapelle, Paris; and also the different places undertaken by the Jablochkoff Company mentioned in pages 228-9.* The result has everywhere been satisfactory. Fontaine's work gives details on the fitting up of these systems of lighting. We shall here merely describe that of the station of the Chemin du fer du Nord, on account of the ingenious method by which a light is obtained not fatiguing to the eye, and capable of illuminating the various parts of the hall by nearly perpendicular rays, thus obviating the shadows thrown by packages, and throwing over them a flood of light like that of the torrid zone. This method consists in arranging round the regulators, which are hung up at various points of the halls, a kind of reflector partly formed by the support of the lamp and partly by a sort of inverted ground-glass funnel, so placed that the luminous arc cannot be seen directly from any part of the hall. The light thus partially stopped is reflected towards the ceiling, as well as that which proceeds from the upper part of the arc; and as the ceiling is painted white, it is capable of forming in its turn an immense reflector, which sends down the luminous rays almost vertically, and thus * Besides these establishments, Fontaine mentioned, in the beginning of 1877, a number of other workshops lighted in this way. Among the rest were the cannon foundry at Bourges ; the workshops of Cail, those of the Mediterranean Iron Works Co. at Havre, those of Crespin and Marieau at Paris, of Beaudet at Argenteuil, Thomas and Powell at Rouen, Ackermann at Stockholm, Avondo at Milan, Quillacq at Anzin ; those of Fives-Lille, of Tarbes, of Barcelona ; the Midi Stations at Brussels ; the workshops at Fourchambault ; the foundries of Besseges and of Fumel ; the dye-works ot Guaydet at Roubaix, of Hannart at Wasquehal ; the weaving shop . ot Baudot at Bar-le-Duc ; the laundry of the Lyons hospitals, &c. 28o ELECTRIC LIGHTING. prevents the very dark shadows which would be cast by the packages. By the adoption of this method, it has been found possible to reduce the number of men on the staff for night services, and the number of small articles lost has been much diminished. E. Reynier has improved this system by making the reflecting apparatus and the lantern which causes the regulator much easier to manage. In his arrange- ment the apparatus is moved like the suspenders in a dining- room. We have not space here to give the details of this interesting arrangement, but we shall, at a later period, pub- lish a more complete description. The Gramme Company has also fitted up for the shops of the Louvre a luminous ceiling, which has proved equally successful. The ceiling is formed in the first place of a large plate of ground unsilvered glass, which constitutes the base of a large hollow pyramid of tin-plate intended to act as the reflector ; an electric light regulator suspended and balanced by a counterpoise is placed within this pyramid, so that the reflected rays may be thrown as uniformly as possible on the ground glass. A second reserve regulator can, moreover, be readily substituted for the first one when the carbons have to be renewed. The light reflecting system, used at the goods station of the Chemin dufer du Nord, has been employed at Vienna to illuminate a skating rink 133 metres long. Two Gramme machines and two Serrin lamps, above which were hoisted jarge reflectors of elliptical curvature, were enough to light up the rink splendidly. This is the most successful of open air applications. Splendid electrical illuminations have recently been pro- vided at the picture exhibition in the Champs-Elysees at Paris, and at the exhibition of electric light apparatus at the Albert Hall, London. The former have been produced by Jablochkoff candles, the latter with these candles combined with Siemens' lamps. The results, although leaving room for some improvement as regards the steadiness and the APPLICATIONS OF THE ELECTRIC LIGHT. 281 appearance of the illuminated objects, have shown the im- mense resources this method of lighting places in our hands. Besides the Chatelet Theatre in Paris, which has for several months been partially lighted by the Jablochkoff system, the Gaiety Theatre in London has been lighted throughout the winter by the electric light, the apparatus employed being Lontin's. To dismiss the subject of electric illuminations, we may here state the results, as noted in the American newspapers* of the lighting of Cleveland Park, where the Brush system was used, According to these newspapers, this system of lighting must have been more economical than those tried in England and France in the proportion of 4 to T. It is asserted that this illumination is supplied by twelve Brush lamps of the pattern described on page 178, excited by one of the same inventor's machines not requiring more than 1 1 H.-P. to drive it, and that each ot these lamps has a luminous power of 2,000 candles, or 200 Carcel lamps. These 12 lamps are represented to have advantageously replaced the 100 gas-jets which had hitherto been used, and the amount of light given off is stated to have been about three times that produced by the gas. We think this account is exaggerated, for if each Brush light represent 200 gas-jets, the 12 should be equal to 2,400 jets. Now, we see that the total illumination is only three times better than the original lighting, which was 100 jets; the Brush illumination would therefore represent only 300 gas-jets, or 25 for each electric lamp. We must here mention, as important adoptions of the electric light, those recently made at the South Kensington Museum, and at the Jardin d* Acclimataiion at Bordeaux. There the Werdermann system has been used, and apparently with much success. Application of the Electric Light to Fishing.— It has p«t yet been decided whether the electric light 22 ELECTRIC LIGHTING. plunged beneath the surface of the water, attracts fish or drives them away. According to some persons, it would be a means of making miraculous draughts, and Jobard, of Brussels, published in 1865 a very ingenious paper on this application ; but we must unfortunately dispel the illusions which were formerly cherished. J. Duboscq has in fact con- structed, to the order of an Anglo-French nabob, a large globe lamp for the electric light, with which experiments were made one fine summer evening on the lac d'Enghitn: the waters were perfectly illuminated, but the fish, instead of coming towards the light, avoided it in alarm ; not one was seen, so the apparatus has been useless. This discomfiture is described as above by FAbbe Moigno ; but we observe that his opinion was not very conclusive, for we read in the journal of Les Mondes, t. XII. , p. 46, /. VI., p. 584, andt. V., p. 374, articles on fishing by the electric light, in which he attaches more importance to the matter. In fact, he quotes an article stating that Fanshawe had been very successful in this way, catching by bait many whitings and mackerel. According to this amateur fisherman, the appearance of the sea during the experiment was splendid ; the reflected light carried the greenish-blue colour of the water from the bottom to the summit of every wave. The sails and rigging of the vessel were also illuminated, and it seemed as if it were floating in a sea of gold. The silvery fish darted all around and constantly rose towards the surface of the illuminated water, presenting the appearance of brilliant jewels in a sea of azure and gold. It is true that in another article the author of the Les Mondes describes the experiments made at Dunkirk with a submarine lamp excited by currents from an Alliance machine, experiments which have left much uncer- tainty on the action of the light on fishes. Electric lights have, however, been constructed for fishing, and Gervais has, according to the journal Les Mondes of the 3oth March, 1865, a rather ingenious one, which is attached to a buoy, and can be let down to any required depth. APPLICATIONS OF THE ELECTRIC LIGHT. 283 Application to Submarine Working. — Since diving, bells and various other apparatus for supporting respiration under water have made it possible to work at the bottom of the sea, several kinds of hydraulic work, and numerous re- coveries of sunken vessels have been executed with ease. When the depth of the water which is to be entered is not great, daylight readily penetrates the liquid layer and affords sufficient light to the workers'; but at a certain depth the light fails, and the submarine explorations, which must always precede the working, become impossible. No doubt, by fitting a lantern with apparatus for renewing the air, a light may be maintained as men's respiration is maintained ; but this necessitates a supplementary pump and special apparatus to prevent the current of air from extinguishing the light. With the electric light the problem is solved in the simplest manner, and the extent of the space illuminated is much greater. The regulator with a globe, which we have already mentioned, may be used, or a special regulator to give the light directly in the water. However, as the light is in this last case much more difficult to control than when it is vacuo, the former method is to be preferred. The experiments made at Dunkirk on fishing by the electric light have allowed the way in which the light behaves under water to be examined, and it has been found that magneto- electric machines, as well as the light they produce, are cer- tainly applicable to submarine working. At a depth of 60 metres the light remained quite steady, and it illuminated a very large surface. The machine, moreover, was placed at more than 100 metres from the regulator. The surface of the glass in the lantern remained perfectly transparent, and the consumption of the carbons was less than in the open air. Applications to the Projections on a Screen of Optical Experiments, Photographic Transparencies, &c,— There are many physical phenomena which require 284 ELECTRIC LIGHTING. to be projected on a screen in order to be visible to a whole .audience. There are certain of these (relating to the nature of light itself) which require an extremely intense light to show them. No doubt, with solar light and a heliostat, the problem may be immediately and cheaply solved. But more often than not, the sun is absent when he is required, and we are forced to forego those experiments which not only impart greater interest and attraction to a course of lectures, but which are much better understood and much better remembered when they have been strikingly presented to the eye. The electric light can be most successfully sub- stituted for the sun in this kind of application, and Duboscq's regulators have, as we have seen, been arranged purposely for that object. The apparatus for projecting the electric light consists of: ist, an arrangement for steadying the light, so that the con- sumption of the two carbons does not displace the luminous point ; 2nd, a closed lantern containing the regulator ; 3rd, a plano-convex lens, for making parallel the divergent rays -coming from the luminous point ; 4th, of a system of optical apparatus, which we cannot here discuss without departing from the subject of this work.* We shall describe only the lantern, as that is a consequence of the electric regulator. . Duboscq's lantern is formed of a kind of bronzed copper box, which surrounds the upper part of the regulator. To •economize space, the column of the regulator is enclosed in a sort of chimney in which the box terminates. To hermeti- cally close the box, small shutters, moved by rackwork, close t'he top and bottom of the box at the same time that its door is closed, so that the openings made in the instrument for introducing the regulator are completely shut. The inside of the lantern is provided with a reflecting mirror and two supports, on which two other mirrors can be fitted, in ordei to throw the light on the lenses of a certain apparatus called * See my account of the method of projecting the principal phenomena of optics by means of Duboscq's apparatus. APPLICATIONS OF THE ELECTRIC LIGHT. 285 \hQpolyorama, adapted to the lantern for certain experiments. Finally, in the side of the lantern is a small bull's-eye, with a violet glass, by which the condition of the light is examined. In order easily to regulate the position of the luminous point,, which in some experiments must be fixed in the most exact manner, the regulator is placed on a stand which can, by means of two screws, be moved in two rectangular directions (up-and-down and sideways). ' The projections may be made at any distance ; only they lose their brightness and clearness where the distance is out of proportion to the intensity of the light : 5 metres is com- monly the most suitable distance for the light from a battery of fifty elements. We show in Fig. 7 2 an experiment of this kind. The magic lantern gives, with the electric light and Levy's- or Favre and Lachenal's transparent photographic views, effects so striking that a spectator would fancy he is tran- sported to the very spot ; and such a perfection in the views is now attained that the objects sometimes seem to stand out in relief, as in the effects of the stereoscope. This method of projection is now much used commercially, and, besides Duboscq's apparatus, which are applicable to every kind of optical experiment, there are those of Molte'ni, which are ex- clusively adapted for this kind of application. Among the projection experiments that have been made by these apparatus, we may particularly mention that of the reading of microscopic despatches forwarded during the siege of Paris by carrier-pigeons ; these despatches, each of which covered less than a square millimetre, were easily read before the multitude of those who were interested in receiving news from the provinces. Fig. 73 represents this application of the electric light. The electric light has also been used for the photographing of places or objects not otherwise illuminated. In this way Levy has reproduced the pretty fountain under the staircase of the Grand Opera, and certain English and American 286 ELECTRIC LIGHTING. artists have succeeded in so truly reproducing the features and details of grottoes and dark caverns, that only by the shadows cast was it possible to distinguish them from places illumi- nated by daylight. Several photographers have desired to FIG. 72. use this means for copying and printing, and Pierre Petit and Liebert have even lately fitted up a complete electrical arrangement for taking portraits in this way. We read in the Scientific Correspondence of the i4th January, 1879, the following news : — i 288 ELECTRIC LIGHTING. "A. Liebert, the well-known distinguished artist, last Saturday invited the press to his artistic and elegant mansion in the Rue de Londres to witness photographic experiments by means of the electric light. We use an ill-chosen word in saying experiments, for they were not experiments, but a real and practical applica- tion of the electric light to photography. The sun is no longer indispensable; I believe Liebert has even dispensed with his services entirely. By the new system the studio is always ready to receive sitters, at midnight as well as at noon, and the opera- tions are carried on regularly and uninterruptedly. " Liebert obtains these interesting results by means of a very simple arrangement. A hemisphere of two metres in diameter is hung from the ceiling, so as to present its concavity towards the sitter. This hemisphere carries two electric light carbons, of which one is fixed, while the other is made movable by a screw connected with its holder. The carbons are brought together at right angles to each other. It is, in fact, an ordinary regulator, with only the difference that there is no mechanism, the carbons being brought together by hand as required by their consumption. At each posing of the sitter the two carbons must be placed at the proper point. The duration of the sitting is so short that this cannot fail during the interval. " The novelty and the improvement of the system consist in the circumstance of the light not falling directly upon the sitter. This light is first of all projected upon a screen, which in turn reflects it to the interior of the hemisphere, which is of dazzling whiteness, so that the luminous rays thus dispersed and divided surround the person whose portrait is to be taken. The illumi- nation is splendid ; the face is softly lighted, without any hard and exaggerated shadows. The sitter's eyes easily support the brilliancy of this light, and do not suffer from any unpleasant glare. "A dozen portraits were taken between 11 o'clock and mid- night with the greatest ease, and all were perfectly successful, to the great satisfaction of the guests so kindly invited by M. and Mme. Liebert. "The electric light used for this purpose is produced by a Gramme machine, driven by a gas-engine of 4 horse-power at the rate of 900 turns per minute." APPLICATIONS OF THE ELECTRIC LIGHT. 289 Public Trials of the Electric Light*— According to a claim advanced by Deleuil in the journal Lcs Mondes of the 26th November, 1863, it was his father who made the first experiment on the large scale with the electric light, and that in 1841 at the Qiiai Conti, No. 7. For this purpose he used a Bunsen battery of 100 cells, and produced the light between two carbons in an exhausted globe. Among the scientific men who were present at this experiment was Cagnard de la Tour, who was able to read a label in the crown of his hat at the base of the statue of Henri IV. Another experiment was made in 1842 by Deleuil pere, on the Place de la Concorde. Nevertheless, it was Archereau who in the first instance most contributed to popularize the electric light, and I shall ever bear in mind that the expe- riments made by him every evening, either in the Rue Rouge- mont or in the Boulevard Bonne-Nouvelle, determined my taste towards electrical science. This excellent pioneer of science I have, therefore, to thank for starting me in the career I have ever since pursued. Since these first public experiments many interesting trials of the electric light have been made, as at Rio Janeiro on the occasion of the anniversary of the independence of Brazil, frequently at London, and for two months the Avenue de I1 Imperatrice was lighted by means of two Lacassagne and Thiers lamps put up on the Arc de Triomphe de rEtoile. Wonderful experiments were also made at Boston, in 1863, to celebrate the victories of the Federal armies (see the full account of these fetes in Les Mondes, tome II., page 165) ; at the ball given at Paris in honour of the Emperor of Russia, equally brilliant experiments were made under the direction of Serrin ; they have also long been carried on at Le Car- rousel, at the Bois de Boulogne, at the Lac des Patineitrs, and in a multitude of other cases where people went to see the electric light as they would go to see fireworks. At the pre- sent time the novelty of all these effects has worn oft", and we are getting so tired of them that they attract only a 290 ELECTRIC LIGHTING. limited degree of attention. On the stage, however, this light produces its full effect, and the play of the Pommes de Terre Malades, in which it was used on the French stage for the first time, the operas of Le Prophete, Mo'ise, Faust, and Hamlet, and the ballets of La Filleule des Fees, La Source, &c., have shown what admirable resources this light has placed at the disposal of the scenic artist. Application of the Electric Light to Theatrical Representations. — The most striking effects that the electric light has produced on the stage have been contrived by Duboscq. For this purpose he has arranged in the new Opera-House a room set apart for the necessary bat- teries and engines. Without stopping to describe the effect o£ the rising sun in Le Prophete, which everybody at once admired, and which was produced by a mere upward move- ment of the regulator — a movement skilfully disguised by a number of more or less transparent curtains ; without speak- ing of the application of the voltaic arc for projecting a bright light on certain parts of the stage, in order to make groups or portions of the scenery stand out brilliantly, we may state that the intense rays of the electric light have served to reproduce upon the stage certain physical pheno- mena in their natural aspect, such as rainbows, flashes of lightning, moonlight, &c. This source of light is the only one which has proved intense enough to produce on the immense stage of the new Opera-House those effects of light and phantasmagoria which the public find so striking. According to Saint-Edme, from whom we borrow these details, the rainbow was produced at the Opera-House for the first time by Duboscq, in 1860, in the revival of Moise. The occasion of the appearance of the rainbow in the first act of that opera is well known. At first, bands of coloured paper in the curtain, representing the sky of Memphis, were illuminated by large oil-lamps simply. Afterwards came the electric light, but only the method of illumination was APPLICATIONS OF THE ELECTRIC LIGHT, 291 •changed, and it was not until many attempts had been made by Duboscq that a real rainbow was obtained, in the follow- ing manner : — " The electric apparatus supplying the arc," says Saint-Edme, "is placed on a stand of suitable height at 5 metres distance from the curtain, and perpendicularly to the canvas representing the sky. The whole optical apparatus is arranged and fixed in a blackened case, which diffuses no light externally. The first lenses give a beam, which afterwards encounters a screen cut out in the form of a bow. This beam is received by a double convex lens of very short focus, which serves the two purposes of increasing the curvature of the image, and of spreading it out. On leaving this last lens the rays traverse the prism, by which they are dispersed and made to produce the rainbow. The position of the prism is not a matter of indifference ; its angle must be at the top of the incident beam, or otherwise the colours will not be displayed in the order in which they appeared in the rainbow. By this method the rainbow appears luminous even when the stage is fully lighted. " It is no difficult matter to imitate peals of thunder in the theatre ; the shops supply tam-tams and elastic sheet-iron for this purpose ; but it is not so easy to make lightnings flash on the stage with anything like a natural effect. At first the pheno- mena was imitated by lighting a narrow zigzag cleft in the scene, with red fire from behind. With the progress of scenic art, ii was necessary to do something better, and by the aid of science the source of light was found in the voltaic arc, which is iden- tical in origin with the lighting itself. But what further had to be found, was some optical arrangement by which the luminous beam could be emitted and cut off at rapid intervals, while giving the zigzag movement characteristic of lighting. For this purpose Duboscq had recourse to a kind of magic mirror, in front of which the electric light was placed. This mirror was concave, and the luminous arc was situated at its locus. The upper carbon was fixed, but the lower carbon could at any re- quired moment be drawn back, when the light would flash out. This could also be done by electro-magnetic attraction, ana as the mirror was held in the hand it was possible,- by shaking it and using the commutator, to obtain currents in various direc J — 2 292 ELECTRIC LIGHTING. lions, by which the zigzags of flashes of lightning and their in- stantaneous apparition were imitated." Very curious effects may also be produced by Colladon's fountain illuminated by the electric light, by reason of the complete illumination of the jets, and various colours they may be made to assume. But the greatest sensation produced by the application of science to the theatre, has been the apparition of spectres on the stage amongst the actors. The reader will recollect the famous apparitions in the piece called Le Secret de Miss Aurore, which drew so many people to the Chdtelet theatre in 1863; and the performances of Robin and Cleverman are not so long past that one cannot recall the deep impressions produced by the spectres they raised and fought with. The whole secret of this display consisted in a plate of unsilvered glass placed in front of the actors, and inclined at an angle of 45 degrees to the stage. This plate of glass re- flected the images of living spectres, strongly illuminated by the electric light, who were placed in an opening below at the front part of the stage. These images were visible on all sides without intercepting the view of the objects, actors, or scenes on the other side of the glass. It was necessary to success, that the position of the persons representing the spectres should be so arranged that their images should appear vertical and seem to stand upon the floor, and also that their movements should accord with those of the actors on the stage; opening and closing the illuminating appa- ratus by means of a movable screen, would cause the appear- ance and disappearance of the spectral images. PART VI.— CONCLUSION. IF all that has been said in the foregoing work be mentally reviewed, and its logical conclusions be sought for, the question of electric lighting may thus be stated : — The peculiar character of the electric light resides in its concentrated power, by which an illumination equal to that from two to four thousand Carcel lamps may be given to a single point. This property may be extremely useful for certain purposes, particularly for lighthouses and ships, but it evidently is an inconvenience as regards public illumina- tion, and for a long time methods have been sought by which this brilliancy may be divided between several lights, in order not only to diminish the glare; but to extend the light over a larger space. Unfortunately, the methods which have been tried for effecting this division have solved the problem only at the cost of a great loss of the intensity produced by a single light But we shall see that by a well-known arrangement it may nevertheless be utilized under sufficiently favourable conditions. It is certain that if the electric arc gives a light too intense to be directly borne by Ae eye, it must be moderated by diffusing globes, by which much light is absorbed and simply lost. This is the case with the Jablochkoff candles, the glass enamel globes of which absorb as much as 45 per cent, of the light But if by any means the light could be so divided that these diffusing globes would be unnecessary, this loss would cease, and it might happen that with a suitable arrange- ment there would still be some advantage in using this system in spite of even a considerable loss of light compared with that produced by a single arc. In the first place, this mode of lighting does not involve, like others, a great heating of 293 294 ELECTRIC LIGHTING. the surrounding atmosphere ; and in the second place, there would be nothing to fear from the chances of explosions and fires, nor would the decorations of an apartment be spoiled. Besides, the white light does not change the natural hues of the illuminated objects, and this may be a great advantage for drapers' and other shops where the effects of colours have to be considered. Lastly, on account of the diminished risks, assurance companies will evidently be able to lower their rates. With regard to expense, it may be that electric lighting will prove cheaper than gas, although the trials already made seem to show the contrary ; but we must remember that these trials are not yet perfect, and we see already that since the erection of the JablochkofT system in the Avenue de rOper a, the cost of each electric light, which was at first said to be five times that of gas, was lowered by one-half in the estimate given by the company to the city of Paris, and we think this might again be halved so as to bring the cost to only 40 centimes per hour for each light. It may be said, it is true, that the cost of gas for an equivalent light is only 27 centimes; but let us suppose that the globes, which ab- sorb 45 per cent, of the light produced, stopped only 24 per cent., as Clemendot believes he can guarantee, the cost would fall below that of gas. These data, it must be under- stood, are merely approximative, and I quote the above figures only in order to show that it would not be impossible to produce the electric light at a cost within reach of prac- tice. In any case the Company which works the Jablochkoff candles has rendered an immense service by showing the possibility of lighting of public thoroughfares by the electric light, which had before that been doubted. We have to thank the initiative taken by the Company, and the beautiful experiments it instituted, for electric lighting having become a question of the day, and in every country new researches have been prosecuted, which will sooner or later lead to the solution of the problem. Several towns in Europe and CONCLUSION. 295 America are about to be illuminated by this system, of which we believe we have not heard the last word. It is already certain that a more complete study of the division of the light will lead to results more satisfactory than those already known. In order that some idea may be formed of the improve- ments within our reach, it will suffice for me to say that, in the investigations hitherto made, the various elements that play an important part in the magnitude of the effect pro- duced have not been sufficiently attended to. Thus, for example, a well-known relation between the resistance of the external circuit and that of the generator can greatly increase the proportion of the useful work. Nor should it be for- gotten that the intensity of the light varies in proportion ever so much greater than that of the intensity of the electric current. It is already known that the calorific action pro- duced by the current varies as the square of its intensity, but the resulting light varies in a still higher ratio; for, accord- ing to Preece, a platinum wire heated to 2,600° F. gives forty times as much light as when it is heated to 1,900° F. This explains why the division of the light is attended by so great a loss ; since with each weakening of the current resulting from this division, there is a loss of light which may, under certain conditions, attain the eleventh power of the ratio of the diminution of the current. All these considerations show that the solution of the lighting problem requires much further investigation before it becomes altogether practical ; but we believe that no one of the questions belonging to it is insoluble, and that before long we shall witness at least a partial transformation in public illumination. NOTES AND APPENDICES. NOTE A. ON THE INDUCTIVE ACTIONS IN THE NEW DYNAMO- ELECTRIC MACHINES. THE inductive actions resulting from the relative move- ments of the inducing and induced circuits are seldom studied with exactitude, and on that account very inaccurate theories of several recently invented dynamo-electric machines have been put forward. The following series of experiments may serve to fix our ideas on the subject : — Let us suppose that on a powerful straight magnet are wound several spires of an insulated wire, the ends of which are connected with a distant galvanometer, and let the coil so formed be capable of taking different positions on the magnet. If this coil is placed at the south pole of the magnet, and a soft iron armature is brought near that pole, a current will be obtained corresponding in duration with a magnetizing current, for it results from the increase of mag- netic energy communicated to the bar by the presence of the armature. This current will give a deviation of 12 degrees to the right, and on withdrawing the armature we shall have a second deviation of 12 degrees to the left. Therefore, in the following experiments a deviation to the right will repre- sent inverse currents, and a deviation to the left, direct cur- rents.* Let us now see what will happen from the various ':> It should be observed that the direction of the currents due to the in- crease or diminution of magnetic intensity is always the same, whether 296 NOTES AND APPENDICES. 297 movements given to the coil when it is passed from the poles towards the neutral line of the magnet, and from the neutral line towards the poles. This is what will be observed : — i°. When the coil is passed from the south pole towards the neutral line, a deviation of 22 degrees to the right will be obtained, and therefore an inverse current, or one of mag- netization. 2°. On making the reverse movement a new current will be produced, and will cause a deviation of 25 degrees towards the left, and therefore the current will be direct. 3°. If, instead of passing the coil from the neutral line towards the south pole, the first movement is continued by bringing the coil from the neutral line towards the north pole, a current will be obtained in the direction opposite to that produced in the first half of the movement, and if the coil is stopped half-way a deviation of 12 degrees to the left will be obtained. 4°. On bringing the coil back from the last position the currents are evoked at one or the other pole of the magnet, or at both together, and whatever may be the position of the coil on the magnet. But the currents produced will be the more energetic as the action takes place nearer to the coil. Thus, by placing the coil at the centre of the magnet on the neutral line, the current due to the increase of magnetization resulting from the approximation of an iron armature to one or other of the two poles will be inverse and of 2 degrees, and that which will result from the removal of the armature will be direct and of the same intensity. By acting simul- taneously on the two poles and developing the armature, these currents will show themselves in the same direction, and will attain an intensity repre- sented by 7 degrees. If the coil is placed at one of the poles, at the south pole for instance, the currents will be of 10 to 12 degrees, when the arma- ture is brought near, or withdrawn from, the south pole ; but they will be of only \ degree when the north pole is acted upon, and but of 9 degrees when the armature acts upon the two poles simultaneously. On placing the coil near the north pole, half-way between that pole and the neutral line, we shall have an inverse current when the armature ap- proaches the poles ; but it will be one of only 5 degrees when the north pole is acted upon, and one of only 2 degrees when the south pole" is acted upon. It will become one of 9 degrees when the armature is made to act upon both poles at once, and the effects will of course be reversed when the armature . is withdrawn instead of being brought near. 298 ELECTRIC LIGHTING. towards the neutral line, a fresh deviation of 10 degrees to the right will be obtained. It follows from these experiments that the induced cur- rents, caused by the movements of the coil along the magnet, will be the same as if the neutral line represented a resultant of all the magnetic actions of the bar. If this resultant were represented by a line through which the whole magnetic current were passing, there would follow from the approach of the movable coil to this line a current which, according to Lenz's law, would be inverse; and this is in fact precisely what takes place, since in bringing the coil back from the south pole, or from the north pole towards the neutral line, deviations to the right are obtained. Besides, it must follow from the same law, that on withdrawing the movable coil from that line, direct currents should be obtained, and this is found to be the case, since the deviations are to the left. It will therefore be understood that, in accordance with these considerations, a small coil movable round a mag-, netized ring, setting out from the neutral line of one of the two semicircular magnets composing the ring, and moving towards the inducer by which the ring is polarized, must give a direct current, and this is precisely what is observed in the Gramme machine. Let us now examine what occurs from the passage of the coil just mentioned, before the inducing pole itself, which I shall suppose to be the south pole of the preceding magnet ; but this time, instead of taking the small coil, of which we spoke at first, we shall take a real bobbin of little thickness and capable of sliding along an iron rod, which acts on its magnetic core. In order to know the directions of the cur- rents that we shall observe, we shall begin by examining the direction of the current produced when we bring near the south pole of the magnet the small bobbin, which we shall present by its anterior extremityj that is to say, by the ex- tremity which in the following experiments goes first. Under these conditions we shall obtain a deviation to the right of NOTES AND APPENDICES. 299 25 degrees, and when we withdraw the coil we shall obtain a deviation of 22 degrees to the left. As this experiment is the reproduction of Faraday's well-known one, we perceivo that the deviations to the right will represent inverse currents, and that the deviations to the left will represent direct currents. Now, if we take the coil in question, and cause it to pass from right to left tangentially before the south pole of the inducer, taking care to produce the movements in two parts, we shall observe : — T°. That in the first half of the motion a current will be produced causing a galvanometric deviation of 8 degrees to the left, and in the second half another current of 5 degrees in the same direction. 2°. That in effecting the movement in the contrary direc- tion, the current will be produced in the inverse direction. It may therefore be concluded that the currents resulting from the tangential movement of a coil before a magnetic pole are produced under conditions altogether different from those which result from a movement effected in the direc- tion of the axis of the magnet. These two movements are, in fact, produced not only in two perpendicular directions, but also under different conditions with regard to the manner in which the induction is produced in the various parts of the spiral. In the case of the tangential movement, the in- duction takes place only on one half of the circumference of the spires, and it acts from the two sides on a different end of the coil. It is not thus in the other case ; the relative positions of the different parts of the coil remain in the same condition as regards the inducing pole, and it is only the position of the resultant that varies with regard to the direc- tion of the motion. It remains to find what occurs when the coil performing the movements just considered is subjected to the action of a magnetic core, influenced by the inducer, and in this case it suffices to cause the coil to pass along the iron rod of 300 ELECTRIC LIGHTING. which we have spoken, while exposing the l-atter to the action of the inducing pole. On proceeding thus the follow- ing effects are observed : — i°. At the first moment, when the inducing pole is being brought near the iron rod, but at a distance sufficient to allow the coil to pass between it and this pole, there is pro duced in the coil, placed on one side, an induced current which results from the magnetization of the rod, and gives a deviation of 39 degrees to the right. The deviations on this side correspond then with inverse currents. 2°. When the coil, placed as in the first series of experi- ments, is moved from right to left, it produces, from the moment it cornes near the inducing pole, a current of 22 degrees to the left, which is, therefore, a direct current, and by continuing the movement beyond the inducing pole a new current is obtained in the same direction of 30 degrees to the left. The effects produced by the passage of the coil before the inducer are therefore in the same direction with or without an iron rod, but are much more energetic with the iron rod.* * The effects produced in this experiment should be carefully noted, for they prove that the magnetic actions are not so simple as is generally sup- posed. In fact, the results which we have just pointed out cannot be esta- blished unless the movable coil is placed on the part of the induced iron rod, intermediate between the inducing pole and its free extremities. Beyond this intermediate part the currents produced are in the opposite direction, which proves that in this case the iron rod has become a true magnet regu- larly constituted. Of course, if the rod is exposed to the inducing pole at one of its extremities, the magnet has only two poles and one neutral line ; but if it is exposed to this inducer at its centre, it forms- a magnet with a consequent point, and has therefore two neutral lines. If, however, the iron rod, instead of being at a distance from the inducing pole, is in contact with it, the effects are quite different. The currents produced by the movement of the coil towards the magnet are always inverse, and those which result from its withdrawal are direct. This shows that the resultant of the mag- netic forces is then concentrated at the inducing pole, which plays the part of a neutral line, as if the two magnetic pieces formed but one. This effect is always produced, on which side soever of the magnetic pole the iron rod is applied. If, however, under these conditions, the rod is separated from the magnet by a magnetically isolating substance, the effects without being NOTES AND APPENDICES. 301 It may therefore be said that the currents produced in con- sequence of the displacement of the coils of a Gramme ring, in relation to the two resultants corresponding with the two neutral lines, are in the same direction as those evoked by the passage of the spires of the coils before the inducing pole in each half of the ring. In order to study the effects resulting from the polar in versions, the experiment may be arranged as follows:— on one of the extremities of the iron rod provided with the in- duction coil mentioned above, a permanent magnet is made to slide perpendicularly to its axis. In this way the rod undergoes successive inversions of its polarities, and it is seen that not only is there produced by this a current more powerful than the magnetization and demagnetization cur- rents which result from the action of the pole of the magnet, but also that this current is not momentary, and appears to increase in energy until the inversion of the poles is complete. The direction of this current varies according to the direction of the movement of the mag- netized bar, and if it is compared with that which results from the magnetization or demagnetization of the magnetic core under the influence of one or other of the poles of the magnetized bar, it is observed that it is exactly of the same direction as the demagnetization current caused by the pole that has first acted; it is therefore in the same direction as the magnetization current of the second pole ; and as, in the movement performed by the magnet, the magnetic core is demagnetized, in order to be again magnetized in the con- exactly those we have analysed with the tangential movement of the coil, somewhat resemble them, and the difference depends upon the currents, re- sulting from the movements of the coil with regard to the magnets, being in the direction contrary to those which result from the magnetization of the rod, and giving rise to a rather feeble differential current, which shows that the last action preponderates. On the other hand, the currents produced from the middle of the rod to its free extremity, being no longer opposed by the direct action of the magnet, possess all their energy. (See my paper on this kind of actions in the Comptes Rendiis for the 24th February, 1879.) 302 ELECTRIC LIGHTING. trary way, the two currents which result from these two consecutive actions are in the same direction, and conse- quently supply a single current during the whole movement of the magnet. Again, the movement in the opposite direction of the magnet, having for its effect at the beginning a demagnetization in the direction opposite to that produced in the first case, the retrograde current resulting from the retrograde movement must be in a direction the reverse of that of the first. If we now return to the effects produced by our magnet, acting on our coils moving perpendicularly to their axes, it will be understood from the preceding that displacement of the magnetic polarity of the core must immediately be pro- duced by the inducing magnet having for its effect the in- version of the contrary polarity of this core before and be- hind the points successively influenced, it must follow that the different parts of the core of the coils will successively constitute a series of magnets with inverted poles, analagous to those the effects of which we have already analysed, and which are able to produce those currents in the same direction, whose presence we have proved. These currents will change in direction according as the coils move from right to left, or from left to right. (See my article on the electrical actions in operation in light-producing machines, in the journal La Lumicre Electriqite of the ist November, 1^79.) NOTE B. • • •." ON THE DUTY OF GRAMME MACHINES ACCORDING TO THE RESISTANCE OF THE EXTERNAL CIRCUIT. WE reproduce below an interesting paper of Hospitaller's, which shows the importance of the remark made at the con- clusion of this work. NOTES AND APPENDICES. 3°3 Dynamo-electric machines, considered as a source of elec- tricity, cannot, by reason of the numerous conditions of their action, be classed with other electric sources, such as liquid batteries or thermo-electric batteries, and therefore, thanks to the kind co-operation of Robert Gray, the engineer of the Indiarubbcr Works Co., we undertook at Silvertown, in the month of July, 1879, a series of experiments to establish, apart from all theoretical considerations, the electric elements of dynamo-electric machines placed under certain conditions of action. Our experiments had reference to Gramme machines of the A pattern, called the workshop pattern (represented on page 75 of this volume.) We must mention that similar ex- periments had been already made in France by Mascart and Angot (Journal de Physique, 1878), and in England by Hopkinson (Institution of Mechanical Engineers"). The former were undertaken more particularly from a theoretical point of view, from which we cannot here regard them ; the latter relates to Siemens' machines, and it is a similar inves- tigation that we desire to have made in France on the machines which are here most in use. Dynamo-electric machines are set up to work at a given speed, which it is convenient to maintain in order to produce from these machines all they can yield without damaging their parts or the solidity of their construction. We suppose, then, that the speed of rotation is constant, and we refer the electric elements to a normal velocity of 1,000 turns per minute. When the register indicates a greater or less velocity it is always easy to reduce it to this standard, for it is found by experiment that, other things being equal, the electro- motive force is proportional to the number of revolutions of the machine even for variations reaching to 300 turns a minute. We were particularly desirous of experimentally explaining the variations of the electric elements of the Gramme machines, by causing the external resistance to vary from 10 ohms to an external resistance of nothing. 3°4 ELECTRIC LIGHTING. Internal Resistance of the Machine. — On testing by the method of Wheatstone's bridge, the resistances of the ma- chine experimented with were found to be : — ohms. Total resistance of the machine before working ... ... ... i'i35 Resistance of the coil when warm, after having been working some time in short circuit 075 Resistance of the electro-magnets under the same conditions ... 072 Total resistance of the machine when warm ... ... ... i'47 These figures show that the internal resistance of the ma- chine varies within very narrow limits, and that these varia- tions are due to the heating of the wire, the resistance of which increases with the temperature. Electro-motivs Force. — Curve I. of the diagram, given in Fig. 74, shows the variation of the electro-motive force when the external resistance increases. The total resistances are referred to the axis of the abscissae, on a scale of Tyhs of a centimetre per ohm ; the electro-motive forces are referred to the ordinates on a scale of TVths of a millimetre per volt. When the total resistance is more than four times the in- ternal resistance, it will be seen that the electro-motive force is nearly constant and very feeble. This is due to the fact that the induction on the coil is produced only by the re- sidual magnetism of the electro-magnets. Then this electro- motive force increases very rapidly between 6 ohms and 4 ohms of total resistance, and it reaches a value of 107 volts, varying then very little. This is caused by the electro- magnets being magnetized to saturation ; the magnetic field remains constant, and as besides the speed of the induced system is constant, the electro-motive force, which is pro- portional to these two quantities, can vary only by the heating of the wire. Intensity of the Current. — The intensity of the current, expressed in webers, is represented by the curve II. of the diagram on a scale of y^ths of a millimetre per weber. NOTES AND APPENDICES. 305 It will be seen that this intensity, at first very feeble, in- creases afterwards in a nearly regular manner for total resist- ances varying between twice and four times the total resist- ance of the machine. These intensities have been calculated by the formula .9* is C } &9 10 ohmt in which Q is the intensity in webers, E the electro-motive* force in volts, R the total resistance in ohms. Work transformed into Electricity. — The figures we have- found are referred to the unit of time, the second. The value of the work transformed into electricity is expressed by Joule's formula, w - 10 Q'-* R, in which Q and R have the same meaning as before, and w is the work in meg-ergs. In 20 306 ELECTRIC LIGHTING. order to reduce to French units or kilogrammetres we must know that the kilogrammetre equals 98*1 meg-ergs. Curve III. of the diagram shows that the work trans- formed into electricity, which is very small when the total resistance exceeds 6 ohms, increases afterwards rapidly and regularly as the resistance diminishes. The scale of curves III. and IV. is T7oths of a millimetre for 2 kilogrammetres. Utilizable Work in the External Circuit. — The current produced by a magneto-electric machine is divided into two parts : the internal work which heats the wire of the coil and the electro magnets, and which cannot be utilized, and the available work produced in the external circuit, and which may be employed either to heat a wire, as in our experi- ments, or to produce different effects. The utilizable work, represented by curve IV. in the diagram, after having been very small, increases as the total resistance diminishes. It reaches its maximum when the external resistance is equal to the internal resistance of the machine, and afterwards diminishes to nothing when the machine is arranged in short circuit. In this case the whole of the work supplied by the motor is transformed into in- ternal work, the machine becomes greatly heated, the brushes burn, and the insulation of the wire may even be damaged. Duty. — Duty in its general sense is the ratio between the work expended and the work utilized If only the work transformed into electricity is taken into account, as we shall do here, by neglecting the passive resistances and the fric- tion of the parts, the duty is the ratio between the total work transformed into electricity (curve III.), and the work utili- zable in the external circuit (curve IV). This ratio, always less than I., is represented by curve V. in the diagram. It will be seen that this ratio increases with the resistance, and tends towards I. for an infinite resistance, in which case there is no longer any current. This assertion seems, in NOTES A AD APPENDICES. 307 contradiction to that which has often been made, that the duty is maximum when the external resistance is equal to the internal resistance. There is in this an error in the words which must be rectified. It is not the duty that is maximum in this last case, for it is only 50 per cent., but the utilizable work in the external circuit. The largest quantity, therefore, of utilizable elec- tricity will be obtained from a -given machine by making the external resistance equal to the internal resistance, but the highest electrical duty will be obtained by making the external resistance equal to 5 or 6 times the internal resistance. Under these conditions the machine will supply very little electricity, but the greatest part of it will be utilized on the external circuit. In practice, it is preferred to lose on the duty, and to cause the machine to produce the most that it can furnish at its normal velocity, by placing it under the conditions of maximum utilizable work, a maximum which is attained when the external circuit is equal to the internal resistance, certain corrections being made, which, according to the experiments of Jamin, Roger, and Le Roux, must be introduced into the value of the internal resistance. In one experiment, the machine turning with a velocity of 1,000 revolutions per minute, with an external resistance of 27 ohms, or i '8 times the internal resistance, a current was produced of 25-5 webers intensity, with an electro-motive force of 107 volts. The total work transformed into elec- tricity was 273 kilogramme tres, or 3-64 horse-power; the utilizable work was 179 kilogrammetres, or 2*38 horse- power. In this case the duty reached 65 per cent. By taking into account friction, passive resistances, &c., the value of which might reach i horse-power, the utilizable work is only about 50 per cent, of the work really used up by the machine. Experiments made under various conditions of resistance on different machines manufactured at Silvertown gave similar results. 20 — 2 3o8 ELECTRIC LIGHTING. Tension and Quantity Machines. — If, on a given dynamo- electric machine, we expend a certain quantity of work w, the expression of this work transformed into electricity may be put into this form : — w = Q E. Now this may be done in these machines in two ways. By making Q very great and E very small, the machine, having then little tension and supplying a large quantity of current, takes the name of a quantity machine. Q may also be made very small and E very great; the machine, having little quantity and a large electro-motive force, takes the name of a tension machine. The former class should work with an external circuit of small resistance ; the latter, on the contrary, requires a con- siderable external resistance to satisfy the relations which should exist between the external and internal circuits in order to obtain the maximum utilizable effect. In conclusion, we give a table showing the different values assumed by the electric elements of a machine, according as it is constructed to supply a so-called quantity current, or a so-called tension current. Elements of the working of Gramme machines, determined by experiments made at Silver tow n, fitly, 1879. Electric Elements, Suantity achine. Tension Machine. n(yj 967 Internal resistance in ohms « I '2O 4-58 External resistance . . . . . . 1*14 4-oo Total resistance of the circuit . . '. 2 '3.1 8'=;8 20*67 I7'sl Electro-motive force in volts . ... « 8rr,8 1 58 '50 Work expended in kilogrammetres "43 277 It will be seen that, according to this table, for the same quantity of work expended, the electric elements, resistance, intensity, and electro-motive force are notably different. NOTES AND APPENDICES. 3°9 It is possible to construct intermediate machines by suitably adjusting the lengths and thicknesses of the wire on the coils j but the figures we have just given show within what limits the Gramme machines used for the electric light are comprised. NOTE -C. THE CRITERIA OF THE ELECTRIC LIGHT. WE here extract the following passages from a paper by W. H. Preece, which we find in the Telegraphic Journal of the 1 5th February, 1879 : — " Heat and light are identical in character, though different in degree; and whenever solid matter is raised to a very high temperature it becomes luminous. The amount of light is dependent upon the height of this temperature ; and it is a very remarkable fact that all solid bodies become self- luminous at the same temperature. This was determined by Daniell to be 980° (F.), by Wedgwood 947°, by Draper 977° ; so that we may approximately assume the temperature at which bodies begin to show a dull light to be 1,000° (F.) The intensity of light, however, increases in a greater ratio than the temperature. For instance, platinum at 2,600° (F.) emits forty times more light than at 1,900°. Bodies when raised to incandescence pass through all stages of the spectrum : as the temperature increases so does the refran- gibility of the rays of light. Thus, when a body is at a tem- perature of — 250° F., it may be called warm. 500° ,, „ hot. 1,000° we have the red rays. 1,200° „ „ orange rays. 1,300° „ „ yellow rays. 1,500° „ „ blue rays. i,/oo° „ „ indigo rays. 2,000° „ „ violet rays. 3-10 ELECTRIC LIGHTING. So that any body raised to a temperature above 2,000° will give us all the rays of the sun. Inversely, the spectroscope may thus be enabled to tell us the temperature of the different lights, and it is, perhaps, because some lights do not exceed 1,300° that we lose all those rays beyond the yellow. " Tyndall has shown that the visible rays of an incandescent wire bear to the invisible rays a much smaller proportion than in the arc, and it is generally assumed that for the same current the arc will give at least z\ times greater light than an incandescent wire. Tyndall's figures are as follows : — Visible Rays. Invisible Rays. Gas i to 24 Incandescent wire ... i to 23 The arc i to 9 " The requirements of a good electric lamp are first, intense brilliancy ; secondly, great steadiness ; thirdly, duration. The Serrin lamp has the first kind of excellence ; all those lamps based on incandescence excel in the second respect; the Wallace Farmer light is the only one that attains the third point. The Rapieff is perhaps that form which, up to the present, most nearly combines the three requisites, but in reality no lamp has yet been introduced which fulfils all these requirements. "The objections to the use of the electric light are : — i°. The deep shadows it throws, 2°. The indifferent carbon that has hitherto been manu- factured for the purpose, which leads to unpleasant sounds, to great variation in the intensity of the light, and to waste. 3°. The difficulty in distributing the light itself. It is so intense, and confined to so small a space, that it does not lend itself to distribution like the gas flame, which occupies a considerable space. 4°. The unsteadiness of the light due to variations in the speed of the engine employed in driving the dynamo machine. There is another cause of variation in the electric arc, and that is the variation in the resistance of the arc itself, for it NO TES AND A PPEN DICES. 3 1 1 has been clearly demonstrated by experiments both in Ame- rica and in England, that the resistance of the arc varies as the resistances in circuit vary. The following table will show this : — Current in Light in Resistance of Arc. Webers. Candles. Ohms. 10 440 277 16-5 705 1-25 21-5 900 1-67 30-12 1,230 -54 " The light in the arc varies directly as the current, and not as the square of the current, as generally assumed. " Now, in the case of light raised by incandescence, the light will increase as the square of the current. It follows that if in the one case — viz., the arc — the light increases as the current only, and in the other case — viz., incandescence —it increases as the square of the current, a point is reached when the light produced by incandescence will equal that produced by the arc. The difficulty in reaching that point is the difficulty of obtaining a conductor with a sufficiently high point of fusion to resist the effect of powerful currents. Iridium is the only metal that is known to do this, and iridium is too scarce and too dear to be used for the purpose. "The multiplication of the light by Gramme's machine upon the Thames Embankment must not be taken as the solu- tion of the problem of the subdivision of the light. Theory shows unmistakably that to produce the greatest effect we must have only one machine to produce one light. We know from absolute measurements that such a machine can be made to produce a light of 14,880 candles, and it is possible to produce 1,254 candles per horse-power. But the moment that we attempt to multiply the number of lights in circuit this power diminishes, so that we have on the Embankment lamps giving us a light of scarcely more than 100 candles. The light of the Rapieff lamp in the Times office appears to be about 600 candle-power, and the Wallace light is equal 312 ELECTRIC LIGHTING. to 800 candle-power. In these two instances six lights are used in one circuit, but we have not here the subdivision of the light ; we have, on the contrary, the multiplication of the light, produced by the increased speed of the engine due to the insertion of additional lamps. It is, however, easily shown that if in a circuit where the electro-motive force is constant we insert additional lamps, then when these lamps are joined up in one circuit, t.e., in series, the light varies inversely as the square of the number of lamps in circuit, and when joined up, as in the multiple arc, the light diminishes as the cube of the number inserted. Hence the subdivision of the light is an absolute ignis fatuus. In the first place no machine has yet been produced which is competent or capable to light over 20 lamps ; secondly, no conductor is known but copper that is capable of conveying the current required to light these lamps, and copper is an expensive .material; thirdly, no electric light has yet been produced combining all the criteria of a good light." We consider this conclusion somewhat premature, and we must confess that on this point we do not quite agree with the learned English electrician. NOTE D. ON A NEW ARRANGEMENT OF THE WERDERMANN LAMP. The new arrangement of the Werdermann lamp, mentioned on page 209, consists in the addition to the arrangement shown in Fig. 56 of a brake acting on the lateral contact, and put in operation under the influence of the end contact, which is for this purpose adapted to the extremity of a lever attached to the jointed guide that brings the brake into action. So long as the pressure exerted by the movable carbon on the end contact is uniform, the brake does not act, NOTES AND APPENDICES. j 1 3 but the moment this pressure begins to diminish or increase, either on account of nodosities in the carbon, or from other causes, the brake is loosened or tightened, and thus the carbon is allowed to advance more easily or less easily. This arrangement of the end contact at the extremity of a jointed lever enabled the lamp to be automatically re-lighted in case it went out. For this purpose it was sufficient to place a contact below the jointed lever. When the latter is no longer kept up by the movable carbon, it falls upon the contact spring, and sends the current into the supplementary lamp. NOTE E. The new metal discovered by Edison, mentioned on page 215, is simply platinum freed from the bubbles of gas enclosed in its pores by being several times heated in a vacuum for pro- longed periods. Under these conditions the metal becomes much harder than in its ordinary state, and less fusible. Edison states that he has succeeded in obtaining a wire of this kind, giving with a radiating surface of -fa of an inch, a light of 8 candles, which would, with the ordinary wires of commerce, have been only i candle. " I can therefore," he says, " by increasing the calorific capacity of platinum, use wires of very small radiating surface, and considerably reduce the electric energy necessary for the production of a light of i candle. / have, in fact, succeeded in obtaining in this way 8 luminous centres, each giving a perfectly fixed light of 18 candles, and yielding a total light 0^ 138 candles, using for the purpose only 36,000 foot-lbs., that is to say, less than i horse- power of steam" TRANSLATOR'S APPENDIX.— No. i. ENGLISH EQUIVALENTS OF THE FRENCH DENOMINATIONS USED IN THIS VOLUME. MEASURES OF LENGTH. i kilometre = 1000 metres = 0*6214 mile FIG. 75- i metre = i '0936 yards — 3*2 809 feet = 39'37 inches. i decimetre = ofi metre — 10 centimetres = 100 millimetres = 3 -937 inches. i centimetre = ooi metre = o'i decimetre — 10 millimetres — 0*3937 inches. i millimetre = 'ooi metre — '01 decimetre = 0*1 centimetre = '03937 inches. MEASURES OF SURFACE. i square metre = \'\^ square yards — 107698 square feet. I square millimetre = 0-00155 square inch. MEASURES OF VOLUME. i litre = i cubic decimetre — 1000 cubic centimetres = 6f '02709 cubic inches- 0*22017 gallon = 0-88066 quart = 1-76133 pints. i cubic centimetre (cc) = '001 litre = 0*06103 cubic inch. WEIGHTS. i kilogramme = 1000 grammes = 2*20462 pounds avoirdupois. i gramme - 15-43235 grains - 0*035274 os. avoirdupois. i decigramme = ofi gramme. i centigramme — 0*01 gramme. I milligramme = 0*001 gramme. TRANSLATOR'S APPENDICES. 315 WORK. I kilogrammetre ~ T^foot-lbs. THERMOMETRIC SCALES. To convert centigrade degrees into Fahrenheit degrees multiply by f or r8, and add 32 to the result. MONEY. i franc = 100 centimes — • 9*6 pence. 25 francs = £i sterling. TRANSLATOR'S APPENDIX.— No. 2. RECENT INVENTIONS. Since the publication of the original edition of this work, the inventors of electrical apparatus have not been idle. Patents without end have been, and continue to be, granted almost daily for alleged improvements in current generators, and for new forms of electric lamps. But none of these in- volves any principle which has not been already illustrated in the course of this work, nor does it yet appear that any- one is destined to supersede such forms of apparatus as are described in the text. Nevertheless, there are some quite recent developments of certain forms of apparatus that greatly affect the problem of electric lighting, and these may here be very briefly described. Incandescent Lamps. — Incandescent lamps of ex- tremely simple construction have been lately brought out by several inventors. These are all identical in principle, con- sisting of a slender filament of carbon enclosed in a vacuous vessel. They differ in such particulars as the dimensions of the carbon filament, the material from which it is prepared, the method of attaching its extremities to the current con- 3 1 6 ELECTRIC LIGHTING. ductors, and the manner in which the vacuous vessels are exhausted and sealed. Such lamps will, without damage, bear a certain maximum of current for a more or less pro- longed period, the amount of light depending, of course, upon their electrical resistance and the energy of the current. In operation, their average endurance, or "life," has been stated to be, under favourable circumstances, about 1,000 hours. The question of the subdivision of the electric light presents no difficulties with these lamps, and they solve the problem of the application of the light to domestic purposes. Swan's Lam^.—The inventor of this lamp has discovered .a method of preparing from cotto^ thread very attenuated filaments of carbon of the tenacity requisite for their sufficiently pro- longed stability and endurance when in use. These extremely thin car- bons are perfectly homogeneous throughout, and are so far from be- coming damaged by use that the effect is to a certain extent an in- crease of their solidity and elasticity. FIG. 76. The arrangement of the lamp, which is shown on Fig. 76, is extremely simple. The filament of carbon, bent round so as to form a spirally circular loop of about one-fifth of an inch in dia- meter, is enclosed in a glass bulb about two inches in diameter. The extremities of the filament are connected in an ingenious manner to two platinum wires, which pass outwards and either form two small loops, or terminate in binding screws, for connecting with the circuit. These platinum wires are fused into the bulb, and are supported by a piece of glass, which descends internally for a certain dis- tance. The bulb is hermetically sealed, after having been completely exhausted by means of a Sprengel pump. The light yielded by these lamps is mild and steady, with an in- tensity depending, of course, on the current of electricity sent TRANSLA TOR'S APPENDICES. 3 T 7 through them, but which may be safely carried as high as 20 candles. It is stated that i horse-power of force absorbed suffices to maintain 10 of these lamps. At the Exhibition of Electrical Apparatus at Paris in 1881, the Swan lamp re- ceived the gold medal as being the best system in its class. Maxim's Lamp. — The carbons for this lamp are prepared from cartridge paper, and the vacuous bulbs contain a residual atmosphere of a hydro-carbon instead of air. It is claimed that by this the durability of the filament is increased, and the irregularities of the resistance at various points become equalized. This form of loop preferred by the inventor has four parallel vertical portions, nearly like a capital M. The resistance of this lamp is stated to be more than twice that of Swan's lamp. The light given out maybe carried to 50 candles. Edison's Lamp and Fox-Lane's Lamp. — Edison prepares his carbon from a filament of bamboo, while Fox-Lane makes use of a string of flax. Both make a single loop of the carbon, which is bent into a horse-shoe form. The re- sistance of Edison's lamp is about the same as that of Maxim's, while the resistance of the Fox-Lane lamp is less than that of the Swan. The Faure Secondary Battery.— Plante's polarization battery, which was invented about 1859, has been mentioned in the text (page 15). This battery was formed by the action of the current from a primary battery on plates of lead im- mersed in diluted sulphuric acid. The nature of the polariz- ing action itself is explained on page 9. Faure also uses thin plates of lead for the elements of his cell, but instead of forming lead oxide by electrolysis, he coats one of the plates with a film of red oxide of lead, and this is separated from the other plate by a layer of felt. The Faure cells, or " accumulators," as they have been called, are made of a large size, and according to Sir W. Thompson, one of these cells, weighing 75 kilogrammes, "can store and give out 3 1 3 F.LRC TRIG LIGHTING. again energy to the extent of an hour's work of one horse- power." The Faure cell may be charged by a voltaic battery or by any generator or dynamo-electric machine giving a direct current. The light of incandescent lamps worked by the Faure accumulator is perfectly steady, being absolutely free from those fluctuations which may usually be detected in the action of the dynamo-electric machines. Its great use for the electric light consists not only in its supplying the means of carrying a store of electricity about, but in affording a regu- lator for theJamps. In fact these might remain lighted for hours, even if the electric supply from the engine were in- terrupted. A train lighted up with incandescent electric lamps, worked by Faure's accumulators, has been running continuously for some months between London and Brighton, without any failure of the light once occurring. This inven- tion imparts to electric illumination as great a degree of readi- ness and certainty in working as has been claimed for gas- lighting. THE END. DALZIEL BROTHERS, CAMDEN PRESS, LONDON, N W. GEORGE ROUTLEDGE AND SONS, POPULAR WORKS ON SCIENCE PUBLISHED BY GEORGE ROUTLEDGE & SONS. A Popular History of Science. By ROBERT ROUTLEDGE, B.Sc. (Lond.), and F.C.S., Author of "Discoveries and Inventions of the Nineteenth Century. " With many Original Illustrations and Portraits, and Sixteen Full-page Plates. Crown 8vo, cloth, 12*. 6d. Discoveries and Inventions of the Nineteenth Century. By ROBERT ROUTLEDGE, B.Sc. and F.C.S. With many Illustrations, and a Coloured Plate, los. 6d, Science in Sport made Philosophy in Earnest. By ROBERT ROUTLEDGE. Post 8vo, cloth, gilt edges, 5*. The Boys' Book of Science. Including the Successful Performance of Scientific Experiments. 470 Engravings. By Pro- fessor Pepper, late of the Polytechnic. Cloth, gilt edges, 7*. 6d. The Book of Metals. Including Personal Narratives of Visits to Coal, Lead, Copper, and Tin Mines ; with a large number of interesting Experiments. 300 Illustrations. By Professor PEPPER, late of the Polytechnic. Post 8vo, cloth, gilt edges, 6s. The Microscope : Its History, Construction, and Appli- cation. Being a Familiar Introduction to the Use of the Instrument, and the Study of Microscopical Science. By JABEZ HOGG, F.L.S., F.R.M.S. With upwards of 500 Engravings and Coloured Illus- trations by TUFFEN WEST. New Edition, crown 8vo, cloth, 7^. 6d. The Common Objects of the Microscope. By the Rev. J. G. WOOD. With Twelve Pages of Plates by TUFFEN WEST, embracing upwards of 400 Objects. The Illustrations printed in Colours. Fcap 8vo, 3*. 6J., cloth, gilt edges. %* A CHEAP EDITION, with Plain Plates, u- , fancy boards. LONDON AND NEW YORK. GEORGE ROUTLEDGE AND SONS, The Orbs of Heaven ; or, The Planetary and Stellai Worlds. A Popular Exposition of the great Discoveries and Theories of Modern Astronomy. By O. M. MITCHELL. With numerous Illustrations. Crown 8vo, 2s. 6d. Popular Astronomy ; or, The Sun, Planet, Satellites, and Comets. With Illustrations of their Telescopic Appearance. By O. M. MITCHELL. 2s. 64. The Story of the Peasant -Boy Philosopher Founded on the Early Life of FERGUSON, the Astronomer. Bj HENRY MAYHEW. Illustrated. Cloth gilt, 3-r. 6d. The Wonders of Science ; or, The Story of Young HUMPHREY DAVY, the Cornish Apothecary's Boy, who taughi Himself Natural Philosophy. By HENRY MAYHEW. Illustrated. Cloth gilt, y. 6d. The Book of Trades, and the Tools used in Them. By One of the Authors of " England's Workshops." With numerous Illustrations. Small 4to, cloth, gilt edges, 3*. 6d. Wonderful Inventions, from the Mariner's Compass to the Electric Telegraph Cable. By JOHN TIMES. New Edition, with additions. Crown 8vo, cloth, 5-r. A Manual of Fret-Cutting and Wood-Carving. By Sir THOMAS SEATON, K.C.B. Crown 8vo, cloth, 2s. 6d. The Laws of Contrast of Colours, and their Appli- cation to the Arts. New Edition, with an important Section on Army Clothing. By M. E. CHEVREUL. Translated by JOHN SPANTON. With Coloured Illustrations. Crown 8vo, cloth, 3^. 6d. Geology for the Million. By MARGARET PLUES. Edited by EDWARD WOOD, F.G.S. With 80 Illustrations. Fcap., picture boards, if. Scientific Amusements. Edited by Professor PEPPER. 100 Woodcuts, is., boards; is. 6d., cloth gilt. Electric Lighting. Translated from the French of Le Comte Th. Du MONCEL by ROBERT ROUTLEDGE, B.Sc. (Lond.), F.C.S., Author of "Discoveries and Inventions of the Nineteenth Century," "A Popular History of Science," etc. With 76 Illus- trations. Crown 8vo, cloth, 2t. 6d. LONDON AND NEW YORK.