GIFT OF fl . THE ELECTRIC LIGHT IN ITS PEACTICAL APPLICATION. THE ELECTEIC LIGHT IN ITS PRACTICAL APPLICATION. BY PAGET HIGGS, LL.D., D.Sc., * > TELFOKD PRIZEMAN AND ASSOCIATE MEMBER OF THE INSTITUTION OP CIVIL ENGINEERS, Author of " Electric Lighting" "Some Eecent Improvements in Dynamo-Electric Apparatus," "Electrical Formula" (Molesworth's Pocket- Book), Papers on Practical Plating, etc. LONDON: E. & F. N. SPON, 46, CHARING CROSS, NEW YORK: 446, BROOME STREET. 1879. J. C. Cebrim, V PREFACE. THE following pages are intended to give the reader an account of what has been effected in the numerous endeavours to obtain a practicable system of electric lighting. But the details have been confined to those necessary to form judg- ment of the advantages of each system. Abstruse discussion has been carefully avoided, and questions have not been raised to which answer could not be found in previous practice. The labours of Du Moncel and Fontaine, the reports of Tyndall, Houston, Thomson, Deacon, Haywood, and others, have been freely utilized, the object having been to give both pro and contra. Much descriptive matter and numerous illustrations have been taken from my translation of Fontaine's " Eclairage Electrique," now out of print; and considerable indebtedness must be acknowledged to other sources, named in the text. Where my own experience has led me to a conclusion, I have ventured to express it, but I have always also stated the reason for the deduction. There must necessarily be, in a technical work of this character, many imperfections. Kecent and untried inven- tions, promising much, cannot be omitted from notice ; nor, from want of knowledge of detail, can a probably correct opinion be held. Electric lighting is, indeed, so far within its period of infancy that, in many cases, suspense of judg- ment is compulsory. Nearly every week marks an important advance, proving the present incomplete state of this branch of engineering. 24890 2 VI PREFACE. With regard to the future of electric lighting, little has been said in this book. Public opinion, if not always strictly accurate, generally approximates to the correct idea of the commercial value of a newly introduced method, and its per- ception of the advantages of the electric light, either future or immediate, has not been greatly misled, however exag- gerated may have been the statements of interested specula- tors. It is beyond doubt that in the present we may look for practical, if not great, improvements, that will cause in no distant future the adoption of electric lighting for very many important, as well as ultimately for general, purposes. Logical sequence has been followed as far as possible, so as to afford aid to the general reader. The first chapter deals with the principles of the voltaic arc, and distinguishes the method of lighting by incandescence. The various forms of lamps employing the voltaic arc are next described, with so- called " candles " and candle-lamps, followed by discussion of most of the proposed systems of lighting by the incandescence of carbon or platinum. The principal magneto- and dynamo- electric machines are then described, with the new multiple- circuit machines, followed by a full consideration of the mechanical efficiency of these machines, and sufficient simple mathematical data to enable the reader to form his own conclusion of the merits of a fresh project. Next the question of cost is entered into. The various well-defined schemes for division of the electric light are commented upon. The book is concluded with chapters on the maritime and military and various applications of the electric light, and descriptions of the several methods of preparing the carbons consumed in the lamps. There is also a chapter on apparatus for main- taining electric currents at constant strength, although this kind of apparatus has not met with practical application. In conclusion, I can only hope that my readers, whether of the press or of the public, will accord me the kindly con- sideration extended to my previous attempt to place before them a synopsis of this subject. PAGET HIGGS. CONTENTS. CHAPTER I. PAGE INTRODUCTORY ... ... 1 CHAPTEE II. LAMPS OR BURNERS, EMPLOYING THE VOLTAIC ARC ... ... ... 8 CHAPTER III. ELECTRIC " CANDLES " AND CANDLE-LAMPS ... ... ... ... 45 CHAPTER IV. LIGHTING BY INCANDESCENCE ... ... ... ... ... 54 CHAPTER V. MAGNETO- AND DYNAMO-ELECTRIC MACHINES ... ... ... 71 CHAPTER VI. MECHANICAL EFFICIENCY OF ELECTRIC-LIGHT MACHINES* 127 Vlll CONTENTS. CHAPTER VII. PAGE SIMPLE MATHEMATICAL CONSIDERATIONS CONCERNING ELECTRIC LIGHTING 158 CHAPTER VIII. ELECTRIC REGULATORS ... ... ... ... ... ... 176 CHAPTER IX. COMMERCIAL ASPECT OF ELECTRIC LIGHTING ... ... ... 184 CHAPTER X. DIVISION OF THE ELECTRIC LIGHT ... ... ... ... ... 214 CHAPTER XI. MARITIME AND MILITARY APPLICATIONS ... ... ... ... 217 CHAPTER XII. VARIOUS APPLICATIONS OF THE ELECTRIC LIGHT .,. ... ... 228 CHAPTER XIII. ELECTRIC CARBONS 233 THE ELECTRIC LIGHT IN ITS PEACTICAL APPLICATION. CHAPTER I. INTRODUCTORY. THE electric light can be obtained, practically, by two methods. Given a sufficiently intense electric source, there are placed in the circuit, at, say, each of its severed ends, two sticks or rods of carbon. These sticks are brought into contact; they become heated, glow, and upon slightly sepa- rating them, the phenomenon known as the voltaic arc appears. This is one method of producing the electric light. It is a property of the electric current that, when passed through a badly conducting substance, it causes that sub- stance to become heated. A carbon rod, a piece of platinum wire, or thin iron wire, interposed in an electric circuit, becomes heated, and glows with an intensity of light de- pendent upon the strength of the current and the resistance offered by the bad conductor. In one case, the circuit is interrupted by the space over which the voltaic arc plays ; in the other, known as lighting by incandescence, the bad conductor is continuous. Upon these two methods of lighting by electricity depend two distinct classes of apparatus, or lamps, for effecting illumination in a practical manner. Light may also be produced by the use of Geisler tubes. B FIG. 1. LEOTIC LIGHT. These are glass tubes, exhausted of air, and filled with rarefied gases, which become luminous upon passage of an induced current of high intensity. But the light developed is very feeble, and unsuited to any ordinary purpose. Consideration will, therefore, be given to only the preceding methods. Brilliancy of the voltaic arc depends upon the strength of the electric current, upon the nature of the electrodes or terminations of the circuit where the light appears, and upon the nature of the atmosphere in which it is produced. With potassium or sodium the light is more brilliant than with platinum or gold ; mercurial vapours and hydrogen reduce the light. The colour of the arc varies with the constitution of the electrodes, being yellow with sodium, white with zinc, mag- nesium, and green with silver. The appearance of the arc depends upon the form and nature of the electrodes. With a carbon point at the positive, and a platinum plate at the negative, electrode, the arc takes the form of a cone, becoming egg-shaped when a second car- bon point is substituted for the platinum plate. The voltaic arc is the result of the incandescence of a jet of particles detached from the elec- trodes, and thrown from one electrode to the other, particu- larly from the positive pole to the negative pole. The positive electrode has a temperature much higher than the other, the negative electrode being barely red when the positive electrode is at white heat. The positive pole is consumed at double the rate at which the negative pole disappears, when the carbons are equal in size. The arc (Fig. 1) appears THE VOLTAIC AEG. 3 as a flickering flame, and brilliant particles are constantly carried between the two electrodes. Molten globules of mineral impurities appear, g g', upon the carbon points. When the carbons are chemically pure, these globules do not appear. This is the appearance of the arc as it burns in air. In a vacuum the carbons are not consumed with such rapidity, the positive point becomes hollowed out and dimin- ished in weight, while the nega- tive point increases in volume (Fig. 2). When the electrodes are un- equal in size, experience has shown that it is advantageous to make the negative the larger, and this electrode can be so extended that its consumption becomes practically nil. Upon this principle several electric lamps have been con- structed, in which metallic electrodes of large surface have been substituted for the negative carbon. The voltaic arc behaves precisely as any other portion of the electric circuit. It is attracted or repelled by magnets in exactly similar manner. Indeed, the incandescent particles constitute between the two electrodes a conductor of great mobility ; and the arc may be regarded as a badly conducting chain of these particles, raised to incandescence, in consequence of the resistance they offer to the passage of the current. The length of the arc follows certain definite laws, and these were determined by Despretz in 1850. He found that the length of the arc increased more quickly than the number of elements employed to produce it ; that this increase is larger with small arcs than with large. The arc produced with 100 Bunsen elements is nearly quadruple that produced by 50 elements ; that resulting from 200 elements is not triple that from 100 ; that from 600 elements is about 7 \ times greater than that from 100. The battery of 600 elements, coupled in a single series, gives as much as 7'87 inches of arc when the positive carbon is the higher ; that, when the elements are 4 THE ELECTRIC LIGHT. coupled in quantity, the length of the arc increases less quickly than the number of elements. The arc from 100 elements being 0'98 inches, it is only 2'7 inches with 600 elements coupled in six series of 100, whilst with the same battery of 600 elements coupled in tension it attains 6*5 inches. Coupling successively in quantity series of batteries of 25 elements in tension, there is obtained : for a single series, nearly no arc at all ; for two series, an arc still too small to be measured ; for three series, 0*039 inches ; and for 24 series, 0'45 inches. The same batteries coupled in tension give an arc of 6*37 inches, that is, a space between the carbons 14 times greater than with 24 series coupled in quantity. When the positive pole is the lower one, the voltaic arc is shorter than when the negative pole occupies this position. With six series of 100 elements coupled in quantity, there is obtained 2*9 inches distance when the posi- tive pole is the higher and 2*2 inches if it be the lower. That when the electrodes are placed horizontally, the arcs are shorter than with vertical electrodes, and then the battery arranged for quantity is more advantageous than that for tension. Thus, six series of 100 elements coupled for quantity give a horizontal arc of 1*6 inches, and 600 elements, end to end, give only a horizontal arc of 1'06 inches. These experiments explain very clearly the futility of con- structing lamps in which the carbons are burnt horizontally, as well as the difficulties that constructors of magneto-electric machines have met in trying to obtain the voltaic light with apparatus of great quantity and low tension. The electric light owes its special value to the creation of great heat in small compass, and to this end the best known means is electricity. This light has great analogy to that of the sun. It causes the combination of chlorine with hydrogen, decomposes chloride of silver, and it imparts phosphorescent properties to susceptible substances. M. Jamin has pointed out that the comparison of the electric arc to the sun, which is our highest conception of brilliancy, may be made in two ways : by the relative times required to produce equal photo- graphic images, or by the direct measure of the illuminating powers. Fizeau and Foucault found by the first process that COLOUR OF LIGHTS. 5 the power of the sun is only two and a half times superior to that of the arc ; the second method has proved that the carhon points with a powerful machine are equal to the sun in lustre. The light of a gas flame is orange yellow when compared to that of the electric light. A light, to be applicable for purposes of illumination, must contain the seven primitive colours of the spectrum in certain proportions. All luminous bodies do not contain these colours in the same proportions. The electric arc, produced between silver and carbon, contains only two green bands, and if the silver be replaced by other metals, the spectrum obtained is always formed of brilliant lines separated by wide, dark spaces. These lights could not be used for illumination. The spectra of gas and oil flames are continuous ; the red, orange, and yellow are very abundant ; there is but little green, almost no blue, and little or no violet. These flames are rich in colours, but slightly refrangible, which gives them their orange tint; poor in highly refrangible rays, and destitute of indigo and violet. The red may be removed, but it is im- possible to add the indigo and violet, and this is the cause of their inferiority. The electric light is more complex ; it pro- ceeds at the same time from the carbons and from the arc, and differs according to the one or the other of the sources. That from the carbons is white, and the same as that of the sun. The light from the arc itself is violet blue, and its spectrum tends towards the most refrangible colours ; it is the opposite of gas or lamp light ; it contains little red, much blue, and a large excess of violet. It is the light of the arc which gives to electric illumination the bluish tint objected to with some reason. But the superfluous rays can be removed from the electric light by the interposition of uranium glass, solu- tion of quinine sulphate, and many other substances. The production of light is ordinarily a secondary pheno- menon which accompanies the chemical combination of the combustible with the oxygen of the air. This combination removes the oxygen from the air, and replaces it by vapour of water and carbonic acid gas. The electric light has the decided advantage of not altering the state of the atmosphere. 6 THE ELECTRIC LIGHT. The electric arc does not heat. This appears astonishing at first, for all bodies fuse or volatilize when introduced into the arc. The reason is that the heat-producing rays "are by far more abundant in gas and lamp flames, than in the arc, which emits the greatest amount of light with the least proportion of heat. The fault, continues M. Jamin, has often been committed of attempting street illumination on the lighthouse system by a beam of light concentrated by reflectors, and thrown along the length of the street. Such experiments have only suc- ceeded in blinding the by-passers, and projecting long shadows behind them. There are cases when such concentration is the only end that is desired. In workshops, it is only neces- sary that the workman shall have a clear view of the work before him. It is the same in dining-rooms, billiard-halls, reading-rooms, etc., and no one pays attention to the obscurity behind him. It is different in depots, theatres, lecture-rooms, and display storerooms ; in these cases a general illumination is required, coming from all directions, and lighting every side of an object. When several electric lights are placed in a hall illuminated by gas, the eye immediately experiences a sort of relief, both by the redoubled brilliancy and by the per- ception of colours which were not before suspected ; and, on the contrary, if the electric lights be suddenly extinguished, the spectators are thrown into the comparative night of the old illumination. The conditions of good electrical lighting must be deter- mined by a study of the general illumination of objects during the day. When the sky is clouded, the sunlight pierces the clouds as through a ground glass, and the whole sky is like an immense illuminated ceiling, radiating light from every point and in all directions. The objects illuminated diffuse in their turn the light which they receive, so that there is an intercrossing of rays, producing the effect of a mean amount of light everywhere. This is general illumination. Such is the model that must be followed. For this pur- pose the ceiling, walls, and floors must be well illuminated, that the diffused light may be radiated into the empty spaces ; PRINCIPLES OF ILLUMINATION. 7 and, that the quantity may be the same everywhere, it will be necessary to multiply the sources of light. That the direct rays may not painfully affect the retina, it will also be neces- sary to diminish their brilliancy by the interposition of ground glass and some fluorescent substance, such as quinine sulphate, in order to transform the violet and ultra-violet rays into white light. Lastly, and especially, it will be necessary to cover all openings by which the light may escape. The exterior light enters by the windows during the day, and it is by them that the nocturnal illumination escapes. M. Jablochkoff introduced electric lighting into the laboratory of the Sorbonne, and the feeble effect it produced was asto- nishing. This laboratory is covered with a glass roof, by which it is well lighted during the day, and by which it allowed the loss of at least one-half of the light produced by the electric candles. This wasted light illuminated the high walls of the surrounding buildings, and gave a brilliant but useless illumination in the court. The same thing happens with gas, and will occur with electricity in the illumination of public places. All of the lamps waste half of their light in radiation towards the sky. A simple reflector would return it to the ground and double the illumination. THE ELECTRIC LIGHT. CHAPTER II. LAMPS OR BURNEKS, EMPLOYING THE VOLTAIC ARC. ELECTRIC lamps are somewhat generally known as " regu- lators," and are also termed "carbon-holders." The ex- pression "lamp or burner" will, however, be sufficient to distinguish that portion of, or contrivance in, the electric circuit employed as a light-centre. Burners may be divided under two heads : (1) Those in which the voltaic arc is caused to exist between the two portions of an interrupted circuit ; (2) Those in which a continuous portion of the circuit, being a bad conductor, is heated to incandescence by the passage of the current. It is, however, highly probable that this division, being arbi- trary and not generic, may at any time be upset by a new invention based upon principles that do not appear in the lamps at present invented. It was not until Leon Foucault, experimenting in 1844 with a Bunsen battery, hit upon the idea of substituting retort carbon for common wood charcoal as the substance for electrodes, that the electric light promised to be of use. This opened up a practical application of the light to photo- graphic purposes. The lamp was simply a holder for the carbon rods, and required help from the hand of the operator. Trials of the light were made in the Place de la Concorde, by M. Deleuil, who had previously experimented with carbons placed in a receiver from which the air had been exhausted. BURNERS USING THE VOLTAIC ARC. 9 STAITE AND EDWARDS' LAMP. T. Wright, in 1845, caused the voltaic arc to play between discs of carbon, the origin of Le Molt's apparatus. Staite and W. Edwards, in 1846, introduced a lamp (Fig. 3), in which FIG. 3. two carbon electrodes are enclosed in small cases, meeting obliquely on a refractory and badly conducting substance. The points are brought into place, as consumed, by springs. A sliding-piece and screw beneath the baseboard enable the length of the voltaic arc to be regulated. Staite and Petrie, as well as Foucault, in 1848, devised plans by which the current itself regulated the distancing of the carbons. These plans were based upon the phenomena (1) that an electric current can cause magnetization according to its strength ; (2) that the voltaic arc as part of the con- ductor, reacts upon the current. 10 THE ELECTRIC LIGHT. FIG. 4. LE MOLT'S LAMP. Le Molt, in 1849, revived Wright's idea. This lamp (Fig. 4) is thus described? "As electrodes producing the light, I patent the use of all carhuretted matter, especially that of retort carbons, and the two combined movements of rota- tion and approximation, at given intervals, of two discs of variable depth and diameter. The discs are maintained, with regard to one another, in a parallel attitude, vertical or horizontal, or, pre- ferably, in positions at right angles, and conveniently distance^ one from the other, to produce the electric light. The discs re- volve regularly upon two metal axles, put into connection with the poles of the generating appa- ratus, and presenting, successively, by the combined rotation and ap- proximation, all the extreme points of their circumferences to the production and emission of the electric light ; in such manner that at each revolution of the discs, the latter approach one another by the distance which they had separated by the combustion of part of the carbon, and thus are always replaced in the same position of invariable distance ; and as the two movements of rotation and approxi- mation, combined with revolving electrodes, may be obtained with the aid of any kind of mechanical system, it is sufficient that I indicate in my design one of these arrangements, to illustrate how the rotation and approximation may be com- bined. I reserve to myself the purification of carburized matter forming the electrodes emitting the light, by more or less prolonged immersion in all kinds of acids, and prefer- BURNERS USING THE VOLTAIC ARC. 11 FIG. 5. ably in nitric and muriatic acids mixed, and subsequently in fluoric acid." This lamp allowed of twenty to thirty hours' continuous light ; but the intensity of the light must be less than that obtainable with vertical carbon rods. AKCHEREAU'S LAMP. Archereau's lamp (Fig. 5) is the basis of many minor ideas, including the Lancaster lamp, and is one of the simplest and most effective of its kind. It consists of a hollow coiled copper wire, with a ver- tical standard, two carbon carriers, and a counterpoise. The upper carbon is carried by a bar, sliding into and turning at the extremity of an insulated horizontal copper bar, in connection with the negative pole of the electric source. The lower carbon rests on a cylinder, half of copper, half of iron, rising or falling in the hollow bobbin. The positive pole of the electric source is attached to one end of the wire coil, and the other end to the interior cylinder of the coil. A weight counterpoises the lower carbon-holder. When the current passes in the ex- terior wire, it produces magnetic action, causing the cylinder to descend into the bobbin, interrupting the current. Under these conditions, the action of the counterweight raises the cylinder. Initially, the carbon points must be brought into contact, to establish the electric circuit. When the voltaic arc is formed, the cylinder remains fixed in the coil, and the counterweight is motionless. As the voltaic arc increases in length, and the current is weakened, the lower carbon rises, until the current again attains sufficient power. 12 THE ELECTRIC LIGHT. LACASSAGNE AND THIEKS* LAMP. Lacassagne and Thiers, observing the expense and incon- venience attending the employment of clockwork in the lamps previously constructed, substituted a float acting in a bath of mercury. Their patent dates in 1855. A cylinder contained a float upon, and in connection with, the mercury ; the carbon electrode rested upon this float. The float was in connection with the positive conductor ; the other carbon electrode was fixed in the same axial line above the electrode supported on the float. As the carbon points consumed, so the float rose ; but a means of arranging the rise to occur at the proper time was necessary, and was thus supplied. Mercury from a reservoir, having entered the float cylinder, passes through a tube placed in an electro-magnet. In this tube is an india- rubber valve, opened and closed by a soft-iron armature, withdrawn by a spring opposing the action of the electro- magnet. The opening of the valve admits mercury to the float cylinder. As the distance between the electrodes in- creases, the magnetic attraction decreases, and the valve opens, the incoming mercury raising the electrode. This lamp was before the Parisian public from 1855 to 1859, and its practical introduction was deterred simply by reason that it was an invention in advance of its time. WARREN'S LAMP. A simplification of the preceding burner has been sug- gested (1878) by Bruce Warren, in which the electro-magnet is omitted. The following is a description of the experimental system : Ebonite or guttapercha tubes are laid in the same way as gas-pipes in a house, and at suitable places are in- serted ebonite taps, which communicate with a reservoir of mercury, on which floats the carbon pencil ; the upper carbon is merely suspended by a good conducting substance to the "earth," etc. The system of tubes is connected with a small tank of mercury at the top of the building, which fills the tubes, and acts as conductor and regulator at the same time. BURNERS USING THE VOLTAIC AEC. 13 When the taps are open, the mercury completes the circuit, the taps when shut simply breaking the column of mercury. On an experimental scale, glass tubing and a few T-pieces con- nected with vulcanized rubber tubing answer; a few pinch- cocks serve the purposes of taps. If a tank of mercury is used for each floor of a building, the tanks themselves, which may be of iron, insulated from the walls, may be all connected together with stout copper conductors, so that one generator supplies the electricity for all the carbons. It is better to have a metal tap to regulate the pressure, and an ebonite tap for extinguishing and relighting, so as not to interfere with the resistance of the circuits. An india-rubber washer is placed on the carbon, through which it slips easily, whilst it prevents the mercury from flowing over. The flotation of the carbon, and the cohesion of the mercury, will enable the friction through the washer to be so adjusted that no inconvenience will arise from a column of mercury the height of any ordinary room. For the convenience of replacing the carbons, a 4-way piece, fitted with a pinch-cock at its lower end, so as to empty the branches carrying the carbon, may be used instead of the T-piece. The accompanying diagram (Fig. 6) illustrates an arrange- ment which was used as a double-light chandelier. M is a small tank containing a few ounces of mercury ; G is a glass tube of small bore, connected with the funnel M and the 4-way piece T by means of small pieces of rubber tubing. This is more convenient than welding the parts together by heat, since it allows the use of pinch-cocks instead of taps. The dis- tance of the ends of the glass from each other in the rubber tubing should be only sufficient, so as to allow the closing of the tubes. As C C' C" C'" are ordinary Mohr's clips, by closing the lower end of T and filling M with mercury, the tube G and arms A A' are filled with mercury. The clip C being closed and that at C'" being opened, the mercury in A A' may be drawn out for FIG. 6. 14 THE ELECTRIC LIGHT. putting in fresh carbons. A wire from a battery, B, dips into the mercury at the funnel M, whilst the other wire from the battery is carried to two pieces of carbon, L- L", lightly suspended over the pencils P P'. The pencils pass through rubber washers, K K', with slight friction only in the top of the arms A A'. The arms A A' may be made by simply bending two short pieces of tubing and contracting them slightly at the top. The whole may be readily extem- porized by using Hoffman's voltameter, or, preferably, a modification of this, supplied with an opening at C'". Of course, the stop-cocks would be removed^ This lamp arrange- ment can be fitted up very efficiently for experiment at the cost of a few pence. A somewhat similar lamp has been shown by F. Higgins. DUCBETET'S LAMP Is similar in principle to the preceding. It (1879) consists in the employment of a column of mercury in which one or more carbons are placed. The different densities of these substances causes the carbon to rise to the surface. To obtain the light, a carbon of somewhat larger section than the one immersed vertically in the mercury is placed horizontally, so that the end of the vertical carbon im- pinges against it. The light obtained is, therefore, similar to the Eeynier or Werdermann light. A battery of six to ten Bunsen elements has given good effects. One advantage of this system is that the resistance of the circuit remains almost constant, that portion of the vertical carbon between the level of the mercury and the horizontal electrode forming what may be termed the real resistance of the circuit. It should be noticed, however, that the resistance is not abso- lutely constant, inasmuch as the carbon being consumed, the level of the mercury will be slightly lowered, and so the resistance is really increased by a constant quantity. GAIFFE'S LAMP. This burner returns to electro-magnetic principles. It is enclosed, as regards the mechanism, in a cylindrical case, BURNERS USING THE VOLTAIC ARC. 15 A B C D (Fig. 7). A cover, F, easily lifted off, allows of examination of the mecha- nism, and is clamped by a screw, G. H is the- upper carbon-holder, and H' the lower. I is a racked copper bar, commanding the carbon- holder H, and moves in the interior of a hollow column, J, fixed vertically on the plate A B. This bar terminates in a stop-piece mounted at right angles, to limit the ascending distance. I is a racked soft- iron bar, with a stop-piece, commanding the carbon- holder H'. This bar de- scends vertically into the interior of the coil I. I, a vertical coil, acting, when the circuit is closed, upon the bar K, which then descends under the influence of the attraction to which it becomes subject. 0, two wheels, toothed and turning freely on the axle N. These wheels are insulated from each other by an ivory disc ; their diameters are as 2:1. The larger engages with the bar I, and the smaller with the bar K ; consequently, when the bar K is raised or lowered to a certain extent, the bar I is raised or lowered to double FIG. 7. 16 THE ELECTRIC LIGHT. the extent. This arrangement compensates the unequal con- sumption of the carbons, under the action of a current of constant direction. A barrel, fixed to the wheels 0, holds a clock-spring, one end of which is fixed to the barrel itself, and the other to the axle N ; this acting spring on the barrel, and consequently on the tooth-wheels, tends constantly to approximate the bars I and K, and with these the carbons. M is a steel axle, on which the wheels and the barrel are mounted. This axis is grasped between bearings, which admit of its revolving for the regulation of the barrel-spring. The end of the axle is squared, for the use of a key. The coil I is pierced centrally, to allow for passage of the bar K. The pinions K are mounted on an axle, N, and these may be displaced parallel to themselves, to actuate the wheels 0, and consequently the bars I and K. By these pinions the luminous arc can be maintained by hand in a given focus, without interrupting the action of the apparatus. V is an adjustable clamp acting on the carbon-holder H. N and P are the negative and positive terminals for the conductors of the electric current. X is a bar conducting the current from the terminal P to the column J. Y is a guide-wheel entering, through a slot in the column J, into contact with the bar I, to insure electric communication. The terminals N, P, and the bar X and column J, are insulated with ebonite. The current, entering by the terminal P, passes through X, J, I, V, H, H', K, the coil I, to the terminal N. When the circuit is not complete, the carbons are maintained in contact with each other by the action of the spring in the barrel. When the circuit is completed, the coil attracts the bar K, the movement of which, combined with that of the bar J, deter- mines the distance of the carbons and the production of the voltaic arc. That this action occur, it is necessary for the attractive force of the bobbin to be slightly biased in favour of the spring. If the spring is too tense, the two carbons remain drawn together, or are brought too near to give light of sufficient intensity ; if not tense enough, the action of the coil predominates, the length of arc becomes too great, and the circuit interrupted. BURNERS USING THE VOLTAIC ARC. 17 DUBOSCQ'S LAMP. Foucault's lamp, as perfected by Duboseq, was for many years the electric lamp. Its results were, in fact, second only to those of the Serrin lamp, sub- sequently introduced. The electro-magnet (Fig. 8) attracts an iron plate at the end of a bent lever. A spiral spring balances magnetic attraction, so that contact is made only under certain conditions of current strength. This spring is attached to a small jockey-lever, which ad- mits of adjustment of the spring, and conse- quently of the sensi- tiveness of the lamp. Above the electro-mag- net is a clockwork movement actuating two carbon-holders, in gearing with wheels of different diameters. A rocking -bar connects this clockwork mecha- nism to the armature, and serves as detent to an escapement that arrests the movement FIG. 8. 18 THE ELECTRIC LIGHT. of the clockwork when the arc has suitable length. When the arc surpasses a normal length, and its resistance in- creases, the armature is attracted, and the detent liberates the clockwork. This lamp has had considerable employment in theatrical displays and in laboratory experiments. SIEMENS' LAMP. This lamp has been the subject of considerable trial. Under its original form, it was invented and constructed by Herr von Hefner-Alteneck, who was also the inventor of the Siemens' dynamo-electric machine. This able mechanician succeeded in producing a lamp that, as regard mechanical principles, was at the time of intro- duction, and for long subsequently, unrivalled ; but that had one inherent defect, not unconsidered in the Serrin lamp, namely, want of promptness in action under sudden variations in current strength. This superiority of the Serrin lamp will be apparent when the immediate attraction or direct pull of the electro-magnet in the Serrin lamp is regarded, as compared with the slow movement of separation of the carbons imparted by the to and fro motion of the ratchet and pawl system adopted in the Siemens' lamp. The position of the carbons is regulated, as in the Serrin lamp, by the weight of the upper carbon-holder, which tends to close the carbons together as consumption accrues. The separation is, in the Siemens' lamp (Fig. 9), effected by a small electro- magnetic motor. The upper carbon-holder, moving freely in a vertical plane, is connected to the lower carbon-holder, by a rackwork and toothed wheels. When the approximation of the carbon rods increases the intensity of the current beyond normal limits, the electro -magnet E attracts the armature A. This armature is withheld by the counter force of the spring /, which also retains the bar T, centred at L, against the stop d. When the electro-magnet overcomes the spring and attracts the armature, contact is established at c ; and as soon as the current ceases in the coils of the electro-magnet, the armature BURNERS USING THE VOLTAIC ARC. 19 returns to its initial position. The vibrations of the bar T are communicated by a pawl and ratchet movement, s U, to the carbons, and cause their separation. When the armature assumes its normal position, a stud, n, com- pels the pawl to leave the teeth of the ratchet wheel, and allows the carbon - holder racks free action. The velocity of ap- proach of the carbon- holders is regulated by a fly, w, actuated by the train!?. This train is controlled by a ratchet stop, which is not carried forward by the wheel U, when the pawl s is in action. If this lamp is to be used with currents alternating in direc- tion, the magnet E works in a similar manner, but the oscil- lations of the arma- ture are produced by the mere change of polarity. A button on the case of the lamp allows of causing the racks to engage with toothed wheels, either in the relation of equal ratios or as 1 : 2. The successful working of this lamp is chiefly due to the employment of one 20 THE ELECTEIC LIGHT. point of support for the armature, instead of two. In the latter case, one point of support corresponds to the period of attraction, and the second to the period of release. There are no clockwork springs in the construction. At the contact only weak sparks appear, consequently there is not much wear. Various other forms of lamps have been introduced by Siemens, notably a form identical with Kapieff's and Wilde's candle-lamps ; but, as the latter inventors appear to have priority of publication, their names have been retained to the apparatus as described in this work. SEREIN'S LAMP. This lamp deserves more than honourable mention in the list of apparatus utilized for the production of the electric light. Indeed, the grand prix should be awarded it, as it stands as a source of electric illumination, both in date and practicability, before any other system giving results of a nature beyond those of mere experimental or laboratory researches. M. Serrin's lamp has effected for electric illumination what M. Gramme's machine has equally effected as an electric source. These two inventions, from their practical nature, have been the stepping-stones that have enabled other inventors to cross the brook of success, and due merit must be accorded them in any record of inventions in this branch of technology. For the practical production of the electric light, in any electric lamp dependent upon the production of the arc as a source of light, it is first necessary that the carbon points should be in contact. When the circuit has been thus estab- lished, the points must be separated to a distance sufficient to produce a constant arc. The lamp must be arranged to bring the carbon points together as they are consumed, either under the influence of atmospheric combustion or conveyance of the electric current. M. Serrin's lamp satisfies these conditions, in the simplest manner consistent with the employment of mechanism. This lamp (Fig. 10) con- sists of an electro-magnet A, bars B and C, armature D, BURNERS USING THE VOLT AW ARG. 21 FIG. 10. 22 THE ELECTRIC LIGHT. stop-piece E, spring F, eccentric G, positive carbon-holder H ; tie-pieces I and J, the one fixed, the other adjustable ; tension lever K L ; a double parallelogram M N P Q, with clockwork movement 0, and adjusting screws K and S; a binding screw T, an ivory stop V, in use only when the lamp is out of action. The positive carbon is held above the negative carbon by a massive bar. The following description is related to the various actions of the lamp, which, of course, in practice occur nearly synchronously. The approximation of the carbon points occurs by the positive carbon-holder tending to descend vertically under its gravitating action. At its lower end the holder has a rack, engaging with a toothed wheel 0, communicating motion to the train. On the same arbor is a pulley, of diameter half that of the wheel. A smaller pulley and a linked chain communicate motion to this larger pulley; the chain is attached to a standard, F, form- ing part of the negative carbon-carrier. By this multiplying arrangement the amount of motion given to the negative ca,rbon-holder is half that of the positive holder, which com- pensates for the inequality in consumption of the two carbons under the action of a constant current. The rate of descent of the positive carbon is regulated by a fly and train of wheels, in connection with which is a radial wheel acting as a detent. The requisite distance between the carbon points is at- tained by a somewhat complicated piece of mechanism. The vertical side M Q of a double-jointed parallelogram is fixed, two other sides M N, P Q, have horizontal motion, and to the side N Q is connected a soft-iron armature D. The effect of gravitation of the upper carbon-holder upon the parallelogram is counterbalanced by two springs, one on the lower horizontal side, and the other on the movable vertical side of the paral- lelogram. The tension of the latter spring can be adjusted by a thumb-screw, K, acting upon a bent lever, L K. A is an electro-magnet. The positive pole of the electric source is connected direct to the body of the apparatus, and the electric current passes from the upper carbon to the lower carbon, BURNERS USING THE VOLTAIC ARC. 23 thence by the holder and insulated conductor S to the electro- magnet, which is at the other end of its coil, in connection with the negative pole of the electric source. The electro- magnet overcomes at a given moment the tension of the springs, causing, by its attraction upon the soft-iron armature, the parallelogram to descend, and with it the negative carbon. The vertical movable side of the parallelogram carries a jockey, E, which, as it descends, enters between the arms of the radial wheel, detaining the movement of the train of wheels, and consequently of the racks. When the current enters the apparatus, the electro-magnet becomes excited, attracts the armature, draws down the paral- lelogram and lower carbon, the point of which is separated from that of the upper carbon. As the carbons are con- sumed or separated, the current becomes weakened, a reverse action takes place, and the carbons are brought together by the weight of the upper holder and train of wheels, which are allowed to come into play by the raising of the detent. These actions constantly occurring or being balanced, a constant voltaic arc is maintained, and with chemically pure carbons and regular electric source, a light may be obtained as steady as that of a gas-burner. CAKKtf'S LAMP. M. Carre introduced, in 1875, an improvement in Serrin's lamp, by employing a double solenoid instead of an electro- magnet, with an S- shaped armature, oscillating around a pivot at its centre, the two ends entering a curved bobbin. When the current is interrupted, the armature is withdrawn by springs, a detent releases the mechanism, and the carbons come into contact, the mechanism being driven by the weight of the upper carbon-holder. When the circuit is complete, the armature ends are sucked into the solenoid, and separa- tion of the carbons results. The lamp does not appear to have met with extended application. LONTIN'S LAMP. M. Lontin's improvement upon the Serrin lamp consists 24 THE ELECTRIC LIGHI. in substituting for the electro-magnet a metallic bar, so ar- ranged that its expansion under the heat produced by the passage of the current through it, causes the separation of the carbon points. Although carried out to great perfection in its results, details of this lamp have not been published, and it is still, we believe, considered as being under trial by its inventor. In another improved form, M. Lontin inverts the action of the Serrin lamp, causing the current to be interrupted where, in the ordinary form of lamp, it was continuous. The inversion will be readily understood from consideration of the illustration of the Serrin lamp. M. Lontin has also introduced a form of lamp in which the action of gravity is dispensed with, and allowing of any length of carbon being employed. This lamp is shown in Fig. 11, in which clockwork or an electro -magnet causes a bar, FIG. 11. running parallel with the carbons, to revolve. The motion is imparted by bevelled wheels to others which cause the carbon- carriers to revolve, and to carry the carbons gradually forward. The carriers hold the carbon about two inches from the end of the carbon point, reducing the resistance to that due to this length of carbon, instead of giving the resistance due to the whole length of carbon, as in many systems. This lamp is arranged to work horizontally, but this arrangement is not to be considered an advantage, for experiments have shown that horizontal carbons, placed end to end, give at least 30 BUR NEE S USING THE VOLTAIG AEG. 25 per cent, less light than the same carbons so placed vertically. The cause has not been explained. GIKOUAKD'S LAMP. This lamp (1876) -consists of two parts the actual lamp, with a clockwork movement for approximating and separating the carbons ; and a kind of relay or regulator, placed near the lamp, and actuated by means of a portable constant battery. These two apparatus have distinct circuits, that upon which the relay is placed controlling the advancement and with- drawal of the carbons. The lamp, from its complication of secondary relay mechanism, has not met with general appli- cation. WAY'S LAMP. Professor Way, in 1856, devised a lamp in which the car- bons were replaced by a fine stream of mercury running from a small funnel, and falling into an iron capsule. One pole of the electric source was put in connection with the funnel, and the other with the capsule. The intensely heated liquid vein was enclosed in a glass chimney of narrow dimensions, to prevent condensation of the mercurial vapours, and as the combustion was thus effected out of contact with atmospheric oxygen, oxidation of the metal did not occur. As with all mercurial lamps, great danger arises from volatilization of the mercury, the inventor himself, in this case, falling a victim to the poisonous action. THOMSON AND HOUSTON'S LAMP. % Having been engaged in an extended series of experimental researches on dynamo-electric machines, and their application to electric lighting, the attention of Professors Houston and Thomson was directed to the production of a system that will permit the use of a feebler current for producing an electric light than that ordinarily required; or, in other words, the use, when required, of a current of insufficient intensity to produce a continuous arc. At the same time, the system should permit the use of a powerful current, in such a manner as to operate a considerable number of electric lamps placed 26 THE ELECTRIC LIGHT. in the same circuit. As is well known, when an electric current, flowing through a conductor of considerable length, is suddenly broken, a bright flash, called the extra spark, appears at the point of separation. The extra spark will appear, although the current is not sufficient to sustain an arc of any appreciable length at the point of separation. In this system, one or both of the carbon electrodes are caused to vibrate to and from each other. The electrodes are placed at such a distance apart, that in their motion towards each other they touch, and afterwards recede to a distance which can be regulated. These motions or vibrations are made to follow one another, at such a rate that the effect of the light produced is continuous ; for, when flashes of light follow one another at a rate greater than twenty-five to thirty per second, the effect produced is that of a continuous light. In practice, instead of vibrating both electrodes, it was found necessary to give motion to but one; and since the negative electrode may be made of such size as to waste very slowly, motion is imparted to it, in preference to the positive. The carbon electrodes can be replaced by those of various substances of sufficient conducting power. The following is a description of one of the forms of electric lamp devised to be used in connection with this system of electric lighting : A flexible bar, b, of metal (Fig. 12) is firmly attached at one of its ends to a pillar p, and bears at its other end an iron armature, a, placed opposite the adjustable pole-piece of the electro-magnet m. A metal collar, c, supports the negative electrode, the positive electrode being supported by an arm, j, attached to the pillar p. The pillar p is divided, by insulation at i, into two sections, the upper one of which conveys the current from the binding-post, marked +, to the armj, and the rod r, supporting the positive electrode. The magnet m is placed, as shown by the dotted lines, in the circuit which produces the light. The pillar p is hollow, and has an insulated conducting wire enclosed, which connects the circuit -closer v to the binding-post, marked . The current is conveyed to the negative electrode, through b and the coils of the magnet m. When the electrodes are in con- BURNERS USING THE VOLTAIC ARC. 27 FIG. 12. tact, the current circulating through m renders it magnetic and attracts the armature a, thus separating the electrodes, when, on the weaken- ing of the current; the elasticity of the rod b again restores the con- tact. During the move- ment of the negative electrode, since it is caused to occur many times per second, the positive electrode, though partially free to fall, cannot follow the rapid motions of the negative electrode ; and, therefore, does not rest in permanent contact with it. The slow fall of the positive electrode is ensured either by properly pro- portioning its weight, or by partly counter- poising it. The posi- tive electrode thus becomes self-feeding. The rapidity of move- ment of the negative carbon may be con- trolled by means of the rigid bar I, which acts, practically, to shorten or lengthen the part vibrating. In order to obtain an excellent but free contact of the arm with the positive electrode, the rod, made of iron, passes through a cavity filled with mercury, placed in electrical con- tact with the arm. Since the mercury does not wet the metal rod, or the sides of the opening through which it passes, free 28 THE ELECTRIC LIGHT. movement of the rod is allowed without any escape of the mercury. This feature could be introduced advantageously into other forms of electric lamps. In order to prevent a break from occurring in the circuit, when the electrodes are consumed, a button, u, is attached to the upper extremity of the rod R, at such a distance that when the carbons are consumed as much as is deemed desirable, it comes into contact with a tripping lever T, which then allows two conducting plugs, attached to the bar v, to fall into their respective mercury cups, attached, respectively, to the positive and negative binding-posts by a direct wire. This action practically cuts the lamp out of the circuit. REYNIEK'S LAMP WITH CONTINUOUS CIRCUIT. If a very intense current of electricity is led through a resisting and refractory conductor, such as a pencil of carbon, the temperature of this conductor rises to a dazzling white heat, and it then emits a vivid light. The principal difficulty to be overcome is to limit the undue waste of the luminous conductors a waste which is very rapid, even in an enclosed space, on account of the volatilization and disaggregation of the carbon pencils, and greatly accelerated in the open air, by the rapid combustion of the incandescent carbon. In the various systems of electrical lamps with continuous conductors, the renewing of the carbon points is performed in the follow- ing manner : The incandescent pencil is placed in the circuit with fixed contacts, and remains until the circuit is broken by the carbon being consumed ; the light is then extinguished. The current now suddenly passes from this carbon to another, which is consumed, the circuit broken in its turn, and so on. This method is open to many objections : there is an inter- ruption of the current, accompanied by an extinction of the light, at every rupture of the pencil ; the luminous intensity varies continually on account of the gradual thinning of the carbon; the conductor only gives its maximum of light at the moment next to that of rupture ; finally, the proposed apparatus can scarcely work, except in an hermetically closed space. BURNERS USING THE VOLTAIC ARC. 29 FIG. 13. In Eeynier's system the renewal of the carbon is pro- gressive. The carbon, incandescent a part of its length, advances almost continuously, till the whole available part has been consumed. This system can operate in the open air. The following is the principle : A cylindrical or pris- matic pencil of carbon forms part of an electrical circuit, continuous or alternate, sufficiently intense to render this part incan- descent. The current enters or leaves at the point of contact (Fig. 13) ; it leaves or enters at the lower contact wheel. The upper contact, which is elastic, compresses the pencil late- rally ; the contact wheel touches it at its end. Under these conditions, the carbon is consumed at its extremity more quickly than at any other place, and tends to dimmish in length. Con- sequently, if the carbon is steadily forced in the direction of the arrow, it will gradually advance as it is con- sumed, sliding through the lateral contact, so as to press continuously on the contact wheel. The rotation of this wheel is made dependent on the progressive movement of the car- bon, so that the weight of the latter, exerted at its end, acts as a brake on the mechanism of the motion. This apparatus gives a clear white light with four Bunsen elements ; with a more powerful electrical source, several lamps of this system may be operated. With a battery of thirty- six elements, grouped in two series of eighteen each, four lamps have been placed in a single circuit. Each of the four lamps could be extinguished and relighted individually, the three others 30 TEE ELECTRIC LIGHT. continuing unaffected. Light has been obtained from one of these lamps by means of the current of a small Gramme machine, worked with treadle, as employed for the labora- tory. Finally, a fine light has been obtained with a battery of three Plante (secondary) elements, which were charged during the afternoon at the establishment of M. Breguet, and carried charged to the hall where they were exhibited. M. Eeynier remarks that this experiment may be considered as a step towards the application of the electric light to domestic purposes. In the most recent arrangement of this lamp, the revolu- tion of the turning contact is obtained from the tangential component of the pressure of the carbon pencil on the cir- cumference of the disc ; thus, the end of the pencil never leaves the revolving contact, and all causes of irregularity in the light are obviated. The brake for retarding the progress of the carbon rod is operated in the following manner : The contact wheel is carried by a lever ; the pressure exerted by the carbon on the wheel causes a shoe to press on the face of a wheel, which is revolved by means of the weight of the heavy rod, through its rack and the pinion. Accordingly, as the point of the luminous conductor presses more or less heavily on the disc, the brake will retard, more or less, the descent of the heavy column, which occurs at almost inappreciable intervals. THE WERDERMANN LAMP. The arrangement of this lamp will easily be understood from Fig. 14. A block of carbon is connected to the nega- tive pole of the electric source, the positive pole being con- nected to the carbon rod, some three or four millimetres in diameter, and of any desirable length. The carbon rod is kept in contact with the block by means of a weight and a cord passing over a pulley. The whole arrangement is exceedingly simple. The method of arranging the lamps in circuit is that known as the multiple-arc system, used to BURNERS USINa THE VOLTAIC ARC. 31 FIG. 14. so great an extent in every branch of electricity, in which, if we suppose the two conductors from the source of elec- tricity to be led away in parallel lines, the lamps will be conductors connect- ing these lines. The experi- ments with this lamp, at the time of going to press, have not been publicly detailed, but there seems to be a large field for a system so simple in construction. It might be argued that this is a lamp in which illumination is due to in- candescence of the carbon simply, but M. Werdermann contends that repulsion be- tween the carbon rod and disc gives a small arc, which adds greatly to the brilliancy of the light. With a 2 h.-p. Gramme plating machine ten lights of 40 candle-power each have been maintained on this system, or two lights of 320 candle-power each. The loss of light, by subdivision, in- creases therefore far more rapidly than in proportion to the number of lights introduced, as might be expected from calculation. But the loss is not sufficient to confirm certain theories, in which it is held that the light would decrease inversely as the square of the number of lamps introduced into the circuit, or rather of multiple circuits added. EAPIEFF'S LAMP. The principal feature of M. EapiefFs lamp consists in the multiple nature of one or both carbon electrodes. Instead of employing, as in most electric lamps, a carbon rod placed THE ELECTRIC LIGHT. FIG. 15. vertically and in the same axis with another carbon rod, M. Kapieff substitutes for the single rod two others, each of half the sectional area. These rods are inclined to one another at an angle of about 20, and meet in V form. The electric arc is pro- duced between the upper and lower pairs of carbons. The position of the carbons is determined by the intersection of the two straight lines or axes of the carbon rods, so that a constant length of arc is neces- sarily consequent, whatever may be the rate and irregularity of con- sumption. Each carbon rod moves freely between guides in the direc- tion of its length, and is drawn through these guides by a cord and weight, to form the apex of the V with the other rod. The motion forward is stopped by the two car- bons impinging against each other. The plane of the upper pair of carbons is at right angles to that of the lower pair. Fig. 15 illus- trates a lamp of this construction, and is shown with the carbons in the position they would occupy when the current is interrupted. Under this condition, the lower pair is kept in contact with the upper pair by the action of a light spiral spring situated in the base of the appa- ratus, and acting through a vertical rod passing up one of the pillars. To the free end of each of the car- bon rods is attached, by a screw BURNERS USING THE VOLTAIC ARC. 33 clip, a silk thread, which, passing over pulleys, is attached to a rectangular weight, sliding vertically up and down the two pillars. The action of this weight is to draw the carbon pairs together. When the lamp is put 'in circuit, and an electric current is established, the lower pair of carbons is drawn away from the upper by the action of the electro-magnet concealed in the base of the lamp. Fig. 16 FIG. 16. represents the base of the lamp as seen from beneath, and shows the electro-magnet and its accessories. This appa- ratus consists actually of two electro-magnets, one of which is fixed, while the other is hinged so that the passage of the current through the coils causes the hinged magnet to approach the fixed one, and thus to lift the sliding rod that passes up the vertical pillar of the lamp, at the same time separating the carbons. The spiral spring withdraws the shifting electro-magnet when the current ceases or become weakened. The base or stand of the lamp contains also another apparatus, consisting of an automatic shunt, which throws into the circuit a resistance equivalent to that of the arc, when, through any cause, the lamp is extinguished. By 34 THE ELECTRIC LIGHT. FIG. 17. means of this apparatus, the brilliancy of the other lamps in circuit is unaffected by accidental or intentional extinguishing of the remaining lamps. When the arc ceases to exist, the current ceases to pass through the coils of the electro-magnet, and this, losing its attractive power, releases the armature, which is drawn back by a spring. The armature carries a contact-piece, which, when the armature is released, falls against a fixed contact-piece, and introduces into the circuit an "artificial resistance" of carbon attached to this latter contact-piece. The lamp thus described has been in practical and public use at the office of the Times newspaper. In this case as many as six lamps have been ignited on a single circuit; but details of cost and power ex- pended have not been published. Fig. 17 represents another form of the Kapieff lamp. In this case the two pairs of carbons are arranged side by side, instead of being placed vertically above each other. Above the arc is a cake of lime, which serves to reflect the rays, as well as to in- crease the illuminating power, by preventing radiation. Either of these lamps can be employed with currents of single or alternate direction. In one instance, the pairs of carbons are equally consumed ; in the other, the positive carbons have to be made twice as long as the negative carbons. At all times, the point of intersection of the axes of the carbons, and conse- quently the length of arc, must remain constant. BURNERS USING THE VOLTAIC ARC. 35 H BRUSH'S LAMP. One of the simplest and apparently most efficient lamps is due to American invention. It is illustrated in Fig. 18. A is a helix of insulated wire, in the form of a tube or hollow cylinder FlG - 18> resting upon an insulated plate, A', upheld by a metallic post or standard. Within the cavity of the helix A is contained the iron core C and the rod B, which passes loosely through the core C. The core C is also made to move very freely within the cavity of the helix A, and p it is partially supported within the cavity by the springs, whose tension is regulated by the set screw. These springs push upward against ears attached to the core C. D is a brass ring. One edge of this ring is over a lifting tongue, which is at- tached to the core C, while the op- posite edge of the ring is a short distance below the crown of an adjustable set screw, D'. The upper standard is fastened to a suitable base, to which is also attached the mechanism for holding the lower carbon F. This mechanism consists of a support, G, terminating in a part similar in construction to the part B'. The lower part of this support is bent at a right angle and rests upon the base, and is fastened by a thumb-screw. It is necessary that the carbons, F F, should present in accurate opposition to each other, and to accomplish this the set screw is made to pass through a hole in the support G, considerably larger than the shaft of the set screw. 36 THE ELECTRIC LIGHT. If one pole of a battery or other source of electricity be attached to the support G, while the other pole is connected to the upper support, the electric current passes from the latter through the helix A, rod B, and carbons F, F, down to the support G, thus completing the circuit. The core C, by force of the axial magnetism thus created, is drawn up within the cavity of the helix, and by means of the finger C', lifts one edge of the ring D, until, by its angular infringement against the rod B, it clamps this rod, and also lifts it up to a distance limited by the adjustable stop D'. While the ring retains this angular relation, the rod B will be prevented from moving. While the electric current is not passing, the rod B can slide readily through the loose ring D and .the core C, and in this condition the force of gravity will cause the upper carbon to rest upon the lower carbon. If a current of electricity is passed through the apparatus, it will effect the lifting of the rod B, and separate the carbons, thus producing the electric arc. The tension of the springs is so adjusted that they, together with magnetic attraction of the helix, shall be just sufficient to support the core C, rod B, and carbon F, in the position for producing the arc. As the carbons burn away, increasing the length of the arc, the electric current diminishes in strength, owing to the increased resistance. This weakens the magnetism of the helix, and the core, rod, and carbon F are moved downward by the force of gravity, until the con- sequent shortening of the voltaic arc increases the strength of the current, and stops this downward movement. After a time, however, the clutch-ring D will reach its floor or sup- port, and its downward movement will be arrested. Then any downward movement of the core C, however slight, will at once affect the rod B, allowing it to slide through the ring D, until it is arrested by the upward movement of the core C, due to the increased magnetism. In continued operation, the normal position of the ring D is in contact with its lower support, the office of the core C being to regulate the sliding of the rod B through it. If, however, the rod accidentally slides too far, it will be instantly BURNERS USING THE VOLTAIC ARC. 37 and automatically raised as at first, and the carbon points thus continued in proper relation to each other. With this lamp, and the machine also designed by Mr. Brush, as many as 16 to 18 lights of 2000 candle power each have been maintained upon a single circuit. With 27 lamps in circuit, upon an experimental trial, the aggregate light was about 10,000 candles ; while, with 16 lamps in circuit, the total light was from 30,000 to 35,000 candles, with an expenditure in the latter case of 13*85 horse power. These are results unattained at present by any other system, and are worthy of record, as being performed under the care of highly reliable engineers. THE WALL ACE -FARMER LAMP. In it no car- Fro. 19. This lamp (Fig. 19) has a peculiar feature, bon rods are employed ; but the carbons take the form of two plates, each about nine inches long and three inches broad, the upper or positive plate being double the thick- ness of the lower plate. The lower plate is fixed, but the upper plate slides in a grooved frame. Above the frame an electro - magnetic apparatus provides for the separation and contact of the plates, as the strength of the current may regulate. The arc, always seeking the position of least resistance, shifts from one end of the plates to the other. This peculiarity, whilst it gives great continuity of action, renders the lamp useless for purposes where a fixed position is required for the light, as in lighthouse illumination. The carbon plates of the dimensions given are stated to 38 THE ELECTRIC LIGHT. last for 100 hours, in conjunction with the current from a Wallace-Farmer machine, with six other lamps in circuit. The intensity of the light afforded, and power expended, have not been detailed publicly. THE KKUPP LAMP. FIG. 20. FIG. 22. FIG. 21. This lamp, invented by Baron von Krupp, includes the application of a brake for the automatic regu- lation of the distance between the two carbon points. A fan or fly revolves in quicksilver, for the purpose of regu- lating the motion of the carbon-holder, this part of the apparatus being designed as a substitute for clockwork. A mag- netic coil, with iron casing and iron bottom, is employed in connec- tion with the brake. Fig. 20 is a side eleva- tion, and Fig. 21 a front elevation of the lamp. A is the holder for the positive carbon, and B the holder for the nega- tive carbon. The upper holder A is suspended from the pulley C by a jointed chain, the lower holder B being similarly attached to a pulley, D, half the size of the former. When the holder A descends a certain distance by its weight, the other holder, B, ascends half the dis- BURNERS USING THE VOLTAIC ARC. 39 tance. Accordingly, the electric arc formed between the carbon points occupies a fixed position. As the weight of the upper holder A must not be too small, because its motion would be easily influenced by dust and dirt, it is necessary to have a brake for retarding and regulating its course or travel. For this purpose a fly or fan, E, revolves in mercury. On the spindle of this fly there is a pinion, F, gearing with a tooth wheel, G, on the spindle X of the chain pulleys C and D. In order that the fly E, by the insertion of a fresh carbon, may not revolve backwards, the tooth-wheel G is fitted with a pawl wheel, H. Fig. 22 is a separate view of the brake acting on the disc I. The brake consists of two parts, K and M, which are jointed together at L. The lower part, M, can turn on the spindle X, and has a hole, M', in which is inserted a small peg, N (Figs. 20 and 21). The peg is loose in the hole, and the backward motion of the brake is limited by it. is a brake block in the upper part, K, of the brake. P is the keeper for an electro- magnetic coil, Q, and this keeper is suspended by a brass rod from the other end of the part K. When the lamp is in action, the keeper P is drawn into the coil Q, and the brake block is pressed against the disc I, turning the latter in its further movement downwards, so far as the set screw K (Fig. 20) will allow. Thus, the upper carbon point will be raised, and the lower carbon point lowered, and the electric arc make its appearance. As the carbon points gradually consume away, the current becomes weaker, and its effect on the electro-magnet Q is lessened. The brake K, supported by the spring S (the action of which can be regulated, in proportion to the strength of the current, by the lever U and set screw V) and by the weight of the carbon-holder, moves slowly back. The brake disc I is enabled to turn forward, and the carbon points to approach each other. When this movement has proceeded as far as the brake disc I moved back before, the lower part of the brake bears against the peg N. By further weakening of the current, the brake now turns in its joint at L, the brake block releases the disc I, and the carbon points move towards 40 THE ELECTRIC LIGHT. each other ; the current is strengthened, and the brake is again applied to the disc I, either simply to hold it when the carbon points are in their right position, or to pull it back when the carbon points are too close together. When in- serting new carbons, the brake is fixed by the set screw W, and work is arrested. The electro-magnetic coil Q rests on the bed-plate T of the lamp, and is surrounded by an iron casing, by which its power of attraction for the keeper is increased. The fixed position of the arc provides for keeping the light in the centre of a reflector. The lamp may be simplified by leaving out the moving parts for the lower carbon-holder. The lamp has been employed by Von Krupp in portions of his factory at Essen, in Germany, and the results have been so satisfactory that the light is being extended to other parts of the establishment. HIGGS' LAMP. With considerable experience in the practical manipulation of most of the existing systems for electric lighting, the author has found that some systems presented considerable, and that others of more delicacy of adjustment necessary to practical employment gave insuperable, difficulties. These difficulties arise chiefly from three causes : that the carbons are not homogeneous, the current inconstant from variations in the resistance of the circuit, and generally from the want of promptness in the mechanism of the lamp to respond to the variations so caused. Most of these electrical apparatus have been devised either with too broad or imperfect views, or with only special application. In the case of carbon-holders carrying seven or eight inches of carbon to be consumed, the resistance of this amount of carbon, if the material is not of the highest quality, is likely to exceed that of the arc and lamp itself ab initio ; and this resistance, constantly varying, has to be compensated for by the mechanism of the lamp. This imperfection has been avoided by Kapieff, Werdermann, and Lontin. Some inventors have recognized it, and have proposed, as a remedy, to electrotype the carbon BURNERS USING THE VOLTAIC ARC. 41 rods with a conducting metal ; but the practical electro- metallurgist is cognizant of the difficulty of obtaining regu- larity in such deposits of metal, and the expense of the coating is, besides, to be taken into account. In each case the long length of carbon becomes heated, and with coppered carbons the mass of metal is insufficient to carry away the heat generated by the passage of the current. The first aim of the electrician who has to maintain an efficient and steady light is undoubtedly to obtain as constant a current as is possible with either battery or electric machine. Battery currents vary, as a rule, very gradually; currents produced by mechanical motion are subject to the irregu- larities of that motion. The slip of belting, the beats due to want of balance in the fly-wheel, are represented in the electric current with too much fidelity for the comfort of the electric-light engineer. But with care these causes of irregularity can be avoided, and the needle of even a delicate galvanometer, interposed in the circuit of a well-set machine, driven by a steady motor, will remain fixed at a degree of deflection representing the current strength. But this steadiness vanishes immediately the electric lamp is introduced into the circuit, so far as the lamp is concerned. This occurs partly because the lamp is, as regards the light it emits, a much more delicate current measurer than the galvanometer. The light power from the carbons of an electric lamp depends not directly upon the current strength, but increases or decreases far more than proportionally. The heat produced by the current varies as the square of the current strength, and the light varies in some such ratio as regards the heat. Thus, a variation in current intensity, measured by the number 2, may be considered in illustration as causing a variation of 16 in the light intensity. It is, therefore, needful to avoid causes of variation in the lamp, for these, it is evident, will have similar effect upon the light intensity to those arising with the machine. Indeed, varia- tions introduced by the lamp cause variations to occur from the machine, unless the latter be extremely well governed. A decrease of resistance in the circuit causes more work to 42 THE ELECTRIC LIGHT. be thrown upon the motor, and vice versa. If the motor, in consequence, momentarily slackens or increases speed, there must elapse several moments before the same conditions as existed before the disturbance are again established, and the largest of these variations are certainly visible as variations in the light. That most of these variations are due to reaction from variations in the lamp itself, is proved by the superior steadiness of the light produced on the principle of incandescence alone. This fact has caused many in- ventors to overlook the cost of the light produced merely by incandescence, and to avoid in their lamps the use of the voltaic arc, with what appears to be its necessarily attendant irregularity. The lamp illustrated in Fig. 23 is an attempt to avoid as much as possible causes of irregularity, and at the same time to produce a light with small expenditure of power. It utilizes the principles of incandescence, of the arc, and of the extra spark. It consists of an electro-magnet, in face of which is an armature mounted on a spring. The armature carries a block of carbon, iron, or compounded material as a negative electrode, which is not consumed, or is consumed with extreme slowness. The positive electrode is a carbon rod, carried in a tube and falling with a certain friction im- posed by a weighted lever, which admits carbon rods of several sizes to be introduced, as may be best suited to the strength of the current. The falling of the carbon can be aided by a weight or spring. The distance of the bottom of the tube from the negative electrode can be adjusted, and limits the length of carbon rod rendered incandescent by the current. When the current passes, through the positive carbon coming into contact with the negative electrode, the armature is attracted, and the voltaic arc and extra spark appear; the current, weakened by this action, fails to keep the armature attracted, and in this manner a constant vibration of the negative electrode is established. This vibration is im- perceptible to the eye. Its advantage, beyond that of pro- ducing the extra spark, which spark itself appears to afford aid in maintaining the voltaic arc, is that the armature has BURNERS USING TEE VOLTAIC ARC. 43 no dead point, and floats as it were above the electro-magnet, in a condition to respond promptly to magnetic effects caused by larger increments or decrements of current strength. FIG. 23. It is preferable to place an insulated spring between the end of the friction-lever and the armature, instead of the 44 THE ELECTRIC LIGHT. weight on the end of the lever; the carbon rod is then allowed to fall freely when required, and as released by the rising of the armature. With only four ordinary Bunsen elements, sufficient light has been obtained to illuminate a shop 60 feet by 40 feet ; and upon the single circuit of a dynamo-electric machine absorbing 2J horse-power, four lights of about 400 candle- power each have been obtained, with sufficient steadiness to read by with comfort. CHAPTER III. ELECTRIC "CANDLES" AND CANDLE-LAMPS. JABLOCHKOFF'S CANDLE. AT the time when the construction of electric lamps was becoming more complicated in each successive instance, and practical men were under the impression that first cost in each lamp would effectually prevent more than special employment of the light, M. Jablochkoif, in March, 1876, brought out an electric "candle." As by this invention mechanism is entirely dispensed with, the impetus given to electric lighting was immense, and whatever objections may be raised against the system, there is no doubt that its intro- duction was the installation of electric lighting as a branch of engineering that has since and will steadily grow in importance. A "candle" consists of two cylindrical carbon rods about three-sixteenths of an inch in diameter, each weighing about 8 grains per inch. These rods, varying from 6J inches to 10 inches in length, are placed vertically side by side, with about three-sixteenths of an inch space between them, which is filled in with plaster of Paris. The combination constitutes a " candle." It is inserted in a holder shown in Fig. 24, and there held merely by a spring clip. To complete the circuit and to start the lighting of the candle, there is laid horizontally, from top to top of the carbon rods, a small piece of graphite or lead from a drawing pencil. When once lighted, combustion is maintained by fusion of the plaster of Paris, and the candle, if once extinguished, cannot be relighted. This want of capability of relighting is a most serious objection to the system. 46 THE ELECTRIC LIGHT. The fusion of the insulating material, the plaster of Paris, absorbs at least 30 per cent, of the electric current, which FIG. 24. is thereby wasted. The relative consumption of carbon in this candle is shown by the following table, compiled from the Journal of the Franklin Institute. Light in Candles. Length consumed in inches per hour. Approximate Weight per in. in grains. Grains of Carbon per hour per Candle. + - 1230 1-78 0-34 36-2 0-062 900 1-91 0-58 36-2 o-ioo 440 2-45 073 20-3 0-146 705 3-15 0-55 20-3 0-106 760* 3-0 3-0 7'5 0-060 * Candle. CANDLES AND CANDLE-LAMPS. 47 The first four lights were obtained with constant current machines, the size of the carbons in the first two being f x f of an inch, and in the third and fourth, one-fourth of an inch square. The shorter Jablochkoff candles average only 1J hour in duration, and at the end of this time another candle has to be put in circuit. This is generally effected by hand, by a switch FIG. 25. or commutator shown in Fig. 25, which represents the arrange- ment of the lamps and machines upon the Jablochkoff- Gramme system, usually employed with the Jablochkoff candle. The 48 THE ELECTRIC LIGHT. single-direction current generated by a small Gramme ma- chine is conveyed to magnetize the electro-magnets of a larger or "distributing" machine, whence currents of alternating direction are taken to the lamp. In each lamp are usually mounted four candles, affording light from six to nine hours. To effect the ignition of a fresh candle, when required, an automatic arrangement has been devised, but has not been generally applied. It consists (Fig. 26) of a pivoted bent FIG. 26. lever, pressed by a spring against the side of the candle. This lever at its other end makes contact with the connection of a second candle, when released by the consumption of the first. DE MERITENS' CANDLE. The objection to the Jablochkoff candle on the score of difficulty in relighting, and loss of current consumed in the fusion of the insulating material, has led to several attempts in remedy. One of these, by M. de Meritens, consists in placing between the two carbon rods, but not in contact with them, a third rod, of about half the diameter, instead of the CANDLES AND CANDLE-LAMPS. 49 insulating substance. The electric arc plays from the outer carhons to the intermediate rod, which is consumed. The arc, thus divided, has less probability of total extinction, and requires lower expenditure of power to produce, as it has not so great distance to leap. RAPIEFF'S CANDLE-LAMP. FIG. 27. This invention (Fig. 27) is a return to mechanical aid. There is no insu- lating material inserted between the carbon rods, and their distance apart can be regulated by a screw adjustment. The holder of one of the carbons is connected to the armature of an electro- magnet concealed in the stand. When no current is passing, the upper ends of the rods are brought into contact by a spring attached to the armature of the movable carbon. When a current passes, the armature is attracted, and the carbons separated to the distance necessary to produce the arc. Upon interruption of the current, the armature is released, and the circuit again completed. WILDE'S CANDLE-LAMP. Mr. Henry Wilde, in a paper read before the Manchester Literary and Philosophical Society, after describing the Jablochkoff candle, says: "My connection with the history of this system of lighting placed me in a position to make some experiments with the Jablochkoff candle, and led to the dis- covery of the following facts : One of the conditions necessary for producing a constant light from the candle, in its most recent form, was that the quantity and intensity of the alter- nating current should be such that the carbons consume at a rate of from four to five inches per hour. If the electric current were too powerful, the carbons became unduly heated, 50 THE ELECTRIC LIGHT. and presented additional resistance to the passage of the current ; the points at the same time lost their regular conical form. If, on the other hand, the current were too weak, the electric arc played ahout the points of the carbons in an irregular manner, and the light was easily extinguished by currents of air. "In the course of these experiments I was struck with the apparently insignificant part which the insulating material played in the maintenance of the light between the carbon points ; and it occurred to me to try the effect of covering each of the carbons with a thin coating of hydrate of lime, and mounting them parallel to each other in separate holders, and without any insulating material between them. The use of the lime covering was intended to prevent the light from travelling down the contiguous sides of the carbons. On completing the electric circuit, the light was maintained between the two points, and the carbons were consumed in the same regular manner as when the insulating material had been placed between them. " Two plain cylindrical rods of carbon, three- sixteenths of an inch in diameter and eight inches long, were now fixed in the holders parallel to each other, and one-eighth of an inch apart. The strength of the alternating current was such that it would fuse an iron wire 0'025 of an inch in diameter and eight feet in length. On establishing the electric current through the points of the carbons, by means of a conducting paste composed of carbon and gum, the light was produced, and the carbons burnt steadily downwards as before. " Four pairs of naked carbons mounted in this manner were next placed in series in the circuit of a four -light machine, and the light was produced from these carbons simultaneously, as when the insulating material was used between them. The light from the naked carbons was also more regular than that from the insulated ones, as the plaster of Paris insulation did not always consume at the same rate as the carbons, and thereby obstructed the passage of the current. This was evident from the rosy tinge of the light produced by the volatilization of the calcium simultaneously with the diminution CANDLES AND CANDLE-LAMPS. 51 of the brilliancy of the light from the carbons. The only function, therefore, which the insulating material performs in the electric candle, as shown by these experiments, is that it conceals the singular and beautiful property of the alternating current to which I have directed attention. "As I have already said, the strength of the alternating current must bear a proper proportion to the diameter of the carbons used ; and when a number of such lights are required to be produced in the same circuit, the quantity and property of the current will remain constant, while the tension will require to be increased with the number of lights. " This simple method of burning the carbons will, I believe, greatly further the development of the electric light, as the carbons can be used of much smaller diameter than has hitherto been possible. They may also be of any desired length, for as they are consumed they may be pushed up through the holders without interrupting the light. One of these developments will be a better method of lighting coal and other mines. In this application the alternating currents or waves from a powerful electro-magnetic induction machine may be used for generating, simultaneously, alternating secondary currents or waves in a number of small induction coils, placed in various parts of the mine. The light may be produced in the secondary circuits from pairs of small carbons inclosed in a glass vessel having a small aperture to permit the expansion of the heated air within. Diaphragms of wire gauze may be placed over the aperture to prevent the access of explosive gas. By generating secondary currents or waves without interrupting the continuity of the primary circuit, the contact-breaker is dispensed with, and the subdivision of the light may be carried to a very great extent." To initiate the light in the Jablochkoff system, it is neces- sary to complete the electric circuit between the carbons by means of some conducting substance, which volatilizes on the passage of the current, and establishes the electric arc between the points. When a number of such lights are produced simultaneously from the same source of electricity, any inter- ruption in the continuity of the current extinguishes all the 52 THE ELECTRIC LIGHT. lights in the same circuit, and each pair of carbons requires to be reprimed before the lights can again be established. This defect, as will be obvious, would cause great inconve- nience when the lights are not easily accessible, or are at considerable distances apart. In the course of Mr. Wilde's experiments, it was observed that when the electric circuit was completed at the bottom of a pair of carbons close to the holders, the arc immediately ascended to the points, where it remained so long as the current was transmitted. This pecu- liar action of the arc was first thought to be due to the as- cending current of hot air by which it was surrounded. This, however, was found not to be the cause, as the arc travelled towards the points, in whatever position the carbons were placed, whether horizontally or vertically in an inverted posi- tion. Moreover, when a pair of carbons were held in the middle by the holders, the arc travelled upwards or down- wards towards the points according as the circuit was estab- lished above or below the holders. The action was, in fact, recognized to be the same as that which determines the pro- pagation of an electric current through two rectilinear and parallel conductors submerged in contact with the terrestrial bed, which was described in the Philosophical Magazine for August, 1868. In all the arrangements in general use for regulating the electric light, the carbon pencils are placed in the same straight line, and end to end. When the light is required, the ends are brought into momentary contact, and are then sepa- rated a short distance to enable the arc to form between them. The peculiar behaviour of the electric arc when the carbons are placed parallel to each other, suggested the means of lighting the carbons automatically, notwithstanding the fact that they could only be made to approach each other by a motion laterally, and to come into contact at their adjacent sides. To accomplish this object, one of the carbon-holders is articulated or hinged to a small base-plate of cast iron, which is so constructed as to become an electro-magnet when coiled with a few turns of insulated wire. The carbon-holder is made in the form of a right-angled lever, to the short hori- CANDLES AND CANDLE-LAMPS. 53 zontal limb of which is fixed an armature, placed over the poles of the electro-magnet. When the movable and fixed carbon-holders are brought into juxtaposition, and the carbons inserted in them, the upper parts of the two carbons are always in contact when no current is transmitted through them. The contact between the carbons is maintained by means of an antagonistic spring, inserted in a recess in one of the poles of the electro-magnet, and reacting on the under side of the armature. One extremity of the coil of the electro- magnet is in metallic connection with the base of the carbon- holder, while the other extremity of the coil is in connection with the terminal screw at the base of the instrument from which it is insulated. The coils of the electro-magnet are thus placed in the same circuit as the carbon pencils. When the alternating current from an electro-magnetic induction machine is transmitted through the carbons, the electro- magnet attracts the armature and separates the upper ends of the carbons, which brings them into their normal position, and the light is immediately produced. When the circuit is interrupted, the armature is released ; the upper ends of the carbons come into contact, and the light is produced as before. When several pairs of carbons are placed in the same circuit, they are, by this arrangement, lighted simultaneously. SIEMENS' CANDLE-LAMP. This is precisely similar to the preceding in principle. A modification consists in employing a bar of iron drawn into an electro-magnet coil, instead of the ordinary electro-magnet and armature, to cause separation of the carbons. 54 THE ELECTEIC LIGHT. CHAPTEK IV. LIGHTING BY INCANDESCENCE. THE production of the electric light by the incandescence of a badly conducting substance has long been a favourite idea with inventors, and it apparently offers many practical advantages. The renewal of carbons, as an instance, is either dispensed with, or the frequency of renewal lessened. Burners on this principle can be made so perfectly automatic that they shall not require attention to initially ignite them, but become incandescent immediately the current is directed through them. Thus, a material saving in the cost of main- tenance is effected. To the present, however, lighting by incandescence has been considered to absorb more power to produce an equal amount of light ; and, although this is undoubtedly true, the relation of light produced to power expended has not been made the subject of direct practical trial, leaving the question open as to the relative economy between lighting by the voltaic arc and by this system. It has been determined that, where light-centres of great brilliancy are desired, the voltaic arc must be employed, because incandescence cannot be made, under any expendi- ture of power, to realize the same intensity of light as is produced by the voltaic arc. Hence, comparison between the value of the two systems is limited to lights of moderate power; and it is probable that, if electricity is to become available for all powers of light required, the two limits will be the voltaic arc for lights of great power, and the use of incandescent conductors for lights of low power ; whilst between these two limits, and tending towards the lower, the two systems will be combined. The system of lighting by incandescence depends upon the principle that a bad conductor becomes heated during the LIGHTING BY INCANDESCENCE. 55 FIG. 28. passage of an electrical current, and when heated, emits light. When a body is at a temperature of 250 degrees, it may be called warm, 500 hot. At 1000 we have the heat rays, 1200 , orange rays, 1300 , yellow rays, 1500 , blue rays, ,,1700 , indigo rays, 2000 , violet rays. So that any body raised to a temperature above 2000 C. will give us all the rays of the sun. Given the resistance of a conductor, it is easy for an electrician to calcu- late the amount of current required to heat it to a certain temperature. The conductors generally employed to afford light are slender carbon rods, platinum wire or strip, or iridio-platinum, which is an alloy of the metals iridium and platinum. Platinized as- bestos has also been proposed. KING'S BUHNER. King, in 1845, appears to have been the first to have considered this method of lighting. The following passages are taken from the specification of the patent : " The invention has for its basis the use of metallic conductors, or of continuous carbons, heated to whiteness by the passage of an electric current. The best metal for this purpose is platinum ; the best carbon is retort carbon. When carbon is employed, it is useful, on ac- count of its affinity for oxygen at high tem- peratures, to cover it from air and moisture, as in Fig. 28. The conductor C rests on a bath of mercury ; the bar B is in porcelain, it serves to support the conductor C ; the conductor D is fixed on the bell by an hermetically sealed joint. The carbon rod A rests at top and bottom on conducting blocks, and becomes incan- 56 TEE ELECTRIC LIGHT. descent by the passage of an electric current. A vacuum is previously established in the bell, and the apparatus veritably forms a barometer with one of the poles of the battery in communication with the column of mercury, and the other with the conductor D. The apparatus, properly closed, may be applied to submarine lighting, as well as to the illumina- tion of powder mills and of mines, especially where the danger of explosion is feared, or the rapid inflammation of very com- bustible substances. When the current is of sufficient in- tensity, two or a larger number of lights may be placed in the same circuit, care being taken to regulate the power of the magneto-electric machines, or the elements of the battery producing the current." STAITE AND PETRIE'S BURNER. In 1846 and 1849 Greener, Staite, and Petrie introduced similar ideas to King's, and Petrie inserts in his patent for a lamp the following suggestion : " A light may be produced by passing an electric current through a short and thin con- ductor, which heats and becomes luminous ; but the majority of substances fuse and burn rapidly : however, I obtain a good light by using iridium, or one of its alloys. Iridium may be fused so as to produce an ingot whilst it is submitted to the heat of the voltaic arc ; afterwards it may be decarbonized and rendered more malleable. It can be cut into small pieces of O'OOl metre diameter and O'OIO to 0'020 metre length, that can be fixed upon two insulated metallic supports, which are in connection with the two wires of a proper galvanic battery. There is then obtained a beautiful light." LODYGUINE'S BURNER. Lighting by incandescence appears to have faded out of view until 1873, when M. Lodyguine reintroduced the system. He employs carbon in a single piece, diminishing the section at the most luminous point. Two carbons are placed in the same apparatus, and are brought into circuit by a small exterior commutator as consumed. Scarcely more credit can be afforded to this inventor than the resuscitation of the idea. LIGHTING BY INCANDESCENCE. 57 KONN'S BURNER. This lamp (Fig. 29) consists of a base, A, in copper, on which are fixed two terminals, N, two bars, C D, in copper, and a small valve, K, opening only outwards. A glass case, B, is retained on the base by a collar, L, pressing on an india-rubber ring. One of the vertical rods, D, is insulated electrically from the base, and communicates with a terminal also insulated. The other rod, C, is constructed in two parts, consisting of a tube fixed directly upon the base without insulation, and of a copper rod split for a part of its length. This split gives elasticity, and admits of the rod sliding in the tube. Five retort carbons, E, are placed between two small plates. Each carbon is introduced into two small blocks, also of carbon, which receive the copper rods at their extremities. The rods are equal in length at their lower ends, and of unequal length at their upper ends. A hammer, I, is hinged on the bar C, and rests only on a single rod of carbon at once. If this lamp is placed in circuit by attaching the two conductors from a battery to the terminals, the bar of carbon E is traversed by the current which passes, by the aid of the hammer I, from the copper bar F, the two carbon blocks 0, the copper bar G, and the plate above the bar D. Vacuum is previously made by putting the cock K in con- nection with an air-pump. The rod E becomes luminous. Its section diminishes, the rod breaks, and the light dis- appears. The hammer I then falls on another rod, and nearly instantaneously lighting is re-established. When all the carbons are consumed, the hammer rests- upon the copper rod H, and the current is not interrupted. Each lamp gives a light of about 160 candle-power. BOULIGUINE'S BURNER. In this burner (Fig. 30), one of the bars is pierced with a small hole from top to bottom, and has a slot admitting the passage of two small lateral lugs. The carbon is introduced into this bar, and is assisted to rise by a counterweight con- nected by cords to lugs in the transverse support on which the carbon rests. The part of the carbon which is to become 58 THE ELECTRIC LIGHT. FIG. 29. LIGHTING BY INCANDESCENCE. 59 FIG. 30. (10 THE ELECTEIC LIGHT. incandescent is held between the clips of two conical blocks of retort carbon. A screw, placed on the base, admits of in- creasing or diminishing the length of the bar which carries the upper conical block, and consequently of giving to the luminous part greater or less length. When the lamp is placed in circuit, the carbon rod illumi- nates until about to break. Then an electro -magnet opens the clips of the carbon-holders, the counterweight above drives out the fragments that remain between the lips, and the carbon rod rises and penetrates the upper block, re-establishing the current. This lamp has not been successful in practice, probably due to the somewhat delicate adjustment required of the upper carbon clips. FONTAINE'S BURNER. This burner, shown in Fig. 31, consists of an arrangement by which the carbons are set in a groove at each of their extremities in rigid contact, and kept fixed, admitting of the burner being placed in any position. The electric current passes automatically from one carbon to the other, under the action of the electro-magnet, which is included in the circuit. The only reliable elucidation of this system of lighting is also due to M. Fontaine, and the description of the following experiments is taken from his work, the result representing the mean of more than twenty trials : Methods of Coupling the Battery. 2 Series parallel of 24 Elements. 3 Series parallel of 16 Elements. 4 Series parallel of 12 Elements. 1 Single Series of 48 Elements in Tension. Slate of the Circuit. B . 12 elements. It became interesting to learn what light could be obtained with 12 elements by diminishing the length of the carbons. This was the object of a new series of experiments. " Five different com- binations were attempted, by varying in turn the coupling of the battery the diameter of the carbon, and its length. "The best results were obtained with a single lamp furnished with Gaudoin carbons of 0*0016 metre diameter, and of 0*015 metre length, in the incan- descent portion. " The light varied between two and eight burners, but it was more often five burners. Each carbon lasted on average 15 minutes." THE SAWYER -MAN BURNER. This burner, the invention of Messrs. Sawyer and Man, is shown in Fig. 32. The light is produced by the incandescence of the slender pencil of carbon. The light-giving apparatus is separated from the lower part of the lamp by three diaphragms, to shut off downward heat radiation. The copper standards are so shaped as to have great radiating surface, so that the con- duction of heat downward to the mechanism of the base is wholly prevented. The structure of the base enlarged is shown in Fig. 33. The electric current enters from below, follows the metallic conductors to the " burner " as shown by the arrows, thence connecting with the return circuit. The light-producing portion is completely insulated, and sealed at the base, gas- tight. 64 THE ELECTRIC LIGHT. This light has attracted considerable attention in America, and its method of working is thus described by the company formed for its supply : " It is well known that an electric current will exactly and readily divide among circuits of equal resistance. Accordingly, if the resistance of a sub-circuit be maintained constant, no matter what may be going on in it, whether a lamp is not lighted at all, or lighted to a mere taper, or to any interme- diate stage up to full brilliancy, it is obvious that no other lamp or lamps in that circuit will be affected. "The lamp has, let us say, a resistance of 0'95 of an ohm. Therefore, if one lamp is out, there should be a resistance of 0*95 of an ohm in its stead. This is the shunt resistance. The resistance of the circuit is maintained constant at 0*95, no matter what may be the change in the proportion of the current given the lamp. The varying resistances required to give the best effect have been worked out by practical trial. " Thus it is seen that the greater part of the illumination is the product of a small part of the current. When the light is well on, a very slight increase in the current increases the light enormously. It is here that the great loss occasioned by dividing a fixed current among several lamps finds its explana- tion. A current that suffices in one lamp to produce a light, say, of 100 candles will, if divided between two lamps, give in each perhaps no more than ten candles, or even five, making a loss of 90 candles in the sum total. But if the current be doubled, each lamp will give a light of 100 candles, and the sum total will be 200 candles instead of ten. Having brought a candle or a system of candles up to the point of feeble incandescence, a (proportionally) small addition to the current will make them all brilliant. If at 6000 F. a given carbon will produce a light of three candles, at 12,000 it will give nine candles, and at 24,000 it will give 81 candles; the illuminating power increasing with vastly greater rapidity than the temperature. " The wires supplying the current may be run through existing gas-pipes, each lamp being provided with a switch placed conveniently in the wall ; and by simply turning a key LIGHTING BY INCANDESCENCE. 65 the light is turned up or down, off or on. So long as the house is connected with the main, it makes no difference to the producer whether all the lights are on or off, since the resistance of the entire (house) circuit must be overcome; though it will to the consumer, since a meter records the time that each lamp is on, and the charge is rated accordingly. The cost of lamps and switches, it is claimed, will not exceed that of gas fixtures. " The meter is a simple clock arrangement, with an attachment designed to throw the dial hands into connection when a light is on. From each switch a pair of conducting wires are run to opposite studs on a wooden disc. When no current passes through the lamp, the revolving spring turns without making any record. When the current is on, one electric connection at each revolution is made through the pins assigned to the particular lamp, the armature of the magnet is moved, and the recording wheel is advanced one notch. This meter does not measure the quantity of elec- tricity passing, but only the time a lamp is on. If two or any larger number of lamps are on, an equal number of connec- tions are made at each revolution of the wheel, and the record wheel is advanced to correspond. This registration is, of course, a mere matter of business detail. In view of the well- founded popular dislike to gas meters, however, it would seem to be desirable to dispense with such devices entirely ; and the nature of electric distribution appears to favour other and less objectionable modes and means of determining the financial relations of producers and consumers. " Where the main is tapped for a sub-circuit, a shunt is introduced so as to throw so much of the current as may be needed into the derived circuit. The resistance of, say, 100 added lamps will be about 100 ohms. By giving to the shunt a resistance of one ohm, l-100th of the current will be diverted, and the lamps supplied. When a large number of lamps are required in a circuit, a combination of the two plans indicated is employed. " The diversion of any portion of the electric supply into an added circuit, whether one house or a group of houses, F 66 THE ELECTRIC LIGHT. necessarily increases the aggregate resistance of the electric district, and calls for more work from the generator. To meet such contingencies automatically, a regulator has been invented, which responds instantly to any increase or diminu- tion in the demand, thereby securing an absolutely uniform volume of current. " This regulator so controls the steam or other power actuating the generator of electricity, that the amount of power supplied is increased or diminished in exact proportion to the demand, either by changing the volume of steam pro- duced, or by coupling on or detaching different generators or parts of a single generator in circuit. " With regard to the cost of this mode of electric lighting no positive figures can be given. It is claimed to be entirely demonstrated that one horse-power will give by the Sawyer- Man system of incandescence a light of 35 foot gas-burners an hour. Where large powers are employed, the cost of steam-power, every item included, is commonly rated at one cent per horse-power per hour. The cost of 150 feet of gas, at New York rates, is 41 cents, which would make the gas over forty-fold dearer than the Sawyer-Man light." JABLOCHKOFtf'S BURNER. M. Jablochkoff, recognizing the advantage of regularity offered by the system of incandescence, designed a burner (Fig. 34) which consists of a strip of kaolin F'" E! i i > i* fri 1 1 _/ S ft r fr i..(H , Ilia 1 i i 4-J i N i FIG. 59. set in motion, they communicate a rotary motion to the two electro-magnets, which influence them continually, and thus generate currents in the series of bobbins which form the discs. The contact becomes more complete between M M' and N N', in proportion as the currents are stronger ; in other words, as the speed of the machine increases. The following references are self-explanatory : A, large electro-magnet, serving as fly-wheel by the crank C, or by a pulley mouned on its axle. B B', cluster formed of the small electro-magnets, d efg h ij k ; C, crank ; L, support. N N', modified Gramme discs, moving the two electro- magnets, M M', which influence them. 0, pulleys to set the axle in motion. P P f P" P'"> friction-springs, collecting the MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 95 currents which are generated in the discs, t t' t", extremities of the coils of the electro-magnets. This machine yields, for each of its discs, a light equivalent to 600 Car eel burners. RAPIEFF'S MACHINE. Very little is really known about this machine. The inventor has patented a subdivision of the core of the electro-magnets, forming, like Trouve, the core as a bundle of very thin electro- magnets. This arrangement is stated to give greater magnetic power. Further, M. Eapieff patents what he terms "two- sided" inductors, in which the two sides of the inducing and induced magnets are utilized. The latter portion of the in- vention has been partially utilized by Brush. Eegarding this part of the principles of the machine, M. Eapieff says : " The induction of currents being generally produced in dynamo -induction machines by means of setting some arma- tures in motion with respect to some inductors, or inversely, coiled rings, cylinders, or prisms can be applied to such machines, either as both electro-magnets and anchors, or only as electro-magnets or as anchors. " The ring-shaped apparatus in which the currents are induced, or armatures, and the inductors or electro-magnets of the same construction through which the currents are sent, can be combined together in different ways ; but these various combinations may be considered as modifications of the follow- ing kind of arrangement : " Several ring-shaped inductors, A A A (Fig. 60), and several armatures of the same shape and arrangement, B B B, are disposed alternately side by side in planes normal to their common axis, the spaces between them being rendered as small as possible. The armatures B B B are fitted in some manner on a common shaft, wherewith they are caused to rotate, while FIG. 60. 96 THE ELECTRIC LIGHT. the inductors A A A, being secured on a frame or stand, remain fixed or inversely. The number and size of both inductors and armatures depend upon the special purpose they are employed for ; the induction being produced in that case from both sides of each armature, it is termed a two- sided one." This machine is considered by some electricians to promise much, but to this date the results of the experiments have not been published. GRAMME'S MACHINE. The machine invented by M. Gramme is essentially different from all others. Since the date of the first applica- tion, success has been on the increase ; the inventor has been rewarded at Lyons, Vienna, Moscow, Linz, and Philadelphia. More than 400 machines have been made. Electric lighting did not exist industrially before M. Gramme's invention. Principle of the machine. To comprehend the principle of the Gramme machine there is required a more complete analysis of the phenomena than is ordinarily attempted. Given (Fig. 61) a magnetized bar, A B, and a conducting helix in reciprocating move- ment, if the helix is brought towards the bar from its position at X, an induced current is produced at each movement. These currents are in the same direction while the helix passes the middle M of the bar A B, until it leaves the opposite pole B. Thus, in the entire course of the helix on to and from the magnet, two distinct periods are to be distinguished : in the first half of the movement the currents are direct, and in the second they are inverted. If, instead of moving from left to right, as we have supposed, the movement is from right to left, everything occurs as before, with the exception that the currents are opposite. MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 97 Let two magnets, A B and B' A' (Fig. 62), be placed end to end, in contact by poles of the same name, B B'. The whole FIG. 62. FIG. 63. forms a single magnet with a consequent point at the centre. If the helix is moved with relation to this system, it is traversed by a positive current during the first movement, between A and B; by a negative current in the second, from B to B'; again by a negative current in the third, from B' to A' ; and finally by a positive current, when leaving A'. Keplacing the straight magnets by two semicircular magnets (Fig. 63) put end to end, the poles of the same name together, there occur the two poles A A', B B', and the results are the same as in the preceding, M M' being the two neutral points. The essential part of the Gramme machine is a soft-iron ring, furnished with an insulated copper helix, wound on the whole length of the iron. The ex- tremities of this helix are sol- dered together, so as to form a continuous wire without issuing or re-entrant end. If the wire is denuded exteriorly, the part bared forms a straight band running round the whole of the circumference. Friction- pieces, M and M', are applied to the bared part of the helix (Fig. 64). When the ring is placed before the poles S and N of a magnet, the soft iron is magnetized by induction, and there occur in the ring two poles, N' and S', opposed to the poles S and N. If the ring revolves between the poles of a permanent magnet, the induced poles developed in the ring 98 THE ELECTRIC LIGHT. always remain in the same relation with regard to the poles N and S, and are subject to displacement in the iron itself with a velocity equal, and of contrary direction, to that of the Fm. 64. ring. Whatever may be the rapidity of the movement, the poles N' S' remain fixed, and each part of the copper helix successively will pass before them. An element of this helix will be the locale of a current of a certain direction when traversing the path M S M', and of a current of inverse direction to the first when passing through the path M' N M. And, as all the elements of the helix possess the same property, all parts of the helix above the line M M' will be traversed by currents of the same direction, and all parts beneath the line by a current of in- verse direction to the preceding. These two currents are evidently equal and opposite, and balance one another. When two voltaic batteries, com- posed of the same num- ber of elements, are coupled in opposition, it is necessary only to put the extremi- ties of a circuit in communication with the poles common to MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 99 the two batteries, and the currents become associated in quantity. M. Gramme collects the currents developed in the ring of his machine by establishing collectors on the line M M', where the currents in contrary direction encounter each other. In practice Gramme does not denude the wire of the ring. Fig. 65 shows the wire and coils. One or two coils, B, are shown in position, and with the iron ring laid bare and cut. Insulated radial pieces, K, are each attached to the issuing end of a coil, and to the entrant end of the following coil. The currents are collected on the pieces E, as they would be on the denuded wire. Their bent parts, brought parallel to the axle, are carried through and beyond the interior of the ring, and are brought near one another upon a cylinder of small diameter (Fig. 66). The friction-brushes on the pieces K are in a plane perpendicular to the polar line A and B; that is, at the middle or neutral points M and M'. The intensity of the current increases with the velocity of rota- tion ; the electro-motive force is proportional to the velocity. 100 THE ELECTRIC LIGHT. Gramme modifies his machine so as to produce effects of tension or of quantity by winding the ring with fine or coarse wire. It appears indisputable that with equal velocities of the ring the tension will be proportional to the number of convolutions of the wire ; but the internal resistance increases in the same proportion, and in the majority of cases the best results are obtained by employing thick wires. Various machines have been constructed on the Gramme MAGNETO- AND principle for experimental purposes. The first type of this apparatus was horizontal (Fig. 66) ; it gave a current equiva- lent to nearly three ordinary Bunsen elements. This was replaced by a more rational arrangement (Fig. 67), which produced the current of five elements without changing the bobbin. Since the invention by M. Jamin of laminated magnets, nearly all the laboratory machines have been con- structed with magnets on this system. Some are turned FIG. 68. by a wooden handle (Fig. 68) ; others with a pedal (Fig. 69). These machines are now equivalent to eight ordinary Bunsen elements. All know that the inconvenience of mounting several Bunsen elements often deters the undertaking of an otherwise simple experiment. The first light-machine constructed by M. Gramme gave a light of 7000 to 8000 candle-power. Its total weight amounted to 2200 Ibs. It had three movable rings and six bar electro- 102 THE ELECTRIC LIGHT. magnets. One of the rings excited the electro-magnet, the other two produced the working current. The copper wound on the electro-magnets weighed 550 Ibs. ; that of the three FIG. 69. rings, 165 Ibs. The space necessitated was 31J inches length, by 4 feet 1J inches height. This machine, which lighted the clock-tower of the Houses MAGNETO- AND DTNAMO-ELECTEIO MACHINES. 103 of Parliament, became slightly heated, and gave sparks between the metallic brushes and the bundle of conductors on which the current was collected. FIG. 70. x -a^ii i^fiimmiiiiiMm...^ EEF^pE?SrT=^FE T/^ - 7> Fig. 70 is a machine with six bar electro -magnets ; but, 104 THE ELECTRIC LIGHT. instead of being in two right lines, these magnets are grouped in triangles. Two rings admit of conveying the total current into the electro-magnet, or of magnetizing the electro- magnets with one of them, or of producing two separate lights. This machine weighs 1540 Ibs. : its height is 35 J inches ; its width, 2 feet 1J inches. The weight of copper wound on the electro-magnet bar is 396 Ibs. ; that of the two rings, 88 Ibs. It produces a normal light of 4000 candles, raised in experiments at great velocity to nearly double. When a current is sent FIG. 71. into two lamps, each gives 1200 candle power, at 400 revolu- tions per minute. No inconvenient heating is incurred in the bobbin or in the electro-magnets. Figs. 71 and 72 are of a superior machine. It consists of two Hanks of cast iron, arranged vertically and connected by four iron bars, serving as cores to electro-magnets. The axle is of steel ; its bearings are relatively very long. The central ring, instead of being constructed with a single wire attached MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 105 by equal fractions to a common collector, is formed of two bars of the same length, wound parallel on the soft iron, and con- nected to two collectors to receive the currents. The poles of the electro-magnet are of large size, and embrace seven-eighths of the total circumference of the central ring. Four brushes collect the currents produced. The electro-magnet is placed in the circuit. The total length of the machine, pulley in- FIG. 72. ^ \\v its eluded, is 31 J inches ; its width, 1 foot 9 J inches ; and height, 23 inches. Its weight is 880 Ibs. The double coil is connected to 120 conductors, 60 on each side. Its exterior diameter is eight inches. The weight of wire wound on is 31 Ibs. The electro-magnet bars have a diameter of 2J inches, and a length of 15f inches. The total weight of wire wound on the four bars is 211 Ibs. The winding of the wires on the ring is effected as if two complete bobbins were put one beside the other, and these two bobbins may be con- 106 THE ELECTRIC LIGHT. nected in tension or in quantity. Coupled in tension, they give a luminous intensity of 6400 candle-power at 700 revolu- Fio.73. MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 107 tions per minute ; coupled in quantity, they give 16,000 candle- power with 1350 revolutions per minute. Fig. 73 represents the type adopted for workshops and large covered spaces. This machine weighs 396 Ibs. ; its height is 23 J inches, its width is 13f inches, and its length, pulley included, 25J inches. The base weighs 260 Ibs., and is 15f inches in height. The copper wound on the electro -magnet bars weighs 62 Ibs., of which the ring weighs 10 Ibs. With so little copper and only 900 revolutions per minute, 11,520 candle-power light is obtained, the axis of the two carbons being exactly in the same plane. The following table gives the results obtained with a Gramme machine of the workshop type, a Serrin lamp, and Gaudoin carbons. The motive power employed did not exceed 2 h.-p. when the machine was making 820 revolutions, and 3 h.-p. at 900 revolutions. The lamp was distant 200 yards from the machine feeding it, and kept at a height of 15 feet. No. of Revolutions. Distance of Observer from Lamp. Candle-power light. Remarks. Feet 820 135 2464 The current was too 820 67 3600 feeble to maintain 820 30 4120 the carbons in. 820 15 4800 apart. 820 7J 4896 870 135 3200 Distance apart, \ in. 870 67 4400 regularly. Work- 870 30 6480 ing satisfactorily. 870 15 8800 870 7i 9040 920 135 3616 Too high tension. 920 67| 5632 The carbons heat 920 30 9656 for considerable 920 15 11360 length. The light 920 n 11520 unsteady. SIEMENS' DYNAMO-ELECTRIC MACHINE. With this machine (Figs. 74 and 75), the electric current is p .oauceu by the rotation of an insulated conductor of copper 108 THE ELECTEIC LIGHT. wire or armature coiled in several lengths, say 8, 12, 16, up to 28, and in several layers, longitudinally, upon a cylinder with a stationary iron core n HI s s lt so that the whole surface of the armature is covered with longitudinal wires and closed at both ends. This revolving armature is inclosed to the extent of two-thirds of its cylindrical surface by curved soft- iron bars. FIG. 74 \ - \\ W" ' PJ M^ The curved bars are the prolongations of the cores of the electro-magnets E E E E. The coils of the electro-magnet form with the wires of the revolving armature one continuous electric circuit, and when the armature is caused to rotate, an electric current (which at first is very feeble) is induced by the remanent magnetism in the soft-iron bars and directed through MAGNETO- AND DYNAMO-ELECTRIC MACHINES. 109 the collecting brushes into the electro-magnet coils, thus strengthening the magnetism of the iron bars, which again induce a still more powerful current in the revolving arma- ture. The electric current thus is increased on the principle of mutual accumulation. At each revolution the maximum magnetic effect upon each convolution of the armature is produced just after it passes through the middle of both magnetic fields, which are in a vertical plane passing through the axis of the machine. The minimum effect is produced when in a plane at right angles or horizontal. According to the law of Lenz, when a circuit starts from a neutral position on one side of an axis towards the pole of a magnet, it has a direct current induced in it, and the other part of the circuit which approaches the opposite pole of the magnet has an inverse current induced in it ; these two in- duced currents are, however, in the same direction as regards circuit. A similar current will also be induced in all the convolutions of wire in succession as they approach the poles of the magnets. These currents, almost as soon as they are induced, are collected by brushes, B, placed in contact with the commutator in the position which gives the strongest current. The position giving the strongest current gives also the least spark at the commutator. The circumference of the revolving armature is divided into an even number of equal parts, each opposite pair being filled with two coils of wires, the ends of which are brought out and attached to a commutator, as shown in Fig. 74. The Siemens machine is stated to give the following results for the various sizes : Revolutions Illuminating power. Horse- Weieht per minute. Standard candles. power. Ibs. 850 1200 2 280 650 6000 4 420 360 14000 8 1288 Only one Siemens or Serrin lamp can be burnt in the circuit of one of these machines. 110 THE ELEGTEIG LIGHT. EDISON'S MACHINE. The extraordinary interest excited by the anticipation of the results likely to attend an invention proceeding from so able a mechanician as Mr. T. A. Edison, has given special importance to this machine, which is thus described! : "It has long been known that if two electro-magnets, or an/electro- magnet and a permanent magnet, be drawn apaftor caused to pass each other,- electric currents will be set up in the helix of the electro-magnet. It has also been known that vibrating bodies, such as a tuning fork or reed, can be kept in vibration by the exercise of but little power. I avail myself of these two known forces, and combine them in such a manner as to obtain a powerful electric current by the ex- penditure of a small mechanical force." As regards this combination of principles, it would appear that Mr. Edison has been misled by analogy in circumstances. Movement in a magnetic field of any closed circuit is always attended by an expenditure of power equivalent to the work done by the current set up in the closed circuit. The well- balanced coil of a Gramme machine can be turned by a child when its circuit is incomplete, but to cause the coil to revolve under the influence of the intense magnetic field existing when the circuit is closed and the machine is in action, requires the exertion of considerable force. The coil, as connected in work- ing a Gramme machine with its belt communicating with shafting on which is a large and heavy fly-wheel, is quickly brought to rest, when the working electric circuit remains closed, Fig. 76 represents a tuning-fork, A 2 , firmly attached to a stand, B 2 . This fork is preferably of two prongs, but only one might be employed, upon the principle of a musical reed. The vibrating fork may be two yards in length, and heavy in proportion. It has its regular rate of vibration, and the mechanism that keeps it in vibration is to move in harmony. A crank or revolving shaft may be employed, but it is preferred to use a small air, gas, or water engine, applied to each end of the fork. The cylinder A 1 contains a piston and a rod, B 1 , MAGNETO- AND DYNAMO-ELECTRIC MACHINES. Ill connected to the ends of the fork, and steam, gas, water, or other fluid under pressure acts within the cylinder, being FIG. 76. admitted first to one side of the piston and then the other by a suitable valve. The valve and directing rod C 2 are shown for this purpose. The fork A 2 may be a permanent magnet or an electro-magnet, or else it is provided with permanent or electro-magnets. An electro-magnet, C, is shown on each prong of the fork, and opposed to these are the cores of the electro- magnets D. Hence, as the fork is vibrated, a current is set up in the helix of each electro-magnet, D, in one direction as the cores approach each other, and in the opposite direction 112 THE ELECTRIC LIGHT. as they recede. This alternate current is available for electric lights, but if it is desired to convert the current into one of continuity in the same direction, a commutator is employed, operated by the vibrations of the fork to change the circuit connections in each vibration, and thereby make the^ulsations continuous on the line of one polarity. A portion of the current thus generated may pass through the helices of the electro-magnets C, to intensify them to .the maximum power, and the remainder of the current is employed for any desired electrical operation. The commutator springs or levers, C 3 and C 4 , are operated by rods 45. When the prongs of the fork are moving from each other, the contact of the levers C 3 C 4 will be with the screws 40, 41, and the current will be from line 1 through C to C, thence to C 3 and to 41, 43, and to the electro-magnets D D ; from these by 42 to 40, C 4 and line, as shown by the arrows. When the prongs A 2 are vibrating towards each other, the circuit will be through C 3 C C 3 42, in the reverse direction through the circuit and magnets D D to 43, and by C 4 to line. Nothing is known of the power or work equivalent of this machine, and indeed it is difficult to see how, with the dimensions given, the machine is to work at all. For instance, a tuning-fork, with its prongs two yards in length, will vibrate less than once in two seconds, so that considerable force must be expended to overcome the rigidity of the prongs to produce the many hundred vibrations per second actually required. Mr. Edison's reputation as an inventor, however, leads to the hope that the principle, which is so far new in its application, may receive extension at his hands to a more practicable form, and of this it is certainly susceptible. DE MEKITENS' MACHINE. This machine needs no illustration, because, from its simplicity, it is readily to be understood, both in practice and principle. Suppose a wheel, the tire of which is divided into segments, each of these segments being wound with insulated copper wire, forming a separate electro-magnet. Thus the wheel is composed of a series of electro-magnets, the north PRINCIPLES OF MACHINES. 113 pole of one following the south pole of the adjacent magnet. Each electro- magnet, when the wheel is without action, stands with its poles beneath the poles of a horseshoe permanent magnet. These permanent magnets are set in a fixed frame around the periphery of the wheel of electro-magnets. The insulated wires on the cores of the armature of the machine are all wound in the same direction, only the outer end of the wire of one coil is connected with the outer end of the wire of a coil next to it ; whilst the inner end of the wire of the one coil communicates with the inner end of another coil next to it. The alternating currents produced are thus of the same sign throughout the whole ring. The two terminals of the wire on the ring, which constitute the two poles, the signs of which change every moment, since the currents are alternating, communicate respectively with two copper rings fixed on the axis of the machine, and insulated from it. Two thick copper wires are in frictional contact with these rings, and are connected to terminal screws, from which the current is obtained, precisely as in the case of the Alliance machine. This machine has given very high results, which will be found detailed in the chapter on " Lighthouse Illumination." The want of continuity in the iron core of the armature of this machine materially aids in strengthening the currents obtained, since the rapid changes due to reversals of mag- netism are added to the rapid realizations of Lenz' law. REACTIONS OCCURRING IN THE PRECEDING MACHINES. Very inaccurate theories have been pronounced in respect to the recently devised dynamo-electric machines, from the want of actual experiment and from too great reliance upon apparently rational deductions. This has been very clearly pointed out by the Count du Moncel. The following experi- ments, easily repeated, lead to more conclusive ideas than those usually described in the text-books. The direction of a current, due to increment or decrement of the strength of a magnet, is the same, whether the north or south pole is operated upon successively or simultaneously, 114 THE ELECTRIC LIGHT. whatever may be the position of the helix upon the magnet. The currents will be stronger as the excitation occurs nearer the helix. If the helix is placed at the middle of the magnet, or at the neutral line, the current of superexcitation which will result from placing an iron armature upon fether of the poles will be inverse, and will produce a currentNof_(sa.y) 2, whilst the current produced by withdrawing the armature will be direct and of the same strength. By operating simul- taneously upon the two poles, with two armatures, the current will be in the same direction, and of strength ex- pressed by 7. If the coil is placed at one of the poles, say south, the currents of superexcitation and reduction will be of 10 to 12 in value, when the armature is placed upon or removed from the south pole; and these will be only 0'25 in value when the north pole is operated upon, and only 9 when the armature operates upon both poles simultaneously. When the coil is placed half-way between the north pole and the neutral line, an inverse current will be produced when an armature is approached to either of the poles; but it will be only 5 in value when the north pole is operated upon, and 2 when the south pole is approached. With an armature simultaneously applied to both poles, the value of the current will be 9, and this value and effect reversed when the arma- ture is withdrawn. Suppose a few turns of insulated wire to be wound around a powerful bar magnet, the extremities of this wire being connected with a distant galvanometer, and let the coil thus formed be easily movable on the magnet. If this helix is placed at the south pole of the magnet, and if an armature of soft iron is approached to this pole, there results a current, the direction of which will correspond to that of a magnetizing current, and which is caused by the increase of magnetic energy due to the presence of the armature. This current has, say, a positive value of 12, and when the armature is withdrawn, a second negative current of equal value will be obtained. It now remains to examine what occurs when the helix is moved in various directions from the poles towards the neutral line of the magnet, and vice versa. The following results have been observed : PRINCIPLES OF MACHINES. 115 1. When the coil is moved from the south pole to the neutral line, an inverse or magnetization current of 22 in value is obtained. 2. When this movement is reversed, another or direct current is generated, of 25 in value. 3. If, instead of bringing the coil back from the neutral line towards the south pole, the first movement is continued towards the north pole, a second current will be obtained in the reverse direction to that of the current due to the move- ment over the first half of the distance between the poles. If the movement is arrested when the coil has reached midway between the neutral line and the north pole, there will result a direct current of 12. 4. By bringing the coil from this last position towards the neutral line, an indirect deflection of 10 will be obtained. The induced currents produced by the movements of the coil along the magnet behave as if the neutral line represents a resultant of all the magnetic actions in the bar. If this resultant were represented by a line in the direction of which the whole magnetic current passes, the current produced by moving the coil towards this line should, according to the law of Lenz, be inverse ; and this is actually the case, since by moving the coil from either the north or the south pole towards the neutral line, the deflection is in accordance. On the other hand, the currents produced by moving the coil away from this line should, according to the same law, be direct ; and this is actually observed. In accordance with these considerations, a small coil movable around a magnetized ring should be traversed by a direct current when it moves from the neutral line in the direction of the inductor, which polarizes one of the semi- circular magnets constituting this ring ; and this is observed in the Gramme machine. What would result from the passage of the coil in front of the inducing pole itself, say the south pole of the magnet ? Instead of a small coil, take a thin real coil, capable of sliding upon a long rod of iron answering the purpose of a magnetic core. To judge as to the direction of the currents to be ob- 116 THE ELECTRIC LIGHT. served, commence by examining the direction of the current generated when the coil is approached towards the south pole of the inducing magnet, the anterior end of which that is to say, the end which, in the following experiments, is in front will be presented first to the pole in questionXjJnster these conditions an inverse current of 25 is obtained, and, by with- drawing the coil, a direct one of 22. This so far reproduces the well-known experiment of Faraday. If the coil is caused to pass from right to left, and tan- gentially, in front of the south pole of the inductor, taking care to produce this movement in two stages, there is ob- served : That in the first half of the movement, a direct current is developed, 8 in value ; and that, in the second half, another current is produced, of 5 in the same direction. That, by reversing the direction of the motion, the direc- tion of the currents is also reversed. We may, therefore, conclude that the currents resulting from the tangential movement of a coil in front of a magnetic pole are produced under conditions altogether different from those which prevail in the case of currents resulting from the movement in the direction of the axis of the magnet. These two movements, in fact, occur not only in two directions per- pendicular to each other, but also under conditions which differ in relation to the mode in which the induction takes place in the different portions of the coil. In the case of tan- gential motion, induction is exerted only upon one half of the circumference of the turns of wire, and it acts on each side through different ends of the coil. In the other case, the relative positions of the different portions of the helix remain under the same conditions in respect to the inductor pole, and it is only the position of the resultant which varies. What occurs when the coil, moved as above described, is subject to the action of a magnetic core influenced by the inductor ? It is only necessary to slide the bobbin upon the long iron rod whilst this is exposed to the action of the in- ducing pole ; the following effects are observed : In the first place, when the iron rod is approached towards PRINCIPLES OF MACHINES. 117 the inducing pole, though maintained at a distance from it sufficient to allow of the movement of the coil between it and the pole, there is produced in the coil, situated on one side of the pole, an induced current resulting from the magnetization of the bar, which gives an inverse current of 39. When the coil, placed as in the first series of experiments, is set in motion from right to left, it produces, at the moment when it arrives beneath the inducing pole, a direct current of 22; and by continuing the movement beyond the inducing pole, a fresh current, of 30, in the same direction is obtained. The effects produced by the passage of the coil in front of the inductor are thus in the same direction, with or without an iron core, but are much more energetic with it. It may, therefore, be said that the currents generated in consequence of the displacement of the helices on a Gramme ring, relatively to the two resultants corresponding to the two neutral lines, are in the same direction as those produced by the passage in front of the inducing poles of the turns of wire in the helices in each half of the ring. In order to study the effects resulting from polar inter- versions, the experiment may be arranged in the following manner : Take the rod of iron provided with the induction coil pre- viously referred to, and slide a permanent magnet over one of its extremities, perpendicularly to its axis. In this manner the rod undergoes successive interversions of polarity, and it is found, not only that by this action alone a more energetic current is produced than the magnetization and demagnetization currents which result from the action of one pole of the magnet, but, further, that this current is not instantaneous, but appears to augment in energy until the interversion of the poles is complete. The direction of this current varies according to the direction of the movement of the magnetized bar, and if we compare it to that which results from the magnetization or the demagnetization of the magnetic core under the influence of one or the other of the poles of the magnetized bar, it will be found that it is exactly in the same direction as the demagnetization current determined by the 118 THE ELEOTE1Q LIGHT. pole which has first acted. It is consequently in the same direction as the magnetization current of the second pole ; and since, in the movement performed by the magnet, the magnetic core becomes demagnetized in ordef to become mag- netized in the contrary direction, the two currents which result from these two consecutive actions are in the same direction, and consequently form one current occurring throughout the whole movement of the magnet. On the other hand, the movement of the magnet in the opposite direction, having the effect of producing at the commence- ment a demagnetization in the contrary direction to that operated in the first case, the current which results from this retrogade movement must be in the direction contrary to the first. As to the effects produced by the magnet acting upon the movable coils perpendicularly to their axes, the result must be that the different portions of the core of the coils successively constitute a series of magnets with interverted poles, which will occasion those currents in the same direction already observed, and these currents will change in direction according as the coils travel from right to left, or from left to right. Kepeating the experiments described at the commence- ment of this section, with a rod of iron converted into a magnet by the influence of two opposite magnetic poles ap- plied to its two extremities, different effects are to be observed. For this purpose take an electro-magnet, with very long arms, one of which, being deprived of its magnetizing helix, can receive the small travelling coil. When a powerful current is passed through this electro-magnet, whilst an armature is applied to its poles, the naked arm becomes a magnet, of which the poles are excited at one end by the base-plate, and at the other by the armature. Consequently, by moving the coil from one to the other end of this naked arm, it might be expected to obtain the same effects as with the persistent magnet. This is not the case, and the following are the results obtained : At the moment when the electro-magnet is excited, a mag- netization current is produced in the system, and, the coil MULTIPLE-CIRCUIT MACHINES. 119 being placed against the base-plate, this current produces an inverse current of 90. By moving the coil towards the middle of the bar, there is a direct current of 5 ; and by continuing the movement in the direction of the armature, the direct current is again 5. By reversing the movement, an indirect current of 5 is obtained when the coil is moved from the armature to the middle of the rod, and one of 4 by the motion from the middle of the rod to the base-plate of the electro-magnet. It appears to result from these experiments that the iron rod, instead of being polarized in inverse direction at the two ends, behaves as though it had only the polarity of the base- plate ; and as the only difference between the two modes of communicating polarity is merely that, on the one hand, the rod was screwed to the base-plate, whilst on the other it was only in simple contact with the armature, it might be con- cluded that the contact of two magnetic bodies does not establish between them a magnetic conductivity sufficient for such contact to be equivalent to one produced by a strong pressure. The same thing is observed in the case of electrical conductivity when two portions of metal are in contact ; the conductivity is perfect only when a strong pressure is applied. By fastening the armature to the poles of the electro- magnet by means of screws, the effects take place as though the rod constituted a true magnet. Upon those various principles, all the effects produced in the Gramme, Siemens, and Meritens machines may readily be explained. MULTIPLE-CIKCUIT MACHINES. The machines described in the preceding section, although in some instances applicable as multiple-circuit machines, are not especially designed to work more than one exterior electrical circuit. This does not include that the preceding machines are capable of maintaining only one light centre ; on the contrary, some of these single circuit-machines will maintain as many lights upon a single circuit, as the machines to be described will in the total of their multiple circuits. 120 THE ELECTEIC LIGHT. These multiple -circuit machines have, however, been designed with a special purpose, that the w&ole- system of lighting should not depend for its existence upon the continuity of one circuit. This is doubtless a great advantage, but at the same time there is necessitated a greater expenditure of motive power to produce the same light power; the merits of each system, however, need very carefully weighing, and in the present state of electrical lighting the balance of judgment is scarcely sufficiently delicate to determine which is superior. LONTIN'S MACHINES. M. Lontin has made some important improvements in dynamo-electric machines. In 1875 he introduced into England a plan for turning the whole of the electricity produced in the revolving armature of a machine, into the exciting electro-magnets, instead of only a portion. This of course rendered the exciting magnets very powerful in a short time, and the magnetic resistance to the rotation of the coil increases in a few moments to such an extent, that it is almost impossible to overcome it. The circuit was then broken by an automatic commutator, and the special working circuit inserted. One great objection to this form of machine was the heat generated in the coils. In 1876 Lontin intro- duced a machine to overcome this objection. He constructs the armature in the form of a wheel provided with a central boss and spokes of soft iron, mounted on a shaft to which rotary motion can be imparted (Fig. 77). Each soft-iron spoke of the wheel has a coil of wire wound on it, and is, in fact, an electro-magnet, which becomes a source of induced electricity when the wheel is revolved between the poles of a fixed electro-magnet. The residual magnetism of the cores of the electro-magnets is sufficient at first to generate a feeble current in the coils when the wheel is revolved ; and a portion of this current, kept in one direction by a commutator, is diverted in the usual manner into the fixed electro-magnets to intensify them. One or several of these induction wheels may be applied on the same shaft, placing them opposite MULTIPLE-CIRCUIT MACHINES. 121 one or more series of permanent or electro-magnets. When two wheels are fixed on the same shaft, one of them can supply currents exclusively for feeding the electro-magnets, and the currents from the other can be used for external work. If the currents are required to be of only one direction, a commutator or collector is used, and one for each coil or pair of coils is placed on the shaft, to each being attached FIG. 77. the two ends of the wire of the corresponding coil or pair of coils. When merely collectors are used, all the coils on the wheel are connected up in series, so as to form a com- pletely closed circuit, as shown in Fig. 77. All the coils approaching a pole of the electro-magnet are inversely elec- trified to those receding from the same pole. A metal strip is placed opposite the pole of the electro-magnet, to collect by contact the electricity generated in the coil at the instant that its polarity becomes reversed; a similar rubber is also applied opposite the other pole of the electro-magnet. To avoid oxidizing effect under the action of sparks, the commu- tators are enclosed in a bath of non- drying oil. The most valuable of Lontin's improvements is the plan of constructing dynamo-electric machines in such a manner THE ELECTRIC LIGHT. 122 that the inducing electro-magnets^ have a rotary motion, whilst the induced coils are stationary. Figs 78 and 79 re- present this machine. The coils of the induction wheel are in this case the inducers, and are transformed into electro- magnets by the current of a spare magneto-electric machine FIG. 78. FIG. 79. passed through them. On rotation of the wheel, they induce in the surrounding coils a series of currents, which can be utilized without employing any collector or contact-ring. In a machine having 50 induced coils, there would be 50 sources of electricity that could be used either separately or combined. The fixed electro-magnet illustrated in the previous figure may have its cores prolonged, so that more than one coil of insulated wire can be placed upon them. Thus, when the wheel in this machine is turned into an inducer, by reason of the currents already induced in it by the electro-magnets, it will in its turn induce currents in the additional coils, and these currents can be utilized for electric lighting. The machine illustrated in Figs. 78 and 79 is used as a " generator," to supply currents to the "dividing" machine (Figs. 80 and 81). This second or dividing machine consists of a revolving drum, carrying a series of radial magnets. The coils of these radial magnets are connected together, so that one magnet has its positive pole at the outside end, and the succeeding MULTIPLE-CIRCUIT MACHINES. 123 magnet its positive pole at the inside end. The radial magnets are thus made to alternate their poles considered as a cir- FIG. 80. FIG. 81. 124 THE ELECTEIC LIGHT. cumference to the wheej^^-feis arrangement admits of the revolving wheel inducing a niilnber of alternate currents, equal to half the number of spokes. By an exterior com- mutator, which may be connected up in many different ways, these currents can be combined as required. This duplex Lontin system supplies a total illuminating power of 12,000 candles. The generating machine is driven at 250 revolutions, and the distributing or dividing machine at about 400, per minute. With an engine of 8 horse-power nominal, 12 light-circuits can be easily maintained. THE GRAMME " DISTRIBUTOR." This machine (Figs. 82 and 83) consists of a ring of iron wound with coils of insulated copper wire, alternately right FTG. 82. and left handed, the wire being coiled in one direction, so as to cover one-eighth part of the ring, then in the opposite direction for the next eighth part, each of the eight sections of the ring being wound in the reverse direction to the winding of the two adjacent sections. This ring may be regarded MULTIPLE-CIRCUIT MACHINES. 125 as eight curved electro-magnets placed end to end, with their similar poles in contact, so as to form a circle ; it is rigidly fixed in a vertical position to the solid framing of the apparatus, the inducing electro-magnets revolving within it. The electro-magnets, of which there are eight, are fixed radially to a central box revolving upon a horizontal shaft, FIG. 83. upon which is a pulley, driven by a band from a motor. These radial, flat electro-magnets are wound alternately right and left handed, and their alternate ends are consequently of opposite polarity. The cores of these magnets are extended by plates, to increase the area of the magnetic field by which currents are induced in the coils of the ring. In this machine there is no self-contained apparatus for producing the current by which the electro-magnets are 126 THE ELECTRIC LIGHT. magnetized, but a small and separate Gramme machine of the continuous -current type is employed, and is driven by a separate strap. The current from this machine is caused to circulate in the coils of the rotating radial electro-magnets, by which these are magnetized to saturation. Each section of the ring is built up of four sub-sections, abed (Fig. 83), and all these sub-sections of any one section are wound in the same direction. This subdivision admits of the connecting up of the sub-sections into 32, 16, 8, or 4 circuits. All the sub-sections marked a are influenced by the rotating magnets in precisely similar manner, because the influence of a north pole upon a coil wound in a right-handed direction is the same as that of a south pole upon a coil wound in a left-handed direction. Similarly, the currents in all the b coils are of one direction, whatever may be the position of the rotating magnets. Thus, all the coils similarly marked can be connected into one circuit, and terminal screws are provided for the required arrangement. The current from the small machine is led to the rotating magnets through the flat brushes of silvered copper wire attached to the framework of the machine, and in rubbing contact with two insulated copper cylinders, one connected to each end of the magnet circuit. The largest size of this machine supplies 16 Jablochkoff candles, each of 1000 candle-power, at a speed of 600 revolu- tions per minute, absorbing 16 h.-p. The cost of working these machines will be found detailed under the proper section. Both for this and the Lontin multiple-circuit systems, the diagram (Fig. 25, p. 47) will illustrate the method of the connections. ( 127 ) CHAPTEE VI. MECHANICAL EFFICIENCY OF ELECTRIC- LIGHT MACHINES. THERE can be nothing done in the intercomparison of any natural force until accurate measurements have been made. For those measurements the electric-light engineer has mainly to look to the labours of the Committee on dynamo-electric machines formed by the Franklin Institute, and to Professors Houston and Thomson's report* as to the ratio of efficiency in the conversion of motive power into electricity. In entering this comparatively new field of research, pecu- liar difficulties occurred, owing to conditions that do not exist in the various forms of batteries used as sources of electrical power. In many battery circuits a high external resistance may be employed, and the electro-motive force remains com- paratively constant, while in dynamo -electric machines, in which the reaction principle is employed, the introduction of a very high external resistance into the circuit must be neces- sarily attended by decided variations in the electro-motive force, due to changes in the intensity of the magnetic field in which the currents have their origin. Moreover, a consider- able difficulty is experienced in the great variations in the behaviour of these machines when the resistance of the arc, or that of the external work, is changed. Changes, due to loss of conductivity by heating, also take place in the machine itself. * This report has been transcribed in these pages with only very slight alterations. The language and statements of the report hare been so carefully considered, and have so much weight of authority, as to render a paraphrase unadvisable. 128 TEE ELECTRIC LIGHT. These variations are also attended by changes in the power required to drive the machine, and in the speed of running, which again react on the current generated. There are certain normal conditions in the running of dynamo-electric machines designed for light, under which all measurements must be made, viz. : 1. The circuit must be closed, since, on opening, all elec- trical manifestations cease. 2. The circuit must be closed through an external resistance equal to that of the arc of the machine. 3. The arc taken as the standard must be the normal arc of the machine. This condition can only be fulfilled by noticing the behaviour of the machine while running, as to the absence of sparks at the commutator, the heating of the machine, the regularity of action in the consumption of carbons in the lamp, etc. 4. The speed of the machine must be, as nearly as pos- sible, constant. 5. The power required to maintain a given rate of speed must be, as nearly as possible, constant. The machines submitted to the Committee for determina- tions were as follows, viz. : 1. Two machines of different size, and of somewhat different detailed construction, built according to the invention of Mr. C. F. Brush, and styled respectively in the report as A 1 , the larger of the two machines, and A 2 , the smaller. 2. Two machines known as the Wallace -Farmer machines, differing in size, and in minor details of construction, and designated respectively as B 1 , the larger of the two, and B 2 , the smaller. In the case of the machine B 1 , the experiments were discontinued after the measurement of the resistances was made, insufficient power being at disposal to maintain the machine at its proper rate of speed. 3. A Gramme machine of the ordinary construction. All the above machines are constructed so that the whole current traverses the coils of the field magnets, being single- current machines, in which the reaction principle is employed. In the case of the machine designated A 2 , the commutators EFFICIENCY OF MACHINES. 129 are so arranged as to permit the use of two separate circuits when desired. For the purpose of preserving a ready measure of the current produced by each machine, under normal conditions, a shunt was constructed by which an inconsiderable but definite proportion of the current was caused to traverse the coils of a galvanometer, thus giving with each machine a convenient deflection, which could at any time be reproduced. As the interposition of this shunt in the circuit did not appreciably increase its resistance, the normal conditions of running were preserved. As indicating the preservation of normal conditions in any case, the speed of running and the resistances being the same as in any previous run, it was found that when there was an equal expenditure of power, as indicated by the dynamometer, the current produced, as indicated by the galvanometer, was in each case the same. Certain of the machines experimented with heated con- siderably on a prolonged run; most of the tests, therefore, were made when the machines were as nearly as possible at about the temperature of the surrounding air. It is evident that no other standard could be well adopted, as under a prolonged run the temperature of the different parts of the machine would increase very unequally ; and, moreover, it would be impossible to make any reliable measurements of the temperatures of many such parts. In measuring the resistance of the machines, a Wheat - stone's bridge, with a sliding contact, was used in connection with a delicate galvanometer and a suitable voltaic battery. In taking the resistances of the machines, several measure- ments were made with the armatures in different positions, and the mean of these measurements taken as the true resistance. It was, of course, a matter of the greatest importance to obtain a value for the resistance of the arc in any case, since upon the relative values of this resistance, and that of the machine, the efficiency would in any given case, to a great extent, depend. In each case, the arc of which the resistance K 130 THE ELECTRIC LIGHT. was to be taken, was that which was obtained when each machine was giving its average results as to steadiness of light and constancy of the galvanometer deflection. The method adopted for the measurement of the arc was that of substitution, in which a resistance of German silver wire, immersed in water, was substituted for the arc, without altering any of the conditions of running. This substituted resistance was afterwards measured in the usual way, and gave, of course, the resistance of the arc. It could, therefore, when so desired, serve as a substitute for the arc. No other method of obtaining the arc resistance appeared applicable, since the constancy of the resistance of the arc required the passage of the entire current through the carbons. It may be mentioned, as an interesting fact in this connec- tion, that when the current flowing was great, the arc cor- responding thereto had a much lower resistance than when the current was small. This fact is, of course, due to increased vaporization, consequent on increased temperature in the arc. In determining the true arc resistance, the resistance of the electric lamp controlling the arc was measured separately, and deducted from the result obtained with the German silver wire substitute. For ease of obtaining a resistance of German silver wire equal in any case to that of the arc, a simple rheostat was constructed, by winding, upon an open frame, such & length of wire as was judged to be in excess of the resistances of any of the arcs to be measured. By means of a sliding contact, successive lengths of the wire were added until the conditions were reproduced. With this arrangement, no difficulty was experienced in reproducing the same conditions of normal running as when the arc was used. The same conducting wires were used throughout these experiments. Being of heavy copper, their resistance was low, viz. about '016 ohm. To determine the value of the current, two methods were selected, one based on the production of heat in a circuit of known resistance, and the other upon the comparison of a definite proportion of the current with that of a Daniell's battery. EFFICIENCY OF MACHINES. 131 In the application of the first method, eight litres of water, at a known temperature, were taken, and placed in a suitable non-conducting vessel. In this was immersed the German silver wire, and the sliding contact adjusted to afford a resist- ance equal to that of the normal arc of the machine under consideration. This was now introduced into the circuit of the machine. All these arrangements having been made, the temperature of the water was accurately obtained, by a deli- cate thermometer. The current from the machine running under normal conditions was allowed to pass, for a definite time, through the calorimeter so provided. From the data thus obtained, after making the necessary corrections as to the weight of the water employed, the total heating effect in the arc and lamp, as given in Table II. , was deduced. Since the heat in various portions of an electrical circuit is directly proportional to the resistance of those portions, the total heat of the circuit was easily calculated, and is given in Table III., in English heat units. For ease of reference, the constant has been given for conversion of these units into the now commonly accepted units of heat. Having thus obtained the heating effect, the electrical current is X 772 Etc where C = the weber current per ohm, W the weight of water in pounds, h the increase of temperature in degrees Fahr., 772 Joule's constant, R the resistance in ohms, t the time in seconds, and c the constant, '737335, the equivalent in foot- pounds of one weber per ohm per second. The currents so deduced for the different machines are given in Table IV. The other method employed for obtaining the current, viz., the comparison of a definite portion thereof with the current from a Daniel!' s battery, was as follows : A shunt was con- structed, of which one division of the circuit was '12 ohm, and the other 3000 ohms. In this latter division of the circuit was placed a low-resistance galvanometer, on which con- venient deflections were obtained. This shunt being placed in the circuit of the machine, the galvanometer deflections were 132 THE ELECTRIC LIGHT. carefully noted. To the resistance afforded by the shunt, such additional resistance was added, as to make the whole equal to that of the normal arc of the machine. These substituted resistances were immersed in water, in order to maintain an equable temperature. Three Daniell's cells were carefully set up and put in circuit with the same galvanometer, and with a set of standard resistance coils. Kesistances were unplugged sufficient to produce the same deflections as those noted with the shunt above mentioned. The shunt ratio, as nearly as could con- veniently be obtained, was 2^00- Then the formula C = s n x 1-079 where C equals the weber current, s the reciprocal of the shunt ratio, n the number of cells employed, 1*079 the assumed normal value of the electro-motive force of a Daniell's cell, and R the resistances in the circuit with the battery, gives at once the current. In comparison with the total resistances of the circuit, the internal resistance of the battery was so small as to be neglected. The results obtained were as follows : Name of Machine. Shunt ratio. Number of Daniell's cells. Resistances unplugged. Speed of Machine. Large Brush ^ 4 _ 3 2710 ohms. 1340 rev. Small Brush jj n 3700 1400 Wallace. Farmer < " " 8320 6980 4800 844 1040 800 The weber currents, as calculated from the above data, are given in Table IV. From the results thus derived, the electro -motive force was deduced by the general formula E = C x B. The electro-motive force thus calculated will be found in Table IV. EFFICIENCY OF MACHINES. 133 a oo 3 " s l PQ ^ o ^ g B H g a H I I g jo ozig jI ojptreo jod paums -noo aoAvod jo spunod-^ooj potansuoo jo spuuod-^oo^ a^nutm aad FIELD MAGNET 02 spunod ui !>. XO O 1O 00 T-H t> 05 XX XX X> CO CO CO n5 O O W3 So 00 (N KlH iH ? O O 134 THE ELECTRIC LIGHT. Statements are frequently made, when speaking of certain dynamo-electric machines, that they are equal to a given number of Daniell's, or other well-known, battery cells. It is evident, however, that no such comparison can properly be made, since the electro-motive force of a dynamo-electric machine, in which the reaction principle is employed, changes considerably with any change in the relative resistances of the circuit of which it forms a part, while that of any good form of battery, disregarding polarization, remains approximately constant. The internal resistance of dynamo-electric machines is, as a rule, very much lower than that of any ordinary series of battery cells, as generally constructed, and, therefore, to obtain with a battery conditions equivalent to those in a dynamo- electric machine, a sufficient number of cells in series would have to be employed to give the same electro-motive force ; while, at the same time, the size of the cells, or their number in multiple arc, would require to be such that the internal resistance should equal that of the machine. Suppose, for example, that it be desired to replace the large Brush machine by a battery whose electro-motive force and internal and external resistances are all equal to that of the machine, and that we adopt as a standard a Daniell's cell, of an internal resistance of, say, one ohm. Kef erring to Table IV., the electro-motive force of this machine is about 39 volts, to produce which about 87 cells, in series, would be required ; but, by Table II., the internal resistance of this machine is about '49 ohm. To reduce the resistance of our standard cell to this figure, when 37 cells are employed in series, 76 cells, in multiple arc, would be required. Therefore, the total number of cells necessary to replace this machine would equal 37 X 76, or 2812 cells, working over the same external resistance. It must be borne in mind, however, that although the machine is equal to 2812 of the cells taken, that no other arrangement of these cells than that mentioned, viz. 76 in multiple arc and 37 in series, could reproduce the same conditions, and, moreover, the external resistances must be the same. The same principles, applied to other machines, would, when the internal resistance was great, require a large EFFICIENCY OF MACHINES. 135 number of cells, but arranged in such a way as to be ex- tremely wasteful, from by far the greater portion of the work being done in overcoming the resistance of the battery itself. The true comparative measure of the efficiency of dynamo- electric machines as means for converting motive power into work derived from electrical currents, whether as light, heat, or chemical decomposition, is found by comparing the units of work consumed with the equivalent units of work appearing in the circuit external to the machine. In Table V. the comparative data are given. In the first column the dynamo- meter reading gives the total power consumed; from which are to be deducted the figures given in the second column, being the work expended in friction, and in overcoming the resistance of the air ; although, of course, it must be borne in mind that that machine is the most economical in which, other things being equal, the resistance of the air and the friction are the least. The third column gives the total power expended in producing electrical effects, a portion only of which, however, appears in the effective circuit, the re- mainder being variously consumed in the production of local circuits in the different masses of metal composing the machines. This work eventually appears as heat in the machine. Columns four, five, and six give respectively the relative amounts of power variously appearing as heat in the arc, in the entire circuit, and as heat due to local circuits in the conducting masses of metal in the machine, irrespective of the wire. This latter consumption of force may be conveniently described as due to the local action of the machine, and is manifestly comparable to the well-known local action of the voltaic battery, since in each case it not only acts to diminish the effective current produced, but also adds to the cost. No determinations made with an unknown or abnormal external resistance can be of any value, since the proportion of work done, in the several portions of an electrical circuit, depends upon, and varies with, the resistances they offer to its passage. If, therefore, in separate determinations with any particular machine, the resistance of that part of a circuit 136 THE ELECTRIC LIGHT. * * g bO ll 11 a R j 8 1 10 10 rH 1C CO 05 rH rH 00 s s ff . "i . s " a - H^ m w PH PQ pq EFFICIENCY OF MACHINES. 137 of which the work is measured be in one instance large in proportion to the remainder of the circuit, and in another small, the two measurements thus made would give widely different results, since in the case where a large resistance was interposed in this part of the circuit, the percentage of the total work appearing there would be greater than if the small resistance had been used. When an attempt has been made to determine the effi- ciency of a single machine, or of the relative efficiency of a number of machines, by noting the quantity of gas evolved in a voltameter, or by the electrolysis of copper sulphate in a decomposing cell, when the resistance of the voltameter or decomposing cell did not represent the normal working resist- ance, it is manifest that the results cannot properly be taken as a measure of the actual efficiency. In Table II. it will be found that, in general, where the machine used had a high internal resistance, the arc resist- ance normal to it was also high, but they are not necessarily dependent upon each other. The arc resistance depends on the intensity of the current, the nature of the carbons, and on their distance apart. Other conditions being the same, the resistance of the arc is less when the current is great. Since all the machines examined were built for lighting, it will readily be seen that, other things being equal, that machine is the most economical in which the work done in the arc bears a considerable proportion to that done in the whole circuit, and since, with any given current, the work is proportional to the resistance, we have in Table II. the data for comparison in this regard. For example, in the second determination of A 1 , the large Brush machine, the resistance of the arc constitutes considerably more than one-half the total resistance of the entire circuit, while in B 2 , the small Wallace-Farmer machine, it constitutes somewhat more than one-third the total resistance. These relative resistances give, of course, only the proportion of the current generated, which is utilized in the arc as light and heat, the conditions of power consumed to produce the current not being there expressed. 138 ELECTRIC LIGHT. PQ M 3 q H d i 1 H 3 I rM K r- not^ouj Snip npui 'ouLpBoa *> Ji> *? i> mra jad 'Aaa jo paadg puooos spnnod m J oT C 8 H sprraod ui 'dra^t pn^ O.TTJ m qnag; -OIT3Q jo ^ Q m o ^ s g^ So O XO 00 IN CO O CO O5 CO rH Tf O CM rH rH rH O HN . Hfcj HN HN O CO CO CM 1O CM i * I rH O rH CM T^ CO CO CO CO ^f* ^* CO CO CO CO CO CO CO CO GO GO CO GO CO CO i-H rH rH rH i-H rH i-H PQ fH EFFICIENCY OF MACHINES. During any continued run, the heating of the wire of the machine, either directly by the current, or indirectly from conduction from those parts of the machine heated by local action, as explained in a former part of this report, produces an increased resistance, and a consequent falling off in the effective current. Thus, in Table II., at the temperature of 73*5 Fahr., A 1 , the large Brush machine, had a resistance of '485 ohm, while at 88 Fahr., at the armature coils, it was '495 ohm. These differences were still more marked in the case of B 1 . In A 2 , the small Brush machine, it will be noticed that two separate values are given for the resistance of the machine. These correspond to different connections, viz. the resistance, 1'239 ohms, being the connection at the com- mutator for low resistance, the double conducting wires being coupled in multiple arc, while 5*044 ohms represent the resistance when the sections of the double conductor are coupled at the commutator in series. Eef erring to Table III., the numbers given in the column headed " Heat in arc and lamp," are the measure of the total heating power in that portion of the circuit external to the machine. They do not, however, in the case of any machine, represent the energy which is available for the production of light, which depends also on the nature and the amount of the resistance over which it is expended. For example, the heat in arc and lamp are practically the same in each of the Brush machines, if the measurement of the smaller of these machines be taken at the higher speed. The amount of light produced, however, is not the same in these two instances, being considerably greater in the case of the larger machine. The explanation of this apparent anomaly is undoubtedly to be found in the different resistances of the arcs in the two cases. In the large Brush machine the carbons are nearer together than when the small machine is used. This suggests the very plausible explanation, that the cause of the difference is to be attributed to the fact that although the total heating effect is equal in each case, when the large machine is used, the heat produced is evolved in a smaller 140 THE ELECTEIC LIGHT. 3 oo o co H 1 * " s .,.* - tiii |*p CO O 00 00 o ^ co S oci o i>- i> ^ -i> O I-H (M OS r-f iH rH ^ O5 \O O5 s rH tH ill!! O CO l& CO 1O CO ? 9 GO rH CO ^^ o g;2 fe . ^.i3 X CO O5 M CO T 1 05 ^ CO CO CO .> CO rH rH rH 05 CO si & 1 M 1*1 3 cs-s rn r Z 3 "? \0 rH O5 05 9 ^ 1 P M in CD So ? S i T 1O O5 t^ 00 O H 5^2 ^1 ! 115 1 1 & fl H H ^ 00 00 Jt> 05 rH 05 CO a ^ K pq J <1 2 a B a P fe fe gs-s II 8-s. II H s i> oo CO* CO 58 S S S S S co as CM Tf< (M M ^ S ll^l 02 l 144 THE ELECTRIC LIGHT. action is not so great. The columns containing the percent- ages of "Power utilized in the arc," and "Useful effect after deducting friction," need no special comment. The determinations made enabled the following opinions to be formed as to the comparative merits of the machines submitted for examination : The Gramme machine is the most economical, considered as a means for converting motive power into electrical current, giving in the arc a useful result equal to 38 per cent., or to 41 per cent, after deducting friction and the resistance of the air. In this machine the loss of power in friction and local action is the least, the speed being comparatively low. If the resistance of the arc is kept normal, very little heating of the machine results, and there is an almost entire absence of sparks at the commutator. The large Brush machine comes next in order of efficiency, giving in the arc a useful effect equal to 31 per cent, of the total power used, or 37J per cent, after deducting friction. This machine is, indeed, but little inferior in this respect to the Gramme, having, however, the disadvantages of high speed, and a greater proportionate loss of power in friction, etc. This loss is nearly compensated by the advantage this machine possesses over the others of working with a high external, compared with the internal, resistance, this also ensuring comparative absence of heating in the machine. This machine gave the most powerful current, and conse- quently the greatest light. The small Brush machine stands third in efficiency, giving in the arc a useful result equal to 27 per cent., or 31 per cent, after deducting friction. Although somewhat inferior to the Gramme, it is, nevertheless, a machine admirably adapted to the production of intense currents, and has the advantage of being made to furnish currents of widely varying electro- motive force. By suitably connecting the machine, as before described, the electro -motive force may be increased to over 120 volts. It possesses, moreover, the advantage of division of the conductor into two circuits, a feature which, however, is also possessed by some forms of other machines. The EFFICIENCY OF MACHINES. 145 simplicity and ease of repair of the commutator are also advantages. Again, this machine does not heat greatly. The Wallace-Farmer machine does not return to the effec- tive circuit as large a proportion of power as the other machines, although it uses, in electrical work, a large amount of power in a small space. The cause of its small economy is the expenditure of a large proportion of the power in the production of local action. By remedying this defect, a very admirable machine would be produced. After careful con- sideration of all the facts, the Committee unanimously con- cluded that the small Brush machine, though somewhat less economical than the Gramme machine, or the large Brush machine, for the general production of light and of electrical currents, was, of the various machines experimented with, the best adapted for the purposes of the Institute, chiefly for the following reasons: It is adapted to the production of currents of widely varying electro-motive force, and produces a good light. From the mechanical details of its construction, especially at the commutators, it possesses great ease of repair to the parts subject to wear. In order to make the measurements as accurate as possible, it was found necessary so to arrange the photometric apparatus that no reflected or diffused light should fall on the photometer, and thus introduce an element of error. The electric lamp was inclosed in a box, open at the back for convenience of access, but closed with a non-reflecting and opaque screen during the experiments. Projecting from a hole in the front of the box was a wooden tube, six inches square inside and eight feet long, with its inner surface blackened to prevent reflection, thus allowing only a small beam of direct light to leave the box. This beam of light passed into a similar wooden tube, placed at a proper distance from the first, and holding in its farther end the standard candle. This tube also held the dark box of a Bunsen photometer, mounted on a slide, so as to be easily adjusted at the proper dis- tance between the two sources of light. A slit in the side of the tube enabled the observer to see the diaphragm. The outer end of the second tube was also covered with a non- 146 THE ELECTRIC! LIGHT. reflecting hood, and the room was, of course, darkened when photometric measurements were taken. The rigid exclusion of all reflected or diffused light is the only trustworthy method of obtaining true results, and will, no doubt, account in a large measure for the lower candle-power obtained by these experiments than that obtained by many previous experi- menters. The difficulties encountered in the measurement of the light arising from the difference in colour, were at first thought to be considerable, but further practice and expe- rience enabled the observer to overcome them to such an extent that the error arising from this cause is inconsiderable, being greatly less than that due to the fluctuations of the electric arc. In determining the amount of light produced by each machine, it was run continuously for from four to five hours, and observations made at intervals, care being taken to maintain the speed and other conditions normal. One of the most important conditions necessary to ensure correct results was the relative position of the carbon points. Great care was taken that the axes of the two sticks or pencils of carbon were in the same line, so that the light produced should be projected equally in all directions. Were the axes of the carbon pencils not in the same line, a much greater quantity of light would be projected in one direction, and the result of calculation of the light produced, based on the inverse square of the distance from the photometer, would be too great or too small, accordingly as this adjustment was in the one or the other direction. Experiments were made to determine what effect on the amount of light was produced by so 'adjusting the carbons that the front edge of the upper one was in line with the centre of the lower one. Fig. 84 shows such an adjustment. Front ... ... ... ... ... 2218 candles. Side 578 578 Back ... Ill 0485 4 = 871. EFFICIENCY OF MACHINES. 147 The light produced by the same machine, under the same conditions, except the carbons being adjusted in one vertical line (Fig. 85), was 525 candles. This would seem to indicate that nearly 66 per cent, more light was pro- duced by this adjust- ment of the carbons; but a close study of the conditions proves that such is not the case, 'i'\ and that there is no advantage to be derived from such adjustment, except when the light is intended to be used in one direction only. The following is a statement upon this point, in the report of Mr. Jas. N. Douglass, Engineer to the Trinity House : "I have found this arrangement of the carbons (the axis qf the bottom carbon nearly in the same vertical plane as the front of the top carbon), and assuming the intensity of the light with the carbons having their axes in the same vertical line to be represented by 100, the intensity of the light in four directions in azimuth, say E., W., N., and S., will be nearly as follows : East or front intensity ... ... ... 287 to 100 North or side South West or back 115 100 38 100 557 -r- 4 = 139 to 100 ****** "In measuring the* candle-power of the light produced by each machine, I have given the mean intensity obtained in the direction of the photometer, the carbons in lamp working with the Holmes and Alliance machines being always arranged with the axes in the same vertical line, and the carbons in the lamp working the Gramme and Siemens' machine being always arranged with the front edge of the top carbon nearly on the centre of the bottom carbon." 148 THE ELECTRIC LiaHT. It is, therefore, evident that the results given by Mr. Douglass must be divided by 2*87 in making a comparison with those obtained by the Franklin Institute Committee. The following abstract from a report of Professor Tyndall, addressed to the Trinity Board, upon experiments carried out to ascertain the relative values of different apparatus, com- pletes the list as regards other machines than the preceding. The machines experimented on were the following : 1. Holmes' machines, which have been already established for some years at the South Foreland lighthouse. 2. Gramme's machine. 3. Two Gramme's machines coupled together. 4. Siemens' large machine. 5. Siemens' small machine. M. Tresca communicated an interesting paper to the Academy of Sciences, containing account of a series of ex- periments made in the establishment of MM. Sautter and Lemonnier, to ascertain the amount of work performed by the Gramme machine for the production of light. The high speed at which the Gramme machine is driven, rendered it difficult to employ a dynamometer, which should not make more than 250 revolutions per minute. The diagrams obtained were, however, satisfactory after some preliminary trials. The work done has thus been accurately determined, but this was not the case with the luminous in- tensity. This latter was measured direct by a photometer with two discs, one illuminated solely by a Carcel lamp, and the other by the electric lamp. One of these discs appeared of a green hue in relation to the other, which was rose-tinted, and amongst the various methods tried, it was found decidedly the best to correct the difference of these tints by the inter- position of two Carcel lamps, burning 1*48 oz. per hour, and at a suitable distance from the photometer, the electric light being placed at a distance of 131*23 feet in the first, and 65'61 feet in the second trial. In spite of the uniformity of the electric current supplied to the regulator, the light, on account of the irregularity in the nature of the carbons, showed oscillations, which for the most EFFICIENCY OF MACHINES. 149 X 00 in IQ 10 O5 to oo t>i> cq J^ ^ \O lO rH ^ O (M 00 O CO co b >n o 61 co cb t eH O O 00 O O (M r-l rH r-| (N 94 93 CO CO r-l CO CO NrHr-iiHO O O fl N O tH iH 9-1 O O iH iH i 1 O O rH Oi O rH CD is determined by the ratio -^ -p ' ^ , where p is the resistance of the cell or cells added or subtracted. Thus, E, E,B E, Thus, in the case of a telegraph circuit, for instance, we have approximately, I oc E. On the other hand, in the dynamo- electric machine, converting into electrical work a given 1 E 2 horse-power, I a =, since, the ratio -g being constant, E 2 oc E, E a VE, and- x ^ = -i- Thus > any variation i\ 1 v Jtv of E in this case necessarily affects E. Again, any variation of E necessarily affects E; and,, the product E I being constant, we have I oc -g , a somewhat start- ling result, which, to some observers, has appeared con- tradictory to the law of Ohm. With this, however, it is in perfect accord in effect, since E oc V E, E oc E 2 , and "CI Tjl -I g- cc jg r = ^ ; or, when E is varied, the current varies in- versely as the electro-motive force, because the resistance varies as the square of this value. It will be seen that Ex E 2 = -p, and that the same quan- tity of work will be done by the current whatever may be the resistance in circuit. 160 THE ELECTRIC LIGHT. If h.p. be taken to express the total horse-power converted into electrical work (in the whole circuit), under the best con- ditions, with a Gramme machine of the form experimented with at the Franklin Institute, H.P. = h.p. x 1-39, and the efficiency of the machine is expressed by ^- = 72 (nearly). Or the machine can convert into electrical work 72 per cent, of the energy expended upon it. Let E = the electro-motive force, in volts, acting in a circuit ; E = the total resistance, in ohms, of the circuit ; r = resistance of the voltaic arc obtained ; H.P. = h.-p. of the prime motor working the dynamo- electric machine; h.p. = the h.-p. absorbed in the production of electrical work in the circuit ; X = the intensity, in standard candles, of the electric light so arranged as to illuminate equally in all horizontal directions ; A == the intensity of the light in one particular direc- tion; the light being arranged to give the maximum illumination (without reflectors) in this direction. The energy of the current, or the mechanical equivalent of the work and heat produced by it per hour, will be TT H TT 2 v 9fi^4- T? 2 v "hi ft W = ^ x J 200 * ft.-lbs. = ^ * llg foot-tons. R K T i i n j.i j. / energy in ft.-lbs. \ .,-, , .P. absorbed in the current ( 00 A nn - ~ ~ ) will be \33,000 x timemmm-/ The ratio ^ is the measure of the efficiency of dynamo- electric machines. In the case of Gramme's machine, under the best conditions, we have H.P. = h.p. x 1-39. MATHEMATICAL CONSIDERATIONS. 161 The horse-power absorbed in the arc itself is h.p.X^. The ratio of this latter value to h.p., or r _h.p. X r X 747 TT~ E a is the measure of the efficiency of the electrical circuit in the production of the greatest quantity of light with a given quantity of electrical energy. In the experiments with Gramme's machine made by the Committee of the Franklin Institute, the light, in standard sperm candles, produced by the voltaic arc was A. = h.p. x ^ x 1044 (candles) . . . (I.) XV when the intensity of the light was approximately equal in every direction. But, when the carbons are so adjusted as to give the best effect with the photometer in a given position, we may multiply the former value by 2*87, and we have A = h.p. x ~ x 2996 (candles) . . . (II.) hi Expressing these equations in a different form, we have \ = I 2 rxT4 (la.) A x I 2 r x 4- (Ila.) It should be remembered that these values are obtainable only under the most carefully arranged conditions. Although the light cannot be subdivided without very con- siderable loss, it is not to be admitted that, if a given total quantity of light be produced with one hundred lamps, it is one hundred times as expensive as if it were produced in one lamp. If we use two lamps instead of one, and place them in series, the original arc resistance, Z, is not necessarily doubled; indeed, it may be preserved constant, in which case we should C 2 1 have -- for each light, and the original value, C 2 1, for the two. And, if we place four lamps in parallel circuit, the total resistance may be reduced nearly fourfold, so that we may obtain twice the original current with half the E.M.F. in E 2 action. Thus C 2 l, or --l, becomes 162 THE ELECTRIC LIGHT. a) 1 '/ C 1 \ C 2 / the theoretical value for each light being ( - ) 1= r-, and that of the four C 2 Z. The loss, when the light is subdivided, is doubtless due to an increase in the quantity of heat, which must be expended before any luminous effect is produced. The voltaic arc is far more economical, as a producer of light, than any devices for the incandescence of solids. Mr. Louis Schwendler has made a report on the electric light, as to the advisability of adopting it for Indian railway stations. The report is exceedingly valuable, but too long to give complete. The following is a precis : Economy of the electric light. The energy of the standard candle was ascertained by direct experiment. It was found that the standard candle, in order to produce the unit of light, does work at the rate of 610 meg-ergs per second at the least. In fact, it is highly probable that the standard candle, in order to produce the unit of light, works up to more than double that amount (1365 meg-ergs per second). Further, by direct experiment, it was ascertained that the unit of light, as produced in an electric arc by any one of the dynamo-electric machines under trial, and through a leading wire offering not more than 0*1 Siemens' unit resistance, is produced at the rate of not more than 20 meg-ergs per second, including all the work transmitted, the light being measured in a line which passes through the centre of the arc, and stands normal to its axis. Hence the probable engineering margin in favour of the electric light is between 30 to 70, or equal to a mean of 50.* Dynamo-electric machine A produces the unit of light at a rate of not more than 10 meg-ergs per second. Hence it may be safely asserted that the electric light produced by dynamo - * A refers to a Siemens machine, medium size. B small C Gramme machine, workshop pattern. D MATHEMATICAL CONSIDERATIONS. 163 electric machines is, as an average, 50 times cheaper than light by combustion. This is, however, true only as long as the light is produced in one arc. If more than one light is produced in the same circuit by the same current, the external or available light becomes rapidly dearer with in- crease of the number of lights produced. For this reason already, if not for many others, the division of light must result in an engineering failure. It is in the nature of the electric light that it should be used in great intensity in one point, instead of small intensities in many points. Current produced by dynamo -electric machines. These currents, as the insertion of a Bell telephone (used as a shunt) will easily prove, are not steady. The dynamo-electric machine with the greatest number of sections in the induction cylinder gives the steadiest current. Twelve sections are found to be necessary and sufficient. Influence of speed. The current produced by any dynamo - electric machine through a given constant total resistance in circuit increases permanently with the speed of the induction cylinder. This increase of current for low speeds is more than proportional to the speed; afterwards it becomes pro- portional, and for high speeds the increase of current is less than proportional to the speed. The current has, however, no maximum for any speed, but reaches its greatest value at an infinite speed. This same law, as the total resistance in circuit is supposed to be constant, of course holds good also for the electro-motive force of the dynamo-electric machine. Influence of external resistance. Keeping the speed con- stant, the electro -motive force of any dynamo-electric machine decreases rapidly with increase of external resistance. This decrease is more rapid the smaller the internal resistance of the dynamo-electric machine is made. Hence the currents must decrease much more rapidly than proportionally to the total resistance in circuit. As in the case of speed, the electro-motive force has no maximum for a certain external resistance, but approaches permanently its greatest value for an external resistance equal nil. It appears that the function which connects E.M.F. and speed is the same as that which 164 THE ELECTRIC LIGHT. connects E.M.F. and external resistance. We have only to substitute for speed the inverse of resistance, and vice versa. Maximum work by a current in the resistance r. As the current decreases much more rapidly than the total resistance in circuit increases, this resistance r should invariably be made smaller than the remaining resistance of the circuit, i.e. smaller than the internal resistance of dynamo-electric machines plus resistance of leading wires. The electro-motive force of a dynamo-electric machine as a function of the resistance and speed. It appears that the following two formulae are most probably correct for all dynamo-electric machines if the loss of current by trans- mission is taken into account : m <* V + r/ E the E.M.F. , m the internal resistance, and r the external resistance, including resistance of leading wire. K and a are independent of m and r, and are functions of the speed of the induction cylinder, and contain also the construction co- efficients, e is the basis of the natural logarithm. Further E' the E.M.F., and v the speed of the induction cylinder. K and a are independent of v, and are functions of m and r only. These two functions, E and E', correspond to all the characteristics of the curves found by experiment, and they also fulfil the limit conditions. Eesistance and electro -motive force of the electric arc. There appears to be no doubt that an appreciable E.M.F. in the arc is established, which acts in opposite direction to the electro -motive force of the dynamo-electric machine. This E.M.F. of the arc increases with the current passing through the arc. The resistance of the arc for constant length is also a function of the current passing through it, i.e. the MATHEMATICAL CONSIDERATIONS. 165 resistance of the arc decreases with the current (see the following table) : Current in Webers. Resistance of the Arc in S.U. E.M.F. of the Arc in Volts. 28-81 0-91 2-02 23-87 1-72 1-91 16-27 1-97 1-86 The E.M.F. in an electric arc, opposite to the electro- motive force of the dynamo-electric machine, constitutes another reason against the unlimited divisibility of the electric light. ^Regularity of the production of currents by dynamo- electric machines at different periods. If the brushes are well set, and if they are placed as nearly as possible in the neutral line of the commutator, the production of current is perfectly regular, and measurements taken through the same external resistance at the most distant periods agree most perfectly with each other, supposing the correction for variation in speed and internal resistance to be applied. Disregarding the heating of the dynamo-electric machine by the current, the time required to arrive at dynamic equilibrium, i.e. when force transmitted, current, and magnetism received are con- stant, is very short indeed, especially for the strong currents, which alone are made use of for lighting. Formula for controlling the test-results. As the power which is represented by the measured current working through a given resistance can never exceed the original power trans- mitted to the machine, we can, from current, resistance, and force measurements, frame a formula which checks the pro- bability of the results. This formula is : C ^ 0-33 V'W'-< r + m W is the total power consumed by any dynamo-electric machine when producing the observed current C in a circuit 166 THE ELECTEIO LIGHT. of resistance r + m; w' is the power consumed by the dynamo-electric machine when producing no current (i.e. driven empty, circuit open, external resistance infinite) ; r is the external resistance, and m the internal resistance. In the above formula C is in webers, W and w' in meg-ergs* per second, and r and m in Siemens' units. Of late, exaggerated statements of the performance of dynamo-electric machines have been made, the absurdity of which would have become evident at once if the above formula had been applied as a check to the results. Co-efficient of transmission. If all the work (W'w f ) were transformed into available current in the external circuit, then W w' ^r = unity, where W is the total work performed by the observed current in a circuit of known resistance. In practice W w' it will be found, however, that ^ > 1 (for many reasons). W w' This expression, = , is called the co-efficient of trans- mission, and designated by the letter K. K is different for the different dynamo-electric machines which have been tried, and decreases with increase of current. Producing currents above 24 webers, the following average values of K have been obtained : Name of Dynamo - electric Machine. K Average Current in Webers. C 1-01 31-0 A and B 1-12 31-1 D 1-28 27-9 W Co-efficient of efficiency. c = ^, _ ^ w is the useful work done in the circuit by the current. As the resistance of dynamo- electric machines and leading wires cannot be made nil, this * The meg-erg = 100 metre -gramme-second units of work = 10*1926 grammetres. MATHEMATICAL CONSIDERATIONS. 167 co-efficient must be always smaller than unity. For currents above 24 webers we have : Name of Dynamo- electric Machine. 6 Average Current. A 0-62 29-5 B 0-53 31-0 C 0-47 32-6 D 0-30 27-9 Hence the dynamo-electric machine A converts 62 per cent, of the total energy transmitted into useful work, while 38 per cent, is lost in heating the machine. Dynamo-electric machine D converts 30 per cent, of the total energy transmitted into useful work, and loses 70 per cent, in heating its own wires. Practical mechanical equivalent of the currents produced by dynamo -electric machines. r? = W- W where C is the current in webers. Above 24 webers the different dynamo - electric machines produce the weber at the following con- sumption of power : Dynamo-electric machines A and B produce one weber at 686*5 meg-ergs per second; dynamo- electric machine C produces one weber at 736 meg-ergs per second; dynamo-electric machine D produces one weber at 920 meg-ergs per second. Regularity of the electric light. If the resistance external to the dynamo-electric machine is represented by the resist- ance of the arc only, i.e. resistance of leading wires equal nil, then although the light is naturally the most powerful, it is the least steady, since any variation of the resistance of the arc has then evidently the largest influence on the current and on the light. By connecting across the electro-magnet of an electric lamp another electro -magnet which acts as a shunt, and adjusting the two electro-magnets in such a manner that they produce equal extra currents when variations in the primary current take place, the regularity of the working of the lamp is greatly enhanced. An electro-static shunt will have 168 THE ELECTRIC LIGHT. a similar effect. For strong lights or strong currents, the electro-magnetic shunt is best ; for weak lights or weak currents the electro-static shunt is best. The lamp should be con- structed mechanically so well and delicately that the carbon points run together with a minimum tension of the spring of the lamp. When making photometric measurements, to get more trustworthy results, it is best to use a flat carbon (two to three millimetres thick) as the positive electrode, and a carbon of the usual form as the negative electrode. The light is to be observed in a line normal to the flat surface of upper carbon, and passing through the centre of the arc. In this manner the largest quantity of total light produced is measured, and, moreover, the ratio between total and external light is more constant. The lower carbon should be invariably of less section than the upper carbon. Further, when producing the light by a short arc, which it is always advisable to do, the lower carbon should be natural carbon. When the arc is long, the flame by combustion of the carbons is large. This appears to be due to the fact that for a long arc the vacuum formed round the carbon points by expansion of the air by heat is less complete than in an arc of shorter length. The con- sumption of the carbon points is due more to combustion than to disintegration. The incandescent part of the carbon points has so much more intensity of light than the flame that the latter causes a shadow. The hissing noise produced by the electric arc is due to the formation of a vacuum round the in- candescent carbon points. The noise is much stronger in a short than in a long arc. It may also be due in part to the disintegration of the carbon points. The noise of the electric light in a quiet room is simply unbearable. This speaks only against the use of the electric light for domestic purposes. There can be no dpubt that one length of arc is best under given circumstances, considering both the intensity and regu- larity of the light. The light permanently decreases with length of arc, hence the arc should be made as short as possible. This would, however, be bad for the constancy of the light, and may also spoil the dynamo-electric machine. Hence adjust the commutator by turning the brushes in the MATHEMATICAL CONSIDERATIONS. 169 direction of the rotation until only small sparks are observed. If this is impossible, make the arc longer by lessening the tension of the spring. In this manner the best length of arc can be experimentally found. This would give the best tension of the spring at the starting-point. Now let the dynamo- electric machine run for several hours, and make the same experiments, when the best tension of the spring will be found somewhat less. Take the mean of the two tensions and fix the micrometer screw. Proportionality of light and current. Although the light produced in the arc must be very nearly proportional to the total energy consumed in the arc (minus the energy expended in giving the disintegrated carbon particles velocity), the resistance of the arc decreasing with increase of current, it follows that the light cannot be proportional to the square of the current. If we make the highly probable supposition that the resistance of an arc of constant length is inversely pro- portional to the current which passes through, then the light produced would be proportional to the current. This appears to be the case. The conduction of the arc appears to be due to two causes, rarefied air and carbon particles flying in both directions. Both causes would point towards an inverse proportionality between current and resistance of arc. As to the divisibility of the electric light, Mr. Preece has contributed to the Philosophical Magazine a very valuable paper, in which he has simply extended the known laws of heating in the galvanic circuit, with respect to the application to the subject of lighting by electricity. He explains that the theory of the electric light cannot be brought absolutely within the domain of quantitative mathematics, for the reason that we do not yet know the exact relation that exists between the production of heat and the emission of light with a given current ; but we know sufficient to predicate that what is true for the production of heat is equally true for the production of light beyond certain limits. The work done in a battery, or any source of current electricity, is expended outside the battery in a closed circuit in the form of heat. When this heat acquires a certain 170 THE ELECTRIC LIGHT. temperature per unit mass, we have light. If the heat be confined to a mass of metal wire like platinum, we have light by incandescence ; if it be expended in the transference of minute particles of incandescent matter like carbon across an air-space, we have the electric arc. The exact relations between current, heat, temperature, mass, and light have yet to be determined by experiment. The arc is thus a form of energy developed in one point of a circuit, which is the exact equivalent of another form of energy expended in another point of the circuit. Thus, if we produce light by a galvanic battery, it is the equivalent of chemical work done in the battery. If it be produced by a dynamo -machine driven by a steam engine, it is the equiva- lent of coal consumed in the furnace. The object to be attained in any economical utilization of this energy is to convert the greatest possible portion of it into light. Now, the relations that exist between the work done, the current flowing, the resistances present, and the heat developed are easily demonstrated. The work done, W, in any circuit varies directly with the electro-motive force E in that circuit, and with the quantity of electricity, Q, that passes through it, or W = E Q ; but by Ohm's law the electro-motive force is equal to the product of the resistance K of the circuit into the current C flowing, or E = C R ; and by Faraday's . law the quantity of electricity passing depends upon the strength of current C and the time it flows, t } or Q = ct. Therefore, substituting these two values in the above equation, we get W = C 2 IU; in which we have what is known as Joule's law, which gives us the work done, W, or its equivalent, the heat generated, H, in any circuit. By regarding the time as constant, we can put the equation H = C 2 R . . . . (i) MATHEMATICAL CONSIDERATIONS. 171 Now let us take the case of a battery whose electro -motive force is E, and whose internal resistance is p. Let the resistance of the connecting wires be r. Let us also have a particular resistance I, which may be a wire to be heated to incandescence, or a lamp to be lit by the arc; then by Joule's law (1), H = C2( P + r + 7); but by Ohm's law, 0= E + r 2 P+r+l Confining our attention for the present to the heat gene- rated, H, this will be distributed throughout the circuit ; and that in the resistance I will be TTv (V\ x ~ + r + l ( P + r + iy Now, if we suppose n resistances in circuit joined up in series, then the total heat generated will be H'=. ^I'li (3) (p + r + n I? If we differentiate this fraction with respect to n I and put it equal to nothing, we can find when the heat generated in these resistances becomes a maximum ; that is, fj TT' 1 r + nO] = 0, whence p + r + n I = 2 n I ; that is, p + r = n l ; or the greatest heat is generated in the resistances when the value of the latter equals the resistances of the rest of the circuit. Let us now assume the n resistances to be connected up in multiple arc; then the joint resistance will become' n and the heat generated will be v l - n 172 TEE ELECTRIC LIGHT. and the maximum amount of heat will occur, as before, when <-<4 Now, in the first case, if the internal resistance of the battery and of the connecting wires be very small compared with n I, we may neglect them ; so that by putting p + r = 0, equation (3) becomes _E 2 m nV or the total amount of heat generated in the resistances will vary inversely as the number of the latter in circuit. In the second case, we cannot neglect p + r; for here the greater we make n, the smaller - becomes with respect to P + r ; so that if eventually - becomes very small, we may IV neglect it in the denominator of the fraction. Then Fi J H" = - " n = - 5!L_ < 5) + r) 2 n (p + r) 2 ' so that in this case also the total heat generated in the resist- ances will vary inversely as the number of the latter in circuit. Now, it must be observed that in each of these cases the total heat is distributed over n resistances ; and, therefore, as compared with one resistance, the heat generated in each is only -g of that generated in one. So that, joined up either in 91 series or in multiple arc, the heat generated in each of a number of resistances varies inversely as the square of their number. With respect to the light emitted, if the amount of heat generated represented exactly the amount of light emitted, then the above equations would indicate the effects produced by multiplying the lights or subdividing the current when a constant battery is employed. But -this is not so. The light obtained is not proportional to the heat generated. Below a certain limit the production of heat is not accompanied by light at all. In the case of incandescence, if the heat be MATHEMATICAL CONSIDERATIONS. 173 distributed over two wires instead of one, inasmuch as the mass to be heated in the one case is double that in the other, the actual temperature to which each of the wires will be heated will be only one quarter of that obtained with one wire, and the total light emitted will be half what it was before. In the case of the arc a similar result probably takes place ; the incandescent matter, which is heated by the current and which gives out the light, is increased by the addition of each lamp, and therefore diminishes the actual temperature of each arc, and consequently diminishes the light given out in direct proportion to the number of lights. Moreover, in the arc the actual disintegration of the carbons, and the transference of matter across the air-space, represent an amount of work done which must be deducted from that converted into heat, and which again tends to dimmish the amount of light emitted. If, therefore, the lamps be joined up in series or in multiple arc, the light emitted by each lamp will vary inversely in a greater ratio than the square of the number in circuit. We have assumed E to be constant ; but if the current be produced by a magneto or dynamo-machine worked by a steam engine consuming a given amount of coal per unit time, E is no longer constant, for it varies with the resistances in the circuit. The constant in this case is the work done in the steam engine in unit time. Calling this Wi, the total heat generated in the circuit when the lamps are joined up in series will be : = (6) nl p and since the light varies inversely as n, the light emitted T \V n ^ f"\ 1 X n (p + r+nl and when joined up in multiple arc, T ffT H fQ\ 174 THE ELECTRIC LIGHT. Or by putting p + r = in equation (7), and - = in the denominator of equation (8), we get Ll = T 1 and So that beyond certain limits, when the current is produced by a dynamo-machine, if n lamps be joined up in series, the total light becomes diminished by - , and the light emitted by ?i each lamp becomes diminished by _^. ii If they are joined up in multiple arc, the total light is diminished by -5, and the light emitted by each lamp !_ i n n* n 3 the latter case the rapid diminution in the light emitted is due to the fact that the heat is developed in the machine itself, instead of in the resistances external to it. We have assumed Wi to be constant ; but this is only the case when a certain limit is reached, and when the velocity of the rotating coils in the dynamo-machine has attained a maximum. This limit will vary with each dynamo-machine and each kind of lamp used. With the Wallace -Farmer machine the limit appears to be reached when six lamps are connected up in series ; with the Gramme alternating machine and Jablochkoff candles, the limit appears to be five lamps. Beyond these limits the above laws will be true. It is this partial success in multiplying the light that has led so many sanguine experimenters to anticipate the ultimate possibility of its extensive subdivision a possibility which this demon- stration shows to be hopeless, and which experiment has proved to be fallacious.* In the reduction of results, the following equational numbers will be found of use : 1 standard candle = J cubic foot of gas per hour. * Vide p. 62. MATHEMATICAL CONSIDERATIONS. 175 1 cubic foot of coal gas = 690 heat units (Ibs. water heated 1 Fahrenheit). The Franklin Institute experiments gave 380 candles = 1 h.-p. force expended. Thirty-eight per cent, of total power in the circuit appeared in the arc. 1 h.-p. = 1,980,000 foot-pounds per hour. 1 heat unit = 772 foot-pounds = 0*252 calorie. Therefore, 1 h.-p. = 2565 units of heat per hour, and s^yy 5 = 6| units of heat per candle of light. 1 Ib. gas coal produces 4 cubic feet of gas, 0*85 Ib. gas coke, and 0*05 Ib. gas tar. In the pound of gas coal there are 15,000 units of heat ; in the coke 13,000, and in the gas tar 20,000, units of heat. The power expended by a dynamo -machine producing the light of one sperm candle is about equivalent to 90 Ibs. falling through one foot in one minute. 1 foot-pound = 0*1380 kilogramrnetre. 1 calorie (kilogramme of water heated 1 Centrigrade) = 424 kilogrammetres = 3*9683 heat units (Fahrenheit). 1 kilogrammetre = 7*2331 foot-pounds. 176 THE ELECTRIC LIGHT. CHAPTEE VIII. ELECTEIC KEGULATORS. THE term ''regulator," derived from the French term regu- lateur, has been very generally applied to the apparatus we have described as electric lamps. The term arose from the regulating mechanism employed in these lamps to main- tain a due distance between the carbon points. We have reserved the term ''regulator" to be applied to the apparatus devised for the maintenance of a constant strength of current in any electrical circuit. In all applications of elec- tricity to lighting purposes, it is of the greatest importance that the strength of current in any of the circuits should be maintained uniform, so that the adjacent circuits in con- nection from the same source may not receive too much nor too little current. Jacobi appears to have devised the first electric regulator. It consisted of a magnetized needle, pivoted, and surrounded by a coil of wire, so arranged in connection with a trough of liquid that increase in the deflection of the needle produced by the current, introduced a greater length of liquid into the circuit. This regulator was impracticable, on account of the variations introduced by polarizations of the electrodes dipping into the liquid, and is evidently inapplicable to such enormous currents as occur in the circuit of a dynamo-electric machine. Staite and Edwards, those really wonderful electricians, who were so much in advance of their time in their electrical inventions, have patented several forms of electric regulators, in which the height of a column of mercury is regulated by the greater or less suction exerted by a magnetic bobbin upon REGULATORS. 177 an iron float or piston, the mercury being included f in the circuit between one end of the bobbin and a long platinum electrode dipping into the other end of the mercurial column. In another form, the expansion of a platinum wire caused the movement of a lever having something like hyperbolic curvature. This lever worked upon a helix of insulated wire, bared in a line parallel to the axis of the helix, where the lever touched; so that the greater the expansion of the platinum wire, the more the lever was raised and the larger the number of turns of wire introduced into the circuit, and vice versa. This idea, as well as that of the lamp by the same inventors, introduced to-day, would be even now considered as applications of the highest order in physical science. Con- sidering that these inventions are more than twenty years old, certain objectors to the fact of progress in the application of electrical laws have at least some ground for what they advance. Lontin appears to have been the first to make practical application of the expansion of metals under the heat generated in them by the passage of the current. He passed the current through the double metallic band of a Breguet or metal thermometer, one end of which was fixed, and the other attached to a helix of wire, the convolutions of which it short-circuited when the current was weakened. This regu- lator was sluggish in its action, because the short-circuiting took place by the movement or pressure of the coils of the helix upon each other laterally to the axis of the wire, and directly with respect to the axis of the helix. It has been pointed out that the essential principle of all good governors, whether of steam engines or electric currents, is that the slightest change in the thing to be controlled pro- duces considerable variation in the supply of energy. The idea of employing the stretching of a wire by the heating of the current passing through it was also suggested by the early contrivers of electric lamps as a simple means of regulating the distance between the carbons, and it has the advantage that, as the heat produced per second is propor- tional to the square of the current flowing through it, a very N 178 THE ELECTRIC LIGHT. slight change in the current will produce a considerable varia- tion in the quantity of heat produced per second, and there- fore in the temperature, if the mass of the substance to be heated is very small. On the other hand, when the current diminishes in strength, it is necessary that the temperature of the heated substance should immediately fall; this requires considerable radiating surface. These conditions Dr. Siemens has attempted to fulfil in the following instrument (Fig. 86). FIG. 86. A B C is a vertical band of metal not more that two-thousandths of an inch thick, passing over the roller B, one end of the band being fixed to the roller A, and the other, C, to the short end of a lever, C D E F, turning on a pivot, D. When by turn- ing the pinion P the thin metallic band is tightened, the upper end, F, of the lever C D E F is pressed against the movable metallic terminals, T, of the resistance coils, the result is these are pushed out of the vertical and pressed together, and all the coils are short-circuited. If now a current, entering at A and leaving at N, passes through the vertical metallic strip ABC, the lever C D E F, and the terminals T T T, it heats the strip, which consequently expands and diminishes the pressure of F on T ; some of the terminals, therefore, separate from one another, and some of the resistance coils are introduced into circuit. In the figure six are shown short- REGULATORS. 179 circuited and three in circuit. Resistance will be thus auto- matically introduced. To provide against accidental changes in the radiating power of the strip, produced by currents of air, the portion A B C is placed under a glass cover. In some trials made with this instrument before the Royal Society, it was shown that the interposition of a certain resistance into the circuit only altered the deflection of a tangent galvano- meter from 40 to 39*5 when the regulator was employed, whereas without it the in- FIG. 87. > sertion of the same extra- fopi )E neous resistance diminished the deflection from 60 to 40. Fig. 87 is another form of governor proposed by Dr. Siemens. The wire A B is stretched until the lever C D E, turning on the pivot D, produces sufficient pressure on a pile of Edison carbon-discs in the glass tube G. This pressure becomes less, and the resistance of the carbons becomes greater, the more the wire stretches by heating with the current passing through it and through the carbon discs ; the stronger, there- fore, the current, the greater is the resistance opposed to it, and equilibrium is maintained. The principle of this apparatus has been described by the inventor in a paper read before the Royal Society, January 30, 1879 :- "It is well known that when an electric current passes through a conductor, heat is generated, which, according to Joule, is proportionate in amount to the resistance of the con- ductor, and to the square of the current which passes through it in a unit of time, or H = C 2 R. The most essential part of the instrument is a strip of copper, iron, or other metal, rolled extremely thin, through which the current to be regulated has to pass as described. Suppose that the current intended to be passed through the instrument is capable of maintaining the sensitive strip at a temperature of say 60 C., and that a sudden increase of current takes place in consequence either 180 THE ELECTRIC LIGHT. of an augmentation of the supply of electricity or of a change in the extraneous resistance to be overcome. The result will be an augmentation of temperature, which will continue until a new equilibrium between the heat supplied and that lost by radiation is effected. If the strip is made of metal of high conductivity, such as copper or silver, and is rolled down to a thickness not exceeding 0*06 millimetres, its capacity for heat is exceedingly small, and, its surface being relatively very great, the new equilibrium between the supply of heat and its loss by radiation is effected almost instantaneously. But, with the increase of temperature, the position of the regulating lever is simultaneously affected, causing one or more contacts to be liberated, and as many additional resistance coils to be thrown into circuit ; the result being that the temperature of the strip varies only between very narrow limits, and that the current itself is rendered very uniform, notwithstanding considerable variation in its force, or in the resistance of the lamp, or other extraneous resistance which it is intended to regulate. " It might appear at first sight that, in dealing with power- ful currents, the breaking of contacts would cause serious inconvenience, in consequence of the discharge of extra current between the points of contact. But no such discharges of any importance actually take place, because the metallic continuity of the circuit is never broken, and each contact serves only to diminish to some extent the resistance of the regulating rheostat. The resistance coils, by which adjoining contact- springs are connected, may be readily changed, so as to suit particular cases ; they are made, by preference, of naked wire, in order to expose the entire surface to the cooling action of the atmosphere. " The apparatus first described may be adapted also for the measurement of powerful electric currents. The variable rheostat is in this case dispensed with, and the lever carries at its end a pencil, pressing with its point upon a strip of paper drawn under it, in a parallel direction with the lever, by means of clockwork. A second fixed pencil draws a second or datum line upon the strip, so adjusted that the lines drawn by the two pencils coincide when no current is passing through the REGULATORS. 181 sensitive strip. The passage of a current through the strip immediately causes the pencil attached to the lever to move away from the datum line, and the distance between the two lines represents the temperature of the strip. This tempera- ture depends, in the first place, upon the amount of current passing through the strip, and in the second place, upon the loss of heat by radiation from the strip, which two quantities balance one another during any interval that the current remains constant. "If C is the current before increase of temperature has taken place, E the resistance of the conductor at the external temperature T, H the heat generated per unit of time at the commencement of the flow, E 7 the resistance and H' the heat when the temperature T' and the current C have been attained ; then, by the law of Joule, H' = E 7 C 72 . But, inasmuch as the radiation during the interval of constant current and tempera- ture is equal to the supply of heat during the same interval, we have, by the law of Dulong and Petit, H 7 = (T'-T) S, in which S is the radiating surface. Then E 7 C' 2 = (T' - T) S, C /2 = (T' - T) S . But T'-T represents the expansion of the strip or movement of the pencil m, and considering that the electrical resistance of the conductor varies as its absolute temperature (which, upon the Centigrade scale, is 274 below the zero Centigrade) according to a law first expressed by Helmholtz, and that we are only here dealing with a few degrees' difference of temperature, no sensible error will be committed in putting the value of E for E', and we have the S condition of equilibrium C 2 = m ^. .-.C'= V m| .... (1) or, in words, the current varies as the square root of the difference of temperature or ordinates. For any other con- Q dition of temperature T 77 we have C //2 = g (T"-T). .*. C" = ^r(T" - T), 182 THE ELECTRIC LIGHT. and (C' /2 - C' 2 ) - (T w - T - T' x T) = (T" - T') ; but for small differences of C" and C' we may put (C" 2 - C' 2 ) = 2 C" (C" C') ; that is to say, small variations of current will be proportional to the variation in the temperature of the strip. " In order to facilitate the process of determining the value of a diagram in webers or other units of current, it is only necessary, if the variations are not excessive, to average the ordinates, and to determine their value by equation (1), or from a table prepared for that purpose. The error committed in taking the average ordinate instead of the absolute ordinates FIG. 88. TO LAMP when the current varies between small limits, is evidently small, the variation of the ordinates above their mean value averaging the variations below the same. " The thin sensitive conductor may thus be utilized either to restrict the amount of electricity flowing through a branch circuit, within certain narrow limits, or to produce a record of the amount of current passed through a circuit in any given time." An example of another form of regulator is shown in Fig. 88. This apparatus, issued with the Siemens machine, is intended to introduce into the circuit a resistance equal to REGULATORS. 183 that of the lamp, should the latter be extinguished. It con- sists of a resistance coil, W, equal to the arc in value, immersed in water in a small tank formed in the base plate. An electro- magnet, E, when the current is passing, attracts an armature, a ; when no current is circulating in the coils of the electro- magnet, a spring, /, draws the armature against a contact- piece, c. Under the normal conditions of the action of the lamp, the magnet draws the armature away from the contact - piece ; when the current ceases, the spring returns the arma- ture to the stop completing circuit through the coil of resist- ance, W. The resistance of the circuit is by this means maintained sufficiently constant to prevent racing of the motor, and influence upon other lights in circuit. The apparatus does not appear to be much used in practice. A somewhat similar device has been described as existing in the base of the Eapieff lamp. 184 THE ELECTRIC LIGHT. CHAPTEE IX. COMMERCIAL ASPECT OF ELECTRIC LIGHTING, THE success of the electric light from a commercial point of view infers that not only must the system of lighting be practicable, but it must as well be practicable with economy. It must be produced cheaper than existing lighting, or its greater expense must be balanced by other advantages. In estimating the true commercial position of the electric system of lighting, the estimator is in a position of great difficulty. He is called upon, in some instances, to form an opinion based upon public exhibitions of one peculiar system, because with other systems the trials have not taken place upon a similarly extensive scale. This system is acknowledged to be the most costly mode of lighting by electricity, yet has been most adopted, partly, we may suppose, from want of energy of the supporters of other systems, and a certain regard for first cost. Numerous reports have been published as to the advantages and disadvantages of electric lighting. Dis- missing those most strongly biased against describing the electric light as " artificial," and predicting its speedy extinction and accepting the details given by manufacturers with every reserve, the conclusion most likely to occur to the investigator who is merely seeking a superior system of illumination is, that the conditions under which the electric light has been produced are so variable as to render an ipse dixit impossible. Its commercial practicability is, in fact, entirely dependent upon circumstances. It is required to light a large space, with no partitions or subdivisions ; then, under any circumstances, electric COMMERCIAL ASPECT. 185 lighting is cheaper than gas lighting, and the cost will vary from one-half to one-twentieth of the latter system. For divided spaces of certain dimensions, as stores, shops, and manufactories of a general character, the economy of lighting by electricity will depend upon the power of the light required. Where a 20-candle burner is sufficient in a single apartment, gas lighting is the cheaper system ; but where 100 candle- power is required, advantages are as much more in favour of electric lighting. Special manufactures and trades in which an equivalent to daylight is necessary to distinguish colours, as in silk-mills, drapers' shops, dyers' establishments, colour works, determine other conditions, and the electric light has then no competitor. It is a substitute for daylight, and enables the manufacturer to double his hours of work, or to work continuously, thereby doubling his production with the same plant. Upon this special basis, no other system of lighting can compete with electric lighting. For theatrical and other displays, there is again a wide and special field. For street lighting, it may be considered that the question has still to be solved. Only a special system has been adopted either in Paris or in London, and this system has, in comparison with other systems, given results that may be regarded as unfavourable to electric lighting. It is, however, much to be deplored that so numerous corporate bodies have not taken the advice of independent electricians, and placed the various systems proposed for public lighting under more direct competitive trial. Every one knows that the monetary interests connected with gas lighting are very great and highly important, and that a very large portion of the investing public, who have taken shares in companies supplying gas, would be more inconvenienced by depreciation of their property than they would gain by the immediate introduction of a superior system of lighting. Hence there is considerably more direct sympathy with existing systems of lighting than with the introduction of novel systems, that might depreciate the value of present property, and not themselves, consequently, offer so profitable 186 THE ELECTRIC LIGHT. an investment. Besides this, public opinion is vastly con- servative, as well as prone to exaggerate probabilities. Not very many years ago, the introduction of railways was thought likely to end in the extermination of the horse. Kailways and cornfields could not co-exist. The sellers of oil protested against the introduction of gas, as strongly as the sellers of gas protest against the electric light. Gas was unhealthy, dangerous ; it would undermine and blow up London ; it was the idea of a madman, and could never be forced through a mile of tubes. It is needless to say that more horses exist now than in the palmiest days of coaching; our rail tracks pass through the most magnificent fields of corn ; more oil is sold than ever was produced ; that the mileage of piping conveying gas through London is counted by the thousands, and that, chimerical and dangerous as gas lighting ever appeared, the system has met with universal adoption. It is quoted against electric lighting that, although it has been known for more than 30 years, it has not been adopted generally. The manufacture of gas was known in the labora- tory for a much longer period prior to its introduction, and lighting by natural gas actually existed in several foreign countries more than 200 years before its artificial adoption. That electric lighting has not progressed will be proved untrue by comparing the following cost in 1857 in France, then the centre of chemical science, with subsequent state- ments. The lamp, one of Lacassagne and Thiers, was fed by 60 Bunsen cells, and was worked for 100 hours: Substance. Consumption in 100 hours. Cost per hour. Actual Cost. Zinc ... 72 -00 kilos. O75 francs. 80 francs per 100 kilos. Sulphuric acid ... 154-00 0-37 12 Nitric 247-00 l'V3 56 Mercury 9-50 0-50 650 Purified carbon . . . 6-61 metres. 0-20 2-5 Total 3-55 M. Becquerel had previously determined the cost to be three francs per hour. COMMEECIAL ASPECT. 187 Le Eoux has prepared the following estimate of the cost of working an Alliance machine, producing 1000 candle-power light, for five hours daily : With special motor : Francs. Interest at 10 per cent, upon 12,000 francs, first cost 3'35 Coal, 50 kilos, at 40 francs ... ... ... 2'00 Stoker ... ... ... ... ... 5'00 Carbons ... ... ... ... ... 1-80 Oil and sundries ... ... ... ... 0'70 Cost per day ... ... ... ... 12'85 The motor used for other purposes : Francs. Interest on 9000 francs at 10 per cent. ... ... 2*50 Coal, 20 kilos. ... ... ... ... 0.80 Carbons ... ... ... ... ... 1-80 Oil and sundries ... ... ... ... 0'40 Cost per day ... ... ... ... 5'50 With 500 hours' lighting a year, the cost is 4*30 francs and 3'0 francs per hour respectively. M. Fontaine deduces the cost of lighting with the Gramme machine to the end of the year 1877, and shows that to this date the Gramme machine yields a light 75 times less expensive than that from wax candles. 55 stearine candles. 16 colza oil. 11 gas at 0'30 franc per cubic metre. 6 gas at 0-15 franc Under the most favourable conditions, this light is 300 times less expensive than that from wax candles. 220 stearine candles. 63 colza oil. 40 gas at 0'30 franc per cubic metre. 22 gas at 0-15 franc M. Fontaine also gives the' following tabulated comparison of cost of various lights : 188 THE ELECTRIC LIGHT. Quantity Cost per Cost for burnt burner * 4000 Observations. per Lour. per hour. burners per hour. grammes. francs. francs. Purified colza oil ... 42 0-07 28-00 Price per kilog. 1'70 francs. Neutral Allaire oil 39 0-06 24-00 1-55 Shale oil 36 0-0468 18-72 1-30 Petroleum oil 30 0-054 21-60 , 1-80 Tallow candle 83 0-141 56-40 1-70 Wax candle 66 0-33 132-00 5-0 Stearine candle .. 82 0-246 98-40 3-0 Voltaic battery .. 0-06 24-00 "Alliance" machine )) 0-024 9-00 For 500 hours per year. " Alliance " machine 5> 0-007 2-80 For 4000 hours per year. Oil gas litres. 140 0-029 11-60 JAt 0*15 franc per cubic metre : \ 500 hours per yard. Oil gas 0-025 10-00 JAt 0*15 franc per cubic metre : \ 4000 hours per yard. Oil gas 0-050 20-00 JAt 0-30 franc per cubic metre : \ 500 hours per year. Oil gas 0-046 17-80 JAt 0'30 franc per cubic metre : \ 4000 hours per year. Gramme machine \ (new type) / 0-0042 1-78 (With steam motor : 500 hours I per year. Gramme machine "1 (new type / 00016 0-56 (With steam motor: 4000 hours I per year. Gramme machine } (new type) / 0-004 1-60 JWith hydraulic power : 500 \ hours per year. Gramme machine j (new type) / 0-0011 0-44 (With hydraulic power: 4000 \ hours per year. Considerable reduction has, however, been made in the cost of electric lighting since these comparatively experimental introductions, and the subsequently quoted costs are those of the present day (1879). The Northern of France Eailway Company introduced the electric light at the goods station of La Chapelle, Paris. The electric sources were six single -light Gramme machines, driven by one 15 nominal h.-p. engine. The work had a nightly duration of an average of ten hours. The first cost of 1640 comprised six lamps, six machines, portable engine with its shed or house, and conductors. In the goods shed, each light illuminated well a radius of 100 feet, and in the yard a radius of 200 feet. The results showed that about 2J h.-p. per hour was required for each light, and that for each lamp * A burner may be accepted as equal to 8-9 normal candle-power in each case. COMMERCIAL ASPECT. 189 three-eighths of an inch square carbons were consumed at the rate of four inches per hour, at a cost of l^d., including waste. Four lights, on an average, for ten hours per night, were found to cost Motive power, 2^ h.-p., 7 -8 Ibs. coals, ten hours, each s. d. lamp ... .. ... ... ... 7 Carbons, ten hours at l^d, each lamp ... ... 4 10 Oil, 8d. ; wood, l\d. ; lighting engine-house, 2%d. ... 1 Engineer ... ... ... ... ... 4 16 10 or 5d. per hour; to which must be added 10 per cent, interest and depreciation on first cost, bringing the total to 7f d. per hour. This represented an actual light of 52 gas-jets, at a cost of 22 gas-jets. In 1876 Messrs. Powell, of Kouen, introduced the electric light into the engineers' shop, a building 125 feet long, 45 feet broad, 40 feet high at the ridge, and 26 feet high at the eaves. Two Gramme single -circuit machines, with Serrin lamps and conductors, cost <196. The motive power was taken from the working engine, and 2J indicated h.-p. was required by each light, having a horizontal intensity of 1900 candles. The actual cost of this illumination is 4d. per hour per light, and, assuming a special engine, the total cost would be 2s. Wd. per hour. The cost of gas in Kouen is 7s. Sd. per 1000 cubic feet, ar>d an equivalent light by gas would cost 7s. Id. per hour. The light has constantly been in use, and is in course of extension throughout the works. "For some months the electric light has been on trial in a large engineering works in the North. The department in which the electric lights are used is 170 feet long by 45 feet broad. Two Gramme machines were used, each machine being of 6000 candle-power, with two Serrin lamps. Two Gramme machines, 6000 candle-power s. d. each, at 100 each ... ... ... 200 Two Serrin lamps, at 18 each ... ... 36 100 metres of cable, at 2s. 9c. per metre ... 13 15 40 metres of carbons, at 2s. Gd. ... 500 Net cost ... 254 15 190 THE ELECTRIC LIGHT. " A two-cylinder engine was fitted up expressly to drive the two machines ; the diameter of the cylinders, 5J inches by 10 inches stroke ; mean pressure, at least 30 Ibs. ; revolutions 200 per minute. At this piston speed it was estimated that fully 8 indicated horse-power could be obtained. On setting up the electric machines, it was found they could not be got up to the speed necessary to produce a full light. It was seen that there was not power enough; besides, by the two machines being driven off the one engine, the light obtained at each lamp was very fluctuating, this arising in a measure from either lamp carbons sometimes splitting off ; and as a conse- quence the engine ran faster, driving the other machine at an increased rate and increasing its brilliancy. On ignition taking place again the engine speed fell. Alternations such as these, together with a limited brilliancy of light, led to the determination to drive one machine with the engine solely, and the other off one of the works engines. This was done, and now the lights in each lamp are working steadily and satisfactorily. From the experience acquired, it is concluded that it requires, at least, 1 horse-power for every 1000 candle- power; also, that each machine should be driven indepen- dently, the whole of the necessary driving gear being and costing as follows : s. d. Engine for driving one 6000 candle-power Gramme machine, cylinders 5J inches by 10 inches stroke ... ... ... ... 40 Piping 1 10 Countershaft, pulleys, and belts ... ... 6 15 Labour fitting up wires, lamps, and machines ... 2 10 50 15 Two machines ... ... ... ... 101 10 Cost of machines, lamps, cables, and carbons ... 254 15 Total net cost ... ... ... 356 5 " In order to arrive at the comparative cost of the electric and gas lights, the data are as follows, taking the cost of the electric light per diem of four hours two in the afternoon COMMERCIAL ASPECT. 191 and two in the morning this being the actual time it is in use : . d. In four hours both lamps burn half a metre of carbons, which at 2s. Qd. per metre is ... ... ... 13 Coals consumed by engines in four hours, at 3 Ibs. per indicated horse-power, 2 cwt. ... ... ... 10 Oil consumed by engines and shaft in four hours ... 02 2. 5 Or equal to l\d. per hour. The actual gaslights dispensed with are 22 on columns and 8 on machines, giving a total of 30 burners. Each burner uses, in the four hours, 13 cubic feet of gas, or 390 feet in all, which costs llf d., or nearly 3d. per hour. The quantity of gas consumed in the department is as- certained from total daily amount used as per gas meter, and is pretty accurate in the rate used per burner during four hours." This report, taken from an engineering paper, is illustra- tive of one of the worst applications of electric lighting, as so far calculated, because no account is taken of the increased light obtained, nor could this light have been properly dis- tributed. In Messrs. Sautter and Lemonnier's workshops 'in Paris, the electric light has been continuously employed, and the illustration on the next page will give some idea of its effect, there being no other lamps or candles used, as in the case just quoted. In a trial of the Farmer-Wallace light at the Southampton Docks, the places selected were the large export shed used by the Peninsular and Oriental steamers, about 340 feet long and 180 feet wide, and the corner of the quay from which the Havre and Channel Islands boats start. The Anglo-American Light Company, of Hatton Garden, supplied the machine and four lamps, the machine being worked by an 8 h.-p. engine. This engine had 9| inch cylinders and 12 inch stroke ; the boiler, a total heating surface of 165 square feet, and grate area of 5*9 square feet. The speed maintained was 125 revolutions per minute, increased by pulleys and belting to 650 revolutions on the machine. Three lamps were placed in the shed, and the fourth at a point on the quay distant 420 feet from the COMMERCIAL ASPECT. 193 machine. The whole area of the shed, 27,000 square feet, was brilliantly illuminated in spite of an asphalte floor and blackened roof. The light at the corner of the quay was mounted on an ordinary lamp-post, and had above it a four- foot reflector. Although so close to the ground, the lamp brilliantly illumined a radius of 250 to 300 yards. The cost of the trial has not been given. The Lontin system has had an extended trial on the Paris terminus of the Paris, Lyons, and Mediterranean Eailway Company. Twenty-eight electric lights here replace 172 gas- jets, and are supplied by one generating and two distributing machines upon the Lontin system, the distributing machines being known as those of 18,000 candle-power. The average cost per hour of the 28 lamps is about Ss. Twelve lamps have, on the same station, been found to cost 5d. per hour per light, the light from each lamp averaging about 1800 candles. At the St. Lazare station of the Western of France Eail- way, at Paris, six lights are maintained from a 12,000 candle- power machine on the Lontin system, the motor being an agricultural engine with 9J inch cylinder and 13J inch stroke, with a working pressure of 80 Ibs. per square inch. The lights are of 480 candles each, and are not covered with any shade. This arrangement has not the least inconvenient effect upon the eye. Serrin's lamps, as improved by Lontin, are em- ployed, there being two in each of the three circuits. The cost of the carbons per hour is 1-J-rf. per light, and the work- ing cost of the six lights : s. d. Coal, at 32s. per ton ... ... ... ... 1 2 Carbons ... ... ... ... ... 8 Attendance ... ... ... ... ... 10 Per hour ... ... ... ... 2 8 or 5^d. per light. Interest and depreciation on first cost will bring the amount to Sd. per light per hour. Amongst numerous other municipal authorities, the Town Council of Liverpool requested their engineer, Mr. George 194 THE ELECTRIC LIGHT. Deacon, to report on the subject. Mr. Deacon visited Paris and investigated the comparative cost. "Regulators," reports Mr. Deacon, " supplied with electri- city by dynamo -electric machines, driven by steam or other motive power, have been extensively used within the last ten years for lighthouses, for naval and military signalling, for tidal and other engineering works, and in certain industrial establishments, and upon the efficiency of such combinations most of the statements laid before the public up to a recent date, as to the relative cost of electric lighting and gas light- ing, have been based. Perfectly true though those statements have been, the inferences to which they have led have been most misleading. "It is true, for example, that the cost of light from ordi- nary regulators having nominal illuminating powers of from 6000 to 15,000 candles is, according to circumstances, only from one-fifth to one-half the cost of gas producing the same candle-power ; but in the case of the electric light a greatly increased candle-power is indispensable to produce the same degree of illumination, as the following considerations will show. " The light from each regulator cannot be efficiently reduced below 1000 candles, and gives a much higher efficiency when increased to 10,000 or 15,000 candles. Regulators giving 6000 candles each are very commonly used. The most favour- able place for the employment of such a regulator would be the centre of a circular space. If such a space were 500 feet in diameter, the light in the centre line of a circular carriage-way just within it would be greater than that in a Liverpool street illuminated in the ordinary way. To illuminate the whole area with gas rather better than the electric light would illu- minate the centre of the supposed carriage-way, would neces- sitate the use. of only about 125 Liverpool street lamps of 16 candle-power each. Thus, with gaslight having an aggre- gate power of only 2000 candles, a better effect is pro- duced than with the electric light having a power of 6000 candles, even in a case particularly well adapted to give the best result attainable from a concentrated light. It is COMMERCIAL ASPECT. 195 obvious, therefore, that any comparison of cost based merely upon the relative candle-power of electric and gas lights must be misleading, and yet upon this principle many of the pub- lished comparative statements have been based. " From the above considerations it is evident that if the illuminating power of each electric light could be greatly reduced without increasing the relative cost, the total expense of illumination by electricity would be much less than that of illumination by gas. It is natural, therefore, that the subject should have attracted the attention of many inventors, though, owing to certain dynamical obstacles, the eminent men who have made electricity a subject for mathematical research have not generally regarded the economical subdivision of the electric light as a promising matter for experimental investi- gation." Mr. Deacon's observations on the cost of the Jablochkoff candle are important, and, as we shall presently see, have been confirmed by other investigators. He found by photo- metric experiments that each naked Jablochkoff candle gave a light, in the horizontal plane passing through the voltaic arc, of 453 standard English candles, on the average, " when the electric candle is placed with its side to the photometer ; when placed with its edge to the photometer the light was somewhat less. In this average the occasional very large diminutions of intensity, lasting for short periods, are not included." . . . " But it must not be forgotten that it is thought necessary to inclose the candles in opal globes, and, though there can be no doubt that this greatly adds to the beauty of the light, it seriously detracts from its available power." Mr. Deacon finds by photometric observations that only 42 per cent, of the light from an electric candle having an illuminating power of 453 standard candles is transmitted in a horizontal direction through the globe, thus reducing the available light from each lamp to 172 candles. Very various statements have been made from time to time with regard to the candle-power of the Jablochkoff system, and they may be tabulated as follows : 106 THE ELECTRIC LIGHT. Standard Illuminating power of candles. Naked Jablochkoff caudle, as given by the pro- moters ... ... ... ... ... 930 Stated by Mr. Deacon as given by the Societe Generale d'^lectricite ... ... ... 465 Determined by Mr. Deacon's photometric observa- tions in horizontal plane ... ... ... 453 M. Allard's measurements in horizontal plane, con- ducted on behalf of the Municipal Council of Paris ... ... ... ... ... 279 Side of Jablochkoff candle enclosed in opal globe, as determined by Mr. Deacon in horizontal plane ... 172 Determined by M. Allard in horizontal plane ... 167*4 Determined by M. Allard for rays reaching the pave- ment ... ... ... ... ... 112-53 Continuing Mr. Deacon's observations, it is necessary to notice that, though Liverpool is supplied with 20-candle gas and London with only 16-candle gas, the Liverpool street lamps only burn four cubic feet per hour, while the London lamps burn five, so that the candle-power of each public lamp is 16, both in Liverpool and London. "In considering what number of the 16-candle gas- burners employed for public lighting in Liverpool would give the same amount of light to a thoroughfare like the Avenue de r Opera not including the two adjoining Places it is necessary to calculate the amount of light along the centre line of the carriage-way, and so to arrange the gas lights that at no points shall the light be less than is produced along that line by the Jablochkoff candles. Having regard to the fact that the intensity of light varies inversely as the square of the distance from the object illuminated, the comparison is readily made, and assuming the too favourable view that the electric light along the carriage-way is equivalent to the concentrated light of 172 candles at each lamp, it is found that to illuminate it equally well would involve the use of 165 Liverpool street gas-jets, in substitution for the 32 Jablochkoff electric lights, each jet having a power of 16 candles, reduced to 14*08 candles by the glass of the lantern.* The gas lamps would be placed two feet from the * The Liverpool street lamps are regulated to give an illuminating power COMMERCIAL ASPECT. 197 curbs and 27 feet 6 inches apart. With this arrangement the footways would be more brilliantly illuminated than with the Jablochkoff candles, and the darkest parts of the carriage- way would be equally, if not better, illuminated. The total candle-power of the electric lights is taken at 172 x 32 = 5504 standard candles, while the total power of the gas would be only 14'08 x 165 = 2323 candles, though its effect in lighting the avenue would be as great, if not greater, showing the benefit of subdivision of the light when it can be effected without increased cost. " Including interest on the cost of the lamps, price paid for the gas and for lighting, repairs, and painting, each public lamp costs the Liverpool Corporation about 0*243cZ. per hour when in use. The comparison for equal illumination of the centre line of the carriage-way in each case would there- fore be 32 Jablochkoff candles, at 7'68d. per hour, 1 Os. Qd. ; 165 Liverpool gas-jets, at 0'243d. per hour, 3s. 4d>. "But such an illumination as either of these methods provides is immensely greater than has ever before been thought desirable for permanent use. To illuminate the Avenue de 1'Opera to the same extent and at the same cost per unit of area as the best lighted thoroughfares in Liverpool would involve the use of only 47 gas-jets, at a cost, as before, of 0*243tZ. each per hour. "The total cost of lighting, according to the highest Liverpool standard, would therefore be ll'42d. per hour. In terms of illuminating power, therefore, the electric lighting in question costs at least six times as much as illumination with Liverpool gas, while in terms of the area lighted, without reference to the amount of light, it costs twenty-one times as much." * Mr. Deacon determines the relative useful effect of lights of different intensities and systems by a very simple, yet of 16 standard candles, with a consumption of four cubic feet of gas per hour, at a pressure of about T 5 o inch of water, the cost of the gas being 3s. 6d. per 1000 cubic feet. * The cost of lighting the Avenue de V Opera above referred to is nearly the same as the estimate published by the company for the illumination of other places, if interest on capital expended and rent of premises be taken into consideration. 198 THE ELECTRIC LIGHT. accurate, method. He selects a few points furthest removed from the surrounding lamps, and adds together the candle- power of each light divided by the square of its distance from the point in question. Let c be the true candle-power of each light, and ^ cl 2 o o o *8ss 888 o co i> O5 - CO CO "^? :O O O O O ,00000 OS ^ to * rH te DO 'SOO 000 CO Tf 1 to , r- r- 71 e os co o o o j j -^fl O5 SCO (N CO O rH ^COXr**tO ^jOSOCO I I S 6 S (M o oo rH XO iH . 5^ ~ "S'd "fl i'S|l ifll p^ a 111 fit lls Annual Wages and al Coal, coke, ca Carriage and liepairs and Interest on fir MARITIME AND MILITARY APPLICATIONS. 223 candles, the cost per unit would prove still more in its favour, no further addition to the working staff being necessary. Statement showing the consumption of coke at the Lizard lighthouses, with the quantity of light produced per Ib. of coke consumed, and per h.-p. absorbed by the electric machines, inclusive of expenditure for reserve of engine-power for acci- dents, doubling the intensity of the light when required, bank- ing fires, etc. : Annual consumption of coke in tons ... ... 95 Lbs. of coke consumed per hour of light exhibited (4412 hours) ... ... ... ... ... 48-2 Intensity of light produced at focus of optical apparatus, in candles : Mean for 365 nights (4412 hours) . . . 8751 Quantity of light in candles produced per Ib. of coke consumed ... ... ... ... ... 182*0 Quantity of light in candles per h.-p. absorbed by electric machines ... ... ... ... ... 1097 The advantages of electric lighting on board ship are not so well understood as they should be. The chief object is to increase safety, by avoiding collisions and facilitating the entrance to ports. It also admits of loading and unloading being effected by night, as well as in daylight. The apparatus usually comprises a beacon, a generator of electricity, a portable lamp, and various accessories. The beacon on board the s.s. Ameriquc, where the light was first introduced, is placed on the top of a small iron-plate turret ascended by steps in the interior, without the necessity of passing along the deck, as the turret is immediately over a hatchway. This arrangement is of great advantage in bad weather, when the fore part of the ship is accessible with difficulty by the deck. The turret was originally 21 feet high. It was reduced to six feet to give it greater stability. Its diameter is three feet. It is fixed on the fore part of the vessel, 45 feet from the bow. The Gramme magneto-electric machine has a power of 1800 candles, and weighs 44 Ibs. It is driven by a three- cylinder Brotherhood engine. The average speed is 850 revolutions a minute, both for the machine and the engine. In the Amcrique, the light is automatically intermitted. 224 THE ELECTRIC LIGHT. This internaittence is effected by a small and very simple mechanism fixed on the shaft of the machine. The current goes alternately through the carbon of the lamp, and through a closed metallic circuit, which becomes alternately heated and cooled. The applications to our own navy, as well as to the navies of foreign countries, generally follow this plan, or do not sufficiently depart from it to need special description. As a torpedo defence, the electric light is likely to prove of the greatest value, because by its aid the approach of a torpedo- boat can be easily detected. To direct and concentrate the light, Messrs. Sautter and Lemonnier have constructed a lenticular projector (Fig. 89), comprising a Fresnel lens, com- posed of three dioptric and six catadioptic lenses. The lamp and the lenses are carried by a cast-iron drum, movable around its vertical axis, and turning on a horizontal axis. The turning and oscillatory movements may be successive or simultaneous ; they direct the luminous beam in all directions and at any inclination, and can be effected by the operator, who has position behind the projector. A small camera lucida, placed on one of the bearings of the cylinder, projects the image of the carbons upon a ground- glass screen, and allows of observation of the working of the lamp without the necessity of opening the cylinder. By means of a screw, the position of the lamp can be altered, when it is required to shift the luminous point beyond or from the focus, to produce greater or less divergence of the beam. A second screw and clamp serves to maintain the beam in a given direction the screw stopping the turning movement; the cramp preventing oscillatory movement. For artillery purposes, a special arrangement permits, by means of tangent screws, of slowly displacing the luminous beam, and of exactly striking a previously given direction. The complete apparatus is placed on a cast-iron socket, which can be affixed to the bridge on board ship, to the interior of a casemate in a fort, or on a movable carriage. By the aid of an interrupter, the current can be suppressed at will without stopping the machine. MARITIME AND MILITARY APPLICATIONS. 225 FIG. 89. 226 TEE ELECTRIC LIGHT. For military operations, a portable machine and engine have been combined and mounted on a trolly. The engine (Fig. 90) is on the three-cylinder system, as designed by Brotherhood. FTG. 90. ^J^!|Wv^ The electro-magnets of the machine are flat and very large ; the coil has two current collectors. A commutator mounted on the armatures admits of the coupling of the machine in tension or in quantity instantaneously. MARITIME AND MILITARY APPLICATIONS. 227 From trials at Mont Valerien with a Gramme machine thus arranged, and with a special projector, an observer at the side of the apparatus could see objects 18,000 yards distant, and clearly distinguish details of construction at 15,000 yards. These trials, which were made in tolerably clear weather, with a transparent atmosphere, have been repeated on dark nights with every success. A great advantage that the Gramme machine possesses for military operations is the power, by simple manipulation of a commutator, of instantly giving twofold more powerful light, or reciprocally. This result is obtained by coupling the machine in tension or in quantity. When the weather is clear, the machine should be coupled in tension; then the expenditure of steam is small, and the carbon rods are slowly consumed. When the weather is foggy or very obscure, the machine is arranged in quantity, the expenditure of steam is increased, and the carbon rods are consumed more quickly. 228 THE ELECT PIG LIOHT. CHAPTER XII. VARIOUS APPLICATIONS OF THE ELECTKIC LIGHT. THE electric light has now been so extensively introduced, that it would be tedious to enumerate the various employers ; a few illustrations will suffice to show some of the accessory ways and means of its adoption. FIG. 91. life Fig. 91 illustrates the use of the electric light in illumi- nating a dockyard in course of construction, in the case of the Spanish Northern Railway, near the Guadurama Mountains, VARIOUS APPLICATIONS. 229 230 THE ELECTRIC LIGHT. in 1862. Serrin lamps and batteries were employed, at a cost of 2s. 6d. per hour per lamp, there being 10 lamps in use for a total period of 9417 hours. This cost was 60 per cent, cheaper than that of torchlight. In the galleries of the mines and pits in connection with these works, the electric light was found of the greatest service, as it did not vitiate nor heat the atmosphere. This application affords an example of the in- troduction of the light under its most expensive condition, and where power for driving a machine is not available. A method of mounting the magneto-electric machine upon the same base as the motor is shown in Fig. 92, as applied to a Gramme machine. The dimensions are in metric measure. It is of considerable convenience to be able to renew the carbons in a lamp without the use of ladders or steps, and for this purpose M. Menier has devised the following suspension for the lamps employed in his extensive factories (Fig. 93). The roller, A, consists of two cast-iron cheeks, mounted on a wooden base, with a drum of vulcanite. The conductors are attached one to each cheek, and these cheeks to the con- ductors within the suspending cable. The cable, B, has an external covering of hemp, upon a coating of india-rubber, which encloses a series of copper, wires braided like a wick, this again enclosing a second sheathing of india-rubber, which insulates the core of a strand of copper . wires. A ratchet prevents the descent of the lamp. This cable, passing over a couple of pulleys, is attached to a small plate, connected to the lamp by two curved bars, C. The current is con- veyed to the terminals of the lamp by the curved bars, one of which is connected to the metallic core of the cable, and the other with the braided wick of wires, which in turn are con- nected with the conductors leading to the terminals of the magneto-electric machine. A peculiar system is adopted in lighting the Louvre at Paris, where a very diffused light is required. The space G (Fig. 94) is above the centre of the room. A lamp, A, is counterpoised by the weight C, and is suspended above a sheet of frosted plate-glass, E, the object of which is to prevent the production of objectionable shadows. Four surfaces, F, are VARIOUS APPLICATIONS. 231 FIG. 93. H 8 nh M ifrt " 232 THE ELECTRIC LIGHT. ranged as the sides of a pyramid, lined with tin plate, reflect- ing the light downwards. Two rods, H H, support the frosted FIG. 94. glass. A second lamp, B, counterpoised at C, and resting upon a bracket, J, is kept in readiness for substitution for the lamp A, when its carbons are consumed. The lighting is thus made nearly continuous. In some cases it is advantageous to project the rays of light from the lamp upon a whitened ceiling by a parabolic reflector, thus diffusing the rays and preventing the casting of shadows. It may, roughly, serve for the purposes of estimate, to remark that a single apparatus will light about 500 square yards of fitters' shops, modelling-rooms, etc.; 250 square yards of spinning-mills, printing-rooms, and the like ; or 2000 square yards of open-air work. The lamps should always be, for these spaces, more than 15 feet from the ground. ( 233 ) CHAPTEK XIII. ELECTKIC CARBONS. DAVY, who made the first experiments on the voltaic arc, used rods of wood carbon quenched in water or mercury. These rods burnt with great brilliancy, very regularly, but too rapidly. Foucault replaced the wood carbon by the deposits collected from the walls of gas retorts. Ketort carbon is far from uniform; it sometimes splits, frequently works irregularly, and produces considerable variations in brilliancy. These varia- tions chiefly depend upon the presence of foreign matters, alkaline or earthy, and notably upon silica. These matters are much less refractory than the carbon, but they vaporize, and form part of the flame which envelops the arc. This flame is a better conductor than the voltaic arc, and has a much greater section; it consequently becomes less heated, and its power of radiation is less than that of the particles of carbon which constitute the arc. Several inventors have endeavoured to substitute purer agglomerates ; others have merely purified retort carbon. Among the processes proposed for the improvement of electric carbons are those of Staite and Edwards, Le Molt, Lacassagne and Thiers, Curmer, Jacquelain, Peyret, Archereau, Carre, Gaudoin, and Sawyer-Mann. STAITE AND EDWARDS' CARBON. In 1846 Staite and Edwards patented a process for the manufacture of carbons for the electric light, from a mixture of pulverized coke and sugar. The coke is first reduced to powder, and a small quantity of syrup added, the mixture 234 THE ELECTRIC LIGHT. pugged, moulded, and strongly compressed. The carbon is then subjected to moderate heat, plunged into a concentrated solution of sugar, and subjected to a white heat. LE HOLT'S CARBON. In 1849 Le Molt patented, for electric carbons, a mixture consisting of two parts of retort carbon, two parts of wood charcoal or of coke, atfd one part of tar. The substances were pulverized, and brought to a stiff paste, then subjected to great pressure. The moulded pieces, covered with a coating of syrup, were placed beside each other in a vessel of retort carbon, and subjected to a high temperature for 20 to 30 hours. LACASSAGNE AND TRIERS' CARBON. Lacassagne and Thiers fuse with the retort carbon a certain quantity of caustic potash or soda. With this bath at a red heat, they digest in it for a quarter of an hour retort- carbon rods. This operation changes into a soluble silicate of potash or soda the silica contained in the carbons. The carbon rods are then washed in boiling water, and subjected, at red heat, for several hours to the action of chlorine, to convert the different earthy matters into volatile chlorides. These carbons give a regular light. CURMER'S CARBON. Curmer calcines lampblack, benzine, and oil of turpentine moulded in the form of cylinders, leaving a porous carbon, which is soaked with resins or saccharine matters, and again calcined. JACQUELAIN'S CARBON. Jacquelain endeavoured to imitate the manufacture of retort carbon. With tars resulting from true distillation, consequently free from all non-volatile impurities, and effecting in special apparatus the conditions of decomposition, retort carbons ought to be reproduced possessing perfect purity. Jacquelain has done this with a tube of refractory earth, in an improvised furnace, and has obtained plates which, cut into rods, give a light steadier, whiter, and of about 25 per CARBONS. 235 cent, greater intensity, than ordinary carbons. These carbons require a considerable amount of manual labour, because the material is so hard that it can with difficulty be cut by the saw, and they produce considerable waste. PEYRET'S CARBON. Peyret prepares carbons by soaking pieces of elder-tree pith, or any other porous body, in liquefied sugar, and after- wards decomposing the sugar by heat. The operation is repeated a sufficient number of times to obtain very dense carbons, which are then submitted to a current of bisulphide of carbon vapour. ARCHEREAU'S CARBON. Archereau mixes carbon with magnesia. The magnesia has the advantage of making the light more steady, and of augmenting its power. CARRY'S CARBON. Carre has made a great number of experiments upon retort carbons impregnated with different salts. Carre proves that potash and soda at least double the length of the voltaic arc, render it more silent, combine with the silica, and eliminate it from the carbons during the action of the current. These substances augment the light in the proportion of 1*25 to 1. Lime, magnesia, and strontia augment the light in the pro- portion of 1*40 to 1. Iron and antimony augment to T60 or 1*70. Boracic acid increases the duration of the carbons by enveloping them with a vitreous layer, which isolates the oxygen from them, but without increasing the light. Carre recommends a composition of powdered coke, cal- cined lampblack, and a syrup of 30 parts of cane sugar and 12 of gum. The following formula is recommended : Coke powder ... ... ... ... 15 parts. Calcined lampblack ... ... ... 5 Syrup ... ... ... ... ... 7 to 8 The whole is strongly triturated, and has added to it from one 236 THE ELEOIRIG^LIGHT. to three parts of water, to compensate for the loss by evapora- tion, according to the degree of toughness to be given to the paste. The coke ought to be pulverized and purified by washing. The coke dust of gas retorts is generally pure enough. The paste is pressed and passed through a draw- plate. The carbons are placed in tiers in crucibles, and are subjected to a high temperature. First, the carbons are placed horizontally in the crucible, resting upon a bed of coke dust, every layer separated by paper to avoid adherence. Between the last layer and the cover is a layer of coke sand, and a layer of silicious sand upon the joint of the cover. The carbons then are put for two or three hours in a concentrated and boiling syrup of cane sugar, or caramel, with two or three intervals of cooling, in order that atmospheric pressure may force the syrup into all the pores. The carbons are then left to drain, by opening a cock placed at the bottom of the vessel, after which they are agitated in boiling water, to dissolve the sugar remaining on the surface. After drying, the carbons are submitted to a second heating, and are manipulated from stage to stage until they have acquired the requisite density or solidity. Carre's carbons are more tenacious, and are harder, than those of retort carbon. They are remarkably similar to, but are better conductors than, retort carbons. GAUDOIN'S CARBONS. Gaudoin also has made numerous experiments upon carbons containing foreign substances. The following sub- stances have been introduced into the carbons: Phosphate of lime from bones, chloride of calcium, borate of lime, silicate of lime, pure precipitated silica, magnesia, borate of magnesia, phosphate of magnesia, alumina, silicate of alumina. The negative carbon being placed at the bottom, M. Gaudoin observed the following results : Complete decomposition of the phosphate of lime under electrolytic action, calorific action, and reducing action of the carbon. The reduced calcium goes to the negative carbon, and burns in contact with the air with a reddish flame. The lime and phosphoric CABBONS. 237 acid are diffused into the air, producing abundant fumes. The light by a photometer is double that produced by carbons of the same section cut from the residue of gas retorts. Chloride of calcium, borate and silicate of lime, are also decomposed, but the boracic and silicic acids escape by vola- tilization from electric action, giving less light than phosphate of lime. Silica in the carbons melts and volatilizes without being decomposed. Magnesia, borate, and phosphate of magnesia are decomposed; the magnesium burns with a white flame. The boracic and phosphoric acids vaporize. Increase of light is less considerable than with the lime salts. Alumina and silicate of alumina are decomposed only with a very strong current, and burn with a blue flame of little lighting power. Carbons intended for the production of the voltaic arc ought to be chemically pure. Eetort carbon, though containing only a small proportion of foreign matter, is not sufficiently pure. Washing in acids or alkalies, with the aim of extracting impurities, is costly and insufficient. Lampblack is pure enough, but its price is high, and its management difficult. M. Gaudoin decomposes by heat, in closed vessels, pitches, fats, or liquids, organic matters capable of yielding carbon sufficiently pure after decomposition by heat. This decomposition is effected in closed retorts, or plumbago crucibles heated to bright red. The bottoms of the crucibles are furnished with two tubes, one for the disengage- ment of gas and volatile matters, the other for the introduction of material. The products of decomposition may be conducted into a condensing chamber, to recover the tars, oils, essences, and hydro-carbons that are produced in this operation. M. Gaudoin utilizes the different sub-products in the manufacture of his carbons. When the material has been properly chosen, carbon, more or less compact, remains in the retort. It is finely pulverized, and mixed with a certain quantity of lampblack, and of the carbides of hydrogen obtained as secondary products. These carbides are completely free from iron, and are much prefer- able to those found in commerce for the compounding of 238 THE ELECTRIC LIGHT. carbons. M. Gaudoin has added to the draw-plate or mould- ing apparatus used in the manufacture of ordinary graphite carbons certain important improvements. Instead of forcing the carbon material through the die vertically, it is caused to issue horizontally from the mould in a descending angle of about 50. The carbon is guided by tubes or gutters. By this means the mould can be completely emptied without interrupting the work, and as the carbon is constantly sup- ported, it does not break under its own weight. In some experiments made to determine the values in lighting power of the various carbons, it was found that when retort carbons produced a light of 824 candle-power, that produced by the artificial carbons varied between 960 and 1620 candle-power for the Archereau and Carre carbons, and between 1600 and 1680 for the Gaudoin carbons. Eeduced to a uniform section, the consumption of the carbons was relatively Retort carbons ... ... ... ... 51 units Archereau ... ... ... ... 66 Gaudoin ... ... ... ... ... 73 Carre 77 Subsequently, M. Gaudoin introduced a process of carbon manufacture in which, instead of carbonizing wood, reducing it to powder, dried wood is taken, shaped in the form of the carbon, which is carbonized and soaked in carbonaceous liquids as previously described. The distillation from the wood is effected slowly, in such manner as to drive out the volatile substances, and the final dessication is effected in a reducing atmosphere, at a very high temperature. The wood is previously washed in acids or alkalies, to remove impurities. By filling up the pores of the wood, by submitting it to the action of chloride of carbon and different carbides of hydrogen under heat, M. Gaudoin proposes to obtain carbons burning at a low rate, and giving a steady light. SAWYER-MANN CARBONS. That the filling up of the pores of the carbon rods by some means is likely to result in the production of the best CARBONS. 239 carbons, has been borne out by the results obtained by Messrs. Sawyer and Mann. These inventors take a carbon rod, and immerse it in olive oil until the oil has thoroughly saturated the pores of the carbon. The carbon rod, still immersed in the oil, is then included in a powerful circuit, the current heating the carbon and carbonizing the oil on its surface and in the pores. Carbons thus produced are extremely hard, and of steel-grey colour on the surface. Used as sources of electric light by incandescence, these carbons give remarkably concordant results. COPPERED AND METALLIZED CARBONS. Numerous suggestions have been made as to a method of rendering carbons more uniform conductors, by coating them with a deposit of metal, as copper or nickel. The same end has been proposed to be attained by incorporating with the carbons copper or iron in powder, by inserting a wire as a core to the carbon rod, or by winding a thin strip of metal around the carbon. Bad carbons are undoubtedly improved by such treat- ment ; and in the very uneconomical method of consuming carbons by forcing the current through the whole of their length adopted, however, in most lamps the regularity of the current, and consequently, in a higher degree, of the light, is dependent upon perfection and continuity of the carbons in the whole length, because a crack or flaw in the carbon may introduce a very variable resistance. Thus, in one of these lamps with automatic action, in which a cracked carbon is mounted and through this carbon the current has to pass, when the two carbons come into contact the pressure closes up the crack, and a sudden increase of conductivity results. Imme- diately the carbons separate, there is correspondingly produced an equally sudden increase of resistance ; presently the carbon becomes heated at the fracture, and flies off,' causing extinction of the light. There is, for this reason, a great advantage due to the Werderman, Eegnier, and Eapieff lamps, in which the current, instead of passing through the whole length of the carbon, is caused to enter the carbon from some point near its burning end. By this arrangement bad carbons will not have 240 THE ELECTRIC LIGHT. so appreciable an effect upon the regularity of the light, nor is it necessary to coat them with metal. Some experiments, conducted by the author, have shown a great increase of light in the voltaic arc when the carbons have been very slightly coated with metallic bismuth, and M. Gramme's experiments with carbon saturated with a solution of nitrate of bismuth, and dried, have confirmed these results. LONDON: PRINTED BY WILLIAM CLOWES AND SONS, STAMFORD STREET AND CHARING CROSS. BOOKS RELATING TO APPLIED SCIENCE, PUBLISHED BY E, & F, N, SPON, LONDON : 46, CHARING CROSS. NEW YORK : 446, BROOME STREET. The Ornamental Penman s, Engraver s, Sign Writer s, and Stone CTitter's Pocket-Book of Alphabets ; including Church Text, Egyptian, Egypt' an Perspective, French, French Antique, French Renaissance, German Text, Italic, Italian Shaded, Italian Hair Line, Monograms, Old English, Old Roman, Open Roman, Open Stone, Orna- mental Roman, Latin, Rustic, Tuscan, etc. Fcap. 8vo, sewn, 6d. Algebra Self-Taught. By W. P. HIGGS, M.A., D.Sc., LL.D., Assoc. Inst. C.E., Author of ' A Handbook of the Differ- ential Calculus,' etc. Crown 8vo, cloth, 2s. 6d. CONTENTS : Symbols and the Signs of Operation The Equation and the Unknown Quantity Positive and Negative Quantities Multiplication Involution Exponents Negative Expo- nents Roots, and the Use of Exponents as Logarithms Logarithms Tables of Logaiithms and Proportionate Parts Transformation of System of Logarithms Common Uses of Common Logarithms Compound Multiplication and the Binominal Theorem Division, Fractions and Ratio Continued Proportion The Series and the Summation of the Series Limit of Series Square and Cube Roots Equations List of Formulae, etc. Progressive Lessons in Applied Science. By EDWARD SANG, F.R.S.E. Crown 8vo, cloth, each Part, 3.5-. Part i. Geometry on Papei Part 2. Solidity, Weight, and Pressure -Part 3. Trigono- metry, Vision, and Surveying Instruments. On Designing Belt Gearing. By E. J. COWLING WELCH, Mem. Inst. Mech. Engineers, Author of ' Designing Valve Gearing.' Fcap. 8vo, sewed, 6r. CONTENTS : What matters may be submitted to Arbitration Of the Submission Of Revocation- Who may Arbitrate Powers of the Arbitrators Of Joint Arbitrators and Umpires Of Evidence Of the Award Of Costs and Charges Advice to Plaintiffs and Defendants- Appendix of Forms. A Handbook of Formula, Tables, and Memoranda, for Architectural Surveyors and others engaged in Building, By J. T. HURST, C.E. Twelfth edition, thoroughly revised and re-written. Royal 3 2mo, roan, $s. CONTAINING : Formulae and Tables for the Strength of Materials, Roofs, Water Supply, Drainage, Gas, and other matters useful to Architects and Builders Information connected with Sanitary Engineering Memoranda on the several Trades used in Building, including a Description of Materials and Analyses for Prices of Builders' Work The Practice of Builders' Measure- ment Mensuration and the Division of Land Tables of the Weights of Iron and other Building Materials Constants of Labour Valuation of Property Summary of the Practice in Dilapidations Scale of Professional Charges for Architects and Surveyors Tables of English and French Weights and Measures. " It is no disparagement to the many excellent publications we refer to, to say that in our opinion this little pocket-book of Hurst's is the very best of them all, without any exception. It would be useless to attempt a recapitulation of the contents, for it appears to contain almost everything that anyone connected with building could require, and, best of all, made up in a compact form for carrying in the pocket, measuring only 5 in. by 3 in., and about \ in. thick, in a limp cover. We congratulate the author on the success of his laborious and practically compiled little book, which has received unqualified and deserved praise from every profes- sional person to whom we have shown it." Ike Dublin Builder. A Treatise on the Use of Belting for the Trans- mission of Power ; with numerous Illustrations of approved and actual methods of arranging Main Driving and Quarter-Twist Belts, and of Belt Fastenings. Examples and Rules in great number for Exhibiting and . Calculating the Size and Driving Power of Belts. Plain, Particular, and Practical Directions for the Treatment, Care, and Management of Belts. Descriptions of many varieties of Beltings, together with chapters on the Transmission of Power by Ropes ; by Iron and Wood Frictional Gearing ; on the Strength of Belting Leather ; and on the Experimental Investiga- tions of Morin, Briggs, and others for determining the Friction of Belts under different Tensions, which are presented clearly and fully, with the Text and Tables unabridged. By JOHN H. COOPER, M.E. I vol., demy 8vo, cloth, 15.$-. Researches on the Action of the Blast Furnace. By CHARLES SCHINZ. Translated from the German by \V. H. Maw and Moritz Miiller. Plates^ crown 8vo, cloth, &r. 6d. Spons Builders Pocket-Book of Prices and Memo- randa. Edited by W. YOUNG, Architect. Royal 321110, roan, 4!-. 6t/. ; or cloth, red edges, 3-r. f> 2/. 2s. PUBLISHED BY E. & F. N. SPON. Hops, their Cultivation, Commerce, and Uses in various Countries. By P. L. SiMMONDS. Cown 8vo, cloth, 4*. 6r. Table of Logarithms of the Natural Numbers, from i to 108,000. By CHARLES BABBAGE, Esq., M.A. Stereotyped edition, royal 8vo, cloth, JS. 6d. To ensure the correctness of these Tables of Logarithms, they were compared with Callett's, Vega's, Mutton's, Briggs', Gardiner's, and Taylor's Tables of Logarithms, and carefully read by nine different readers ; and further, to remove any possibility of an error remaining, the stereotyped sheets were hung up in the Hall at Cambridge University, and a reward offered to anyone who could find an inaccuracy. So correct are these Tables, that since their first issue in 1827 no error has been discovered. BARLOW'S Tables of Squares, C^lbes, Square Roots, Cube Roots, Reciprocals of all Integer Numbers up to 10,000. Post 8vo, cloth, 6s. CAMUS (M.) Treatise on the Teeth of Wheels, demon- strating the best forms which can be given to them for the purposes of Machinery, such as Mill-work and Clock-work, and the art of finding their numbers, translated from the French. Third edition, carefully revised and enlarged, with details of the present practice of Millwrights, Engine Makers, and other Machinists. By ISAAC HAWKINS. Illustrated by i% plates, 8vo, cloth, $s. The Practical Sugar Planter; a complete account of the Cultivation and Manufacture of the Sugar Cane, according to the latest and most approved processes, describing and comparing the different systems pursued in the East and West Indies and the Straits of Malacca, and the relative expenses and advantages attendant upon each, being the result of sixteen years' experience as a Sugar Planter in those Countries. By LEONARD WRAY, Esq. 8vo, cloth, IQJ. 6d. Laying and Repairing Electric Telegraph Cables. By Capt. V. HOSKICER, Royal Danish Engineers. Crown 8vo, cloth, The Practice of Hand Turning in Wood, Ivory, Shell, etc., with Instructions for Turning such Work in Metal as may be required in the Practice of Turning in Wood, Ivory, etc., also an Appendix on Ornamental Turning. By FRA.NCIS C AMPIN. Second edition, with wood engravings, crown 8vo, cloth (a book for beginners), 6s. CONTENTS : On Lathes Turning Tools Turning Wood Drilling Screw Cutting Miscellaneous Apparatus and Processes Turning Particular Forms Staining Polishing Spinning Metals Materials Ornamental Turning, etc. PUBLISHED BY E. & F. N. SPON. 15 Health and Comfort in House Building, or Ventila- tion with Warm Air by Self-Acting Suction Power, with Review of the mode of Calculating the Draught in Hot- Air Flues, and with some actual Experiments. By J. DRYSDALE, M.D., and J. W. HAYWARD, M.D. Second edition, with Supplement, demy 8vo, with plates, cloth, "js. 6\ 248903