SEVENTEENTH LECTURE ARC LIGHTING W"^HILE incandescent lamps can be operated on constant potential as well as on constant current, the arc is —^ essentially a constant current phenomenon. At con- stant length, the voltage consumed by the arc decreases with increase of current, as shown by curve I in Fig. 47. If, there- fore, an attempt is made to operate such an arc on constant potential, for instance on 80 volts — which would correspond to 3.9 amperes on curve I — then any tendency of the current to increase — as by a momentary drop of the arc resistance — would lower the required arc voltage, and so increase the cur- rent, at constant supply voltage, hence still further lower the arc voltage, etc., and a short circuit would result. Vice versa, a momentary decrease of arc current, by requiring more volt- age than is available, would still further decrease the current, increase the required voltage, etc., and the arc would extin- guish. Therefore only such apparatus is operative on constant potential, in which an increase of current requires an increase of voltage, and vice versa; and so limits itself. While therefore arcs can be operated on a constant cur- rent system, to run arc lamps on constant potential, some cur- rent limiting device is necessary in series with the arc, as a resistance; or, in an alternating current circuit, a reactance. The voltage consumed by the resistance is proportional to the current, and a resistance of 8 ohms inserted in series to the arc would thus consume the voltage shown in straight line II in Fig. 47. The voltage consumed by the arc plus the resistance then is given by the curve III, derived by adding I and II. As 2l8 GENERAL LECTURES "" "~ " mm ^ ^ -m r "" "" r MB ■"T* "T~" ~ P '/7 ^ ? T" r \ 1/ ^(i /? 9/ it p ^7 'C ^ \ \ / rrt \ w ^ / :«i \ ^ / il^ \ \ • \ \ ■ 7 ?^ \ \\ \ ^ -- \ s, \ 1 ?rt \ s \ r ^ --'' V s S iZ r. •-- ■* ■ \ «^ 'sl ih. ,- „ 1 '<3 \ S V, in S ^ \ "-^ T' ' $ C 'V s a i " I V v / * ---. ^ vi 0 i ^ ***■ -, K ■M 7 0 ^ *■"- — ■ / 0 "■ — ^' « n ^ ^-^ ■^ X "^ &" < A Jl f^ ^ -« fr ^ f < n ^ ^ * ^ ^ /Ij %* £/ ^•{ ," ^^ e: ^ < A > L. I ^ £ < f i t „ ^ Fig. 47. seen, below 3.35 amperes, the total required voltage still decreases with increase of current, and the arc is still unstable ; that is, the resistance is insufficient. Above this current, an increase of current requires an increase of voltage and so limits itself; that is, the arc is stable; with 8 ohms series resist- ance, 3.35 amperes therefore is the limit of stability of the arc; and attempting to operate it at lower current, as for instance at 2 amperes and 116 volts supply, the arc either goes out, or ARC LIGHTING 219 the current runs up to 5.5 amperes, where the arc becomes stable on 116 volts supply. With a higher series resistance, the arc remains stable to lower currents, and vice versa. It follows herefrom, that for the operation of an arc lamp on constant potential, a higher voltage is required than that consumed by the arc proper. At every value of series resistance therefore a point a in Fig. 47, is reached, at which for decreasing current the arc becomes unstable; and all these points, for different resistance values, give a curve IV, which is called the "stability curve" of the arc curve I. The supply voltage required to operaite the arc represented by curve I must therefore be higher than that given by the stability curve IV. For instance, at 4 amperes, the arc cannot be operated at less than 104 volts supply. At 104 volts supply the limit of stability is reached ; that is, a change of current does not require a change of voltage, but the arc voltage decreases as much as the resistance voltage increases and the current thus drifts; and for supply voltages higher than 104, the arc is stable, the more so, the higher the supply voltage is above 104. The difference in voltage between the supply voltage and the arc voltage thus is consumed by the "steadying resistance" of the arc. High reactance in series with the direct current arc retards the current fluctuations and so reduces them ; so that with reactance in series to the direct current arc, the arc can be operated by a supply voltage closer to the stability curve IV than without reactance ; reactance therefore is very essential in the steadying resistance of a direct current arc. Obviously, no series reactance can enable operation of the arc I on a supply voltage below that given by the stability curve IV. 220 GENERAL LECTURES The arc characteristic I is far steeper for low currents than for high currents, and is the steeper the greater the arc length. Low current arcs and long arcs therefore require, that on a constant potential supply, a greater part of the supply voltage is consumed by the steadying resistance (or steadying reactance with alternating arcs) than high current arcs, or short arcs; and are therefore less economical on constant potential supply. Constant potential arc lamps are necessarily less efficient than constant current arc lamps, due to the power con- sumed in the steadying resistance. A large part of this power is saved in alternating constant potential arc lamps, by using reactance instead of resistance, but the power factor is there- fore greatly lowered ; that is, the constant potential alternating arc lamp rarely has a power factor of over 70%. Where therefore high potential constant current circuits are permissible, as for outdoor or street lighting, arc lamps are usually operated on a constant current circuit, with series connection of from 50 to 100 lamps on one circuit. With the exception of a few of the larger cities, all the street lighting by arc lamps in this country is done by constant current systems, either direct current or alternating current. For direct current constant current supply, separate arc light machines have been built, and are still largely used. In these machines, inherent regulation for constant current is produced by using a very high armature reaction and relatively weak field excitation; that is, the armature ampere turns are nearly equal and opposite to the field ampere turns, and thus both very large compared with the difference, the resultant ampere turns, which produce the magnetic field. A moderate increase of current and consequent increase of armature ampere turns therefore greatly reduces the resultant ampere turns and ARC LIGHTING 221 so the field magnetism and the voltage, (that is, the machine tends to regulate for constant current. Perfect constant current regulation then is secured by some governing device, as an auto- matic regulator varying a resistance shunted across the series field. It must, however, be understood .that the "regulator" of the arc machine does not give a constant current regula- tion, but the armature reaction of the machine does this, and the regulator merely makes it perfect ; but even with the regu- lator disconnected, arc machines give fairly close constant cur- rent regulation. As the voltages produced by arc machines are very high — 4,000 to 10,000 — commutation of the current, with the ordinary commutator, which is limited to a maximum of 40 to 50 volts per segment — is not well suited, but rectification is used. The Brush arc machine therefore is a quarter-phase alternator with rectifying commutator. That is, the commuta- tor shifts the connection over from the phase of falling e. m. f. to that of rising e. m. f., and thereby is able to control as high as 3,000 volts per commutator ring. With the development of the mercury arc rectifier, which converts constant alternating current into constanit direct current, arc machines are going out of use. The arc machine necessarily must be a small unit, since 100 to 150 lamps in series give already as high a voltage as is safe to use in arc circuits, but do not yet represent much power; and when supplying thousands of arc lamps a large number of small machine units are required, which are uneconomical in space, in attendance and in efficiency. The mercury arc rectifier in combination with the stationary constant current transformer enables us to derive the power from the alternating current con- stant potential supply system. 222 GENERAL LECTURES Constant alternating current is derived by a constant cur- rent transformer or constant current reactance. Diagram- matically, the constant current transformer is shown in Fig. 48. The primary coil P and the secondary coil S are movable Fig. 48. with regard to each other (which of the two coils is movable, is immaterial, or rather, is determined by consideration of design). Fig. 48 shows the coil S suspended and its weight partially balanced by counter-weight W. With the secondary coil S close to the coil P, that is, in the lowest position, most of the magnetism produced by the primary coil P passes through the secondary coil S, and the secondary voltage therefore is a maximum. The further the secondary coil moves away from the primary coil, the more of the magnetism passes between the coils, the less through the secondary coil, and the lower therefore is the secondary voltage, which becomes a minimum (or zero, if so desired), with the ARC LIGHTING 223 secondary coil at a maximum distance from the primary, that is, in the top position. Primary current and secondary current are proportional and in opposition to each other, and repel each other, and the repulsion is proportional to the product of the two currents; that is, proportional to the square of the secondary current. The weight of the secondary coil is balanced by the counter- weight W and the repulsion from the primary coil, at normal secondary current. Any increase of secondary current by a decrease of load, increases the repulsion, in this way pushing the secondary coil further away from the primary and thereby reducing the secondary voltage and thus the current ; and vice versa, a decrease of secondary current, by an increase of load, reduces the repulsion and so causes the secondary coil tc come nearer to the primary, that is, increases its voltage and so restores the current. Such an arrangement regulates for con- stant current between the voltage limits given by the two ex- treme positions of the movable coil. These usually are chosen from some margin above full load, down to about one-third load. The constant current reactance operates on the same principle: the two coils P and S are connected in series with each other into the arc circuit supplied from the constant potential source, and by separating or coming together, vary in reactance with the load, and thereby maintain constant current. While the alternating current arc lamp is less efficient, that is, gives less light for the same power, than the direct cur- rent arc lamp, the disadvantages of the use of numerous arc machines have led to the extended adoption of alternating cur- rent series arc lighting before the development of the mercury 224 GENERAL LECTURES arc rectifier, which enabled the operation of direct current arc circuits from constant current transformers. While incandescent lamps give the same efficiency for all sizes except such small sizes where mechanical difficulties appear in the filament production, the efficiency of the arc decreases greatly with decrease of current ; that is, the arc is at the greatest efficiency only for large units of light, but rather inefficient and not so well suited for small units of light. Even in large units, the efficiency of light production of the direct current carbon arc lamp is not superior to that of the tungsten incandescent lamp ; that of the alternating current carbon arc lamp is inferior to the tungsten lamp ; and the carbon arc lamp thus finds its field mainly where large units of light are required, especially as long as the cost of renewal of the metal filament lamps is still very great. Entirely different, however, are the conditions developed in the last years, with the luminous arcs, as the flame carbon arc, the mercury lamp, and the magnetite and titanium carbide arc. In these, efficiencies of light production have been reached which no incandescent lamp can hope -to approach. In the carbon arc, practically all the light comes from the incandescent tips of the carbons, very little from the arc flame. Then by using materials, which in the arc flame give an intense- ly luminous spectrum, the efficiency of \the arc lamp has been vastly improved. So far only three materials have been found, which in luminous arcs give efficiences vastly superior to incandescence : mercury, calcium (lime), and titanium. All (three even in moderate sized units, give efficiencies of one-half watt or better per candle power. The mercury arc has the advantage of perfect steadiness, a long life — requiring no attention for thousands of hours — ARC LIGHTING 225 and high efficiency over a fairly wide range of candle powers ; but it is seriously handicapped for many purposes by its bluish- green color. In the flame carbon lamp carbons impregnated with cal- cium compounds, usually calcium fluoride (fluorspar) are used, and the arc then has an orange-yellow color. Other com- pounds which give red or white color to the arc are so much inferior in efficiency that they are used only to a very limited extent. The compounds, after coloring the arc and giving it efficiency, escape as smoke; the arc therefore must be an open arc. This, however, means short life of the carbons and fre- quent trimming. The open arc lamp, which was used formerly, has, how- ever, been almost entirely superseded by the enclosed carbon arc, in spite of the somewhat lower efficiency of the latter ; and the inconvenience of daily attendance required by an open arc, and the large consumption of carbons, makes a return to this type improbable. For this reason the flame carbon lamp has not proven suitable for general outdoor illumination, as street lighting, where the cost of carbons and trimming would usually far more than ofl^set the gain in efficiency. Flame carbon lamps, however, have found a field for decora- tive lighting, for advertising purposes, etc., for which the glare of light and its color makes them very suitable. They are generally used on constant potential circuits with two or three lamps in series. To eliminate the objections of short life and consequent frequent trimming and high cost of carbons, and thereby make the luminous arc able to enter the field of general outdoor il- lumination, carbon had to be eliminated altogether as electrode material, and its place was taken by magnetite, while titanium compounds give the high efficiency. This lead to the long 226 GENERAL LECTURES burning luminous arc of the white color of the titanium-iron spectrum as represented by the magnetite arc, the metallic oxide arc, and other types still in development. In all these long burning luminous arcs, some efficiency had to be sacrificed in developing sufficiently small units for general illumination. While the substitution of -the flame car- bon in the open arc has quadrupled the light at the same power consumption, and the substitution of the magnetite electrode for carbon at the same power consumption would in the same manner increase the light, for street illumination the main problem was, to decrease the power consumption rather than increase the amount of light given ; and so in the long burning luminous arcs, which are now beginning to replace the carbon arcs of old, the power consumption has been reduced by from 30 to 60% with a sufficient increase of light to be marked. In the arc lamp, the current is carried across the gap between the terminals by a stream of vapor of the electrodes ; thus the electrodes consume more or less rapidly. Some feeding mechanism is therefore required to move the electrodes towards each other during their consumption. This arc lamp mechanism may be operated by the current, or by the voltage, or by both ; this gives the three different types of lamps : the series lamp, the shunt lamp, and the differential lamp. In the series lamp, an electromagnet energized by the lamp current, and balanced against a weight or a spring, moves the carbons towards each other when by their burning off, the arc lengthens and the current decreases. Obviously, this lamp cannot be used on constant current circuits, or with several lamps in series, but only as single lamp on constant potential circuits, and therefore has practically disappeared. In the shunt lamp, the controlling magnet is shunted across the arc, and with increasing arc length and consequent ARC LIGHTING 227 arc voltage, moves the electrodes towards each other. In con- stant current circuits, this lamp tends towards hunting, and therefore requires a very high reactance in series; it thereby gives a lower power factor in alternating current circuits, and has therefore been superseded by the differential lamp. It has, however, the advantage of not being sensitive to changes of current. In the differential lamp, an electromagnet in series with the arc opposes an electromagnet in shunt to the arc, and the lamp regulates for constant arc resistance. It is the lamp now universally used in constant potential and constant cur- rent systems, is most stable in its operation; but in constant current systems, it requires that the current be constant within close limits : if the current is low, the arc is too short, and the lamp gives very little light, and if the current is high, .the arc becomes so long as to endanger the lamp. From the operating mechanism the motion is usually transmitted to the electrode by a clutch, which releases and lets the electrodes slip together. In the carbon arc lamp, the mechanism is "floating" ; that is, the upper carbon, held by the opposing forces of shunt and series magnets, moves with every variation of the arc resist- ance, and so maintains very closely constant voltage on the arc. In the long burning luminous arc, as the magnetite lamp, the light comes from the arc flame, and thus constant length of arc flame is required for constant light production. The floating mechanism, which constantly varies the arc length with the variation of the arc resistance, has therefore been superseded by a mechanism which sets the arc at fixed length, and leaves it there until with the consumption of the electrodes the arc has sufficiently lengthened to cause the shunt coil to 228 GENERAL LECTURES operate and to reset the arc length. Thus in some respects, these lamps are shunt lamps. During the early days of the open carbon arc lamp, 9.6, 6.6 and 4 amperes were the currents used in direct current arc circuits, with about 40 volts per lamp. The 4 ampere arc very soon disappeared, as giving practically no light. In the enclosed arc lamp, the carbons are surrounded by a nearly air tight globe, which restricts the admission of air and thus the combustion of the carbon, and so increases the life of the carbons from 8 or 10 hours to 70 to 120 hours. In these lamps, lower currents and higher arc voltages, that is, longer arcs, are used: in direct current circuits, 6.6 amperes and 5 amperes, with 70 to 75 volts per lamp; in alternating current circuits, 7.5 and 6.6 amperes are used with the same arc voltage. In the direct current magnetite arc lamp, 4 amperes and 75 to 80 volts per lamp are used ; in the alternating current titan- ium carbide arc lamp, only 2.5 amperes and 80 to 85 volts per lamp are used. ' I 1 1 i APPENDIX I LIGHT AND ILLUMINATION Paper read before the Illumlnotln^ Engineering Society, December 14, 1906. Revised to date. I. C"~" OMPARED with other branches of engineering, as the transformation of electrical power into mechanical — ^ power in the electric motor, or the transformation of chemical into mechanical energy in the steam engine, we are at the disadvantage when dealing with light and illumination, that we have not to do any more strictly with a problem of physics, but that we are on .the borderland between applied physics, that is engineering, and physiology. Light is not a physical quantity, but it is the physiological effeot exerted upon the human eye by certain radiations. There are different forms of energy, all convertible into each other, as magnetic energy, electric energy, heat energy, mechanical momentum, radiating energy, etc. The latter, radi- ating energy, is a vibratory motion of a hypothetical medium, the ether, which vibration is transmitted or propagated at a velocity of about 188,000 miles per second; and it is a transverse vibration, differing from the vibratory energy of sound in this respect, that the sound waves are longitudinal, that is, the vibration is in the direction of the beam, while the vibration of radiation is transverse. Radiating energy can be derived from other forms of energy, for instance, from heat energy by raising a body to a 230 GENERAL LECTURES high temperature. Then the heat energy is converted into radi- ation and issues from the heated body, as for instance an incan- descent lamp filament, as a mass of radiations of different wave lengths, that is, different frequencies. All kinds of frequencies appear : from very low frequencies, that is only a few millions of millions of cycles per second, up to many times higher frequencies. We can get, if we desire, still very much lower fre- quencies, as electromagnetic waves, such as the radiation sent out by an oscillating current or an alternating current ; but the radiations which we get from heated bodies are all of extremely high frequency, compared with the customary frequencies of electric currents. At the same time they cover a very wide range of frequencies, many octaves, and from all this mass of radiations, from all the frequencies of radiating energy, some- what less than one octave can be perceived by the human eye as light. Light, therefore is the physiological effect exerted upon the human eye by a certain narrow range of frequencies of radiation. Frequencies lower than those visible to the eye, and frequencies higher than those visible to the eye, are again invisible. We frequently speak of those frequencies which are lower than the visible ones, as radiating heat, and of those frequencies higher than the visible ones as chemical rays. This, however, is misleading, and there is no distinction in character between radiations of different frequency. There are no heat rays differing from light rays or chemical rays. Any form of energy when destroyed gives rise to an exactly equivalent amount of some other form of energy. If there- fore we destroy radiating energy by intercepting the beam of radiation by interposing an opaque body in its path, then the energy of radiation is converted into some other form of LIGHT AND ILLUMINATION 131 energy, usually into heat. That means that any radiation when absorbed produces heat and the amount of heat pro- duced merely represents the amount of energy which was con- tained in the radiation. If the radiation contains a very large amount of energy, the heat evolved by intercepting it may be sufficient to be felt by putting your hand in the beam. If the amount of energy is less, it may not be possible to feel it, though with a sensitive instrument, as a bolometer, we may still be able to measure the heat. All radiations therefore are convertible into heat: the visible light waves as well as the invisible ultraviolet rays, and the — usually more powerful — long ultrared waves ; but none of the radiations can be called heat, no more than the mechanical momentum of a flywheel is heat, because when destroyed, it produces heat. If we consider the infinite range of radiation issuing from heated bodies, we find that those rays which are of lower fre- quency than the visible rays will be felt as heat, because they contain a very large amount of energy. The rays which are visible represent very little energy — and therefore (they do not give as much heat. For instance, in the case of a hot steam boiler, although we get no light, we can feel the radia- tion from it by the heat which it produces when intercepted by our hand held near it. We do not feel the radiation as heat which issues from the green light of the mercury lamp, merely because the energy of radiation in the latter is less than the amount of energy in the radiation from a hot steam boiler ; but while it is less in the former case, it happens to be of that fre- quency which affects the eye and is visible. As a consequence, when we speak of cold light, this does not mean that it is different from hot light — from the light, for instance, given by a hot coal fire, where we feel the radia- tion as heat; it merely means that what is usually called cold 232 GENERAL LECTURES light (as the light of the firefly is supposed to be) is radia- tion containing to a very large extent rays of the visible frequencies and not much energy outside of the visible range; i. e., containing very little total energy, so that the energy when destroyed, that is, converted into heat, cannot be felt easily, but requires more delicate methods of determination; while a very inefficient light, as a coal fire for instance, which gives most of its energy as invisible radiation of low frequency, very little as visible radiation, can be felt by the heat pro- duced by the interception of the rays, mainly the energetic low frequency rays. As stated, then, there is no essential differ- ence between so-called heat waves and light waves, but any radiation can be converted into other forms of energy, the so- called chemical rays of ultraviolet light, the X-ray, as well as the ultrared and the visible rays, and when converted into heat can be noticed as such. Now it just happens that most of our means of producing radiating energy give high intensi- ties of radiation only for very low frequencies, invisible ultra- red rays, but we are not able to produce anywhere near the same intensities of radiation for higher frequencies. So also, when we speak of ultraviolet, or short, high frequency waves, as chemical waves, thait does not mean that they have a distinctive character in producing chemical action — any form of energy, naturally, can be converted if we know how, into chemical energy, the long ultrared waves just as well as the short ultraviolet waves. It just happens that those chemical compounds which are easily split up by radiating energy, are silver salts or salts of gold and platinum ; they are especially affected by the ultraviolet and violet rays. We observe, then, the chemical action of these rays, but do not observe so well the chemical action of other rays. There may, however, be some feature in the constitution of matter, which LIGHT AND ILLUMINATION 233 accounts for the high chemical action of the ultraviolet and violet rays. It is obvious that if we intercept and destroy radiations and so convert their energy into other forms of energy, if the energy is only great enough, we get a high temperature, and thus a high chemical action, merely by the effect of temperature. But we may also get a chemical effect by what probably is some kind of a resonance phenomenon. The particles of a body, atoms or molecules, must have some rate of vibration of their own. If, then, a ray of radiation impinges upon them which is of a frequency of the same magnitude as the inherent rate of vibration of the atom, by resonance this vibration of the atom must rapidly increase in intensity until the atom breaks away from the others, or the molecule breaks up, that is, the chemical combination is split up. The inherent frequency of oscillation of the atom seems to be of about of the same magnitude as the visible radia- tion, or rather of a little higher frequency; that is, if the atoms are left to vibrate freely as under the influence of an electric current in the arc, then we get radiations of the frequency inherent to the atom. The general tendency then is toward the violet or short wave end of the spectrum. If we assume that the mass of the silver atom is such as to give a rate of vibration in the range of the violet and ultraviolet, it is easy to understand that radiation of this frequency splits up the silver salt by increasing the vibration of the atom by resonance, and that shorter or longer waves have no effect, or much less effect. So it may be a mere incident that those chemical compounds on which we observe the chemical action of radiation just happen to be sensitive to the violet end of the spectrum. It is indeed a fact that other chemical changes brought about by radiating energy, as the formation of ozone 234 GENERAL LECTURES from oxygen, that is, the splitting up of the oxygen molecule and reforming of the ozone molecule from the atoms, do not take place in the violet or ultraviolet, but requires frequencies very much higher, about the highest frequencies which the mercury arc at low temperature gives. Possibly, since the oxygen atom is so much lighter than the silver atom, its fre- quency of vibration is much higher, which means that resonance effects and destruction of the molecules take place only with a much shorter wave length of radiation, or much higher frequency. Vice versa, it seems that these frequencies which are chemically active in organic life, which give the energy absorbed from radiation by plants, and so the chemical activity utilized in building up the growth of vegetation, are not at the violet, but at the red end of the spectrum. It appears that the red and ultrared rays produce growth of plants and the chemical activity which we call life, while the violet and ultraviolet rays kill. Even this we can well under- stand if we consider the chemical activity as a resonance phenomenon, because the metabolism of protoplasm which we call life, is based on the existence of unstable structures of carbon compounds. We have here not atoms combining with each other, but groups, chain and ring formations, which are of larger mass and therefore have a lower rate of vibration and so should be expected to respond to lower frequencies or to red light, as indeed seems to be the case. The violet and ultraviolet light does not split up the organic matter into groups, which recombine to form complex bodies, and so represent the changes called life; but due to its higher fre- quency, resonance occurs with the atoms, that is, the organic compound splits up into atoms and so disintegrates, which means death. LIGHT AND ILLUMINATION 235 So it can be understood that the chemical activity of different radiations may be different; the chemical activity of long rays gives life to the vegetation and the short waves, The popular distinction between heat waves, chemical waves and light waves, therefore is not a physical distinction, destruction down to the atom. death; one splits up into carbon groups and the other carries but all are radiating energy of the same character, differing merely in wave length, and the visible range is somewhat less than one octave, rather at the upper end, at the higher fre- quencies, which are difficult to produce. This makes the prob- lem of investigating and dealing with light difficult for the engi- neer, because it is not any more a physical quantity which can be measured accurately, as in the case of power or velocity, but it is a physiological effect. We can, indeed, measure very accurately the total energy of radiation from a heated body, but the total energy of radiation is not light : only a very small part of it is visible. We can go further and split up the total radiation issuing from a hot body, as the incandescent lamp fila- ment, into its different wave lengths and different frequencies ; as for instance, we can resolve the total radiation into the spectrum by using a prism to separate the different frequencies, and then collect the total of the radiation within the visible range, by a lens or other means, and measure all the energy of the visible radiation. Or, still simpler, although approximate, we may interpose in the beam of radiation some medium which absorbs the invisible long rays and invisible short rays, and which transmits, all or rather most, of the visible rays, as for instance glass and water. In this manner one could easily measure the energy of the visible radiation, and compare the energy of the visible radiation with the total energy producing this radiation. That would give a physical measure of the 236 GENERAL LECTURES efficiency of producing visible radiation but it would not be a measure of the efficiency of producing light, since unfortunately the different wave lengths of visible radiation are very differ- ent in their physiological effect. The same amount of energy as visible radiation, giving the effect of green light, represents an entirely different amount of light, a many times greater physiological effect than the same amount of energy as red rays, that is, rays of the wave lengths which give the impres- sion of red light. That means, the physiological effect or light-equivalent of mechanical energy within the visible range — is a function of the wave length and varies with the wave length, that is, with the color. That really is obvious, if you think of it: if you follow the range of frequency from a low frequency to a high frequency, you see that energy radiating at low frequency represents no light whatever, has no physiological equivalent, is invisible. When you come into the visible range it has a physiological effect. Therefore, when you pass from the invis- ible into the visible range, the physiological equivalent must pass from zero into a finite value and must necessarily pass continuously, that is, at the extreme end of the visible range ; the light equivalent of energy must be extremely low, and the further you go into the visible range, the greater it is, reaching the maximum in the middle of the visible range, in the green and yellow, and decreasing again down to zero at the violet end of the visible rays ; beyond that, at still higher frequencies, the physiological equivalent of energy is zero again ; or, vice versa, if we consider the mechanical equivalent of light, it is a mini- mum in the middle of the visible range, where one candle power of light represents the lowest amount of energy, and increases toward the ends of the spectrum of the visible range, to infinity at the ends of the visible range. LIGHT AND ILLUMINATION 237 Now, that means, in plain language, that the efficiency of light production is a function of the wave length, that is of the color, and that it is at its maximum in the middle of the spectrum, where the same amount of power, measured in watts, gives the largest amount of light measured in candle power. So the efficiency of light production is a function of the wave length. Unfortunately, the physiological equivalent of power, or the physiological effect of light varies not only with the wave length, but also with the absolute intensity. Suppose we undertake to compare red, yellow and green lights, or any lights of different colors. First we meet great difficulties in comparing them. We want one candle power in light, as red, yellow, or green. You cannot compare different colors of light directly, since there is no physical measure of light. Lights are compared with a standard lamp, which has a cer- tain color, yellowish white. A light of the same color we can compare exactly; if the color is not much different, we still get an approximate comparison; but with widely different colors, we obviously can not get even an approximate com- parison, can not say when the two sides of the photometer screen, one illuminated by green light, the other by red light, are equal in intensity. There is thus no direct comparison of differently colored lights. You have then to go one step farther and consider that light is used for illumination, is used to see by, and this gives you a fair comparison : you observe at what distance from the two lights, red and green, you can read with the same convenience, read the same kind of print, or to meas- ure more exactly, get the maximum distance at which you can just read a certain size of print, by either light. At that distance the two illuminations are the same, and the two lights so have an intensity inversely proportional to the square of 238 GENERAL LECTURES these distances. In this manner lights of different color can be compared. Necessarily, the comparison has not the accuracy of photometrical comparison. This cannot be expected, since you do not compare physical quantities, but only physiological effects on the eye, and different observers may have different personal contants. The eye of the one may be more sensitive to green, and the eye of the other may be more sensitive to red, and therefore the comparison may be different. However, these individual differences are not great, and different observers, even with widely different colors of light, do not give results differing much from each other, so that a com- parison of intensities of differently colored lights, and thereby a measurement of the intensity of differently colored lights in candle power is feasible, by some such method, that is, of observing the illumination produced by the different lights. You find, however, if you have a green light and a red light, which at a certain distance appear equally brilliant to the eye, then when you get nearer to the two lights the orange red light appears much brighter than the green, and when you go further away the green light appears brighter, and at still greater distances you still see the green light fairly brightly, while the red light is almost invisible. That is, the relative physiological effect of different wave lengths varies, not only with the wave lengths, but also with the absolute intensity of illumination, and while throughout the whole range the sensi- tivity of the eye for green light is much greater than for red light, the difference is far gfreater for low than for high illumination, that is, the ratio of sensitivity for green com- pared with that for red is greater for faint illumination than for intense illumination. If you desire to express lights of different colors in candle power it therefore seems necessary LIGHT AND ILLUMINATION 239 also to give the distance, or the intensity of illumination at which you have observed; in other words, the light from the middle and the short wave end of the spectrum gives a better and more efficient illumination where the total intensity of illumination is low, while the long wave or low frequency of the red and orange and yellow light gives a much more bril- liant effect at high intensity than the same volume of light of shorter wave length. This is of importance for the illuminating engineer, because where you desire to get high intensity effects, as in decorative lighting or in advertising, better results are given by the low frequency end of the spectrum, by orange and yellow light, whereas when you are satisfied with low intensity of illumination, as in street lighting, you get better results from the short wave end or the middle of the spectrum, from the greenish-yellow of the Welsbach gas light and the bluish- green of the mercury lamp, and not from the orange-yellow of the old incandescent lamp. Therefore the white light of the carbon arc gives better results in street lighting than the yel- low of the incandescent lamp, even at equal intensity of illumination. These features have been of less importance until a few years ago, since the available sources of light were all approximately of the same color, varying from the orange- yellow, to yellow and yellow-white, to white; from the g^s lamp, kerosene lamp and tallow candle of orange-yellow color> to the yellow incandescent lamp and the yellowish-white arc, yellowish-white sunlight, to the white diffused daylight. This was a fairly limited range. It is only in the last few years that illuminants of high efficiency have been brought out, which give marked and decided color differences, and are available in units of suitable size and of high efficiency, as the greenish- yellow of the Welsbach gas lamp, the bluish-green of the mer- 240 GENERAL LECTURES cury lamp, and the orange-yellow of the flaming arc, and hence these questions are increasing in importance. IL This brings us to the consideration of the methods of producing light. Until a few years ago, until the develop- ment of the Welsbach gas mantle, practically all methods of producing light were based on incandescence. That is, by impressing energy on a solid body, either the chemical energy of combustion, or electric energy in the incandescent or car- bon arc lamp, the temperature is raised to such a high degree that amongst the total radiation issuing from the heated body a certain very small percentage appears within the fraction of an octave of visible light. With increasing temperature of the radiating body, the average wave length of radiation decreases, that is, the average frequency of radiation increases and so approaches nearer to the visible range, although still at the very highest temperature which can be produced the average wave length of radiation is very far below the visible. This means that the higher a temperature is reached by an incandescent body, the higher is the average frequency of radiation, and therefore the larger a percentage of the total energy of radiation is within the visible range, as light. The problem of efficient light production by incandescence therefore is the problem of reaching as high a temperature as possible in the luminous body. In the gas flame and the kerosene lamp, this temperature is the temperature of combustion, rather lim- ited. In the incandescent lamp it is limited also. In the latter case the temperature which can be reached is limited by self- destruction of the incandescent body. The highest temperature probable which can be reached is the boiling point of carbon ; it is reached in the crater of the LIGHT AND ILLUMINATION 241 carbon arc lamp, and therefore the carbon arc gives the most efficient incandescent light. It is incandescent light, because it comes from the incandescent crater, and the arc flame or the vapor conductor does not appreciably contribute to the amoimt of light issuing from the arc lamp. Very much lower, neces- sarily, is the temperature of the incandescent lamp, of the car- bon filament. The problem is to find materials which can stand very high temperatures, to increase the temperature of the gas flame as well as of the incandescent filament. We have increased the temperature of the gas flame by using a gas of higher chemi- cal energy, as acetylene. The acetylene flame is white; the ordinary gas flame is yellow. We have increased the tempera- ture of the carbon filament by replacing the carbon with some more refractory material, such as tantalum, osmium, tungsten, etc., and thus getting a higher efficiency. We can increase -the temperature of the gas flame by increasing the rapidity of com- bustion. We can increase the temperature of the carbon fila- ment in the incandescent lamp by increasing the energy input, but if we increase the temperature of the carbon filament, it is more rapidly destroyed. If we increase the temperature of the gas flame by more rapid combustion — to a certain extent we have done it already, by having the gas issuing not from a round hole, but from a flat slit, so as to give a larger surface to the flame; if we go still further and mix the gas with air, we get a still higher temperature, a more rapid combustion, but we lose the incandescent body, because the incandescence of the gas flame is the light given by carbon or heavy hydro- carbon particles, floating in the gases of combustion. We could increase the efficiency of the gas flame by mixing the gas with air, as in the Bunsen flame, but we have then to insert a luminous body of some other material, as no carbon is pro- 242 GENERAL LECTURES duced by the gas in its dissociation. We can do it by a skele- ton of platinum wire. In no case, however, can we reach very high efficiencies by incandescence, due to the temperature limit. We could, however, increase the efficiency of light pro- duction if we could find an incandescent body which would not radiate in the same manner as the carbon filament or the so- called black body, but which would give an abnormally low rad- iation in the low frequency range, or an abnormally high radia- tion in the high frequency or visible range. Such a body may be said to give selective radiation, because the distribution of ener- gy in the spectrum amongst the different frequencies of radi- ation is not the same as it would be with an ordinary black body of the same temperature. If we found a body which would give an abnormally low radiation in the visible range, or abnormally high radiation in the invisible range, this body would be an abnormally inefficient light producer. Vice versa, if we found a body giving abnormally high radiation of short wave lengths, in the visible range, or abnormally low radiation of long waves, of low frequency, this would give an abnormally efficient incandescent body. Such bodies exist and the enor- mous progress in gas lighting made by the introduction of the Wellsbach mantle is based on selective radiation, that is, the oxides do not radiate the same range and intensity of waves as a black body, the incandescent carbon, but give an abnorm- ally large amount of visible rays compared with invisible rays ; that is to say, — they give a larger percentage of high frequency light rays compared with the low frequency invisible rays. Possibly and even probably some of these highly efficient fila- ments like the tungsten filament, also owe some of their high efficiency of one watt per candle power to selective radiation. When discussing selective radiation, we have first to come to an agreement on what we understand by selective LIGHT AND ILLUMINATION 243 radiation. The question whether an illuminant owes its high efficiency to selective radiation, depends largely on the defini- tion of the term "selective radiation". We have here a simi- lar case to that of the much discussed problem of the "counter electromotive force of the electric arc". Whether the electric arc has a counter e. m. f. or not, entirely depends on the defini- tion of counter e. m. f. In the same way, the decision on the question of selective radiation depends upon what you define as selective radiation. If you define as selective radiation any radiation in which the intensity of radiation is distributed through the total spectrum differently from that of the theoreti- cal black body, then the Welsbach mantle has selective radia- tion. If, however, you define selective radiation as the radia- tion of a body which gives spectrum lines, or bands, or absorp- tion lines and bands, that is, sharply defined narrow ranges in the spectrum, of abnormally high or abnormally low intensity, then the Welsbach mantle has no selective radiation. So all discussions and statements on selective radiation have rather little meaning, if the writer does not give his definition of selective radiation. In the following, I define as selective, any radiation which differs in the distribution of its intensity from the radiation of the theoretical black body. In an incandescent lamp filament we do not get a definite pitch, or definite frequency of vibration, but we get an infinite number of different waves. The reason is perhaps, that in a solid or liquid body the vibrating particles are so close together as to interfere with each other. If you could set a body in vibration, in which the vibrating particles, atoms or molecules, are so far apart as not to interfere with each other, as in a gas at low pressure, then they would execute their own periods of vibration, and then the light from such a body would not be a radiation of all wave lengths, but we would get radiations of 244 GENERAL LECTURES only a few definite wave lengths, or a line spectrum. So in- candescent or luminous sodium vapor gives only one kind of light, a yellow spectrum line, and in addition thereto a number of utrared and ultraviolet rays. Since the spectrum light is based on the non-interference of the vibrating particles, it is easy to understand, that when you bring the atoms or molecules closely together — as at atmospheric pressure — interference may begin, and the lines of the spectrum become more confused and blurr into bands. Therefore, we see in the mercury arc spectrum, which is at low vapor pressure, a small number of definite, sharply definite lines. In the calcium spectrum of the flame carbon arc, we get a large number of lines blurring into each other to an almost continuous spectrum ; so also in the white spectrum of the magnetite-titanium arc. When we set a gas or vapor in vibration, it vibrates at its own frequency, independent of the temperature, and it is merely a question of the character of the material whether a very large percentage of the total energy of radiation happens to be within the visible range or outside -the visible range. Temperature does not come in as factor, because the frequency of radiation is no longer a function of the temperature, but independent of the temperature. Sodium vapor gives the same frequency of radiation, the same yellow line when the sodium vapor is at low temperature or at high temperature. Some spectrum lines may increase in intensity with an increase of temperature faster than others, and the color of light may change with the temperature, change from yellow to white or blue, or from green toward white, and red, as the mercury light does with increasing temperature, but that is merely a characteristic feature of that particular body, and not a general character of the temperature effect; the possibility there- LIGHT AND ILLUMINATION 245 fore exists of finding materials which, when energized, as vapor or gas, give a spectrum with a large amount of energy in the visible range, thus giving an efficiency of light produc- tion far in excess of that available by incandescence. So far the only materials which give these characteristics are mercury, calcium and titanium. These three metals give spectra which contain such a large percentage of the total radiation in the visible range, that the amount of light meas- ured physiologically in candle power is far in excess of that which possibly could be produced by incandescence, even with the assistance of selective radiation. Their industrial appli- cations are represented by the mercury arc, the yellow flame carbon, and the white magnetite and titanium arc, and these are of such very high efficiency as to be of higher magnitude than any incandescent light. Even if we consider only these three illuminants, we have quite a color scale. From the orange-yellow of the flame carbon, which is about the longest wave length we could use, to yellow and yellow-white, in the acetylene flame and the tungsten filament. Then we have the greenish-yellow of the Wellsbach mantle, by selective radiation. We have the bluish- green in the mercury arc, and the yellowish-white of the car- bon arc, as well as the clear white of the titanium arc. Each of these can be modified. We can modify the titanium arc, giving all colors from yellow-white to bluish-white, by the addition of other materials which give either yellowish or bluish spectra. We can modify the yellow calcium arc, from the orange-yellow of calcium fluoride down to the yellowish- white of calcium borates. You can modify each color over a certain range, and you can get pretty nearly any color, with the exception perhaps of a clear blue and violet : no means have been found to get approximately the same efficiency in those 246 GENERAL LECTURES colors of very short wave length, as in the other colors of lights. This feature makes the effect of color which I discussed before, the variation of the physiological effect with the bril- liancy of illumination, of more importance now than years ago, when the only method of producing colored light was by the absorption of all other colors. III. After all, however, it is not light, that is wanted, but illumination ; it is not the amr.»unt of visible rays issuing from the source of light, the incandescent lamp or gas flame, which is of importance, but the amount of light which reaches the objects we desire to see, that is, the illumination produced by light. In this respect, I believe a mistake has been made by the gas industry as well as the electric lighting industry, for many years, by devoting all energy to the production of light, the development of the lamp, while they have almost entirely left out of consideration, that the production of an efficient light is not the only important problem, but that of the same import- ance is the arrangement of the light so as to get efficient illum- ination, that is, get the greatest benefit from the light produced ; and this feature has been usually left to the tender mercies of the architect or the decorator, who placed the lights where- ever he thought they would look artistic, regardless of the requirements of effective illumination. If you look around, you find cases everywhere of artificial illumination where the lights have been arranged, so that you get a very poor illumin- ation from a large amount of light. To overcome these defects, it is necessary to study the problems involved between the production of light and the physiological effect produced by the light upon the eye, and it requires a careful study, just as LIGHT AND ILLUMINATION 247 any other engineering problem. It is of importance to con- sider not only the amount of light issuing from the source, but the amount of light which reaches the object to be seen by the illumination. The demands of illumination are mainly of two classes, general illumination, and local, or concentrated illumination. Many cases require general illumination, as a meeting room, where it is desired to see equally well everywhere; that is, to get the same intensity of illumination throughout the whole illuminated area. So also a draughting room, a school room, the hall of a house and the streets of a city require general illumination, a uniform fairly high intensity in a draughting room or school room, a relatively low but as nearly as possible a uniform intensity in the streets of a city. It is true, street lighting is usually very far from uniform, but that merely means that the problem of proper street lighting is usually not solved in the most efficient and satisfactory manner. In other cases concentrated lighting is required, as in domestic lighting, in the dining room, the living room, etc, where light is desired on the table where we work, eat, read, etc. In such cases, the general illumination of the room is of lesser importance ; it is not needed to any extent, or is frequently undesirable, because a room with a very low intensity of general illumination fre- quently is considered more homelike, especially by the feminine part of the human race. In still other places general illumina- tion may be directly objectionable, as in a sick room. Most cases, however, require a general illumination of moderate intensity, and a far more intense local illumination, as over the desks in an office, or the reading tables in a library. In such cases merely a general illumination would be sufficient, if very intense, but this is uneconomical and to some extent objectionable on account of the blinding glare, which is disa- 248 GENERAL LECTURES greeable; and so a combined general and local illumination is more efficient and more satisfactory. In producing illumination either direct lighting or indi- rect lighting may be used. That is, the rays issuing from the source of light may either pass directly to the illuminated objects, or they may pass to a reflecting surface, and be reflected from this surface to the object, or may pass through a refracting body, as the frosted incandescent lamp globe, or opal globe of the arc lamp, and so reach the illuminated object. In general, it is obvious that any method of indirect lighting by refraction or reflection wastes a considerable amount of light. That means, the total amount of light which reaches the illuminated object must necessarily be less with indirect light- ing, as compared with direct lighting, with the same amount of light. Indirect lighting can be done by reflection or refraction by some attachment to the lamp, as a reflector or a holophane or frosted globe, or by reflecting the light from the ceilings and walls of the room, on the objects to be illuminated. In the latter case, it is obvious that white walls give the highest efficiency of reflected light. It is easy to observe that the same source of light in a room with white walls gives several times the intensity of illumination which it gives in a room with black or non-reflecting walls. That means that the total amount of illumination is increased several-fold by reflection from white walls. So in a draughting room, or school room, by using as light walls as possible, we get the best efficiency of illumination. It is not always feasible to have light walls, especially when you come to machine shops or foundries, and other places where the walls do not remain white, but change to some darker color. The question is, what color do these walls LIGHT AND ILLUMINATION 249 assume? The color of almost everything which is changed by age is due to either iron or carbon. In most cases of discolora- tion by age you see the reddish-brown color of iron and the brownish-yellow color of carbon. This is the color most sub- jects gradually assume. This color of age is in the long wave or low frequency end of the spectrum. To get the benefit of reflected light from walls which cannot be kept perfectly white, a source of light rich in the long low frequency waves, or of a yellowish tinge is therefore more efficient by giving more reflection from the walls than a source of light rich in short or high frequency waves, that is, bluish-white. This effect is very marked when you compare the mercury lamp with the flame carbon lamp. The illumination given by the mercury lamp in a draughting room is very satisfactory. The same illumination in a foundry or machine shop is far less satisfac- tory, and you notice there a marked absence of reflected light, that is, the walls and ceilings all gradually assume a color which is rich in red and yellow, and so reflect very little light of the violet end of the spectrum. Even the black- begrimed walls of a blacksmith shop reflect a considerable amount of light with an orange-yellow source of light; prac- tically none with the bluish-green of the mercury lamp. Thus the shade of color of the illuminant may be very essential in getting efficient illumination. In the interior of a city, the walls usually have a reddish-yellow color. In that case white or yellowish lights are superior. When outside of a city, the greenish-yellow of the Welsbach lamp, the bluish- green of the mercury arc, gives a much larger amount of reflected light from vegetation than the yellow of the incandes- cent light and so a better illumination. Vegetation absorbs the long waves, the low frequency radiation; so with a yellow source of light there is practically no reflection from 250 GENERAL LECTURES living vegetation, but only reflection from dead vegetation, and in the light of the incandescent lamp or the flame arc all vegetation appears very poorly, the dead parts are very promi- nent, while the reverse is the case where the light is deficient in the red and yellow, and rich in the green and blue, as with the mercury arc : the green shows prominently, while the dead leaves, etc., are not visible. IV. It is, however, not the amount of light which reaches the illuminated objects which is of importance, that is, not the physical intensity of illumination, but the amount of light which from these illuminated objects reaches the human eye. With the same intensity of illumination, the same amount of light reaching the illuminated object, of the same color, the amount of light entering the eye may vary widely with the opening or contraction of the pupil of the eye. The eye is automatically adjusting for intensity of light. This is the reason we see well at sun light and at the light of the full moon, although the former is many thousand times greater in inten- sity than the latter. The eye can accommodate itself to intens- ities varying over an enormous range. It does this partly by the fatigue of the nerves of vision, partly by the contraction or opening of the pupil. This is undoubtedly a protective device developed in the human race. It means that if we have in the field of vision a source of light of high intrinsic bril- liancy, the eye protects itself by contraction of the pupil and so it receives very much less light in the field of vision where we want to see objects, than if the source of light were taken out of the field of vision. By eliminating the source of light from the field of vision and eliminating the contraction of the pupils resulting from the high intrinsic brilliancy of the illuminating LIGHT AND ILLUMINATION 251 body, we get actually a much larger amount of light into the eye with the same amount of light striking the illuminated object ; that is, we get a higher physiological efficiency. Even with a much smaller amount of light reaching the illuminated objects, we still get more light in the eye. That means if we reduce the intrinsic brilliancy of the illuminant by indirect lighting, by diffusing the light, we may lose a considerable amount of light, actually get a considerably reduced quantity of light on the object which we desire to see, but still we get a larger amount of light from these objects into the eye, because the eye is open further and admits more light and is less fatigued. It follows from this that in efficient illumination, it is of foremost importance to arrange the illuminants so as not to have excessive intrinsic brilliancies in the field of vision when looking at the objects we desire to see. That means that the proper field for the illuminant is outside of the field of vision, or where you cannot get it out of the field of vision, that its intrinsic brilliancy should be reduced by diffusion : thereby we actually get a much higher physiological effect. This is the reason for indirect lighting. We may have a very large amount of light thrown on any object in a room, but if the eye is fatigued by seeing the source of light in the field of vision, we get very little light in the eye, while with a properly arranged indirect lighting, with a much lower amount of light reaching the object, we get a higher physiological effect, that is a better and more efficient illumination. It appears, however, that this automatic protective faculty of the eye was developed through the ages as a protection not against light, but against energy; apparently the eye is pro- tecting against the energy of radiation, not the physiological intensity, and since the energy of radiation is mainly in the asa GENERAL LECTURES ultrared, in the long waves, the frequency which causes the protective reaction is the frequency of the long wave end of the spectrum, the red and yellow waves; they make the pupil contract. This action is much less for the green and blue rays. That is the reason the eye does not react on the mercury lamp to any great extent. It means a green light, like the mercury or Welsbach lamp, can be in the field of vision to a much greater extent without causing the contraction of the pupil and so reducing the physiological effect. This is of importance in places where the light cannot well be taken out of the field of vision, as in street illumination. In this case, all the sources of light must be arranged along the street and so must be in the field of vision. By cutting off the red end of the spectrum you eliminate the contraction of the pupil, and get the full benefit of the light between the illuminants, while with a yellow source of light, as with the incandescent or arc lamp of old, you do not get the benefit, that is, the physiological effect of the illumination by a green illuminant in such cases is superior to that by a yellow illuminant, the illumination appears brighter and more uniform. A light devoid of red and yellow rays is at the same time the safest and most harmless, and also the most harmful. It is the safest and most harmless, and gives the most uniform illumination, if its intrinsic brilliancy is sufficiently low to be below the danger limit of energy of radiation, but it is harmful if above that, because the eye does not protect itself against it, probably because these lights have not existed throughout all the ages when this protective action of the eye was devel- oped, and sunlight and fire were the only sources of light, both rich in red rays. This accounts for the rather contradictory effect observed, that green or blue light, as the Welsbach mantle or mercury lamp, is a very good light to work by, LIGHT AND ILLUMINATION 253 superior to the yellow kerosene lamp, and at the same time there is some suspicion that it is harmful to the eye. It may well be that where it is of very high intensity, the automatic protection of the eye is not sufficient to protect with such light. Where you use such sources of light you can get the benefit of the absence of the contraction of the pupil, but it devolves upon you to arrange the illumination so as not to get the harm- ful effects against which the automatic protection of the eye fails. That means all these lights are superior for illumination if they have low intrinsic brilliancies, but somewhat question- able if they have extremely high intensities. V. It is, however, not even the amount of the light which enters the eye which is of importance in illumination, but the difference in the amount of light. If in the illuminated area the light were of uniform intensity, and everything of the same color, we would see nothing but a glare of light. The seeing takes place by a difference in color, and difference in intensity. Difference in intensity includes shadows. Shadows are thus an essential feature in seeing things. Considering, then, the seeing by shadows and seeing by color differences, you observe that by this feature we can divide illumination into directed and diffused illumination. In diffused illumination light comes in all directions with approximate uniformity, and shadows do not exist ; in directed illumination, shadows exist. In some cases shadows are objectionable, and in other cases shadows are necessary for clear distinction, and diffused illumination in such cases would not be satisfactory. As regard to seeing differences in color, it is obvious that where definite color distinctions are required, you can intensify 254 GENERAL LECTURES the sharpness of vision by selecting the color of your light best suited to bringing out the colors desired. Where the color conditions you want to distinguish are those due to age, iron and carbon, then the light which is deficient in red and yellow, which therefore shows the colors given by iron and carbon, as black, gives a much sharper distinction, and the mercury lamp shows blemishes and dirt much more pronounced than the white light. Again, the sources of light which are very rich in red and yellow rays show these colors due to iron and car- bon very much less, and therefore show blemishes or a slight amount of dirt much less, soften them; and where the color distinctions are those due to these two most prominent ele- ments, in the yellow light their appearance will be greatly softened, and under the green light they will be made harsh and sharp. If you desire to soften effects, as in a ballroom, it would be fiendish to use mercury lamps, but where you want to search out a spot that is soiled, it would be very wrong to use a dull, yellow incandescent lamp or a gas flame, but rather to use the green Welsbach light, or better still the bluish- green mercury arc, which gives in such case sharp distinction, where white light shows little, and yellow light nothing. Where you desire to see all colors in about the same relation as by daylight, you obviously desire a white light. It is therefore important for the illuminating engineer to select the shade or color of the light and study the require- ments of each case which comes into his charge. It would be just as wrong in one case to use an incandescent lamp, where the mercury lamp would be better, as to do the reverse. We have to distinguish then between general illumina- tion and local illumination, between direct illumination and indirect illumination, and between directed illumination and diffused illumination. These three different classes or distinc- LIGHT AND ILLUMINATION 255 tion to a certain extent overlap. It would be very wrong, how- ever, to mistake them, and a very serious mistake in the design of a system of illumination can very easily be made; for instance, by mistaking general illumination and diffused illumination for each other. The problem may be to get uni- form intensity all over. You can get that by distributing a large number of small units all around the cornices and reflect the light from white walls and ceilings and get a very diffused illumination, or you could get general illumination, where the intensity of illumination all over is the same, in a moderate sized room, from one source of light by using one of these sources of light as an incandescent lamp with a holophane reflector, which gives the proper distribution, or you could get light from any other source by controlling the distribu- tion curve of the light so as to get uniform distribution. The former arrangement gives diffused light, the latter directed light. You may get the same intensity of illumination all over the room, in both cases; but in the former case no shadows, in the latter case absolutely black shadows. Probably in the former case for domestic use the lighting will be unsatisfac- tory and trying to the eyes, because you do not see well, you do not have any shadows, objects around you are not so dis- tinct, because you lose the distinguishing feature of the shadow. In the latter case with the directed lighting from one source, the lighting will be unsatisfactory because you get very dark shadows, and you do not see anything in the shadows, and the eyes will be made tired by trying to see in the very dark shadows. You have to consider how much directed light and how much diffused light you require. In some cases you may desire only diffused light. In the general lighting of a draughting room you do not want any directed light, since 256 GENERAL LECTURES you must have no shadows, because if the ruler casts a shadow, it is trying to the eyes to distinguish between the edge of the ruler and the edge of the shadow, and mistakes may be made. In this case, you see only by differences in color and in the intensity, and not by shadows. You therefore get satis- factory illumination from many small units, or by indirect lighting, reflected light from white walls and ceilings, but you get unsatisfactory illumination from a few units even when properly distributed so as to give uniform intensity all over, but giving little reflected light. In other cases, you may also require a general illumination equal in intensity all over, but you need directed illumination so as to see by the shadows. So for instance, a good draughting room illumination would not be suitable for a foundry. In a foundry, where all the objects assume more or less the same color, you require shadows to see by. Then you need a number of units of light to give directed illumination, but you must not go so far as to be unable to see in the shadows ; you must have some diffusion, or overlapping of the different beams of light. So if you take a satisfactory foundry illumination and put it in the draughting room, even if the intensity were satisfactory, it would be entirely unsatisfactory, and so would be the reverse. It is therefore not merely the distribution of the intensity of the light, which is essential, but also the character, whether diffused or directed light, or how divided between diffused and directed light. In the different lighting problems you therefore meet the question of concentrated and general illumination, of directed and diffused illumination. In domestic lighting, by reflected light from white walls and ceilings we can get a high intensity, and can increase the illumination several-fold over that given directly from the source of light, such as the incandescent LIGHT AND ILLUMINATION 257 lamp or gas flame. Still the illumination would be unsatis- factory and tiring to the eyes. We all know that in the home a room with white walls is not as agreeable as one with darker walls. We say we have too much light. But we do not have too much light, because we do not have anywhere near the same amount of light as we get during the daytime out of doors. We have too large a percentage of diffused light. The intensity of diffused light is too great as compared with the directed light. We lose the shadows and that is tiring to the eyes. The problem of domestic lighting then is, to get sufficient directed and not too much diffused light- ing so as to get the best vision, that is, to get sufficient shadows to see by, but the shadows must not be so dark as to make seeing objects in the shadows tiring to the eyes. During the daytime we get directed light from the win- dows, diffused light reflected from the walls. To get the proper proportion between directed and diffused light, fixes the shade of the walls, and in general we have to use walls of somewhat darker color. When you come to lighting in the evening, with a source of light like the incandescent lamp or gas lamp, sending out light in all directions, the diffused light compared with the concentrated or directed light is a higher percentage than in the daytime for the same color of walls, partly due to the color of the light, which is yellow, and is more reflected from the walls, largely, however, because with the daylight through the window the directed light is a much larger percentage of the total light than in the lamp, where only a small part is concentrated light. It is not comfortable to have this strong diffused light, and so we put shades on which absorb three-quarters of the light, but which give us a more comfortable illumination in the room. That means waste, however, and you pay for light which you do not use. 258 GENERAL LECTURES The proper illuminating engineering then is to secure the cor- rect distribution curve of the source of light, so as to give the desired amount of concentrated lighting on the dining or read- ing table, and give only as much diffused lighting as is com- patible with the amount of direct light used, to see in the shadows. The problem of domestic lighting, from the illum- inating engineering point, is to determine the illumination over the entire area, and also the character of illumination, whether directed or diffused; how large an amount of light should be concentrated, and how large an amount should be directed ; then the question of colors and shades also comes in as an important factor, as was discussed before. Practically nothing has yet been done in this direction systematically and intelligently, but all has been done by trial which at the best usually means producing more light than necessary, and throw- ing away the excess of diffused light by absorption. APPENDIX II LIGHTNING AND LIGHTNING PROTECTION Paper read before the Annual Convention of the National Electric Light Association, 1907. Revised to date. L LIGHTNING PHENOMENA IN THE CLOUDS. /n^ HE first man who attacked the problem of lightning and I lightning protection, a century and half ago, was our -^ great citizen, Benjamin Franklin. He gave us the lightning rod, which is now universally recognized as the most effective and only protective device for isolated points, as steeples, chimneys, etc. The next step in advance was made by Faraday : he showed that in the interior of a perfectly con- ducting body no electric disturbances can be produced by out- side electric forces. This led to the most effective protection possible against lightning or electric disturbances, the use of a grounded metal cage, "Faraday's cage", enclosing the struc- ture which is to be protected, whether a building against lightning, or a delicate instrument against electric fields. In its simplest form, Faraday's cage, applied to a trans- mission line, is the ground wire above the line, and the pro- tection afforded by it is the more complete, the more the over- head ground wires represent the condition of an enclosing cage of perfect conductivity. That is, a system of wires above and on the sides of a transmission line is superior to a single wire, a wire of high conductivity superior to a small iron wire. Here I specially desire to draw attention to the second requirement of the Faraday cage, high conductivity. Thus it is not sufficient merely to have any kind of overhead 26o GENERAL LECTURES grounded wire regardless how small, but high conductivity of the grounded conductor is essential in many cases of atmospheric disturbances. In the last ten years, transmission voltages have crept higher and higher, transformers have been built of consider- able size, of still higher voltages, so that exact data on the action of voltages up to 300,000 are now available, and approximate data for potentials above a million volts. It was found that air has a definite and fixed breakdown strength, that is, just as a beam breaks mechanically as soon as the stress in it exceeds a definite value, the breaking strength of the material, so air breaks down by a disruptive spark, as soon as the electric stress in the air, or the potential gradient, exceeds a certain value, which is about 100,000 volts per inch. The disruptive strength of air is, over a wide range, propor- tional to the pressure, that is, at a pressure of two atmospheres it is twice as high, or 200,000 volts per inch; at one-quarter atmosphere it is one-quarter, or 25,000 volts per inch.* The striking distance in air between needle points has been investigated up to 300,000 volts, and found that for high voltages it is very nearly 10,000 volts per inch, that is, a dis- charge of 30" length between needle points requires 300,000 volts. If between two needle points the potential difference is gradually increased, already at relatively low voltages the dis- ruptive strength of the air at the needle points is exceeded, the air at the points breaks down and becomes conducting, and luminous, as "brush discharge", so that the terminals are not the needle points any more, but the whole space, of approxi- mately spherical shape, which is covered by the brush dis- charge. As result thereof, for high voltage, no appreciable * Only at very low pressures, where the distance between air molecules become appreciable, this law ceases, and the disruptive strength increases again, and seems to become infinitely great in a perfect vacuum. LIGHTNING AND LIGHTNING PROTECTION 261 difference exists in the striking distance between needle points and between spheres, the centers of which approximately coincide with the needle points, as long as the diameter of the spheres is small compared with their distance apart apart. With increasing potential difference between needle points, the brush discharges spread out against each other, until only about 40% of the space between the needle points is free, and then a dis- ruptive spark passes. Naturally, as soon as determinations of spark voltages became available, attempts were made to estimate the voltage of a lightning flash. Considering, in a lightning flash, the dis- charge as that in an ununiform field, similar to that between needle points, and so requiring about 10,000 volts per inch. In this case, a lightning flash of two miles, or about 10,000 feet length, would require a potential difference of about 1200 million volts. The existence of such voltages in the clouds does not appear possible: a potential difference of 1000 mil- lion volts would produce a brush discharge of about one-half mile in length, before the final lightning flash occurs. In the brush discharge the air is electrically broken down, and becomes conducting. But it is also mechanically and chemically broken down, that is, the molecules are dissociated and recombine after the discharge, in all possible combinations. That is, we get ozone and nitric acid, and a lightning flash produced by a thousand million volts would thus be followed by a deluge of nitric acid. This fortunately is not the case. An estimate of the voltage and the current in a lightning flash would not yet give the energy, if the duration of the dis- charge is not also known. We can, however, get an approxi- mate estimate of the magnitude of the energy of the lightning flash indirectly, from photometric considerations, and elimi- nate the consideration of the duration of the flash by -the TynOPERTY Of ELECiniCAL LABOHAlOKY, j FACULTY OF A^-fLlE* SCIENCE. j 262 GENERAL LECTURES integrating feature of the human eye for impressions of very short duration: an impression on the human eye persists for some time, about .i seconds, and any phenomenon of shorter duration than . i seconds so appears to last . i seconds. Hence the effect on the eye by a lightning flash would be about the same whether the flash lasted .i seconds, or if it were of a thousand times greater intensity but lasting a thousandth of the time. This means that the eye would see a lightning flash about in the same manner as if its light, and so probably its energy were spread uniformly over .i seconds. The illumination given by a brilliant lightning flash is about of the same magnitude as good artificial illumination, perhaps one foot candle, since at night time in a well lighted room, the light of a lightning flash is still quite appreciable. Estimating roughly one watt per candle foot, a lightning flash illuminating a space of two miles square or lo' square feet, with one foot candle would consume lo* watts, and as this is the illumination as averaged, by the human eye over .i seconds, the energy is lo' watt-seconds, or 10,000 K. W. seconds. The energy of a large lightning flash, estimated from its light, would thus be of the magnitude of 10,000 K. W. seconds. This value, while considerable when expressed in electric quanti- ties, is by no means so very great: reduced to heat measure, it only equals the latent heat of evaporation or condensation of about 9 lbs. of water. As seen, an estimation of the voltage of the lightning flash from length and disruptive potential gradient of the air, does not give reasonable values, that is, the lightning flash cannot be a single discharge as that of a Leyden jar. An estimation of the voltage may then be attempted in a differ- ent manner. LIGHTNING AND LIGHTNING PROTECTION 263 Lightning flashes usually occur within thunder clouds and only rarely from cloud to cloud or from cloud to ground. They therefore seem to be rather due to equalization of potential differences within the cloud, than to dicharges between oppositely charged bodies. Lightning occurs mainly when rapid condensation of moisture takes place in the air and the electric phenomena seem to be the more intense, the greater the rapidity of condensation, or rain formation. Thus the atmospheric electric disturbances seem to be connected with the condensation of water vapor to clouds and rain. There exists normally a potential gradient in the air. That is, a potential difference exists between the air at different elevations, reaching sometimes several hundred volts per foot, so that we can estimate as a fair average, a natural potential gradient in the air, in vertical direction, of about 100 volts per foot. A point 100 feet above ground may show a potential difference of about 10,000 volts against ground. Usually the higher strata of the air are positive against the lower. The cause of this potential gradient, whether terrestrial or cosmic, is of no interest to us here, but merely its existence. It is of interest to investigate, what effect must be expected, from our well-known physical laws, from the con- densation of moisture, and agglomeration of the moisture particles to rain drops, in an atmosphere having such a poten- tial gradient. Assuming water vapor in a higher stratum of the atmosphere to condense to moisture particles, these moisture particles have the potential of the air in which they float, that is, have a considerable potential difference, perhaps hundred thousands of volts, against ground, and so contain an electric charge against ground. These moisture particles conglomer- ate with each other to larger moisture particles and ultimately 264 GENERAL LECTURES rain drops. By the collection of n* particles into one, the diameter of the particle has increased n fold. Its capacity has also increased n fold (the capacity of a sphere being pro- portional to the diameter). The particle contains, however, the accumulated charges of n* smaller particles, and n' times the charge, with n times the capacity, gives ii^ times the poten- tial. It follows herefrom that with the conglomeration of the water particles, their potential must increase rapidly, propor- tionately to the square of their diameter. The conglomeration of moisture particles in the clouds is, however, very uneven, due to the uneven distribution of moisture, as is plainly seen by looking at any cloud : dense or dark parts representing consid- erable condensation and so considerable moisture content, alternate with light parts, in which little or no condensation occurs. As a result thereof, starting with a uniform potential in the stratum of the air, where condensation begins, differ- ences of potential distribution by necessity result from the differences in the condensation of water vapor to moisture and the accumulation of the moisture particles to larger ones, that is, the denser portions of the cloud are at a higher potential than the lighter portions. Thus, starting with uniform poten- tial, and thus zero potential gradient in the air at the moment of the beginning of condensation, potential differences and thus potential gradients appear. Such potential differences in the clouds increase with increasing agglomeration of moisture particles to rain drops, and so the potential gradient rises. Assuming even as low a potential gradient as lOO volts per foot in the cloud at the beginning of agglomeration of moisture particles, the collec- tion of n* such particles to one rain drop of n times the diameter and so n times the capacity, but containing the static charge of n' particles, gives n^ times the potential, and since the dis- LIGHTNING AND LIGHTNING PROTECTION 265 tances between the particles are now n times as large, the potential gradient has increased n fold. That is, by conglom- eration of water particles, the potential gradient rises propor- tionately to the diameter of the particles. Estimating then the average diameter of moisture particles as lO"* inches at the be- ginning of agglomeration, when the potential gradient in the cloud is about 100 volts per foot, then the breakdown potential of the air, of between 100,000 and 200,000 volts per foot, would be reached when the drops have reached about .1 to .2 inches diameter, that is, the size of rain drops. Potential gradients in the cloud thus gradually rise, until somewhere the disruptive strength of the air is reached, and a discharge passes, equalizing the voltage at this spot. This, however, causes a greater potential gradient at the ends of the discharge, exceeding the breakdown strength of the air, and so causes a second discharge, following partly over the path of the first, then a third and so on, until all of the potential differences or inequalities of the potential distribution in the cloud, are leveled down by a series of successive discharges. The phenomenon thus is similar to that of a landslide, setting off another and another landslide. Or it can best be pictured by representing the unequal moisture distribution in the cloud by a relief map built of wet sand, the dense portion of the cloud, and therefore the portions of high potential, being repre- sented by the hills, the light or low potential portions of the cloud by the valleys of the relief map. As soon as the sand dries, somewhere, where the declivity is very steep, that is, the potential gradient is very high, a slide occurs, this causes another slide and so on, until the whole configuration of sand settles down to a flat and smooth shape, the hills are leveled off and the valleys filled. 266 GENERAL LECTURES The existence of such successive discharges, following each other after appreciable intervals of time in the same path, has been shown by the photographs of lightning flashes taken with a rotating camera. In this case, by the motion of the camera the successive flashes are recorded side by side, and sometimes more than forty successive discharges have been counted, the whole phenomenon lasting about .6 seconds, that is, quite an appreciable time. Oscillograms of lightning discharges from (dead) trans- mission lines also showed the frequent occurrence of multiple strokes, or strokes following each other within a fraction of a second. It follows herefrom, that lightning flashes in the clouds, of several miles' length, occur without any considerable poten- tial difference between the ends of the flash, but result from the disruptive equalization of the unequal potential distribu- tion in the clouds, caused by unequal vapor density and so unequal condensation and conglomeration of moisture particles. This also explains the relatively small tendency to dis- charges between cloud and ground, across a space in which no condensation takes place and so no unequal potential distri- bution supplies the power of the discharge : although the dis- tance between cloud and ground is smaller than the distance traversed by a lightning flash in the clouds, and the average potential differince between cloud and ground probably is greater than the potential differences in the clouds, a discharge to ground probably occurs in general only where by a heavy downpour of rain a range of high potential is carried bodily part ways down to ground. This also may explain, that light- ning discharges to the ground are usually followed by a heavy downpour of rain. LIGHTNING AND LIGHTNING PROTECTION 267 The potential gradient in the air may rise to disruptive values in still another, slightly different manner, and lead to lightning discharges without being accompanied or followed by rain. By conglomeration of moisture particles the potential gradient rises, as described above, but before the water drops have reached sufficient size to precipitate as rain, evaporation again sets in : for instance by the drops falling to a lower and warmer stratum of the air, or by intercepting the heat of the sun's rays, and the drops thus dwindle away. The decrease in size of the drops represents a decrease of capacity, the capacity being proportional to the diameter, and as each drop retains the same charge, its potential increases with the decrease of size, without limit, and so also the potential gradient until its disruptive value is reached and the lightning discharge occurs. This phenomenon is frequently observed towards the evening of a hot summer day, and is called "heat lightning", and, being the result of evaporation, thus does not lead to rain. Estimating then as disruptive strength of air under dis- charge conditions in a non-uniform field, and at the reduced air pressure in the clouds, 100,000 volts per foot, the average potential gradient in the path of the lightning discharge through the clouds would be about 50,000 volts per foot. This gradient, however, would not be unidirectional, but the poten- tial would rise from a low, or even negative value at a light portion of the cloud, to a maximum value af i dense position, then decrease again, that is, give a gradient in opposite direc- tion, to a light position, etc., and the potential gradient would vary from nothing at a maximum potential point, to a maxi- mum, equal to the breakdown strength of air at the starting point of the discharge, to zero at a minimum potential point, etc 268 GENERAL LECTURES To estimate the current which discharges in the lightning flash, the conductivity of air in the path of the discharge, and the diameter of the discharge are required, and as both are unknown, any estimate must be very approximate only. The specific resistance of gases and vapors decreases with increasing temperature and with decreasing pressure. It is a few ohm centimeters at atmospheric pressure and the high temperature of the magnetite or carbon arc, and is also a few ohm centimeters at the low temperature and low pressure of a high current Geissler tube discharge. The mercury arc stream also gives a specific resistance of a few ohms. The temperature of the air in the lightning discharge probably is moderately high, but the pressure is also not far from atmos- pheric, so that lOO ohm centimeters may not be very far from the true magnitude of the resistance. Estimating one to two feet as the diameter of the discharge path, and lOO ohm centi- meters as the specific resistance, and allowing for the induct- ance, gives, with an average potential gradient of 50,000 volts per foot, a current of about 10,000 amperes. The heating effect and the magnetic effect of lightning strokes also point to the passage of currents of some thousand amperes. Assuming then the average potential gradient in the light- ning flash as 50,000 volts per foot, the current as 10,000 am- peres, a lightning flash of two miles' length would represent a power of 5 X 10* K. W. Estimating the energy of the discharge, as approximated from the photometric consideration, as 10,000 K. W. seconds, the duration of the discharge would be: 10V5 X 10* = 2 x 10"* sec, or two-millionths of a second. The discharge probably is oscillatory. In view of the high resistance of the discharge path, the damping effect must be very great, that is, a very large part or nearly all the energy LIGHTNING AND LIGHTNING PROTECTION 269 is expended in the first half-wave ; that is, the discharge consists of only one or very few half- waves. With a duration of the discharge of 2 x 10*' seconds, assuming two half- waves as an average, gives 500,000 cycles. The frequency of oscillation of the lightning flash thus appears to be of the magnitude of half a million cycles. Since the velocity of propagation of electric disturbances is the velocity of light, or 188,000 miles per second, the wave length of a discharge of 500,000 cycles is ^qq'qqq = g miles, or about 2000 feet. A wave length of 2000 feet means that the current in the discharge flows in one direction for 1000 feet, in the opposite direction, that is, with opposite potential gradient, in the next thousand feet, etc. That is, in our former discussion, the average distance through which the potential gradient has the same direction, or the distance between maximum and mini- mum, between densest and lightest parts of the cloud is about 1000 feet. This agrees fairly well with the appearance of the clouds to the eye, and it also agrees in magnitude with the dis- tance over which the wind velocity varies, in gusts, as shown by Prof. Langley in his investigation on the "internal energy of the wind". It appears herefrom, that the varying wind velocity as measured by Prof. Langley, that is, the gusty character of the air currents, results not only in an internal mechanical energy, which the bird utilizes for soaring, but also results in unequal moisture distribution, and so, when condensation occurs, in an "internal electrostatic energy" of the thunder cloud, which dis- charges as lightning. With an average length of the half -wave of 1000 feet, and 50,000 volts per foot as potential gradient, the potential 270 GENERAL LECTURES differences in the clouds would be of the magnitude of fifty million volts. These are values which appear reasonable. Assuming that a lightning flash drains the electric energy of the cloud within a radius of about lOO to 200 feet from the path of the discharge, this affords a different method of esti- mating the magnitude of the energy of the lightning flash: assuming for instance saturated air at 40° C mixing with air at o°C, condensation of a part of the moisture occurs which can easily be calculated. Assuming that this moisture has conglomerated to rain drops of .1" to .2" diameter, the num- ber of such drops in a space of two miles' length, and 200 to 400 feet diameter, can be calculated, and also their electro- static capacity. With a wave length of 2000 feet, and a potential gradient of 50,000 volts per foot, from the capacity follows the energy of the electrostatic charge, which dis- charges as lightning flash. This is found under the above assumption, as of the magnitude of 10,000 K. W. seconds, so agrees with the results derived from the photometric considera- tions. To conclude then, as approximate values of magnitude of the electric quantities in a lightning flash may be estimated : Average potential gradient : 50,000 volts per foot at the moment of discharge. Average potential difference between different points of the cloud : 50 million volts. Average current in the discharge 10,000 amperes. Average duration of the discharge ^q^qqq sec. Average frequency of discharge : 500,000 cycles. Average energy of the discharge: 10,000 K. W. sec., or seven million foot pounds. LIGHTNING AND LIGHTNING PROTECTION 271 II. LIGHTNING IN ELECTRIC CIRCUITS. Of greatest importance to an electrical engineer are the high potential phenomena produced in electric circuits by atmospheric lightning as well as by other causes, frequently internal to (the circuit, which give the same or similar effects to such an extent, that it has become customary when dealing with electric circuits, to distinguish between external or atmospheric lightning, and internal lightning, as caused by electric circuit disturbances or defects, such as sudden changes of load, or arcing grounds, etc. While a very large amount of data on high potential phenomena in electric circuits has accumulated, the possible variety of phenomena is so g^eat that an intelligent under- standing of the phenomena, as is required for effective pro- tection of the circuits, is feasible only by a theoretical investi- gation of the high potential phenomena which may be expected in electric circuits, and a comparison thereof with the observed effects. In general, the high potential phenomena possible in electric circuits are the same three classes of phenomena which can occur in any medium, as a body of water, which is the seat of energy. 1. Steady electrostatic stress, that is, a gradual rise of potential of the total circuit against ground, until a discharge occurs somewhere ; just as in a body of water, as a river, the pressure, that is, the water level, may gradually rise, until it breaks through the embankment. 2. Impulses, or traveling waves, similar to the ocean waves rolling over the surface of the water. 3. Standing waves, or oscillations or surges, similar to the oscillation of a tuning fork, or a violin string. 272 GENERAL LECTURES A more extended discussion on the three forms of electric disturbances, and their causes, is given in a paper read before theA.LE.E.* Steady electrostatic stress obviously can occur only where the circuit is very well insulated from the ground, but not in a grounded circuit, or a leaky circuit, as low voltage circuits usually are, and such static stresses can be eliminated by a permanent leak, that is, a high resistance connection between the circuit and the ground. As sources of impulses or traveling waves only two characteristic phenomena may be considered here: the light- ning flash, or induction by the clouds, as external, and the arc- ing ground as internal cause. Assuming a thunder cloud to pass over the line. The ground below the cloud then assumes an electrostatic charge, corresponding to the opposite charge of the cloud. The trans- mission line, as part of the ground, thus also assumes a static charge, higher than that of the ground, since it projects above it. Any equalization of the potential distribution in the cloud by a lightning flash, as discussed in the preceding, requires a change in the electrostatic charge of the line, corresponding to the changed potential difference between ground and cloud above the ground, and the static charge thus set free on the line rushes as an impulse or wave along the line. The wave shape of such impulses induced by cloud discharges is in general not a smooth sine wave, but may be very irregular : during the equalization of the cloud potential by the lightning flash, the potential difference against ground, of the part of the cloud above the electric circuit, may vary in almost any conceivable manner, thus giving rise (to very different wave shapes of the impulses. So some impulses may rise very rapidly, with •A. I. E. E. Transact. March, 1907: "Lightning Phenomena in Electric Circuits." LIGHTNING AND LIGHTNING PROTECTION 273 extremely steep wave front, and slowly die down. Others may rise slowly, then suddenly fall and reverse, or a series of oscillations may occur in the impulse, etc. If the lightning flash is parallel with the line, simultaneous impulses of different directions may be produced, corresponding to the different directions of the potential gradient in the different parts of the lightning flash, and these waves, of different directions, intensity and wave length, traveling over each other, then pro- duce a very complex system of phenomena. So for instance, by the interference of two impulses of nearly equal wave length, moving in opposite directions, a high voltage point may be produced, traveling slowly along the line, and visible to the eye as a luminous streak. The frequencies of these impulses then are those corres- ponding to the frequencies of cloud discharge, that is, of the magnitude of hundred thousands of cycles per second. With the velocity of light, 188,000 miles per second, they travel along the line, until they gradually fade out by the dissipation of their energy, or are reflected at an open end of the line, or at the entrance to the station are broken up by partial reflection, in reactances, and interference between the reflected waves, the incoming waves and the waves passing over the reactances, and so give rise to systems of standing waves or oscillations, similarly as an ocean wave rolling on to a sloping beach breaks up into surf. Where a traveling wave is reflected, the combination of the reflected wave and the incoming wave produces a standing wave or oscillation, that is, a wave in which the voltage maxi- ma and the zero points or nodes have fixed positions on the line. By superposition of the wave maxima of incoming and reflected wave, the standing wave rises to a maximum double 274 GENERAL LECTURES that of the traveling wave. Where different oscillations or standing waves superimpose upon each other, their maxima substract at some places and add at others, and thus again double the voltage, that is, a traveling wave or impulse, break- ing up into systems of oscillations at a station, doubles and quadruples the potential; so that a traveling wave of moder- ate potential may cause dangerous voltages when breaking up into oscillations, just as in the ocean surf, the waves rise to far greater heights than in the on-rolling ocean wave before it reaches the beach. If we consider that the impulses traveling along the line are not sine waves, but of very irregular shape, that is, can be considered as consisting of a fundamental of some hundred thousand cycles, and numerous higher harmonics of still greater frequency, and each of the components when breaking up at the station gives rise to a set of oscillations at every inter- ference point, that is, at every reactance, the complexity of the phenomenon can be imagined. Since the equalization of cloud potential usually occurs by a series of successive discharges in short intervals, a small fraction of a second, and each discharge gives rise to an impulse in the line, and so a system of oscillations at the sta- tion, whatever protective device is used, must restore itself instantly after a discharge, so as to receive the next following discharge. Any device depending on mechanical motion to restore itself after a discharge to operative position, therefore fails to protect, when a series of discharges follow each other in very rapid succession, as discussed above. Traveling waves very similar in character to those due to induction from the clouds, but frequently of far greater volume, sometimes occur in an electric circuit from internal causes, as arcing grounds, or spark discharges. LIGHTNING AND LIGHTNING PROTECTION 275 Let, for instance, a spark occur in an insulated under- ground cable system between one of the conductors and the gfrounded cable armor, through a weak spot in the insulation, as a faulty joint or a cable bell. Normally a potential difference exists between the cable conductor and the ground, equal to the Y potential of the system, and so an electrostatic charge on the conductor corresponding thereto. A spark passing between conductor and ground, connects it to ground, and the charge of the conductor so passes over the spark as arc to ground. As soon, however, as the conductor is discharged and at gfround potential, the arc between conductor and ground ceases, since there is no voltage left to maintain it, and so the conductor disconnects from ground. The conductor then charges itself again to its normal Y potential and during the in-rush of the charge, momentarily the potential builds up to double voltage. Thereby a spark again passes between con- ductor and ground, discharges it again, opens after discharge, again causes a spark to pass, etc. So a series of successive sparks occur between conductor and ground, discharging the conductor by currents which momentarily rise to very high values, the discharge current of the capacity of the conductor against ground, over a path of practically no resistance. Each spark discharge sends out an impulse or traveling wave, and thus a spark discharge between conductor and cable armor, or in the same manner an arcing ground on an overhead transmis- sion line, as is for instance caused by a broken insulator, pro- duces a continuous series of impulses or traveling waves, which follow each other with the rapidity of charge and discharge of the cable or the line, that is, many thousands per second, and so give what has been called a recurrent surge. In a long distance transmission line, the frequency of the recurrent surge usually 276 GENERAL LECTURES is somewhat lower than in an underground cable system, but is still thousands of impulses per second. The frequency of oscillations occurring in electric cir- cuits varies over an enormous range: from low frequencies, very little above alternator frequency, up to hundreds of mil- lions of cycles per second ; and the effect of the oscillations in the system therefore varies accordingly: from the relatively harmless static displays; brush discharges, streamers, sparks, etc., of extremely high frequencies, down to the disastrous high power low frequency short circuit oscillations, in which even in 10,000 volt system*^, currents ^i many thousands of amperes may surge, which voltages approaching 100,000, and with which no protective device can cope, which does not have unlimited discharge capacity, that is, contains no resistance whatever in the discharge path. III. LIGHTNING PROTECTION OF ELECTRIC CIRCUITS. From the preceding considerations it follows that the problem of protecting electric circuits from lightning is two- fold: 1. To guard against high potential disturbances enter- ing the circuit from the outside or originating in the circuit. 2. To discharge harmlessly to ground, whatever high potential phenomena may appear in the circuit. From atmospheric electric disturbances, complete protec- tion can be secured by putting the circuit under ground, or, where this is not feasible, to put the ground over the electric circuit. This means the use of grounded overhead wires. The overhead ground wires so protect the circuit the more com- pletely, the more they realize a complete shield interposed between line and sky. While complete protection thus would LIGHTNING AND LIGHTNING PROTECTION 277 require a system or network of grounded conductors above, beside, and also below the transmission line, very good protec- tion in most cases is secured by a single ground wire of good conductivity, installed well above the line; and in no place of electric transmission systems can money be more efficiently spent, than in securing good overhead ground wire protection. To guard against the appearance of internal lightning requires constant watchfulness in the design, construction and operation of the system, to avoid all conditions which may lead to the formation of oscillating arcs. Thus poor contacts, loose joints, masses of insulated metal near high potential con- ductors, etc., should be carefully avoided. The disturbances which have to be taken care of by the lightning arresters proper, are steady accumulation of static pressure; impulses or traveling waves; oscillations or surges ; occurring singly or in groups, and of frequencies vary- ing between many millions of cycles and ordinary machine frequencies; and recurrent surges, that is, impulses and oscil- lations, usually of high frequency, following each other in very rapid succession, usually thousands per second. It is necessary that the discharge over the lightning arrester should occur with the least possible disturbance to the system, that is, the discharge current should be as small as per- missible without causing a voltage rise due to the resistance of the discharge path. At the same time, the protective devices must be able to discharge practically unlimited currents, that is, currents of the magnitude of the momentary short circuit current of the system. This obviously requires that the pro- tective devices should have no appreciable resistance in the discharge path. Any lightning arrester containing series resistance obviously fails to protect as soon as the discharge current is so large that the ohmic drop across the resistance 278 GENERAL LECTURES becomes serious, and the maximum discharge current which may occur, is the short circuit current of the system, that is, extremely large. Three types of protective devices are at present available. I. The circuit is connected to ground by a single spark gap set for a voltage exceeding the normal operating voltage by a safe margin : the so-called horn gap, or goat horn light- ning arrester. As soon as the voltage rises beyond the value for which the spark gap is set, it discharges, and the system is short circuited to ground, until the arc rises and gradually blows itself out. As this requires an appreciable time, motors and converters have usually dropped out of step, and the gen- erators have broken synchronism, that is, the system is shut down and has to be started up again. This type of protection therefore is not particularly favored in systems which require reasonable continuity of service, but if used, it is considered rather as an emergency device in addition to other arresters and is then adjusted for much higher discharge voltage. A reduction of the current over the horn gap by series resistance is not per- missible, since it correspondingly reduces the protective value, as explained above, and the arrester ceases to protect against a high power surge. While such surges are relatively infre- quent, their destructiveness is such that protection against them is especially needed. Fuses in series with the horn gap, if they open slowly, would still shut down the system, and if opening very rapidly, the shock of the explosive opening of the fuse on the short circuit current of the system may be disastrous. Obviously, the use of series fuses require a multi- plicity of spark gaps to give continuity of protection. 2. The type of lightning arrester now almost universal- ly used is the multi-gap arrester, which short circuits the system for one-half wave only. It consists of a large number LIGHTNING AND LIGHTNING PROTECTION 279 of spark gaps between metal cylinders, in series with each other. As now designed, different sections of the gaps are shunted with different resistances, for the purpose of affording equal protection against all frequencies, and adjusting automatically the resistance of the discharge path to the volume of the dis- charge, as for instance, discharge slow accumulations of poten- tial over a very high resistance, short circuit surges over a path of zero resistance, and thus pass a discharge with the minimum shock on the system. The operation of the multi-gap — ^which by the way is suitable only for alternating current systems — depends on the non-arcing character of cer- tain metals. Metals of low boiling point, as mercury or zinc, cannot maintain an alternating current arc, but the arc goes out when at the end of the half wave, the current falls to zero, and a very much higher voltage is required to again start an arc for the next half-wave.* Alloys of such metals, usually zinc, with metals of high melting point, as copper, are therefore used as terminals in the multi-gap arrester. A discharge over the multi-gap arrester short circuits the system for the rest of the half- wave during which the discharge occurs. At the end of the half-wave, the current falls to zero, and the reverse current cannot start, that is, the circuit of the arrester is opened. A short circuit on the system, for a fraction of a half- wave, does not interfere with the operation of synchronous apparatus, that is, the operation of the system is not affected by a discharge over the multi-gap arrester. In a large system, the short circuit current is very consid- erable ; its power, and thus the heating effect produced by it, is enormous. The energy, and thus the heat produced by the short See paper A. I. E. E. Transact. 1906, p. 789. "Transformation of Electric Power into Light." 28o GENERAL LECTURES circuit current during the fraction of the half-wave, which the discharge over the multi-gap arrester lasts, is moderate, due to its very short duration, and can easily be absorbed and radiated by the arrester; so that even if lightning discharges rapidly follow each other for some time, they can be taken care of by the arrester with moderate temperature rise : assuming a vicious thunder storm, in which lightning flashes succeed each other practically continuously, several per second. Each discharge causes a short circuit over the lightning arrester, varying in duration from nearly a half-wave — if the discharge occurs at the beginning of a half-wave — to practically nothing — if the discharge takes place near the end of a half-wave — ^that is, in average, for one-half of one-half wave, or :— • sec, in a 60 cycle system. Therefore from two to three lightning dis- charges per second would still short circuit the system over the multi-gap arrester only for i % of the total time, and the heat- ing effect, caused by a short circuit during i % of the time, can be taken care of by the arrester for a considerable period. Let us see, however, what happens to the multi-gap light- ning arrester in case of the appearance of a recurrent surge, as an arcing ground, that is, discharges following each other in rapid succession, thousands per second. The first discharge, passing over the lightning arrester, short circuits the system for the rest of the half-wave, and at the end of the half-wave, the arrester functionates properly, that is, opens the circuit. At the next moment, however, at the beginnirg of the next half- wave, the next oscillation of the recurrent surge again dis- charges over the arrester, and thus again short circuits. That is, with a recurrent surge, the multi-gap arrester at the end of every half-wave opens the circuit, at the beginning of the next half-wave, the next oscillation of the recurrent surge short circuits again. As far as the effect on the operation o? LIGHTNING AND LIGHTNING PROTECTION 281 the system, and the heating of the arrester is concerned, a recurrent surge causes a permanent short circuit on the system, except that at the beginning of every half-wave, for a short period, the circuit is opened and free for the appearance of disruptive voltages elsewhere, and so apparently, simul- taneous with the short circuit, destructive high potentials may appear in the system. The heating effect of the short circuit current, which occurs at every half-wave, rapidly destroys the arrester. In such cases, to save the arrester, it has been cus- tomary to insert a series of auxiliary gaps, which are thrown in by the blowing of a fuse shunting them, and raise the dis- charge voltage of the arrester so that the recurrent surge does not pass over it. It is obvious, that in this case the arrester ceases to protect the system against the recurrent surge: but if left in circuit, the destruction of the arrester would put it out of operation anyway. It is obvious now, that no lightning arrester, which func- tionates by short circuiting the system for the rest of the half- wave, during which a discharge occurs, can take care of and protect against a recurrent surge, since the proper functionating of the arrester, with a recurrent surge, represents a permanent short circuit on the system over the arrester, and so a destruc- tion of the arrester, no matter whose make it may have been, and a shutdown of the system. 3. To take care of a recurrent surge, a protective device would thus be required, which does not short circuit the system even for one half-wave, but which never allows the normal voltage of the system to pass a current over the arrester, but acts as a short circuit for any excess voltage above the normal voltage. The possibility of such a device we can understand by considering the effect, which in a direct current circuit a storage battery would have, when shunted between the 282 GENERAL LECTURES circuit and the ground. Assuming for instance in a 500 volt trolley circuit, a 500 volt storage battery of very high capacity, that is, negligible internal resistance, permanently connected between line and ground. With the normal line potential of 500 volts, no current would pass over the battery to ground, except the very small current required to maintain the battery charged. No rise of voltage, however, could occur in the system by lightning or any other cause, since any voltage above 500 volts, the counter e. m. f. of the battery, would be short circuited to ground through the battery, and such a battery would thus give perfect protection against any high voltage dis- turbances in the system. In case of a recurrent surge, the cur- rent discharging over the battery would be the short circuit current of the excess voltage, that is, the surge potential, and the heating effect of this current is negligible, since high potential high frequency phenomena are of limited power and especially of limited current, as condenser discharges. A storage battery obviously is not suitable for alternating current and would not be practical in any case, as it requires a cell for every two volts. The same effect, however, is pro- duced at a much higher voltage, in an alternating current cir- cuit, by the aluminum cell. If such a cell, consisting of two aluminum plates in certain electroljrtes, is exposed to an alter- nating voltage, a film forms on the aluminum plates, which holds back the impressed voltage, that is, acts like a counter e. m. f. equal to the impressed e. m. f., so that practically no current passes through the cell, or only the small current required to maintain the film, of a magnitude of about .01 amperes per square inch plate surface, while for any sudden rise of voltage the cell acts as a short circuit for the excess voltage. Over the storage battery, the aluminum cell has the advantage of higher voltage: a single cell can take care of LIGHTNING AND LIGHTNING PROTECTION 283 300 to 400 volts and even more, and also that it does not have a fixed counter e. m. f., but a counter e. m. f., which adjusts itself to equality with the impressed voltage, at any value up to about 6cK) volts per cell. Assuming for instance an alumi- num cell connected across an alternating e. m. f. of 300 volts. With the film formed, a negligible current passes " through the cell, for instance, ^ of an ampere, maintaining the integrity of the film. If now the voltage is suddenly raised to 330 volts, in the first moment the cell acts as a short circuit of the excess voltage, in this case, 30 volts, and for an instant a very large current, possibly hundreds of amperes if the supply source is capable of giving such a current, rushes through the cell. This current very rapidly decreases, by the film of the aluminum plates forming for higher voltage, so that in a few seconds the current is already small, and in a few minutes the normal current of 1-4 ampere again passes, but now at 330 volts impressed, and the film has formed to a counter e. m. f. equal to this higher voltage, probably has thickened. If now we again lower the voltage suddenly to 300, in the first moment the current in the cell practically disappears, and then graually rises again, and after a few minutes is again normal at ^ ampere, that is, the film has built down again to 300 volts. In this manner the aluminum cell adjusts its counter e. m. f. to changes of impressed voltage, by the film building up or build- ing down. This adjustment, for moderate voltage variation, as may be expected when varying the generator voltage of the system, is quite rapid, most of the change occurring within less than a second, but is still extremely slow compared with the rapidity of lightning phenomena, and for lightning phenomena the aluminum cell therefore acts as a short circuit of the excess voltage above the normal machine voltage. Thus the recurrent surge, with a system of aluminum cells in series with 284 GENERAL LECTURES each other connected directly across the circuit, cannot produce any rise of voltage, but the excess voltage over the normal, or the surge potential, is short circuited through the aluminum cell, so causing a small increase of the current in the cells, by the superposition of the high frequency surge current over the normal leakage current of the cell, but no rise of voltage. Since the recurrent oscillations are intermittent, obviously the film of the aluminum cells cannot build up to their voltage, but remains corresponding to the machine voltage, that is, the aluminum cell can permanently discharge a recurrent surge without any short circuit of the main voltage, or any disturb- ance on the system. a y i 4 PLEASE DO NOT REMOVE CARDS OR SLIPS FROM THIS POCKET UNIVERSITY OF TORONTO LIBRARY IK ENGIM.