ELECTRICITY AT HIGH PRESSURES BY ELIHU THOMSON A Lecture delivered before the New York Electrical Society March 29, 1899 At the Society of Civil Engineers, New York City \ Vcvn^^cfc o ELECTRICITY AT HIGH PRESSURES BY ELIHU THOMSON A Lecture delivered before the New York Electrical Society March 29, 1899 At the Society of Civil Engineers, New York City ELECTRICITY AT HIGH PREgSlIRES. MY interest in the subject we shall consider this evening .is th t e more pronounced from the fact that I began my ex'peji^)ice,\\ ; it:K electrical phenomena by working with relatively high pressures or potentials. My first machine, constructed when I was eleven years of age, was a high potential apparatus, giving about 30,000 volts or thereabouts. It was a frictional electric machine, the main part being a wine-bottle revolved upon an axis, to which was added the usual rubber, silk flap, and prime conductor. Electrical development on the large scale has in the past few few years been going on in the direction of increase of electrical pressures, or increase of potential differences, and this fact gives a renewed interest. The prime cause of all electrical manifesta- tions is, of course, difference of pressure or potential. That we have much more to learn in this fascinating field is evidenced by the condition of our knowledge in regard to the phenomena of lightning, the aurora, comets' tails, and possibly, also, of the solar corona. It is about one hundred years since Volta studied the voltaic battery and gave to the world a source of steady currents at low pressure. Long before that, however, the older experi- ments of the development of electric charges by friction, and the properties of charged bodies had been studied and wondered at. Priestley, in his "History of Electricity," a work of nearly 800 pages, has even given expression to the opinion that, "In elec- tricity there is the greatest room to make new discoveries. It is a field but just opened," etc. His book was published in the latter part of the last century. It dealt entirely with electricity at high pressures. Even the old experiment of rubbed amber would give 10,000 to 20,000 volts, and the old glass globe machines, such as Franklin used, a much higher pressure; nothing, however, in com- parison with that of a lightning stroke. Still I must say that I have no sympathy for those who speak without hesitation of hun- dreds of millions of volts even in such cases, or who affect to describe apparatus as in existence to-day developing pressures of several millions of volts. A few years ago (in 1892) I constructed a high potential, high frequency coil which gave a torrent of 64-inch 679936 4 ELECTRICITY AT HIGH PRESSURES. sparks, estimated by me at the time as representing about three- fourths of- a million volts, and since thought by Professor Trowbridge of Jefferson Physical Laboratory, Harvard University, to have required one million volts. Professor Trowbridge has since con- structed apparatus from which he has obtained some sparks of seven feet in length, a discharge of perhaps something more than a million and a 'half volts. This apparatus was calculated to give about three millions of *vci'ts,'but proba'bly- before that pressure was obtained, the leakage, brush discharges to the walls, floor, and surrounding objects, limited the pressure to that which gave the discharge of seven feet. To illustrate the development of electrical pressures by friction, I may describe an experiment of many years ago. I poured a layer of turpentine varnish upon a clean tin plate and allowed it to dry into a flexible sheet closely adhering to the tin. Upon stripping it off carefully so as not to cause friction, it came away with an electrical pressure, or charge, sufficient to cause small sparks, etc. The varnish, in close association with the tin surface, had taken up an opposite electrical state to that of the tin. The pressure or potential difference was not, however,' high while the surfaces- remained in contact, but, upon separating them, the pressure rose gradually during separation. This was owing to lessened capacity,, there being no more electricity in the one case than in the other. The positive and negative electricities were neutralized or bound by being close together, and the pressure was very low. When the film was raised, the electrical pressure rose and energy was expended, just as it is expended in lifting a weight. I may illustrate the conditions by laying a thin sheet of rubber upon a smooth zinc plate, and rubbing with a piece of fur the upper surface of the thin rubber film. A bound charge is accumulated upon the rubber, and the zinc and the rubber exhibit, as tested by the electroscope, only a slight electrification. The charge is so nearly neutralized by the opposite charge brought up by the zinc that the potential exhibited is very low. Upon stripping the thin rubber from the plate, work is done in overcoming the attraction due to the opposite charges, and the potential of the rubber at once rises so that it gives sparks of two or three inches to the knuckle held near it. When, as in this case, the electricity is developed by friction, the rubbing only acts to secure area of contact, or to insure the condition that all parts of the rubbed surface shall have been in close contact with the rubbing body in this case the fur. If a ELECTRICITY AT HIGH PRESSURES. 5 piece of ebonite be rubbed by the fur, wherever the two touch bound charges are developed and would remain bound and neutral- ized. They would exist at low voltage until they were separated. The high pressure charges only come into existence as the two bodies are gradually separated. How inefficient the friction process must be can be understood when it is pointed out that the actual work which produces electricity is not that of the rubbing, but only that of the contact and separation of the two bodies. The friction proper produces heat; the pulling apart of the bound -charges gives electrical energy at high pressure. The phenomena of development of high electrical pressures by belts, such as those of high speed machinery, are illustrations of the same principles. Instructions are sometimes given to discharge the belt by a row of points connected to earth, so as to avoid the belt electricity endangering the insulation of a dynamo or motor by -charging its frame. Such a procedure is the very one to enhance or increase such danger, and the proper procedure is to connect the points to the frame itself and to the shaft which serves as the countershaft belted to the dynamo shaft. Some of the larger belts used in transmission give very striking exhibitions of electricity at high potentials. Sparks over three feet long were obtained under a large belt at the Narragansett Electric Lighting Company's station, in Providence, some years ago. The old electrophorus was an instance of the utilization of similar principles of alternate binding and freeing of a charge pro- duced by friction upon a resinous plate contained in a sheet metal tray. When the insulated disk of the electrophorus is lifted off the resin surface, the charge on the resin binds itself to an opposite charge induced in the tray, and when the disk is returned and con- nected to the tray or to ground, the charge mostly binds itself with an opposite charge induced in the disk now resting upon the rubbed resin. Upon disconnecting the disk from earth and lifting it off the resin surface, it is found to be possessed of a charge of opposite name to that upon the resin surface, which latter again binds itself to its opposite in the tray. The work done is the lifting off of the disk against electrical attraction between the disk and resin surface, the energy used in lifting manifesting itself in a high pressure charge in the disk. The apparatus known as Armstrong's Hydro-Electric Machine is an instructive example of the operation of similar principles. It enables a high potential to be produced from the effects of a slight charge produced in minute drops of water which are driven ELECTRICITY AT HIGH PRESSURES. along with steam, through a jet lined with box-wood. The globules from partly condensed steam, on leaving the jet form a small cloud into which collecting points from a prime conductor project, and a high potential charge is thus given to the con- ductor, from which long sparks may be taken. Since the actions occurring in this steam electrical machine serve to throw light on the accumulation of high pressures in a thunder cloud, I may be pardoned for dwelling upon the matter for a few minutes. Let S, Fig. i, represent an insulated sphere, charged to, say, e% S :>e 000 FIG. i. FIG. 2. 10,000 volts potential with respect to earth. It will have a certain capacity depending upon the proximity of surrounding conducting objects. Let this sphere be now surrounded by a number of hollow spheres, such as by bringing hollow hemispheres together over the center sphere, properly supported and insulated. Let the hollow spheres be well insulated from S, and from each other, by a sufficient layer of air or dielectric. If each hollow sphere be like S, charged to, say, 10,000 volts before being brought over S, the potential ex- hibited by the outer sphere will gradually rise as the spheres are added. This is owing to the diminished capacity of the concentric shells as compared with their capacity when separately charged. If we now substitute for the hollow spheres and central sphere a large number of floating water globules in air as electrified and driven out from a jet J, Fig. 2, of a Hydro-Electric Machine, we have a similar condition, the result of which is that, although the ELECTRICITY AT HIGH PRESSURES. 7 globules so driven together in a mass but are slightly charged when leaving the jet, the combined effect of all results in greatly en- chanced potential, owing to the fact that the water globules within the small cloud have virtually been deprived of capacity. Add to this the effect of diminished total surface of the globules due to their coalescing, and further the reduction of the surface of the globules themselves by gradual evaporation in the air, and we need not be surprised that the potential shown by the mass is high, though it was formed by driving forward into the cloud globules originally at comparatively low potentials. Applying the same considerations to the thundercloud itself, we seem to understand why it is that globules of water suspended in air may act inductively, when only slightly electrified at the start, FIG. 3. and so give rise to the exhibition of enormous potentials. Assume a horizontal layer of air L, Fig. 3, saturated with moistureand possess- ing an electrical charge, however it may have been accumulated. If such a layer be uplifted by warm uprising currents from below, it will condense into a cumulus cloud containing electrified globules of water insulated from each other, but massed together in the cloud with diminishing capacity due to the various causes before men- tioned: the inductive effect outward of the globules within the mass; the coalescence of numerous small globules into larger ones; and possibly also, in some portions of the cloud, the diminishing size due to evaporation. The cloud, as a whole, will show a high state of electrification, but no one part of it need at any time be very highly charged in the sense that an insulated body is charged. In 8 ELECTRICITY AT HIGH PRESSURES. the latter the charge is on the surface. In the cloud each globule is charged and also acts inductively with its neighbors, outwardly and toward the earth. In discharging, the spark forks and extends or ramifies through the cloud masses, often occupying an appreciable time in this process. The discharge of one portion of cloud opens a good path of hot gas for other discharges to follow, and the disturbance thus spreads. Frequently discharges are multiple in character, several successive discharges following the same path to earth under the cloud, but, doubtless, within the cloud, ramifying in different directions and discharging different portions of the cloud mass. If this idea of the actual conditions which exist in a thundercloud be true, it may easily be understood that attempts to represent the action of lightning by the discharge of small condensers, such as Leyden Jars, may utterly fail of their purpose. Neither can any arguments avail, based upon the discharges of lightning having a rapid oscillatory character. That the breakdown of the air under a thundercloud when a stroke to earth occurs is very sudden is doubtless true, but that the flow or flash is oscillatory or that it en- dures for an excessively short time may not be true; at least, not in all cases. The conditions must be very variable, and the discharges, occurring under wide variations of these conditions, are not likely to be alike. It is useless, therefore, to attempt to calculate the voltage or the current in a flash of lightning, and no estimate of the energy expended can be more than a guess. Such devices as lightning arresters cannot be fully tested as to their action or effectiveness without actual practice with them dur- ing thunderstorms, and tests made with ordinary static discharges from Leyden Jars may not, and probably do not, represent the real conditions under which the devices themselves will be required to work. Hertz showed that the higher pitch rays, such as those of the violet or ultra violet order in the spectum, served to precipi- tate electric discharges between terminals set so far apart that the potential between them was insufficient to cause a spark. The thought occurs that possibly such phenomena as that of the return stroke, as well as that of multiple strokes or flashes of lightning, oc- curring practically simultaneously over different paths, may some- times depend on the sudden illumination of the air by a first discharge provoking other discharges around it. I am, moreover, led to suspect that very long discharges in cloud, such as those which occur horizontally, over distances of perhaps five to ten miles in a cloud layer, actually take an appreciable time ELECTRICITY AT HIGH PRESSURES. 9 to develop or progress. There seems indeed to be a progressive breakdown of the cloud, extending on and on during a time which can be noted. I have in fact often watched the development of such discharges in the east or toward the eastern horizon, and have been able to turn the head and follow the discharge to its finish in the west or toward the western horizon. I have noticed also that these long discharges will start as a single streak, ramify, and extend their rami- fications before the eyes. I am persuaded that this is no optical illusion; the effect seems too definite for that. It would be of deep interest to photograph such discharges on two plates, one fixed and the other revolving, and compare the images. This is quite a diffi- cult matter, however, as lenses having a sufficiently wide angle to take in the whole path, or practically the whole sky at once, are not to be had, and one never knows just when, and at what portion of the cloud, the desired discharge may take place. Moreover, for such work the possibility of pointing a camera upward or nearly vertical may be prevented by rain occurring at the time. How much greater is the contrast that existed in Franklin's time between the lightning's effects and those of the feebly acting ap- paratus at his command, than can be said to exist to-day with modern apparatus capable of yielding long sparks! With the latter, Franklin's genius would not have been needed to discern electricity in lightning. The old frictional machines possess now only an historical inter- est. The principle of the electrophorus was applied about 1866 by Holtz in his famous influence machine, which has undergone various modifications from time to time during the past thirty years. It is not necessary to discuss the details of its construction and the theory of operation. The various forms of static influence ma- chines are in reality forms of continuous electrophorus, the more recent machines embodying the principle of the multiplier or doubler used by Lord Kelvin many years ago in charging his electrometers. A very large machine has recently been constructed under the direction of Dr. Francis H. Williams of the City Hos- pital, Boston, for use in exciting Crookes tubes in his Roentgen- ray work. It has four revolving glass plates six feet in diameter, and the effects obtainable are very excellent for Roentgen-ray excitation. We have before us, as a loan for this lecture, a splendid example of a Holtz influence machine as manufactured by the Galvano-Faradic Mfg. Company of New York. It has eight revolving glass plates 10 ELECTRICITY AT HIGH PRESSURES. of 30 inches in diameter, in addition to a pair of plates which can be revolved oppositely by hand for charging the sectors of the main machine. It is now driven by an electric motor of 1/6 H. P., and works with great smoothness. The construction of the machine is certainly a credit to its makers, who have apparently given careful attention to every detail. They have produced, also, a very hand- some piece of electrical apparatus. The foregoing apparatus de- pends chiefly upon the principle of increasing potential by diminishing capacity. If a number of Leyden jars, or other con- densers of nearly equal capacity, be charged in multiple and then connected in cascade or in series, the terminal difference of pressure or potential will be found to be increased in proportion to the num- ber of jars so connected. In this case the capacity of the set, when in multiple and acting as one large condenser, is virtually reduced at FIG. 4. the moment of connection in series and the pressure of the con- denser rises accordingly. This is the principle of the Plante rheo- static machine, to be alluded to later. To illustrate the actual change which takes place in the Rheo- static machine, we may represent by A, Fig. 4, four condensers charged in parallel. The single thickness of dielectric is equally strained; there being four units of surface -j- and . When these are connected in series the end foils only need to be considered: the dielectric becomes four times as thick, as shown at B, and the dielectric is throughout under the same electric stress as before. Such a condition cannot, however, be maintained without an in- ELECTRICITY AT HIGH PRESSURES. II crease of potential difference between the single pair of condenser foils in B. There is thus a multiplied potential according to the number of plates so placed in series. Besides the method of raising potential by diminishing capacity as above, electromagnetic methods are well known and are now em- ployed extensively. When a coil of wire is wound upon an iron core and a second coil of finer wire is wound alongside of it or around the same iron core, a periodic current sent through the first wind- ing induces a periodic electromotive force wave, or a current, as the case may be, representing an increase of pressure or potential nearly in the relation of the turns in each coil. This relation is called, in alternating current transformers, the transforming ratio. In high potential transmission this ratio maybe even as great as 200 to i in extreme cases. Usually the ratio is not so great. The same transformers, being completely reversible in function, are used both for step up and step down transformation. A very large amount of power is now transmitted over varying distances by utilizing this principle of transformation. Many large plants are in operation, using for the line 6000 to 10,000 volts; some involve pressures of 20,000, and one plant in California uses 40,000 volts on the line. This last is, properly speaking, a very high pressure for such work. Whether it will be much exceeded in the further development of the art remains to be seen. Mr. Chas. F. Scott in a recent paper has ably dealt with the question of in- crease of voltage in systems of transmission, and the subject is too extended to be taken up in detail now. Suffice it to say that there are indications of a limit being reached at which the losses from the wires may become so great as to neutralize the benefits attained in raising potential. Besides, the capacity of a very long line at high pressures introduces new problems and requires additional compli- cation to balance it. We have very much to learn in the field of high pressure transmissions. Difficulties which seem insurmount- able to-day may disappear in the technical advance of a few years. This has always been the case in the development of new arts, and it would be unwise at present to assign any limits to increase of pressure. Back in 1878 I used to wonder whether I should live to see elec- tric currents used of as much as 25 to 50 amperes with a line of a potential, or tension as we then called it, such that a discharge would leap from the line to earth over as much as half an inch. I discussed the possibility of such transmissions with friends, and tried to picture to myself the conditions of such a line, even figuring out 12 ELECTRICITY AT HIGH PRESSURES. the energy transference. I am free to confess that I had little hope of eve.r being a witness to the actual realization of such dreams. But we are indeed far beyond those early meditations in our actual accomplishment. We build transformers for 10,000 to 30,000 volts of pressure and send hundreds of amperes into the line at such pressures; and we transmit power in thousands of kilowatts over distances up to eighty miles. Transformers are built which singly have a capacity of 2000 kilowatts, and 10,000 volts are em- ployed on lines, as at Niagara, where the distance is, say, two and a half miles only, because copper can be saved. A transformer of such large capacity as to have a voltage of sixty for each turn around its core no longer surprises us; and transmissions of power involving 15,000 to 20,000 horse power, to use the old unit now gradually being displaced by the kilowatt, are beginning to be ordi- nary achievements of electrical engineering. The potential which is developed in any coil through the axis of which a magnetic circuit is or can be formed will depend upon the maximum rate of change in the number of lines of force of such circuit, in entering or leaving the core. Each line of force will, in being introduced or removed from the area inclosed by the wire coil, develop electromotive force in each turn so cut by the mag- netic line, which electromotive force depends on the rate of cutting. We of course use the phrase "line of force " as a convenience. The first contact, so to speak, between the old so-called statical electricity or electricity of "high tension" and "galvanic or voltaic electricity," which were at one time regarded as distinct species, was made by the invention of the Ruhmkorff coil or induc- tion coil. It bridged the gap existing, and served to break down a barrier and destroy distinctions between different kinds of elec- tricity, which distinctions are never entertained now. In such an ordinary induction coil used for obtaining high pressure sparks or discharges from low pressure currents, as of a few cells of battery in series, the ratio of transformation is made very high and the rate of change of the magnetism of the field or iron core is also made as great as possible. The turns of the sec- ondary coil may bear to those of the primary the relation of 1000 to i, and the interruption of the current sent through the primary is made as sudden as possible. The simple addition of a condenser across the break in the primary circuit greatly increased the potential obtainable from a given sized coil. The condenser not only prevents spark at the contacts of the interrupter of the coil, which spark would prolong the current and make the interruptions. ELECTRICITY AT HIGH PRESSURES. 13 much less sudden, but it also acts to receive the self-inductive dis- charge of the circuit or extra current of Faraday, and becomes itself so highly charged that it can send a reversed current through the primary circuit, and so increase the rate and extent of change of magnetism of the core upon which the potential developed in the secondary largely depends. The condenser in this case, in fact, gives an oscillatory character to the primary discharge, as is easily noted in the changing sound of the secondary spark when the capacity of the condenser is varied. I have in my possession a coil which shows this phenomenon very clearly. The insulation is of supreme importance in these cases of devel- opment of very high electric pressures in coils of this type. A new interrupter, Fig. 5, remarkable for its extreme simplicity and for the sharpness of its interruptions, has been recently FIG. 5. described by Dr. Wehnelt. As no condenser is needed with this device, it leads to a great simplification of the coil apparatus. The interrupter consists simply of a vessel containing an electrolyte, such as sulphuric acid of density 1.2, into which a lead plate or surface is plunged and connected as a cathode in the circuit. Dip- ping into the liquid is a glass tube, into the end of which has been sealed a platinum wire which projects into the liquid a short dis- 14 ELECTRICITY AT HIGH PRESSURES. tance. The wire is attached to a cable within the glass tube, or its inner end is surrounded by mercury into which a wire is dipped for connection to the platinum. The platinum is made an anode. The induction coil primary is connected without condensers, in series with the interrupter across battery terminals or lighting mains, and FIG. the action is such that the current in the interrupter circuit is periodically cut off. Figs. 6 and 7 are photographs of the stream of sparks produced by an induction coil so operated. Fig. 7 is in reality a flaming arc like a high potential alternating current discharge. I conceive the action of the Wehnelt interrupter to be about as follows: The gradual increase of the current through the primary on closing the circuit causes the platinum anode to evolve oxygen gas bubbles, which tend to cut down the section of the liquid in con- FIG. 7. tact with the anode. The inductance of the primary circuit tends to maintain the current through the limited section of the liquid in contact with the wire. This, in turn, leads to more rapid gas evolution and, possibly, disassociation of the liquid, whereby ELECTRICITY AT HIGH PRESSURES. 15 there is a very sudden stripping of the liquid layer from the platinum anode surface, which, being now surrounded only with gas, is out of contact with the liquid. This action occurs so suddenly as to give rise to small but vigorous explosions of vapor or gas. The condensation into liquid, or reunion of the particles after this separation and rupture of contact at the platinum surface, causes a sudden rush of the liquid conductor toward the platinum, re-estab- lishing the circuit, which is thereafter very quickly interrupted as at first, and so on. The period of these interruptions will naturally be higher the higher the potential of the circuit supplying the cur- rent, the lower the inductance of the circuit of the interrupter, and the smaller the platinum anode surface. In illustration of this fact, I have brought before you a small apparatus (Fig. 8), which is used to produce varying rates of inter- ruption. It consists of a small interrupter in series with a small coil wound upon an iron core. The interrupter is made of an ordi- FIG. 8. nary tumbler and has four platinum anodes of varying size dripping in the acid solution. I shall use three of these in parallel. Out- side the small coil on the iron core, or primary coil,, is a single layer of secondary with taps from each turn, thirty-five taps in all, taken to stationary contacts arranged in an arc of a circle. These are traversed by a radial arm, which enables me to short circuit more or less of the secondary turns. I am thus, as you see, enabled to actually .play a tune on a sort of electrical organ, not, it is true, very remarkable for the fine quality of its tones. I now place in series with the primary an electro-magnet similar in con- struction to one of those used by Dr. Elisha Gray in his electro- harmonic telegraph years 'ago. This magnet is fastened to the r6 ELECTRICITY AT HIGH PRESSURES. bottom of a resonating box. I adjust the pitch of the interruptions until I get the condition of resonance, and a very loud sound testifies not only to the vigor of the interruptions but to the energy going into sound waves. We may have here an electrical substitute for a steam whistle or fog siren. Even for signaling on moving trains, electricity need not, it is evident, be dependent upon com- pressed air. A form of induction coil depending upon a very high rate of change, as well as upon the transforming ratio of primary to sec- ondary windings, is the well-known high frequency coil. In a sense, the ordinary induction coil, with condensers, as above stated, is a high frequency apparatus, as the breaking of the primary circuit charges the condenser across the break, then in discharging gives a relatively high frequency oscillation through the primary circuit. If separate means be employed to charge the condenser to a high potential, and it be permitted to discharge over a spark gap through a few turns of wire, well insulated, we have a high frequency dis- charge in these turns. The lower the capacity of the condenser and the less the inductance of the circuit, the frequency will be greater, according to the law enunciated by Lord Kelvin. In 1889, Professor Henry A. Rowland of Johns Hopkins Uni- versity, in a discourse at the annual meeting of the American Insti- tute,* drew attention to the high frequency effects, and employed experimentally a Ruhmkorff coil, the discharge of the secondary of which charged some Leyden jars or condensers C l (Fig. 9), which were discharged over a spark gap, a, through a few turns of wire carried on a frame as a large open coil with an air core as at, A, in the figure. Opposite to this and parallel in plane, that is in inductive relation to it, was another coil, JB, acting as a secondary circuit receiving the high frequency induction, and between the terminals of which at b, sparks were discharging at every discharge of the jars through the primary. He showed also that by attuning the two circuits, as by connecting jars or condensers, <7 2 , to the secondary circuit, B, the results were such that the frames or coils could be placed far apart without preventing the inductive action of the high frequency currents. Professor Rowland had all of the elements of a high frequency coil except the step up ratio between primary and secondary. As far back as 1877 it was my custom in lecturing on Electricity to show the effects of the discharge of Leyden jars through the primary of an ordinary induction coil upon * " On Modern Views with Respect to Electric Currents," pages 344, 345, 346, Transactions of American Institute of Electrical Engineers. ELECTRICITY AT HIGH PRESSURES. I/ the secondary as a step up apparatus, and the reverse; having, in fact, received the first suggestion of electric welding while discharg- ing a battery of Leyden jars through the fine wire secondary, while noting the extremely heavy current set up in the primary circuit. By abolishing the iron core and winding only a single layer of FIG. 9. secondary upon an insulating frame placed inside of a primary layer of but few turns, and immersing the whole in a tank of oil, we were enabled to produce the now well-known high frequency, high potential coil. It uses an alternating current of ordinary frequency with a step up transformer to charge the condensers, sometimes with a strong air blast blowing upon the spark gap to insure regu- 18 ELECTRICITY AT HIGH PRESSURES. larity of discharges. Professor Rowland had employed the ordinary induction coil for charging his condensers and an ordinary spark gap. Tesla employed a high frequency dynamo to charge his con- densers, and a blow-out magnet over the spark gap, while in his high frequency coil he at first retained the iron core, discarding it later. A well-arranged, high frequency, high potential coil is adapted to reproduce many of the well-known beautiful effects of Ruhmkorff coils using a rapid break, but with increased brilliancy and effect, while other actions are peculiar to the very rapidly reversed discharges of the high frequency order. The construc- tion of high frequency, high potential apparatus may, of course, be greatly varied. As an example of one form which has been em- ployed by me, I may here insert a description of an apparatus giving 3o-inch sparks between its terminals. A step up transformer, whose secondary gives 20,000 volts alter- nating current, is connected to charge the condensers, the discharge from which passes by air gaps through the primary coil of the high frequency apparatus. This primary consists of ten turns of wire wound on a wooden frame. The conductor is of two No. 6 wires placed side by side. This open coil is 18 inches long and i$$4 inches' diameter. Its resistance is .0088 ohm and inductance .0076 millihenry. The secondary coil has 396 turns of No. 26 wire, wound as a single layer in notches on a hard rubber frame, the wires being spaced apart to form a coil 18 inches long. The di- ameter of the secondary is 12 inches and the weight of the wire about one pound. The wire of the secondary coil layer equals 1250 feet, representing a resistance of 41.6 ohms and inductance of 25.2 millihenrys. These coils are immersed and supported concentrically in a vat of oil, and the secondary has its terminals carried to the brass rods and balls which form the discharge terminals of the apparatus. There are used two condensers, all or portions of which may be connected for the primary coil discharge, each contained in a box 7 inches x 15^ inches inside, and 17^ inches deep. Each box con- tains 84 built up mica sheets 15 in. x 15 in. and .075 in. thick; 42 of these are coated with tin foil 10 x 1 1 inches 1 10 sq. in. Effec- tive foil surface=45io sq. in. The condensers are immersed in oil in the boxes. The capacity of each condenser box is about .03 microfarad. In the apparatus which gave 64-inch sparks between the termi- nals, the primary coil has a length of 28 inches and diameter of 22 inches. It has 15 turns of double No. 6; 85 feet of wire, doubled. ELECTRICITY AT HIGH PRESSURES. 19 Its resistance is .0147 ohm, and inductance .09 millihenry. The secondary coil length is 28 inches, its diameter being 17 .inches. The wire. forms one layer with the turns spaced apart and carried in notches in a hard rubber frame, 580 turns in the layer. The wire is No. 26, about 2^ Ibs. or 2600 feet total length. The condensers used with the primary were in three boxes, the dielectric being of mica plates and oil. When all were in use the capacity was .046 microfarad. These were charged by a large step up transformer to 30,000 volts and discharged across air gaps through the primary. An air blast was kept blowing on the gaps. The greatest distance at which it was possible on account of the construction to set the terminals was 64 inches, which was crossed with ease. The probability is that if the terminals could have been more widely separated, longer discharges could have been obtained. The current used in the primary of the step up transformer being of 125 cycles, there were at least 250 of the 64-inch discharges in each second. We should be cautious in accepting some statements in regard to high frequency currents. It has been claimed that insulators are conductors for such currents, and experiments have been shown in illustration thereof. Thus the two terminals of a high frequency coil, when placed in connection with metal plates on each side of a hard rubber sheet, seem to be short circuited, or if the sheet be double and the two parts be separated a short space, an intense blue dis- charge is seen in the space between the sheets so separated, as if the current got through the dielectric. Even glass in the same way has been regarded mistakenly as conducting the discharges. It is, of course, only an effect of capacity which gave rise to the miscon- ception. The insulating power of oil for high frequencies is as much as ten to twenty times what it is for low frequencies, and it is possible that many other insulators show similar relations. Some years ago 1 investigated some of the properties of oil as an insulator. I found that for certain kinds of currents, such as high frequency currents, oil was from forty to sixty times as good an insulator as air; that is, a 5o-inch spark would be insulated by using an inch of oil. We had terminals in air fifty inches apart, and other terminals in oil one inch apart, and in this case there was an even chance of its jumping under the oil or through the air. We thought this fact favored the use of oil ; but it turned out when we tried these same experiments, or similar ones, with ordinary dis- charges of low frequency, that the very same oil was only about two 20 ELECTRICITY AT HIGH PRESSURES. and one-half times as good an insulator as air that a two and one- half inch spark was all that would be insulated by one inch of oil, varying according to the condition or quality of the oil, which in- sulation was nothing as compared with the insulating power for the higher frequencies. We found also a most striking thing with points. If we put two points opposite to each other in air they of course facilitate the discharges, but we found that points under oil were far better than polished balls, and polished balls better than plates; that oil between plates was broken down easily. It would seem that to get good insulation under oil we would have to make the con- ductor full of points. Whether this experimental fact would hold good when a great many points were clustered together, I do not know. We used single points. That brilliant physicist Plante, whose name is so well known in connection with the early Plante cell or accumulator, invented what he called a rheostatic machine, consisting, in substance, of a set of condensers charged in multiple (as from a great number of battery cells connected in series and giving, say, 5000 volts), with commutating or switching apparatus for connecting the condenser plates in series or in cascade, thus reducing the capacity of the series set and raising the terminal potential in proportion. If ten such condensers were used in the apparatus, and the charging potential were 5000 volts, then the series connection would yield discharges of, approximately, 50,000 volts. Professor Trowbridge of Harvard has in recent years greatly magnified the effects of the Plante machine. He has arranged small storage cells to be charged in sets or In multiple by a dynamo, and then connected in series so as to give 20,000 volts or more of continuous potential at the terminals. With this battery of rela- tively high potential he charges in multiple a range of condensers consisting of glass plates coated on the two sides with tinfoil, ex- cept near the edges. By a large shifting frame worked by a lever he throws the connection of the condensers, after charging, into series order or cascade. The end condenser foils of the set are connected with the discharge terminals of the apparatus. In this way he obtains with great ease discharges of three or four feet in length, and of great brilliancy and beauty, one spark at each throw of the commutator lever. With a large machine of this type he has obtained sparks of seven feet in length, the distance covered being in fact limited by too close proximity of the walls of the building, its floor, its roof, etc., and not by the real capability of the apparatus, which, without the leakages and lateral discharges, ELECTRICITY AT HIGH PRESSURES. 21 would be capable of producing potential differences of 3,000,000 volts. It was while studying the action of Professor Trowbridge's appara- tus that it occurred to me to dispense with the large series of battery cells used in charging, and employ instead thereof the tops of the waves of a high potential alternating current suitably delivered in one direction. I was thus enabled to produce a new machine for obtaining high potential discharges. It consists partly of a motor- dynamo, such as is obtained by taking an ordinary continuous cur- rent motor and tapping the winding so as to obtain alternating cur- rents. To this end two of the commutator segments or leads are connected to a pair of insulated metal rings on the shaft. Brushes resting on these rings yield alternating currents of a periodicity de- pending upon the speed and number of field poles. This machine may be driven by continuous current, or as a synchronous alternat- ing current motor after starting, or it may be driven by mechanical power as a self-exciting alternating current dynamo. The alternat- ing current brushes are connected to the terminals of the primary winding of a step up transformer, giving in the secondary a poten- tial of, say, 10,000 to 15,000 volts. Driven by the shaft of the machine is a frame of insulating material as wood, having at one side a pair of metal strips which periodically connect the high potential secondary terminals of the step up transformer to the plus and minus foils of a set of condensers (eleven glass plates coated with tinfoil) in parallel. These connections, to avoid noise and friction are made without actual contact, that is, over a small spark gap. The revolving frame is so adjusted that the charging shall be completed only at the tops of the waves, and thus a high potential be available for charge. Moreover, in the particular machine before you (Fig. 10), the tops of only every third alternat- ing wave of same polarity are employed, and thus the condenser plates or foils are always given the same polarity. In this way the chopped up alternating discharge becomes, for charging the conden- sers, a substitute for a high potential battery. The revolving frame carrying the charging strips also carries a set of series connectors whereby, after the charging strips have withdrawn from proximity to the stationary contacts led from the condenser foils, these con- tacts are connected in series as in the Plante or Trowbridge ma- chines, and the terminals or end foils discharge across a wide air gap at high potentials, giving a rapid series of sparks of about 12 inches in length in the apparatus here shown. The length of spark is governed by the number of condenser plates and the potential of 22 ELECTRICITY AT HIGH PRESSURES. the charging current. This may be made to correspond with the maximum potential of the wave in the secondary of the high poten- tial or step up transformer. I find shellacked wood sufficiently in- sulating for the parts of the machine outside of the condensers and metal connections. The machine has been further developed and improved by me until it actually becomes a practical substitute for a static machine, independent of the weather. It, therefore, needs no outer case nor means for [drying. I have added a simple attachment whereby FIG. 10. Leyden jar batteries may be charged or a stream of thin sparks ob- tained, and with this attachment the machine may be used to excite the sectors of large influence machines, if desired, in all states of the w r eather. This attachment consists of a revolving connector covering a wide gap between one of the terminals or end condenser foils and a stationary insulating ball or conductor. This connector itself, con- sisting of a pair of balls or rounded surfaces connected by a wire, is insulated and synchronously bridges the gap of several inches be- tween the end condenser terminal and the insulated ball. The time ELECTRICITY AT HIGH PRESSURES. 23 of making this connection coincides with the series connection made by the revolving frame. The insulated ball or conductor thus is synchronously charged, while the opposite terminal of the apparatus may be put to earth. The ball may be made either posi- tive or negative by changing the alternating current connections from the collector rings to the primary of the step up transformer or in other ways. A Leyden jar battery or condenser may be charged by connecting its interior coating with the insulated ball and its exterior to earth, or to the opposite terminal of the appara- tus, and from the charged jar condenser a string, dipped in very weak acid or rubbed in graphite, may be made the means for convey- ing the jar charge slowly to a prime conductor for weak or thin sparks, whereby the effects of a static machine may be closely reproduced. The whole apparatus combines so many features of transform- ation of energy as to become a highly instructive apparatus for schools, and its capabilities are quite varied, as can be readily understood. I call it my Dynamo-Static Machine. I may mention as an interesting fact that the new machine may be used to charge a set of condensers to a high potential in a second rheostatic machine, after which series connections being made, as in Professor Trowbridge's apparatus, a second multiplica- tion of potentials would result. Very high potentials may thus be obtained. There are other parts of the subject which can only be touched upon now, as, for example, the varied uses to which electricity at high pressures is put, the question of insulation, dielectric strength, striking distance, leakage, etc., etc. I may call attention also to the difficulty experienced in measurement of high potentials, owing to the striking distance demanding that the parts of measuring instruments for high potential be widely separated. I have at- tempted to overcome this difficulty by sealing in a glass bulb the parts of a small electrostatic voltmeter and in a vacuum so high as to be non-conducting, in which case the parts may be closely placed with the result of greatly increased torque and freedom from short circuits. When two terminals as of a circuit are at high differences of poten- tial, a discharge or neutralization across the gap may, as is well known, take place and may give rise to a variety of effects depend- ing upon conditions. The striking distance in air, or the distance over which a -spark or arc discharge can be formed, is itself very variable. When the terminals are polished balls and the parts of 24 ELECTRICITY AT HIGH PRESSURES. the circuit discharging have some capacity, the spark is sharp and distinct; if, however, after this initial spark the high pressure be maintained an arc or flame takes the place of the first spark, and may continue indefinitely, as in electric power transmissions in the case of a short circuit occurring. It is important to note also that the effect of heating the terminals is to increase the striking dis- tance to a marked degree. Apparatus at high potentials becomes more safe from discharge if the parts are not overheated. A blast of air or a magnetic field between the discharging ter- minals tends to stop any arc by lengthening it so that it breaks, but neither of them has any effect upon the striking distance proper nor upon the effect of hot terminals, so far as is determined. Ter- minals between which an arc has been playing may become so heated that, although they are far enough apart not to be crossed by a given potential when cool, the discharge easily renews itself, and does this repeatedly; in spite of extinction by a magnetic field, for example. With alternating currents such as are used in power transmission at 10,000 to 50,000 volts, it is found that the striking or arcing dis- tance between needle points may be used as a fair measure of the maximum potential difference. Mr. C. P. Steinmetz has investi- gated this matter up to over 200,000 volts, and has published his results, together with a curve showing the relation of striking dis- tance to potential in such cases. In such a curve, if ordinates rep- resent striking distances and abscissas potentials of discharge, the curve on leaving the origin is at first considerably convex to the base; above 10,000 to 20,000 volts becomes nearly straight; and at about 200,000 volts tends upward more rapidly. The general tend- ency is to a more rapid increase in striking distance than in direct proportion to the voltage, although, for a considerable range, as from about 20,000 to 200,000, the two are nearly proportional. I have already mentioned the anomalous condition as to striking distance under oil when points were compared with balls and plates, a condition the reverse of that which obtains, of course, in air and gases generally. It would seem that a perfect vacuum might require an infinite potential to send a discharge even over quite a small gap between terminals in the vacuous space. It would take us too far afield to consider the very beautiful and instructive effects produced by dis- charges in high vacua and in spaces not so completely exhausted. This is a field of scientific investigation which has since the early experiments of Hittorf and Crookes yielded a rich harvest of ELECTRICITY AT HIGH PRESSURES. 2$ scientific facts and to which a renewed interest has been in recent years imparted by the work of Lenard, and Roentgen, and that of many others, among them Professor J. J. Thomson's investigations. Much more is doubtless to be learned in this field of work. When a solid or plastic dielectric is used as insulation around a conductor, or between two conductors with high pressure stress- ing the dielectric layer, discharge may take place by leakage or by puncture. The latter means usually a breakdown; the former, if not great, may not be serious. The higher the temperature of the insulating layer the less, as a rule, is its insulating power. It is well known that all insulators without exception are either destroyed by decomposition or become conducting at high tem- peratures. This fact has been commonly recognized for many years past, although, curiously enough, it has been put forward as new in connection with the much-talked-of Nernst lamp. So also the dielectric strength, or power to resist puncture under electric stress, is generally weakened by increase of temperature. When the electric stresses are alternated rapidly, as in condensers used on the higher periodicity alternating circuits or with high frequency apparatus, the presence of air or gas bubbles in the die- lectric layer is usually very objectionable, not only on account of the loss of energy involved, but from the great risk of puncture, provoked, it may be, by heating. Each little bubble of air inclosed in the dielectric, or between it and the conducting metal, is filled with blue discharges which waste energy and produce local heat and liability to puncture. Some dielectrics, even when solid, seem to be unsuited as insula- tors where the frequency is high. In my experience glass is easily broken down in such work unless kept under low stresses. It would be an interesting study to follow the effects of high periodicity in the use of such dielectrics as glass as compared with oil and the like, increase of frequency seemingly conducing to the destruction of the one and adding value to the other. The phenomenon of soakage may have more or less to do with these different actions. Some specimens of lightly glazed earthen- ware exhibit very striking power of absorbing a charge. I have found dishes which, when subjected to the high potential discharges of a powerful frequency coil, or even to those of a Holtz machine, would continue to sparkle in the dark for a considerable time after- ward, and which would yield innumerable short sparks to the finger tip from various parts of the surface. The experiment is both curious and striking, being made all the 26 ELECTRICITY AT HIGH PRESSURES. more so by placing a layer of oil on the dish before subjecting it, to the discharge. The oil afterward moves about on the dish, owing to the charge working out from its interior. There can be no doubt that porcelain insulation which is not so thoroughly vitrified, as to be non-porous, will behave in the same way, and will in consequence be easily broken down upon the con- tinued application of high electric pressures, though at first it may appear to have a sufficient dielectric strength. The phenomenon of surface creeping, or imperfect conduction over surfaces of insulating material, is naturally of very great importance in limiting the insulation which can be maintained for high pressures. When a condenser such as a Leyden jar is charged beyond a certain potential difference between its foils, we note the appearance of thin purplish sparks at the edges of the foils, which sparks are all the more marked when the charging current is alter- nating or periodic. The tendency of any electric charge is to increase the capacity by spreading or other action. This spread- ing of a charge at high potentials is in fact one of the most important factors in limiting the pressure which may be maintained. It may extend in the case of the Leyden jar condenser so far over the glass above the coating as to cause discharge. This fact is, of course, well known. The same action, however, must be recognized as occurring at the edges of the foils, even when condensers are immersed in oil or surrounded by insulating liquids or solids. Its effect, however, under such conditions may be quite serious. I have noticed, for instance, that condensers consisting of coated glass or mica plates, subjected to alternating current charge and discharge at about 20,000 volts while immersed in oil, are subject to deterioration at the edges of the tinfoil coatings to such an extent that the oil is partly decomposed by minute sparks, and the solid dielectric cut away or etched at the same time. This eventu- ally leads to breakdown. In high pressure work, particularly in that involving large amounts of energy in transfer, as in power transmissions, the avoid- ance of dust or dampness upon the surfaces of insulation, over which a discharge or leak may form by creeping, is of the greatest impor- tance. So, also, the avoidance on either side of a circuit of sharp projecting edges or points near metal connected to ground, or to the opposite sides of the system, is imperative. To this end the high pressure conductors, where near together, are incased as much as possible in a sufficient thickness of insulator to withstand from two to three times the normal pressure for a certain time, such as ELECTRICITY AT HIGH PRESSURES. 2/ ten minutes, selected for the duration of test. Where the con- ductors are not incased, they are given a large separation far in excess of the normal striking distance between points at the potential difference involved. The practice of subjecting electric apparatus to tests with pres- sures much higher that those of normal use is now quite general, and conduces more than any other thing to security from break- downs due to loss of insulation. Apparatus intended to work normally at 40,000 volts may thus have to stand test pressures which involve actual striking distances of several inches. Besides the enormous extension of systems of power transmission by polyphase currents involving pressures of from five thousand to forty thousand volts, measured as alternating current, which really involve a much higher pressure depending upon the form of the wave, a number of other uses of high pressure electricity may be briefly alluded to. The production of Roentgen rays involves pressures ranging from about 40,000 to 150,000 volts, but the current is of course insignifi- cant. The higher the vacuum in the Crookes tube the higher the potential demanded to pass the discharge, though this condition may be complicated by other phenomena. I have in fact frequently found a tube in work to apparently pass instantly from the condition regarded as indicating lower vacuum to one of very high vacuum and the reverse. Certain constructions of tube permit the operator to control it instantly or choose in what way the tube be worked. The Roentgen rays, emitted from tubes demanding the highest potentials to work them, are recognized as the more penetrating or the less easily absorbed in metals or other substances. It is indeed an interesting speculation as to what might be the character of rays produced by potentials of a million volts, for example, passing through tubes the. vacua in which are made pro- portionately high. If the rule holds good, such rays should pass freely through most metals in considerable thickness; they should hardly affect fluorescent materials such as calcium tungstate or bar- ium platinocyanide; and the photographic plate should stop such a small percentage as to remain practically unaffected. The test of their presence would probably be their power, if still existent, of ionizing gases and causing electric conduction or convection through them. An old use of high pressure electricity to which more and more attention appears to be given is that of generating ozone in ozon- izers. This is so well known that it is needless to dwell thereon. 28 ELECTRICITY AT HIGH PRESSURES. Apparatus for cutting glass by spark perforation in the line of a desired fracture has been recently worked, and furnishes a curious instance of the application of high pressures accompanied by cur- rents or discharges so small as to be almost insignificant. The development of the wireless telegraph in the hands of Signer Marconi is another example which tends to a renewed interest in electrical apparatus for developing high pressure discharges. The possibility of signaling without wires and through the agency of electric radiation or ether waves was, I believe, pointed out by Hertz himself in connection with his beautiful researches in that field. The radiations sent out and picked up by the vertical wire of Marconi, and the response of the sensitive coherers of Lodge and Branly have already been the means of communication over dis- tances so great as to leave room for hope that hundreds of miles may not be insurmountable by this latest important development in the field of electricity at high pressures. It was early seen that the Hertz experiments would probably make it possible to signal at sea, despite storm or fog, and this expectation has been fully realized. The next ten years will doubtless witness a very great development in this fascinating field and mankind will be greatly benefited. There is very much yet to be learned concerning those vast natural exhibitions of electrical actions at high pressures which fill us with wonderment and awe. I allude especially to the light- ning and the aurora. And what have we to say. about that mystery " globular lightning" ? Its existence seems to be so well attested that we can scarcely doubt. Personally, I came so near seeing it in one instance that, if my eyes had been turned toward the north- west instead of to the northeast, I would have been a witness of it, for a companion who was looking the other way did see it, and called my attention just a moment too late. I did hear, however, the noise of the explosion of the ball which he saw slowly fall. The noise was not thunder, but merely a single explosion without rumble, and resembled the boom of a cannon of the old type. It is well known that the cause of the rattle and roll of thunder from ordinary flashes is the length of the spark discharge, the sounds from the various twistings and angles of which reach the ear suc- cessively on account of their varying distance from the observer. From this it follows, by the way, that, as with light in the case of the rainbow, no two observers receive exactly the same impression of sound in thunder. ELECTRICITY AT HIGH PRESSURES. 29 Auroral displays are shown to be probably dependent upon solar disturbances; an earthly coronal stream, perhaps, developing in response to some unusual coronal development on the sun, or to some vast sun spot disturbance. I am tempted to think that pos- sibly the flame gases of the sun actually reach the upper atmos- phere of the earth, and break down the insulation of the layers already under electric stress, or themselves bring electricity which disturbs the condition of our air. I am disposed to favor the former idea, however, on account of the temporary character of the discharges. The earth may in fact be brushed by an invisible prolongation of a coronal streamer, the effect of which acting like ionized gas, or flame gases, or gases through which an electric dis- charge has been passed, is to make the upper thin air conduct, and relieve its accumulated electric stresses in an hour or two, after which follows a period of comparative quiescence. Besides the solar corona, the streaming light of the comet's tail may indicate electric redistribution as the comet approaches the sun a view which has gained ground in the past few years; and there are not wanting astronomers who suspect that some of the light of nebulae may be electrical in its origin, or similar to that of the comet's tail. Nature does things on a grand scale, and her celestial electrical manifestations may not be unlike those which we are wont to call terrestial, although dependent perhaps upon the grander actions outside the earth's gas envelope at even higher pressures. RETURN IO the circulation desk of any University of California Library or to the NORTHERN REGIONAL LIBRARY FACILITY Bldg. 400, Richmond Field Station University of California Richmond, CA 94804-4698 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 2- month loans may be renewed by calling (510)642-6753 1-year loans may be recharged by bringing books to NRLF Renewals and recharges may be made 4 days prior to due date DUE AS STAMPED BELOW DEGI81995 I 20,000(4/94) PAT. JAN. 21. 1908 679936 UNIVERSITY OF CALIFORNIA LIBRARY