US2056660A - Self starting gaseous electric discharge device - Google Patents

Description

Oct. 6, 1936. 'r. E. FOULKE 2,056,660

SELF STARTING GASEOUS ELECTRIC DISCHARGE DEVICE I Filed July 1, 1932 2 She aets-Sheet 1 61 9. 1 c7119. 2

INVENTOR Oct. 6, 1936 T. E. FOULKE SELF STARTING GASEOUS ELECTRIC DISCHARGE DEVICE I Filed July 1, 1932 2 Sheets-Sheet 2 i a I Z9 54 Patented Get. e, ose

lose-e60 SELF STARTING GASEOUS ELECTRIC DESGEARGE DEVICE Ted E. Foolke, Nutlev, N. 3., essimor to General Electric Vapor Lump (lompany, Hobokeu, N. J,

a corporation of New olemy miollootiou duly i, 1932, Serial No $20,515

- discharge devices-generally, end more particularly to discharge devices operating with e long positive column discharge. it

A particular object of the invention is to provide a gaseous electric discharge device of the positive column type in which a, discharge may be initiated by the application off a. potential oi. less than lid volts D. C. between the electrodes thereof. A lurther object of my inventlou is to provide 7 s, novel method of operating on electric discharge device. Another object of my invention is to pro-- vide a novel method of forming e, coating of low work function. Still other objects and advantages of the invention will appear from the following detelled description thereof, or from an inspection of the accompanying drovvinos.

The invention consists in s, new end hovel elec tric gaseous discharge device, end ,in e, novel method of production and of operation thereof, es hereihofter set forth and claimed.

It is well known that electric gaseous olischoi ge tubes in theilnormal tin-ionized state ofifer it very considerable resistance to the pessege of on elec= trio discharge therethrough, it being necessary to ionize the gas or vapor before a positive column discharge can take place therethrough ot low voltage. To attain this ionization it is obviously necessary to provide moons for pioducins ions ct e rote which exceeds the loss of ions through recombination with electrons in space end at the wells of the deviw. In the post this has involved either application of on extremely high potentied to produce the required ionization, or the production of ions toy means of a. discharge to on auxiliary electrode across o gap which is suf= ficleutlv short to permit the initiation of as dis-- charge at the line pctentiel. The 'llrst'ei zpedient is obviously undesirable, since it involves the use auxiliary apparatus such a high voltage transformer of poor regulation or the equivuleut thereof; or a high frequency, high voltege mecliehism such as e lsheltou coil; or means to produce an inductive slug-e of the required potential.

Hence the use of the eiudlieiv discharge, ovoid,

leg as does the use of auxiliary eppeestus, hes theoretically been the more desirable oi these expedieuts. In practice, however, it leech found impossible to start oositlve column dis charges of eny'epprecielole length on the usual commercial circuits having potentials of the order of .110 volts A. fi-orD. (3., log the reason that the auxiliary discharge has beenlocailzed at the end of the tube, difiusion of the ions clone losing i'e-= lied upon to reduce the potential required to (@l, lid-124%) initiate the positive column in the tube. This difiusioo, however, tells so rapidly with the dis tame, the number of ions which reech e, distence of but three tube dlometers irom the discharge heme insignificant; Thus this method has reouired the use of potentials oi the order oi 220 volts or more, except where the length of the positive column was less then three tube diametei's. This mode of ionizing the gas has, there'- lore, never gone into commercial use due to the ohvious limitotious inherent therein.

I have new discovered, that by means of a novel orronoement of on electrlcclly conducting body within the dlschoree device the ouxiliery discharge may he made to travel along the tube, with the resultthot the entire discharge path is ionized with out little dependence upon difiuslon of the ions from the eusllioi'v discharge. In some coses this electrically conducting body may be electrically isoloted from the electrodes; in other cases it constitutes eh auxiliary electrode of novel sheoe sod disposition. But in either case o glow discharge is produced at potentials of llli volts or less, this slow hell 1e" ceused to screed throughout the length oi the tube with the i'esulthig ionization of the discharge path Fig. fl is on elevetionol view, in port section, of e further modification of the device oi l lo, 11, using it hot cathode, I

l ienxi is do elevotlouel view, in port section, of u modification of the device of 3, and

Fig. 5 is e Slldili&l view. oi u modification oi the device of i designed for use on alternating current In these dram-vines, with special ceieaeuce to Fig-r, i, there is shown on electric gaseous dis charge device having e tubular euvelogoe i of glass or other sultohle vitreous materiel" At one end of sold envelope i there is a cathode 2, of nickel or the like, which is prelemhlv tuloulel in. shape. Said cathode is supported lay on iuleed which is sealed through the end of sold envelope l. At the opposite end of sold envelope there is on ohode l of graphite, iron, or other suitable meteuse the applied potential.

' which is sealed through the cathode end of the envelope i, and preferably consists of a wire of nickel, molybdenum, or the like of suitable rigidity, say of 30 mils diameter. Said electrode 6 passes through the cathode,2 and extends up the arc tube to a point within a few millimeters of the anode l.- A support 8 which is welded to the electrode 6 near .the free end thereof serves to maintain the axial position of said electrode.

Additional supports of like nature may also be used at intermediate points along the electrode 6' if desired, in which case said electrode may be: of lighter construction. The inlead 5 iscon= nected directly to the positive terminal of. a suitable source of direct current, while the negative terminal of said source is connected through-a ballast resistance '9 to the inleadl and to the inlead I, the latter connection including a high resistance Ill.

The device of Fig. 2 has an envelope l which is similar to that of Fig. 1, and likewise has a -tubu-' lar cathode 2, supported by an inlead 3, at one end thereof. An identical tubular cathode l2,

. supported by the inlead i3, is located at the opposite end of the envelope I, however, replacing the anode of Fig. 1. A wire electrode 6, forming an extension of the inlead I likewise extends through the cathode 2, while a similar wire electrode l8, formed as an extension of the inlead l1, extends through thecathode il'said electrodes 8 and I s terminating with a short gap of the order of a, centimeter therebetween at -a point near the center of the envelope 1. Suitable supports-8 are provided to hold the free ends of each of the electrodes 6 and IS in the desired position; The

'inlead 3 is connected to'one terminal of a suitable source of alternating current throughthe ballast resistance, 9, and the inlead l'l is'com nected to the same terminal through said resist-' ance 5 and an added high resistance Ill; Simi-. larly the inlead i3 is connected directly to the pure barium by the use of heat after the device has been evacuated. The cathodes 2 and I 2 are,

7 of course, easily heated in the-usual manner by an induction furnace to the temperature neces sary to decompose the azide coating, but the wires 6 and I6 have so little mass that such heating thereof is diflicult. I have discovered, however, that the gas evolved during the treatment of the cathodes is sufficient, even though the device be still connected to an exhaust system, to support an inductive discharge" about the adjacent part of the wires 6 and "i by which said wires-are heated to the required temperature, the gas evolved during the reduction of the aside on this part of the wire being utilized to support a dis charge on the adjoining part thereof, and so on,-

the induction coil or furnace being moved along the device at a pace which keeps up with the evolution of 'gas'from the azide coating on the wire. This procedure leaves a coating of metallic barium on the surface of the cathodes 2 and II and of the wire electrodes 6 and 16, this barium beingintermixed with some nitride and a trace of unreduced azide which function to cement the barium to the surface of the electrode. The surface of these electrodes is then further improved by distilling metallic caesium or other it alkali metal thereonto. This is conveniently done by heating the pellet II, which contains a mixture of a caesium'or other alkali metal compound and a reducing agent, such as s. mixture, of'caesium dichromate and silicon. This heating l0 obviously causes caesium vapor to be produced, some of this vapor condensing on the desired electrode surfaces. When so preparedthese surfaces arefound to have an extremely low work function, throughout'a long useful life of a. thousand hours or more. As an alternative to this coating process the electrodes and the wires'may be coated with a mixture of barium and caesium azides, as

disclosed in my Patent 1,965,588, granted July 10, 20 n 1934. According to the novel procedure there disclosed the barium azide is first broken up by heating to 180 C. following which the caesium azide is partly broken up by 'a' gaseous'dlscharge, producing a surface useful life. I V

The envelope I of the" devices of both Fig. 1 and Fig.- 2 is filled witha suitable gaseous atmosphere such as neon, helium. argon, nitrogen,

mercury vapor or the like. These gases may be 30 used alone or in combination with each other, as desired. For certaini purpose, however, I prefer to use a combination of gases, such as proposed in my Patent 1,990,175, granted February 5, 1935, inwhich the principal gas is modified 35 by a trace of another gas having an ionizing potential which is less than the meta-stable potential of the principal gas. For example, in a neon lamp 'I prefer to use a mixture of neon with about 0.2% of argon. In any case the pres- 40 sure of the gas is a compromisebetween the low pressure at which the positive column. is most efllcient and the somewhat higher pressure .at

which a glow discharge is most easily produced.

In practice I use a pressure of the order of 3.6 45

m. m. of mercury, although for specialpurposes' somewhat higher or lower pressures may be used if desired.

' The lamp shown in Fig. 3 has a tubular envelope 20 of glass or other suitable vitreous 60 material. At one end of said envelope 20 there is an anode 2| of nickel, iron, tungsten, graphite or the like, which is supported by an iniead 22.

This anode may be of any desired shape, but has been illustrated as cup-shaped, since in certain 55 cases a lower anodev fall results from the use of such an anode. At the .opposite .end of the envelope 20 there is a thermionic cathode 23 which preferably consists of a cellular electrode which closely'surrounds a heater 24. Said cathode II o0 is itself surrounded by the tubular heat conserving shields 25 and 20 to which it is electrically connected, the heat shield 28 preferably extending some distance beyond said cathode 23. Said heater 2 is designed for operation on 110 volts, c5

hence the heating coil thereof (not shown) is completely embedded in a suitable refractory material, such as alundum, in order to prevent the formation of an are between different-portions thereof. connected to an inlead'fl, which is also shielded in a suitable manner to prevent the formation of an arc thereto, while the other end of said heatlng'coil is electrically connected to the cathode 23. The inlead is attached to the heat shield this characteristic being retained 15 having an unusually long 25' One end of said'heating coil is 70 26, and hence is electrically connected to the cathode 23. Said cathode 23 and the heat shields 25 and 26 are conveniently formed of nickel or the like. The inner surfaces of the cathode 23 are preferably coated with an alkaline or alkaline earth metal or compound thereof, in order to reduce the work function ofthese surfaces. For instance, I prefer to use a mixture of barium and strontium oxides, these oxides being conveniently reduced from the carbonates in situ by the use of heat in a well known manner. In some cases I also coat that portion of the inner surfaces of the heat shields 25 and 26 which extends beyond the cathode 23 with a material, such as powdered aluminum, which is known to resist activation by particles sputtered from the active cathode surface, in order to permanently confine thedischarge to the desired surfaces within the cathode 23. A third inlead 20 which is sealed into the cathode-end of the envelope 20 extends beyond the heat shield '26, the end thereof terminating in a ring 30 which is concentric with, and of about the same diameter as, the heat shield 26 and which lies in a plane close to the end of said shield. Said ring serves as an auxiliary anode, and hence is preferably made or" nickel, iron, tungsten, or other suitable material. to confine the discharge to the ring 250 a vitreous sleeve at is preferably placed about that portion of said inlead 2d which extends alongside of the heat shield 255. A conductor 32 of suitable rigid= ity is attached, preferably by welding, to said inlead 29 at a point near the ring 30, said con ductor extending up the arc tube to a point adjacent to the anode 2i, the free-end of said conductor being held in the desired position within the envelope 20 by a support The inleads 22 and ii are connected directly to the positive terminal of a suitable direct current source, while the inlead 29 is connected to the same terminal through a high resistance 3 .1. The inlead 28 is in turn connected to the negative terminal of said source through the ballast resistance oil.

The conductor 32 is preferably coatedwith a substance having a low work function, such as barium, this coating being conveniently produced by the same method, set forth in detail hereinbefore, by which the coating is produced on the wires and it.

The device shown in Fig. 4 is in most respects identical'with that shown in Fig. 3. The conductor 32 is, however, electrically isolated from the inlead 29, although it preferably extends to a point within a few millimeters thereof. Said conductor 32- is conveniently held in place by the ring support 36 and the support ii, to the latter of which it is welded, said supports being fused into the wall of the envelope Said conductor 32' is coated with barium or the like in the same manner as the conductor 32 of Fig. 3. An alternate arrangement of the operating circuit is also shown in this figure, the ballast resistance 35 here being included in the positive lead, while the inlead 2g is connected through the high resistance lid to a point between the resistance 35 and the inlead 22.

The device shown in Fig. 5 is identical with the device of Fig. 4, save that the anode 29 of Fig. 4 is replaced by a thermionic electrode assembly which is similar to the cathode assembly of Fig. I

i. In this modified structure the inlead 22 is connected to one end of the heater 38, the other end of said heater being connected to the outer shield 39 about the thermionic cathode (not shown).

In order previously referred to herein.

The heater 38 is furthermore so proportioned as to carry the main discharge current. The inlead- '22 is connected to one terminal of a suitable alternating current source through a resistance 35, of less resistance than the resistance 35, it being understood, however, that in some cases,

where the heat shield 39 has sufiicient heat rahigh resistance 36, while theinlead 21 is also connected directly to the same terminal. The other terminal of said source is connected to the inlead zfl. 1

In each of the devices shown in Figs. 3-5 the envelopes 20 are filledwith a suitable gaseous filling such as has been described in connection with the devices of Figs. 1 and 2. Due to the greater facility with which the initial breakdown occurs with this type of device it is, however, possible to use somewhat lower pressures than are employed with the devices of Figs. 1 and 2, with a corresponding increase in luminous efficiency, pressures of 2-4 m. m. of mercury being found especially effective where neon is employed, and 7 even lower pressures, of the order of 1-3 m. m. of mercury being found suitable if a trace, of say .01 to 1% argon is added thereto, in accordance with the principles outlined in Patent 1,990,175,

low pressures are preferably used where other gaseous fillings are employed.

In the use and operation of the device of Fig. 1 upon application thereto of a suitable direct current potential, say of 110 volts D. 0., breakdown of the gas space occurs between the anode i and the adjacent end of the conductor 6 with a resultant glow discharge, this breakdown occurring at a relatively low potential, despite the low pressure of the gas, due to the low work function of the coating on said conductor. Due to the well known phenomenon that a glow discharge always tends to extend over such an area on thecathode as to maintain a fixed current density (for any given gas pressure) at which the normal cathode fall is maintained, the glow discharge immediately extends down the length of the conductor b toward the cathode 2, the resistance l0, of the order of 10,000 ohms, being so proportioned as to limit the current to substantially that value which will cause the'glow to completely cover said conductor 6. This glow discharge, of course, causes a copious ionization of the gas about the conductor 6 throughout the length of the arc tube with the result that a positive column discharge occurs virtually instantly between the' cathode 2 and the anode 4, this latter discharge having the'usual characteristics of an arc whose current is limited to a desired value by the ballast resistance 9. .Where a mixture of gases is employed, as disclosed in my co-pending application,

.Serial No. 541,021, which has been hereinbefore referred to, it is obvious that the radiations from the glow discharge will be effective to produce Correspondingly.

4 I the aforesaid neon-argon gas. mixture is employed.

While I have described the device of Fig. 1 as being operated on direct current it is obvious that it can be operated on alternating current if so. desired. In this case a positive column discharge will occur every alternate half cycle, this halfwave discharge being reinitiated once each cycle in the manner which has been described above, as is readily apparent.

Upon application of a suitable alternating current potential to the device of Fig. 2 a glow discharge starts on each half cycle at three difler-.

'' ent points: between the ends of the conductors 6 and I6 between the electrode 2 and the condoctor 6, and between the electrode 12 and the conductor l6. Assuming that at a given moment the electrode 2 is negative with respect to the electrode l 2, it is obvious that a cathode glow will coven said electrode 2, and that the glow discharges at each end of the conductor IE will spread along the entire length of said conduc- Where especially long tubes, of the order of tube diameters in length, are operated on 110 volts the starting of the main discharge is sometimes delayed, however, until the next half cycle. The electrode I2 is then covered with a cathode glow, as is the entire conductor 6, with a resulting ionization of the gas thereabout. In case the main discharge was initiated on the first half cycle, it is obvious that it will be immediately reinitiated in the reverse direction as a. result of this ionization in the same manner described above. Where the discharge is not initiated during the first half cycle, however, there is considerable ionization still remaining about the conductor l8 after the potential has reversed, this residual ionizatipn cooperating with the positive ionswhich difluse into that portion of the tube to permit the initiation of the main discharge from the acting anode 2 to the acting cathode i2. The vreinitiation of the discharge oneach successive cycle thereafter is, of course, obvious. Where mixtures of gas, such as the neon-argon mixture hereinbefore mentioned, are used even longer positive column discharges, of the order of 24 .tube diameters inlength, may be readily initiated on 110 volts A. C. In this case the radiations from each of the glow discharges pass up the tube exciting atoms of the main gas, many of these excited atoms passing into a metastable level as a result of collisions. Upon collision of any of these metastable atoms with an atom of a gas of lowerionizing-potential the energy is transferred, with'a resulting ionization of the gas.

The ionization thus produced throughout the tube, added. to the ionization directly produced by the glow discharges, greatly facilitates the initiation ,of the main discharge and permits even longer positive column discharges to be initiated on a given voltage than would otherwise be possible.

In the 'operation of the device of Fig. 3, upon energization of the leadsthereto with directcurrent of suitable potential, say 110 volts'D. C., the cathode heater 24 is immediately energized and that there will be no appreciable discharge therebetween until the cathode 2 has been heated to its normal operating temperature. In this way premature initiation of the main arc discharge, before the free electron emission from the cathode 2-3 is sumcient to support the normal discharge 10 current without the formation of destructive localized hot spots on said cathode, is avoided. As the temperature of the cathode rises the electron emission therefrom rapidly increases, altering the potential distribution between said cathode and 15 the auxiliary anode 30 to such an extent that eventually the gas therebetween is sufliciently ionized to permit an arc discharge therethrough at line potential. As soon as this occurs the potential of the auxiliary anode 30 is lowered by the resistance dropin the resistance 34 to substantiallyspace potential. Where this auxiliary electrode is made relatively large, as is preferred, so that the random current which could be collected thereby, if it were not for the resistance 34, is greater than the current actually collected thereby, there is a negative anode fall, as is well known, with the result that said auxiliary anode operates at a potential somewhat less than that of the surrounding space. For example, ina neon lamp the space potential about said anode is about 21 volts, while the negative anode fall is easily made of the order of 5 or 6 volts, with the result that said auxiliary anode operates at a potential of the order of 15 volts above that of the cathode.

This low potential of the auxiliary' anode 30 is highly important, since the lowerthe potential thereof the greater is the potential difference available to initiate a glow discharge. between the anode 2| and the wire 32 40 ;as a cathodeas the next step. in the initiation of the main discharge. Where the potential of the auxiliary anode 30 is thus 15 volts, 95 volts is obviously impressed between said anode 2| and the wire 32, where the line potential is 110 volts. Due

to the low work function of the coating on the surface of said wire 32 this potential is adequate to initiate a glow discharge therebetween, this glow discharge. rapidly spreading down said wire toward the cathode 23, and thus ionizing the gas throughout the length of the tube. As a result of this ionization the main positive column-discharge, of the order of 18 tube diameters in length, is immediately formed between the mainanode 2| and the cathode 23, and continues so long as potential is applied to the device.

In the case of pure gases the initiation of the aforesaid cathode glow discharge to the wire 32' is facilitated by the radiations from the discharge between the auxiliary anode 30 and the cathode 23'. These radiations reach the end portion of the wire 32 and as a result of their photo-electric effect thereon cause a certain amount of electron emission therefrom", and thus permit the initiation of the desired glow discharge at a potential 55 lower than would otherwise be possible. As a result my novel device readily functions on 110 .6 volt direct current lines despite appreciable voltage fluctuations therein. By the addition of a trace of another gas, such as argon where the main gas is neon, having an ionizing potential slightly less than a metastable potential of the main gas, as disclosed in my copending application, previously referred to, the radiations from the auxiliary discharge to the hot cathode ,23

aosaoeo also produce an appreciable cunt of ionization throughout the length of the tube, further iacili toting the initiation of the cathode glow discharge to the wire 82, and then cooperating with said glow discharge to produce the necessary ioniza= tion for the initiation of the main positive column discharge Thus even longer positive column discharges may be initiated on 110 volts D. C. with such a mixture of gases than is possible with the use of a pure gas alone in my novel lamp. Due to the inclusion of the ballast resistance as in the cathode lead the voltage on the heater to is reduced as soon as the main discharge is initiated thus reducing the energy input thereto.

While the various steps in the starting of the discharge in this device have been described as taking place sequentially it is to be understood that once the cathode as has reached such a temperature that the discharge irom the auxiliary anode til is initiated the remaining phenomena take place virtually instantly, there be no perceptible time lag. As a. result this device may be satisfactorily operated with a hali wave dis= charge on alternating current, if desired. In.

this case somewhat longer positive column discharges can be readily initiated on 110 volts, due to the greater penis voltage available.

The operation of the device or mg. l is in many respects similar to that of the device oi? Fig. 3. Upon application or a suitable potential, such as no volts D. 6., the cathode as is heated in the same manner, with resultant discharge thereto from the auxiliary anode to as soon as the then mionic emission therefrom has reached the desired value. This discharge ionines' the gas in the immediate vicinity thereof, creating a plasma whose space potential, in neon, is oi the order oi 21.5 volts above that of the cathode. The end of the wire 232' projects into this plasma, and has been found to assume a potential loss than that of space of the order of e times the average energy, in volts, oi the random electrons inthe 4 gas. 'ihus in neon the end of this wire has a potential oi from 3-7 volts less than that or the surrounding space, or 14.5-48.5 volts above that .of the cathode 23. Due to the fact that the wire 32 is a good conductor the entire wire assumes this potential, so that in a no volt lamp from so-os volts is impressed on the gap between said wire and the anode 28. This potential causes the formation or. a cathode glow discharge therebetween, this discharge extending down said wire 32' and ionizing the gas throughout the length of thetube, permitting the main discharge to be initiated by the line potential in the same manner as in the device of Fig. 3., From the foregoing it is obvious that it is desirable that the wire 32 should float at as low a potential as is consistent with the collection of enough electrons from the cathode 23 to support the glow discharge.

Upon applying a suitable alternating current potential to the device of Fig. 5 a halt-wave discharge is initiated therein on every half cycle that the cathode as is negative in exactly the same manner that the discharge is initiated in the device of Fig. d, as described above. The discharge current passes through the heater to, heating the associated electrode assembly, and there is also some anodic heating thereof. its a result the oxide coated portion of this assembly soon freely emits electrons whenever it is-negative, permitting the initiation at line potential of the main discharge in the reverse direction through the one column, which retains some ionizatl ir'om the previous halt cycle. Thus this device starts with n half-wave discharge and then changes over to a iull-wavedischarge de vice after the second thermionic electrode has mixture hereinbefore mentioned may obviously be used in either the device of Fig. 4. or that of Fig. 5 with beneficial results, such as increase in the maximum length of discharge which can be initiated with a given voltage. The operation oi the device in such a case is obvious from the fore going description and hence will not be further described.

While I have described my invention by reference to specific embodiments thereof, it is to be understood that it is not limited thereto, but that various changes, substitutions, or omissions. within the scope of the appended claims, may be made therein without departing from the spirit thereof. I

I claim as my invention:

1. An electric gaseous discharge device comprising a tubular sealed vitreous envelope containing a gaseous atmosphere, an electrode sealed into said envelope at each end thereof, a conduc ing wire extending within said envelope from a point near one of said electrodes toward the other or said electrodes, said wire being coated with an alkaline metal,- and means to electrically connect said wire to a single one of said electrodes, whereby substantially all of a potential which is impressed between said electrodes is concentrated over a short path to initiate a discharge in said device.

2. An electric gaseous discharge device comprising a tubular sealed vitreous envelope con-- taining a gaseous atmosphere, an electrode sealed into said envelope at each end thereof, and a conducting wire extending within said envelope from a point near one of said electrodes toward the other or said electrodes, said wire being coated with an alkaline metal and connected to one or said electrodes through a resistance which is isolated from the path of a discharge between said electrodes.

3. An electric gaseous discharge device com- 18 prising a tubular sealed envelope, a main electrode at one end of said envelope, said electrode being adapted to emit electrons when heated,

means to heat said electrode, an auxiliary anode adjacent to said main electrode, another main electrode at the opposite end or said envelope, and an electrical conductor extending within said envelope from a point near said lastmentioned electrode to a point near said auxiliary anode.

4. An electric gaseous discharge device comprising a tubular sealed envelope, a main electrode at one end of said envelope, said electrode being adapted to emit electron when heated, means to heat said electrode, an auxiliary anode adjacent to said main electrodaanother main electrode at the opposite end of said envelope,

and an electrical conductor extending within said envelope from a point near said last mentioned electrode to said auxiliary anode, said conductor being electrically connected to said anode.

5. The method of producing a coating of low work function on an electrode which comprises coating said electrodewith an alkaline compound which decomposes with the evolution of gas when heated, sealing said electrode into an envelope, evacuating said envelope, heating a portion of said electrode to a temperature sufioient to de= compose'said compound and then producing anlid inductive discharge in the gasps evolved to heat more of said electrode to a temperature suillcient to decompose said compound.-

'6. An electric gaseous discharge device comprising a tubular sealed envelope containing a gaseous atmosphere, an electrode sealed into said envelope at each end thereof, a conducting wire extending within said envelope i'rom a point near one of said electrodes to a point near the other thereof, said wire being connected to a single one of said electrodes through a resistance which is isolated from the path oi a discharge between said electrodes.

"I. An electric gaseous discharge device comprising a tubular sealed envelope containing a gaseous atmosphere, 2. thermionic cathode at one end of said envelope, an auxiliary anode adjacent to said cathode, another electrode at the oppoplasma of a discharge between said auxiliary 1o anode and said cathode.

Images