On the Nature of Magneto-Cathodic Rays Proceedings of the ... · their production, have led Villard and Righi, ainong others, to regard then1 as a new kind of rays, and the name magneto-cathodic

On the Nature of Magneto-Cathodic Rays

H. Thirkill

Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical andPhysical Character, Vol. 83, No. 563. (Mar. 2, 1910), pp. 324-334.

Stable URL:

http://links.jstor.org/sici?sici=0950-1207%2819100302%2983%3A563%3C324%3AOTNOMR%3E2.0.CO%3B2-A

Proceedings of the Royal Society of London. Series A, Containing Papers of a Mathematical and Physical Character iscurrently published by The Royal Society.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtainedprior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/journals/rsl.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.orgMon Mar 10 12:05:35 2008

0.12 the Na tu~e of Magneto-catlzodic Rays.

By H. TIIIRKILL,B.A. (Cantab.), B.Sc. (Lond.), Research Scholar, Clare College, Cambridge.

(Communicated by Prof. Sir J. J. Thomson, F.R.S. Received December 20, 1909,-Bead January 27, 1910.)

I j ~ t ~ o d z ~ c t i u , ~ .

As early as 1535, Pluckerx observed that, when a tube in which an electric discharge is passing is placed longitudinally in a strong magnetic fielcl, a hand of light is seen, which has the cathode for its section, and for its direction the line of inagnetic force passing through the cathode. Until compal.atively recelltly the nsually accepted explanation of this light was as follows : Under the influence of the magnetic field the cathode rays are bent into the form of a spiral or helix round the lines of magnetic force. Now these primary cathode rays produce secondary rays by collisions with the molecules of the gas, and both the primary and seconclary rays produce fluorescence in the gas. The fornler alone would show llie spiral natnre of the path, the latter cause a decrease in the definition of the light, which IIOW appears as a band.

I n the meantime, othcr physicists have observed the same phenomenon. Attempts have been niacle to detect an electric charge on the rays, but ~vitll- out success. This lack of charge and tlie necessity of a magnetic fielcl for their production, have led Villard and Righi, ainong others, to regard then1 as a new kind of rays, and the name magneto-cathodic rays (Villard) or magnetic rays (Righi), has been assigned to them to indicate their origin.

In a recent meinoir, Righi has put forward the hypothesis that these rays consist of neutral doublets : a negative corpuscle revolving rouncl a posilive ion. This would explain their lack of charge, their behaviour in a magnetic field, arid some other properties of the rays.

I t was the aim of the author, at the suggestion of Prof. Sir J. J. Thomson, to investigate the nature of these magneto-cathodic rays, and especially to see if they carry an electric charge.

The experiments described below seem to show tliat magneto-cathodic rays consist of slowly inoviilg negatively electrifiecl particles. Under the influence of the powerful magnetic field, these have been coiled up to forrri a very fine spiral along the lines of magnetic force passing through the cathode.

* 'Pogg. Ann.,' vol. 103, p. 88, 1868.

325 On the ivature of Magneto-cathodic Rays.

The reinaining part of this paper is described under the following heads :-(1) Historical Note ; (2) Experiments with Magnetic Fields ; (3) Charge

carried by the Rays ; (4) Discussion of Results ; (5) Summary.

I t will be useful to give a short account of the previous work on this subject, in order to see what has led to the idea of a new kincl of rays.

Reference has already been made to Plucker's work. Using a simple tube with aa electrode at each end, placed longitudina,lly in a magnetic field, he observed a band of light which assr~med the forxu1 of a tnbe of magnetic force limited by the outline of the cathode.

Soine years later EIittorf* continued these experiments. As theory would lead us to expect, Hittorf found that if the magnetic field be uniform, the path of the cathode rays is a helix drawn on a circular cylinder, the axis of which coincides with the line of magnetic force passing through the starting point of the cathode particle. When the field is not uniform the path is still a helix, following approximately the lines of magnetic force. These forrns were observed by Hittorf, and, in addition, the light along a tube of magnetic force as described by Pluclter.

Brocaf and Birlteland,; working on the same lines, observed a sudden change in the character of the discharge for a certain value of the variable inagnetic field. A band of light appeared along the line of magnetic force passing through the cathode, and this was accompanied by a great fall in the potential difference between the terminals of the tube.

Villards has deecribed some interesting experiments on the influence of a magnetic field on this part of an electric discharge. He noticed that the distance the light extends into the tube increases with the magnetic field. The latter appeared to be not only the source but also the motive power of the rays, and the name magneto-cathodic was assigned to them. Villard also attempted to discover if the rays carry any electric charge, and though in his experiments he could detect none, he admits the possibility that this may be due to the high conductivity of the gas through which they pass. He further showed that an electric field, acting at right angles to the magnetic field, deflected the rays at right angles to the directions of both fields.

I n Righi's experiments11 a tube was placed in a magnetic field, the intensity

* 'Pogg. Ann.,' vol. 136, p. 221, 1869. t ' Comptes Rendus,' p. 736, 1898. 1 'Comptes Rendus,' vol. 126, p. 586, 1898. $ ' Le Radium,' April, 1906. j j 'Journal de Physique,' vol. 7, pp. 589-617, August, 1908.

of which could be varied at mill. In the abeence of the nlagnetic field the discharge tube presented very little luniinosity, bnt as the magnetic field was increased, for a certain intensity the gas became luminous. The appearance of this lulninofiity was usually, though not always, suclden. The extent of this light, as Villard observed, Ivai affectcil. by the magnetic field. Further, he fonncl tliat, xvhile thc potential difference between the electrodes was always affoctecl when these rays appeared, the direction of the charge was not constant, there Leing sollietimes an increase, at other times a decrease, in value. 1i;xperirnentiag on the charqe carricd by the rays, he found that in thc absence oi the niagnetlc field, and for slnall values of the same, his electrometer inc1ic:tted that an electric charge was cntering the Faraiiay cylincle~ placed in the tulue ; but when the magnetic rays as he calls them appeared, a charge could no longer be detected. From this he concl~icled that the rays convey no charge.

S~mlnling np, then, i t may be said that the chief points indicating the appearance of tl. ilew liincl of rays are :-

(1) The clixcontinuons passage of one phenomenon into the other. (2) The rays only appear in a magnetic. field, which also seems to be the

motive power. (3) .A change in the potential of discharge acconlpanies the appearance of

the rays though the vacuum is unallerecl. (4) Tile rays do not appear to charge a Fara~lay cylinder when an electyo-

meter is used. (3) They arc deflected by an electric field in a direction at right angles to

both the elcctric and magnetic fields. These will be discussed in the latter portion of the paper.

It was tlrouglit that some light might be thrown on the nature of magneto- cathoclic rays by using a vely fine beam of cathode rays and a i~lagnetic field the intensity of which could he varied very gradually.

App"lvt~is.-The apparatas used in the earlier experiments is s h o ~ n in fig. In. The cathocle and anode are at C: ancl A respectively. In some experiments the cathocle consisted of all aluminium wire, 1 to 2 111111. in diameter, in a glass tube, the encl of the wile being flush with the end of the tube, fig. l h . Another form is shown in fig. l c : the aluminium disc in this case varied from 3 to 6 mm. in diameter. Both fornls of cathocle fnrliish a fairly fine beam of cathode rays when the pressuie is low enough. The anode was an alumii~ium wire placed in the side tube A near C , so that there

327 1909.1 On the Nc~ture of Magneto-cathodic Rccys.

would be only a small electric force acting on the rays in G ; such an electric force would oppose the formation ancl existence of the cloublets which Righi supposes to exist.

The glass tube G was about 4 cm. in diaiiieter and the encl was closecl by a plate of glass on which was fixed a willemite screen S. The cathode rays impinging on this screen give rise to a small patch of fluorescence which enables one to follow the niovernents of tlie rays. D was a mica diapliragm used in sonie experiments. I t was pierced centrally and excentrically, ancl the surface next the cathode was covered with willeii~ite. The tube B connected the apparatus to a Topler pump, a McLeod gauge, and a P2Q5bulb,

FIG.la.

FIG.l b . FIG.1 ~ .

The apparatus was placed between the poles of ail electromagnet. The latter was a large horse-shoe magnet. The pole pieces were 6 cni. square and their distance apart varied from 6 to 1 0 cm. One pole piece was drilled to admit the tube C. The current flowing round the nagn net could be varied by ilieans of a wire resistance and a liquid resistance, the latter allowing very sniall and continuous variations in the current, and hence in the magnetic field. The intensity of the field in the aeighbourhood of the cathode reached a value of 1100 lines per square centimetre. I t may be nieiltioned that this is equal to the largest value used by Biglii.

An induction coil was used to supply the electroiiiotive force. The ordinary discharge not being very steady, a small rectifier was put in the circuit.

328 Mr. H. Thirkill. [Dec. 20,

The gas used ill the majority of the experiments was air, but oxygen was also tried*; the pressure varied froin 0.6 to 0.1 mm.

Action of ct zViujnetic Pidicld.-The pressure of the gas having heen sufficiently reduced, a Cairly fine beam of cathode rays passes down the tube G ; the positive column occupies the side tubc A.

IVhen the current ia the magnet circuit is started and graclnally increased, tlie cathode rays are bent into the form of a spiral, the diameter of which graclually diminishes as the strength of the nlagnetic field is increased. The coiling of the rays could be observed directly, since the rays render the gas fluorescent; or by observing the lnovemcnts of the fluorescent patch on the willemite screen. As the strength of the field was still further increased, the spiral gradually closed up until a band of light was left, the direction of which was sensibly that of a tube of force having the cathode for section, There were in addition other spirals of greater radius. These represent either rays of different velocity or different initial direction. Ry introducing the screen I), it was possible to have the bundle along the line of force and. in addition, a xx~ell marked spiral bundle passing through the second aperture.

When the cathode bundle was not so fine, the effect of the niagnetic field was to concentrate the light along the line of force through tlze cathode. I11

the absence of the field the tube was comparatively dark, but on putting on a strong field, the intensity of the light greatly increased, though its lateral extent decreased. The effect appeared to be a gradual one and no discon- tinuity could be observed. When the pressure \\-as fairly high and tlie beam did not reach the extremity of the tube, the extellt of the beam increased as the field was increased.

The discharge from a coil is not very steady and is not uniclirectional. This difficulty can be obviated by using a valve of the Villard type, or by using a lime cathode. A speck of lime when heated gives off an abundant supply of negative corpuscles, and when used as a cathode, the discharge is very steady and is unidirectional. By this means i t was also possible to obtain a very fine beam of cathode rays, the cathode being a very snzall speck of lime on a narrow piece of platinum foil. The apparatus is shown in fig. 2, where C is the lime cathode. The rest of the apparatus is the same as before.

The rays were coiled up by the magnetic field, but on account of the fineness of the bearn, i t was possible to observe the coiling much further; even when the beam along the line of force had a diameter of the order of a millimetre, i t was possible to observe the spiral nature of the beam. With a

* Slnce sealing wax was used in certain cases, the gas present would contain carbon monoxide.

1909.1 0%the 2CTaturc of ilday~cto-cathodicRays. 329

number of tiny specks of lime, it was possible to exhibit the magnetic lines of force in the form of very close spirals.

Efeet of cn S2~bsidinl.yMagnetic Field o n the Beam of Rcbys.-As has been stated, the bean1 of light follows the lines of magnetic force. If a piece of soft iron is brought near, the bean1 is attracted. The iron distorts the lines of force, tlse rays following the new direction. I n a similar manner by bringing np a niagnet, attraction or repulsion can be produced a t will.

Thns these bundles of rays act very much like those of Villard and Righi. G~aclzbalAppearance of the 3cbys.-In these experiments, the appearance of

the rays was always gradual. Using an ordinary discharge and a diffuse beam, the first effect of the field was to concentrate the rays along the line of force-in this case parallel to the axis of the tnbe. The lulninosity in this way was greatly increased, but, in every case, by gradually increasing t l ~ e magnetic field, this increase of luislinosity was observed to be gradual. The concentration continued until a fairly fine heam was obtained along the lines of magnetic force in which the spiral nature could no longer be detected.

Using a fine beam of cathode rays from a linie cathode, the coiling could be observed with greater ease, both directly and by observing the patch of phospl*oresceiit light on tlse willemite. At no time mas there observed ilny discontinuity, the ljght along the tnbe of force consisting of slow cathode rays in the form of a veyy fine spiral.

I11 many, but not all, of Righi's experiments, a discontinuity was observecl,

330 Mr. H. Tllirkill. [Dee. 20,

the magnetic rays appearing sucldenlg. It will be shown later that this sudden change in the character of tlie discharge call he explainecl on the assumption that the rays consist of catliotlc rays in spirals.

The experimellts described above sceril to show that rays acting likp magneto-cathodic raps consist of slowly inoving cathode rays which ha\-e been bent into spirals of very sinall step ancl mclius. Such being the case, the rays ought, to carny an electric charge, and experiments were made to see if any charge coulcl be detected. The apparatus used for this purpose is shown in fig. 3. C is the lime cathotle, A the anode, F is a Paraday

cylinder. I t consists of ail inner insnlatecl brass cylinder F, sl~rrounded by an outer earthed cylinder G of the same material. There mas a small circular opening in G, 1mm. in diameter, and a slightly larger one in I?. The insulated cylinder F mas connected to a Nalder galvano~neter (1 nim. a t 1illet,re corresponcling to 10-l%mpere). The use of a galvanometer instead or an electrometex has sonie advantages.

In a discharge tnbe there are present ioils of both signs ; these give rise to the condnctivity of the gas. A. charged conductor placcd in such n gas mill gmdnally lose its charge ; ions of sign opposite to that of the charge on the concluctor will approacli uncler the influence of the electrical fieltl set up in the aeighbonrhood. If the cylinder in the experinlent receives a charge while insulated, we have two processes at work : a charging up clue lo the

1909.1 On the N a t z ~ r e o f L?rlagr~eto-cuthodic Rays. 331

arrival of chargetl particles, ancl a discharge due to the condrtctiuity of the gas. An equilibrium will be reaclieil when the rate of discharge is eclual to the rate of charging, and the greater the number of ions present the lower will be thc linliting potential. If there is a large number of ions present ; in other words, if the conductivity of the gas is very great, the linliting potential will soon he reached.

An electrometer was first used, being connec'ced to F by wires enclosed in earthed metal tub~js. The front part of G was covered with willernite, and the rays could be deflected into and out of the cylinder at will, by means of an external magnet. I t was not lourid possible to detect, with any certainty, a charging up of the cylinder.

I t was thought that this rniglit be due to the high conductivity of the gas as explained ahove (the pressure was fairly high), and the Nalder galvanometer was substituted for the electrometer. One terminal was conr~ected to F and the other to earth. The cylincler now loscs a portion of its charge through the galvanometer, and the current through the latter will be proportional to the potential of the fornzer. The conductivity of the gas is now not so important.

It was now found on deflectii~g the rays into the cylinder that the galvanoniete~ gave a deflection indicating a negative charge. These rays then carry a negative electric charge, and therefore consist of slo\vly moring negatively charged particles which have been bent into the form of a very fine spiral by the nlagiiet field.

It may be nlentioned that $he spectrnin of the dischalge was observed, but no trace of any new rays was fonnd, the ordinary cathode-ray spectr~un being obtained.

(4) Discussion of Results.

The exyeriinents described above seem to s~lpport the view that the magneto-cathodic rays consist of slowly moving negatively electrified particles which have been coiled into a fine spiral. How can we explain tbe results of Villard and Righi on this hypothesis? Briefly summarised, these are :-

(1) The rays only exist in the presence of a magnetic field. ( 2 ) I n many cases, but not always, the appearance of the rays is sudden,

and the distance they extend from the cathode increases when the field is increased.

(3) Siniultaneously with this change in the character of the discharge there arises a difficulty in detecting any charge on a Baraday cylinder placed in the tube (an electrometer being used).

332 Mr. H. Thirkill. [Dec. 20,

(4) The appearance of the rays is acconlpanied by a change in the potelltial of discharge.

(6) The magnetic rays are deflectecl by an electric field in a direction a t right angles to both magnetic and electric fields.

On this hypothesis the cathode rays are coiled up hp the magnetic field, the magnitucle of the field required depending on the velocity of the rays.

Consider a simple clischarge tube such as that shown in the fig. 4, with an electrode at each end, acted on by a longitudinal magnetic field. Cathode rays spcecl along the tube. Except in the case of those particles inoving in a

ciirectioiz parallel to the field, tlie effect of tlie latter is to bend the rays into spirals round the lines of force, the dian~eters of the spirals cliininishing as the strength of the field increases. There will, therefore, be a concen-tration of the rays along the axis of the tuhe; in other worcls, the cu'rrent density along the axis will be increased.

Now Prof. Thomson, many years ago, showed that the effect of such an increase is not to increase in tlie same proportion the n~unlber of lunlinous ato-rns. An increase in current density is often sccoinpanied by an abrupt change in the l~mlinositp, out of all proportion to the increase in current clensity. The discharge may be luminous when the nlagilet is acting and son-lu-rninous when it is off.

It seems " that the clischarge produces a c~ull~ulative effect on tlie molecules of the gas in its irnnlediate neighbourhooc!, and that the luminosity of an atoll1 aeed not be the result of a single collision between a corpuscle and the atom." A similar effect can be procluced by raising the tetriperature of the liriie cathode and thereby increasing the supply of corpuscles.

Tllis would explain the discontinuity often observed. If,as in Righi's experiments, the magnetic field be not unifornl, but

cjecrcases as 11e pass away from the cathode, then, uucler tihe above conditions, i t ~ ~ o ~ ~ l d be expected thst the lurnillosity worrld cease at a certain clistancc f ~ o n lthe cathode, due to tlie declease in (urrent clensity.

The effect of an increase in I-riagnctic field mill be to increase the current dellsity at points remo~ed from tlie cathocle, anrl, therefore, the loilgituclinal extent of the luminosity will be inclea~ed. This *I.las been founcl to be the case experimentally.

1909.1 4312 the Natzcrc o f iVagneto-cathodic Rays. 333

Again, luminosity is acconlpanied hy ionisation, and a sudden increase in bhc former implies an abrupt increase in the latter. Therefore, the appearance of l~uu-~inosity will be accorilpaniecl by a suclden increase in the conductivity of the gas. This is of importance in conneetioil mith experi- ments on the charge carlied by the rays. The potential of the Faraday cylinder will remain constant when the rate a t which it receives a charge is equal to the rate of leali of its charge. The latter depends very largely on the conductivity of the gas. If the conductivity be sncldenly increased, the limiting lsoteiitial ill be suddenly lowered. This was found to be the case in Itiglri's experiments : he observed that when the increa.;e in luminosity occurred he could no longer detect a charge on the rays.

The cliange in the potential of discharge does not appear to loe very important. Siinilat- effects are produced by a transverse field, and as Almy* has shown, they can be produced without the aid of a niagnet by covering hhe outside of Lhe tube in the neighbou~hood of the cathode with tinfoil conrlected mith the cathode.

The deflection of the rays in an electric field lias still to be considered. An electrified particle moving in an electrost~tic field is cleflected in a direction parallel to tlie field ; hut as Prof. Tliomson has shown, if the particles have already been coiled up into a spiral by a magnetic field, then the effect of tlie electrostatic field is to deflect the axis of the spiral in a direction a t right angles to both fields. The bnnclle of rays will therefore appear to be deflected in this direction.

The value of this deflectioll is given by tile relation

tan 8= E/Hc,

where E is the electric force and I1 tlie magnetic force aching on the particle, and .u is the initial velocity of the ~ ~ a l t i c l e in a di~.ectjoa parallel to the magnetic: force. E'iom this equation it can 1)e seen that the rotation is directly ploportioaal to tlie electric field nncl inrersely ploportionxl to the magnetic field, and a change of sign accoinpanies a reversal of either the elertric or nlagllctic field. This is in agreement nit11 Villarcl's xesults.

A siniilar agreeirielit i~ fo~uld for tlie order of tlle magnitntle of the deflec- tion. Xnbstitilting the values I? = 1000 volts per centinzetre, M = 200 gauss, end 2* = 3 x l o n cnl. per scconcl, we lravo tan 8= -i-$, which agrees mitt1 Villard's results.

i7f'ngij~tzcJiaps t t z llns s11own tliat magnetic [ire Crii.iznl-~5?1~nJ,Z~~n.-ltiglii rays are ppocluced whcn a niagrletic field acts on a discharge tnbc eontainirlg Canal-strahlen. Tlie :\ppenr9a,nce of tliesc lays call be explained as follows :

* 'Eroc. Carnb. Pl~il.Boc.,' rol. 11, p. 153, 1901.

VOL. LXXXIII.-A. 2 a

334 01%thc i i a t ~ w e yf" Magi~eto-cccthodicRays.

Tt has been shown by Prof. Thoillson that in a bundle of canal-strahlen there are some negatively electrified particles. These have been bent into spirals by the n~agnetic ficld in the same nianner as the ordinary cathode rays would be.

The experiments described above show that rays acting like magneto- cathodic rays consist of slowly moving negatively electrified particles svhich ]lave been coiled into very fine sl7irals. The charge carried by these particles call be detected by using a sensitive galvanometer. On this hypothesis the results obtained by 1:ighi and others carr be explained.

The conditions in an electric discharge tube are exceedingly coinplex ; the liincls of rays already clistinguislled are increasing ill number and del?encl very much on the conditioils of tlie experinlent. To prove the non-existence of any kind of rays is, therefore, qomewllat clif6cult, but i t may be said that until further experinleiltal evitlence is forthconling i t seems unilecessary to postulate in this case that the rays fclllo~~isig the lines of magnetic force are otlier than slosvly nioring cathode rags,

I n collclusioii, I desire to express rily appreciation of the liindly and coilstant interest which I'rof. Sir J. J . Thonlson has shown throughout the course of this work. I have pleasure also in tlianli-ing ;\IraE. Everett for the assistance he gave ine cluring the early 1?art of tlie nlorli.