J J Thomson and the discovery of the electron

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

1997 Phys. Educ. 32 226

(http://iopscience.iop.org/0031-9120/32/4/015)

Download details:

IP Address: 129.78.139.28

The article was downloaded on 27/08/2013 at 06:56

Please note that terms and conditions apply.

View the table of contents for this issue, or go to the journal homepage for more

Home Search Collections Journals About Contact us My IOPscience

226

J J T H O M S O N ’ S E L E C T R O N

One experiment, more than any other, is oftenassociated with the ‘discovery of the electron’ in1897. This is J J Thomson’s determination of themass to charge ratio (m/e) of cathode rays bydeflecting them in magnetic and electric fields.Yet this experiment was performed two monthsafter Thomson first announced that cathode rayswere very small, negatively charged particles. Sowhy was it important? I look at Thomson’s routeto, and conduct of, the experiment, and then athow his ideas were received.

Born in 1856, Joseph John Thomson was the son ofa Manchester bookseller. He was educated atOwens College, Manchester, and Trinity College,Cambridge. In 1884, after four years of research,mainly theoretical, he was elected CavendishProfessor of Experimental Physics at Cambridge atthe early age of 28. Upon his election Thomson began experiments ongaseous discharge. He saw this as a way ofuntangling the relationship between the ether andchemical atoms. In 1896, with his student Ernest

J J Thomson and the discovery ofthe electronIsobel Falconer St Andrews, UK

Figure 1. J J Thomson giving a lecture demonstration of his m/e experiment, around 1900. (Photographcourtesy of the Cavendish Laboratory)

issue 4 (july) 1997 4/7/97 9:40 am Page 226

227

J J T H O M S O N ’ S E L E C T R O N

Rutherford, he achieved his most outstandingsuccess to date. Working with the newly discoveredx-rays, they established the theory that electricconduction through gases took place by splitting thegas molecules into oppositely charged ions.

Cathode rays

Thomson then began to assimilate other phenomenainto his newly successful theory. Cathode rays werean obvious target. They had just sprung intoprominence as the origin of x-rays, produced whencathode rays hit a target. Cathode rays werediscovered in 1857 by Julius Plücker. They areobserved at very low pressures in an electricdischarge tube. Although invisible, they causefluorescence where they hit the wall of the tubeopposite the cathode.

By 1896 it was known that they could be deflectedin a magnetic field, and that they cast strongshadows, suggesting that they travelled in straightlines from the cathode. In 1893 Heinrich Hertz hadshown that they could pass through thin metallicfilms. This topic was pursued by his student PhilippLenard. Such rays became known as Lenard rays. In1895 Jean Perrin had demonstrated that the rayscarried an electric charge.

With the discovery of x-rays, much discussion wassuddenly focused upon cathode rays. Two views oftheir nature came into open conflict. One was thatthey were negatively charged particles, probably

atoms. The other was that cathode rays were aphenomenon of the ether, akin to light.

In late 1896 Thomson turned his attention tocathode rays. He examined their magneticdeflection, and then modified Perrin’s experiment toshow more conclusively that an electric charge wasan indispensable property of the rays. He alsoconsidered Lenard’s results on the properties andabsorption of Lenard rays.

On 30 April 1897, at a Royal Institution FridayEvening Discourse, Thomson announced hisconclusion that cathode rays are small negativelycharged particles which are a universal constituentof atoms. He supported his suggestion by the resultsof his first m/e experiment, which relied on theheating effect of the rays. His results gave a mass tocharge ratio about 1000 times smaller than that forthe hydrogen ion, hitherto the smallest known. Hecalled the particles 'corpuscles', but they have sincebecome known as 'electrons', and Thomson hasbeen hailed as their discoverer.

The m/e experiment

The experiment upon which this article focuses wasThomson’s second method of determining the massto charge ratio of the corpuscle. He performed it inlate June or July 1897, two months after his firstsuggestion that cathode rays were corpuscles.

The experiment relied on electrostatic deflection ofcathode rays. Hertz’s inability, 13 years previously,

Figure 2. Thomson’s electric and magnetic deflection method of measure m/e for cathode rays (Thomson1897b).

issue 4 (july) 1997 4/7/97 9:40 am Page 227

228

J J T H O M S O N ’ S E L E C T R O N

to show an electrostatic deflection had been held tosupport the ether view of the rays.

Cathode rays were produced at the cathode, on theleft of the apparatus (figure 1), and travelled to theright, passing through the anode which acted as acollimating tube. They entered a region where eitherelectric or magnetic deflection fields might beapplied, and then hit the end of the tube, causingfluorescence. Deflection measurements, made in thedark, were obtained by moving a luminous needlewith a screw over the measuring scale until itcoincided with the fluorescent patch.

The electrostatic deflection was given by

u 5 Fel/mv 2

(u 5 electric angular deflection, F 5 applied electricintensity, l 5 length of electric plates). If a magneticforce was applied, extending over the same area asthe electric plates, then

f 5 HEl/mv

(f 5 magnetic angular deflection, H 5 appliedmagnetic field intensity). If the magnetic field wasvaried until the magnetic deflection was the same asthe electric deflection, the equations simplified to

Fel/mv 2 5 Hel/mv or v 5 F/H

m/e 5 H 2l/Fu.

u was measured by measuring the electric deflectionalone.

Thomson made measurements on air, hydrogen andcarbonic acid gas, and tried both aluminium andplatinum electrodes. His results for m/e varied from1.1 3 1027 to 1.5 3 1027 grams per coulomb, andwere in general agreement with his earlier results.

He published his results in October 1897. Hereiterated his conclusion that the cathode rays werevery small charged particles which were theconstituents of atoms, and he proposed an atomicmodel based on them.

Thomson’s route to the m/e experiment

In identifying cathode rays as corpuscles, Thomsonadhered to the particle view of cathode rays,ostensibly because this was definite and its

consequences could be predicted, whereas we wereignorant of the laws governing the ether (Thomson1897b, p293). His attitude was symptomatic of theMechanical Philosophy through which many Britishphysicists sought a unified theory of nature. TheMechanical Philosophy was the belief that allphenomena could be described ultimately in termsof matter in motion. Thus, in investigating anunknown phenomenon such as cathode rays,Thomson’s first interest was in their velocity andmass.

Earlier, during the 1880s, Thomson had pursued theMechanical Philosophy to its mathematicalconclusion. His results led him to believe that 'atheory of matter is a policy rather than a creed. Itsobject is to connect or coordinate apparently diversephenomena, and above all to suggest, stimulate anddirect experiment' (Thomson 1907, p1). Thomsonmanipulated his theories in just this way. He seldomallowed them to be tied closely to experimental'facts' which might limit their scope for imaginativeextension to other phenomena or furtherexperiment.

His cathode ray work shows this clearly. Hiscorpuscle hypothesis seems based on only twoexperimental results. First, experiments by Lenardand by himself showed that the magnetic deflectionof cathode rays was independent of the electrodes orgas through which they passed. This suggested thatthe particles were the same in all cases. Second,Lenard had shown that Lenard rays travelled muchfurther through a gas than one would expect for anatomic sized particle. Furthermore, their absorptionwas inversely proportional to the density of the gas.This suggested that the particles might be very smalland be interacting with individual constituents ofthe gas molecules.

Rather than introduce two new particles, Thomsoncharacteristically made one do both jobs. Thecathode rays, he said, were small corpuscles whichwere themselves the building blocks of atoms. Hehad been playing with ideas of structured and/ordivisible atoms for the previous 15 years, so thisexplanation in 1897 is hardly surprising. It is typicalof Thomson’s attempts to unify physics by seekingtheories which would explain as many differentphenomena as possible.

On 30 April 1897 Thomson’s theory met withgeneral scepticism. Even George FitzGerald, a most

issue 4 (july) 1997 4/7/97 9:40 am Page 228

229

J J T H O M S O N ’ S E L E C T R O N

sympathetic listener, considered that he had gonemuch further than his experimental data warranted.FitzGerald made the alternative suggestion that thecathode ray particles were free electrons, as providedfor by the theory of Joseph Larmor. Such electronswere supposed to be centres of strain in the ether.They explained discrete electric charges, but wereindependent of matter. Thus FitzGerald acceptedthe particulate nature of cathode rays. But hedissented from the idea that these same particlesmight make up atoms. In Thomson’s terms he wasintroducing an unnecessary extra hypothesis.By July 1897 Thomson had several reasons for hism/e experiment. He had extended the corpuscletheory in many directions: he explored itsimplications for ionization and discharge, he devisedatomic models, and he speculated about thestructure of molecules. His aim was to extend thescope of corpuscles and demonstrate (contrary toFitzGerald) that they were constituents of atoms. Henoted that the specific inductive capacities of gaseswere approximately additive, implying that theelectrical moment of each atom was very high. Thisled him to suggest that the small value of m/emight be due to a large charge as well as a smallmass. These speculations pre-dated his attempts todeflect cathode rays electrostatically and provide amotive for them.

Obtaining an electrostatic deflection

Initially Thomson did not invest much effort in theattempt to deflect cathode rays electrically, hemerely cannibalized the apparatus previously usedto measure the magnetic spectrum of the rays. Hefound no deflection. But he noticed that when therays were on, a discharge passed readily between thetwo deflecting plates, indicating that the cathoderays turned the gas into a conductor. He realizedthat the conducting gas screened out the appliedfield.

The secret to obtaining an electrostatic deflectionwas clearly to get rid of the residual gas in the tube.The difficulty was to obtain a low enough pressureto prevent ionization and conduction, and yet toinitiate a discharge.

One source of residual gas was mercury vapour fromthe pump itself. Another potential source of vapourwas the pressure gauge, and Thomson seldom usedone, recording his pressures simply as 'low', 'very

low' etc, judging them by the behaviour of thedischarge tube.

By 1897 the pump generally used in the Cavendishwas a Topler mercury pump worked by hand. It hadto be worked for half a day before a good cathoderay vacuum was obtained. Before Thomsonmanaged to observe an electric deflection of therays his assistant Ebeneezer Everett ran the tube forseveral days pumping all the time.

The purpose of running the discharge tube whilepumping was to get rid of all the gas adsorbed onthe walls and electrodes. Otherwise this wasreleased throughout the experiment and destroyedthe vacuum. 'Baking' the discharge tube wasbecoming a recognized technique for getting rid ofthis adsorbed gas, but Thomson’s attitude wasinconsistent. He used it only when the necessity fordoing so was forcibly brought to his attention.

Having attained very low pressure, a further problemarose: below a certain critical pressure the potentialneeded for discharge rose rapidly and dischargesoon became impossible because the dischargetubes broke. In 1883 De la Rue and Muller found

Figure 3. Hertz’s 1883 apparatus to try to detectelectrostatic effects of cathode rays (Hertz 1883).His tube to try to deflect the rays electrically wassimilar, but charged electric deflection plates wereintroduced.

issue 4 (july) 1997 4/7/97 9:40 am Page 229

230

J J T H O M S O N ’ S E L E C T R O N

that increasing the diameter of the discharge tubelowered the critical pressure. Thomson appears tohave designed his cathode ray tube of 1897 withthis in mind — the bulb is about 10 cm wide aroundthe cathode before narrowing to collimate the beam.Hertz, working 13 years earlier, was unaware of Dela Rue and Muller’s results. His discharge tube wasa uniform width of about 2.5 cm as shown in figure3. Had he attained a low enough pressure toobserve electrostatic deflection, he may not havebeen able to obtain a discharge.

Results of the experiment

Eventually, then, Thomson observed the electrostaticdeflection he had been seeking. He was now able tomeasure m/e by his second method, as outlinedabove. Thomson judged that, ‘This method ofdetermining the values of m/e is much lesslaborious and probably more accurate than theformer method’ (Thomson 1897b, p310) and thephysics community has concurred with hisjudgement. Nevertheless, his experiment was farfrom accurate.

Thomson discussed only two sources of error in hisexperiment. First, the magnetic force was assumedto be confined to the space between the electricplates, which was only approximately true. This wasa systematic error which increased the measurementof m/e. He made no attempt to assess the increase.

Secondly, under either deflection, the cathode raysspread out into a spectrum. Rather than a brightfluorescent spot on the end of his tube, Thomsonwas observing a patch several millimetres long,introducing an error of up to 20%. The significanceof this depends on whether he measured theelectrostatic and magnetic deflections one after theother, or whether he opposed them to get a nulldeflection, an inherently more accurate method.Accounts written much later state that he opposedthe forces, and this was certainly a subsequentrefinement of the method. But there is nothing inthe 1897 paper to suggest that he did so — theimplication is that he did not.

For Thomson’s purposes, attempts at precision werea waste of time: in October 1897, having reiteratedhis corpuscle suggestion, he plunged intospeculations on atomic structure, viewing the atomas an aggregation of corpuscles. The theory was too

complex to allow more than a qualitative discussion.Thomson was confident of his m/e results to withinan order of magnitude or so, and this he consideredgood enough.

Corpuscles to electrons

Other physicists, however, did not agree. Thesuggestion that atoms might be composed ofcorpuscles smacked too much of alchemy to bereadily accepted and Thomson’s experiments werenot sufficiently definite to establish this. The mainreason why the results of the m/e experimentrapidly became important was that they alsosupported FitzGerald’s alternative electronsuggestion.

In this form, that cathode rays were 'electrons',independent of matter, Thomson’s work wasaccepted rapidly, especially after his experiments of1899 which showed that the charge was equal tothe unit of electrolytic charge. H A Lorentz, whoseelectron theory was similar to Larmor’s and was farmore influential on the Continent, seized upon theelectron interpretation and incorporated it into histheory. Moreover, a 'free electron' was some sort ofstructure in the ether. Hence this suggestion wasacceptable even to the protagonists of the etherview of cathode rays. By 1900 the cathode raycontroversy had virtually died out.

By the time it was realized that Thomson was right,and that cathode ray particles were an essential partof atomic structure, the word 'electron' wasinextricably associated with the particles.Thomson’s term 'corpuscle' was forgotten, but itcontributed an added meaning to 'electron', that ofbeing an elementary particle, the first to bediscovered, and a fundamental constituent of atoms.

This realization waited upon three developments inphysics. First came appreciation of the enormoustheoretical possibilities if matter was composed ofelectrons. The mass of Lorentz and Larmor’s electronswas electrical in origin (an idea first suggested byThomson in 1881). If these electrons were also thefundamental particles from which atoms were made,then the entire mass of the universe might beelectrical. This idea promised a great advance in thesearch for a unified theory of physics and wasextensively developed in the early twentieth century.

Experimentally, Thomson’s work, and the newelectron ideas, were confirmed by many others.

issue 4 (july) 1997 4/7/97 9:40 am Page 230

231

J J T H O M S O N ’ S E L E C T R O N

Most notable was Walter Kaufmann, a highly skilledexperimentalist. Several people, including theCuries, nominated Kaufmann jointly with Thomsonfor the Nobel Prize, reasoning that Thomson’stheories would not have been accepted withoutKaufmann’s supporting evidence.

The third development was also largelyexperimental: the discovery and investigation ofradioactivity. By the early 1900s radioactivity wasproviding ample evidence that atoms could and didsplit up and change their chemical nature.Thomson’s atomic model was the only one whichgave an explanation for this. Moreover,measurements of m/e for beta rays showed them tobe the same as cathode rays.

The importance of the m/e experiment

For Thomson, his cathode ray work was one stepalong the way to establishing a coherent theory ofgaseous discharge. It was for this, rather than thecathode ray work, that he won the Nobel Prize in1906. Neither cathode rays, corpuscles nor electronswere mentioned in the citation. Indeed, for many atthe time, Thomson was not the clear-cut 'discovererof the electron'. Alternative accounts, in whichThomson was of only minor importance, vieweddevelopments by Lorentz, Larmor, Zeeman orWiechert as the significant steps which establishedthe existence of electrons (e.g. Kaufmann 1901).

Viewed with hindsight, though, it was Thomson whomade the nineteenth century electron 'real'.Arriving at the theoretical idea of an electron wasnot much of a problem in 1897. But Thomsonpinpointed an experimental phenomenon in whichthe electron could be identified, manipulated andexperimented upon. He did this most clearly in them/e experiment, showing how electrons could bedeflected magnetically and electrically, how

measurements could be made upon them, and howto attach meaning to those measurements. Throughthe m/e experiment electron theory changed froman abstract mathematical hypothesis to an empiricalreality, expanding its meaning in the process.

Received 26 March 1997PII: S0031–9120(97)83221–2

References and further reading

Davis E A and Falconer I 1997 J J Thomson andthe Discovery of the Electron (London: Taylor andFrancis)

Falconer I 1987 Corpuscles, electrons and cathodeRays: J J Thomson and the 'Discovery of theElectron' Br. J. Hist. Sci. 20 241–76

FitzGerald G 1897 Dissociation of atomsElectrician 21 May 103–4

Hertz H 1883 Experiments on the cathodedischarge Ann. Phys., Lpz. 19 782–816

Kaufmann W 1901 The development of theelectron idea Electrician 8 November 95–7

Rayleigh 4th Lord 1942 The Life of Sir J J Thomson(Cambridge: Cambridge University Press; reprinted1969, London: Dawsons of Pall Mall)

Thomson G P 1964 J J Thomson and the CavendishLaboratory in his Day (London: Nelson)

Thomson J J 1897a 'Cathode Rays' RoyalInstitution Friday Evening Discourse, 30 April 1897,published in The Electrician 21 May 1897 104–9

—— 1897b Cathode rays Phil. Mag. 44 293–316

—— 1907 The Corpuscular Theory of Matter (London:Constable)

—— 1936 Recollections and Reflections (London:Bell; reprinted 1975, New York: Arno Press)

issue 4 (july) 1997 4/7/97 9:40 am Page 231