A Brief History of Magnetospheric Physics Before the Spaceflight Era

Home , Eugene Parker, Kristian Birkeland, Space physics, Terrella

A BRIEF HISTORY OF MAGNETOSPHERIC PHYSICS BEFORE THE SPACEFLIGHT ERA

David P. Stern Laboratoryfor ExtraterrestrialPhysics NASAGoddard Space Flight Center Greenbelt,Maryland

Abstract.This review traces early resea/ch on the Earth's aurora, plasma cloud particles required some way of magneticenvironment, covering the period when only penetratingthe "Chapman-Ferrarocavity": Alfv•n (1939) ground:based0bservationswerepossible. Observations of invoked an eleCtric field, but his ideas met resistance. The magneticstorms (1724) and of perturbationsassociated picture grew more complicated with observationsof with the aurora (1741) suggestedthat those phenomena comets(1943, 1951) which suggesteda fast "solarwind" originatedoutside the Earth; correlationof the solarcycle emanatingfrom the Sun's coronaat all times. This flow (1851)with magnetic activity (1852) pointed to theSun's was explainedby Parker's theory (1958), and the perma- involvement.The discovei-yof •solarflares (1859) and nent cavity which it producedaround the Earth was later growingevidence for their associationwith large storms named the "magnetosphere"(1959). As early as 1905, led Birkeland (1900) to proposesolar electronstreams as Birkeland had proposedthat the large magneticperturba- thecause. Though laboratory experiments provided some tions of the polar aurora refleCteda "polar" type of support;the idea ran into theoreticaldifficulties and was magneticstorm whose electric currents descended into the replacedby Chapmanand Ferraro's notion of solarplasma upper atmosphere;that idea, however, was resisted for clouds (1930). Magnetic storms were first attributed more than 50 years. By the time of the International (1911)to a "ringcurrent" of high-energyparticles circling GeophysicalYear (1957-1958), when the first artificial the Earth, but later work (1957) reCOgnizedthat low- satelliteswere launched, most of the importantfeatures of energy particlesundergoing guiding center drifts could the magnetospherehad been glimpsed, but detailed have the same effect. To producethe ring current and understandinghad to wait for in situobservations.

INTRODUCTION EARLYWORK ON GEOMAGNETISM

This is an account of early researchon the Earth's The historyof geomagnetismbegins with the magnetic distant magnetic environment,work that led to mag- compass,invented in China around the year 1000 and netosphericphysics and to space plasma physics. It tellsc•f quickly adoptedby Arabsand Europeans[Mitchell, 1932]. a sciencein its earliest,most primitive stage, when Gradually,it wasrealized that the magneticneedle did not explanationswere qualitativeand full of speculation.The point to truenorth; Columbus observed during his crossing early stagelasted here a long time, becauseremote sensing of the Atlantic that it shifted from one side of true north to of the spaceenvironment from the grounddid not tell the other [Mitchell, 1937]. enoughfor a full understanding.Researchers relied mainly Magnetism was the avocationof William Gilbert, on global magneticdata with some help from solar and Queen Elizabeth I's personalphysician. Gilbert gave a auroralobservations. Their prime tools were insight and convincingexplanation of the action of the compass:the imagination,and their mathematicalskills could only Earth was a greatmagnet. He reachedhis conclusionwith occasionallybe broughtto bear. With all thoselimitations the help of a sphericalmagnet, a modelof the Earth which it is remarkablehow many of our fundamentalideas on he named the "terrella," or "little Earth." Moving a spacephysics were glimpsedduring those early years. compassover the surfaceof the terrella, he observedthat Note thatfor early work, recentand relatively accessible its needlepointed toward the magneticpoles, and he also publicationsare sometimesgiven and readersseeking the demonstratedthis before the queen. Gilbert's book De originalpapers will find themcited there. The Journal of Magneteappeared in 1600 and describedall that was then GeophysicalResearch was known prior to 1957 as knownabout magnetism and electricity [Gilbert, 1958]. It TerrestrialMagnetism and Atmospheric Electricity. was one of the importantscientific books of the age of

Thispaper is not subjectto U.S. copyright. Reviewsof Geophysics,27, 1 / February1989 pages 103-114 Publishedin 1989 by the AmericanGeophysical Union. Paper number 89RG00097

ß 103ß 104 . REVIEWS OF GEOPHYSICS / 27,1 Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE

Galileo and amongother thingscontained the first use of With such tools it was observed that the Earth's field the term "electric force" which led to the later term was occasionallydisturbed for a day or so: theseevents "electricity." were termed "magneticstorms," but no one knew their Important advancesin geomagnetismfollowed in the cause. Celsiusfound that the large magneticdisturbance next two centuries[Chapman and Bartels, 1940, volume2, of April 5, 1741, was detectedsimultaneously by him in chapter26; Nelsonet al., 1962]: Uppsalaand by Grahamin London[Chapman and Bartels, 1. The discoveryby Gellibrandin 1635 of the slow 1940, section26.10], demonstratingthe nonlocalnature of variation of the Earth's field [Malin and B ullard, 1981; magneticstorms. The magneticnetwork started by Gauss Brushand Banerjee, 1988]. and Weber later showed the storms to be a worldwide 2. The discoveryby Graham[1724] (seeChapman and phenomenon. Bartels [1940, section26.9]) of "magneticstorms" (later term),large irregular disturbances of the compassneedle. 3. The first magneticsurvey of the AtlanticOcean by THE SUNSPOT CYCLE Halley, in 1699 [Bullard, 1956; Ronan, 1969; Evans, 1988]. Enter the Sun. In the first half of the nineteenthcentury 4. The discoveryby Oerstedin 1820 that electric there lived in the German town of Dessaua pharmacist currentsproduced magnetic forces [Shamos, 1959; Dibner, named Samuel Heinrich Schwabe whose hobby was 1962]. astronomy[Newton, 1958]. Every day when the Sun was 5. The laws of electromagnetism,by Ampbrein 1821 not obscured,Schwabe observedit, paying attention to [Williams, 1965, 1989]. sunspots,noting their numbers,and keeping a tally of days 6. Electromagneticinduction, by Faraday in 1831 when they were absent [Meadows, 1970]. He started [Faraday,1952; Williams,1963, 1965]. observingin 1826 and 10 yearslater publisheda reportof In 1839 Carl Friedrich Gauss [Gauss, 1839, 1877; his results: no one seemedto pay attention. In 1843 he Dunnington,1955] publisheda methodfor mathematically publisheda more completeaccount, suggesting a 10-year describingthe Earth's field B by means of a scalar cycle: at first, again,no response.Eventually, however, potential¾, Schwabe'swork caughtthe eye of Alexanderyon Hum- (1) boldt,naturalist and promoter of the sciences,who in 1851 included Schwabe's results in his third volume of Kosmos, expandedat anypoint (r, O,•) in sphericalharmonics: an encyclopaediccompilation of information about the physicalworld [Schwabe,1851]. Suddenly,sunspots and 7= a 2;(a/r)n+l P7 (0) [g7 sinm•) + h• cosm•)] their cycle became a hot topic: astronomersbegan countingsunspots and studyingthem, earlier cycleswere + a 2;(r/a)'• P7 (0) [G7 sinm•) + H• cosm•)] (2) reconstructedfrom old observations,and searchesbegan for terrestrialeffects which correlatedwith the sunspot The first sumrepresents sources inside the Earth, and the cycle. second one external sources. Gauss and his associate Very soon such a correlation was found. Edward Wilhelm Weber then went on to found a network of Sabine, a British scientist and the main architect of a observatories,greatly expandedby British and Russian worldwidenetwork of magneticobservatories (an expan- help [Cawood, 1979; Malin, 1969]. From data thus sion of an earlier effort by Gaussand Weber), announced obtained, the coefficients due to sourcesinside the Earth in 1852 that the frequencyof magneticstorms rose and fell were derived[see Barraclough, 1978]; as for the external with the numberof sunspots[Sabine, 1852; Meadows and coefficients,the calculationgradually continned what had Kennedy,1982] (seealso Lamont [1852]). beensuspected, that better than 99% of the field originated Evidencewas soonalso found that the polaraurora was inside the Earth. more frequentlyseen (at relatively low latitudes)near the However,as Graham'swork suggested,some magnetic peakof the sunspotcycle. Here too a magneticconnection effectsdid originateon the outside. Observationsof such existed: as early as 1741 the Swedishscientist Celsius effectswere advancedby the work of CharlesCoulomb, reportedthat during auroraldisplays the magneticneedle who in 1777 greatlyincreased the sensitivityof magnetic was disturbed [Stoermer, 1955, section 6; Eather, 1980]. measurementsby suspendinga magneticneedle from a The actual discoverymay have been due to Hiorter, a fine string[Gillmor, 1971; Shamos,1959]. Suchinstru- studentof Celsiuswho later wrote that when he reported mentscould be made even more sensitiveby attachinga the magneticeffect of the aurora to his mentor, Celsius small mirror which moved a spot of light, and this type saidthat he too had observedthe phenomenonbut had not dominatedgeomagnetism for closeto 200 years[Nelson et mentioned it in order to see whether his student would find al., 1962;Multhauf and Good, 1987]. it independently. Stern: HISTORY OF PRE-SPACEAGE MAGNETOSPHERE 27,1 / REVIEWSOF GEOPHYSICSß 105

SOLAR FLARES electrically charged particles whose properties were measuredby J. J. Thomsonand which were eventually How did sunspotsexert their influence?The first clue named electrons [Thomson, 1967; Shamos, 1959]. cameon September1, 1859,in an unexpectedobservation Electronbeams propagated at greatspeed, which led to the by the distinguishedBritish astronomerRichard Car- plausiblesuggestion that the sourceof observeddistur- rington[Meadows, 1970, p. 181]. Carringtonwas in the bances was streams of electrons emitted from sunspot middleof an 8-yearstudy of sunspotsand was observing a regions. largesunspot group when "two patches of intenselybright The first serious study of this phenomenon was andwhite light broke out... thebrilliancy was fully equal performed by the Norwegian Kristian Birkeland to thatof directsunlight." Noting that the spotwas rapidly [Birkeland,1901, 1908;Egeland, 1984, 1986;Devik, 1968; brightening,Carrington rushed off to find a witness,but BostrOm,1968]. In 1896 Birkelandaimed cathoderays at comingback only 60 secondslater he foundthe spotof a magnetand foundthat the magnetapparently "sucked in" light "muchchanged and enfeebled" and soon afterward it cathoderays: he suggestedthat the Earth's field did the fadedaltogether [Carrington, 1860]. same to beams from the Sun. He communicated his As luck had it, the astronomerHodgson [1860] (see findingsto his former teacher,the French mathematical Meadows[1970, p. 187]) observedthe sameevent from physicistHenri Poincar6,who showedthat rather than anotherpart of England. An unusuallyintense magnetic beingattracted, charged particles were guidedby magnetic stormfollowed 17 hoursafterward, accompanied by polar field lines [Pointcard, 1896]. Poincar6 calculated the aurora that could be seen far from the polar regions motionof an electronin the field of a magneticmonopole, (anothersuch storm had occurreda few days earlier, a completelysoluble problem, and found that the electron probablyfrom the samesunspot group). Carringtonnoted spiraledaround a cone boundedby field lines, gradually the coincidence but added "one swallow does not make a losingheadway until at a certaindistance it was reflected summer." backward [Rossi and Olbert, 1970, section 2.5; Mitchell We now know that Carringtonhad seena solarflare, a and Burns, 1968]. rapidrelease of energyprobably drawn from the sunspot's Birkelandthen built a large vacuumtank, placedin it a magneticfield, capableof acceleratingelectrons and ions sphericalmagnet--like Gilbert he calledit a terrella--and to highenergies. Flares rank among the most rapid of the aimedat it beamsof cathoderays (Figure 1). Bright spots Sun'sobserved phenomena: they canextend over tensof appearedwhere the beamshit the terrella,generally in the thousandsof kilometers, and their fastest features have polarregions. In someexperiments there were even bright time scalesof seconds,though the wholesequence usually ringsaround the magneticpoles. lasts tens of minutes to an hour. By that time, appreciableinformation had accumulated Only rarely do flares emit intensewhite light, as aboutthe aurora. EliasLoomis of Yale publisheda map of Carrington'sdid, but they are readilyobservable through contours of equal auroral frequency in the northern filterswhich isolate the red Ha brighteningsnear sunspots, hemisphere[Loomis, 1860], showingthat they centeredon andin 1892George Ellery Hale [Wright,1966] devised the the magneticpole (ratherthan the geographicone) and that spectroheliograph,which produced images of entireareas theirfrequency was highest in an ovalband about 20 ø from on the Sun usingonly a singlespectral wavelength. On the pole [Eather, 1980]. Hermann Fritz conducteda July 15 of thatyear, Hale produced a seriesof photographs similar study with far greater precision [Fritz, 1881]. documentingthe evolutionof a large flare, which was FollowingBirkeland's work thepieces suddenly seemed to followed19 hourslater by a largemagnetic storm [Hale, fall into place: flares (or sunspots)apparently emitted 1892]. electronstreams, which were steeredby the Earth's field More suchcorrelations soon followed, leaving no doubt toward the auroral zones--and since a stream of electrons that somethingwas propagating from the Sun to the Earth carriedan electric current,a magneticdisturbance would at about 1000 km/s (or faster,as in the two eventscited alsobe produced[Stoermer, 1917]. here), causinga magneticdisturbance upon its arrival This view was supportedby Maunder [1904], who had [Fitzgerald,1892]. deduceda tendencyof stormsto recurat 27-day intervals, the rotationperiod (relative to Earth)of low solarlatitudes where sunspotstend to occur. He noted that recurrent ELECTRON BEAMS FROM THE SUN? stormswere hard to correlatewith solarphenomena (see furtherdiscussion below), but he still believedthat "solar What was it? One clue seemedto comefrom discharges streams"were responsible and wrote as follows: in low-pressuregases and from beamsof "cathoderays" That, therefore,which Lord Kelvin spokeof twelve years ago propagatingbetween electrodes in evacuatedvessels. as "the fifty years' outstandingdifficulty" is now rendered Laboratorystudies showed that these"rays" consisted of clear. Our magneticdisturbances have their originin the Sun. 106 ß REVIEWS OF GEOPHYSICS / 27,1 Stern' HISTORY OF PRE-SPACE AGE MAGNETOSPHERE

The solar actionwhich givesrise to them doesnot act equally terrellaat low and middle latitudeswere trapped,so that in all directions,but along narrow, well-definedstreams, not particlesarriving from a distantsource, like the electrons necessarilytruly radial. These streamsrise from active areas in Birkeland's experiment,never reachedthose latitudes of limited extent. Theseactive areasare not only the source of our magneticdisturbances but are also the seatsof the but were always steeredto the polar regions or turned formationof sun-spots.... away, in full accordwith Birkeland'sobservations. In the Earth's field, Stoermer'stheory worked well for Birkelandcertainly did his bestto promotethe notionof cosmicray particlesin the Gev range,but not for auroral solar electron streams. He also asked a colleague,the electrons.It wasrealized quite early that the atmospheric young mathematicianCarl Stoermer, to calculate the density at 100 km, where auroral electronsgenerally motionof electronsin a dipolefield, and Stoermerspent a stopped,was so low that the energyof suchelectrons had large part of his careerattacking that problem [Stoermer, to be low too. For instance,tlarang [1951, table 31] 1955; Nutting, 1908]. Unfortunately,motion in a dipole estimatedtheir speedat 0.3c, correspondingto about 23 field (unlike the monopoleproblem) has no analytical keV. At suchlow energies,Stoermer's theory predicted solutionbut is besetby pathologiesresembling those of the impactsvery close to the magneticpoles, contraryto notorious three-body problem of celestial mechanics observationsthat showed the midnight aurora peaked [Dragt and Finn, 1976], so that Stoermernever achieved aroundmagnetic latitude 68 ø . Neithercould it explainthe what he had sought,though he did integratemany orbits observationthat auroraswere scarcenear the magnetic numerically. pole itself. He did, however,manage to prove that a wide classof The theory of solar electronstreams soon hit another orbitsexisted in the dipolefield that were trappedand did snag: Arthur Schuster[Schuster, 1911; Chapman, 1934; not extend to infinity. He furthermoreshowed that for Bartels, 1934b] showedthat electrostaticrepulsion would sufficiently low particle energiesall orbits hitting the quicklydisperse any streamof solarelectrons.

Figure 1. Birkeland(left) in his laboratorywith largeterrella experiment. Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE 27,1 / REVIEWS OF GEOPHYSICS ß 107

THE CHAPMAN-FERRARO CAVITY that thereforea sufficientdensity of interplanetaryplasma existed for such a shock to form. SidneyChapman, a relative newcomerto the field of geomagnetismwho was apparentlyunaware of Schuster's work, againraised the idea of solarelectron streams in a 1918paper on magneticstorms [Chapman, 1918; Akasofu et al., 1969]. He was pouncedupon by FrederickLin- demann,Oxford professorof physics (Lord Cherwell, Winston Churchill's controversial World War II science adviser), who pointed out that the negative charge accumulatedon the Earth would disrupt the process [Lindemann,1919]. Lindemannthen suggestedthat any cloud or streamexpelled from the Sun would have to be electricallyneutral, containing equal charge from ions and electrons. It took more than 10 yearsbefore Chapman figured out how a neutralbeam could cause magnetic disturbances. In 1927 he wasjoined in his questby VincentC. A. Ferraro, newlygraduated [Cowling, 1975]. The two hadrealized that an electricallyneutral mixture of ions and electrons--what would nowadaysbe called a plasma--wouldbe a very good conductorof electricity. Therefore,when a cloud of such matter approachedthe Earth, electriccurrents would be inducedin it, creatinga magneticdisturbance. But how could such currentsbe calculated?Chapman felt that as an approximationto a Figure 2. The Earth's dipole field (left), flattened by the three-dimensionalcloud one might start with a two- additionof the field of an image dipole (right), as proposedby dimensionalconducting sheet, approaching the Earthin its Chapmanand Ferraro. equatorialplane [Ferraro, 1969]. He knew thatMaxwell had calculatedcurrents induced in conductingsheets and advisedFerraro to look up that work. However, when The Earth's magneticfield also exerts a force on the Ferraro saw Maxwell's calculation, he realized that a induced currents, and that force grows stronger as the differentsheet approximation would be evenbetter. cloud draws nearer. Ultimately, Chapman and Ferraro If a large plasma cloud nears the Earth, its front argued,it becamestrong enough to stopany furtherfrontal boundary appears like an approachingwall--as the advance of the cloud toward Earth; however, the flanks elephantdid to one of the blind men in the parable[Saxe, continuedto advance,so that soon a cavity was formed, 1936]. Furthermore,if the cloud is a perfect electrical envelopingthe Earth. That was known for many yearsas conductor,all inducedcurrents flow on the surfaceof that the "Chapman-Ferrarocavity," the region from which the "wall." Maxwell had shown that when a perfectly plasmaof the cloud was excludedby the action of the conductingflat plane approacheda dipole, its externally Earth'smagnetic field (Figure3). inducedfield was the sameas the field of an equal "image Solar flareswere one obvioussource of plasmaclouds. dipole" located symmetricallyon the other side of the However, as Maunder had already noted, many storms plane. Thusthe initial magneticdisturbance caused by the couldnot be tracedto any clearsource, not evento definite cloudshould resemble the field of an imagedipole at twice sunspots.This held especiallyfor moderatestorms with a the distanceof the cloud,rushing toward Earth at twice the 27-day recurrence,extensively studied by Bartels [1932] cloud'sspeed (Figure 2). and Newton [1932], who named their elusive sources That was how Chapmanexplained the "suddencom- "M-regions." The mysterydeepened with the realization mencement,"a rapid steplikeincrease in the magneticfield [Bartels, 1934a] that in one carefully studiedsolar cycle, heraldingthe onset of many (though not all) magnetic recurrent storms tended to cluster around solar minimum, storms[Chapman and Ferraro, 1930, 1931, 1932]. There includingperiods when no sunspotswere visible at all. was a postscript[Dungey, 1979]: muchlater, Gold [1955] The answer was delayed until Mariner 2 detectedhigh- pointedout thatthe fact thatthe cloud maintained a sharply speed streamsin the solar wind [Snyder et al., 1963; defined front boundary long after it had left the Sun Neugebauerand Snyder, 1966], which seem associated suggestedthat this boundary was a collision-freeshock and with recurrentstorms. Still later it was shown by solar 108 - REVIEWSOF GEOPHYSICS / 27,1 Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE field (at low and middle latitudes)gradually weakens over 6-12 hours, followed by a slow recovery of the field lasting1-3 days. This partof the stormdisturbance can be far more intensethan the "suddencommencement," yet the Chapman-Ferrarocavity providedno goodexplanation. Stream In the early 1900sthe idea arosethat a "ringcurrent" of trappedparticles might exist aroundthe Earth's equatorial plane. Electronsand positive ions of sufficiently high energycould circle the Earth's equatorialplane in opposite directions,each contributing an electriccurrent in the same sense,which always weakensthe Earth's main field as observed. Carl Stoermerproposed such a ring current [Stoermer, 1910, 1911, 1912] to overcomea discrepancyin his theory, whichpredicted the aurorafar closerto the magneticpole than where it was observed[Smith, 1963; Chapmanand Bartels, 1940, section 24.13]. Soon afterward, however, Adolf Schmidtsuggested that a ring currentwas also the cause of the main phase of magnetic storms [Schmidt, 1924]. The main problem was that the energy requiredfor motionslike thosesuggested by Stoermerwas rather high: such orbits, when close to the Earth (distantorbits have other problems)are now recognizedas appropriatefor Figure 3. The formation of the Chapman-Ferrarocavity. Arrowstrace the pathsof ionsand electronswhich Chapmanand cosmicray particles.As partof theirtheory, Chapman and Ferraroproposed to accountfor ring currenteffects. Ferraroalso proposed their own versionof the ring current concept,set up (somehow)inside the Chapman-Ferraro cavity [Chapmanand Ferraro, 1933; Smith, 1963]; the observationsfrom spaceand especiallythose of Skylab curvedarrows in Figure 3 are relatedto their theory. This [Bohlin, 1977, section la] that such streams came from was later expandedby Martyn [1951] and Stoermer[1955, "coronal holes," most prevalent near solar minimum section60]. But as Chapmanremarked [cited by Hulburt, [Zirker, 1977; Hundhausen,1979]. 1937], An additional problem was posed by the imperfect The whole theoryis necessarilyboth speculative and difficult; correlation between sudden commencements(sc) and probablythe mostdoubtful feature is that relatingto the ring magneticstorms: often an sc is followedby no storm,and current,the existenceand formationof which are still very stormsoften occur without any sc. This problem, too, uncertain. requiredspace data for its resolution: it was only ex- plainedin the 1960swith the realizationof the major role Other evidencefor plasma in the distantgeomagnetic played by the interplanetarymagnetic field (IM• in field came from low-frequency radio emissions and solar-terrestrialinteractions, in particularby the north- especially from whistlers [Helliwell, 1965, chapter 2; southcomponent of the IMF. When the directionof the Alpert, 1980]. Startingwith the work of Preece [1884] IMF is not favorable,the arrival of a plasmacloud may note was taken of clicks and whistleson long telephone well produceno storm. Nowadaysa suddensteplike rise lines: the causewas later identified as electromagnetic of the magnetic field is termed an ssc (storm sudden wavesin theaudio frequency range, picked up by thelines, commencement)if it is followed by a storm and an si which acted as antennas. Such sounds were also noted on (suddenimpulse) if not. field telephonelines during World War I and included a sound like "piou," descendingin frequency. The phe- nomenonwas studiedby Barkhausen[1919, 1930] and THE RING CURRENT later by Eckersley[1925], and the descendingtones were named "whistlers." Owen Storey [Storey, 1953, 1956] If the cloud advancesat 1000 km/s, the cavity will be definitely identifiedtheir sourceas lighming,sometimes fully formedin a few minutes. A typicalmagnetic storm, occurringin the oppositehemisphere: the waves were however,lasts much longer. Its main featuresare a "main guided along magnetic field lines and often oscillated phase"in which the north-southcomponent of the Earth's severaltimes between hemispheres before decaying. Their Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE 27,1 / REVIEWS OF GEOPHYSICS ß 109 dispersionsuggested an appreciable plasma density even in ALFV•.N'S THEORY AND ELECTRICFIELDS the mostdistant portions of the field lines,and that subject was widely studiedby 1957, the year the first artificial HannesAlfv6n in Swedenwas an early investigatorof satelliteswere orbited. However, it shouldbe realized that plasmasin space. In the last yearsbefore World War II he theplasma involved here was mostly thermal, more related proposedthat ring currenteffects (and aurorasas well) to the ionosphericplasma than to the more energetic were due to the entry into the Earth's field of particles particlesof thering current. from the solarplasma cloud, convected there by an electric Shortlybefore the discovery of theradiation belt, Singer field due to the cloud's motion. If the cloud has a high [1957] pointedout that trappedparticles of low energy electricconductivity, then the localelectric field E* inside couldalso carry a ring current,even thoughtheir motion it vanishes, was more complex. An ion confinedto the equatorial E* = E + vxB = 0 (3) plane, for instance,tends to circle locally aroundfield lines,but its circle will be slightlytighter where it comes yieldingthe so-called MHD condition.The electricfield E closestto Earth, becausethe field thereis slightly stronger. enablesparticles to flow perpendicularto magneticfield This causesthe meanposition of the ion to drift slowly in lines,imparting to theiraverage position ("guiding center") longitude,gradually carrying it aroundthe Earth (Figure the velocity 4); ions and electronsdrift in oppositedirections, and thereforea neutralplasma yields a net circulatingcurrent. v = ExB/B2 (4)

Alfv6n did not believe in the Chapman-Ferrarotheory, which treated the cloud as a continuousfluid [Alfvdn, 1951] but rather viewed the cloud as a collection of individuallymoving particles. Those particles would flow togetherwith the above bulk velocity v until they came closeto the Earth'sdipole; there the guidingcenter motion (in a mannersomewhat similar to what was later invoked by Singer [1957] for trappedparticles) would move ions and electrons in opposite directions, creating a cavity aroundthe Earth and also leadingto the ring currentfield (Figure5). Alfv6n's argument was somewhatmore involvedand also includedan explanationof the aurora. Chapman,as might be expected,strongly disagreed[Dessler, 1970; Akasofu, 1970], and in the end, Alfv6n's article on his theorywas not acceptedby any majorjournal but appeared in Swedenin a relativelyobscure format at the end of 1939 [Alfvdn,1939; Cowling, 1942] (see also Stoermer[1955, Figure 4. Schematicdrift path of an equatorialring current section61] and Stern [1977]). Though it made some protonaround the Earth,viewed from north of theequator. importantpoints, in 1957 it was still poorly known and appreciatedoutside Scandinavia.

The conceptof suchdrift motionwas first notedby Gunn [1929] and was describedby Alfvdn [1950] in Cosmical INTERPLANETARY PLASMA Electrodynamics,where he presentedthe equationsof guidingcenter motion. Slow motionaround the Earthwas Chapmanand Ferrarohad assumedthat exceptfor their also found in Stoermer'snumerically integratedorbits plasmaclouds, interplanetary space was relativelyempty, [Stoermer,1955]. Singer [1957] notedthat owing to this but evidenceto the contrarycame from observationsof motion,a ring currentcould also be carriedby a belt of cometmils. For many yearsit was held that the long mils trappedparticles of relatively low energy; he suggested of cometswere adequatelyexplained by the pressureof that suchbelts were formedin magneticstorms and lasted sunlight,but Hoffmeister[1943, 1944] found that many up to a few daysbefore decaying. comet tails deviatedby severaldegrees from the radial And yet, if thoseorbits were truly trapped,both entry direction,in a way suggestingthat they were shaped not by andescape would be impossible.If theybecame populated sunlightbut by solar particlespropagating at a lower duringstorms, how did particlesreach them? velocity. After World War II this was picked up by 110 ß REVIEWSOF GEOPHYSICS/ 27,1 Stern: HISTORY OF PRE-SPACEAGE MAGNETOSPHERE

i i i i i

...... ••-•. • ------••• ...... +

•OR TH AURO•AL

AURORALZON• ; •'• /ON• CDC•O•aaOUNDARYOF•OR•/OO•N RE$/ON• Cg ANDC• • ARf •OUNO•QIE•O• $•AO0WOF •AR TN WNER• NO tONS

Figure 5. The motion of ions (dark lines) and electrons(light lines) accordingto Alfv6n's theory[after Cowling, 1942]. The Sunis to the left.

Biermann [1951], who notedthat dust tails, whosespectra was only settledby observationsfrom space. Gringauzet resembledscattered sunlight, could be explainedby light al. [1960] (see also Gringauz [1961]) mountedcharged- pressure,but that the distinction mils often showedhuge particle trapson Lunik 2 (September1959) and later on accelerationswhich could only be accountedfor by a Lunik 3 (October1959), and theydetected far from Earth a "solarcorpuscular radiation." For a long time, however, flow of energeticpositive charges,consistent with solar moredirect evidence was lacking. wind ions and also displayingappropriate modulation due When it wasdiscovered, from spectraof highlyionized to spin of the spacecraft. In 1961 the Massachusetts species [Grotrian, 1939; EdlJn, 1941, 1942, 1945; Instituteof Technologyparticle trap aboardExplorer 10 Shapely, 1960; Billings, 1966, chapter 1; Lang and obtained more derailed evidence for the solar wind [Rossi, Gingerich,1979] that the Sun's coronahad a temperature 1984;Bonetti et al., 1963], and informationconcerning the around106 øK, the question arose of how the Sun's gravity continuous nature of the solar wind came in 1962 from the couldkeep sucha hot atmosphereattached [see Last, 1962; flight of Mariner 2 to Venus [Snyder et al., 1963; Parker, 1964]. Coronal temperaturenear the Sun was Neugebauerand Snyder,1966]. It thenbecame clear that observed not to decrease with height, and this was the Chapman-Ferrarocavity was not a temporaryfeature explainedby the high heat conductivityof the plasma, but existedat all times,and it receivedthe name "magneto- which seemedto precludea stratifiedatmosphere like the sphere,"coined by Gold [1959]. Rapidlyspreading plasma Earth's, with temperature decreasing with height. cloudsproduced by solarflares, like thoseenvisioned by Chapmanproposed a theory in which a static equilibrium Chapman and Ferraro, are sometimessuperposed on the was still possible,yielding moderately lower temperatures solar wind flow. We now know that when the expansion at the Earth's orbit. EugeneParker, however,derived an velocity of such cloudsgreatly exceedsthat of the solar alternative solution in which the corona was not in wind, they are indeedpreceded by collision-freeshocks. equilibrium but insteadcontinually streamedaway from

the Sun to form a high-speed"solar wind" [Parker, 1958; POLAR MAGNETIC STORMS Dessler, 1967;Brandt, 1970]. The processconverted heat to kineticenergy rather efficiently. One additionalpiece of the picture deservesmention: The debate between proponentsof a static corona, magnetic disturbancesassociated with the aurora, like Parker's solar wind theory, and an alternative "solar thoseobserved by Hiorter and Celsius. Suchdisturbances breeze"theory of Chamberlain[1960, 1961;Dessler 1967] are far more intense,rapid, and frequent than magnetic Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE 27,1 / REVIEWS OF GEOPHYSICS ß 111 storms observed at low and middle latitudes. Birkeland detail the main linesof investigation,and many detailsand studied them in 1902-1903 using a network of four namesare by necessityabsent. Written with hindsight,it stations--in Norway and on Iceland, Spitzbergen(Sval- also paints a far tidier picture of magnetosphericphysics bard), and Novaya Zemlya [Birkeland, 1908, 1913; than what actually existed: only throughthe original Bostrbm, 1968]--and concluded that there existed a articlescan the readerrecapture some of the uncertainty, distinct type of magnetic storm, the "elementarypolar confusion,and high "noiselevel" which often obscuredthe magneticstorm" with a typical time scaleof less than an modest achievements described here. B irkeland did not hour,associated with the auroraand with electriccurrents claim to have observedone type of polar magneticstorm which descendedalong auroral field lines and flowed but four or five: only later was it recognizedthat they all horizontallyalong auroral arcs. reflected the same phenomenon. Theories we now Birkeland died in 1918; his work was not followed up recognizeas false, for example,some theoriesof the ring for many years,and in the decadesthat followed,relatively Currentand of the interplanetaryplasma, often drew gmat few magneticstudies were performedat high latitudes. attention, and where investigatorsdid find a reasonable Chapmandid not believe that Birkeland's "polar storms" explanation for one facet, for example, the Chapman- existed. He realizedthat they were muchshorter than the Ferraro cavity or Alfv6n's electric field, they often felt nonpolarstorms with which he was familiar, and in his compelledto fill the rest of the patternwith guesswork encyclopedic two-volume treatise on geomagnetism, which generallydid not standup to the test of time. This jointly written with JuliusBarrels [Chapman and Bartels, sense of confusion often marks work near the limits of data 1940] (see also Chapman [1968]), he suggestedthat and understanding,and it may explain the long delays Birkeland's eventswere probablyjust isolatedphases of which often occur before the truth of a discovery is magneticstorms. He noted there that "a great magnetic generallyacknowledged. storm is a unitary phenomenon,going through regular The picturechanged considerably after 1957, the startof phases"and maintainedthat Birkeland's polar storms"... the InternationalGeophysical Year (IGY). The IGY was seem to be clearly part of a singlephenomenon, waxing an international effort which included the launch of the and waning in unison with the non-polar disturbance first artificial satellites, and it formed a natural transition in field." the historyof magnetosphericphysics. .- The name "substorm,"coined by Chapman for this The implicationsof that transitionare best appreciated phenomenonabout 20 years later, reflected that attitude, in the contextof other researchon our physicalenviron- thoughby then Chapmanmust have begun to realize the ment. The surface of the Earth, the oceans, and atmos- importanceof Birkeland's early observations[Akasofu, phere are completely accessibleand can be directly 1970, p. 603] (see also Siscoe [1980]). We now view studied,even experimentedupon: in the jargon of Earth substormsas impulsiveacceleration events, quite possibly observationfrom space, we have "ground truth." The terrestrialanalogs of solarflares. realm of the astrophysicist,on the contrary, can only be By the 1950s this realizationwas slowly forming, and sensed remotely and imperfectly, and the amount of there was considerableinterest in "magneticbays," large informationwe can ever hopeto receivefrom it is severely magneticdisturbances in the auroral zone which would limited [Harwit, 1981]. By necessityour explanationsof nowadaysbe classifiedas substorms[Silsbee and Vestine, astrophysicalphenomena are lacedwith guesswork,and in 1942]. Currentsflowing into the auroralzone and out of it, many cases(e.g., the origin of cosmicrays) it is quite however,were observedonly in the late 1960s,and their likely that such guesseswill never find convincing global pattern was first mapped in 1974 [Zmuda and confirmation. Armstrong,1974; Iijima and Potemra, 1976]: they are Magnetosphericphysics standshalfway between those now known as Birkelandcurrents [Schield et al., 1969, p. extremes.Until the IGY it was very much like astro- 247]. Contraryto Birkeland'sinterpretation, much of the physics:the magnetospherecould only be sensedremotely, horizontalpart of their circuit, in the ionosphere,flows not and much of what was believed about it was merely alongauroral arcs but perpendicularto them,for by a quirk intelligentguesswork. Then cameartificial satellites and of electrodynamics[Fukushima, 1969, 1976] the main provided some "ground truth," and it is interesting to circuit producesonly a weak magneticsignature on the compare what they revealed with what was believed ground.What Birkeland observed was mostlythe signature beforehand. of an associatedHall current,the "auroralelectrojet" which Many importantmagnetospheric features had indeed parallelsauroral arcs. been inferred before spacecraftwere available, but in almost every case some importantdetail was missingor ASSESSMENT wrong. The Chapman-Ferrarocavity was predictedas a The readershould be cautionedhere that the preceding temporaryrather than permanent feature, and the same was discussionis in no way a completeaccount of pre-space true for the radiation belt. Alfv6n's convection contained a age magnetosphericphysics. It merelydescribes in austere nucleus of truth, but electric field effects supplemented 112 ß REVIEWS OF GEOPHYSICS / 27,1 Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE

ratherthan supplantedthe Chapman-Ferraropicture, and Bohlin,J. David, An observationalder'tuition of coronalholes, in the convectionwhich they producedwas found to flow CoronalHoles and High Speed Wind Streams, edited by Jack B. Zirker,pp. 27-69, ColoradoAssociated University Press, from the tail sunward,opposite to its directionin Alfv6n's 1977. theory. Birkeland's auroral currentsdid exist, but their Bonetti,A., H. S. Bridge,A. J. Lazarus,B. Rossi,and F. Scherb, configurationwas not the one predicted. The existence Explorer10 plasmameasurements, J. Geophys.Res., 68, andimportance of themagnetospheric tail generallywent 4017-4063, 1963. Bostr/Sm,Rolf, Currents in theionosphere and magnetosphere, in unsuspected,and so did the existenceof parallelelectric The BirkelandSymposium on Auroraand Magnetic Storms, fieldsalong auroral arcs, although Alfv6n later developed editedby A. Egelandand J. Holtet, pp. 445-458, Centre the theory of quasi-neutralequilibria, relevant to such Nationalde la RechercheScientifique, Paris, 1968. fields. All this underscores the essential role of in situ Brandt,John, Introduction to the Solar Wind, W. H. Freeman, New York, 1970. observations:one can only speculatehow muchof this Brush,Stephen G., andSubir K. Banerjee,Geomagnetic secular mightbe paralleled in astrophysics. variationand theoriesof the Earth'sinterior, in Past, Present andFuture Trends in Researchin Aeronomyand Geomag- netism,edited by W. Schroeder,Peter Lang, Frankfurt,in press,1988. ACKNOWLEDGMENTS. M. Neugebauerwas the Editor for Bullard,Edward, Edmond Halley (1565-1741), Endcavour,15, this paper. She thanksK. Hutbauerand anotherreferee for their 189-199, 1956. assistancein evaluationof the technicalcontent of thispaper and Carrington,Richard C., Descriptionof a singularappearance S. Kirby for servingas a cross-disciplinaryreferee. seenin the Sun on September1, 1959,Mon. Not. R. Astron. Soc.,20, 13-15,1860. (Reprintedby A. J. Meadows,Early SolarPhysics, p. 181,Pergamon, New York, 1970.) REFERENCES Cawood,John, The magneticcrusade: Science and politics in early VictorianBritain, Isis, 70, 493-519, 1979. Akasofu,S.-I., In memoriamSydney Chapman, Space Sci. Rev., Chamberlain,Joseph W., Interplanetarygas, 2, Expansionof a 11,599-606, 1970. modelsolar corona, Astrophys. J., 131, 47-56, 1960. Akasofu, S.-I., B. Fogle, and B. Haurwitz (Eds.), Sidney Chapman,Sidney, An outlineof a theoryof magneticstorms, Chapman,Eighty, University of ColoradoPress, Boulder, Proc. R. Soc.London, Set. A, 97, 61-83, 1918. 1969. Chapman,Sidney, Arthur Schuster, 1851-1934, J. Geophys. Alfv6n, Hannes,A theoryof magneticstorms and of the aurorae, Res., 39, 341-345, 1934. K. Sven. Vetenskapakad.Handl., Set. 3, 18(3), 1939. Chapman,Sidney, Historical introduction to auroraand magnetic (Reprintedin part with commentsby Alex Desslerand John storms,pp. 21-29, in The BirkelandSymposium on Aurora Wilcox in Eos,51, 180-194, 1970.) and MagneticStorms, edited by A. Egelandand J. Holtet, Alfv•n, Hannes,A theoryof magneticstorms and of the aurorae, CentreNational de la RechercheScientifique, Paris, 1968. II, The aurorae;Ill, The magneticdisturbances, K. Sven. Chapman,Sidney, and Julius Bartels, Geomagnetism, 1049 pp., Vetenskapakad.Handl., Set. 3, 18(9), 1940. OxfordUniversity Press, New York, 1940. Alfv•n, Hannes,Cosmical Electrodynamics, 1st ed., 237 pp., Chapman,Sidney, and Vincent C. A. Ferraro,A newtheory of OxfordUniversity Press, New York, 1950. magneticstorms, Nature, 126, 129-130, 1930. Alfv•n, Hannes,The theory of magneticstorms and auroras, Chapman,Sidney, and Vincent C. A. Ferraro,A new theoryof Nature, 167, 984, 1951. magneticstorms, I, The initial phase,J. Geophys.Res., 36, Alpert,Yakov, 40 yearsof whistlers,J. Atmos.Terr. Phys.,42, 77-97, 171-186, 1931. 1-20, 1980. Chapman,Sidney, and Vincent C. A. Ferraro,A new theoryof Barkhausen, Heinrich, Zwei mit Hilfe der Neuen Verst•rker magneticstorms, I, Theinitial phase (continued), J.. Geophys. entdeckteErscheinungen Phys. Z., 20, 401-403, 1919. Res., 37, 147-156, 421-429, 1932. Barkhausen,H., Whistlingtones from the Earth,Proc. IRE, 18, Chapman,Sidney, and Vincent C. A. Ferraro,A new theoryof 1155-1159, 1930. magneticstorms, I1, The main phase,J. Geophys.Res., 38, Barraclough,David R., Spherical haxmonicmodels of the 79-96, 1933. geomagneticfield, Geomagn.Bull. 8, Inst. of Geol. Sci., Cowling,Thomas Gilbert, On Alfv6n's theory of magnetic Edinburgh,1978. (Also available from Her Majesty's stormsand of the aurorae,J. Geophys.Res., 47, 209-214, StationeryOffice, London.) 1942. Bartels,Julius, Terrestrial-magnetic activity and its relationsto Cowling,Thomas Gilbert, VincentFerraro as a pioneerof solarphenomena, J. Geophys.Res., 37, 1-52, 1932. hydromagnetism,Q. J. R. Astron.Soc., 16, 136-144, 1975. Bartels, Julius, Twenty-sevenday recurrencesin terrestrial- Dessler,Alexander J., Solarwind andinterplanetary magnetic magneticand solar activity,1923-33, J. Geophys.Res., 39, field,Rev. Geophys., 5, 1-41, 1967. 201-202, 1934a. Dessler,Alexander J., Swedish iconoclast recognized after many Bartels, Julius, Arthur Schuster'swork on periodicities,J. years of rejection and obscurity,Science, 170, 604-606, Geophys.Res., 39, 345-346, 1934b. 1970. Biermann, Ludwig, Kometenschweife und solare Kor- Devik, Olaf, KristianBirkeland as I knew him, in The Birkeland puskularstrahlung,Z. Astrophys., 29, 274-286, 1951. Symposiumon Auroraand MagneticStorms, edited by A. Billings,Donald E., A Guide to the Solar Corona,323 pp., Egelandand J. Holtet, pp. 13-19, Centre National de la Academic,San Diego, 1966. RechercheScientifique, Paris, 1968. Birkeland,Kristian, Exp6dition Norvegienne de 1899-1900pour Dibner,Bern, Oersted and the Discoveryof Electromagnetism, l'•tude des aurores bordales, Skr. Nor. Viderisk. Akad. Kl. 1 85 pp., Blaisdell,Waltham, Mass., 1962. Mat. NaturvMensk.Kl., No. 1, 1901. Dragt,Alex J., andJohn M. Finn,Insolubility of trappedparticle Birkeland,Kristian, The NorwegianAurora Polaris Expedition, motionin a magneticdipole field, J. Geophys.Res., 81, 1902-1903,vol. 1, section1, H. Aschehoug,Oslo, 1908. 2327-2340, 1976. Birkeland,Kristian, The NorwegianAurora Polaris Expedition, Dungey,James W., First evidenceand early studiesof the 1902-1903,vol. 1, section2, H. Aschehoug,Oslo, 1913. Earth'sbow shock,Nuovo Cimento C, 2, 655-660, 1979. Stern: HISTORY OF PRE-SPACE AGE MAGNETOSPHERE 27,1 / REVIEWS OF GEOPHYSICS ß 113

Dunnington,G. W., Gauss--Titan of Science,479 pp., Hafner, Grotrian,Walter, The questionof the significanceof the lines in New York, 1955. the spectrumof the solar corona,Naturwissenschaftern, 27, Eather, Robert H., Majestic Lights, 323 pp., AGU, Washington, 214, 1939. (Translatedby B. Doyle, in Source Book in D.C., 1980. Astronomyand Astrophysics,1900-1975, editedby Kenneth Eckersley, T. L., Note on musical atmosphericdisturbances, R. Lang and Owen Gingerich,p. 1122, Harvard University Philos.Mag., 49, 1250-1259, 1925. Press,Cambridge, Mass., 1979. Ed16n,Bengt, An attemptto identify the emissionlines in the Gunn,Ross, An electromagneticeffect of importancein solarand spectrumof the solarcorona, Ark. Mat. Astron.Fys., 28B, no. terrestrialmagnetism, Phys. Rev., 33, 832-836, 1929. 1, 1-4, 1941. Hale, George Ellery, A remarkablesolar disturbance,Astron. Ed16n,Bengt, Die Deutungder Emissionlinienim Spektrumder Astrophys.,11, 611-613, 1982. Sonnenkorona,Z. Astrophys.,22, 30-64, 1942. Harang,Leiv, The Aurorae,John Wiley, New York 1951. Ed16n,Bengt, The identificationof the coronallines, Mon. Not. Harwit, Martin, CosmicDiscovery, 334 pp., Basic Books, New R. Astron Soc., 105, 323-333, 1945. York, 1981. Egeland,Alv, Kristian Birkeland: The man and the scientist,in Helliwell, Robert A., Whistlers and Related Ionospheric MagnetosphericCurrents, Geophys. Monogr. Ser., vol. 28, Phenomena,349 pp., Stanford University Press, Stanford, editedby ThomasA. Potemra,pp. 1-16, AGU, Washington, Calif., 1965. D.C., 1984. Hodgson,R., On a curiousappearance seen in the Sun,Mon. Not. Egeland, Alv, and E. Leer, Kristian Birkeland: His life and R. Astron. Soc., 20, 16, 1860. work, IEEE Trans. Plasma Sci., PS-14, 666-677, 1986. Hoffmeister,C., PhysikalischUntersuchungen auf Kometen,I, Evans, Michael E., Edmond Halley, geophysicist,Phys. Today, Die Beziehungendes prim'firenSchweifstrahl zum Radius- 41, 41-45, 1988. vektor,Z. Astrophys.,22, 265-285, 1943. Faraday, Michael, Experimental researches in electricity, Hoffmeister,C., PhysikalischUntersuchungen auf Kometen, reprintedin Great Booksof the WesternWorld, vol. 45, pp. Die Bewegungder Schweifmaterieund die Repulsivkraftder 255-898, Universityof ChicagoPress, Chicago, II1., 1952. Sonnebeim Kometen,Z. Astrophys.,23, 1-18, 1944. Ferraro,Vincent C. A., The birth of a theory,in SidneyChap- Hulburt, EdwardO., Terrestrialmagnetic variations and aurorae, man, Eighty, edited by S.-I. Akasofu, B. Fogle, and B. Rev. Mod. Phys.,9, 44-68, 1937. Haurwitz,pp. 14-18, Universityof ColoradoPress, Boulder, Hundhausen,Arthur J., Solar activity and the solar wind, Rev. 1969. Geophys.,17, 2034-2048, 1979. Fitzgerald,G. F., Sunspotsand magnetic storms, Electrician, 30, Iijima, Takesi and ThomasA. Potemra,The amplitudedistribu- 48, 1892. tion of field-aligned currents at northern high latitudes Fritz, Hermann,Das Polarlicht,Brockhaus, Leipzig, 1881. observedby Triad, J. Geophys.Res., 81, 2165-2174, 1976. Fukushima,Naoshi, Equivalence in groundgeomagnetic effect of Lamont,J., Ann. Phys.Leipzig, 84, 580, 1852. Chapman-Vestine'sand Birkeland-Alfv6n's current systems Lang, KennethR., and Owen Gingerich(Eds.), SourceBook in for polar magneticstorms, Rep. Ionos. SpaceRes. Jpn., 23, Astronomy and Astrophysics,1900-1975, pp. 122-124, 219-227, 1969. HarvardUniversity Press, Cambridge, Mass., 1979. Fukushima,Naoshi, Equivalencein groundgeomagnetic effect Lindemann,Frederick A., Note on the theoryof magneticstorms, of vertical currents connected with Pedersen currents in the Philos.Mag., 38, 669, 1919. uniformconductivity ionosphere, Rep. Ionos. Space Res. Jpn., Loomis,Elias, On the geographicaldistribution of aurorasin the 30, 35-40, 1976. northemhemisphere, Am. J. Sci.Arts, 30, 89, 1860. Gauss,Carl Friedrich,Allgemeine Theorie des Erdmagnetismus, LSst, Reimar, Interplanetary plasma, Space Sci. Rev., 1, Resultateaus den Beobachtungendes magnetischen Verein im 522-552, 1962. Jahre 1838, 1839. (Translatedinto Englishby Mrs. Sabine, Malin, Stuart, British world magneticcharts, Q. J. R. Astron. revisedby Sir John Herschelin ScientificMemoirs Selected Soc., 1 O, 309-316, 1969. From Transactionsof Foreign Academies and Learned Malin, Stuart, and Edward Bullard, The direction of the Earth's Societiesand From Foreign Journals,vol. 2, pp. 184-251, magneticfield at London, 1570-1975, Philos. Trans. R. Soc. 1841.) London, Ser. A, 299, 357-423, 1981. Gauss,Carl Friedrich,Werke vol. 5, pp. 119-193, KSniglichc Martyn, David F., The theory of magneticstorms and auroras, Gcscllschaftder Wisscnschaftcn,GSttingcn, 1877. Nature, 167, 92-94, 1951. Gilbert,William, De Magnete1600, (On the Magnet,)translated Maunder, E. Walter, Magnetic disturbances,1882 to 1903, as by SilvanusP. Thompson,reprinted from the 1900 editionby recorded at the Royal Observatory,Greenwich, and their Basic Books, New York, 1958. associationwith sun-spots,Mon. Not. R. Astron. Soc., 65, Gillmor, C. Stewart,Coulomb and the Evolutionof Physicsand 2-34, 1904. Engineeringin the EighteenthCentury, 328 pp., Princeton, Meadows,Arthur J., Early Solar Physics,312 pp., Pergamon, 1971. New York, 1970. Gold, Thomas, Comment, in Gasdynamicsof Cosmic Clouds, Meadows,Arthur J., and J. E. Kennedy,The origin of solar- edited by H. C. Van dc Hulst and J. M. Burgers,p. 103, terrestrial studies,Vistas Astron., 25, 419-426, 1982. Interscience, New York, 1955. Mitchell, Crichton, Chapters in the history of terrestrial Gold, Thomas,Motions in the magnetosphereof the Earth, J. magnetism,1, On the directiveproperties of a magnetin the Geophys.Res., 64, 1219-1224, 1959. Earth's field and the origin of the nautical compass,J. Graham,George, Phil. Trans.R. Soc.London, no. 383, 96, 1724. Geophys.Res., 37, 105-146, 1932. Gringauz,K. I., Some resultsof experimentsin interplanetary Mitchell, Crichton, Chapters in the history of terrestrial spaceby means of chargedparticle traps on Soviet space magnetism,2, The discoveryof the magneticdeclination, J. probes,Space Res., 2, 539-553, 1961. Geophys.Res., 42, 242-280, 1937. Gringauz, K. I., V. V. Bczmkikh, V. D. Ozerov, and R. E. Mitchell, ThomasP., and J. A. Bums,Relativistic dynamics of a Rybchinskii,A studyof the interplanetaryionized gas,high point chargein a magneticmonopole field, J. Math. Phys. energyelectrons, and corpuscularradiation from the Sun by N.Y. 9, 2016-2017, 1968. meansof the three-electrodetrap for chargedparticles on the Multhauf, Robert P., and Gregory Good, A Brief History of second Soviet cosmic rocket (in Russian),Doklady Akad. Geomagnetism, 87 pp., Smithsonian Institution Press, Nauk. SSSR,131(6), 1301-1304, 1960. (Englishtranslation, Washington,D.C., 1987. Soy.Phys. Dokl. 5, 361-364, 1960.) Nelson, James H., Louis Hurwitz, and David G. Knapp, 114 ß REVIEWSOF GEOPHYSICS/ 27,1 Stern: HISTORY OF PRE-SPACEAGE MAGNETOSPHERE

Magnetismof the Earth, U.S. GovernmentPrinting Office, Siscoe, George L., What's in the name 'Magnetospheric Washington,D.C., 1962. Substorm'?,J. Geophys.Res., 85, 1643-1644, 1980. Neugebauer, Marcia, and Conway W. Snyder, Mariner 2 Smith, Edward J., Theoreticaland experimentalaspects of ring observationsof the solar wind, 1, Average properties,J. currents,in SpaceScience, edited by Donald LeGalley, pp. Geophys.Res., 71, 4469-4484, 1966. 316-373, JohnWiley, New York, 1963. Newton, Harold W., The 27-day period in terrestrialmagnetic Snyder,Conway W., M. Neugebauer,and U. R. Rao, The solar disturbances,Observatory, 55, 256-261, 1932. wind velocity and its correlationwith cosmicray variations Newton, Harold W., The Face of the Sun, 208 pp., Penguin andwith solarand geomagnetic activity, J. Geophys.Res., 68, Books, New York, 1958. 6361-6370, 1963. Nutting, P. G., Stoermer'swork on the physicsof the aurora,J. Stem, David P., Large-scaleelectric fields in the Earth's Geophys.Res., 13, 23-24, 1908. magnetosphere,Rev. Geophys.,15, 156-194, 1977. Parker, Eugene N., Dynamics of the interplanetarygas and Stoermer, Carl, Sur la situation de la zone de fr6quence magneticfields, Astrophys. J., 128, 664-676, 1958. maximumdes autoresbor6ales d'apr•s la th6orie cor- Parker,Eugene N., The solarwind, Sci. Am., 210, 66-76, April pusculaire,C. R. Acad.Sci. 151, 736-739, 1910. 1964. (Reprintedin Comets,edited by J. C. Brandt, pp. Stoermer,Carl, Sur les trajectoriesdes corpusculeselectrises 75-84, W. H. Freeman,New York 1981.) dans l'espace sous l'actions des magnetismeterrestre avec Poincar6,Henri, Remarquessur une experiencede M. Birkeland, applicationaux auaroresbor6ales, seconde memoire, Arch C. R. Acad. Sci.,123, 530-533, 1896. Sci. Phys. Nat., Ser. 4, 32, 117-123, 190-219, 277-314, Preece,William H., Earth currents,Nature, 49, 554, 1884. 415-436, 505-509, 1911. Ronan,Colin, EdmondHalley, 251 pp., Doubleday,New York, Stoermer,Carl, Sur les trajectoriesdes corpusculeselectrises 1969. dans l'espace sous l'actions des magnetismeterrestre avec Rossi, Bruno, Discovery of the solar wind, in Plasma application aux auarores bor6ales, seconde memoire Astrophysics,ESA Publ. SP-207,pp. 27-31, EuropeanSpace (continued), Arch. Sci. Phys. Nat., Ser. 4, 33, 51-69, Agency,Paris, 1984. 113-150, 1912. Rossi,Bruno, and StanislawOlbert, Introductionto the Physics Stoermer, Carl, Corpusculartheory of the aurora borealis, J. of Space,454 pp., McGraw-Hill, New York, 1970. Geophys.Res., 22, 23-34, 97-112, 1917. Sabine, Edward, On periodicallaws discoverablein the mean Stoermer,Carl, The Polar Aurora, 403 pp., Oxford University effectsof the largermagnetic disturbances, Philos. Trans. R. Press,New York, 1955. Soc. London, 142, 103-124, 1852. Storey,Owen L. R., An investigationof whistlingatmospherics, Saxe, John Godfrey (1816-1887), The blind men and the Philos. Trans. R. Soc. London, Ser. A., 246, 113-141, 1953. elephant,Best Beloved Poems of the AmericanPeople, edited Storey,Owen L. R., Whistlers,Sci. Am., 194, 34-37, Jan.1956. by Hazel Felleman,Doubleday, New York, 1936. Thomson,George, The septuagenarianelectron, Phys. Today, 20, Schield, Milo A., J. W. Freeman, and A. J. Dessler, A source for 55-61, 1967. field-alignedcurrents at aurorallatitudes, J. Geophys.Res., Williams, L. Pearce,Faraday's discovery of induction,Contemp. 74, 247-256, 1969. Physics,5 (1), 28, 1963. Schmidt,Adolf, Das erdmagnetischeAussenfeld, Z. Geophys.,1., Williams, L. Pearce,Michael FaradaymA Biography,531 pp., 3-13, 1924. Basic Books, New York, 1965. Schuster,Arthur, The origin of magneticstorms, Proc. Phys.Soc. Williams,L. Pearce,Andre-Marie Ampere, Sci. Am., 260, 90-97, London Sect. A, 85, 61, 1911. Jan. 1989. Schwabe, Heinrich, Solar observationsduring 1843, Astron. Wright, Helen, Explorerof the Universe,480 pp., E. P. Dutton, Nachr., 20(495), 1851. (Reprintedin Kosmos,edited by A. New York, 1966. yon Humboldt. Also reprintedby A. J. Meadows,in Early Zirker, Jack, Coronal holes and high-speedwind streams,Rev. SolarPhysics, pp. 95-97, Pergamon,New York, 1970.) Geophys.,115, 257-270, 1977. Shamos,Morris H., Great Experimentsin Physics,370 pp., Zmuda, Alfred, and JamesArmstrong, The diurnal patternof Henry Hold, New York, 1959. field-aligned currents,J. Geophys.Res., 79, 4611-4619, Shapley,Harlow (Ed.), SourceBook in Astronomy,1900-1950, 1974. pp. 41-45, Harvard University Press, Cambridge,Mass., 1960. Silsbee,H. C., and ErnestH. Vestine, Geomagneticbays, their frequency and current systems, J. Geophys. Res., 47, D. P. Stern, Laboratoryfor ExtraterrestrialPhysics, Code 195-208, 1942. Singer, S. Fred, A new model of magneticstorms and aurorae, 695, NASA GoddardSpace Flight Center, Greenbelt,MD Eos Trans. AGU, 38, 175-190, 1957. 2O771.