Originally published in:
Eos, Transactions, American Geophysical Union, Vol. 75, No.52,
December 27, 1994, Pages 609, 612-613
An earlier article published in the December 6, 1994, issue of Eos traced the history of solar-terrestrial relations from 1852, when Sabine discovered that the 11-year sunspot and geomagnetic activity cycles waxed and waned in unison, to the early 1890s. The narrative here picks up in 1892, a year of fomentation that led to the development of the corpuscular hypothesis by Maunder Birkeland, Chapman, Ferraro, and others.
Magnetism was "in the air" in 1892. This seminal year saw several remarkable developments, not all of which were positive or bore immediate fruit, but each provided stimulus and a basis for subsequent progress.
E. Walter Maunder, a sunspot expert from Greenwich Observatory, was inspired by the sunspot group of February 1892 (the largest observed up to that time) and an attendant severe storm, to look for similar storms that were associated with earlier large sunspot groups. He noted that the three largest magnetic storms during the preceding 19 years were associated with the three sunspot displays that "stand out with equal distinctness far above all other similar displays." In looking at the second tier of great spots and storms, however, Maunder noted that such spot groups could rotate across the Sun with scarcely a "flutter" of the magnets, and that storms in this class could occur when there were only a few small spots upon the visible disk of the Sun. "Though sunspots are the particular solar phenomenon most easily observed," he concluded, "we must not ... infer that their number and extent afford the truest indication of the changes in the solar activity which produce the perturbations we remark in our magnetic needles."
Sabine had noted that, in the observations made hourly at Toronto and Hobarton, magnetic storms were remarkably synchronous. A few subsequent observations showed that storm onsets at widely separated stations might be virtually "simultaneous." In 1892, Ellis reexamined the question of the simultaneity of storms over the globe with data for 17 storms that were preceded by or commenced with "sudden distinctive movements." Ellis found that onset times of storms in North and South America, Europe, Asia, Africa (Mauritius), and Australia coincided within a few minutes (the timing uncertainty of the recordings). Precisely timed measurements of magnetic storms and active Earth currents in England showed that both phenomena began within a few tenths of a minute. Ellis suggested, contrary to Airy, that both storms and Earth currents had a common cause external to the Earth, "Some sudden action, apparently from without, sets up as it seems, both Earth currents and magnetic activity simultaneously, or nearly so, over the whole Earth." Ellis noted that the daily magnetic variation was associated only with weak and intermittent Earth currents, in contrast to storms. He concluded that while both the daily variation and storms were "probably solar in origin," they were not produced in the same way.
In his Presidential Address to the Royal Society in November, 1892, Lord Kelvin expressed his long-held view regarding the Sun's role in producing magnetic storms. Kelvin was motivated by the work of Ellis, Schuster, and others who had prompted renewed interest in geomagnetism. He allowed that atmospheric currents, as suggested by Schuster, could give rise to the daily variation of Earth's magnetism, but he attacked the notion that direct action of the Sun "as a magnet" could cause magnetic storms. He concluded that during "eight hours of a not very severe magnetic storm, as much work must be done by the Sun in sending magnetic waves out in all directions through [the vacuum of] space as he actually does in four months of his regular heat and light. This result ... is absolutely conclusive against the supposition that terrestrial magnetic storms are due to magnetic action of the Sun; or to any kind of dynamical action taking place within the Sun, or in connexion with hurricanes in his atmosphere, or anywhere near the Sun outside." He made an even stronger inference: "The supposed connexion between magnetic storms and sun-spots is unreal, and the seeming agreement between the periods has been a mere coincidence."
Kelvin did not speculate on the origin of geomagnetic activity but drew attention to the "two good and sure connexions between magnetic storms and other phenomena: the aurora above, and the Earth currents below ... " He challenged future magneticians to persevere in this work.
Earlier that year, Schuster seemed to anticipate Lord Kelvin's misgivings about the Sun as the ultimate source of geomagnetic activity. He asked, "May not the periodicity of sun-spots, and the connection between two such dissimilar phenomena as spots on the Sun and magnetic disturbances on the Earth, be due to a periodically recurring increase in the electric conductivity of the parts of space surrounding the Sun? Such an increase of conductivity might be produced by meteoric matter circulating around the Sun." This hypothesis represents a position "halfway" between those struck by the close agreement of the solar and magnetic 11-year cycles and Kelvin's stance that the Earth's magnetic storms did not originate on the Sun.
In the same month as Kelvin's Presidential Address, the Irish physicist Fitzgerald proposed a corpuscular hypothesis for the source of magnetic storms, suggesting that "a sunspot is a source from which some emanation like a comet's tail is projected from the Sun .... Is it possible, then, that matter starting from the Sun with the explosive velocities we know possible there, and subject to an acceleration of several times solar gravitation, could reach the Earth in a couple of days?" Fitzgerald's work on the corpuscular hypothesis, which displays keen intuition by a highly respected scientist, had little immediate import [cf. Dessler, 1967].
Also in 1892, the spectroheliograph was invented by the American astronomer George Ellery Hale. Hale's spectroheliograph made it possible to isolate narrow emission lines and build up an image of the Sun by scanning a slit across the disk. In this way Hale was able to obtain the first photograph to be published of a solar flare on July 15,1892. The flare, referred to by Hale as a "luminous phenomenon" or "eruption," was observed in white light and was followed within about a day by a severe magnetic storm and an aurora seen as far south as Chicago. Hale declined to draw any definite conclusions from his comparison of solar and magnetic activity in mid-1892, but he did favor a suggestion by Tacchini from Rome that exceptional disturbances in prominences or faculae anywhere on the visible disk could cause geomagnetic activity. Hale found no support for either Veeder's theory that favored "eastem limb" sunspots as sources of magnetic stomms, or for a competing theory attributed to Marchand that regarded sunspots at central meridian as most geoeffective.
In 1898, Ellis made the first major study of solar and geomagnetic variability following Kelvin's address. He extended his 1880 comparison of the diumal range of magnetic declination and horizontal force with sunspot activity through 1896. The results shown in Figure 1 for the 1841-1896 interval were incontrovertible regarding some sort of connection between sunspots and Earth's magnetism. As Ellis stated, "...in the face of such evidence... there would appear to be no escape from the conclusion that such close correspondence, both in period and activity, indicates a more or less direct relation between the two phenomena..." Yet, Ellis was reluctant to interpret his data as unequivocal evidence of a direct connection and, following Schuster, proffered "the existence of some common cause producing sunspots and geomagnetic variation" as a viable alternative. In 1900, Ellis demonstrated that the occurrence frequency of the larger magnetic storms also varied with the sunspot cycle for the 50-year interval from 1848-1897, although the correspondence was more apparent in some intervals than in others.
Father Sidgreaves embraced the idea that a non-solar agency controlled the 11-year synchronous variations of sunspots and storms. On the basis of a statistical study of spot areas and magnetic disturbances from Greenwich for the years 1881-1898, he concluded that, ... the real connexion must be through a common cause of both-something moving, which may pass by near the Sun or near the Earth, and at times near enough to both to produce the two effects together."
Increasingly, attention focused on the details of the
sunspot-storm association. Sidgreaves' colleague, Father Cortie,
was impressed by the failure of a large "naked eye" sunspot in
May 1901 to produce a significant geomagnetic disturbance during
its disk passage. Conversely, the greatest storm of the first
half of 1902 occurred when the visible disk was free of spots.
Ellis drew attention to the semiannual variation of magnetic
activity that had been noted previously by Sabine in the Toronto
data, to show that "the Earth or its near surroundings has a part in
determining the form in which the external influence reaches us."
Such was the state of affiars when Maunder (Figure 2) made yet another association study in 1904. He hoped to get beyond the general link between sunspots and storms that had been demonstrated by the 50-year study of Ellis but which always failed in the details. Like Ellis, he separated storms according to their sizes. Maunder found that the greatest geomagnetic storms occurred when the most important visible spot group was located between 19o E and 47o W solar longitude. Of the 19 "great" storms in his sample, he found that 16 were associated with the presence of a large spot group within the above confines near central meridian and three other cases for which a spot group, at one time very large, "had returned in a diminished form to the central meridian." These three exceptions provided Maunder with a point of departure for a follow-on study in which he considered storms ranging in size from "moderate" to "great" and established the 27-day recurrence tendency of geomagnetic activity.
That analysis (Figure 3) caused a sensation when it was first presented. Maunder argued that magnetic disturbances arranged in patterns corresponding to the synodic solar rotation period of 27 days could only mean that the Sun, not some outside agency that produced both spots and storms, was the source of the disturbances. Maunder drew attention to the extended sequence of six recurrent storms in 1889: the first three were accompanied by central meridian spots and the last three were not. He concluded that sunspots were not a necessary condition for magnetic storms; areas on the Sun could be "magnetically active" before the formation and after the disappearance of sunspot groups. Maunder also considered faculae and prominences as indicators of magnetic activity, but neither of these phenomena distinguished themselves from sunspots. All phenomena tracked each other.
Maunder went further. "The areas of the Sun giving rise to our magnetic disturbances are definite and restricted areas, as the definiteness with which certain longitudes are indicated [in Figure 3] proves." He argued that the "sharp" movements with which most large storms began indicated that "the influence proceeding from the Sun, whatever its character, ... does not radiate like heat or light, but its action is confined to a definite and very restricted direction." He added that the repetitive nature of storms at synodic rotation intervals could only be explained "by supposing that the Earth has encountered, time after time, a definite stream, a stream which, continually supplied from one and the same area of the Sun's surface, appears to us, at our distance, to be rotating with the same speed as the area from which it rises."
From the typical duration of a storm, Maunder deduced an average stream diameter of ~20o. This reduced Lord Kelvin's estimate of the work that must be done by the Sun to produce a moderate magnetic storm by a factor of 130, since Kelvin had assumed that the disturbance was isotropic. "That, therefore, which Lord Kelvin spoke of twelve years ago as 'the fifty years' outstanding difficulty [in the way of believing the Sun to be the direct cause of magnetic storms in the Earth] is now rendered clear. Our magnetic disturbances have their origin in the Sun."
Moreover, Maunder suggested that, while the physical cause of the coronal streams deemed responsible for storms was unknown, the corpuscular hypothesis in the form proposed by Arrhenius that invoked radiation pressure to drive charged particles from the Sun was consistent with the observations.
Maunder's research was strongly attacked. Cortie disputed his observations and logic. Schuster argued against Maunder's conclusion that the energy expended during magnetic storms was supplied by the Sun and chided him for "his somewhat boastful claim" that he had resolved "the 50 years outstanding difficulty" referred to by Lord Kelvin. In a meeting of the Royal Astronomical Society in January 1905, Maunder, who joked beforehand of the "severe heckling" he expected his paper to receive, answered his critics masterfully. Equally as important, early, more sophisticated mathematical analyses conducted by Schuster and Dyson supported the existence of the periodicity.
The Englishman Chree, who would later provide conclusive evidence for the 27-day recurrence tendency with his superposed-epoch analysis, provided a thorough critique of Maunder's work. Chree concluded that although further evidence was required to justify the final acceptance of Maunder's views his paper was a most important one. "The theory advanced hardly touches the physical side of the problem," Chree wrote, "but it is clear and definite so far as it goes, and is obscured by no mystifications of language. It is certainly incomplete, and may be wholly erroneous, but it is exceedingly suggestive, and indicates a number of lines of research which can hardly fail to lead to valuable results."
Suggestions and claims for the 27-day periodicity had been made by several scientists during the second half of the 19th century, but it was Maunder's remarkable figure and his recognition of its implications that provided an effective counterweight to Lord Kelvin's objection. Maunder's work not only made it difficult to argue against the Sun as the source of magnetic storms, it also gained support for the corpuscular theory as the means by which the Sun exerted its influence.
Kr. Birkeland of Oslo was the primary proponent of the corpuscular theory at the turn of the century. He directed cathode rays at a "terella", a magnetized globe, to produce phosphorescence patterns at the poles that were analogous to terrestrial aurorae. His experiments supported the idea that solar disturbances were propagated to Earth as beams of electrified particles. Birkeland inspired Stormer to undertake calculations of orbits of charged particles in the Earth's magnetic field. The work of Birkeland and Stormer marks the beginnings of magnetospheric physics.
Other developments in the early 20th century supported the fledgling corpuscular theory of magnetic storms. In his epochal discovery of solar magnetism in 1908, Hale pointed out that the fields in sunspots were inadequate to account for storms by direct magnetic action, and that the origin of storms "may be sought with more hope of success in the eruptions shown on spectroheliographic plates in the regions surrounding spots." Additional evidence was not long in coming. Great storms in 1908 and 1909 were preceded by spectroheliograph observations of chromospheric eruptions. Perhaps the best indication that the field was turning to the corpuscular hypothesis was a priority claim for the hypothesis made by Lodge in 1909 that provoked a response by Chree in support of Birkeland.
The corpuscular hypothesis championed by Maunder and Birkeland was not unopposed, however. An alternative hypothesis was proposed by Schuster in 1905 and later adopted by Bauer. In this picture, solar emanations were viewed as a "trigger" that released energy stored in upper atmospheric currents as a result of the Earth's rotation. The key difference between Maunder's and Schuster's hypotheses was in the ultimate source of storm energy-either in the Sun or in the Earth.
In 1918-1919 the Englishman Sydney Chapman, a leading figure of solar-terrestrial physics for the next half century, suggested that streams of alpha-particles emanating from the Sun, rather than electrons, as Birkeland and others hypothesized, caused geomagnetic storms. Chapman also argued for a separation of the solar sources of the diumal and storm variations, concluding that the daily variations were almost certainly controlled by the Sun's ultraviolet emissions.
In 1919, Lindemann contested Chapman's alpha particle model of storms because the stream could not propagate cohesively for more than 1-2 solar radii because of mutual repulsion and could not approach the Earth after the first few seconds because of charge build-up. Lindemann then proposed that the expelled cloud of particles must contain an approximately equal number of positively and negatively charged particles, specifying what is now referred to as a "plasma."
Following Lindemann's suggestion, Chapman revised his model and, with Ferraro in 1931, proposed the first "modern" theory of magnetic storms. The region of space surrounding Earth from which the corpuscular stream or cloud was excluded by the action of the Earth's magnetic field came to be known as the Chapman-Ferraro cavity and eventually, as the magnetosphere.
The Chapman and Ferraro paper in 1931 established the corpuscular hypothesis as the means by which the Sun gave rise to geomagnetic disturbances. The notion of two different types of streams, originating in different sources on the Sun, remained unknown, however.
A subsequent article will show how sunspots-which provided the original evidence for a solar activity-magnetic storm link-were replaced by flares and complemented by M regions as sources of geomagnetic activity. That picture, in turn, was altered and refined with the modern discoveries of coronal mass ejections and coronal holes.
Chapman, S., and V.C.A. Ferraro, A new theory of magnetic storms, 1, The initial phase, J. Geophys. Res., 36, 77, 1931.
Cliver, E.W., Solar activity and geomagnetic storms: The first 40 years (1852-1891), Eos, 75, 569, 1994.
Dessler, AJ., Solar wind and interplanetary magnetic field, Rev. Geophys., 5, 1, 1967.
Maunder, E.W., Magnetic disturbances, 1882 to 1903, as recorded at the Royal Observatory, and their association with sunspots, Mon. Not. R Astron. Soc., 65, 2, 1905.