|
|
|
|
Birkeland, in 1896, first suggested that aurorae and magnetic storms were due to streams of high-speed electrons emitted from the sun and deflected poleward by the earth's magnetic field. Schuster criticized such a one-sign theory on the grounds that a cloud of electrons could not hold together against the electrostatic repulsion of its parts. In 1918, Chapman added at the end of his paper on the average characteristics of magnetic storms an atmospheric-type theory of the origin of such storms, which he ascribed to the action of a stream of charged particles from the sun, mainly of one sign. Lindemann in 1919 criticized Chapman's numerical development of the theory chiefly on the grounds that it involved an accumulation of charge on the earth's atmosphere which would, by electrostatic repulsion, prevent the supposed continued entry of further charges.
Lindemann added to his criticism a suggestion that Chapman's atmospheric theory might be preserved in substance if the supposed stream of charged particles, mainly of one sign, were replaced by a neutral but ionized stream. Lindemann suggested how such a stream might be emitted from the sun by the action of radiation pressure on the atoms and ions, and might travel to the earth, without appreciable recombination, with a velocity of about 800 km/sec. He did not, however, consider its terrestrial effects except to suggest that, on entry, the electrons would be stopped at a higher level than the heavier positive ions, resulting in a vertical separation of charge.
In 1923, Chapman made the first attempt to examine whether such a strictly neutral ionized stream could be detected by the earth's magnetic field so as to impinge mainly in the polar regions of the earth. He concluded that the stream would be deflected only slightly and not in such a way as to produce aurorae. Furthermore, contrary to Lindemann's suggestion, the entry of particles into the atmosphere would not produce the effects predicted by Chapman's theory. No alternative explanation appeared to result from the motion of a neutral ionized stream near the earth.
This was the state of the theory when, after graduating in 1927, I began my researches with Professor Chapman. The close parallel between solar and geomagnetic activity, and the tendency of magnetic storms to recur after one or more solar rotations, seemed to provide strong evidence that streams of solar particles were the cause of magnetic storms. A reexamination of the electrical state of such solar streams during their passage from the sun to the earth removed the lingering hope that the streams might carry a volume charge sufficient to explain the production of aurorae by the bending of the streams in the earth's magnetic field.
Our work confirmed Lindemann's earlier conclusion that the solar streams were electrostatically neutral to a high degree of approximation.
After many abortive attempts to develop a corpuscular theory of magnetic storms, we returned to a consideration of whether, as Chapman had suggested in 1920, the velocities of the ions and electrons differed so that the solar stream could carry an electric current. We devised an ideal problem, which I solved, and which convinced us that the stream could carry no appreciable current or possess any appreciable magnetic field. Thus the only kind of stream on which the corpuscular theory could rest was a neutral ionized stream with the same streaming velocity for both ions and electrons.
The properties of such streams were also largely hypothetical, since at that time there was no direct evidence of the existence of solar streams. The work of Lindemann and Milne on the emission of solar atoms and ions by selective radiation pressure suggested that the streaming velocity would be at least of the order of 1000 km/sec. If it were assumed that the duration of a storm (about two days) was the same as the time during which the earth was enveloped by the stream, then estimates of 1012 to 1013 cm were obtained for the size of the stream.
Although Chapman's first attempt to develop a theory of magnetic storms on the basis of Lindemann's hypothesis was faulty, it was correct in one of its main conclusions: that the ions and electrons within the stream can move approximately together in rectilinear paths without spiraling and with only a slight deflection by the field. However, the phenomena of importance in the production of a magnetic storm are associated with the approach of the stream toward the earth. Moreover, the theory was based on the hypothesis that the earth's magnetic field permeates the stream and polarizes it at every point. At first sight, the assumption appeared reasonable, since the magnetic field tends to deflect the ions and electrons in opposite directions and so separate them. This tendency is opposed by the electrostatic field set up, resulting in a slight polarization of the stream. By considering the case of a very massive stream which would, ipso facto, be almost undeflected by the field, we showed that the hypothesis led to a contradiction.
Our next task was therefore to consider afresh the problem of the advance of a neutral ionized stream in the earth's magnetic field. Our first attempts were still guided by Chapman's hypothesis that the earth's magnetic field would polarize the stream. However, no self-consistent solution could be obtained, and it was at this stage that we began to question the validity of the hypothesis. We did not question the assumption that the electric field experienced by a moving charge in the stream vanishes: if E and B are the electric and magnetic field intensities in an inertial frame of reference, in which the particles have a velocity v, then E + (l/c)(v x B) = 0. Our doubts were confirmed by the special case considered above in which v is constant.1
In looking at the problem again, we argued that since a neutral ionized stream is a good electrical conductor, electric currents would be induced in the stream by its motion in the earth's magnetic field, and this might account for the production of magnetic storms. We envisaged these currents to flow mainly in the body of the stream, and because of the mathematical difficulties inherent in such a problem, Chapman suggested that we should first consider the case of a neutral sheet of charged particles moving in the equatorial plane of the earth's magnetic field, thereby reducing the number of dimensions by one. Chapman advised me to look at Maxwell's papers on the induction of electric currents in plane current sheets. On reading about the induction of electric currents in a plane current sheet by a magnetic system moving toward the sheet, I realized that the field on the side of the sheet containing the magnetic system would be increased. This field was, in fact, represented by the positive image of the system reduced in a certain ratio depending on the electrical conductivity of the sheet. The field of the other side of the sheet would be diminished, and, if the sheet were perfectly conducting, the far side would be completely shielded from the effect of the magnetic system.
It seemed to us that the first phase of a magnetic storm (during which the horizontal component of the earth's magnetic field is increased) could be explained along these lines, since the currents induced appeared to flow near the surface of the stream, so as to reduce practically to zero the intensity of the field within the stream and enhance the field outside. The magnetic field exerts a mechanical force on the surface currents, which tends to oppose the advance of the stream into the stronger parts. Since this retardation is greatest for those parts of the stream surface nearest the earth, a hollow space is formed in the stream around the earth, which compresses the magnetic field. This space is the geomagnetic cavity.
The theory of the first phase described above was sketched by us in the course of a morning. The solution which had eluded us for over 18 months was now firmly in our grasp, and I believe Chapman was as excited as I was on that morning. Much remained to be done, but this was largely a question of developing the details of the theory; over the next three years (1930-33) we encountered new difficulties, some of which were later shown to be illusory and related principally to the reduction of the electrical conductivity of the stream by the earth's magnetic field.
Some months previously we had again considered the Stoermer-Schmidt hypothesis of the ring current, and had shown that for mechanical equilibrium the current had to be westward. This, we argued, would produce the necessary diminution in the earth's magnetic field associated with the main phase of a magnetic storm. We showed that to produce a diminution of, say, 50y during a moderate storm, the mean radius of the ring had to be a few earth radii.
We had not, however, indicated how such a ring current could be formed. We supposed that the ions and electrons would circulate together around the earth in large circles. When the first phase of the theory of magnetic storms was developed, we showed that the wall of the hollow facing the morning side of the earth would be positively charged, and that the wall facing the evening side of the earth would be negatively charged, because of the tendency of the positive and negative charges to be deflected eastward and westward respectively. Such charges would be unstable, and we imagined that the ring current could perhaps be formed by charges escaping from the walls of the hollow and bridging the gap.
Our ideas about the ring current, unlike those relating to the first phase, were not based on hydromagnetic concepts, and our theory of the main phase is today generally considered unacceptable. As Singer, and later Parker and Dessler, showed, the ring current is likely to arise from the trapping of the charged particles escaping from the main body of the stream by the earth's magnetic field, the positive charges spiraling westward and the electrons eastward.
As was reasonable at the time, we assumed that interplanetary space was a complete void. Although we now know this not to be the case, it was indeed fortunate for us; if we had had to deal with the complexities introduced by the presence of an interplanetary medium and magnetic field our task would have been insuperably difficult. But such blissful ignorance is not an uncommon occurrence in science, and it would seem as if Providence had a hand in it. Speaking for myself I now realize how lucky I was to have started my researches in this field at the time when I did. But more than this, to have worked with a scientist as eminent and inspiring as Sydney Chapman has been both an honor and a privilege.
|
|
|
|
