Here is a excerpt of a paper I found
Chapman's interest in geomagnetic, ionospheric, and auroral observations was only part of his interest in understanding their origin. He was also instrumental in developing the dynamical theory of the origin of the various phases of the geomagnetic storm. In 1882, Balfour Stewart pointed out that the daily variation of the geomagnetic field at the surface of Earth must be caused by electric currents in the upper atmosphere. It was recognized by Stormer that Maxwell's equations require that the observed magnetic storm variations be associated with electric currents. In 1924 Schmidt pointed out that the main phase of a magnetic storm was equivalent to the addition of a southward field of 0.5 x 10-3 to 3.0 x l0-3 gauss across Earth, and postulated that the current associated with this field was in the form of a ring with the ions circulating westward around Earth and the electrons eastward. There was, at the time, a great variety of ideas on the cause of the westward ring current, invoking ultraviolet effects, solar corpuscular emission, etc., in association with solar activity. In 1930, Chapman and Ferraro set up the specific line of thinking that has evolved into the present-day understanding of the geomagnetic storm. They pointed out that the delay of geomagnetic and auroral activity of one or more days following a solar outburst rules out the ideas based on electromagnetic radiation from the sun. Corpuscular emissions with velocities of the order of 103 km/sec must be responsible. Adopting the point made by Lindemann that the solar corpuscular emission must be electrically neutral (because of the enormous electrostatic forces which arise when even very small numbers of charges of one sign are carried away from the sun), Chapman and Ferraro pointed out, and illustrated with a number of idealized examples, that the impact of a cloud or stream of electrons and ions against the geomagnetic field compresses the geomagnetic field. The compression produces an increase in the horizontal component at the surface of Earth. They pointed out that this must be the origin of the initial phase of the magnetic storm. They showed that the incident stream of particles and the geomagnetic field do not interpenetrate, the boundary layer being only 100 km or less in thickness. This picture was verified thirty years later by direct satellite observations of the magnetic field and the solar wind plasma. It is interesting to note, too, that the work of Chapman and Ferraro on the impact of charged particles against the boundary of a magnetic field has been taken up and vastly extended since 1950 in connection with the dynamics of laboratory plasma.