Aside from the wrath of gods and demons, it was popular by the 1600s to propose that they were vapors 'exhaled' by the Earth. By the 1700's, Sir Edmond Halley developed many ideas about them having to do with the magnetic properties of the Earth as discovered by his countryman, William Gilbert in the 1600's. Halley proposed that atoms of magnetic matter circulated between the north and south poles of the Earth as a form of 'effluvia' of 'very subtle nature' which produces these auroral displays as a consequence of their motion. The magnetic nature of aurora was confirmed in 1741 by a Swedish observer Olof Hiorter who noted the motion of a compass needle at the time of an auroral display. This led to the idea that aurora were produced by magnetic, or iron- like particles flowing through the air because only magnetic iron or steel could produce magnetic fields.
Jean Jacques de Mairan proposed that aurora were vapors from the Sun that entered Earth's atmosphere, but no one believed this idea in 1733 save for himself. Then in 1779, Benjamin Franklin announced they were an electrical discharge phenomenon...he being the world expert on electricity at the time.
The 1800's were a time of great experimentation, and many expeditions to the polar regions to measure the auroral phenomenon with increasingly sophisticated tools. By 1860, the correlation between solar sunspots and storms, magnetic disturbances, and auroral frequency was established, but it still was not known why the aurora glowed as they did. By 1896, Sir William Crooks had examined 'cathode rays' which also produce a glow on a phosphor screen, and Kristian Birkeland proposed that streams of electrons emitted by the Sun impacted the Earth's atmosphere after being directed by the Earth's magnetic field to the polar regions. Sydney Chapman revised this idea in 1933 by proposing that the Sun emits a plasma, not just electrons, and that vast electromagnetic circuits were set up between the Earth and the Sun, along which the plasma moved. Hans Alfven, in the 1940's revised Chapman's idea by proposing that the Earth's field is swept back by the solar field, and that the distorted field acts like a generator to pump electrons in and out of the upper atmosphere. Aurora and solar streams were not separate, but intimately related. Once satellite observations began in the late 1950's, much of Alfven's theory was confirmed, and in 1970 he received a Nobel Prize for this work.