Action he has erected a consistent scheme of molecular physics in which he finds an explanation of most observed facts.
The discovery by Zeeman of the effect of a strong magnetic field in triplicating or multiplicating the lines in the spectrum of a flame placed in a magnetic field meets with an obvious explanation when we remember that the effect of a magnetic field on an electron in motion is to accelerate it always transversely to its own motion and the direction of the field. Hence it follows that a magnetic field properly situated will increase the velocity of an electron rotating in one direction and retard it if rotating in another. But a linear vibration may be resolved into the sum of two oppositely directed circular motions and accordingly a magnetic force properly applied must act on a single spectral line, which results from the vibration of an electron in such manner as to create two other lines on either side, one representing a slightly quicker and the other a slightly slower vibration.
The notion of an electron or point charge of electricity as the ultimate element in the structure of matter having been accepted, we are started on a further enquiry as to the nature of the electron itself. It is obvious that if the electron is a strain center or singular point in the æther, then corresponding to every negative electron there must be a positive one. In other words electrons must exist in pairs of such kind that their simultaneous presence at one point would result in the annihilation of both of them.
On the view that material atoms are built up of electrons we have to seek for a structural form of atom which shall be stable and equal to the production of effects we find to exist.
The first idea which occurs is that an atom may be a collection of electrons in static equilibrium. But it can be shown that if the electrons simply attract and repel each other at all distances according to the law of the inverse square no such structure can exist. The next idea is that the equilibrium may be dynamic rather than static. That an atom may consist of electrons, as suggested by Larmor, in orbital motion round each other, in fact that each atom is a miniature solar system.
Against this view, however, Mr. T. H. Jeans ('Mechanism of Radiation,' Proc. Phys. Soc. Lond., Vol. 17, p. 760) has pointed out that an infinite number of vibrations of the electrons would be possible about each state of steady motion and hence the spectrum of a gas would be a continuous one and not a bright-line spectrum.
If we are to assume an atom to consist wholly of positive and negative electrons or point charges of electricity, Mr. Jeans has indicated that we may obtain a stable structure by postulating that the electrons, no matter whether similar or dissimilar, all repel each other at very small distances.