THE ELECTRIC AND LUMINIFEROUS MEDIUM. 
807 
One way of bringing about this desired interaction of magnetic with elastic energy, 
at the same time safeguarding the permanence of the atomic current, would be to 
make it a current of convection, i.e. to suppose the core of the vortex-ring to be made 
up of discrete electric nuclei or centres of radial twist in the medium. The circulation 
of these nuclei along the circuit of the core would constitute a vortex which can move 
about in the medium, without suffering any pressural reaction on the circulating nuclei 
such as might tend to break it up ; the hydrodynamic stability of the vortex, in fact, 
suffices to hold it together. But its strength is now subject to variation owing to 
elastic action, so that the motion is no longer purely cyclic. A magnetic atom, 
constructed after this type, would behave like an ordinary electric current in a non- 
dissipative eircuit. It would for instance be subject to alteration of strength by 
induction when under the influence of other changing currents, and to recovery when 
that influence is removed ; in other words the Weberian explanation of diamagnetism 
would now hold good. 
The monad elements (§ 70) out of which a magnetic molecule of this kind is built 
up are electric centres or nuclei of radial rotational strain. From what is known of 
molecular magnitudes, in connexion with electrochemical data, it would appear that 
to produce an intensity of magnetization of 1700 c.g.s., which is about the limit attain¬ 
able for iron, these monad charges—or electroiiii, as we may call tliem, after Dr. John¬ 
stone Stoney —must circulate very rapidly, in fact with velocities not many hundred 
times smaller than the velocity of radiation.* Even a single pair of electrons revolving 
round each other at such a rate as this would produce a practically perfect secular 
vortical circulation in the medium ; so that a magnetic molecule may quite well be 
composed of a single positive or right-handed electron and a single negative or left- 
handed one revolving round each other in this manner. We may in fact rigorously 
apply to the present problem the principle used by Gauss for the discussion of 
secular effects in Physical Astronomy. Instead of proceeding by addition of the 
elementary effects produced by a planet as it moves from point to point of its orbit, 
Gauss pointed out that the secular results as distinguished from mere periodic 
alternations are the same as if the mass of the planet were supjDosed permanently 
distributed round its orbit so that the density at any point is inversely proportional 
to the velocity the planet would ha.ve when at that point. Just in the same way 
here, the steady flow of the medium, as distinguished from vibrational effects, is the 
same as if each electron were distributed round its circular orbit, thus forming effec¬ 
tively a vortex-ring, of which however the intensity is subject to variation owing to 
the action of other systems.! 
* Let q be the ionic charge, v its velocity, A the area of the orbit and I its length, n the nnniber of 
atoms in 1 cnb. centim.; then n.qjl.v.A = 1700. From electrochemical data we may take nq — 10'’, 
and from molecular dimensions A/Z = D10“®; whence r = 3.10®, which is of the order of about 
one hundredth of the velocity of radiation. This wonld make the periodic time come out about 10 times 
the period of luminous radiations. 
t It may be observed that for the case of a simple diad molecule, composed of two ecjual and opposite 
