Memoirs presented to the Cambridge philosophical society on the occasion of the jubilee of Sir George Gabriel Stokes, bart., Hon. LL. D., Hon. SC. D., Lucasian professor.

394 MR LARMOR, ON THE DYNAMICS OF A SYSTEM OF ELECTRONS OR IONS:
this kind of explanation cannot be of universal application: it would be interesting to
ascertain whether the Zeeman effect is the same for the two sets of constituents of a
double series such that the difference of frequencies is the same all along it. At any
rate, uniformity in the Zeeman effect along a series of lines is evidence that they
are all connected with the same vibrating group: identity of the effect on the two
constituents of a doublet is evidence, as Preston pointed out, that these belong to
modifications of the same type of vibration.
NATURE OF MAGNETIZATION.
6. The proposition above given determines the changes in the periods of the
vibrations of the molecule in the circumstances there defined. But it is not to be inferred
from it that the imposition of the magnetic field merely superposes a slight uniform
precessional motion on the previously existing orbital system. That orbital system will be
itself slightly modified in the transition. For instance, in the ideal case of the magnetic
field being imposed instantaneously, the velocities of all the electrons in the system will
be continuous through that instant: hence the new orbital system on which the precession
is imposed will be the one corresponding to velocities in that configuration which are
equal to the actual velocities diminished by those connected with the precessional motion.
On the usual explanation of paramagnetic induction, the steady orbital motion of each
electron is replaced by the uniform electric current circulating round the orbit which
represents the averaged effect: the circuit of this current is supposed to be rigid so
that the averaged forcive acting on it is a steady torque tending to turn it across the
imposed magnetic field. This mode of representation must however à priori be incomplete: for example it would make the coefficient of magnetization per molecule in a gas
increase markedly with length of free molecular path and therefore with fall of density,
because this torque would have the longer time to orientate the molecule before the
next encounter took place. It appears from the above that the true effect of the imposed
magnetic field is not a continued orientation of the orbits but only a slight change in
the orbital system, which is proportional to the field; and in the simple circumstances
above discussed is made up of a precessional effect of paramagnetic type, accompanied
by a modification of the orbital system which is generally of diamagnetic type, both
presumably of the same order of magnitude and thus very small.
The recognition of this mode of action of the magnetic field also avoids another
discrepancy. If the field acted by orientating the molecules it must induce dielectric
polarization as well as magnetic: for each molecule has its own averaged electric moment,
as revealed by piezoelectric phenomena, and regular orientation would accumulate the
effects of these moments which would otherwise be mutually destructive. But there is
nothing either in the disturbance of the free orbital system into a slightly different
free system, or in the precession imposed on that new systemi-nor in a more general
kind of action of the same type,-which can introduce electric polarization.