round the central nucleus, is always directed _away from_ the central
point, and here it seems to me is the solution of the difficulty which
Newton failed to solve. For when a luminous corpuscle enters any medium,
assuming it to do so, it would have to overcome the pressure due to the
increased elasticity of the denser aetherial envelopes, and as the two
motions, viz. that of the incident ray, and the pressure due to the
elasticity of the elastic envelope, would be in opposition to each
other, the result would be that the luminous corpuscle, if it entered
the medium at all, would be retarded and not accelerated as suggested by
Newton, and such a result is perfectly in harmony with experiment. So
that by our theory of an atomic and gravitating Aether, it seems to me
that it now becomes possible to reconcile the two theories.

There is another difficulty that the emission theory had to contend
with, and that was, how was it possible for the same surface of any
substance to reflect and refract a corpuscle at one and the same time?
Newton overcame this difficulty by suggesting, from the results of his
observations on certain coloured rings, that each particle had what he
called certain phases or fits, of easy reflection or refraction, so that
at certain times they would be refracted, and at other times they would
be reflected.

Boscovitch has suggested that the fits were due to the fact that each
luminous corpuscle possessed polarity; which, by rotating, alternately
offered their different sides to the refractive and reflecting surfaces,
so that sometimes they would be reflected or repelled, and at other
times attracted or refracted.

A similar hypothesis has also been suggested by Biot. Now if such a
hypothesis will satisfactorily account for the fact that the same medium
will reflect or refract the luminous corpuscles, as the case may be,
then in our aetherial atom we have the very conditions which would
satisfy both Boscovitch and Biot's hypothesis. For one of the properties
that we suggested regarding our aetherial atom was, that it possessed
rotation like our own earth, and that it also possessed polarity.

The harmonizing of the two theories, therefore, seems to rest upon the
atomicity or non-atomicity of the Aether.

It is absolutely certain that the electro-magnetic theory of light
demands the recognition of some form of atomicity for the Aether. For if
light be really an electro-magnetic phenomena, as has been