s oscillation be a sufficient origin of
wave-motion, each distinct particle of the wave _m_ _n_ ought to give
birth, to a series of circular waves. This is the important point up
to which I wish to lead you. Every particle of the wave _m_ _n_ _does_
act in this way. Taking each particle as a centre, and surrounding it
by a circular wave with a radius equal to the distance between _m_ _n_
and _m'_ _n'_, the coalescence of all these little waves would build
up the large ridge _m'_ _n'_ exactly as we find it built up in nature.
Here, in fact, we resolve the wave-motion into its elements, and
having succeeded in doing this we shall have no great difficulty in
applying our knowledge to optical phenomena.

[Illustration: Fig. 17.]

Now let us return to our slit, and, for the sake of simplicity, we
will first consider the case of monochromatic light. Conceive a series
of waves of ether advancing from the first slit towards the second,
and finally filling the second slit. When each wave passes through the
latter it not only pursues its direct course to the retina, but
diverges right and left, tending to throw into motion the entire mass
of the ether behind the slit. In fact, as already explained, _every
point of the wave which fills the slit is itself a centre of a new
wave system which is transmitted in all directions through the ether
behind the slit_. This is the celebrated principle of Huyghens: we
have now to examine how these secondary waves act upon each other.

[Illustration: Fig. 18.]

Let us first regard the central band of the series. Let AP (fig. 18)
be the width of the aperture held before the eye, grossly exaggerated
of course, and let the dots across the aperture represent ether
particles, all in the same phase of vibration. Let E T represent a
portion of the retina. From O, in the centre of the slit, let a
perpendicular O R be imagined drawn upon the retina. The motion
communicated to the point R will then be the sum of all the motions
emanating in this direction from the ether particles in the slit.
Considering the extreme narrowness of the aperture, we may, without
sensible error, regard all points of the wave A P as equally distant
from R. No one of the partial waves lags sensibly behind the others:
hence, at R, and in its immediate neighbourhood, we have no sensible
reduction of the light by interference. This undiminished light
produces the brilliant central band of the series.

Let us now consider th