the eye is contained within the great
stratum somewhere before the separation, and not far from the place
where the strata are still united. Then this second stratum will not
be projected into a bright circle like the former, but it will be
seen as a lucid branch proceeding from the first, and returning
into it again at a distance less than a semicircle. If the bounding
surfaces are not parallel planes, but irregularly curved surfaces,
analogous appearances must result."

The Milky Way, as we see it, presents the aspect which has been just
accounted for, in its general appearance of a girdle around the heavens
and in its bifurcation at a certain point, and HERSCHEL'S explanation of
this appearance, as just given, has never been seriously questioned. One
doubtful point remains: are the stars scattered all through space? or
are they near its bounding planes, or clustered in any way within this
space so as to produce the same result to the eye as if uniformly
distributed?

HERSCHEL assumed that they were nearly equably arranged all through the
space in question. He only examined one other arrangement, _viz._, that
of a ring of stars surrounding the sun, and he pronounced against such
an arrangement, for the reason that there is absolutely nothing in the
size or brilliancy of the sun to cause us to suppose it to be the centre
of such a gigantic system. No reason, except its importance to us
personally, can be alleged for such a supposition. Every star will have
its own appearance of a Galaxy or Milky Way, which will vary according
to the situation of the star.

Such an explanation will account for the general appearances of the
Milky Way and of the rest of the sky, supposing the stars equally or
nearly equally distributed in space. On this supposition, the system
must be deeper where the stars appear most numerous.

HERSCHEL endeavored, in his early memoirs, to explain this inequality of
distribution on the fundamental assumption that the stars were nearly
equably distributed in space. If they were so distributed, then the
number of stars visible in any gauge would show the thickness of the
stellar system in the direction in which the telescope was pointed.
At each pointing, the field of view of the instrument includes all the
visible stars situated within a cone, having its vortex at the
observer's eye, and its base at the very limits of the system, the angle
of the cone (at the eye) being