Let us try to conceive this universe before we study its
evolution. I do not adopt any of the numerous devices that have
been invented for the purpose of impressing on the imagination
the large figures we must use. One may doubt if any of them are
effective, and they are at least familiar. Our solar system--the
family of sun and planets which had been sheltered under a mighty
dome resting on the hill-tops--has turned out to occupy a span of
space some 16,000,000,000 miles in diameter. That is a very small
area in the new universe. Draw a circle, 100 billion miles in
diameter, round the sun, and you will find that it contains only
three stars besides the sun. In other words, a sphere of space
measuring 300 billion miles in circumference--we will not venture
upon the number of cubic miles--contains only four stars (the
sun, alpha Centauri, 21,185 Lalande, and 61 Cygni). However, this
part of space seems to be below the average in point of
population, and we must adopt a different way of estimating the
magnitude of the universe from the number of its stellar
citizens.

Beyond the vast sphere of comparatively empty space immediately
surrounding our sun lies the stellar universe into which our
great telescopes are steadily penetrating. Recent astronomers
give various calculations, ranging from 200,000,000 to
2,000,000,000, of the number of stars that have yet come within
our faintest knowledge. Let us accept the modest provisional
estimate of 500,000,000. Now, if we had reason to think that
these stars were of much the same size and brilliance as our sun,
we should be able roughly to calculate their distance from their
faintness. We cannot do this, as they differ considerably in size
and intrinsic brilliance. Sirius is more than twice the size of