lar statistics millions
of stars are classified as if each taken individually were of no more
weight in the scale than a single inhabitant of China in the scale of
the sociologist. And yet the most insignificant of these suns may, for
aught we know, have planets revolving around it, the interests of whose
inhabitants cover as wide a range as ours do upon our own globe.

The statistics of the stars may be said to have commenced with
Herschel's gauges of the heavens, which were continued from time to
time by various observers, never, however, on the largest scale. The
subject was first opened out into an illimitable field of research
through a paper presented by Kapteyn to the Amsterdam Academy of
Sciences in 1893. The capital results of this paper were that different
regions of space contain different kinds of stars and, more especially,
that the stars of the Milky Way belong, in part at least, to a
different class from those existing elsewhere. Stars not belonging to
the Milky Way are, in large part, of a distinctly different class.

The outcome of Kapteyn's conclusions is that we are able to describe
the universe as a single object, with some characters of an organized
whole. A large part of the stars which compose it may be considered as
divisible into two groups. One of these comprises the stars composing
the great girdle of the Milky Way. These are distinguished from the
others by being bluer in color, generally greater in absolute
brilliancy, and affected, there is some reason to believe, with rather
slower proper motions The other classes are stars with a greater or
less shade of yellow in their color, scattered through a spherical
space of unknown dimensions, but concentric with the Milky Way. Thus a
sphere with a girdle passing around it forms the nearest approach to a
conception of the universe which we can reach to-day. The number of
stars in the girdle is much greater than that in the sphere.

The feature of the universe which should therefore command our
attention is the arrangement of a large part of the stars which compose
it in a ring, seemingly alike in all its parts, so far as general
features are concerned. So far as research has yet gone, we are not
able to say decisively that one region of this ring differs essentially
from another. It may, therefore, be regarded as forming a structure
built on a uniform plan throughout.

All scientific conclusions drawn from statistical data require a
critical i