that
the stars which form the agglomerations of the Milky Way have, beyond
doubt, the same average motion as the stars in other regions of the
universe. The difficulty is that these stars appear to us so faint
individually, that the investigation of their spectra is still beyond
the powers of our instruments. But the extraordinary feat performed at
the Lick Observatory of measuring the radial motion of 1830
Groombridge, a star quite invisible to the naked eye, and showing that
it is approaching our system with a speed of between fifty and sixty
miles a second, may lead us to hope for a speedy solution of this
question. But we need not await this result in order to reach very
probable conclusions. The general outcome of researches on proper
motions tends to strengthen the conclusions that the Keplerian sphere,
if I may use this expression, has no very well marked existence. The
laws of stellar velocity and the statistics of proper motions, while
giving some color to the view that the space in which we are situated
is thinner in stars than elsewhere, yet show that, as a general rule,
there are no great agglomerations of stars elsewhere than in the region
of the Milky Way.

With unity there is always diversity; in fact, the unity of the
universe on which I have been insisting consists in part of diversity.
It is very curious that, among the many thousands of stars which have
been spectroscopically examined, no two are known to have absolutely
the same physical constitution. It is true that there are a great many
resemblances. Alpha Centauri, our nearest neighbor, if we can use such
a word as "near" in speaking of its distance, has a spectrum very like
that of our sun, and so has Capella. But even in these cases careful
examination shows differences. These differences arise from variety in
the combinations and temperature of the substances of which the star is
made up. Quite likely also, elements not known on the earth may exist
on the stars, but this is a point on which we cannot yet speak with
certainty.

Perhaps the attribute in which the stars show the greatest variety is
that of absolute luminosity. One hundred years ago it was naturally
supposed that the brighter stars were the nearest to us, and this is
doubtless true when we take the general average. But it was soon found
that we cannot conclude that because a star is bright, therefore it is
near. The most striking example of this is afforded by the absence of
me