pot where our solar system is
situated. And the many small arms which stretch out from the Milky Way
would have to be either planes seen edgeways or the convexities of
curved surfaces viewed tangentially. The improbability of these various
assumptions is very great. But evidence is not wanting that the
relatively bright stars are crowded together along the same zone where
the excessively faint ones are so closely packed. The late Mr. Proctor
plotted all the stars which occur in Argelander's great atlas of the
northern hemisphere, 324,198 in number, on a single chart, and though
these stars are all above the tenth magnitude, and thus superior in
brightness to that innumerable host of stars of which the individual
members are more or less lost in the galactic zone, and on the
hypothesis of uniform distribution ought to be relatively near to us,
the chart shows distinctly the whole course of the Milky Way by the
clustering of these stars. This disposes sufficiently of the idea that
the Milky Way is nothing but a disc-like stratum seen projected on the
heavenly sphere; after this it is hardly necessary to examine Professor
Barnard's photographs and see how fairly bright and very faint regions
alternate without any attempt at regularity, in order to become
convinced that the Milky Way is more probably a stream of stars
clustered together, a stream or ring of incredibly enormous dimensions,
inside which our solar system happens to be situated. But it must be
admitted that it is premature to attempt to find the actual figure of
this stream or to determine the relative distance of the various
portions of it.

[Illustration: PLATE XVI.

NEBULAE

OBSERVED WITH LORD ROSSE'S GREAT TELESCOPE.]

CHAPTER XXIII.

THE PHYSICAL NATURE OF THE STARS.

Star Spectroscopes--Classification of Stellar Spectra--Type I.,
with very Few Absorption Lines--Type II., like the Sun--Type III.,
with Strongly Marked Dark Bands--Distribution of these Classes over
the Heavens--Motion in the Line of Sight--Orbital Motion Discovered
with the Spectroscope: New Class of Binaries--Spectra of Temporary
Stars--Nature of these Bodies.

We have frequently in the previous chapters had occasion to refer to the
revelations of the spectroscope, which form an important chapter in the
history of modern science. By its aid a mighty stride has been taken in
our attempt to comprehend the physical constitution of the sun. In the