as we have seen, some thirty years before
anything was known of it on the earth.

In a recent eclipse photographs were taken of the spectra of the lower
part of the sun's atmosphere by itself, and it was found that the
spectrum of this restricted area taken by itself gave the lines which
specialize the spectra of so different a star as Procyon. "I recognize
in the result," says Professor Lockyer, "a veritable Rosetta Stone which
will enable us to read the celestial hieroglyphics presented to us in
stellar spectra, and help us to study the spectra and to get at results
much more distinctly and certainly than ever before."

But the most striking confirmation which the meteoritic hypothesis has
received has come to hand through study of the spectrum of the new star
which appeared in the constellation Perseus in February, 1901, and which
was so widely heralded everywhere in the public press. This star was
discovered on the morning of February 22d by star-gazers in Scotland,
and in America almost simultaneously. It had certainly not been
visible a few hours before, and it had blazed up suddenly to a greater
brilliancy than that of a first-magnitude star. At first it was
bluish-white in color, indicating an extremely high temperature, but
it rapidly subsided in brilliancy and assumed a red color as it cooled,
passing thus, in the course of a few days, through stages for which
ordinary stars require periods of many millions of years.

The most interesting feature of the spectrum of this new star was the
fact that it showed both light and dark lines for the same substances,
the two lying somewhat apart. This means, being interpreted, that some
portions of a given substance are giving out light, thus producing
the bright lines of the spectrum, and that other portions of the same
substance are stopping certain rays of transmitted light, thus producing
the dark lines. The space between the bright and dark lines, being
measured, indicated that there was a differential motion between the
two portions of substance thus recorded of something like seven hundred
miles a second. This means, according to theory--and it seems hardly
possible to explain it otherwise--that two sidereal masses, one at least
of which was moving at an enormous rate of speed, had collided, such
collision, of course, being the cause of the incandescence that made the
mass suddenly visible from the earth as a new star.

New stars are by no means every-day affa