ugh utilization of the multiform
records accumulated during years of study, to define the various typical
stages of the sidereal evolution; and not merely to define them but
to illustrate them practically by citing stars which belong to each
of these stages, and to give them yet clearer definition by naming the
various elements which the spectroscope reveals as present in each.

His studies have shown that the elements do not always give the same
spectrum under all conditions; a result quite at variance with the
earlier ideas on the subject. Even in the terrestrial laboratory it
is possible to subject various metals, including iron, to temperatures
attained with the electric spark at which the spectrum becomes different
from that, for example, which was attained with the lower temperature
of the electric arc. Through these studies so-called series-spectra
have been attained for various elements, and a comparison of these
series-spectra with the spectra of various stars has led to the
conclusion that many of the unknown lines previously traced in the
spectra of such stars are due to the decomposition products of familiar
elements; all of which, of course, is directly in line of proof of the
dissociation hypothesis.

Another important result of Professor Lockyer's very recent studies has
come about through observation of the sun in eclipse. A very interesting
point at issue all along has been the question as to what layers of the
sun's atmosphere are efficient in producing the so-called reverse lines
of the spectrum. It is now shown that the effect is not produced, as
formerly supposed, by the layers of the atmosphere lying just above the
region which Professor Lockyer long ago named the chromosphere, but by
the gases of higher regions. Reasoning from analogy, it may be supposed
that a corresponding layer of the atmosphere of other stars is the
one which gives us the reverse spectrum of those stars. The exact
composition of this layer of the sidereal atmosphere must, of course,
vary with the temperature of the different stars, but in no case can
we expect to receive from the spectroscope a full record of all the
substances that may be present in other layers of the atmosphere or in
the body of the star itself. Thus, for example, the ordinary Freuenhofer
spectrum of the sun shows us no trace of the element helium, though
through other observations at the time of eclipse Professor Lockyer had
discovered that element there,