can be certain
that gases at low pressure are present in the object under examination.
And this was precisely what Sir William Huggins discovered to be the
case in many nebulae. When he first decided to study the spectra of
nebulae, he selected for observation those objects known as planetary
nebulae--small, round, or slightly oval discs, generally without central
condensation, and looking like ill-defined planets. The colour of their
light, which often is blue tinted with green, is remarkable, since this
is a colour very rare among single stars. The spectrum was found to be
totally different to that of any star, consisting merely of three or
four bright lines. The brightest one is situated in the bluish-green
part of the spectrum, and was at first thought to be identical with a
line of the spectrum of nitrogen, but subsequent more accurate measures
have shown that neither this nor the second nebular line correspond to
any dark line in the solar spectrum, nor can they be produced
experimentally in the laboratory, and we are therefore unable to ascribe
them to any known element. The third and fourth lines were at once seen
to be identical with the two hydrogen lines which in the solar spectrum
are named F and g.

[Illustration: PLATE E.

NEBULAE IN THE PLEIADES.

_From a Photograph by Dr. Isaac Roberts._]

Spectrum analysis has here, as on so many other occasions, rendered
services which no telescope could ever have done. The spectra of nebulae
have, after Huggins, been studied, both visually and photographically,
by Vogel, Copeland, Campbell, Keeler, and others, and a great many very
faint lines have been detected in addition to those four which an
instrument of moderate dimensions shows. It is remarkable that the red
C-line of hydrogen, ordinarily so bright, is either absent or
excessively faint in the spectra of nebulae, but experiments by Frankland
and Lockyer have shown that under certain conditions of temperature and
pressure the complicated spectrum of hydrogen is reduced to one green
line, the F-line. It is, therefore, not surprising that the spectra of
gaseous nebulae are comparatively simple, as the probably low density of
the gases in them and the faintness of these bodies would tend to reduce
the spectra to a small number of lines. Some gaseous nebulae also show
faint continuous spectra, the place of maximum brightness of which is
not in the yellow (as in the solar spectrum), but about the green. It is
p