THE OUTSIDE OF A STAR 
343 
/ 
is aggravated for ‘'subordinate lines 55 which are absorbed by only a small 
proportion of the atoms present. On the other hand any degree of contrast 
can be admitted in the H and K lines and a few other prominent lines 
which belong to the chromosphere rather than the reversing layer and are 
formed under conditions different from those laid down in § 234. 
Emission Lines. 
239. The spectra of certain stars contain bright lines superposed on 
the continuous spectrum. The Balmer series of hydrogen and the enhanced 
lines of iron are specially liable to this reversal, some of the lines appearing 
bright instead of dark in a spectrum which would otherwise be considered 
normal. The phrase “star with peculiar spectrum” usually has reference 
to bright lines. In type 0 bright lines are fairly common, the stars with 
this feature being called Wolf-Rayet stars; but emission lines occur also 
in other types. We shall not here consider the bright lines in long-period 
variables where the phenomenon is evidently associated with the varying 
physical conditions and is not so mysterious as in an apparently static 
star; nor do we consider such phenomena as dark lines with bright centres 
where the brightness is not compared with the regular photospheric 
background. 
It is very difficult to account for a bright line in a static star. We 
can scarcely attribute it to special abundance of the atom or ion concerned 
since that is already invoked to explain strong absorption lines. 
To contribute to an emission line of frequency v an atom must emit 
without having previously absorbed this frequency. The double process of 
absorption followed by emission adds nothing to the amount of radiation 
of frequency v present; it merely helps to equalise the flow in all directions, 
so that the outward flow is at any rate not strengthened by it. There are 
only three ways in which emission can occur without previous absorption 
of the same frequency— 
(1) The atom may be brought by a collision into the state necessary 
to emit. 
(2) It may be brought to the required state by absorption or emission 
of other lines of its spectrum. 
(3) It may capture an electron in an excited orbit. 
As regards (1), since the emitting material must clearly be in front of 
the photosphere the speeds of the electrons correspond to a temperature 
near T 0 , or at any rate less than T e . By collision they excite normal atoms 
and de-excite excited atoms, and would by themselves just maintain the 
number of excited atoms corresponding to equilibrium at their temperature. 
To give a line showing bright against the photosphere there must be many