/ 
THE OUTSIDE OF A STAR 
345 
These arguments are not affected by postulating an extended chromo 
sphere—within reasonable limits. If the extension is comparable with or 
greater than the radius of the star the appendage is more naturally 
described as a nebulosity surrounding the star. An emission spectrum is 
then more intelligible. For example, the integrated spectrum of a planetary 
nebula with its central star would show bright lines; this must still be 
explained by the process (2), but the process is afforded more scope by the 
great deviation from thermodynamical equilibrium in the more distant 
part of the nebula. 
We conclude provisionally that bright lines in the spectrum of a static 
star indicate that either (a) the star is greatly disturbed by “thunder 
storms,” or (6) it is a nebulous star. 
Spectra and Temperature. 
240. After the first use of spectroscopy for identifying the elements 
present in a star, the most important advance was the discrimination of 
“enhanced” lines, i.e. lines which are strengthened in the spark spectrum 
compared with the arc spectrum. This leads naturally to the use of spectra 
as a clue to the physical conditions (as well as the chemical conditions) 
in the outside of a star; without going further than the empirical relations 
we can decide that in certain stars the conditions are akin to those pre 
vailing in the electric arc, in others to the spark. This progress was mainly 
due to Sir Norman Lockyer and A. Fowler. 
It is now known that the arc lines are usually due to un-ionised atoms 
and the spark lines to ionised atoms. The spectra not only of singly ionised 
but of doubly and trebly ionised atoms can now be disentangled. For 
example, the well-known H and K lines of calcium are due not to neutral 
Ca but to the singly ionised atom Ca + . Calcium is divalent, that is to say, 
it has only two moderately loose electrons; hence the next stage Ca ++ is 
a very compact ion difficult to disturb. But for the quadrivalent element 
silicon Fowler has identified the spectra of doubly and trebly ionised atoms 
in the stars*. 
A very fertile line of investigation was initiated when M. N. Sahaf 
first brought together the observational indications of the state of ionisa 
tion in the outside of the stars and the modern thermodynamical theory of 
ionisation. By Saha’s theory all the details of stellar spectra become 
quantitatively connected with the temperature and pressure in the revers 
ing layer. From the historical point of view it is to be remarked that the 
thermodynamical theory of ionisation was at that time regarded as a bold 
and tentative generalisation admitting of no terrestrial test. It is due to 
* Proc. Roy. Soc. 103 a, p. 413. 
t Phil. Mag. 40, pp. 472, 809 (1920); Proc. Roy. Soc. 99 a, p. 135 (1921).