350 
THE OUTSIDE OF A STAR 
The radiation pressure supports the atoms above the photosphere in 
the chromosphere; elements not subject to the strong radiation pressure 
do not rise to this level. Thus the difference of pressure is accounted for. 
Whilst this explanation also lays stress on the fact that the elements with 
strong principal lines show the deviation (since it is the absorption by these 
lines which makes them ascend to the chromosphere), it does not assume 
complete opacity above the photosphere—which was the objection to 
Fowler and Milne’s explanation. 
It is rather doubtful at present what the observed “intensity” of a 
line should be taken to signify. In the foregoing it has been supposed to 
be a measure of the blackness. Actually our information depends almost 
wholly on eye estimates of intensity, which presumably combine blackness 
and breadth. Now breadth depends on entirely different considerations 
from blackness. It would seem that photometric measurements of black 
ness at the centres of the lines are essential for a proper application of the 
theory, and the present use of eye estimates is a very uncertain make 
shift. 
What is the temperature referred to in these investigations? The 
absorbing material is in front of the photosphere and presumably nearer 
to the boundary temperature T 0 than to the effective temperature T e . The 
molecular speeds and the density of the radiation will correspond to a 
temperature near to T 0 ; but the quality of the radiation corresponds to T e 
or indeed to a temperature rather higher. The ionising power of the radia 
tion depends on its intensity for high frequencies, and this will be more 
nearly represented by T e than by T 0 . Probably T in the foregoing results 
may be considered to correspond fairly nearly to T e . 
Molecular Sjpectra. 
244. In the reversing layers of cooler stars chemical compounds can 
exist and the molecules give rise to band spectra. For diatomic molecules 
such as cyanogen, CN, the structure of the bands has been unravelled, and 
a satisfactory theory of the distribution of intensity in the bands has been 
developed. So far as we are aware the only attempt to use the band spectra 
as a clue to the physical conditions in the reversing layer is contained in a 
paper by R. T. Birge*. 
The cyanogen band observed in the solar spectrum near 3883 A is 
composed of five overlapping series. Each series starts from a missing 
line labelled m = 0 and runs in both directions; the positive (“P”) branch 
runs towards the red, the successive lines being denoted by m = + 1, 
+ 2, ... ; the negative (“P”) branch runs towards the violet and is denoted 
Astrophys. Journ. 55, p. 273.