( 994. ) 
scattering layer, the intensity of the emergent radiation diminishes, 
but at a slower rate than it would do if scattering acted in the 
same way as absorption. Putting for instance, s.¢== 98, we obtain 
R, —0.02S; and then doubling the laver, we find R, = 0,01 5; 
while, if in the original layer an equal loss of 98 percent had 
been caused by absorption, the layer of double thickness would only 
have transmitted 0.0004 5. 
In a vast mass of gas, like the solar atmosphere, even an exceed- 
ingly small absorption-coefficient would suffice to produce a very 
sensible attenuation of the light. We therefore think it much easier 
to understand the narrowness of most of the Fraunhofer lines, and 
their appearance in general, if we assume the absorption coefficient 
to vanish at a very short distance from the middle of each line, so 
that in the rest of the spectrum the distribution of the light only 
depends on scattering’) and other influences (refraction, diffraction, ete). 
Our confidence in the validity of the hypothesis is, however, 
chiefly based en the fact, that it enables one to explain concisely 
and in mutual coherence a great many astrophysical phenomena, e.g. 
the systematic displacements of the Fraunhofer lines, and, if also 
refraction effects are considered, several irregularities in the behaviour 
of the lines, together with many yrrticulars revealed by the speetro- 
heliograph. 
Let us now substitute the value of the scattering coefficient as 
given by (10) into the equation (15); it thus becomes 
on m 
== — ed. 
07 32 4 167? A ft. R? 
We wish to investigate how R, varies with 2. If however, we 
only consider a small part of the spectrum at once, comprising no 
more than a few Angstrom units, we are free to treat 4* and S as 
K 
constants, and may write 
a 
Rese ee Bo, ee FE 
vo atb ( ) 
1) The question may arise whether there are perhaps indications from which 
one might obtain some idea about the magnitude of scattering effects, reasonably 
to be expected in a gaseous medium of the dimensions of the solar atmosphere. 
Now, according to RAYLEIGH’s theory, the average sunlight loses about 5 °/9 of 
its intensity by molecular scattering in passing through our terrestrial atmosphere. 
Substituting R,=0,95S in our formula (15), we find s.t=0,1. If we make the 
very rough estimate, that the solar atmosphere is 50 times as thick as the 
atmosphere of the earth, and has the same average density, we must write for 
the sun: s.t=5, and, consequently, Ro ==; S. This is not an unreasonable result. 
It proves that even with a much smaller density the solar atmosphere would be 
able to produce sensible scattering effects, especially near absorption lines. 
