136 
TChe Constitution of the Sun • 
[March, 
explain the appearance of their absorption lines in the solar 
spedtrum ; and secondly, as to whether or not the prepon- 
derance of one kind of light produces increased absorption 
of that light. 
From spedtroscopic observation we learn that while the 
spedtra of the sun’s chromosphere and prominences are 
made up of a few permanent bright lines , there are others 
which seem to appear and disappear, as if their presence 
depended on the locality under examination. Are we to 
draw from this the inference that the sun’s gaseous envelope 
has not a uniform composition, but that while a few of its 
elements — among which we must class hydrogen — obey the 
laws of gaseous diffusion, there are others which do not, 
owing to their being possessed of some unaccountable pro- 
perty that allows them to be present with, while yet not 
diffused through, their more obedient companions ? Or 
might we not reason from these fadts that the constant 
bright lines indicate an atmosphere as uniform and simple 
as our own, and that the presence of the hydrogen lines 
points to that atmosphere as being impregnated with aqueous 
vapour ? And, further, that the fleeting bright lines do not 
belong to gaseous matter in the sun’s atmosphere, but to 
his discontinuous light refledted to us from his clouds ? 
Lockyer (“ Solar Physics ”), in describing a prominence, 
says — “Judge of my surprise and delight when, upon 
sweeping along the spedtrum, I found hundreds of the 
Frauenhofer lines beautifully bright at the base of the pro- 
minence.” This is exadtly where they should appear if 
they are to be explained by refledtion from the surface 
underneath. 
If the fleeting bright lines do not belong to the sun’s 
atmosphere, then the solid body of the sun does not give a 
continuous spedtrum, as supposed by Kirchhoff, but spedtra 
varying according to the composition of the matter from 
which the light is emanating. Neither is it probable that 
the Frauenhofer lines are due to absorption in a metallic 
gaseous medium, but more probably to a mere crossing of 
the same kind of light rays, as is indicated by the point of 
maximum crossing in our hypothetical spot, corresponding 
with the observed point of maximum absorption in sun- 
spots. 
There is still another point in connedtion with our spot 
hypothesis that requires investigation, namely, the sheaf- 
like rays of light seen to proceed from the sun during total 
eclipse. On the hypothesis of incandescent solid and glow- 
ing vapour, how are we to explain this clubbing together (if 
