130 
ASTRONOMY: ADAMS AND BURWELL 
If we compare the intensities, of these lines on photographs taken with 
and without an eclipse with the intensities of lines due to such elements 
as iron, we find the Unes of the heavy elements relatively stronger 
on the plates taken without an eclipse. The heavier elements, there- 
fore, are at a comparatively low level in the solar atmosphere, a result 
in harmony with that found from investigations of solar rotation and of 
radial motion in sun-spots. 
The close agreement in wave-length between the bright lines of the 
flash spectrum and the dark lines of the limb spectrum is difficult of 
explanation on the basis of anomalous refraction in the solar atmos- 
phere. If hypothetical irregular density gradients are used to account 
for the displacements of the dark lines at the sun's limb as has been 
done by Juhus, it would seem necessary to conclude that they would 
have a marked effect upon the wave-lengths of the bright lines in the 
flash spectrum for which the path of the light is very similar. No 
such effect is found. Neither the regular density gradient nor the 
presence of irregular density gradients appears to be capable of ex- 
plaining simultaneously the displacements at the sun's limb and the close 
correspondence in wave-length of bright and dark lines without the 
introduction of additional assumptions. The symmetry of the double 
reversals of the flash spectrum fines is also opposed in general to a theory 
which ascribes their origin to anomalous refraction and dispersion in 
the solar atmosphere. 
Some peculiar characteristics of the dark fine spectrum of the sun's 
limb are seen on the photographs of the flash spectrum. In some regions, 
especially near X 5050, the intensities of the solar lines are so greatly 
modified as to render identifications difficult. A number of lines of con- 
siderable intensity make their appearance, which are not seen at all 
in the solar spectrum, and conversely, strong fines in the solar spectrum 
are weakened very greatly at the sun's limb. The identification of the 
elements to which these fines belong should prove to be a matter of 
decided interest. 
The full details of this investigation with tables showing the results 
for 1027 bright lines will appear as Contributions from the Mount Wilson 
Observatory No. 95, in the current volume of the Astrophysical Journal. 
