278 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1907. 
our present apparatus. Whether they can be photographed in the 
future remains to be seen.* But you will certainly agree that the 
visual observer has a superb opportunity, which the photographic 
observer can not by any possibility take away from him. 
I now wish to speak rather more particularly of another phenome- 
non mentioned here the other night, which is peculiarly adapted for 
investigation with a small solar image. I refer to the differences be- 
tween the spectrum of the center of the sun and the spectrum of the 
sun’s disk near the limb, as shown in the next photograph. Here is 
the spectrum of the center of the sun, and here is the spectrum of 
the sun at a point a short distance inside of the limb. You will see 
at once the remarkable changes that take place. The broad H, and K, 
lines (or bands) are greatly reduced in width; and the same thing oc- 
curs, I think, in the case of all lines that are accompanied by wings. 
In this region of the ultra-violet many of these lines have wings, 
which are lost or greatly reduced near the edge of the sun. This 
causes a remarkable change in the appearance of the spectrum. Sev- 
eral other curious things occur. Not only do these wings change in 
intensity very much, but the central part of the line, which seems to 
be sharply distinguished from the wings, undergoes a decided change 
of intensity also, so that we find. from a preliminary examination of 
the plates that the lines that are strengthened in sun spots are gener- 
ally strengthened near the edge of the sun, while the lines that are 
weakened in sun spots are generally weakened near the edge of the sun. 
This is true, I think, in the great majority of cases. Again, we find 
another curious thing: Almost all of the lines derived from points 
near the sun’s limb are shifted toward the red in the spectrum with 
reference to lines from the center of the disk. But there are some 
striking exceptions, and one of them is most significant: The lines in 
this fluting of cyanogen are not appreciably displaced. As we know 
from laboratory experiments that flutings are not displaced by pres- 
sure, whereas lines are thus displaced, we seem to have an interesting 
confirmation of the conclusion previously reached by Halm from his 
visual observations of two lines in the red—that the displacement of 
these lines is to be ascribed to pressure.” 
@Some of these ‘ reversals,’ since photographed on Mount Wilson, have 
turned out to be Zeeman doublets (or doublets showing the same polarization 
phenomena). Thus it seems possible that these double lines are produced by an 
intense magnetic field in spots. If so, an important new field of research will 
be open to visual observers having telescopes of moderate size. 
6 This conclusion is further confirmed by the fact that lines of a given ele- 
ment, which exhibit unequal displacements at a certain pressure in the labora- 
tory, in general show corresponding displacements near the sun’s limb. It 
remains to be seen, however, whether some other hypothesis may not be equally 
capable of accounting for the observed phenomena. 
