24 
ASTRONOMY: C. E. ST. JOHN 
few of the best plates have now been remeasured with the view of ob- 
taining data relative to such lines, and also a few plates of a new series of 
observations on spots belonging to the approaching spot-maximum. 
The displacements of 144 lines have been measured; of these 16 are within 
one angstrom of very strong lines showing great anomalous dispersion. 
For standards of reference the mean displacements for all lines of a given 
intensity are used. Within one angstrom of these strong lines there are 
eight lines to the violet which give a mean residual of + 0.0006 A based 
upon 51 measures, and eight lines to the red which show a mean residual 
of + 0.0003 A based upon 62 measures. The result is practically zero. 
There is, moreover, no systematic variation, as the influencing lines are 
approached from a distance of five angstroms to their near neighborhood. 
The displacements so far considered are of lines in the reversing 
layer and indicate an outflow from the spot- vortex tangential to the solar 
surface. The important chromospheric lines, H and K of calcium, the 
hydrogen lines, the D Hues of sodium and the b group of magnesium, 
give displacements of the opposite sign, interpreted as inflow of the high 
level vapors. Up to the present the theory of anomalous dispersion 
has not been able to suggest an explanation of these negative shifts. 
All these displacements are obtained with the slit of the spectrograph 
parallel to the radius of the solar image passing through the spot. Dis- 
placements of the high level lines are also found when the slit is normal 
to the radius, which on the line-of-sight interpretation indicate vortex 
motion. These displacements have always been troublesome to the 
dispersion theory and are disposed of as occasional phenomena due to 
unequal refraction at the opposite edges of spots under exceptional con- 
ditions. The displacements are, however, a practically constant feature 
of all spots, persist for many weeks, and characterize regular spots during 
at least the greater part of their existence. 
It is probably possible to imagine a density distribution capable of 
explaining these displacements, "Since almost any peculiarity in the 
appearance of spectral lines may be explained by anomalous dispersion, 
if only we are at Kberty to assume the required distributions," as 
JuHus remarks; but it seems questionable that such an artificial condi- 
tion characterizes a regular spot and the mind accepts more readily the 
obvious explanation based upon motion in the line of sight. 
There is one more consideration bearing upon an explanation of these 
displacements based upon the anomalous-dispersion theory, namely, 
the differences characteristic of the elements. In examining the stand- 
ing of a theory, facts that cannot be explained by it are of particular 
