ASTRONOMY: C. E. ST. JOHN 
459 
companions at an average distance of 0.314 A to the red is +0.0042A; 
for 59 lines with companions at an average distance of 0.270 A to the 
violet it is +0.0040 A. The displacements for the three cases are equal 
within the limits of accidental error, and, within such limits, the equality 
shows the absence of mutual influence. 
Since a detectable mutual repulsion between adjacent solar lines 
should increase their separation over that determined from terrestrial 
sources, another approach to the question is found through such a 
comparison. It is usually assumed that repulsion occurs between all 
closely adjacent solar lines. For 45 pairs, mean separation 0.29 A, 
the differences AX Sun — AX Arc are positive for 17, negative for 21 and 
zero for 7 lines, the mean differences being zero. It has recently been 
suggested^ that repulsion is to be expected only if the line in the solar 
spectrum has a closely adjacent line due to another substance not 
present in the arc. The following data from recent determinations do 
not show the differences required by the theory, though the lines are 
due to different elements and originate in separate arcs. 
A Rowland 
AX Arc 
AXSun — AXArc 
4058.915 
Fe 
.081 
Mn 
.170 
.173 
-0.003 
4226.904 
Ca 
.606 
Fe 
.700 
.710 
-0.010 
4315.138 
Ti 
.262 
Fe 
.115 
.115 
0.000 
4427.266 
Ti 
.482 
Fe 
.215 
.213 
+0.002 
4454.552 
Fe 
.953 
Ca 
.399 
.399 
0.000 
4489.911 
Fe 
.253 
Mn 
.340 
.340 
0.000 
5208.596 
Cr 
.776 
Fe 
.169 
.170 
-0.001 
5446.797 
Ti 
.130 
Fe 
.332 
.334 
-0.002 
Mean 
.305 
.307 
-0.0018 
Recently Albrecht^ found by comparing the Rowland and Inter- 
national wave-lengths of iron lines that the violet and red components 
of a solar pair showed apparent displacements to the violet and red of 
0.007 and 0.005A, respectively. He interpreted these as effects of 
anomalous dispersion, an interpretation accepted by JuKus, who says,^ 
^'Only if the Fraunhofer lines are mainly due to anomalous dispersion 
will they be able to show a mutual influence of the observed kind and 
magnitude. " The absence of mutual influence v/ould, within the limits 
