ASTRONOMY: C E. ST. JOHN 
25 
interest. The existence, then, of such differences becomes of great 
importance, since as yet it has not been possible to suggest an expla- 
nation of them in harmony with the dispersion theory. As standards 
of reference the displacements of the iron lines of hke solar intensity 
measured upon the same plates have been used. The residuals (that is, 
displacement for the element minus the mean displacement of iron lines 
of the same intensity) are given in thousandths of angstroms in the 
accompanying table: 
Element 
Ba 
Ce 
Cd 
La 
Nb 
Nd? 
Pb 
Yt 
T 
Number of lines 
2 
1 
1 
9 
1 
3 
1 
1 
47 
137 
142 
112 
138 
94 
144 
207 
173 
48 
Mean residual in O.OOIA. 
+5 
+7 
+5 
+2 
+ 10 
0 
+6 
+6 
-3 
The residuals for the group of heavy elements are systematically 
positive and are undoubtedly real, while the residuals for titanium are 
systematically negative and likewise exceed the errors of measurement. 
The displacements considered in this paper may all be interpreted in 
terms of motion and level as shown in contributions already pubhshed.^ 
The purpose of the present discussion has been to show their bearing upon 
the theory of anomalous dispersion in its appHcation to certain solar 
phenomena of which the theory has suggested partial explanations and 
to which it has appealed for evidence of its influence in the solar at- 
mosphere. The only quantitative criterion, mutual influence, fails to 
confirm the theory, and in my opinion there are well estabHshed re- 
sults of these observations for which the theory has no explanation. 
^W. H. Julius: Radial motion in sun-spots. Astrophys. J., 40, 1 (1914). 
2 Contrib. ML Wilson Solar Obs., No. 69. 
3 Ibid, No. 93. 
4 Ibid, Nos. 69 and 74; Astrophys. J., 37, 322 (1913); 38, 341 (1913). 
