506 
PHYSICS: ST. JOHN AND BABCOCK 
standard and are thus freed from the systematic errors of the Rowland 
system. 
For the main purposes which the new wave-lengths should serve, 
namely, the comparison of the wave-lengths in cosmical and terrestial 
sources, the determination of the constants in series formulae, and the 
investigation of the effect of changed terrestial conditions, thestandard 
and derived lines should be accurately reproducible and their wave- 
lengths should represent what may be considered the fundamental or 
unperturbed vibration of the emitting centers. Recent work at this 
observatory has shown, however, that the 6 mm.-6 amp. iron arc adopted as 
the source for International Secondary Standards, ^ while fulfilling to a 
moderate degree the first condition, gives for large classes of lines wave- 
lengths vitiated by pole effect.^ 
The investigation has involved the development of a standard source 
from which under ordinary working conditions both reproducible and 
fundamental wave-lengths are obtainable. The methods followed and 
the tests applied in its development are given in a Contribution from 
Mount Wilson Solar Observatory where the investigation will be reported 
in full. 
The presence of pole effect in the International arc was shown by direct 
comparison between its center and the center of the proposed or provi- 
sional standard and also by comparing the wave-lengths of the Inter- 
national Secondaries belonging to groups c and d with their wave-lengths 
obtained from the new standard by means of the interferometer. The 
two independent methods gave the same result, namely, a systematic 
error of 4-0.006 A in the International wave-lengths of the lines of these 
two groups. 
The importance of this result lies in the systematic character and the 
magnitude of the error, which is of the order of sun-arc displacements 
upon which our intimate knowledge of solar conditions must depend. 
If the solar wave-lengths of such lines are compared with the published 
International values, the solar lines apparently shift to the violet; but 
when compared with wave-lengths representing the fundamental vibra- 
tions, the same lines are displaced to the red. Such comparisons have 
already led and must necessarily lead to mistaken interpretations of 
solar phenomena.* The normal behavior of the great majority of iron 
lines, the stable lines of groups a and h is displacement to the red in the 
solar spectrum, whatever form of arc is used as the source. When the 
International or the 6 mm. -6 amp. Pfund arc is used, the iron lines of 
groups c and d are displaced to the violet and those of group e abnormally 
to the red, but with the proposed source the lines of these groups yield 
