ASTRONOMY: C. E. ST. JOHN 
451 
terpretation of astrophysical data. The problem of determining stellar 
motions in the line of sight, a matter of fundamental importance, would 
be confronted with difficulties of a high order, depending as it does upon 
line displacement in stellar, relative to terrestrial spectra. Our knowl- 
edge of the motions, pressure, and many other phenomena in the solar 
atmosphere must be obtained from line displacements in the spectrum, 
but here it would be possible to apply definite corrections, this would 
in many cases, however, modify our interpretations. 
The question of confirmation may be approached through direct ob- 
servation of the displacement of a star near the sun^ but the conditions 
of observation are beset with great difficulties, statistical study of stellar 
masses and motions* or the determination of displacements of solar lines'^ 
under conditions that eliminate other possible causes. 
Some results of an investigation bearing upon the relativity displace- 
ment of lines in the sun's spectrum are here communicated, a full ac- 
count will appear later in a Contribution from the Mount Wilson Solar 
Observatory. 
The substance of the investigation is the behavior of lines in the nitro- 
gen (cyanogen) bands, X 3883, at the center and at the limb of the sun, 
these lines owing to their freedom from pressure shift are in so far well 
fitted for such an investigation, but their compound character in the most 
important series and the frequent superposition of the lines of different 
series may introduce disturbing factors. The line density in the bands 
is high, above ten lines per angstrom. The possible occurrence of blends 
with metallic lines and the closeness of adjacent lines are important con- 
siderations in the selection of lines for observation. As the probability 
of blends is least for the narrowest lines and the precision of measure- 
ment is highest for lines showing no evidence of duplicity and sufficiently 
separated from adjacent lines, ^ a greater weight is attributed to the fines 
of group A than to the broader lines of group B in the accompanying 
table. 
The wave-lengths of the lines were measured in the arc and at the 
center and limb in terms of identical iron standards. The limb-center 
shifts were found by two independent determinations. The sun-arc 
displacements at the center were obtained by direct comparison and by 
three indirect methods. The 'Standard' (R — I) is the mean differ- 
ence, Rowland minus International for lines whose wave, lengths in the 
sun and arc are equal, determined from the iron lines in this region by 
taking account of their sun-arc displacements. 
The mean, zero, given by 43 lines for center-arc requires from the rela- 
tivity point of view a radial movement in the solar atmosphere just 
