l8o REPORTS ON INVESTIGATIONS AND PROJECTS. 



The first tests with the compound quarter-wave plate were made by the 

 Director with the 6o-foot tower in October, but no appreciable shifts of the 

 solar lines could be detected. They were continued in the third-order spec- 

 trum of the 75-foot spectrograph of the 150- foot tower in January. The 

 slit was set on the sun's axis, with its center at 45 ° north or south latitude, 

 the purpose being to detect any slight displacements of the lines due to the 

 extinction of their elliptically polarized red and violet edges by the nicol and 

 successive quarter-wave strips. Telluric lines on the same plates afford 

 excellent checks, as their displacements rarely exceed 0.0002 Angstrom. 

 It may be doubted whether so high a degree of precision has been obtained 

 in any previous measures of lines in the solar spectrum. Most of the lines 

 of solar origin are much less sharp, and consequently the errors of measure- 

 ment are decidedly greater, especially in the less refrangible region, where 

 the largest magnetic effects are to be sought. 



A selected list of lines in the region A 5810 to A 5930, having large separa- 

 tions in the laboratory, showed no displacements exceeding 0.0006 Angstrom. 

 Two iron lines, and a line of unknown origin, nevertheless gave displace- 

 ments as great as 0.0024 Angstrom. Photographs were accordingly taken 

 north and south of the equator, and at different latitudes. The plates, as 

 measured by Miss Lasby, gave opposite signs for the displacements in the 

 two hemispheres, and values decreasing from a maximum near 45 ° to about 

 zero near the equator. As a check on the method, a half- wave plate, inserted 

 between the quarter-wave and nicol, was rotated to different positions, so 

 as to reverse the direction of the shift for a given quarter-wave strip, or 

 eliminate it altogether. In other cases the compound quarter-wave plate, 

 used without the half-wave plate, was inverted, to determine whether the 

 sign of the displacements depended upon its position. Almost without ex- 

 ception both the sign and the order of magnitude of the shifts agreed with 

 the theory, all of the measurements being made by an observer who knew 

 nothing of the conditions under which the plates were taken. 



The same plates were then measured by other observers, who failed to 

 confirm the above results. An extensive series of photographs was then 

 undertaken in the second order of the Michelson grating, as the lines in 

 question were more diffuse in the third order. A form of parallel-plate 

 micrometer, suggested by Mr. Pease, was employed for measuring the plates. 

 The means of a great number of measures, however, failed to confirm the 

 results previously obtained by a single observer on the third-order plates. 

 At present it is impossible to account for these results, if they do not actually 

 represent the general field of the sun. 



The investigation is being continued, but on account of the exceedingly 

 small quantities involved, and the fact that the lines showing apparent shifts 

 are not quite as sharp as some others in the solar spectrum, the work of 

 measurement is difficult.* 



* Dr. Van Maanen has since confirmed Miss Lasby's results for the plates of the 

 first series, and a new series of photographs will be made soon. 



