332 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1908. 



TLiis result has been confirmed by otber photographs, which indicate that the 

 direction of the displacement always depends upon the direction of revolution 

 in the vortex. 



PLANE POLARIZATION ACROSS THE LINES OF FORCE. 



So far we have confined our attention lo polarization phenomena observed 

 along the lines of force. But it is well known that the doublets are, in general, 

 transformed into triplets when observed in a magnetic field at right angles to 

 the lines of force. The components of the triplets are plane polarized, the 

 central line in a plane at right angles to the plane of polarization of the side 

 components. It should be possible to detect similar phenomena in spot spectra, ^ 

 if they are produced in a magnetic field. 



It natui-ally happens that these spectra are most commonly observed when 

 the spots are not vei*y f^i" removed from the center of the sun, because fore- 

 shortening near the limb reduces the umbra to a narrow strip diflicult to keep 

 on the slit. This n)ay partially explain why our photographs of spot spectra, 

 taken without polarization apparatus, show the doublets without a trace of a 

 central component. But it does not account for the failure of the central line 

 to appear in the spectra of spots well removed from the center. It is true that 

 a few triplets occur in all of our spot spectra, such as X5781.97, X60G4.85, and 

 \6173.r)5. But these I have regarded as probable examples of an exceptional 

 type of lines, observed in the laboratoiy as triplets along the lines of force. 

 * * * * * * * . 



LABORATORY TESTS. 



If the widened lines and doublets in spot spectra are produced by a magnetic 

 field, an equal degree of widening and an equal ser)aration of the components 

 of doublets should be found in the laboratory when the same lines are observed 

 in a field of equal strength. As the necessary apparatus was fortunately avail- 

 able, the work was at once undertaken in our Pasadena laboratory by Doctor 

 King. A brilliant spark is produced by a high potential transformer between 

 electrodes supported in the field of a large Du Bois magnet. The light, passing 

 through the pierced pole pieces, falls on a lens, which forms an image of the 

 spark on the slit of a vertical spectrograph, after reflection on a mirror mounted 

 at an angle of 45° above the slit. This spectrograph, which is precisely similar 

 to the 30-foot spectrograph used with the tower telescope, also stands in a con- 

 stant temperature well, with the slit about 3 feet above the floor of the 

 laboratory.'* It may be used as an instrument of 30 feet focal length, or, as in 

 the present case, a 5-inch (13 centimeters) objective of 13 feet (4 meters) 

 focal length, with a 5-inch plane grating, having 14,438 lines to the inch (5G7 

 to the millimeter), can be swung into the axis of collimation 13 feet below the 

 slit. With this shorter focal length the dispersion in the second or third order 

 of the grating is amply sufficient for the present purpose. 



If all of the doublets observed in spot spectra could be photographed in the 

 laboratory it would be easy to make a satisfactory comparison. Unfortunately, 

 however, most of these lines are very faint in the spark, and as the great 

 majority of them occur in the less refrangible part of the spectrum, exposures 

 of from flfteen to twenty hours are sometimes recpiired to bring out even the 

 stronger doubtlets. The results hitherto obtained for the iron doublets are 



« Hale, " The Pasadena Laboratory of the Mount Wilson Solar Observatory," 

 Contributions from the Mount Wilson Solar Observatory, No. 27; Astrophys- 

 ical .Tournnl, Vol. XXVIII, p. 244, 1908. 



