SOLAR VORTICES AND MAGNETISM IN SUN SPOTS ABBOT. 337 



If subsequent work proves tliis to be the case, it will apiiear very iuiprobable 

 (as indicated by theory) that terrestrial magnetic storms are caused by the 

 direct effect of the magnetic fields in sun spots. We have some reason to tliink 

 that their origin may be sought with more hope of success in the eruptions 

 shown on spectroheliograph plates in the regions surrounding spots. * * * 



Mount Wilson Solar Observatory, 



October 7. t90S. 



ADDENDUM. 



The fact that the doublets in the sun-spot spectrum do not change to triplets, 

 even when the spot is as much as G0° from the center of the sun, appeared, 

 when the proof of the above paper was corrected, to be a serious argument 

 against the magnetic field hypothesis. Thanks to the recent work of Doctor 

 King, this difficulty no longer exists, at least in the case of several iron and 

 titanium lines. Photographs of the spark spectrum in a strong magnetic field, 

 taken at right angles to the lines of force, show that the iron lines X\G21.S.14, 

 6301.72, and 6337.05 are doublets, with no trace of a central component. As 

 these lines are also doublets when observed parallel to the lines of force, it is 

 only natural that they should be double in spots, wherever situated on the solar 

 disk. XG173.55, which is a fine triplet in spots, is a triplet when observed at right 

 angles to the lines of force. But the line X6302.71 is the most interesting of all. 

 In Table I this is classed as a spot doublet. In the spot spectrum the line ap- 

 pears as a triplet, but so decidedly asymmetrical that I supposed the intermediate 

 line to be due to some element other than iron, greatly strengthened in the spot. 

 It now turns out, however, that this is an asymmetrical triplet in the spark, 

 when observed at right angles to the lines of force. Moreover, the displace- 

 ment of the intermediate line from the center is toward the red, both in the 

 spot and in the spark. As soon as a suitable photograph can be taken in a 

 higher order of the grating, it will be possible to measure the asymmetry in 

 the spark, as has already been done in the spot spectrum. 



The titanium lines \XG.303.9S and 6312.46, which are double in spots, are also 

 double in the spark, when observed at right angles to the lines of force. 

 XG0G4.S5, already mentioned as a triplet in spots, with a rather faint central 

 component, is a triplet, with strong central component, in the spark under the 

 above conditions. 



The titanium spot doublets XX5903.56 and 5938.04 (Table II) have not yet 

 been observed at right angles to the lines of force. 



These results leave no doubt in my mind that the doublets and triplets in the 

 sun-spot spectrum are actually due to a magnetic field. As I am now design- 

 ing a spectrograph of 75 feet (23 meters) focal length, for use with a tower 

 telescope of 150 feet (46 meters) focal length, I hope it may become possible to 

 investigate small spots, as well as large ones, and to resolve many of the close 

 doublets and triplets in their spectra. 



Let ITS sum up the principal parts of Mr. Hale's experimental 

 evidence: (1) That some spectrum lines which are single in the or- 

 dinary solar spectrum become double in sun spots; (2) that these 

 double lines are found to be circularly polarized in opposite direc- 

 tions; (3) that, as shown by Zeeman, this is a characteristic of 

 spectrum lines produced in a powerful magnetic field and observed 

 along the magnetic lines of force; (4) that the different doublets in 



