BAUER : ANALYSIS OF THE SUN's MAGNETIC FIELD 515 



In Hale's published report, the Zeeman displacement which 

 had been detected with sufficient certainty, was tabulated for 

 various heliographic latitudes for the three lines: X5812.139, 

 X5828.097 and X5929.898, all of which probably originate at com- 

 paratively low levels on the sun. The present analysis is based 

 entirely on these published data and, since these are designated 

 by Hale as preliminary ones, the results derived accordingly from 

 this analysis must be regarded as wholly preliminary and as 

 likely to be superceded by later ones which may be based upon 

 more extensive observations. The prime purpose here has been 

 to ascertain if precisely the same method of analysis be employed 

 for the solar magnetic data, as for the terrestrial magnetic data, 

 some indications, at least, are obtained to show whether or not 

 the magnetic fields of the two bodies follow somewhat analogous 

 laws. In order to obtain strictly comparable data, it was neces- 

 sary to make at the same time a fresh analysis of the earth's 

 magnetic field on the basis of the same magnetic component in- 

 volved in the solar data, and applying to the region 60° north 

 to 60° south. 



The chief results deduced from this first analysis of the sun's 

 general magnetic field are : 



1. The magnetic axis, as determined from the published data between 

 the parallels 60° north to 60° south and for the four series of observations 

 between January 1912 to February 1913, is found to be inclined, for the 

 sun, 9° to 12° to the axis of rotation. The same angle of inclination 

 for the earth was 11?6, in 1885. 



2. The sun's magnetic field is asymmetrical about the equator in 

 much the same manner and in the same direction as is that of the 

 earth. It is quite possible that the sun's actual magnetic poles, or 

 equivalent points, as in the case of the earth, will be found not to be 

 diametrically opposite each other. 



3. The analysis determined four instants, distributed over a year, 

 when the north end of the sun's magnetic axis was on the central me- 

 ridian. Since a whole number of rotations of the sun must have oc- 

 curred during the intervals between the four instants, it was possible 

 to determine a period of rotation which applies, perhaps, to the sun as 

 a whole, instead of to the surface, as is the case with the methods hith- 

 erto used. The synodic period of nearly 33 days which represented 

 the present magnetic data best, differs considerably from the period 

 usually regarded as the solar rotation period. It will therefore be highly 



