506 



NATURE 



[July 17, 19 13 



extent, in Prof. Hale's own words. As a general 

 problem of physics, Schuster's suggestion that 

 every rapidly rotating body may produce a mag- 

 netic field is of fundamental importance. A direct 

 test by laboratory experiments cannot be made 

 because of the limitations of size and rotational 

 velocity, but advantage may be taken of the 

 heavenly bodies where these limitations do not 

 obtain. The most promising object for such an 

 investigation is the sun. It is here that the direct 

 method of determining the magnetic field by 

 observation of the Zeeman effect is most readily 

 employed, since the sun is bright enough to per- 

 mit the use of the very high dispersion required. 

 Further, it is possible to observe at a great number 

 of points on the surface, and since observations 

 may be made in both hemispheres the most perfect 

 test of the Zeeman effect can be applied by look- 

 ing for a reversal of the sign of the displacement 

 with the polarity. The present minimum of solar 

 activity has furnished a particularly favourable 

 opportunity for the investigation, in consequence 



of the general 



absence of local 



pEB|j strong fields due 

 to spots and 

 other disturb- 

 ances. 



Assuming the 

 jf&SH sun's magnetic 

 lgj|5l| field to be similar 

 to that of a mag- 

 netised sphere, 

 w i t h magnetic 

 poles coincident 

 with the poles of 

 rotation, the 

 lines of force 

 would appear as 

 in Fig. 1, the 

 angle 8 between them and the solar surface being 

 given by tan 5 = 2 tan r/>, where c/> is the heliocen- 

 tric latitude. If the field were strong enough, and 

 if the observer could look along the sun's axis and 

 form an image of one of the poles on the slit of a 

 spectroscope, certain solar lines would appear as 

 doublets with components circularly polarised in 

 opposite directions. If a Nicol prism were placed 

 in front of the slit, with its long axis parallel to 

 the slit, in combination with a quarter-wave plate 

 set with its principal section at an angle of 45 , 

 one of the components would be extinguished, 

 while the other would be transmitted by the Nicol. 

 Assuming the red component to be transmitted, a 

 rotation of the quarter-wave plate through 90 

 would cause this to be extinguished and the violet 

 component to be transmitted. If from the same 

 place of observation the slit were directed to a 

 point in 45 lat. , the effect would still be clearly 

 observable, though the transformation of the circu- 

 larly polarised light of the components into ellipti- 

 cally polarised light would result in less complete 

 extinction by the Nicol. 



In the actual case the terrestrial observer is close 

 to the plane of the sun's equator and must look 

 rCO. 2281, VOL. 91] 



in a direction nearly at right angles to the lines 

 of force at the sun's poles. 2 He therefore 

 cannot take full advantage of the fact that 

 the total intensity of the sun's magnetisation 

 is twice as great at the poles as at the equator. 

 The angle between the lines of force and the line 

 of sight, however, is reduced to zero at 35 north 

 and south latitude ; but the most favourable position 

 for observation is 45 lat. , where the effect of the 

 ellipticity of the light is overcome by increased 

 strength of the field. 



On account of the weakness of the sun's mag- 

 netic field, complete separation into doublets is not 

 to be expected, and the investigation must, there- 

 fore, depend upon the possibility of detecting very 

 slight displacements of lines to red or violet, 

 according to the position of the quarter-wave plate, 

 with reversal of the sign of the displacements in 

 passing from the northern to the southern hemi- 

 sphere. 



;. 2. — Region of A5930 photogtaphed in the third orc'er with the 

 75-ft. spectrograph showing the division of the spectrum into 2 mm. 

 strips by the compound quarter-wave plate. The heavy horizontal line 

 marks the junction of two sections of the Nicol. The fifth stripe 

 below is the "marked strip" used for reference purposes. 



The first attempts to detect the solar magnetic 

 field were made in 1908 with the 60-ft. tower tele- 

 scope at Mt. Wilson, but it was not until the new 

 164-ft. tower telescope and 75-ft. spectrograph be- 

 came available, in 1912, that definite results were 

 obtained. With the latter instrument the sun's 

 image is about 16 in. in diameter, and about 

 A.5900 in the third order spectrum, where much of 

 the work was done, the linear dispersion is 

 1 Angstrom = 4*9 mm. ; on this scale the distance 

 between the D lines is 29 mm.' For determination 

 of focus and investigation of resolving power the 

 extremely fine lines in the absorption spectrum of 

 iodine were employed with advantage, and lines as 

 close as C025 A.U. were found to be just resolved. 



The polarising apparatus consists of a Nicol 

 prism 18 mm. wide, built up of four sections, each 

 32*5 mm. long, so as to give a total length of 

 130 mm. The impossibility of rotating it is easily 



2 The trii lets produced by light from the poles would, of course, be too 

 for observation as such, and the use of a Nicol in different positions 

 lot affect the symmetry of the lines. 



