258 CARNEGIE INSTITUTION OF WASHINGTON. 



edge of the penumbra. These are apparently due to the difference in 

 the inchnation of the hnes of force to the hne of sight on opposite sides 

 of the spot. 



The dispersion and resolution of the 75-foot spectrograph are 

 sufficiently high to show as many as five well-separated components in 

 some lines, while in other cases the compound quarter-wave plate 

 leaves no doubt as to the presence of components not completely 

 resolved. In this way the detailed comparison with Mr. Babcock's 

 admirable laboratory photographs of the Zeeman effect, which is now 

 in progress, can be made very satisfactorilj'^, in spite of the compara- 

 tively weak fields in spots. 



A remarkable characteristic of certain photographs is a distinct shift 

 of the central line of some triplets on alternate strips of the quarter- 

 wave plate. This would naturally be taken to indicate that the 

 edges of this central component are cu'cularly polarized in opposite 

 directions, but I am not aware that any such effect has been observed 

 in the laboratory. The phenomenon will be fully investigated. 



SUN-SPOT POLARITIES. 



The polarity of a sun-spot, which may be assumed to give the direc- 

 tion of rotation of the spot vortex, is indicated by the transmission 

 of the red or violet component of a Zeeman triplet by a given strip of 

 the quarter-wave plate. With the large solar image it is possible to 

 apply this test, both visually and photographically, to the smallest 

 spots, and thus to check the tentative conclusions announced in the last 

 aimual report. These have been abundantly confirmed, both as regards 

 the general characteristics of bipolar spot groups and the direction of 

 their rotation in the northern and southern hemispheres of the sun. 



With rare exceptions, possibly confined to the region of the solar 

 equator, the spots lying near the opposite ends of a physically united 

 group are of opposite magnetic polarity. Minor companions of either 

 sign may be present, but opposite polarities strongly predominate at 

 opposite ends of the group. A quantitative study is being made of the 

 weaker fields present in the entire region surrounding spot-groups. 



This result, which is now based on the observation of over 100 spots, 

 has rendered possible a much more complete study of the rule of polari- 

 ties, which is found to be as stated last year. In bipolar groups of low 

 latitude the preceding spot vortices rotate counter-clockwise in the 

 northern hemisphere and clockwise in the southern hemisphere, as in 

 the case of terrestrial cyclones and tornadoes. In high latitudes, on 

 the contrary, the preceding spot vortices rotate clockwise in the northern 

 hemisphere and counter-clockwise in the southern hemisphere. In all cases 

 the direction of rotation for the following members is opposite in sign.^ 



^These conclusions, which refer to the low-lying spot vortices, are based on the assumption 

 that the sign of the revolving charged particles is always the same. This assumption, however, 

 may prove to be incorrect, as results obtained since this report was written indicate that the 

 vortices shown on spectroheliograms taken with Ha are rotating in the same direction as before 

 the minimum. 



