250 



CARNEGIE INSTITUTION OF WASHINGTON. 



sun-spot vortex would be found to have opposite directions of rotation 

 in the northern and southern hemispheres was therefore not fulfilled. 



The difficulty has now been cleared up by the necessary recognition 

 of the characteristics of bipolar spots, first detected on Mount Wilson 

 in 1909. It has long been known that sun-spots frequently occur in 

 pairs, often accompanied by numerous minor companions. The dis- 

 covery that the two principal members of such groups are of opposite 

 polaritj' suggested the existence of a physical connection between them. 

 It is important to note that even in the absence of one member of the 

 group, calcium and hydrogen flocculi may occupy its place. The char- 

 acteristic structure of the Ha flocculi may even suggest the existence 

 of a magnetic field, such as the second spot would produce if present. 



It will suffice here, without entering into a discussion of such groups, 

 to note that whenever the necessary criteria exist every spot should be 

 classified as the preceding or following member of a bipolar group. 

 Adopting this plan, and separating the spots whose polarities were 

 observed before and after the recent sun-spot minimum into two groups, 

 we have the following results : 



Thus in the old cycle, 5 preceding spots in the northern hemisphere 

 were characterized by the violet component of the Zeeman triplet, while 

 one following spot showed the red component. In the southern hemi- 

 sphere the polarities, with two exceptions, were reversed, 15 preceding 

 spots showing the red component, while the violet component appeared 

 in the case of 7 following spots. 



After the minimum, we find for the new cycle the same rule of 

 opposite polarity for the preceding (or following) spots in the northern 

 and southern hemispheres, but, quite unexpectedly, the polarities in a 

 given hemisphere are now reversed. It is too early to explain this 

 remarkable change, but probably the polarity is determined by the lati- 

 tude. In accordance with the well-known law, the late spots of the old 

 cycle occur in low latitudes, while the advent of the new cycle is indi- 

 cated by the appearance of high-latitude spots. As the two groups 

 here in question are thus distinguished by a marked difference in lati- 

 tude, the simplest explanation of the anomaly lies in assuming the 



