228 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1955 



maximum occurred in 1947 and the last minimum in 1954. The next 

 maximum is likely to occur during the operations of the International 

 Geophysical Year, July 1, 1957-December 31, 1958. 



Sunspots occur only within two zones extending northward and 

 southwrird from the equator to a latitude of about 35°. At the com- 

 mencement of a new cycle, spots begin to appear in the outer parts of 

 these zones ; as the cycle advances they appear in progressively lower 

 latitudes and in increasing numbers, completely leaving the higher 

 latitudes. The mean latitude of the spots in each zone progressively 

 diminishes until, toward the end of the cycle, few spots are found more 

 than 10" from the equator. As the spots near the equator gradually 

 die out toward the end of a cycle, other spots begin to appear in high 

 latitudes, being the first spots of the new cycle. Thus there is always 

 some overlapping between successive cycles. These features of the 

 sunspot cycle are well illustrated by the so-called butterfly diagram 

 (fig. 1) furnished by E. W. Maunder, Greenwich, in which is indicated 

 the latitude of each spot appearing through a sequence of years. The 

 figure shows the butterfly diagram from 1874 to 1954, compiled from 

 the Greenwich Observatory data. Below is shown the mean area of 

 the spots for each solar rotation, together with a curve representing 

 the same data smoothed. The lower section of the diagram will be 

 referred to later. 



Strong magnetic fields are associated with sunspots. The intensities 

 of these fields can be determined from the splitting of the lines in the 

 sunspot spectra which they produce, known at the Zeeman effect. 

 During the course of a cycle the polarities of the sunspot fields, which 

 are opposite in the two hemispheres, remain unchanged, with few ex- 

 ceptions ; during the next subsequent cycle the polarities in each hemi- 

 sphere are reversed. Thus it seems that the length of the cycle is 

 really 22 years, divided into the two periods of 11 years. 



Various attempts have been made to represent the sunspot numbers 

 over the past two centuries or so by the combination of a number 

 of periodic terms. By taking sufficient terms it has been possible to 

 provide a satisfactory representation of the observed numbers ; but the 

 predictions made from these representations have invariably rapidly 

 diverged from the subsequent sunspot numbers and it can be con- 

 cluded that the representation has no physical reality. 



It is not possible to say when or where a spot will appear on the 

 sun. Many spots are small and short-lived, lasting for not more than 

 a few days; others grow rapidly and attain a great size, becoming 

 easily visible to the naked eye ; a large spot may last for several weeks, 

 or even for a few months. The five hirgest spots on the sun during the 

 80 years of observation at Greenwich all appeared in the course of the 

 recent cycle; the greatest of all appeared in April 1947 (pi. 1, fig. 1) 



