

PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. III. 517 



THE TOTAL STORMINESS AS A FUNCTION OF TIME AND ITS RELATION TO SOLAR 



ACTIVITY. 



95. The main object of the investigation with regard to the total storminess was to find any possible 

 regularity in the occurrence of magnetic storms, especially as regards a monthly period. The existence 

 of such a period has been recognised by many authorities, but various opinions are held with regard 

 to its length. 



Mr. E. W. MAUNDER ( l ), from records made at Greenwich, deduced a period of 27.275 days. Mr. ARTHUR 

 I IARVKY ( 2 ), from a study of storms at Toronto, found independently about the same period namely, 

 27.246 days. Dr. AD. SCHMIDT ( 8 ), however, from observations at Potsdam, deduced a period of 29.97 days. 



A period of about the same length - - about 26 days( 4 ) --is found for most magnetic elements, 

 lor atmospheric electricity and northern lights, and for a great variety of phenomena connected with 

 meteorology. All these periods may in some way be related; but it is not my intention to overload 

 the problem by treating such possible connections. We shall in the following pages confine our atten- 

 tion to the treatment of the period for the polar storms, as this period has actually manifested itself 

 during the period of our observations. 



The variation of total storminess at the Norwegian stations is given in tables 88 and 89 and graphically 

 represented by the curves fig. 189. As might be expected, the curves for the four stations show almost 

 exactly the same variation, there being merely a difference as regards absolute magnitude. If, for one 

 station, the absolute storminess for each component were represented by curves, we should see what 

 seems almost a matter of course -- that the storminess varies in very nearly the same way for all 

 three components. 



We notice that a period of about one month is extremely well marked at all four stations. 

 The maxima seem to fall into two groups, the first of these having the first maximum at the end of 

 September, and the last at the end of January, while the second group has its first maximum at the 

 beginning of February and the last observed maximum at the beginning of May. 



If we do not divide the maxima into groups, but consider the two occurring at the end of January 

 and the beginning of February as belonging to the same maximum, we deduce a period of 30.7 days as 

 the average of 7 periods, the first period beginning with the maximum on September 30 and the last 

 one ending with the maximum on May 3. 



Considering each group separately we get : 



From the first group 30.0 days, mean of 4 periods 

 second 28.3 3 



Mean 29.2 days. 



I 1 ) E. W. MAUNDER: The .Great" Magnetic Storms, 18751903, and their Association with Sun-spots, as recorded at the 



Royal Observatory, Greenwich. Monthly Not. 64. 1904. 



,,Magn. Disturb. 1882 1903, etc." Monthly Not. 65, 1904. 



.The Solar Origin of Disturb, of Terr. Magn." Astron. Nachr. 167. 1904. 



,,Magn. Disturb, as recorded, etc." Monthly Not. 65, 538-559 and 666-681, 1905. 



.The Solar Origin of the Disturb, of Terr. Magn." Astroph. Journ. ar. 1905. 



Journ. Brit. Astr. Assoc. 16. 1905. 

 I-) Nature. Vol. 83, p. 354. 1910. 



( 3 ) AD. SCHMIDT: Ergebnisse magnetischer Beobachtungen in Potsdam im Jahre 1907, p. 29. Published 1910. 

 ( 4 l ARRHF.NIUS: Lehrbuch der kosmischen Physik p. 146. 



Hirkflaiul. The Norwegian Aurora Polaris Expedition, 19021903. 



