1904.] Yearly Variation of Magnetic Storms and Aurorce. 93 



may be. Both the magnetic and auroral curves represent four of the 

 set of curves which Mr. Ellis* has recently published. 



It need scarcely be pointed out that the low minima of the auroral 



JULY AUQ SEP OC7 NOV DEC JAN FEB MAR APR MAY JUNE 



DIRECTION 



EARTH'S POLES ^5- 



RELATIVETO THE 



SUN I<H 



s-\ 



DiRECTSON 



RELATIVETOTHE 



EARTH 



o- 1 - 



, 



MAGNETIC 10 n 

 DISTURBANCES 



(FREQUENCY) 

 GREENWICH 



(5O YEARS) 



MAGNETIC 

 DISTURBANCES 



(FRE.aU ENO) 



PARIS 



(5 YEARS) 



AURORAE , . 



EDINBURGH 

 (si YEARS) 



AURORAE 



SCANDINAVIA 

 (SOUTH OF LAT.65} 

 (116 YEARS) Q 



O- 1 - 



10- 

 5- 



N 



\ 



V.x 



X 



\ 





X 



\ 



V 



FIG. 2. Curves showing the relationship between the positions of the Sun's north 

 and south poles with regard to the earth and the frequency of magnetic 

 disturbances and aurorse throughout a year. 



curves during the summer months are due in great part to the short- 

 ness of the nights, and therefore to the restriction of the time 

 available for aurora observations. 



The coincidence in time between the epochs of the maxima of the 



* ' Monthly Notices, R.A.S.,' vol. 64, p. 229. 



