450 retort — 1884. 



In addition to the more regular changes in the magnetic needle we 

 observe occasionally greater magnetic disturbances, which have been called 

 magnetic storms by Humboldt. These magnetic storms are both more 

 violent and more frequent at times of maxima than of minima of sunspot 

 frequency. Loomis ' has tried to determine whether particular magnetic 

 storms are coincident in time with increased solar activity. He took over 

 a hundred magnetic disturbances, and determined the spotted area on the 

 sun on the days of the storms as well as on the preceding and succeeding 

 days, and found an increased area on the same, and the four or five 

 preceding days. It would thus seem that a greater number of spots was 

 accompanied or shortly followed by magnetic disturbances. 



Oue or two occurrences, though not directly connected with our subject, 

 deserve mentioning. On September 1, 1859, Mr. Carrington at Reclhill, 

 and Mr. Hodgson at Redhill, both observed two brilliant patches of intense 

 white light in the neighbourhood of a great spot. The phenomenon was 

 an altogether exceptional one, and was accompanied by a magnetic dis- 

 turbance, which was repeated with greater energy the same evening, and 

 brillant aurora? were seen over a large part of the terrestrial globe. 



Professor Young 2 has also on one occasion noticed an exceptionally 

 strong outbreak of protuberances on the sun to be accompanied by a 

 magnetic disturbance. 



o 



Connection between Sunspots and the Aurora Borealis. 



It was suspected in the first half of last century already that the 

 aurora appeared more frequently in years in which the number of spots 

 oii the sun was peculiarly great. After the discovery of the periodicity of 

 sunspots the connection was placed beyond doubt by Wolf, 3 Fritz, 4 and 

 Loomis. 5 We owe to Fritz 6 the most detailed investigation of the subject, 

 and we shall here give a short summary of his results. In the first place, 

 the eleven-yearly period of sunspots can be distinctly traced in the fre- 

 quency of auroras observed in our atmosphere. This holds for the aurora? 

 seen in Europe as well as in America. It is true for those observed in 

 the arctic circle as well as those observed in the temperate zone, and, as 

 far as the scarcity of the material allows us to judge, the same connection 

 exists for the aurora australis. The sunspot curve presents the appear- 

 ance of a phenomenon similar to, if not identical with, that of beats, 

 which at certain intervals raises the maxima values high above their 

 means. This phenomenon appears in an exaggerated form in the curve 

 of frequency of the aurora borealis. It is, indeed, much more marked 

 tban the eleven years period itself, a fact which it is important to bear 

 in mind in any attempt to speculate on the nature of the connection, for 

 it shows that the number of aurora? is not proportional to the number of 

 sunspots, like the increase in amplitude of the daily variation, but that 

 vvhile a small number of sunspots hardly shows any effect, a slight in- 

 crease in them is often accompanied by a very large increase in the 

 number of aurora?. The inequalities in the sunspot curve show thus a ten- 

 dency to be exaggerated in the curve of aurora?, and the exaggeration is 

 the greater, the greater the number of sunspots. Plate I., which is copied 



1 American Journal of Science, vol. 1- 2 The Sun. 



» Wolf's Mitthcihmgen, No. X. (1859>. 4 Ibid. No. XV. (1863). 



'' Ancr. Rep. (18G5). 6 Ueber die Beziehungen, &c. 



