432 Prof. E. Edlund on Atmospheric Electricity 



auroras are often simultaneous in the northern and the south- 

 ern hemisphere. 



If we compare the theoretic deductions above formulated 

 with the results of observation, we shall find that there exists 

 a satisfactory accordance between them. 



According to Loomis, in North America, under the meridian 

 of Washington and at the 40th parallel, 10 auroras per year 

 are seen ; under the 42nd parallel the number amounts to 20 ; 

 and near the 45th the number is 40. In the latitude of 50°, 

 the number of aurorae boreales yearly is stated at 80 ; and 

 between this latitude and that of 62° the aurora appears nearly 

 every night. Between the last two latitudes the aurora 

 borealis appeared quite as often to the south as to the north. 

 Here, then, is situated the zone, properly so called, of the 

 auroras boreales, of which about 56° may be considered 

 the mean latitude. To the north of 62° the auroras appear 

 almost exclusively on the south side, and they diminish in 

 number and brightness as we advance northward. At 67°lat. 

 their number has fallen to 20, and is only 10 in the vicinity of 

 78°. The same fact presents itself at the meridian of St. Pe- 

 tersburg ; but here the zone of the auroras is situated much 

 more northward than in America : it is only between the 66th 

 and 75th degrees of north latitude that the annual number of 

 auroras boreales is stated at 80*. 



A multitude of measurements upon the position of the auroral 

 arc were made at Bossekop in the winter of 1838-39, by the 

 French Expedition to Spitzbergen and Norway. The result 

 of more than 200 measurements was, that the crown of the 

 arc was situated 10° west of the magnetic meridian. Arge- 

 lander had arrived at results nearly the same by his observa- 

 tions made at Abo, in Finland. In consequence of accidental 

 circumstances, the crown of the auroral arc appeared several 

 times to the east of the magnetic meridian. The position indi- 

 cated being the mean of all the observations, is cleared of acci- 

 dental perturbations")". The fact that the crown of the auroral 

 arc must in those regions appear, on the average, to the west 

 of the magnetic meridian flows directly from the theory, if we 

 take into consideration the geographical situation and the mag- 

 netic declination of the locality, as well as the form and situa- 

 tion of the annular space of the maximum of electric density. 

 It is easy, in the same way, to understand the accuracy of 

 Argelander's observations. In North America, on the con- 

 trary, as in Siberia, the crown of the auroral arc must more 

 nearly coincide, on the average, with the magnetic meridian ; 



* Loomis, Annual Report of the Smithsonian Institution, 1866. 

 t Sw les Aurores boriales rues a Bossekop et a Jupvig : Paris. 1840. 



