OF THE AURORA BOREALIS. 



23 



72 J. Whence it appears that, during this period, the number of auroras was 

 greater when the earth was in the part of its orbit most distant from the sun than 

 in the nearer part, in the ratio of 724 : 57£, or nearly of 24 to 19. This result is 

 very different from that obtained by Mairan, who found the number for the six 

 winter months greatly to exceed that for the six summer months ; but the cases 

 collected by Mairan were those only of the higher classes, or of the greatest 

 intensity. We shall see that there is no fixed relation between intensity and 

 number. 



4. The monthly average is nearly 4 ; but that for April is about 5i, and that for 

 September is about 5 ; while that for December is but little more than 2, and for 

 June is only 21. 



Remarks. 



1. The average per annum for the whole period, is nearly 3. 



2. Maximum year, 1838; maximum month, September. The year 1837 was 

 distinguished for three examples of the first class; 1838 was distinguished for a 

 greater number, but they were chiefly of the second class. September, 1838, pro- 

 duced no less than 5 of this class. 



3. There is no apparent connection between number and intensity. The year 

 1835 was the least productive year of the series; but the exhibition of November 

 17th, of that year, was one of the highest class. In 1S3G there were 61 auroras, 

 and in 1837 only 49 ; yet in the latter year there were three of peculiar grandeur; 

 while in the former there was only one of the highest order. 



4. The auroras of the first class, during the period under review, occurred at the 

 following dates : — 



1833. May 17th. 1839. September 3d. 



1835. November 17th. 1S47. March 19th, April 7th. 



1836. April 22d. 184S. November 17th. 



1837. January 25th, July 1st, November 14th. 



