OF THE AURORA BOREALIS. 25 



fact has since been observed repeatedly, varying as to the time of night from 1 to 

 3 o'clock. 



III. End. — Ordinary auroras commonly end before midnight, bid those of the first 

 class frequently continue all night. In the aurora of July 1, 1837, streamers were 

 observed an hour after daybreak. Auroral waves are sometimes the most conspicu- 

 ous feature of grand exhibitions after midnight; as was particularly the case in the 

 great displays of November 17th, 1835, 1 April 22d, 1836, and November 18th, 1841, 

 in which these impressive phenomena made their appearance after the corona had 

 vanished, but lasted the greater part of the night. 



The foregoing propositions relating to the time of the beginning, maximum, and 

 end of auroras of the first class, have been derived from the comparison of a large 

 number of instances, but are especially exemplified in the six greatest of the series, 

 namely, those of November 17th, 1835; April 22d, 1836; January 25th and 

 November 14th, 1837; September 3d, 1839; and November 18th, 1841. Those 

 exhibitions which extend to an unusually low latitude, sometimes arrive at their 

 maximum at an earlier hour than common. Thus, the aurora of September 3d, 

 1839, is stated in some reports, to have attained its greatest magnificence between 

 9 and 10 o'clock, and that of September 29th, 1851, formed its first corona as early 

 as 7 o'clock; but, with these exceptions, there is much uniformity in regard to the 

 times of beginning, middle, and end, in most cases the exhibition commencing 

 near the end of twilight, coming to its maximum splendor about 11 o'clock, and 

 continuing, with an increased display of auroral waves, the greater part of the night. 

 Also, if we examine the accounts of great auroras of other times, and of foreign 

 countries, there appears to be much uniformity in these results. Thus Mairan, 2 in 

 his account of the aurora of 1726, that first drew his attention to the subject which 

 he afterwards investigated with consummate ability, states that its maximum 

 occurred a little before 11 o'clock. Indeed, in the polar regions, when the aurora is 

 observed during the period of the year when the sun is constantly below the 

 horizon, the exhibition is not indifferently at all hours of the 24, but, according to 

 the French Commission at Bossekop, consisting of Lottin and others, the usual display 

 is from 10 P. M. to 4 A. M.; and during the short days, the auroras begin there as 

 elsewhere at a certain definite interval after sunset, the hour varying with the time 

 of year. 3 Something of the nature of auroral clouds is occasionally seen in the 

 daytime, but that great exhibitions do not occur by day, may be inferred from the 

 fact that the extraordinary variations of the magnetic needle which attend such 

 displays, are seldom if ever seen in the daytime. 



IV. Extent. — Auroral exhibitions of the higher order are commonly of great extent. 

 The arch of August 28th, 1827, was seen all over the Northern States of the Union, 

 and also in Scotland. The display of November 17th, 1848, was witnessed with 

 very similar appearances, in Asia, Europe, and America, accounts of it having 

 reached us from Odessa on the east to San Francisco on the west, through 150° of 



1 Amer. Journal of Sci., XXIX, 388. 



5 Traite Phys. et Hist, de lAurore Boreale. 



3 Voyages en Scandinavie, etc. Tar MM. Lottin, &c. 



