0± ON THE RECENT SECULAR PERIOD 



With the foregoing data before us, compared with such as have been recorded 

 in different countries and in different ages, we now proceed to classify the leading 

 facts appertaining to this mysterious phenomenon, with the view of ascertaining 

 its laws. 



LAWS OF THE AURORA BOREALIS. 



I. Beginning. — An aurora of the first class usually commences near the end of 

 evening twilight, in the form of a northern light resembling the dawn. The greatest 

 exhibitions are always found in their incipient and never in their maturer states, at 

 this time of the evening. No instance is on record of an aurora suddenly bursting 

 upon the view, as the light of day disappeared, with its corona or its crimson columns, 

 or its magnificent arches, or its waves fully formed, and in complete operation. 1 



Moreover, the beginning at different places is nearly at the same instant of local 

 time at points widely different in longitude. Thus the exhibition of November 

 17th, 1835, was first seen at Boston and Cincinnati at 6 o'clock, and in the incipient 

 stage, although these places differ nearly an hour in longitude; and that of November 

 17th, 1848, distinguished above all others for its extent from east to west, began, 

 arrived at its maximum, and, in short, performed all its evolutions at nearly the 

 same hour of the night, in Western Asia, in Europe, and in the United States, even 

 to the shores of California. 



II. Maximum. — An aurora of the first class commonly arrives at its maximum at all 

 places from 10 to 11 o'clock, and more frequently a little before ll. 2 At this period, 

 the columns suddenly increase in size and splendor, the bloody hue becomes most 

 conspicuous, and there is a general rush from the northeast and northwest towards 

 the common point of concourse, forming the corona. 



In some cases, the meteor rallies and forms a second maximum, though of inferior 

 intensity to the first, at a later hour of the night, often about 2 o'clock, a fact first 

 remarked by Mr. Herrick, in the auroral display of July 1, 1837; 3 and the same 



1 That auroras usually begin* at a certain hour of the evening, and that great auroras begin earlier 

 than others, was noticed by Mairan in the following terms: Le commencement du phcnomene arrive 

 commencement deux, trois, ou quatre heures tout au plus apros le concher du soleil, c'est-a-dire qu'il 

 arrive presque toujours le soir, et jamais, que je sache, le matin apres minuit, lorsque les nuits sont un 

 pen longucs. Les grandes aurores boreales commencent ordinairement de bonne heure, peu de temps 

 apri-s la fin du crepuscule, et qnelques fois auparavant. — Traite Phys. et Hist. dul'Au. Bor., p. 115. 



a Capt. Lefroy (Preliminary Report on the Observations on the Aurora Borealis, made in Canada, in 

 Regents' Rep. for 1850) recognizes this law, and remarks that he had observed it as early as 1843-4. 

 In my account of the great aurora of Nov. 18, 1841, published the next day after its occurrence in the 

 New Haven Daily Herald, I remarked as follows: In all these respects, as well as by the grand display 

 of auroral waves, this exhibition of the aurora resembled that of November, 1835. Indeed, both came 

 to their maximum and formed their corona around the pole of the dipping needle at almost precisely 

 the same time of the night, namely, about 1 1 o'clock. Nor does this appear to have been a casual coinci- 

 dence. It was the case, also, with a great exhibition of the aurora seen in France about a century ago, 

 (particularly described by Mairan), and with so many others as to constitute a striking and important 

 feature in relation to the true theory of the phenomenon. 



3 Amer. Journal of Sci., XXXIII, 143. 



