152 Atmospheric Origin of the Aurora, Sj'c. 



as well as others which he has believed to exist, between the au- 

 rora and a certain class of clouds seen in the day time, do not im- 

 ply an usual identity of location. He had stated, that the aurora 

 is usually higher than clouds, even than cirrous clouds, which are 

 often many miles above other clouds, and many miles above the 

 highest mountains. It by no means follows, that its origin is 

 above crystals of the invisible kind. That the latter may be form- 

 ing and descending for many hours, and in some instances a day, 

 before they attain such a number, magnitude and complexity, as 

 to form visible haze, is evident from the phenomena of halos and 

 vertical solar and lunar columns in a clear sky. But these crys- 

 tals, in their nascent state, must have had a still earlier and higher 

 existence. Should it then be thought surprising, that minute 

 crystals, in a region far above halos, should require a day longer 

 for their aggregation and descent ? 



It is not my present purpose to discuss at length the question 

 as to the intimate nature of the aurora ; but I am of opinion that 

 in some region, usually high, a crystallization takes place on the 

 evening of an aurora, and that the latter originates in the atr<ios- 

 phere. In the publication above referred to, I have ventured to 

 speak of such a thing as " atmospheric magnetism," and to re- 

 gard it as the direct cause of the needle's disturbance, and as loca- 

 ted in a kind of auroral vapor ; although it was the prevalent 

 opinion of philosophers, that the aurora, so far as it was magnetic, 

 was connected with changes in tellurian magnetism alone, that 

 is, the magnetism of the solid earth. The variations of the nee- 

 dle were thought to afford evidence of variations in the latter ; 

 and this view was thought to be corroborated by some simulta- 

 neous disturbances of the needle in distant parts of the globe. 

 Numerous facts might be cited, in corroboration of the atmosphe- 

 ric location. Let one at present suffice. During the brilhant and 

 extensive red aurora of Jan. 25, 1837, I observed at Schenectady, 

 N. Y., a variation of the needle of 1^° in eighteen minutes, of 

 2i° in two hours, and 2^° during the night. At New^Haven, the 

 variations were, at one hour, still more rapid, that is, 45' in two 

 minutes ; but the whole extent observed was only 1°. About 

 thirty miles north of New Haven, no change whatever could be 

 detected ; whilst at Annapolis, the needle varied to the astonish- 

 ing amount of 10° during the night.* Are not these facts wholly 



"" See this Journal, Vol. xxxii, p. 180. 



