of the Aurora Borealis in North America. 297 



there are other forms which may be used for parallax no 

 less than a common cloud, as the upper line of a well-de- 

 fined arch, or some single fragment of an auroral cloud, 

 unconnected with other parts of the exhibition, and of a form 

 so peculiar as to arrest the attention of observers at some 

 distance from each other. From an extensive comparison of 

 data, we arrive at the conclusion that the height of the au- 

 rora is very variable, sometimes reaching an elevation of at 

 least 160 miles, and at other times rising but little above 

 the regions of the clouds ; still we attempt to prove that the 

 testimony of Farquharson, that the aurora is sometimes seen 

 below the clouds, is not conclusive ; nor that of Parry, Rich- 

 ardson, and others, who have asserted that, in the Polar seas, 

 auroras sometimes appear between the spectator and ice- 

 bergs, at a little distance. 



It is required of a theory that it be a legitimate deduction 

 of well-established truths; and it is required of a hypo- 

 thesis that it explain the leading facts, and that it be not in- 

 consistent with any known facts, although its application in 

 certain cases may not be readily perceived. An explana- 

 tion which unites the characters of both, which is at once an 

 inference from acknowledged truths, and which affords an 

 adequate solution of the leading phenomena — such an ex- 

 planation is deemed peculiarly worthy of confidence until a 

 better can be proposed. In conformity with these principles, 

 we endeavour, first, to infer the cosmical origin of the aurora 

 borealis from known facts ; secondly, we investigate, as far 

 as we can, the nature of the body, or bodies from which the 

 material of the aurora is derived ; and, thirdly, we apply the 

 theory thus formed to the general facts or laws of the pheno- 

 mena. In the first place, we argue the cosmical origin of the 

 aurora from the extent of the exhibitions, which is greater 

 than could arise from any terrestrial emanations, or atmo- 

 spheric precipitations ; from the velocity of the motions, which 

 is too great for any terrestrial forces ; from the occurrence 

 of the different stages of an aurora (the beginning, maximum, 

 and end) at the same hour of the nighty in places difi^ering 

 widely in longitude, — a fact which indicates that successive 

 portions of the earth's surface, in the diurnal rotation, come 



