seen in London, September 25, 1827. fff^ 



exclusive of appearances of the Aurora in respect of which 

 the month is not particularised, eight of the different months 

 of the year occur by name ; that is to say, the months of 

 September, October, November, and December, January, 

 February, March, and April ; leaving only four months (May, 

 June, July, and August, the identical summer-months of the 

 Polar regions, or months during which the sun visits the 

 Polar horizon !) hitherto undistinguished by the phenomenon 

 of the Aurora, and almost establishing, as the season of its 

 occurrence, not the middle point of the winter solstice, but 

 the whole period extending, in general terms, from the au- 

 tumnal equinox to the vernal, beginning at or before the 

 first, and ending at or after the last ; or, what may be called 

 the entire winter of the northern hemisphere, or the period 

 during which the sun's course is to the southward of the tropic 

 of Cancer ; a deduction from the scanty data offered by such 

 archives of the phenomenon as we possess, not, perhaps, of 

 trifling importance toward the establishment of the true theory 

 of the cause, as well as of the purpose of its being. 



2. The third sentence, where it describes the Aurora Bore- 

 alis as the constant attendant of clear evenings in the Shetland* 

 Islands, and thereby a great relief to the gloom of the long 

 winter-nights, is probably tainted with errors in regard to the 

 phenomenon, such as affect its whole history and philosophy. 

 The suggestion has just been hazarded above, that at least 

 considerable displays of the Aurora are probably almost as 

 rare, even in the Arctic regions, as in climates further south ; 

 and the truth of this persuasion, as the writer anticipates, 

 will fully appear below. In the sentence now referred to, 

 the word " constant" should, at least, give way to " fre- 

 quent," if not to " often ;" and a distinction should be allowed 

 for, between those feeble appearances which alone, it may be 

 suspected, are even often beheld in the Shetland Islands, and 

 those extraordinary displays which make themselves visible to 

 their southward. 



•' 3. The fourth of the above sentences, in which the Aurora is 

 said to appear commonly at twilight, will have been seen to 

 agree with the time assigned for the commencement of the 

 Aurora in the late example ; and this, when coupled with the 



