Obsetvaiions of the Aurora Borealis. 459 



It is, however, probably Jess than the truth, as far as it ex- 

 presses the actual frequency of the pha^nomenon, as 1 have 

 considered observation to have been possible whenever nothing 

 to the contrary is stated, which is most likely more than the 

 facts would warrant; moreover, when we consider the short 

 duration of some of the displays, and how close to the horizon 

 others of them occur, it is difficult to believe that we have 

 noted every one, even on nights when the sky was clear; it is 

 probably set down as clear in many instances when it was 

 sufficiently clouded near the northern horizon to prevent a 

 feeble display from appearing. The dates included in the list 

 at which it was seen at all the stations, which extend along a 

 line of 500 miles, are Jan. 11,16; Feb. 2 1 , 23, 24 ; March 1 6, 

 24; April 1, 2, 5, 6, 7, 29; Aug. 21; Nov. 16. On several 

 other occasions it was seen at every station at which the state 

 of the sky permitted it; but there are one or two instances of 

 clear sky at stations not recording aurora which was seen 

 elsewhere. 



Aurora, or auroral light, was observed at Toronto in 1849 

 on 63 nights, exclusive of 5 entries of an uncertain character, 

 the observations terminating at midnight throughout the year. 

 The other stations, including Newfoundland and Halifax, add 

 70 more, exclusive of 2 doubtful ones; making a total of 133 

 certain, and 6 uncertain appearances, in Canada, Nova Scotia, 

 and Newfoundland*. The area included this year, measured 

 from London, C.W., to Newfoundland, extends about 1150 

 miles from east to west ; and measured from Quebec to Ha- 

 lifax about 140 miles from north to south. Owing to this 

 great extent, there are but few nights (24) clouded at all the 

 stations ; and omitting these, the proportion is 39 per cent., 

 or exactly the same as before. We have at — 



Newfoundland, in 13 monthsjl 



or from Nov. 26, 1 848, to \ 59 obs. to 178 practicable nights, or 33 p. c. 



Dec. 31, 1849 J 



Halifax, in 10 months, or from "I qn i q« ao 



Jan. 14 to Oct. 31, 1849.../'*" "• ^'^^ '" ^^ "' 



Quebec, in 12 months of 1849 44 ... 182 ... 24... 



Montreal 26 ... Descriptions imperfect. 



Kingston 34 ... 178 ... 19... 



Toronto 63 ... 199 ... 31 ... 



London 26 ... 172 ... 15... 



In this list there are but two auroras seen at all the stations 

 without exception ; they occurred on Feb. 27 and July 23. 

 There are eleven dates on which it was seen at Newfoundland 

 and London or Toronto, but missed at some of the interme- 



* One of the uncertain appearances at Toronto is confirmed by other 

 observations. 



2H 2 



