32 ON THE RECENT SECULAR PERIOD 



the height of the aurora in three different instances, namely, the auroral cloud of 

 December, 1835; the auroral arch of August, 1836, and of May of the same year. 1 

 In each case there were at least three scientific observers, all much conversant 

 ■with observations of this kind. The identity of the object seen by all was proved 

 by a number of coincidences, such as the time of formation and of disappearance — 

 the correspondence of position in azimuth — in the peculiar appearances and success- 

 ive changes of aspect, described by one observer as parallel fleeces, and by another as 

 ridges into which the arch broke up. These coincidences apply more particularly 

 to the arch of August, 1S36, observed by Professor Twining and myself at a 

 distance from each other of twenty-two miles. On comparing notes, no doubt has 

 remained on the mind of either of us that we were both gazing at one and the 

 same object, and our minutes were sufficiently copious and definite to determine 

 the parallax. The calculation was made by Professor Twining, with all possible 

 accuracy, and gave an altitude of 144 i miles. The same authority, in his observa- 

 tions upon the auroral cloud of December, 1835, remarks that it was not his object 

 to determine the exact height, but to find the limit below which it could not have 

 been, and this limit he fixes at 42 J miles; but he still believes this to have been, 

 iu fact, an elevation far less than the true one. In regard to the arch of May, 183G, 

 he concludes that the largest parallax which could be assigned, consistently with 

 the observations, made the height more than 100 miles; and he supposes that it 

 might have reached the elevation which I had deduced from my observations 

 compared with those of Dr. Ellsworth, of Hartford, namely, 160 miles, although 

 he believes that this was the utmost limit. 



On the whole, I think we are authorized to infer from all the foregoing authori- 

 ties, that the auroral arches seldom, if ever, fall below an elevation of 70 miles 

 above the earth, and that they do not often exceed a height of 160 miles. The 

 probability is that they vary between these extremes. In fixing these limits, however, 

 it may be thought proper to take notice of certain observations on record, which 

 indicate a much lower elevation, descending even to the region of the clouds, or 

 below it. I allude particularly to the views of Ptev. Mr. Farquharson, of Scotland, 

 and of Captain Parry, and one or two other navigators of the Polar Seas. 



To nearly all who have attentively observed tins phenomenon, a difference of 

 stations of a few miles or even of a few degrees, has made but slight changes in 

 the position of an auroral arch ; to an inhabitant of Montreal and to one of 

 Washington City, the same exhibition has been still north; but, according to Mr. 

 Farquharson, an aurora changed its place to the view of an observer, from north to 

 south, merely by crossing a hill in the opposite direction. Moreover, according to 

 the same authority, accurate trigonometrical measurements at the extremities of a 

 base line only 6,810 feet long, afforded so great a parallax as to give a perpen- 

 dicular height above the lower place of observation of only 5,693 feet, or a little 

 more than a mile ; and of only 1,500 feet above the summit of the neighboring 

 hills. These results, and other similar ones reported by the same authority, are so 



Aracr. Jour, of Science, XXXH, p. 211. 



