Auroral Arches. 177 



March, 1826, near Nottingham, and on the 22nd of March, 

 1841, at York, Durham, etc., when the height was estimated 

 at more than 150 miles from the observations of Professors 

 Chevahier, Phillips, andStevelly (Lowe's Treatise on Atmospheric 

 Phenomena, p. 144, et seq.). A similar arch was observed at 

 Edinburgh on the 27th August, 1846, and an auroral band of 

 the same kind observed by Mr. E. J. Lowe at Nottingham, 

 and by Sir J. Herschel, at Hawkhurst, in Kent, on the 9th 

 March, 1861, was estimated by the latter to have been eighty- 

 three miles high midway between the two places, or over a 

 place thirty miles north of London. Finally, a similar arch 

 was observed by Professor Challis, at Cambridge, on the 21st 

 February, 1862, which was vertical at Newark, in Lincolnshire, 

 but other observations were wanting to determine the height 

 of this arch. 



It appears from these instances that such arches are most 

 common in the months of February and March, and that they 

 vary from 80 to 160 miles in their height above the earth. 



We append a description, furnished by Mr. Herschel, of 

 the appearance of the Aurora Borealis, as seen in Kent in the 

 month of January last. 



"1865, Jan. 28th, 7'35 p.m., Edenbridge, Kent, Two 

 arches of the Northern lights appeared in the N.TT., bright 

 and colourless, the uppermost 7° or 8° above the horizon. 

 Both preserved their brightness for a few minutes only, when 

 the uppermost arch merged into the lower arch, resting 

 together upon a black segment not more than 2° or 3° high. 

 This black segment was cloud, for stars of the third magnitude 

 setting behind it were immediately eclipsed. 



"9*35 p.m., Hawkhurst, Kent. In four or five seconds of 

 time an army of small streamers shot upwards from the border 

 of the dark segment to a height of 10° in the north. These 

 were yellowish, and covered an area of 10° or 12° in azimuth; 

 they immediately began to course each other rapidly across 

 the area from west to east, so as to run over the whole width 

 of the area in less than one or two seconds, when, disappearing 

 in the east, they were replaced by new streamers in the west. 

 This running motion was renewed at intervals (for a few 

 seconds at a time), until, in less than two minutes, the 

 streamers subsided as quickly as they came, and disappeared. 

 The arch and segment afterwards became less distinct, and no 

 other display occurred until 11*30 p.m., excepting a wide 

 diffuse streamer of brief duration at 10'50 p.m. 



