POLAR LIGHT, OR AURORA. 157 



in Asia in 1829, by the rotation, or movement in 

 azimuth, of the points of convergence. When the 

 phenomenon is very complete, the two apparent points 

 of convergence do not remain fixed, the one in the 

 north-east, the other in the south-west, (in the direction 

 of the line which unites the highest points of the 

 nocturnally bright auroral arches,) but the line turns, so 

 as gradually to bring those points more towards the east 

 and west.( 212 ) A perfectly similar movement in the line 

 which joins the summits of true auroral luminous arches 

 (the ends of feet changing in azimuths from E. W. 

 towards IS". S.) has been observed with great exactness 

 more than once in Finmarken. ( 213 ) In the view here 

 taken, the cirri arranged as "polar bands " correspond in 

 position to the " streamers " or " bundles of rays " which 

 shoot upwards towards the zenith from the auroral arch 

 of which the span is most often E. W., and are not 

 therefore to be confounded with the arch itself; the 

 latter was once seen by Parry in the arctic regions 

 during daylight, it having continued visible after a 

 night of aurora. A similar phenomenon was witnessed 

 on the 9th of September 1827 in England. It was 

 even possible to recognise the streamers ascending from 

 the arch in daylight.(- 14 ) 



It has been often stated that a perpetual evolution of 

 light, or aurora, takes place around the north magnetic 

 pole. Bravais, who observed without intermission for 

 200 nights, in which 152 auroras admitted of exact 

 description, does indeed assure us that nights without 

 any aurora are very exceptional ; but yet he sometimes 

 either saw no trace of auroral light during a very clear 

 night, and with an uninterrupted horizon, or the " mag- 



