observed at Leith in 1825. 87 



may be allowed to describe in detail. About 8 o'clock p. m., 

 after a pleasant day, there was seen in the north the diffuse 

 light which generally precedes the Aurora Borealis, and which 

 has been compared, with perfect justice, to the twilight. The 

 sky was clear ; the stars were sparkling vividly ; the air was 

 calm and serene. The light in the north continued to increase 

 in intensity, till about half-past 9 ; when, suddenly, extreme- 

 ly brilliant coruscations began to play along the horizon, and 

 dart towards the zenith in great numbers. The colour of 

 these was generally white, or yellowish white, but blue and 

 green were at times discernible. " Immediately below the 

 constellation of Cassiopeia, the illumination was most vivid, 

 and resembled a cylinder of light, white and glowing below, 

 and gradually becoming bluish as it ascended towards the 

 pole." The horizontal extent of the illumination did not ex- 

 ceed 90°, and none of the beams rose higher than 60° or 65°. 

 This very fine display continued about 15 or 16 minutes; 

 after which the intensity of the illumination diminished gra- 

 dually, and the beams became less numerous and less vivid. 

 But before 10 o'clock another phenomenon of equal interest 

 appeared. This was a luminous arch, which passed through 

 the zenith of this place, and descended towards either hori- 

 zon, in the direction of about N.E. and S.W. It was of a 

 white colour, vivid, and well defined. Its breadth in the 

 zenith was about 7°, whence it tapered almost to a point to 

 about 5° or 6* from either horizon, beyond which it was not 

 visible ; its lustre was more intense at the extremities than in 

 the zenith, and throughout its whole extent it was perfectly 

 continuous. Stars of the first and second magnitude oidy 

 could be seen through it. This state of things continued till 

 about 20 minutes from 11, when the centre of the arch seem- 

 ed suddenly to grow very vivid, and a narrow stripe of light, 

 about 30° in length, was seen to extend across the arch, not 

 passing beyond its edges, which were still well defined. This 

 stripe of light soon began to have a distinct motion, and re- 

 taining the same general direction and position with regard to 

 the arch, it traversed, with a moderately rapid motion, its 

 whole western limb, and disappeared below the horizon ; soon 

 after, the arch broke into fragments and disappeared. Dur • 



