94 The Relation of Aurora Borealis 



the appearance of a lunar rainbow. It afterwards became of 

 a reddish tinge, and threw out columns nearly to the zenith. 

 It did not last more than an hour. The following day we 

 had a heavy shower of rain, resembling in intensity that of a 

 thunder storm. 



The aurora again appeared on Oct. 13., and was one of the 

 most beautiful I have ever seen. It first appeared as a white 

 cloud from n.w. to n.e., occasionally throwing up bright pil- 

 lars to the zenith. The cloud finally disappeared ; and was 

 succeeded by waves of light rising from the northern horizon, 

 and flowing after each other in rapid, uniform, and beautiful 

 succession, and vanishing in the zenith. The effect was 

 beautiful in the extreme ; and appeared as if it were occasioned 

 by a light reflected in the heavens from some flickering fire 

 beneath. On Oct. 14-. we had wind and rain. 



I did not observe the aurora again until Nov. 3. 1834. It 

 then showed itself in a most extraordinarily brilliant column, 

 rising from a dark cloud near the horizon, at w.n.w. As it 

 advanced in the heavens, the light became less intense, and 

 took a little turn towards the north ; and finally vanished at 

 e.n.e. I hardly need say that it has been succeeded by very 

 tempestuous weather. 



Whether or not the aurora affects the atmosphere, so as to 

 cause a change in the weather, is a matter, I think, not suffi- 

 ciently proved : indeed, meteorological science is hitherto in 

 its infancy. I am inclined to think that the changes I have 

 above recorded, in connection with aurora, are too slight to 

 allow us to trace them to that as a cause : indeed, I find that, 

 in January, 1831, there was no change whatever in connection 

 with it. The subject is certainly interesting, and deserving 

 of farther observation. With this design, I have hastily 

 thrown these few remarks together : they are of but very 

 slight importance, certainly ; but, if they shall serve to draw 

 the attention of others to the subject, my object will be ob- 

 tained, and the space they occupy will not be entirely wasted. 



Pontnewydd Works, near Newport, Monmouthshire, 

 Nov. 12. 1834. 



Art. IX. A Notice of Instances of the Appearing of the Aurora 

 Borealis, seen from Dundee, Perthshire, in 1833 and 1834; and 

 of the State of the Weather subsequent to the several Instances. 

 By William Gardiner, Jun., Esq. 



These northern illuminations have been very frequent of 

 late, as the subjoined list of dates of their occurrence in 1833 



