AURORA BOREALIS. 447 



founded. We sometimes saw this cloud before any light was 

 visible, and observed it afterwards become illuminated at its 

 upper surface, and exhibit all the appearances above men- 

 tioned. It was the general opinion that the lustre of all the 

 stars was diminished by the Aurora, but particularly by this 

 part of it. Captain Parry, however, observes in his Journal, 

 p. 142, that the stars in this dark cloud were unobscured, ex- 

 cept by the light of the Aurora. He, however, agrees with 

 us in the lower part of the arch being always well defined 

 and in the upper being softened off, and gradually mingled 

 with the azure of the sky. It is worthy of notice, that we 

 never observed any rays shoot downwards from this arch, and 

 I believe the remark will apply equally to the observations of 

 Captains Parry and Franklin. We frequently observed the 

 Aurora attended by a thin fleecy-cloud like substance, which, 

 if not part of the meteor, furnishes a proof of the displays 

 having taken place widiin the region of our atmosphere, 

 as the light was decidedly seen between it and the earth. 

 This was particularly noticed on the 28th of September, 

 1827. The Aurora on that night began by formino- two 

 arches from W. by N. northward to E. by N., and about 

 eleven o'clock threw out brilliant coruscations. Shortly after 

 the zenith was obscured by a lucid haze, which soon condensed 

 into a canopy of light clouds. We could detect the Aurora 

 above this canopy by several bright arches being refracted, 

 and by brilliant colours being apparent in the interstices. 

 Shortly afterwards the meteor descended, and exhibited a 

 splendid appearance, without any interruption from clouds, 

 and then retired, leaving the fleecy stratum only visible as at 

 first. This occurred several times, and left no doubt in my 

 own mind of the Aurora being at one time above and at 

 another below the canopy formed about our zenith. I must 

 not omit to observe here that, on several occasions, when the 

 light thus intervened between the earth and the cloud, bril- 

 liant meteors were precipitated obliquely toward the south 

 and south-west horizons. 



This supposition of the light being at no great elevation is 

 strengthened by the different appearances exhibited by the 



