AURORA BOREALIS, 1870. 447 



an irregular band from NNE. to W., most prominent at 

 about 20 to 30 above the horizon. This increased in height 

 and breadth until it nearly reached a point SW. of the zenith, 

 and about 15 WNW. of the star Vega. At this time the 

 northern part of the sky was perfectly free from aurora ; 

 gradually that part and the whole dome of the heavens, with 

 the exception of a section from W. to nearly S., became filled 

 with luminous streamers. These, for about 20 on each side 

 of N., were white, the others crimson striped with white or 

 rather greenish light ; but the green I believe to be an effect of 

 contrast, as where similar streamers were distant from the red 

 light they were white. 



The white or green streamers appeared to eclipse the red 

 light ; they changed their size, shape, and position, while the 

 red continued comparatively unchanged. There were also dark 

 streamers, which might have been mere spaces without light, 

 and be caused by the darkness beyond, but they certainly ap- 

 peared to form a part of the phenomenon itself. These 

 streamers or long brushes could be seen beyond and clear of 

 the luminous portion of the aurora, leaving the normal light 

 of the sky between them and it, and hanging like long horse- 

 tails, or like the fringes of rain seen on the edges of a distant 

 rain-cloud, changing their shape and position just as the 

 luminous streamers are seen to do. 



The most remarkable part of the phenomenon, however, 

 was the circle of sky, or what may be called the pole of the 

 aurora, to which the streamers converged. It appeared to em- 

 brace about from 7 to 10 of space. To an ordinary observer 

 it might have seemed occasionally to shift its position to some 

 extent, but, as far as I could judge during an hour's observa- 

 tions, this was not really the case ; flickerings at times covered 

 portions of it, and at other times the whole became faintly 

 luminous ; but by marking its position with reference to some 

 small stars the position of the circle seemed to me to be un- 

 altered. Most singular were the terminations of the streamers 

 as they culminated at this circle, not being undefined or gra- 

 dually evanescent, but having angular tips far brighter than 

 the portions immediately beneath, the nearest illustration to 

 which I can give is an inverted fish-tail or bat's-wing gas- 



