^56 



watching the motions of the magnetometers in the Obser- 

 vatory. 



At half-past eleven o'clock, the streamers were very 

 splendid, and covered the whole sky, appearing to spring, 

 however, chiefly from the N. E. They were remarkable for 

 the intensity of their light ; the irregularity of their forms (sel- 

 dom affecting the usual rectilinear form ;) and their incessant 

 dancing motion. At first, the flashes of light appeared 

 in broad irregular masses, at considerable intervals over the 

 sky, like scattered clouds illuminated by the moon, except 

 that their appearance was momentary ; or (as they have been 

 described) like the jets of illuminated vapour, shot from 

 the boiler of a locomotive engine. About twelve o'clock, 

 they spread themselves over the face of the sky, and ex- 

 hibited a nearer approach to their usual form. At this 

 period, a distinct point of convergence, a little to the S. E. 

 of the zenith, was occasionally exhibited ; and a marked 

 contortion of the auroral clouds, at this point, showed the 

 tendency to the formation of the corona. There was like- 

 wise a stationary luminous cloud, to the S. E., which ap- 

 peared to be connected with the phenomenon . 



After twelve o'clock, the brilliancy of the phenomenon 

 in the upper part of the sky gradually lessened ; but a very 

 intense auroral light, with streamers, still remained in the 

 N. W. The atmosphere was remarkably clear, and the stars 

 very bright ; the cloud-like patches of the aurora not 

 seeming to present any obstacle to the transmitted light. 

 There was a cold cutting wind, which came in gusts ; and 

 it seemed as if these gusts were simultaneous with the flashes 

 of the aurora. 



Mr. Lloyd then presented a table, exhibiting the results 

 of observation with the two magnetometers, one of which mea- 

 sures the changes of declination, and the other those of the 

 horizontal part of the earth's magnetic force. The observa- 

 tions commenced at lO'^ 25"^, and were continued, at intervals 



