V 

 Wo. III.] Dr. RICHARDSON ON THE AURORA. ^609 



At 11 Jh. p. m., the sky, which had previously been clear, was covered by 



thin stratum of clouds, belonging to that modification of cirrus which forms the 

 mackarel sky of sailors, conjoined with small portions of what are termed by 

 the same people mares '-tails. Between the bars of the former, and *the long 

 fringes of the latter, streaks of deep-blue sky appeared. 



These clouds were not dense enough to hide the larger stars completely ; and 

 from their first appearance, until they spread entirely over the sky, not more 

 than a quarter of an hour elapsed. 



On attentively regarding the sky for some time, the more rounded parts of 

 the mackerel sky were observed to send shoots across the blue spaces, to unite 

 with similar processes of the neighbouring masses. At the moment of junction, 

 a yellowish light, with a slight tinge of red, was emitted, most brightly from 

 the centres of the two clouds, but extending, though more faintly, to their 

 margins. A longer space of time had not elapsed, than was required to note 

 down these appearances, when an arch of light was observed to cross the zenith, 

 its extremities bearing east and west, and terminating about 50° from the 

 horizon. It was from 3° to 4° broad, and had a pale gold-yellow colour. When 

 it ceased to emit light, its site was seen to be occupied by a range of small 

 fleecy clouds, similar to those already described, but more closely aggregated. 

 The moon now bore nearly south, and shone brightly, strongly illuminating the 

 arch-formed range of clouds just mentioned ; but their rarity was such, that 

 they showed no dark sides. Winds very variable from S. W. to W. 



About a quarter of an hour after the last observation, a round mass of 

 cloud in the S.E. was observed to assume, suddenly, an appearance of greater 

 density, at the same time emitting from its centre a yellowish light. Imme- 

 diately after which, it shot forth towards the S.E. several bright parallel hori- 

 zontal streaks of light, which, crossing the near face of a neighbouring mass of 

 clouds, became slightly curved from the south. They were about 8° or 10 

 above the horizon, and were prolonged after passing before the clouds, through 

 a portion of clear sky. A few degrees beneath them, there were two or three 

 dark layers of cirro-stratus. 



The clouds, in their general arrangement at this period, had that appearance 

 of convergency in opposite points of the horizon, which has been frequently 

 noticed in a sky covered with cirri. In the present instance, these points were 

 at right angles to the magnetic north and south. In the zenith, the mackerel 



4 I 



