No. III.] LIEUTENANT HOOD ON THE AURORA. 581 



they separate into parts. Those arches, which were bright at the horizon, in- 

 crease their brilliancy in the zenith, and discover the beams of which they are 

 composed when the interior motion is rapid. This interior motion is a sud- 

 den glow, not proceeding from any visible concentration of matter, but burst- 

 ing out in several parts of the arch, as if an ignition of combustible matter had 

 taken place, and spreading itself rapidly towards each extremity. In this 

 motion the beams are formed, such as are described in the last observations 

 upon the subject. They have two motions, one at right angles to their length 

 or sideways ; and the other, a tremulous and short vibration, in which they do 

 not exactly preserve their parallelism to each other. By the first, they pro- 

 ject themselves into wreaths, serpentine forms, or irregular broken curves. 

 The wreaths, when in the zenith, present the appearance of Coronae Boreales. 







The second motion is always accompanied with colours ; for it must be ob- 

 served, that beams are often formed without any exhibition of colours ; and I 

 have not, in that case, perceived the vibratory motion. The beams, in differ- 

 ent Aurora, and sometimes in the same, are of different magnitudes, arising, 

 probably, from their various distances. These evolutions, often repeated, 

 destroy the shape and coherence of the several arches ; though they, doubtless, 

 retain the arched appearance to the eye of a spectator at the southern horizon. 

 For it would be absurd to suppose, that these changes occur only in the zenith 

 of one particular place. The observations, at different places, in 1820, afford 

 satisfactory proof to the contrary. And the number of arches often increased 

 or diminished, in their advance to the zenith, by a dismemberment of which, 

 from their distance, we could not distinguish the particulars. However, their 

 several parts passed gradually to the southward, where I have seen them re- 

 assume the form of an arch. They are also sometimes distributed into flashes, 

 and other detached portions, which pass to the south eastward. The revolu- 

 tion of an arch, from north to south, occupies a space of time, varying from 

 twenty minutes to two hours. At Cumberland-House, the arches were in 

 many instances almost stationary for several hours ; a proof, that if their mo- 

 tion was not slower, their distance from the earth was greater than at Fort 

 Enterprise. 



The arches which are faint at the horizon, very frequently pass the zenith 

 without any increase of brilliancy or apparent internal motion. I conceive 

 them to be more elevated than any other description of Aurorae. 



