190 COSMOS. 



according to accurate observations made on this subject,* 

 not generally in the magnetic meridian itself, but from 5 to 

 18* towards the direction of the magnetic declination of the 

 place. f In northern latitudes, in the immediate vicinity of the 

 magnetic pole, the smoke-like conical segment appears less 

 dark, and sometimes is not even seen. Where the horizontal 

 force is the weakest the middle of the luminous arch deviates 

 the most from the magnetic meridian. 



The luminous arch remains sometimes for hours together 

 flashing and kindling in ever- varying undulations, before rays 

 and streamers emanate from it, and shoot up to the zenith. 

 The more intense the discharges of the northern light, the 

 more bright is the play of colours, through all the varying 

 gradations from violet and bluish white to green and crim- 

 son. Even in ordinary electricity excited by friction the 

 sparks are only coloured in cases where the explosion is very 

 violent after great tension. The magnetic columns of flame 

 rise either singly from the luminous arch, blended with black 

 rays similar to thick smoke, or simultaneously in many 

 opposite points of the horizon, uniting together to form a 

 flickering sea of flame, whose brilliant beauty admits of no 

 adequate description, as the luminous waves are every moment 

 assuming new and varying forms. The intensity of this light 

 is at times so great, that Lowenorn (on the 29th of June, 1786) 

 recognised the coruscation of the polar light in bright sun- 

 shine. Motion renders the phenomenon more visible. Round 

 the point in the vault of heaven which corresponds to the 

 direction of the inclination of the needle, the beams unite 

 together to form the so-called corona, the crown of the 



* For an account of the results of the observations of Lottin, Bravais, 

 and Siljerstrom, who spent a winter at Bosekop on the coast of Lapland 

 (70 N. lat.), and in 210 nights saw the northern lights 160 times, see 

 the Comptes rendus de I'Acad. des Sciences, t. x. p. 289, and Martins' 

 Meteoroloyie, 1843, p. 453. See also Argelander, in the Vortragen 

 geJi. in der Konigsberg Gessellschaft, bd. i. s. 259. 



f [Professor Challis, of Cambridge, states that in the aurora of Oc- 

 tober 24th, 1847, the streamers all converged towards a single point of 

 the heavens, situated in or very near a vertical circle passing through 

 the magnetic pole. Around this po ; nt a corona was formed, the rays of 

 which diverged in all directions from the centre, leaving a space tree 

 from light : its azimuth was 18 41' from south to east, and its altitude 

 9 54'. See Professor ChnU*. in the Athenceum, Oct. 31, 1847.] Tr. 



