According to some authorities, the phenomenon is accompanied by noises resembling the discharge of fireworks, or the crackling of silk when one piece is folded over another; but this statement is discredited by the most trustworthy observers. Mrs. Somerville’s description is worth quoting, as taking up more emphatically some points to which we have already alluded:-- The aurora, she says, is decidedly an electrical (or, more strictly speaking, a magneto-electrical) phenomenon. It generally appears soon after sunset in the form of a luminous arc stretching more or less from east to west, the most elevated point being always in the magnetic meridian of the place of the observer; across the arc the coruscations are rapid, vivid, and of various colours, darting like lightning to the zenith, and at the same time flitting laterally with incessant velocity. The brightness of the rays varies in an instant; they sometimes surpass the splendour of stars of the first magnitude, and often exhibit colours of admirable transparency,--blood-red at the base, emerald-green in the middle, and clear yellow towards their extremity. Sometimes one, and sometimes a quick succession of luminous currents run from one end of the arc or bow to the other, so that the rays rapidly increase in brightness; but it is impossible to say whether the coruscations themselves are actually affected by a horizontal motion of translation, or whether the more vivid light is conveyed from ray to ray. The rays occasionally dart far past the zenith, vanish, suddenly reappear, and, being joined by others from the arc, form a magnificent corona or immense dome of light. The segment of the sky below the arc is quite black, as if formed by dense clouds; yet M. Struve is said to have seen stars in it, and so it would appear that the blackness of which several observers speak must be the effect of contrast. The lower edge of the arc is evenly defined; its upper margin is fringed by the coruscations, their convergence towards the north, and that of the arc itself, being probably an effect of perspective.