The number of auroras which develop a corona near the zenith is comparatively small in our latitudes; but many of them, although not exhibited on so grand a scale, are nevertheless very interesting. On some very rare occasions the auroral display has been confined almost exclusively to the dark segment, which appeared then as if pierced along its border by many square openings, like windows, through which appeared the bright auroral light.
PLATE IV.—AURORA BOREALIS.
As observed March 1, 1872, at 9h. 25m. P.M.
Among the many auroras which I have had occasion to observe, none are more interesting, excepting the type first described, than those which form an immense arch of light spanning the heavens from East to West. This form of aurora, which is quite rare, I last observed on September 12th, 1881. All the northern sky was covered with light vapors, when a small auroral patch appeared in the East at about 20° above the horizon. This patch of light, gradually increasing westward, soon reached the zenith, and continued its onward progress until it arrived at about 20° above the western horizon, where it stopped. The aurora then appeared as a narrow, wavy band of light, crossed by numerous parallel rays of different intensity and color. These rays seemed to have a rapid motion from West to East along the delicately-fringed streamer, which, on the whole, moved southward, while its extremities remained undisturbed. Aside from the apparent displacement of the fringes, a singular vibrating motion was observed in the auroral band, which was traversed by pulsations and long waves of light. The phenomena lasted for about twenty minutes, after which the arch was broken in many places, and it slowly vanished.
The aurora usually appears in the early part of the evening, and attains its full development between ten and eleven o'clock. Although the auroral light may have apparently ceased, yet the phenomenon is not at an end, as very often a solitary ray is visible from time to time; and even towards morning these rays sometimes become quite numerous. On some occasions the phenomenon even continues through the following day, and is manifested by the radial direction of the cirrus-clouds in the heights of our atmosphere. In 1872 I, myself, observed an aurora which apparently continued for two or three consecutive days and nights. In August, 1859, the northern lights remained visible in the United States for a whole week.
The height attained by these meteors is considerable, and it is now admitted that they are produced in the rarefied air of the upper regions of our atmosphere. From the researches of Professor Elias Loomis on the great auroras observed in August and September, 1859, it was ascertained that the inferior part of the auroral rays had an altitude of 46 miles, while that of their summits was 428 miles. These rays had, therefore, a length of 382 miles. From the observation of thirty auroral displays, it has been found that the mean height attained by the summit of these streamers above the Earth's surface was 450 miles.
But if the auroral streamers are generally manifested at great heights in our atmosphere, it would appear from the observations of persons living in the regions where the auroras are most frequent, as also from those who have been stationed in high northern and southern latitudes, that the phenomenon sometimes descends very low. Both Sabine and Parry saw the auroral rays projected on a distant mountain; Ross saw them almost at sea-level projected on the polar ice; while Wrangel, Franklin, and others observed similar phenomena. Dr. Hjaltalin, who has lived in latitude 64° 46' north, and has made a particular study of the aurora, on one occasion saw the aurora much below the summit of a hill 1,600 feet high, which was not very far off.
The same aurora is sometimes observed on the same night at places very far distant from one another. The great aurora borealis of August 28th, 1859, for instance, was seen over a space occupying 150° in longitude—from California to the Ural Mountains in Russia. It even appears now very probable that the phenomenon is universal on our globe, and that the northern lights observed in our hemisphere are simultaneous with the aurora australis of the southern hemisphere. The aurora of September 2d, 1859, was observed all through North and South America, the Sandwich Islands, Australia, and Africa; the streamers and pulsations of light of the north pole responding to the rays and coruscations of the south pole. Of thirty-four auroras observed at Hobart Town, in Tasmania, twenty-nine corresponded with aurora borealis observed in our hemisphere.