tint, lightly washed with green, was shown. Although the meteor was not visible continuously, there was often observed in the spectroscope, and even quite high above the horizon, the characteristic band of the auroras without the eye perceiving any trace of their light. Since this fact was remarked even when there was no snow, it could not be attributed to reflected auroral flashes. Moreover, the observers not rarely saw during the nights a light yellowish, diffuse, and phosphorescent light that illuminated the horizon and paled the stars. The effect produced was compared to that of the moon half veiled by clouds. M. Lenström and his associates attempted, on the 8th of December, 1882, to measure the height of an auroral arc above the surface of the earth. They divided themselves into two groups, and took with a theodolite the angular distance from the crest of the arc to the horizon. The two stations were four and a half kilometres apart on the same magnetic meridian, and correspondence was had during the observations by a telegraphic wire previously arranged for the purpose. They endeavored to look in concert at the same point of the meteor, but, after reiterated essays, they recognized that any particular ray visible to one party could not be seen by the other. The results of the views were irreconcilable, for the angle obtained was greater for the southern post than for the northern one, although the latter post was, a priori, nearer to the meteor. M. Lenström concluded from this, as M. de La Rive had done, that every observer sees his own aurora the same as every one sees his own rainbow, and that the phenomenon is produced at the height of only a few thousand metres. He also calls attention to the results obtained in Greenland by the engineer Fritze, which lead, in certain cases at least, to numbers twenty times as small. During the Swedish Polar Expedition of 1868, faint flashes or phosphorescent lights were remarked around the summits of the mountains. This fact, with which M. Lenström did not become acquainted till 1871, related as it was to some of the descriptions given by other travelers, decided him to try to provoke or facilitate the appearance of the meteor by artificial means. The first attempts date from 1871, and, like those that followed them, were made in Lapland. The enterprise being successful from the first, the experiments were resumed during the Finnish Polar Expedition of 1882, and were renewed twice on two different peaks, called respectively Oratunturi and Pietarintunturi. Oratunturi, rising more than five hundred metres above the level of the sea, is situated in latitude 67° 21', near the village of Sodankylä. Near the topmost height of the mountain was placed a long copper wire, so bent upon itself as to form a series of squares within squares, having a total surface of nine hundred square metres, supported by insulated posts. Tin points or nibs bristled out from this connecting net at distances of half a metre apart, and the whole was connected by an insulated wire running along on stakes with a galvanometer fixed in a cabin at the foot of the peak. The galvanometer was connected with the earth by the other ex-
