time over the whole sky, their terminations being in the same eastern and western portions of the horizon, and this simultaneously with a brilliant auroral light in the southern quarter. The sky at the same time was free from clouds, and the air calm. It is worthy of note that though on five out of six times on which these bands have been observed the moon was above the horizon, on this occasion it was invisible, as it was new moon on the 23rd. This circumstance removes some of the complexity which might otherwise arise in accounting for this phenomenon, lunar halos of any description being inadmissible.

A third display in the same year took place on the 13th May, the arches extending from east to west, a beautiful lunar halo occurring at the same time, the moon being not far from the northern quarter. The sky was slightly hazy, as only large stars were seen, and the air was calm.

Another display (and the earliest observed at this station) took place on the 10th July, 1867, and in this case the bands, instead of extending from east to west, stretched from north to south, and it was compared at the time to the streamers of an aurora australis meeting the streamers of aurora borealis. An apt comparison, and one which would perhaps illustrate the phenomenon more forcibly than a description, might be made by supposing these bands as they appeared on this occasion to represent a number of imaginary meridians at some uniform distance above the globe. A very fine lunar halo was seen at the same time, the moon being somewhat west of the meridian; the air was calm. Lunar halos were also seen at Sydney on the 8th, 9th, and 10th. A similar phenomenon occurred on the 29th March, 1869, in which the bands extended from north to south, accompanied by a lunar halo, the moon being near full and over the east.

I shall instance another case by making the following extracts from recorded meteorological observations. “On the 1st May, 1871, symmetrical bands of light seen at 9.45 p.m. similar to those of 1870. They had their terminations east and west; were very distinct; stars could be seen through them. They radiated so regularly from the east and west (fanlike) that a luminary might have been supposed to have been under the horizon at these points. On this occasion also the moon was near the meridian and surrounded by a large halo; the air was calm.” “A partial display occurred on the 25th September, 1871, in the eastern quarter, seen in broad daylight, about 5.45 p.m. The radiations were straight, and spread, fanlike, for about 35° upward from the horizon. The edges of these bands were well defined, and not gradually shaded off. They were unmistakably distinct from, and not to be confounded with, light cirrus clouds which were also

observed covering part of the sky in streaks which roughly coincided with these bands.”

It may be interesting to give Captain Cook's remarks on what was evidently a partial display of these luminous streamers adorning the night sky of the south. In the course of his second voyage he remarks that “on the 17th February, 1773, a beautiful phenomenon was observed in the heavens. It consisted of long colours of a clear white light shooting up from the horizon to the eastward almost to the zenith, and spreading gradually over the whole southern part of the sky. These columns sometimes bent sideways at their upper extremity, and though in most respects similar to the northern lights, yet differed from them in being always of a whitish colour. The stars were sometimes hid by, and sometimes faintly to be seen through, the substance of these southern lights. The sky was generally clear, the air sharp and cold, the ship being in latitude 58° south.”

I shall instance another display occurring recently, on the evening of Sunday, the 28th of July, 1901, from my notes. “This evening the whole sky from 6.30 p.m to 9.30 p.m. was lit up with beautiful symmetrical bands of light, which appeared to radiate from the S. W. horizon, and after crossing the sky converged to the opposite or N.E. part of the horizon. The only motion observed was a slow lateral one of the whole bands towards the S.E. The moon (which was within three days off the full) was surrounded by a very distinct and perfect halo of 30° diameter, of the same white colour as the bands across the sky. The apparent convergence of these bands at their ends as they approach the horizon is possibly only the result of perspective. The sky was clear and the wind N.E.

“Observations as early as 1859 tended to connect solar energy with intense magnetic action. Moreover, brilliant auroral streamers were seen in both hemispheres on the night following the solar disturbance, proving that a relation exists between the aurora, terrestrial magnetism, and the central luminary of our system.”

It is remarkable that auroras and halos often occur together or near the same time, indicating a similarity of electrical and atmospheric conditions as necessary for their production. The year 1870 was characterized by the frequency of these phenomena, and in the summary of sun-spot observations made at Kew the observers remark that “the year 1870 was characterized by an exuberance of solar energy which is without parallel since the beginning of observations in 1825.”

I am enabled to give another instance of a daylight exhibition of these bands or arches, which occurred in the morning of the 11th October, 1901. On this occasion it was noticed