with ph. 13, and greatly corroborates and confirms the explanations we have given of this meteor.

PH. 26. Aurora Boreales are seen very frequently in the northern regions; they are of more rare occurrence as we advance towards the equator, so that in latitudes lower than 35° or 40° they are scarcely ever, if at all observed.

Exp. As we advance towards the north, the spiral currents of electric fluid moving north-west, (ph. 10 and 11), are more and more arrested, as above explained, and much contracted by their shorter revolution near the pole, and hence the Aurora will be more frequently seen the nearer we come towards the northern magnetic pole. But the causes we have assigned, and which as we have shewn evidently exist, indicate that they ought not to be seen in the parts of the earth near the torrid zone.

PH. 27. These meteors have been observed at all seasons of the year, but are most common at the spring and autumn, and even during winter, when the weather is serene and clear.

Exp. The changes in the weather are most frequent at the spring and autumn, and often after serene weather in winter a period of warm days ensues; but there is a more uniform warmth and moisture in summer: hence (ph. 13 and 25) these meteors should be seen as above stated.

PH. 28. In some years they are more frequent than in others.

Exp. The weather in some years is found to be far more uniform than in others, and consequently less favourable for producing these phenomena.

PH. 29. As the motions of the Aurora cease, the light approaches more and more towards the horizon, contracting from the east and west, and terminating near

the north; the dark segment, becoming luminous, disappears with greater or less rapidity.

Exp. This close of a brilliant meteor, depending, as above explained, on the electric fluid, and taking the positions specified, is such precisely as the circumstances would lead us to anticipate.

PH. 30. On Sept. 29, 1828, a fine rainbow-like arch was observed in England: at Bristol it was seen to rise a little before eight o'clock in the evening from an apparent cloud, or thin vapour, nearly at the WSW point of the horizon, it passed a few degrees to the north of Altair, in the Eagle, and nearly at right angles to the line produced which joins ẞ and y in that constellation, it proceeded along by Sagitta towards the ENE point of the horizon, passing a little to the south of the Pleiades in the east. It was of a yellowish pale white colour; it was seen in its greatest splendour at about a quarter past eight, and gradually faded, so that about nine it was scarcely discernible. It shifted its position a little; its breadth towards the west was about six degrees, but towards the east it was about three or four degrees broad. Nothing more of the Aurora was observed at the time, except a faint light, and a few feeble flashes of light which appeared towards the north, after the arch became invisible.

Exp. This is only a particular case of these meteors, and agrees with ph. 22, where it is shewn, that when such arches exist, they are first seen; sometimes there are two or three of them seen at once, and they are frequently unaccompanied by most of the other parts of the Aurora. Mr. Dobbs mentions several similar arches seen in Ireland, Sept. 24, 25, 26, 1725. Phil. Trans. No. 395,

p. 128.

Mr. Howard, in his important work on the Climate of

London, records the following account of a similar Aurora. Edinburgh, March 6. "A little past eight o'clock, p. m. a beautiful Aurora Borealis, nearly resembling that which appeared in September last, was distinctly visible here for a considerable time. A similar beautiful arch of bright light stretched across the heavens. It sprung from a point nearly ENE, and passing the zenith, terminated in the opposite point of the horizon. Its eastern limb was the brightest and best defined. The horizon in almost every point was obscured by dark broken clouds, which rendered both its beginning and termination less distinct than the last."

The three following phenomena are taken from Professor Hansteen's paper, Phil. Mag. Nov. 1827, p. 336.

PH. 31. The columns of light shooting up from the northern horizon towards the zenith are unconnected, short parallel rays, or cylinders of light, whose direction nearly coincides with that of the dipping needle. When these columns pass the zenith, they seem broken off, and form the crown.

Let N CS, fig. 73, be a part of the magnetic meridian, Ff, E e, and Dd, parallel rays forming the Aurora. If the observer at C turn his eyes in the direction of CF, CE, CD, a part of each column is covered by that in front, and the whole mass of light from F to Z, and Z to D, seems connected. But in the direction CZ one only sees the transversal section of the column, and hence the blue. sky appears. The light also seems connected, if we look towards the east or west; hence the whole seems to rise to the magnetic zenith.

Exp. What is here stated and illustrated, agrees with the explanations we have advanced in the preceding phenomena, and requires no other principles or explanation.

PH. 32. In the magnetic meridian these rays seem per

pendicular to the horizon, but towards the east or west they have an inclination towards the south, as represented in fig. 74; thus every observer will see the crown in his magnetic zenith, and therefore each sees a different crown; hence the altitude cannot be found by observations at different places.

Exp. The facts here noticed are fully explained in ph. 18, and agree with all that has been said on this subject.

PH. 33. The rays are frequently seen to form into a regular ring, DE, fig. 75, darting from a small zone of the earth's surface, the center of which is 'somewhere north of Hudson's Bay. If BC be the horizon at C, the angle FCB is the elevation of the lowest arch, and ECB of the highest.

Exp. Since the electric fluid, as shewn ph. 18, will find its way to the earth most copiously in an irregular zone, and particularly at some certain places in that zone, it will follow that in some cases of the Aurora, the appearances above mentioned will occur; and the space occupied by the ring is evidently the place in the irregular zone, (ph. 18), where the ethereal matter is discharged from the atmosphere in the greatest abundance.

The four following phenomena are taken from Mr. Richardson's paper in the Edin. Phil. Jour. for JulySept. 1828; the observations were made in the land expedition in 1825, 6, and 7.

PH. 34. The Aurora Borealis is generally most active when it seems to have emerged from a cloud near the earth.

Exp. The nearer the thin watery vapour (ph. 18) is to the earth, the more rapidly it will convey the electric fluid, and the more brilliant will be the coruscations of the Aurora, which agree with the facts here stated.

PH. 35. When the Aurora is very active, a haziness is perceived about the coruscations, though the sky in other parts is very clear.

Exp. The vivid light of the Aurora will render the slender mist, by which it is conveyed, sufficiently opaque to exhibit the apparent haziness.

PH. 36. A low temperature seems favourable for the production of brilliant and active coruscations.,

Exp. What has been advanced on ph. 25, will shew the reason of this fact.

PH. 37. The gold leaf electrometer was not affected by the appearance of the Aurora.

Exp. It is not to be expected that it should affect this instrument; for the Aurora takes place in the elevated regions of the air, and the air at those places of observation is not irregularly, or suddenly, or in any very great degree affected by its action.

Obs. Mr. Richardson observed that brilliant and active coruscations of the Aurora Borealis cause a deflection of the magnetic needle, and that almost invariably, if they appear through a hazy atmosphere, and exhibit the prismatic colours. This will be understood from the magnetic phenomena which follow. The altitude of these meteors is considerable, but by no means so great as many have imagined, as appears from the observations of Mr. Richardson, and others. There are many sources of fallacy in observations on these lights.

PH.38. The corona is not always formed exactly in the point of the heavens towards which the dipping needle is directed, nor is it always stationary during the short time of its appearance, but approaches towards the point just now specified: thus, that which was noticed by Dr. Halley, Phil. Trans. No. 347, p. 406, first appeared a