deed, informs us, that instances are recorded, where the northern lights have been seen to join, and form luminous balls, darting about with great velocity, and even leaving a train like the common fire-balls. It would seem, therefore, that the highest regions of the aurora borealis are the same with those in which fire balls move.

With regard to the cause of the aurora borealis naturally occurring was, that it was occasioned by the ascent of inflammable sulphureous vapours from the earth. To this supposition Dr Halley objects the immense extent of such phenomena, and that they are constantly observed to proceed from north to south, but never from south to north. This made him very reasonably conclude, that there was some connexion between the poles of the earth and the aurora borealis; but being unacquainted with the electric power, he supposed, that this earth was hollow, having within it a magnetic sphere, which corresponded in virtue with all the natural and artificial magnets on the surface; and the magnetic effluvia passing through the earth, from one pole of the central magnet to another, might sometimes become visible, in their course, which he thought was from north to south, and thus exhibit the beautiful coruscations of the aurora borealis. Had Dr Halley, however, known that a stroke of electricity would give polarity to a needle that had it not, or reverse the poles of one that had it before, he would undoubtedly have concluded the electric and magnetic effluvia to be the same, and that the aurora borealis was this fluid performing its circulation from one pole of the earth to the other. In fact, this very hypothesis is adapted by S. Beccaria; and by the supposed circulation of the electric fluid he accounts for the phenomena of magnetism and the aurora borealis in a manner perfectly similar to that of Dr Halley, only changing the phrase magnetic effluvia for electric fluid. The following is the account given us by Dr Priestley of Beccaria's sentiments on this matter.

"Since a sudden stroke of lightning gives polarity to magnets, he conjectures, that a regular and constant circulation of the whole mass of the fluid from north to south may be the original cause of magnetism in general.

"That this ethereal current is insensible to us, is no proof of its non-existence, since we ourselves are involved in it. He had seen birds fly near a thunder-cloud, as he was sure they would not have done had they been affected by its atmosphere.

"This current he would suppose to arise from one source, but from several, in the northern hemisphere, of the earth; and he thinks that the aurora borealis may be this electric matter performing its circulation in such a state of the atmosphere as renders it visible, or approaching the earth nearer than usual. Accordingly, very vivid appearances of this kind have been observed to occasion a fluctuation in the magnetic needle."

A direct disproof of this circulation, however, is furnished by the observation of Mr Forster already mentioned: with which, though neither Dr Halley nor S. Beccaria could be acquainted, they might have thought of it as a final proof either of the truth or falsehood of their hypothesis.—If the aurora borealis is no other than the electric fluid performing the above-mentioned circulation, it ought to dart from the horizon towards the zenith in the northern hemisphere, and from the zenith to the horizon in the southern one: but Mr Forster plainly tells us, that the columns shot up from the horizon towards the zenith as well in the southern hemisphere as in the northern; so that if the aurora borealis is to be reckoned the flittings of electric matter, its course is plainly directed from both poles toward the equator, and not from one pole to the other.

Concerning the cause of this phenomenon, Mr Canton has the following query: "Is not the aurora borealis the flashing of electrical fire from positive towards negative clouds at a great distance, through the upper part of the atmosphere where the refraction is least?" But to this we must reply in the negative; for in this case it would flash in every direction according to the position of the clouds, as well as from north to south. Besides this query, he conjectures, that when the needle is disturbed by the aurora borealis, that phenomenon proceeds from the electricity of the heated air; and supposes the air to have the property of becoming electric by heat, like the tourmalin. But neither does this hypothesis appear at all probable; because, in such a case, the aurora borealis ought to be most frequent in summer when the air is most heated, whereas it is found to be the reverse. Lastly, with these electrical hypotheses we shall contrast that of Mr Mairan, who imagined this phenomenon to proceed from the atmosphere of the sun, particles of which were thrown off by its centrifugal force acquired by his rotation on his axis; and that these particles falling upon the atmosphere of the earth near its equatorial parts, were from thence propelled by the diurnal motion of the earth towards the polar regions, where they formed the aurora borealis. This hypothesis, besides its being a mere supposition unsupported by one single appearance in nature, is liable to the objection already mentioned; for in this case the light should dart from the equator to the poles, and not from the poles to the equator: or if we should suppose this matter to be gradually accumulated at each of the poles, we must then make other suppositions equally vague and ill founded, concerning its getting back with such surprising rapidity in direct opposition to the power which once brought it thither.

The first person who seems to have endeavoured to find any positive proof of the electrical quality of the aurora borealis, was Dr Hamilton of Dublin. He observes, that though this phenomenon is commonly supposed to be electrical, yet he had not seen any attempt to prove that it is so; but the only proof he himself brings is an experiment of Mr Hawkesbee, by which the electric fluid is shown to put on appearances somewhat like the aurora borealis, when it passes through a vacuum. He observed, that when the air was most perfectly exhausted, the streams of electric matter were then quite white; but when a small quantity of air was let in, the light assumed more of a purple colour. The flashing of this light therefore from the dense regions of the atmosphere into such as are more rare, and the transitions through mediums of different density, he reckons the cause of the aurora borealis, and of the different colours it assumes.

Dr Hamilton's proof, then, of the electricity of the the aurora borealis consists entirely in the resemblance the two lights bear to one another; and if to this we add, that, during the time of an aurora borealis, the magnetic needle hath been disturbed, electric fire ob- tained from the atmosphere in plenty, and at some times different kinds of rumbling and hissing sounds heard, we have the sum of all the positive evidence in favour of the electric hypothesis.

Was the aurora borealis the first natural pheno- menon the solution of which had been attempted by elec- tricity, no doubt the proofs just now adduced would be very insufficient; but when it is considered, that we have indisputable evidence of the identity of the phe- nomena of thunder and of electricity; when we also consider, that the higher parts of our atmosphere are continually in a strongly electrified state; the analogy becomes so strong that we can scarce doubt of the au- ora borealis arising from the same cause. The only difficulty is, to give a good reason why the electricity of the atmosphere should be constantly found to direct its course from the poles towards the equator, and not from the equator to the poles; and this we think may be done in the following manner.

1. It is found that all electric bodies, when consi- derably heated, become conductors of electricity; thus hot air, hot glass, melted rosin, sealing-wax, &c., are all conductors, till their heat is dissipated, and then they again become electrics.

2. As the converse of every true proposition ought also to be true, it follows from the above one, that if electricity when heated become conductors, then non- electricity when subjected to violent degrees of cold ought to become electric. In one instance this has been verified by experience; water, which is a conduc- tor when warm or not violently cooled, is found to be- come electric when cooled to 20° below 0 of Fahrenheit's thermometer. With regard to metallic substan- ces, indeed, no experiments have as yet been made to determine whether their conducting power is affected by cold or not. Very probably we might not be able to produce such a degree of cold as sensibly to lessen their conducting power; but still the analogy will hold; and, as we are by no means able to produce the greatest degree of cold possible, reason will always sug- gest to us, that if a certain degree of cold changes one conductor into an electric, a sufficient degree of it will also change all others into electrics.

3. If cold is sufficient to change conducting sub- stances into electrics, it must also increase the electric power of such substances as are already electric; that is to say, very cold air, glass, rosin, &c., provided they are dry, will be more electric than when they are warmer. With regard to air, which is most to our present purpose, this is rendered extremely probable, by considering that clear frosty weather is of all others the most favourable for electric experiments. They may be made indeed to equal advantage almost in any state of the atmosphere, provided sufficient pains are used, but in dry hard frosts they will succeed much more easily than at any other time.

These three axioms being allowed, the cause of the aurora borealis is easily deduced from them. The air, all round the globe, at a certain height above its sur- face, is found to be exceedingly cold, and, as far as experiments have yet determined, exceedingly electro- cal also. The inferior parts of the atmosphere between the tropics, are violently heated during the day-time by the reflection of the sun's rays from the earth. Such air will therefore be a kind of conductor, and much more readily part with its electricity to the clouds and vapours floating in it, than the colder air towards the north and south poles. Hence the prodig- ious appearances of electricity in these regions, show- ing itself in thunder and other tempests of the most terrible kind. Immense quantities of the electric fluid are thus communicated to the earth; and the inferior warm atmosphere having once exhausted itself, must ne- cessarily be recruited from the upper and colder re- gion. This becomes very probable from what the French mathematicians observed when on the top of one of the Andes. They were often involved in clouds, which, sinking down into the warmer air, ap- peared there to be highly electrified, and discharged themselves in violent tempests of thunder and light- ning; while in the mean time, on the top of the mountain, they enjoyed a calm and serene sky. In the temperate and frigid zones, the inferior parts of the atmosphere never being so strongly heated, do not part with their electricity so easily as in the torrid zone, and consequently do not require such recruits from the upper regions; but notwithstanding the difference of heat observed in different parts of the earth near the surface, it is very probable that at considerable heights the degrees of cold are nearly equal all round it. Were there a like equality in the heat of the under part, there could never be any considerable loss of equili- brium in the electricity of the atmosphere; but as the hot air of the torrid zone is perpetually bringing down vast quantities of electric matter from the cold air that lies directly above it; and as the inferior parts of the atmosphere lying towards the north and south poles do not conduct in any great degree; it thence follows, that the upper parts of the atmosphere lying over the torrid zone will continually require a supply from the northern and southern regions. This easily shows the necessity of an electric current in the upper parts of the atmosphere from each pole towards the equator; and thus we are also furnished with a reason why the au- ora borealis appears more frequently in winter than in summer; namely, because at that time the electric power of the inferior atmosphere is greater on account of the cold than in summer; and consequently the abundant electricity of the upper regions must go al- most wholly off to the equatorial parts, it being impos- sible for it to get down to the earth; hence also the aurora borealis appears very frequent and bright in the frigid zones, the degree of cold in the upper and un- der regions of the atmosphere being much more nearly equal in these parts than in any other. In some parts of Siberia particularly, this meteor appears constantly from October to Christmas, and its coruscations are said to be very terrifying. Travellers agree, that here the aurora borealis appears in greatest perfection; and it is to be remarked, that Siberia is the coldest coun- try on earth. In confirmation of this, it may also be observed, that, from the experiments hitherto made with the electrical kite, the air appears considerably more electrical in winter than in summer, though the clouds are known to be often most violently electrified in the summer time; a proof, that the electricity natu- rally Aurora Borealis.

turally belonging to the air is in summer much more powerfully drawn off by the clouds than in the winter, owing to the excess of heat in summer, as already observed.

A considerable difficulty, however, still remains from the upright position which the streams of the aurora borealis are generally supposed to have; whereas, according to the hypothesis above mentioned, they ought rather to run directly from north to south. This difficulty occurred to Dr Halley; but he answers it by supposing his magnetic effluvia to pass from one pole to another in arches of great circles, arising to a vast height above the earth, and consequently darting from the places whence they arose almost like the radii of a circle; in which case, being sent off in a direction nearly perpendicular to the surface of the earth, they must necessarily appear erect to those who see them from any part of the surface, as is demonstrated by mathematicians. It is also reasonable to think that they will take this direction rather than any other, on account of their meeting with less resistance in the very high regions of the air than in such as are lower.

But the greatest difficulty still remains: for we have supposed the equilibrium of the atmosphere to be broken in the daytime, and restored only in the night; whereas, considering the immense velocity with which the electric fluid moves, the equilibrium ought to be restored in all parts almost instantaneously; yet the aurora borealis never appears except in the night, although its brightness is such as must sometimes make it visible to us did it really exist in the daytime.

In answer to this it must be observed, that though the passage of electricity through a good conductor is instantaneous, yet through a bad conductor it is observed to take some time in passing. As our atmosphere therefore, though very violently heated, is but a bad conductor of electricity; though the equilibrium in it is broken, it can by no means be instantaneously restored. Add to this, that as it is the action of the sun which breaks the equilibrium, so the same action, extending over half the globe, prevents almost any attempt to restore it till night, when flashes arise from various parts of the atmosphere, gradually extending themselves with a variety of undulations towards the equator.

It now remains to explain only one particularity of the aurora borealis, namely, that its streams do not always move with rapidity; sometimes appearing quite stationary for a considerable time, and sometimes being carried in different directions with a slow motion. To this indeed we can give no other reply, than that weak electric lights have been sometimes observed to put on the same appearance at the surface of the earth; and much more may we suppose them capable of doing so at great heights above it, where the conductors are both fewer in number and much more imperfect.

When M. de Romas was making experiments with an electric kite in Italy, a cylinder of blue light about four or five inches diameter was observed surrounding the string. This was in the daytime; but had it been night, he imagined it must have been four or five feet in diameter; and as the string was 780 feet long, it would probably have seemed pyramidal, pointing upwards like one of the streams of the aurora borealis.

A still more remarkable appearance, Dr Priestley tells us, was observed by Mr Hartman. He had been making electrical experiments for four or five hours together in a very small room; and upon going out of it, and returning with a light in his hand, walking pretty quick, he perceived a small flame following him at about three feet distance. Being alarmed at this appearance, he stopped to examine it, upon which it vanished. This last instance is very remarkable, and singular in its kind: from both, however, we are sufficiently warranted to conclude, that final portions of our atmosphere may by various causes be so much electrified as to shine, and likewise be moved from one place to another without parting with the electricity they have received, for a considerable time.

The corona, or circle, which is often formed near the zenith by the aurora borealis, is easily accounted for in the same manner. As this corona is commonly stationary for some time, we imagine it would be a very proper mark whereby to determine the distance of the meteor itself. If an aurora borealis, for instance, was observed by two persons, one at London, and the other at Edinburgh; by noting the stars among which the corona was observed at each place, its true altitude from the surface of the earth could easily be determined by trigonometry.

Under the article Atmosphere it was suggested, that no good proof had been as yet brought for the extreme rarity of the air usually supposed to take place at no very great heights above the earth. The brightness of the meteor there mentioned at 70 miles perpendicular from the surface, as also its figure, seemed to prove the air considerably denser at that distance from the earth. Though the height of the aurora borealis has never been determined, we can scarce imagine it to be greater than that of this meteor, or indeed so great; but although its streams resemble the passage of electric light through a vacuum, it cannot be from thence inferred, that the air is at all in a state similar to the vacuum of an air-pump in those places where the aurora borealis is produced; seeing we have instances of similar appearances being produced in very dense air. The plate of an electrophorus is often so highly electrified, as to throw out flashes from different parts as soon as it is lifted up, and by proper management it may be always made to emit long and broad flashes which shall scarcely be felt by the finger, instead of small, dense, and pungent sparks; so that, though long flashes may be produced in rarefied air, it by no means follows, that the same may not also be produced in denser air. As little can we infer anything from the colours; for we observe the electric spark sometimes white, sometimes blue, and sometimes purple, in the very same state of the atmosphere, and from the same substance.

The aurora borealis is said to be attended with a peculiar hissing noise in some very cold climates; Gmelin speaks of it in the most pointed terms, as frequent and very loud in the north-eastern parts of Siberia; and other travellers have related similar facts. Gmelin's account is very remarkable. "These northern lights (says he) begin with single bright pillars, rising in the north, and almost at the same time in the north-east, which gradually increasing, comprehend a large space of the heavens, rush about from place to place with incredible velocity, and finally almost cover the whole sky." sky up to the zenith. The streams are then seen meeting together in the zenith, and produce an appearance as if a vast extent was expanded in the heavens, glittering with gold, rubies, and sapphire. A more beautiful spectacle cannot be painted; but whoever should see such a northern light for the first time, could not behold it without terror. For however fine the illumination may be, it is attended, as I have learned from the relation of many persons, with such a hissing, cracking, and rushing noise throughout the air, as if the largest fireworks were playing off. To describe what they then hear, they make use of the expression, Spolachi ebodjat, that is, 'the raging holt is puffing.' The hunters who pursue the white and blue foxes on the confines of the icy sea, are often overtaken in their course by these northern lights. Their dogs are then so much frightened, that they will not move, but lie obstinately on the ground till the noise has passed. Commonly clear and calm weather follows this kind of northern lights. I have heard this account, not from one person only, but confirmed by the uniform testimony of many, who have spent part of several years in these very northern regions, and inhabited different countries from the Yenesei to the Lena; so that no doubt of its truth can remain. This seems indeed to be the real birthplace of the aurora borealis."

The hissing or rushing noise above described, Dr Blagden is inclined to attribute to small streams of electric matter running off to the earth from the mists or accumulations of electricity by which the northern lights are supposed to be produced.

We shall conclude this article with an account of a paper presented to the Royal Society by Mr Winn, in 1772, wherein he says that the appearance of an aurora borealis is a certain sign of a hard gale of wind from the south-west. This he never found to fail in 23 instances; and even thinks, that from the splendour of the meteor, some judgment may be formed concerning the ensuing tempest. If the aurora is very bright, the gale will come on within twenty-four hours, but will be of no long duration; if the light is faint and dull, the gale will be less violent, and longer in coming on, but it will also last longer. His observations were made in the English channel, where such winds are very dangerous; and by attending to the aurora, he says he often got easily out of it, when others narrowly escaped being wrecked. This is an exceedingly useful observation for sailors: but it cannot be expected that the winds succeeding these meteors should in all places blow from the south-west; though no doubt a careful observation of what winds succeed the aurora borealis, and other meteors, in different parts of the world, might contribute in some measure to lessen the dangers of navigation.

That the aurora borealis ought to be succeeded by winds, may be easily deduced from the hypothesis last mentioned. If this phenomenon is occasioned by the vast quantity of electric matter conveyed to the equatorial parts of the earth, it is certain that the earth cannot receive any great quantity of this matter at one place without emitting it at another. The electricity, therefore, which is constantly received at the equator, must be emitted nearer the poles, in order to perform its course, otherwise there could not be a constant supply of it for the common operations of nature. It is observed, that electrified bodies are always surrounded by a blast of air, which is sent forth from them in all directions; hence, if the electric matter find a more ready passage through one part of the earth than another, a wind will be found to blow from that quarter. If therefore one of these places happens to be in the Atlantic ocean near the coast of France, or in the bay of Biscay, the electric matter which has been received at the equator during an aurora borealis will be discharged there some time after, and consequently a wind will blow from that quarter, which will be from the south-west to those ships which are in the English channel. It cannot be imagined, however, that all the matter can be discharged from one place; and therefore according to the different situations of these electrical vents, winds may blow in different directions; and thus the same aurora borealis may produce a south-west wind in the English channel, and a north-west one in Scotland.