f the Compass, is the deviation of the magnetic or mariner's needle from the meridian or true north and south line. On the continent it is called the declination of the magnetic needle; and this is a better term, for reasons which will appear by and by.

We have given the general facts relating to magnetic variation under the article MAGNETISM, No 19; and under the articles COMPASS, and Azimuth COMPASS, we have noticed the methods of ascertaining the variation at any particular time or place. We shall here only give a short historical account of the progressive discoveries respecting magnetic variation, and notice the the explanations that have been offered to account for this phenomenon.

About the time that the polarity of the magnet was first observed in Europe, the magnetic direction, both in Europe and in China, was nearly in the plane of the meridian. It was therefore an indefinable present to the mariner, giving him a sure direction in his course through the pathless ocean. But by the time that the European navigators had engaged in their adventurous voyages to far distant shores, the deviation of the needle from the meridian was very sensible even in Europe. The son of Columbus positively says, that it was observed by his father in his first voyage to America, and made his companions so anxious lest they should not find the way back again to their own country, that they mutinied and refused to proceed. It is certain that Gonzales Oviedo and Sebastian Cabot observed it in their voyages. Indeed it could not possibly escape them; for in some parts of their several tracks the needle deviated above 25° from the meridian; and the rudest dead reckoning, made on the supposition of the needle pointing due north and south, must have thrown the navigators into the utmost confusion. We know that spherical trigonometry was at that time abundantly familiar to the mathematicians of Europe, and that no person pretended to take the command of a ship bound to a distant port that was not much more informed in this science than most masters of ships are at present. The deviation of the compass, however, was not generally allowed by mathematicians, who had not yet become sensible of the necessity of quitting the Aristotelian trammels, and investigating nature by experiments. They chose rather to charge the navigators with inaccuracy in their observations than the schoolmen with errors in principles. Pedro de Medina at Valladolid, in his Arte de Navigar, published in 1545, denies the variation of the compass. But the concurring reports of the commanders of ships on distant voyages, in a few years, obliged the landmen in their closets to give up the point; and Martin Cortez, in a treatise of navigation, printed at Seville before 1556, treats it as a thing completely established, and gives rules and instruments for discovering its quantity. About the year 1580 Norman published his discovery of the dip of the needle, and speaks largely of the horizontal deviation from the plane of the meridian, and attributes it to the attraction of a point, not in the heavens, but in the earth, and describes methods by which he hoped to find its place. To the third, and all the subsequent editions of Norman's book (called the New Attractive), was subjoined a dissertation by Mr Burroughs, comptroller of the navy, on the variation of the compass, in which are recorded the quantity of this deviation in many places; and he laments the obstacle which it causes to navigation by its total uncertainty previous to observation. The author indeed offers a rule for computing it, a priori, founded on some conjecture as to its cause; but, with the modesty and candour of a gentleman, acknowledges that this is but a guess, and intreats all navigators to be assiduous in their observations, and ready in communicating them to the public. Accordingly observations were liberally contributed from time to time, and were published in the subsequent treaties on navigation.

But in 1635 the mariners were thrown into a new and great perplexity, by the publication of a Discourse Mathematical on the Variation of the Magnetical Needle, Variation by Mr Henry Gillebrand, Greatham professor of astronomy. He had compared the variations observed at London by Burroughs, Gunter, and himself, and found that the north end of the mariner's needle was gradually drawing more to the westward. For Norman and Burroughs had observed it to point about 11 1/2 degrees to the east of north in 1580; Gunter found its deviation only 6° in 1622, and he himself had observed only 4° in 1634; and it has been found to deviate more and more to the westward ever since, as may be seen from the tables given under Magnetism.

Mr Bond, teacher of mathematics in London, and employed to edit and improve the impressions of the popular treatises of navigation, about 1650, declared, in a work called the "Seaman's Kalendar," that he had discovered the true progres of the deviation of the compass; and published in another work, called the "Longitude Found," a table of the variation for 50 years. This was, however, a gratuitous prognostication, not founded on any well-grounded principles; and though it agreed very well with the observations made in London, which showed a gradual motion to the westward at the rate of —12' annually, by no means agreed with the observations made in other places. See Phil. Trans. 1668.

But this news soon lost its credit: for the inconsistency with observation appeared more and more every day, and all were anxious to discover some general rule, by which a near guess at least might be made as to the direction of the needle in the most frequented seas. Halley recommended the matter in the most earnest manner to the attention of government; and, after much unweared solicitation, obtained a ship to be sent on a voyage of discovery for this purpose. He got the command of this ship, in which he repeatedly traversed the Atlantic ocean, and went as far as the 50th degree of southern latitude. See his very curious speculations on this subject in the Phil. Trans. 1683 and 1692.

After he had collected a prodigious number of observations made by others, and compared them with his own, he published in 1700 a synoptical account of them in a very ingenious form of a sea chart, where the ocean was crossed by a number of lines passing through those planes where the compass had the same deviation. Thus, in every point of one line there was no variation in 1700; in every point of another line the compass had 20° of east variation; and in every point of a third line it had 20° of west variation. These lines have since been called Halleyan lines, or curves. This chart was received with universal applause, and was undoubtedly one of the most valuable presents that science has made to the arts.

The polarity of the magnetic needle, and a general though intricate connection between its positions in all parts of the world, naturally makes the philosopher speculate about its cause. We see that Cortez ascribed it to the attraction of an eccentric point, and that Bond thought that this point was placed not in the heavens, but in the earth. This notion made the basis of the famous Theory of Magnetism of Dr Gilbert of Colchester. See Magnetism, No. 71.

Gilbert's theory may be understood from the following general proposition.

Let NS (fig. 1.) be a magnet, of which N is the north Fig. 1. Variation. north and S the south pole: Let ns be any oblong piece of iron, poited on a point c like a compass needle. It will arrange itself in a position nc s precisely the same with that which would be assumed by a compass needle of the same size and shape, having n for its north and s its south pole. And while the piece of iron remains in this position, it will be in all respects a magnet similar to the real compass needle. The pole n will attract the south pole of a small magnetised needle, and repel its north pole. If a paper be held over ns, and fine iron-filings be strewed on it, they will arrange themselves into curves issuing from one of its ends and terminating at the other, in the same manner as they will do when strewed on a paper held over a real compass needle. But this magnetism is quite temporary; for if the piece of iron ns be turned the other way, placing n where s now is, it will remain there, and will exhibit the same phenomena. We may here add, that if ns be almost infinitely small in comparison of NS, the line nc will be in such a position that if sa, sb, be drawn parallel to nc, sc, we shall have sa to sb, as the force of the pole N to the force of the pole S. And this is the true cause of that curious disposition of iron-filings when strewed round a magnet. Each fragment becomes a momentary magnet, and arranges itself in the true magnetic direction, and when so arranged, attracts the two adjoining fragments, and co-operates with the forces, which also arrange them. We throw this out to the ingenious mechanician as the foundation of a complete theory of the magnetical phenomena. When the filings are infinitely fine, the curves nc s have this property, that, drawing the tangent nc s, we always have sa : sb = force of N : force of S; and thus we may approximate at pleasure to the law of magnetic attraction and repulsion. The theory, of which an outline is given under MAGNETISM, is founded on this principle, and applies with success to every phenomenon yet observed.

Now, to apply this theory to the point in hand.—Let ns (fig. 2.) be a small compass needle, of which n is the north and s the south pole: let this needle be poited horizontally on the pin cd; and let ns' be the position of the dipping needle. Take any long bar of common iron, and hold it upright, or nearly so, as represented by AB. The lower end B will repel the pole n and will attract the pole s, thus exhibiting the properties of a north pole of the bar AB. Keeping B in its place, turn the bar round B' as a centre, till it come into the position A'B' nearly parallel to ns'. You will observe the compass needle ns attract the end B' with either pole n or s, when B'A' is in the position B'a perpendicular to the direction ns' of the dipping needle: and when the bar has come into the position B'A', the upper end B' will show itself to be a south pole by attracting n and repelling s. This beautiful experiment was exhibited to the Royal Society in 1673 by Mrs Hindshaw.

From this it appears, that the great magnet in the earth induces a momentary magnetism on soft iron precisely as a common magnet would do. Therefore (says Dr Gilbert) it induces permanent magnetism on magnetifiable ores of iron, such as loadstones, in the same manner as a great loadstone would do; and it affects the magnetism already imparted to a piece of tempered steel precisely as any other great magnet would.

Vol. XX. Part II.

Therefore the needle of the mariner's compass in every part of the world arranges itself in the magnetic direction, so that if poited as a dipping needle should be, it will be a tangent to one of the curves nc s of fig. 1. The horizontal needle being so poited as to be capable of playing only in a horizontal plane, will only arrange itself in the plane of the triangle nc s. That end of it which has the same magnetism with the south pole S of the great magnet included in the earth will be turned towards its north pole N. Therefore what we call the north pole of a needle or magnet really has the magnetism of the south pole of the great primitive magnet. If the line NS be called the axis, and N and S the poles of this great magnet, the plane of any one of these curves nc s will cut the earth's surface in the circumference of a circle, great or small according as the plane does or does not pass through the centre of the earth.

Dr Halley's first thought was, that the north pole of the great magnet or loadstone which was included in the bowels of the earth was not far from Bassin's bay, and its south pole in the Indian ocean south-west from New Zealand. But he could not find any positions of these two poles which would give the needle that particular position which it was observed to assume in different parts of the world; and he concluded that the great terrestrial loadstone had four irregular poles (a thing not unfrequent in natural loadstones, and easily producible at pleasure), two of which are stronger and two weaker. When the compass is at a great distance from the two north poles, it is affected so as to be directed nearly in a plane passing through the strongest. But if we make it approach much more to the weakest, the greater vicinity will compensate for the smaller absolute force of the weak pole, and occasion considerable irregularities. The appearances are favourable to this opinion. If this be the real constitution of the great magnet, it is almost a desperate task to ascertain by computation what will be the position of the needle. Halley seems to have despaired: for he was both an elegant and a most expert mathematician, and it would have cost him little trouble to ascertain the places of two poles only, and the direction which these would have given to the needle. But to say what would be its position when acted on by four poles, it was necessary to know the law by which the magnetic action varied by a variation of distance; and even then, the computation would have been exceedingly difficult.

In order to account for the change of variation, Dr Halley supposes this internal magnet not to adhere to the external shell which we inhabit, but to form a nucleus or kernel detached from it on all sides, and to be so poited as to revolve freely round an axis, the position of which he hopes to discover by observation of the compass. Dr Halley imagined that the nucleus revolved from east to west round the same axis with the earth. Thus the poles of the magnet would change their positions relatively to the earth's surface, and this would change the direction of the compass needle.

The great Euler, whose delight it was always to engage in the most difficult mathematical researches and computations, undertook to ascertain the position of the needle in every part of the earth. His dissertation on this subject is to be seen in the 13th volume of the Memoirs of the Royal Academy of Berlin, and is exceedingly Variation. irgely beautiful, abounding in those analytical tours d'adref/e in which he surpassed all the world. He has reduced the computation to a wonderful simplicity.

He found, however, that four poles would engage him in an analysis which would be excessively intricate, and has contented himself with computing for two only; observing that this supposition agrees so well with observation, that it is highly probable that this is the real constitution of the terrestrial magnet, and that the coincidence would have been perfect if he had hit on the due positions of the two poles. He places one of them in lat. 76° north, and long. 96° west from Teneriffe. The fourth pole is placed in lat. 38° south, and long. 158° west from Teneriffe. These are their situations for 1757.—Mr Euler has annexed to his dissertation a chart of Halleyan curves suited to these assumptions, and fitted to the year 1757.

It must be acknowledged, that the general course of the variations according to this theory greatly resembles the real state of things; and we cannot but own ourselves highly indebted to this great mathematician for having made so fine a first attempt. He has improved it very considerably in another dissertation in the 22d volume of these memoirs. But there are still such great differences, that the theory is of no use to the navigator, and it only serves as an excellent model for a farther prosecution of the subject. Since that time another large variation chart has been published, fitted to a late period; but the public has not sufficient information of the authorities or observations on which it is founded.

The great object in all these charts is to facilitate the discovery of a ship's longitude at sea. For the lines of variation being drawn on the chart, and the variation and the latitude being observed at sea, we have only to look on the chart for the intersection of the parallel of observed latitude and the Halleyan curve of observed variation. This intersection must be the place of the ship. This being the purpose, the Halleyan lines are of great service; but they do not give us a ready conception of the direction of the needle. We have always to imagine a line drawn through the point, cutting the meridian in the angle corresponding to the Halleyan line. We should learn the general magnetic affections of the globe much better if a number of magnetic meridians were drawn. These are the intersections of the earth's surface with planes passing through the magnetical axis, cutting one another in angles of 6° or 10°. This would both show us the places of the magnetic poles much more clearly, and would, in every place, show us at once the direction of the needle. In all those places where these magnetic curves touch the meridians, there is no variation; and the variation in every other place is the angle contained between these magnetic meridians and the true ones.

The program of a work of this kind has been published by a Mr Churchman, who appears to have engaged in the investigation with great zeal and considerable opportunities. It is pretty certain that the north magnetic pole (or point, as Mr Churchman calls it) is not far removed from the stations given it by Halley and Euler; and there seems no doubt but that in the countries between Hudson's bay and the western coasts of North America the needle will have every position with respect to the terrestrial meridian, so that the north end of a compass needle will even point due south in several places. Almost everything that can be derived in this inquiry would be obtained by a few well-chosen observations made in those regions. It would be of immense advantage to have the dips ascertained with great precision. These would enable us to judge at what depth under the surface the pole is situated; for the well-informed mechanician, who will study seriously what we have said about the magnetical curves, will see that a compass needle, when compared with the great terrestrial magnet, is but as a particle of iron-filings compared to a very large artificial magnet. Therefore, from the position of the dipping needle, we may infer the place of the pole, if the law of magnetic action be given; and this law may be found by means of other experiments which we could point out. See MAGNETISM, No 89, &c.

Mr Churchman has adopted the opinion of only two poles. According to him, the north pole was (in 1800) in Lat. 58° N. and Long. 134° W. from Greenwich, very near Cape Fairweather; and the south pole lies in Lat. 38° S. and Long. 165° E. from Greenwich. He also imagines that the north pole has moved to the eastward, on a parallel of latitude, about 65 since the beginning of the 19th century (from 1600), and concludes that it makes a revolution in 1096 years. The southern pole has moved less, and completes its revolution in 2280 years. This motion he ascribes to some influences which he calls magnetic tides, and which he seems to consider as celestial. This he infers from the changes of variation. He announces a physical theory on this subject, which, he says, enables him to compute the variation with precision for any time past or to come; and he even gives the process of trigonometrical computation illustrated by examples. But as this publication (entitled The Magnetic Atlas), published for the author, by Denton and Harvey, 1794) is only a program, he expresses himself obscurely, and somewhat enigmatically, respecting his theory. He speaks of the influence of one pole being greater than that of the other; and says, that in this case the magnetic equator, where the needle will be parallel to the axis, will not be in the middle between the poles. This is true of a common magnet. He must therefore abide by this supposition in its other consequences. The magnetic meridians must be planes passing through this axis, and therefore must be circles on the surface of the earth. This is incompatible with the observations; nay, his charts are so in many places, particularly in the Pacific ocean, where the variations by his chart are three times greater than what has been observed.—His parallels of dip are still more different from observation, and are incompatible with any phenomena that could be produced by a magnet having but two poles. His rules of computation are exceedingly exceptionable. He has in fact but one example, and that so particular, that the mode of computation will not apply to any other. This circumstance is not taken notice of in the enunciation of his first problem; and the reader is made to imagine that he has got a rule for computing the variation, whereas all the rules of calculation are only running in a circle. The variation computed for the port of St Peter and Paul in Kamtchatka, by the rule, is ten times greater than the truth.

For our own part, we have little hopes of this problem ever being subjected to accurate calculation. We believe, Variation, believe, indeed, that there is a conical change going on in the earth, which will produce a progressive change in the variation of the needle; and we see none more likely than Dr Halley's motion. There is nothing repugnant to our knowledge of the universe in the supposition of a magnetic nucleus revolving within this earth; and it is very easy to conceive a very simple motion of revolution, which shall produce the very motion of the sensible poles for which Mr Churchman contends. We need only suppose that the magnetical axis of this nucleus is not its axis of revolution. It may not even be fixed that axis; and this circumstance will cause the two poles to have different degrees of motion in relation to the shell which surrounds it.

But this regular progress of the magnet within the earth may produce very irregular motions of the compass needle, by the intervention of a third body susceptible of magnetism. The theory of which we have just given a hint comes here to our assistance. Suppose NS (fig. 3.) to represent the primitive magnet in the earth, and n s to be a stratum of iron ore susceptible of magnetism. Also let n' s' be another small mass of a similar ore; and let their situations and magnitudes be such as is exhibited in the figure. The fact will be, that n will be the north pole and s the south pole of the great stratum, and n' and s' will be the north and south poles of the small mass or loadstone. Any person may remove all doubts as to this, by making the experiment with a magnet NS, a piece of iron or soft tempered steel n s, and another piece n' s'. The well-informed and attentive reader will easily see, that by such interventions every conceivable anomaly may be produced. While the great magnet makes a revolution in any direction, the needle will change its position gradually, and with a certain regularity; but it will depend entirely on the size, shape, and situation, of these intervening masses of magnetizable iron ore, whether the change of variation of the compass shall be such as the primitive magnet alone would have produced, or whether it shall be of a kind wholly different.

Now, that such intervening disturbances may exist, is past contradiction. We know that even on the film of earth which we inhabit, and with which only we are acquainted, there are extensive strata or otherwise disposed masses of iron ores in a state susceptible of magnetism; and experiments made on bars of hard tempered steel, and on bits of such ores, assure us that the magnetism is not induced on such bodies in a moment, but propagated gradually along the mass.—That such disturbances do actually exist, we have many relations. There are many instances on record of very extensive magnetic rocks, which affect the needle to very considerable distances. The island of Elba in the Mediterranean is a very remarkable instance of this. The island of Cannay also, on the west of Scotland, has rocks which affect the needle at a great distance.

A similar effect is observed near the Feroe islands in the North sea; the compass has no determined direction when brought on shore. Journ. de Spavans, 1679, p. 174.

In Hudson's straits, in latitude 63°, the needle has hardly any polarity. Eller's Voyage to Hudson's Bay.

Bouguer observed the same thing in Peru. Nay, we believe that almost all rocks, especially of whim or trappe stone, contain iron in a proper state.

All this refers only to the thin crust through which the human eye has occasionally penetrated. Of what may be below we are ignorant; but when we see appearances which tally so remarkably with what would be the effects of great masses of magnetical bodies, modifying the general and regularly progressive action of a primitive magnet, whose existence and motion is inconsistent with nothing that we know of this globe, this manner of accounting for the observed change of variation has all the probability that we can desire. Nay, we apprehend that very considerable changes may be produced in the direction of the compass needle, even without the supposition of any internal motion. If the great magnet resembles many loadstones we are acquainted with, having more than two poles, we know that these poles will act on each other, and gradually change each other's force, and consequently the direction of the compass. This process, to be sure, tends to a state of things which will change no more.—But the period of human history, or of the history of the race of Adam, may make but a small part of the history of this globe; and therefore this objection is of little force.

There can be no doubt of the operation of the general terrestrial magnetism on every thing susceptible of magnetic properties; and we cannot hesitate to explain in this way many changes of magnetic direction which have been observed. Thus, in Italy, Father de la Torre observed, that during a great eruption of Vesuvius the variation was 16° in the morning, at noon it was 14°, and in the evening it was 10°, and that it continued in that state till the lava grew so dark as no longer to be visible in the night; after which it slowly increased to 13½, where it remained. Daniel Bernoulli found the needle changed its position 45' by an earthquake. Professor Muller at Mainz observed that the declination of the needle in that place was greatly affected by the earthquake in Calabria. Such streams of lava as flowed from Hecla in the last dreadful eruption must have made a transference of magnetic matter that would considerably affect the needle. But no observations seem to have been made on the occasion; for we know that common ironstone, which has no effect on the needle, will, by mere cementation with any inflammable substance, become magnetic. In this way Dr Knight sometimes made artificial loadstones.—But these are partial things, and not connected with the general change of variation now under consideration.

We have said so much on this subject, chiefly with the view of cautioning our readers against too languine expectations from any pretensions to the solution of this great problem. We may certainly gather from these observations, that even although the theory of the variation should be completed, we must expect (by what we already know of magnetism in general) that the disturbances of the needle, by local causes intervening between it and the great influence by which it is chiefly directed, may be so considerable as to affect the position of the compass needle in a very sensible manner: for we know that the metallic substances in the bowels of the earth are in a state of continual change, and this to an extent altogether unknown.

There is another irregularity of the mariner's needle that we have noticed under MAGNETISM, p. 365, namely, the daily variation. This was first observed by Mr George Graham in 1722 (Philosophical Transactions, No 383), and reported to the Royal Society of London. It usually moves (at least in Europe) to the westward from 8 morning till 2 P.M. and then gradually returns to its former situation. The diurnal variations are seldom less than 6° 5', and often much greater. Mr Graham mentions (Philosophical Transactions, No 428.) some observations by a Captain Hume, in a voyage to America, where he found the variation greatest in the afternoon. This being a general phenomenon, has also attracted the attention of philosophers. The most detailed accounts of it to be met with are those of Mr Canton (see Magnetism), in Philosophical Transactions, vol. ii. part 1. p. 399, and those of Van Swinden, in his Treatise on Electricity and Magnetism.

Mr Canton attempts to account for these changes of position, by observing that the force of a magnet is weakened by heat. A small magnet being placed near a compass needle, ENE from it, so as to make it deflect 45° from the natural position, the magnet was covered with a brass vessel, into which hot water was poured. The needle gradually receded from the magnet 45°, and returned gradually to its place as the water cooled. This is confirmed by uniform experience.

The parts of the earth to the eastward are first heated in the morning, and therefore the force of the earth is weakened, and the needle is made to move to the westward. But as the sun warms the western side of the earth in the afternoon, the motion of the needle must take the contrary direction.

But this way of explaining by a change in the force of the earth supposes that the changing cause is acting in opposition to some other force. We do not know of any such. The force, whatever it may be, seems simply to produce its own effect, in deranging the needle from the direction of terrestrial magnetism. If Æpinus's theory of magnetic action be admitted, we may suppose that the sun acts on the earth as a magnet acts on a piece of soft iron, and in the morning propels the fluid in the north-west parts. The needle directs itself to this constricted fluid, and therefore it points to the eastward of the magnetic north in the afternoon. And (to abide by the same theory) this induced magnetism will be somewhat greater when the earth is warmer; and therefore the diurnal variation will be greatest in summer. This change of position of the constricted fluid must be supposed to bear a very small ratio to the whole fluid, which is naturally supposed to be confounded in one pole of the great magnet in order to give it magnetism. Thus we shall have the diurnal variation a very small quantity. This is departing, however, from the principle of Mr Canton's explanation; and indeed we cannot see how the weakening the general force of the terrestrial magnet should make any change in the needle in respect to its direction; nor does it appear probable that the change of temperature produced by the sun will penetrate deep enough to produce any sensible effect on the magnetism. And if this be the cause, we think that the derangements of the needle should vary as the thermometer varies, which is not true. The other method of explaining is much better, if Æpinus's theory of magnetic attraction and repulsion be just; and we may suppose that it is only the secondary magnetism (i.e. that of the magnetizable minerals) that is sensibly affected by the heat; this will account very well for the greater mobility of the fluid in summer than in winter.

A great objection to either of these explanations is the prodigious diversity of the diurnal variations in different places. This is so very great, that we can scarcely ascribe the diurnal variation to any change in the magnetism of the primitive terrestrial magnet, and must rather look for its cause in local circumstances. This conclusion becomes more probable, when we learn that the deviation from the meridian and the deviation from the horizontal line are not affected at the same time. Van Swinden ascribes them solely to changes produced on the needles themselves. If their magnetism be greatly deranged by the sun's position, it may throw the magnetic centre away from the centre of the needle's motion, and thus produce a very small change of position. But if this be the cause, we should expect differences in different needles. Van Swinden says, that there are such, and that they are very great; but as he has not specified them, we cannot draw any conclusion.

But, besides this regular diurnal variation, there is another, which is subjected to no rule. The aurora borealis is observed (in Europe) to disturb the needle exceedingly, sometimes drawing it several degrees from its position. It is always observed to increase its deviation from the meridian, that is, an aurora borealis makes the needle point more westerly. This disturbance sometimes amounts to fix or seven degrees, and is generally observed to be greatest when the aurora borealis is most remarkable.

The observation of the connection of the polarity of the needle with the aurora borealis occurred to the late Professor Robison in 1759, when a midshipman on board the Royal William in the river St Lawrence. The point of the heavens to which all the rays of light converged was precisely that which was opposite to the south end of the dipping needle.

This is a very curious phenomenon, and we have not been able to find any connection between this meteor and the position of a magnetic needle. It is to be observed, that a needle of copper or wood, or any substance except iron, is not affected. We long thought it an electric phenomenon, and that the needle was affected as any other body balanced in the same manner would be; but a copper needle would then be affected.

We see the needle frequently disturbed both from its general annual position, and from the change made on it by the diurnal variation. This is probably the effect of aurora borealis which are invisible, either on account of thick weather or daylight. Van Swinden says, he seldom or never failed to observe aurora borealis immediately after any anomalous motion of the needle; and concluded that there had been one at the time, though he could not see it. Since no needle but a magnetic one is affected by the aurora borealis, we may conclude that there is some natural connection between this meteor and magnetism. This should farther incite us to observe the circumstance above mentioned, viz. that the south end of the dipping needle points to that part of the heavens where the rays of the aurora appear Variation to converge. We wish that this were diligently observed in places which have very different variation and dip of the mariner's needle.

For the diurnal and this irregular variation, consult the Diffrations of Celsius and of Hörter, in the Memoirs of Stockholm; Wargentin, Philosophical Transactions, vol. xlviii.; Braun (Comment. Petropol. Novi, tom. v. vii. ix.); Graham and Canton as above.