Magnetical Experiments and Observations. By Mr. Tiberius Cavallo, F. R. S.

III. *Magnetical Experiments and Observations.* By Mr. Tiberius Cavallo, F.R.S. The Lecture founded by the late Henry Baker, Esq. F.R.S. Read Nov. 16, 1786. THE Bakerian Lecture, which last year I had the honour to deliver to the Royal Society, contained the account of some magnetical experiments, particularly concerning the magnetism of brass, from which it appeared, that most brass becomes magnetic, so far as to attract the magnetic needle, by being hammered, and loses its magnetism by annealing or softening in the fire; but that there is some brass, which possesses no magnetism naturally, nor acquires any by hammering. Several experiments, made since the reading of that paper, having shewn a few particulars, which tend to correct what was advanced in the said Paper, I shall in the present lecture mention, first, those particulars, and shall then proceed to relate other experiments and observations relating to other branches of the same subject of magnetism. In performing the experiments on the magnetism of brass, I generally used a magnetic needle suspended in a peculiar manner, as it is described in my last lecture; viz. a common sewing needle, or a piece of steel wire rendered magnetic, and suspended suspended by means of a chain of hair; which sort of suspension I find not only from the experiments then made, but also by several subsequent trials, to be the nimblest hitherto contrived; because some substances which seem to be quite destitute of magnetism, by not attracting any of the magnetic needles otherwise suspended, will sensibly affect this. However, notwithstanding the nicety of this method for discovering a very low degree of magnetic attraction, it was found still inferior to that of exploring substances floating on the surface of quicksilver, as used by M. BRUGMAN*. It seemed, therefore, necessary to repeat some of those experiments on brass, and also upon platina, by examining their magnetism by this means, viz. by putting the pieces of brass or grains of platina upon the surface of quicksilver, and then presenting a strong magnet near them. The result of those experiments was, that very seldom a piece of brass, or grain of platina, occurred, which was not affected by the magnet; and even when they were not affected by it, their indifference, as may be expressed, was not very clear and decisive; and indeed there are very few substances in nature which, when examined by this means, are not in some degree attracted by the magnet, so general is the dispersion of iron, or such is the tendency which most bodies have towards the magnet. Such brass which in the former experiments appeared to have no magnetism naturally, nor to acquire any by hammering, was now found to be mostly magnetic, though in so very small a degree as to be discoverable only when floating upon quicksilver. The same was the case with the grains of platina before they were hammered; but after hammering, * See his Magnetismus, seu de Affinitatibus Magneticis, printed at Lyons in 1778. their attraction towards the magnet became more evident; whereas those pieces of brass, which naturally had not any degree of magnetism sufficient to affect the needle, nor acquired any by hammering, but yet shewed some tendency towards the magnet when floating upon quicksilver, never, or very seldom, had that tendency increased by hammering. As in the course of those experiments it naturally occurred to observe several particulars, which may be of use to those persons who wish to repeat these experiments, I shall now subjoin the principal of them. It is necessary first of all to observe, that the vessel wherein the quicksilver is put for the purpose of examining the magnetism of divers bodies, must be at least six inches in diameter; otherwise the substances that are set to float upon the mercury, will be continually running towards the sides of the vessel, on account of the curvature of the surface of that metal, which in narrow vessels begins from a greater distance from the edge, than in vessels of a larger diameter. It is necessary likewise to keep the quicksilver very clean, and also very pure; the want of which precautions will render the event of the experiments precarious. I have observed a very remarkable phenomenon, with respect to the surface of the mercury that is used for this purpose. It is, that though substances will float upon it with wonderful nimbleness, when the mercury is first poured out of the bottle into the open vessel, yet a short time after, viz. after having remained for an hour or two, and sometimes for a shorter time, exposed to the atmosphere, a piece of brass or other substance will by no means float upon it with equal facility; so that some pieces of brass, or grains of platina, which, after first pouring the quicksilver into the open vessel, were evidently attracted by the magnet, about about an hour after were not in the least attracted by it. The method which, when the surface of the quicksilver is rendered thus sluggish, will effectually purify it, is to pass the quicksilver through a funnel of paper, viz. a piece of clean writing paper rolled up conically, and having at its apex an aperture of about a fiftieth part of an inch in diameter; which operation is sometimes necessary even on first pouring out the quicksilver, and which I have often been obliged to repeat three or four times in the course of an hour. There seems to be formed a kind of crust upon the surface of the mercury when exposed, which, though invisible by mere inspection, may be perceived by moving the floating body; for if it be tried immediately after having passed the quicksilver through the paper funnel, the floating substance will seem to proceed by itself; whereas, some time after, the same body, when moved, seems to communicate that motion to the adjacent quicksilver, and to drag it along with itself, somewhat like when one moves a body, which floats upon the surface of a liquor that begins to coagulate. The formation of this crust seems to be mostly owing to the imperfect metals, which in various quantities are usually amalgamated with the common sort of quicksilver; because that amalgamation tends to dephlogisticate those metals, and the semi-calcined part floats at the top, and it is not unlikely that the dephlogistication goes on much quicker in the open air. The reality of this supposition is corroborated by observing, that the purer the quicksilver is, the smaller is the crust formed, or opposition made to the floating bodies. However, I have observed it in some measure even in the purest quicksilver that can be procured; and am inclined to think, that it must be partly partly owing to moisture, or to very fine dust that adheres to the quicksilver when exposed to the atmosphere. In performing such experiments care should be had to keep the air and the quicksilver as much undisturbed as possible, and also to present the magnet in a proper manner, viz. so as not to touch the surface of the mercury, nor to strike against the floating body, especially when that is in motion; for after the impulse, though that may be very slight, the floating body will be impelled backwards, which may be often mistaken for magnetic repulsion. The least exceptionable method is the following: first, if the floating body be in motion let it stop, then hold a strong artificial magnet nearly in a perpendicular direction, and with one pole just over one side of the floating body, or rather so that the perpendicular, let fall from the pole of the magnet to the surface of the quicksilver, may be about one-tenth of an inch distant from the body to be tried. The height of the magnet above the quicksilver should be just sufficient to let the floating body pass under it without touching it. In this situation the magnet must be held steady; and if the floating body has any magnetism, it will be soon drawn directly under the magnet. In these experiments it will be generally found, that one part of the floating body is more magnetic than the rest, which appears from that particular part being constantly drawn directly under the pole of the magnet; whereas, when the magnetism is diffused equably, the center of gravity of the body (provided its shape be not very irregular) becomes stationary just under the pole of the magnet. It is not every magnet that will discover the very weak magnetism of certain substances; for sometimes a powerful mag- net will evidently attract what a weaker one will not move in the least. I shall lastly observe, with respect to the experiments of last year's lecture, that though then I thought to have fused and incorporated together brass and iron, yet some subsequent trials gave reason to believe, that the iron is concealed in some part or other of the melted brass, rather than equably diffused through the substance of the latter; and the principal reason for this suspicion is, that when those pieces of mixed metal are tried upon the quicksilver, some points in their surfaces are generally attracted by the magnet in preference to others. Experiments and observations relating to the attraction between ferruginous substances and the magnet, in their different states of existence; to which are added, some thoughts concerning the cause of the variation of the needle. It is a proposition well established in magnetics, that soft iron, or soft steel, acquires magnetism very easily, and loses it with equal facility; but that hard steel acquires that power with difficulty, and afterwards retains it obstinately. From the consideration of those properties I was led to imagine, that if a piece of steel, whilst red-hot, were placed between magnetic bars, and whilst standing in that situation cold water were to be suddenly poured upon it, so as to harden it, there might, perhaps, be obtained an artificial magnet, much more powerful than what can be produced by the ordinary way; because the magnetic bars, employed for such purpose, would communicate a great degree of magnetic power to the steel, when red hot and consequently soft, which power would be fixed upon the steel by the hardening. In order to put this project to the trial, six magnetic bars were so disposed in an oblong earthen vessel, as that the north poles of three of them might be opposite the south poles of the three others, forming two parcels of bars, lying in the same direction, and about three inches asunder, which nearly was the length of the steel bar which was intended to be rendered magnetic. Things being thus disposed, the steel bar was made quite red hot, and in that state was placed between the magnetic bars; after which cold water was immediately poured upon it, which rendered it so hard as not to admit being filed: its magnetism was found to be considerably strong, but by no means extraordinary. From repeated trials with steel bars of different sizes, and by using a greater or less number of magnetic bars, I found, that short steel bars acquire a proportionably greater degree of magnetism by this method, than those which are longer; that the magnetism in the longer bars is not proportionably as strong, principally because the artificial magnets being placed at their extremities have very little power on those parts of the pieces of steel which are nearer its center; and, lastly, that when, in order to remedy the just mentioned impediment, more magnets are placed nearer the middle of the steel bar, then this piece of steel generally acquires many successive magnetic poles. Upon the whole it seems, that though this method alone be not sufficient to communicate to steel bars an extraordinary degree of magnetism, yet it may be of great use in constructing large artificial magnets; for if those bars, instead of being hardened in the usual way by plunging them when red-hot in water, be hardened whilst standing between powerful magnets, they will thereby acquire a considerable degree of magnetic power without any additional trouble to the workmen. men. They may then be polished, after which they may be rendered more strongly magnetic by the usual method of touching them with other magnetic bars; whereas it is a very laborious operation to render magnetic large bars of hardened steel from the very beginning, viz. when they have none of that power. In the course of performing those trials, I frequently observed, that the pieces of steel, whilst they were red hot, seemed not to be attracted by the magnets; so that the least shock, and even the pouring of the water, could remove them from the proper situation, which rather surprised me; because it has been asserted, by some authors, that the magnet attracts red-hot iron as well as cold. Kircher especially says, that he tried the experiment*, and found that a piece of iron heated so as to be hardly discerned from a burning coal, was attracted by the magnet as easily as when cold; and he even assigns a reason why the power of a magnet is destroyed by a great degree of heat; whereas the red heating of the iron will not prevent its being attracted by the magnet. The reason he gives is, that the fire corrupts and calcines the magnet, but purifies the iron. The following experiments were made in order to ascertain this matter. I kept a piece of steel in the fire till it was quite red hot, and in that state presented the magnet to it, so as to touch it repeatedly in various places; but no sign of attraction could be perceived before the redness disappeared. I mean, however, such redness as may be evidently seen in the clear daylight; for, as was shewn by other experiments, when the magnet begins to attract the heated iron, the redness of the latter can still be seen in the dark. * De Magnete, lib. I. p. II. theorem xxxi. Having repeated the experiment with different pieces of iron and of steel, the result was constantly the same, viz. whilst the iron or steel remained quite red hot or white hot, the magnet did not attract it; but the attraction began when that degree of redness, which is clearly perceivable in the daylight, began to disappear; and it was as strong as ever when the iron was cooled a little more than when the redness quite disappeared in the dark. In regard to this limit or maximum of attraction, I think I have observed, as well as the nature of the experiments would permit, a difference between steel and iron, which is, that in the steel the maximum of attraction follows the disappearance of the red heat sooner than in iron. This experiment is subject to two sources of mistake, which perhaps misled Father Kircher, and which are necessary to be mentioned for the sake of others who wish to repeat this experiment. The first is, that when a piece of iron of no great extent is red hot, or even white hot, in one place, and below a red heat in other parts, the magnet will frequently attract it, though the red-hot side be presented to it. The second cause of mistake is, that when a small piece of iron or steel, as a common sewing needle, is made red-hot, and is then presented to the magnet, if the magnet touch it, that contact cools it instantly below the necessary degree of heat, and of course the attraction takes place. It is owing to this last cause that I have not yet been able to ascertain, whether the attraction between the magnet and the iron be quite annihilated, or only diminished to a great degree, by rendering the iron red or white hot; so that I can only say with certainty, that a magnet will not attract a certain piece of iron red hot or white hot; whereas it will attract another piece of iron at least fifty times bigger, if it be cold or below a red heat. To try this property in a different and more convincing way, I heated a large iron nail till it was white hot, and in that state placed it upon an earthen support near one pole of the magnetic needle, so as to lie not in the same direction, but on one side of it. Then, looking attentively on the graduated circle of the compass, I observed, that the needle was not in the least moved from its natural situation, whilst the nail remained red-hot; but, as soon as the redness began to disappear, the needle advanced towards the nail, and a few seconds after the needle pointed directly towards it. I tried whether in this experiment any difference was occasioned by the magnets being natural or artificial; but, as it might be expected, there was none. In pursuance of those magnetic experiments wherein heat is concerned, I tried the effects which took place when the magnet was heated; but as the diminution of its power by heating, and an increase of it by cooling, were observed and described by the late Mr. Canton, I shall only add a circumstance, which may perhaps be new. It is that an artificial magnet, after having had its power diminished by heating, does not recover it entirely again by cooling; having constantly found, that the magnets which had been heated, after cooling would never hold as great a weight of iron as they did before. The heat to which those magnets were exposed never exceeded that of boiling water. This was rendered more evident by the following experiment. A magnetic bar was placed in an earthen vessel at some distance from the south pole of the needle of a very good compass; by the action of which magnet that end of the needle was drawn several degrees from the magnetic meridian, or the direction in which it stood before. In this situation of the apparatus apparatus boiling water was poured into the vessel wherein the magnet stood, in consequence of which the needle went back two degrees and a half. Some time after, when the water was quite cold, the needle was found nearer to the magnet, but not so near as it stood before the hot water was poured into the vessel. Next to the effects of heat, I was desirous of trying what could be effected by decomposing the iron; and for this purpose an earthen vessel, containing about two ounces of iron filings, was placed near the south end of the needle of the compass, by which the needle was drawn a little out of its natural direction. Having marked where the needle now stood, some water first, and then some vitriolic acid, were poured upon the filings, which occasioned a brisk effervescence, and a copious production of inflammable air; but soon after the beginning of the effervescence, I was surprised to observe, that the needle came nearer to the vessel, shewing that the attraction between the needle and the filings had been increased by the action of the vitriolic acid upon the latter, which is contrary to what could have been expected; for if we consider that the power of a magnet is diminished by heat, and that red-hot iron has either no attraction at all, or an exceedingly small degree of it, towards the magnet, we might have concluded, that the action of the vitriolic acid upon iron would immediately diminish its attraction, besides the other strong reason arising from the dephtlogistication of the iron occasioned by the effervescence; and in fact some time after, when the violence of the effervescence, and of course the production of inflammable air, begins to abate, the attraction between the needle and the filings begins likewise to diminish; and at last, when the effervescence is hardly perceptible, the needle is found to stand stand farther from the vessel containing the filings, &c. than it stood before the vitriolic acid was added, which diminution of attraction is certainly owing to the loss of phlogiston; it being well known, that iron is less and less attracted by the magnet in proportion as it approaches nearer to the calcined state. Here follows the particular account of the above-mentioned experiment. The south end of the needle coincided with the $285^\circ 15'$ on the divided circle. When the pot with the iron filings was placed on one side of it, the needle pointed to $286^\circ 15'$, being drawn just one degree nearer. Having added the diluted vitriolic acid to the filings, the needle came nearer, and stood at $286^\circ 45'$. Ten minutes after the beginning of the effervescence it stood at $286^\circ 35'$, having receded a little; and a few minutes after this observation it stood at $286^\circ 30'$. An additional quantity of diluted vitriolic acid was now added, which increased the effervescence considerably; and on observing it a short time after, the needle was found at the same point at which it stood before, from which time it began to go back very gradually; so that about three hours after it stood at $285^\circ 50'$, viz. farther from the effervescing mixture than it stood before any vitriolic acid was poured upon the iron filings. As a single experiment ought not to be depended upon when an error may be occasioned by many concurring circumstances, I repeated this experiment with great precaution, taking care that nothing could shake the needle, or the rest of the apparatus; but the result was nearly the same, the attraction between the iron filings and the needle being increased by the action of the vitriolic acid. In order to ascertain that this effect was not owing to the heat generated by the effervescence, the pot, with some iron filings, was placed near the magnetic needle, as before; then some boiling water was poured upon the filings, which heated them much more than the diluted vitriolic acid could have done; but the magnetic needle was not moved in the least from its original situation. The suspicion which occurred next was, whether the effervescence might not agitate the iron filings so as to bring a greater number of them to that side of the vessel, which stands contiguous to the magnetic needle. In order to obviate this objection, the experiment was repeated with a single piece of iron instead of the filings; but as this piece of iron presented a very small surface to the diluted acid, the effervescence was very weak, and of course the magnetic needle did not move when the acid was poured into the pot. At last, in order to remedy this second inconvenience, arising from the want of surface, I used a long piece of small steel wire, which was twisted in various directions, so as to be admitted into the pot; in which case the metal presented a large surface to the acid, and it was not subject to be moved by the effervescence. The result was similar to that of the first experiment, viz. the attraction was increased by the action of the acid on the wire; and here follows the particular account of the experiment. About six yards of clean steel wire, somewhat less than one-fiftieth of an inch in diameter, being twisted in various directions, was put into an earthen vessel, which was placed near the south end of the magnetic needle, which in consequence of that was drawn from its natural situation, viz. from $281^\circ$ to $280^\circ$. After adding the diluted vitriolic acid, a strong effervescence commenced, and the needle came to $279^\circ 47'$. About five minutes after, it stood at $279^\circ 35'$. Five minutes after this, it stood at $279^\circ 30'$. And a little after this observation, it ap- peared to be even somewhat nearer to the pot than the above- mentioned point. The experiment was then discontinued, and on removing the pot, the needle went back to its original situa- tion, viz. $281^\circ$; which shewed, that its alteration during the process was occasioned by the action of the acid on the iron, and not by any extraneous cause. On examination the wire was found only blackened on its sur- face, but not nearly consumed; I had therefore the curiosity of trying the same wire again, and accordingly it was placed in the same vessel near the magnetic needle, which attracted the latter from its original situation $281^\circ$ to $280^\circ$. After adding the acid, the needle came nearer, as in the preceding experiment; and a short time after it stood at $279^\circ 30'$, at which time the pot was removed, there being no occasion to continue the expe- riment any longer. On pouring the liquor out of the pot, the wire did not appear to be much wasted. The pot was then replaced near the needle, so as to attract it a little nearer as before; but on pour- ing boiling water upon the wire, a pretty brisk effervescence took place, and the needle was in consequence attracted still nearer. This experiment shewed, that though the diluted acid had been poured out, yet there remained a quantity of it adhering to the wire, which was sufficient to renew the effe- vescence, when assisted by the heat of the boiling water. Upon the whole, it appears, that the action of vitriolic acid upon iron or steel increases their attraction towards the mag- net; that this increase of attraction has a limit, after which it begins to decrease; and that this limit seems to come sooner when iron than when steel is used; but, however, in respect to this last particular I am not yet quite certain, since, in the Mr. Cavalló's Magnetical experiments hitherto made, the variety in the shape or bulk of the iron or steel may have occasioned a considerable difference. After the result of those experiments, it was natural to examine the effect which other acids might have on iron and steel; therefore the above-mentioned experiment of the steel wire was repeated with nitrous instead of vitriolic acid; the result of which was that the attraction between the magnetic needle and the wire was increased, but not so much as when vitriolic acid had been used. Here follows the particular account of the experiment. About six yards of clean steel wire, near one-fiftieth of an inch in diameter, being twisted in various directions, was placed in the usual vessel near the south end of the needle; in consequence of which the needle was attracted from its natural situation, viz. from $283^\circ 32'$ to $282^\circ 50'$. About two ounces of water were then poured over the wire, and immediately after, near one ounce of nitrous acid was added, which produced an effervescence: the magnetic needle, however, hardly moved from its former situation; but in about a minute's time, the effervescence being increased very much, the needle came to $282^\circ 42'$. About a quarter of an hour after, the violence of the effervescence abated a little, and the needle went back again to $282^\circ 50'$. A short time after, it stood at $283^\circ 2'$. At last, when about half an hour had elapsed since the beginning of the operation, the effervescence was hardly perceivable, the liquor was become red like the colour of red ochre, and the needle stood at $283^\circ 15'$, viz. farther from the vessel than it did before the acid was added to the wire. It appears, therefore, that the effervescence occasioned by the nitrous acid produced a similar effect, though not in so great a degree as the vitriolic. The maximum of the attraction seems to come sooner when nitrous than when vitriolic acid is used, after which limit the attraction decreases much faster in the former than in the latter case, which is evidently owing to the metal being more easily dephlogisticated and dissolved by the nitrous than by the vitriolic acid. The marine acid was tried next; but, notwithstanding all the precautions I could take, it always occasioned a very weak effervescence, and the needle was not in the least affected by it. A strong effervescence seems to be necessary to increase the attraction between the magnet and iron or steel; for when I tried the experiment by putting a small quantity of acid into the vessel, the effervescence was exceedingly weak; the magnetic needle was not at first affected by it, but several hours after it stood a little farther from the pot, which was evidently owing to the dephlogistication of the iron. After observing the action of acids on iron, I next tried to decompose that metal by means of fire, to observe what effect would take place with respect to its magnetism. For this purpose two ounces of iron filings were mixed with an ounce of flowers of sulphur and an ounce of nitre. This mixture was put into a small and shallow earthen vessel, and was placed near the south extremity of the magnetic needle, which attracted the needle nearer than its natural situation by about one degree and a half. A pane of glass was interposed between the magnetic needle and the vessel. Things being thus disposed, the mixture was fired, and it burned rather slowly, the fire sometimes going out, so as to require being fired again; till at last it would burn no longer. During this combustion the needle was once attracted somewhat nearer; but its motion that way was so little, that I could not be quite certain of it. This happened not long after the fire was first communicated to the mixture, after: after which the needle generally vibrated backwards and forwards, but upon the whole it gradually receded. On repeating the experiment with a larger quantity of the mixture, and also with different proportions of ingredients, I could not observe any particular attraction. The needle vibrated rather more than before, but gradually receded; so that at last it stood farther from the vessel than it did before the mixture was set on fire, though not quite in its natural direction. The vibration of the needle in this experiment, or its waving motion, was probably owing to the irregular burning of the mixture, and perhaps to the heating of one part of it whilst the other was burning. The gradual receding of the needle was certainly owing to the dephlogistication of the iron. After having thus related the result of experiments, I shall now beg leave to add a few thoughts concerning the application of those observations towards accounting for the variation of the magnetic needle. This wonderful phenomenon has, since it was first discovered, employed the thoughts of very able philosophers; many hypotheses having been offered, not only for its explanation, but even to foretel the future variations in various parts of the world. I need not detain this learned Society with a particular history of those hypotheses, but shall only observe in general, that neither their predictions have answered, nor any of them was founded upon evident principles. The supposition of a large magnet being inclosed within the body of the earth, and of its relatively moving with respect to the outward shell or crust; the supposition of there being four moveable magnetic poles within the earth; the hypothesis of a magnetic power, partly within and partly without the surface of the earth; together with several other hypotheses on the same subject, are not only only unwarranted by actual experiments, but do neither seem analogous to the other operations of nature. The late ingenious Mr. Canton, Member of this Society, was the first, who endeavoured to account for the daily variation of the magnetic needle by the heating and cooling of the magnetic bodies in different parts of the earth's surface; which was in consequence of his having first observed, that the action of the magnet on the needle was diminished by heating, and increased by cooling*. Following Mr. Canton's judicious method of deriving the explanation of natural appearances from properties actually proved by experiments, I think, that the increase and diminution of magnetic attraction by heating and cooling of the magnet, as observed by Mr. Canton, together with the result of the experiments recited in this Paper, seem fully sufficient to account for the general variation of the needle. If we collect under one point of view all the causes hitherto ascertained, which can increase or diminish the attraction between magnetic bodies, we shall find, that the attraction between the magnet and iron, or between magnet and magnet, is increased by cooling, by a regeneration of iron or phlogistication of its calx, and within certain limitations by the action of acids upon the iron; that this attraction is diminished by heating, and by the decomposition of iron; and, lastly, that it is probably annihilated by a very great degree of heat. These truths being premised, it must be considered, first, that, according to innumerable observations and daily experience, the body of the earth contains almost everywhere ferruginous bodies in various states and bulks; secondly, that the magnetic needle must be attracted by all those bodies, and its * Philosophical Transactions, vol. LI. p. 398. situation or direction must be determined by all those attractions considered together, viz. from their common center of action; thirdly, that by removing or altering the degrees of attraction of some of those bodies which are situate on one side of the magnetic meridian, more than of those situated on the other side, the above-mentioned common center of attractions, and, of course, the direction of the magnetic needle must be altered, which in fact is the variation of the needle; and, lastly, that this alteration in the attractions of some of the ferruginous bodies in the earth must undoubtedly take place, it being occasioned by the parts of the earth being irregularly heated and cooled, by the action of volcanoes which decompose or otherwise alter large masses of ferruginous substance, by earthquakes which remove ferruginous bodies from their original places, and we may add also by the aurora borealis; for though we are as yet ignorant of the cause of that surprising phenomenon, it is however certain, that the magnetic needle has been frequently disturbed when the aurora borealis appeared very strong. The magnetic needle, therefore, being necessarily affected by those causes, it seems unnecessary to have recourse to other hypothetical causes which are not established on actual experience. In order to exemplify this explanation of the variation in a familiar manner, I made the following experiment, with the account of which I shall conclude this Paper. Four earthen vessels were disposed round the magnetic needle, two near its south, and the other two near its north pole, but not at equal distances. In one of those vessels there was placed a natural magnet; the second contained several small bits of magnetic steel mixed with earth; and in each of the other two there were put put about four ounces of iron filings. Things being thus disposed, and left undisturbed for about half an hour, the needle remained unaltered. Then the pieces of magnetic steel and earth were stirred with a stick, in consequence of which the needle was agitated. After this, some diluted vitriolic acid was poured upon the filings in one of the vessels, the action of which attracted the needle that way; but whilst the needle remained in that situation, some diluted vitriolic acid was poured upon the iron filings in the other vessel, which stood on the other side, in consequence of which the needle went back again towards its former direction. Whilst the effervescences were going on in the two vessels, the magnet in the first vessel was heated by means of boiling water, which occasioned another alteration in the direction of the magnetic needle; and thus, by altering the state of the ferruginous substances in the vessels, the needle's direction was altered, in evident imitation of the natural variation. T. CAVALLO.