In regard to the latter point, namely, the generation and increase of the mag netic attraction, we shall endeavour to give a brief but distinct view of what re lates thereto ; observing, that where vol canic eruptions are frequent, and in those latitudes where the aurora borealis is dis tinctly seen, the needle or magnet is sen sibly affected. Previously to earthquakes, as well as during their action, and while the northern lights are in full display, no, reliance can be placed on the compass ; of which the card will appear much agi tated. This has given rise to the opinion held by some, that the power is a fluid : to this, however, there appear so many objections, that we are more disposed to reject than to favour it, although under the necessity of confessing that we are not able to offer one that may account sar tisfactorily for the various phenomena at tendant upon magnetism.
We have already stated, that every magnet has two poles ; that is, one end is called the north, the other the south, pole : the former being considered as-ca pable of attraction ; the other, as we shall infer from the subjoined explana tions, being far more inert, if at all pos sessed of an attractive power. When two magnets are brought together with their north poles in contact, they will, in stead of cohering, be obviously repelled to a distance corresponding with their re spective powers of attraction, when ap plied individually to unmagnetised nee dles. The south poles will, in like man ner, repel each other ; but the north pole of one, and the south pole of the other, will, when approximated, be evidently at tracted, and will cohere so as to sustain considerable weights. Iron is the only metal, hitherto known, which is capable of receiving and communicating the mag netic power ; but quiet, and the absence of contact, in some respects, are indis pensably necessary towards its perfect re tention. Thus, when a bar has been ina, pregnated, however abundantly, with the magnetic principle, if it be heated or hammered, the power of attraction will be dissipated ; or if a tube filled with iron filings have their surface magnetised, by shaking the tube the magnetic influ ence will likewise be lost. In some re spects, the magnetic influence resembles caloric ; for it very rapidly communicates to iron, devoid of magnetism, a certain portion of its own powers ; which, how ever, appear to be reproduced instanta neously. As various small fires under one large vessel will thereby heat it, and cause the water it contains to boil, though either of them individually would not produce that effect ; so many weak mag nets may, by causing each to communi cate a power equal to its own, be made to create an accumulated power, larger than that contained by either of them in dividually there is, however, a seeming contradiction to be found in some au thors, who recommend that the weakest Magnets should be first applied, and those more forcible in succession, according to the power they may possess ; the reason assigned being, that the weaker magnets would else, in all probability, draw off some of the accumulated power from the new magnet. Of this there appears no
danger, since experience proves that magnets rather gain than lose efficiency by contact, not only with each other, but even with common iron. In fact, the magnetic power may at any time be cre ated by various means : the friction of two pieces of flat and polished bars of iron will cause them for a short while to attract, and to suspend, light weights. Soft iron is more easily influenced, but steel will retain the influence longer. Lightning, electricity, and galvanism, be ing all of the same nature, equally render iron magnetic. It is also peculiar, that when two or more magnets are left for any time with their several north poles in contact, the whole will be thereby weak ened ; whereas, by leaving a piece of common iron attached to a magnet, the latter will acquire strength. It is also well known, that some pieces of steel quickly receive the magnetic influence, while others require considerable labour, and after all are scarcely impregnated. The oxide of iron cannot be impregnated, and those bars that have been so, when they become partially oxydized, lose their power. Hence we see the necessity of preserving the needles of compasses from rust.
Magnets have the power to act, not withstanding the intervention of substan ces in any degree porous between them, and the body to be acted upon : thus, if a needle be put on a sheet of paper, and a magnet be drawn tinder it, the needle will follow the course of the magnet. The peculiar affinity of the load-stone for iron is employed, with great success, by those who work in precious metals, for the separation of filings, &c. of iron from the smaller particles of gold, &c. A mag net being dipped into the vessel, in which the whole are blended, will attract all fer ruginous particles.