Magnetism acts, without any diminution of its force, through any medium; and iron not magnetical will have that power while it is in connexion with a magnet, or rather the power of the magnet is extended through the iron.
Steel filings gently thrown upon a magnet, adhere to it in a curious manner; and the filings, acquiring magnetism by the contact, adhere together, and form a number of small magnets, which arrange themselves according to the attraction of the poles of the original magnet. This experiment is made to the most advantage upon a piece of pasteboard, or paper, placed over the magnet.
Magnetism is communicated by the friction, or the near position, of a magnet to a piece of steel of a size less than it. For this reason a combination of magnetical bars will have a greater effect than a single one; and in the following manner, beginning without any magnetism at all, the greatest quantity may be procured. Six bars of steel may be rendered slightly magnetical by fixing each of them successively to an upright poker, and stroking it several times from the bottom to the top with the lower end of an old pair of tongs. If then four of these bars be joined, the magnetism of the remaining two will be much increased by a proper method of rubbing with them; and by changing their places, joining the strongest, and acting upon the weakest, they may all be made as magnetical as they are capable of being.
The strength of a natural magnet may be increased by covering its polar extremities with steel. This is called the _arming_ of the loadstone.
To account for the variation of the needle, Dr. Halley supposed the earth to consist of two parts, an external _shell_ and an internal _nucleus_, detached, and having a revolution distinct from it; and that the action of the poles of the shell and of the nucleus would explain all the varieties in the position of the needle. But others think that the cause of the magnetism of the earth is not _within_, but _without_ itself. One reason for this opinion is, that a magnet is liable to be affected by a strong aurora borealis; and another is, that the variation of the needle proceeds in such manner as supposes that the motion of the nucleus must be quicker than that of the shell of the earth; whereas, since it is most natural to suppose that motion was communicated to the nucleus by the shell, it would be slower.