The properties of the magnet are best observed in magnetised steel; and when we proceed to test its magnetic power, it will be found that it is most active at the extremities of the bar, which are hence called its poles, and hardly, if at all, at the centre; that while both poles attract certain substances and repel others, the one always points nearly north and the other nearly south when the bar is horizontally suspended; and that, when we break the bar into two or any number of pieces, however small, each part forms into a complete magnet with its virtue active at the poles, which, when suspended, preserves its original direction; so that of two particles one is, in that case, always north of the other; nay, it is probable that each of these has its north pole and its south, as constant as those of the earth itself, which, too, is a large magnet. The magnet acts through media and at a distance, as well as in contact; and it has an especial attraction for iron, the more so when the conducting medium is solid, such as a table; and so when the magnet is horizontally suspended, or poised, in the vicinity of iron, its tendency to point north and south is seriously disturbed. The disturbance of the bar, or needle, in such a case, is called its _deflection_; and it is corrected by so placing a piece of soft iron or another magnet in its neighbourhood as to neutralise the effect, and leave said bar, or needle, free to obey the magnetism of the earth. The needle, it is to be remarked, does not point due north and south, neither, when poised freely on its centre, does it lie perfectly horizontal; in our latitude it points at present 20° west of north, which is called its _declination_, and its north pole slopes downwards at an angle of 68°, which is called its _dip_. By holding a rod of iron, or a poker, for a length of time parallel to the direction of the needle, so as to have the same declination and the same dip, it will gradually assume and display magnetic virtue, and this will ere long become fixed and powerful under a succession of vibratory shocks. There is a beautiful experiment in which a needle, when magnetised, can be made to float on water, when it adjusts itself to the magnetic meridian, and will incline north and south the same as the needle of the compass.