The discovery of the properties possessed by fragments of a ferrous mineral (magnetite) to attract small pieces of iron to two areas of their surface (called poles) appears to have taken place long before Taletes from Mileto (600 BC) discovered the property of amber to attract sufficiently small bodies when rubbed with wool or silk.
However, the first scientific study of phenomena connected with this property (magnetism) was published only in 1600 in De Magnete by William Gilbert (1544-1603), a physician at the court of British Queen Elizabeth I.

The name of magnetite given to that mineral, that of magnet, referring to its fragments, and that of magnetism referring to phenomena connected to it, comes from the name of a town in Asia Minor (Magnesia) near which this mineral was found. Since these magnets (called natural) were usually weak, their force of attraction was enhanced by an armature of soft iron which, when suitably arranged, brought the maximum load they could carry up to two hundred times more than they could carry prior to application of the armature.

Once the poles of a natural magnet had been found, its surfaces were worked to produce two smooth faces normal to the axis of the poles. Two plates of very soft iron, which covered the faces completely, were applied to these faces.
The plates, which were held closely to the faces of the magnet by a copper frame, ended on the bottom with two short, well-flattened appendices, usually having a rectangular section, which served as polar expansions. A soft iron anchor was applied to the polar expansions. It had a hook from which weights were hung to demonstrate the great force of attraction of the magnet.
In such a way the magnet's force of attraction was greatly increased; besides exploiting the attraction of both poles, all the lines of force of the magnetic field were conveyed to the soft iron plates and then to the polar expansions to which the anchor adhered perfectly.