Clerk Maxwell was so fascinated with this beautiful concept that he made it the work of his life to study the field of force due to electrified bodies, to magnets and to conductors conveying currents; his powerful intellect visualized those lines and gave them accurate mathematical expression in the great treatise on electricity and magnetism which he gave to the world in 1873. Gilbert observes that the lodestone may be spherical or oblong; "whatever the shape, imperfect or irregular, verticity is present; there are poles," and the lodestones "have the selfsame way of turning to the poles of the world." He knows that a compass-needle is not drawn bodily towards the pole, and does not hesitate in this instance to give credit to his countryman, Robert Norman, for having clearly stated this fact and aptly demonstrated it. Following Norman, he floats a needle in a vessel by means of a piece of cork, and notices that on whatever part of the surface of the water it may be placed, the needle settles down after a few swings invariably in the same direction. His words are: "It revolves on its iron center and is not borne towards the rim of the vessel." Gilbert knew nothing about the mechanical couple that came into play, but he knew the fact; and, with the instinct of the philosopher, tested it in a variety of ways. We explain the orientation of the compass-needle by saying that it is acted upon by a pair of equal and opposite forces due to the influence of the terrestrial magnetic poles on each end of the needle and by showing that such a couple can produce rotation, but not translation. We find Gilbert working not only with steel needles and iron bars, but also with rings of iron. He strokes them with a natural magnet and feels certain that he has magnetized them. He assures us that "one of the poles will be at the point rubbed and the other will be at the opposite side." To show that the ring is really magnetized, he cuts it across, opens it out, and finds that the ends exhibit polar properties. A favorite piece of apparatus with Gilbert, as with Peregrinus, was a lodestone ground down into globular form. He called it a terrella, a miniature earth, and used it extensively for reproducing the phenomena described by magnetizers, travelers and navigators. He breaks up terrellas, in order to examine the magnetic condition of their inner parts. There is not a doubtful utterance in his description of what he finds; he speaks clearly and emphatically. "If magnetic bodies be divided, or in any way broken up, each several part hath a north and a south end"; _i.e._, each part will be a complete magnet. [Illustration: FIG. 8 Behavior of compass-needle on a _terrella_ or spherical lodestone]