1600 saw the light under the title of _De Magnete_, to which was added the sub-title: _magneticisque corporibus, et de magno magnete tellure, physiologia nova_. That which Gilbert had in fact perceived, and which none before him had glimpsed even dimly, was that the globe of the earth itself acted as a great loadstone, and that the tendency of the needle to point in a polar direction was due to the globe acting as a whole. So he boldly put into his title-page the statement that his new philosophy was concerning the great magnet the earth: and in chapter after chapter he set himself to describe the experiments upon which he founded his famous induction. The phrase _terrestrial magnetism_ does not occur in any of the prior treatises, because the idea had not presented itself. Gilbert piled proof upon proof, sometimes most cogently, as when he constructed loadstone globes, or _terrellas_ to serve as magnetic models of the earth; sometimes with indifferent logic, as when he pointed to the iron ore in the earth and reasoned that the magnet tended to conform to (_i.e._ turn itself toward) the homogenic substance of the body from which it had been dug. The local deviations of the compass he sought to account for by the irregularities of the earth's crust, and maintained that the compass tended always, at places off the coast of a continent, to be deflected somewhat toward that continent. His syllogism was based on the fact that at that date all the way up the Atlantic seaboard of Europe, from Morocco to Norway, the variation was eastward. He argued that this was a universal law. But even within one generation, as may be seen in _Purchas his Pilgrims_, in the narrative of the voyage of Bylot and Baffin, the generality of the law was questioned. Gilbert reasoned on such knowledge as he had, and this did not include any notion of the secular changes in the declination. In his time, as he tells us, the variation of the compass at London was 11-1/3 degrees. What he did not know was that this was a diminishing quantity which in fifty-seven years would be reduced to zero, to be succeeded by a westward declination that would last for nearly three hundred years. For the facts as known in the thirty years succeeding Gilbert's death, see the remarkable and scarce volume of Gellibrand: _A Discourse Mathematical on the Variation of the Magneticall Needle_ (1635).