He discusses its observation and measurement, the influence of islands, the results obtained by travellers to distant parts, Nova Zembla, the Guinea coast, the Canary Isles, Florida, Virginia, Cape Race, and Brazil. Then he recounts his experiments with terrellas having uneven surfaces to represent the irregularities of the earth's crust. He points out errors arising from the fallacious practice of setting the needle obliquely under the card. He considers in separate chapters the variations in Nova Zembla, in the Pacific, in the Mediterranean, and in the Eastern Ocean. The fifth book is on the Dip. Gilbert seized with avidity on Norman's discovery of this effect, and devised an improved form of dipping-needle. He experimented on the dip of compass-needles placed at different points over his terrella, and evolved a theory on the proportion which he conceived to exist between the latitude and the dip. Arguing from all too imperfect data, he propounded the view that the dip was the same in any given latitude; and proposed that seamen should ascertain their latitudes by simply observing the dip. He was aware that local irregularities might occur, as they do in the declination; but was not deterred by this knowledge from propounding his theory with much circumstance and considerable geometrical skill. After the publication of his book he developed the theory still further and gave it to Blundevile for publication. At Gilbert's suggestion Briggs of Gresham College calculated out a table of dip and latitude. It was, however, soon found that the facts deviated more or less widely from the theory. Further observations in other lands showed the method to be impracticable; and Gilbert's hope to give to the mariner a magnetic measure of latitude remained unfulfilled. Book 5 closes with an eloquent passage in which Gilbert affirmed the neo-Platonic doctrine of the animate nature of the universe, and asserted that Thales was right when he held (as Aristotle relates in the _De Anima_) that the loadstone was animate, being part of and indeed the choice offspring of its animate mother the earth. Book 6, as already mentioned, is devoted to Copernican ideas, and contains Gilbert's one contribution to the science of Astronomy, in his remark that the fixed stars (previously regarded as fixed in the eighth of the celestial spheres at one common distance from the central earth) were in reality set in the heavens at various distances from the earth.