Recent magnetic research has shown that while there are no magnetic mountains that would account for the declination of the compass in general, yet there are minor local variations that can only be accounted for by the presence of magnetic reefs or rocks. The reader is referred to the account of the magnetic survey of Great Britain in the _Philosophical Transactions_ (1890) by Professors Rücker and Thorpe. The well-known rocky peak the Riffelhorn above Zermatt, in Switzerland, produces distinct perturbations in the direction of the compass within half a mile of its base. Such local perturbations are regularly used in Sweden for tracing out the position of underground lodes of iron ore. See Thalén, _Sur la Recherche des Mines de Fer à l'aide de Mesures magnétiques_ (Soc. Royale des Sciences d'Upsal, 1877); or B. R. Brough, _The Use of the Magnetic Needle in exploring for Iron Ore_ (_Scientific American_, Suppl. No. 608, p. 9708, Aug. 27, 1887). Quite recently Dr. Henry Wilde, F.R.S., has endeavoured to elucidate the deviations of the compass as the result of the configurations of land and sea on the globe, by means of a model globe in which the ocean areas are covered with thin sheet iron. This apparatus Dr. Wilde calls a _Magnetarium_. See _Proc. Roy. Soc._, June, 1890, Jan., 1891, and June, 1891. {16} An actual magnetic rock exists in Scandinavia, the following account of it being given in the _Electrical Review_ of New York, May 3, 1899: "The island of Bornholm in the Baltic, which consists of a mass of magnetic iron ore, is much feared by mariners. On being sighted they discontinue steering by compass, and go instead by lighthouses. Between Bornholm and the mainland there is also a dangerous bank of rock under water. It is said that the magnetic influence of this ore bank is so powerful that a balanced magnetic needle suspended freely in a boat over the bank will take a vertical position." [29] PAGE 5, LINE 35. Page 5, line 43. _Josephus Costa._--This is unquestionably a misprint for _Acosta_ (Joseph de), the Jesuit, whose work _Historia natural y moral de las Indias_ was publisht at Seville in 1590. An Italian edition appeared at Venice in 1596. The English edition, translated by E. Grimestone, _The Naturall and Morall Historie of the East and West Indies_, was publisht in London in 1604 and 1878. There are in Gilbert's book references to two writers of the name of Costa or Costæus, Joannes Costa of Lodi, who edited Galen and Avicenna (see pp. 3 and 62), and Filippo Costa of Mantua, who wrote on antidotes and medicaments (see p. 141). The passage to which Gilbert refers is in Acosta's _Historia_ (ed. 1590, p. 64).