Mandeville': 'In an isle clept Crues, ben schippes withouten nayles of iren, or bonds, for the rockes of the adamandes; for they ben alle fulle there aboute in that see, that it is marveyle to spaken of. And gif a schipp passed by the marches, and hadde either iren bandes or iren nayles, anon he sholde ben perishet. For the adamande of this kinde draws the iren to him; and so wolde it draw to him the schipp, because of the iren; that he sholde never departen fro it, ne never go thens.'[1] Now it seems that accounts of the magnetic mountain have been given not only as belonging to the southern seas, but also to the north, and that men have connected with such notions the pointing of the magnetic needle, as Sir Thomas Browne says, 'ascribing thereto the cause of the needle's direction, and conceeving the effluxions from these mountains and rocks invite the lilly toward the north.'[2] On this evidence we have, I think, fair ground for supposing that hypotheses of polar magnetic mountains were first devised to explain the action of the compass, and that these gave rise to stories of such mountains exerting what would be considered their proper effect on the iron of passing ships. The argument is clenched by the consideration that Europeans, who colloquially say the needle points to the north, naturally required their loadstone mountain in high northern latitudes while on the other hand it was as natural that Orientals should place this wondrous rock in the south, for they say it is to the south that the needle points. The conception of magnetism among peoples who had not reached the idea of double polarity may be gathered from the following quaint remarks in the 17th century cyclopædia of the Chinese emperor Kang-hi. 'I now hear the Europeans say it is towards the North pole that the compass turns; the ancients said it was toward the South; which have judged most rightly? Since neither give any reason why, we come to no more with the one side than with the other. But the ancients are
