world-picture of great size and magnificence, possessing in its time considerable official importance, must not lead us to take it as an example of cartographical perfection. [Footnote 281: Cp. also the elliptical Florentine example of 1447 (Nordenskjöld, _Facsimile Atlas_, 116), or Leardus' Mappemondes of 1448 and 1452 (_ibid._ 61).] [Footnote 282: "Sinus Ethiopicus:" very similar to that depicted on the Leardus of 1448. On the southern side of this is "Fundan."] [Footnote 283: Perhaps a Ptolemaic concession.] [Footnote 284: Still more is this the case with Asia, where Fra Mauro is in some ways more satisfactory than anywhere else, and contrasts well even with the "Harleian" or Dieppe Map of _c._ 1536, and many other similar works.] * * * * * The use of the magnetic needle is essentially connected with the portolan type of map; this instrument was well known to Prince Henry's sailors, and is referred to by the Infant himself as being, like the sailing chart, a necessity for navigators.[285] But it could hardly come into general employment till men reached beyond the mediæval stage of a magnetic needle enclosed in a tube so as to float on water. [Footnote 285: _Azurara_, ch. ix.] In the Discovery of the Compass four stages may be distinguished: (1) The discovery of a species of stone with polar-magnetic qualities, _i.e._, with the power of attracting iron. (2) The discovery that steel or hardened iron could be made polar-magnetic by rubbing it with a lode-stone. (3) The discovery that the magnet (or magnetised iron) possessed the quality of definite direction, one of its poles always indicating the north, if it were so supported or suspended that it could move freely. (4) The discovery of using the magnetised iron needle as a compass. The first dates from a high antiquity, and is noticed by Plato, Theophrastus, Pliny, Ptolemy, Claudian, and many writers of the Mediæval as well as of the Classical period. The subsequent advances we cannot date, for Europe, earlier than the twelfth century; when Alexander Neckam and Guyot de Provins (_c._ 1190-1200) show us that some investigators had advanced as far as the third of the stages above recounted.