| Egerton 2803 maps | |
| Caption1: | The Egerton 2803 world map |
| Headerstyle: | background:lightgrey |
| Header1: | General |
| Label3: | Type |
| Data3: | Portolan charts |
| Label5: | Date |
| Data5: | ca 1508 or ca 1510 |
| Label7: | Attribution |
| Data7: | Visconte Maggiolo |
| Header20: | Details |
| Label22: | Drafted |
| Label24: | Drafter |
| Label26: | Published |
| Label28: | Publisher |
| Label29: | Printer |
| Label30: | Location |
| Data30: | Egerton MS 2803, Egerton Collection, British Library |
| Label31: | Number of charts |
| Data31: | 20 |
| Label32: | Medium |
| Data32: | multichrome ink and pigment on 11 vellum folios |
| Label34: | Dimensions |
| Data34: | 11 × 8 in (30 × 20 cm) |
| Label36: | Coverage |
| Data36: | World |
| Label38: | Known for |
| Belowstyle: | background:gainsboro |
| Below: | cf |
The Egerton 2803 maps are an atlas of twenty Genoese portolan charts dated to around 1508 or 1510 and attributed to Visconte Maggiolo. The manuscript maps depict various regions of the Old and New Worlds, blending both Spanish and Portuguese cartographic knowledge. They have been noted as the earliest non-Amerindian maps of Middle America, and, jointly, as one of the oldest portolan atlases of the Americas. The maps were acquired for the Egerton Collection in 1895, published in facsimile form in 1911, and are now held by the British Library in London, England.
Very little is definitively known of the atlas's provenance, as its containing manuscript collection, Egerton MS 2803, entitled Atlas of Portolan Charts, is neither signed nor dated. The historian Edward L Stevenson suggested Maggiolo in 1508 as the atlas's possible origin, based on certain author-characterising features in several charts, and dates in the astronomical tables which follow the atlas within the collection. Prior to Stevenson, the historian Henry Harrisse had suggested 1507, while Johannes Denucé had suggested 1510, arguing that its toponyms indicate a post-Pinzon–Solis voyage composition.
The manuscript collection containing the atlas was acquired by the British Museum in 1895. Facsimile copies of all folios were first taken by Stevenson and published in 1911 by the Hispanic Society of America.
Each portolan features a central 32-wind compass rose, but its windrose network lacks the usual ring of 16 vertices, and is imprecisely drawn. Some parallels are marked. Toponyms are written in Greek, Latin, Italian. Coastlines are rather faithfully rendered for the Old World, and somewhat less accurately for the New World.
The maps are thought to depict recent discoveries from the fourth voyage of Columbus, Pinzon–Solis voyage, Vespucci voyages to South America, Corte-Real voyages to Labrador, and Gama–Cabral voyages to Africa and the Indian Ocean. The nomenclature of Central and South America, in particular, 'is infinitely richer and more complete than any other map of the Americas known to us until those of Diego Ribeiro of 1527 and 1529.' Denucé showed the maps included, without omission, all toponyms from the Pinzon–Solis voyage, the Peter Martyr map, and still 'dozens more whose precise source is unknown.'
Stevenson suggested the atlas might be 'not only the oldest known Portolan Atlas on whose charts any part of the New World is laid down, but the oldest known atlas in which the coast regions of a very large part of the entire world are represented with a fair approach to accuracy.' David W Tilton deemed it the earliest known map to 'show a coastline west of Hispaniola that is recognisable as part of Central America.' Arthur Davies concluded the atlas 'provides in its charts of the world the first complete and up to date summary of Portuguese and Spanish explorations to that time.'
Stevenson notes a 'striking resemblance' of the Indian subcontinent and Far East charts to relevant portions of the Cantino, Canerio, and Waldseemüller Carta Marina maps. Siebold notes the maps seem to imply that the Americas are joined onto Asia, which concept 'is utterly different from Portuguese cosmography and maps,' thereby suggesting 'a Spanish and not a Portuguese origin.' Simonetta Conti similarly notes, 'it is clear that they [the mapmaker] must have been very familiar with the work of the Padron Real's first authors, as can be seen from the large number of toponyms stretching from the area near Yucatan to the lands of Santa Cruz.'