Riding the wave


The British Library ms Egerton 2803
Vesconte Maggiolo or Francesco Rosselli?

At the British Library in London there is a 16th-century manuscript nautical atlas, labelled "Ms. Egerton 2803", anonymous and undated, which seemed to have been forgotten for at least fifty years.

Recently, the Italian geographer dr Lorenzo Bagnoli, during his studies to obtain his PhD in "geo-environmental sciences and cartography" at the University of Genoa, subjected the atlas to a comprehensive analysis ( cf. L. Bagnoli, Il manoscritto Egerton 2903 della British Library e il Nuovo Mondo (The "The British Library Egerton 2803 manuscript and the New World"), which was published in Studi e ricerche di Geografia, xxv, 2002, pp. 81-110) of which we present a summary.

The atlas, which has been perfectly preserved throughout time, consists of 12 sheets of parchment of which the recto of the first and the verso of the last one are glued to the leather binding of evident Venetian origin. The subjects represented in the sheets are:

Table II. fol. 1b
Table III. fol. 2nd
Table IV. fol. 2b
Table V. fol. 3rd
Table VI. fol. 3b
Table VII. fol. 4th
Table VIII. fol. 4b
Table IX. fol. 5th
Table X. fol. 5b
Table XI. fol. 6th
Table XII. fol. 6b
Table XIII. fol. 7th
Table XIV. fol. 7b
Plate XV. fol. 8th
Plate XVI. fol. 8B
Plate XVII. fol. 9th
Plate XVIII. fol. 9b
Plate XIX. fol. 10th
Plate XX. fol. 10b
Plate XXI. fol. 11th
Plate XXII. fol. 11b
Plate XXIII. fol. 12th

Caspian Sea
Black Sea
Eastern Mediterranean
Aegean Sea
Adriatic Sea
Central Mediterranean
Western Mediterranean
Strait of Gibraltar
Bay of Biscay
Baltic Sea
British Isles
Central America
Northern coast of South America
North Atlantic
Central Atlantic
Gulf of Guinea
Western Indian Ocean
Eastern Indian Ocean
East Asian coast
Moon Phases and ecclesiastical calculations
Astronomical instrument

The size of the sheets is 20.5x28 cm and on each is drawn a panel of 18x24,5 cm, for a total of 22 charts. On the edge of the charts there is a graphic scale, of different size on each chart. The orientation of each regional map is shown on the edge of the charts with the letters and symbols that on Italian charts represent the four cardinal points, i.e. a wedge for the north, a Greek cross for the east, a capital O (Ostro) for the south and a capital P (meaning "ponente") for the west.

The atlas mentions over 4000 place-names, of which 17 are in Greek, 11 in Hebrew and 2 in Ethiopian. If the place-names in Hebrew are rare in the cartography of the time, those in Ethiopian are a feature practically limited to this atlas. They would seem, however, to have been added later, probably by a buyer of great culture. The high quantity and originality of place-names, although their transcription is not always correct and often italianized, are undoubtedly the main feature of this most interesting atlas.

None of the tables provide a date or a signature enabling unequivocally to assign the manuscript to an author, but this is not surprising. Anonymity is in fact non unusual as regards navigational charts of the sixteenth and seventeenth centuries, and actually reaches about 40%: this is explained by the presence, in most workshops, of skilled artisans and apprentices whose charts could be sold at a lower price, because lacking the signature of the owner of the map shop.

The fact that the atlas is also undated may be attributed, on the other hand, to the habit of cartographers of the time not to date their works, if not sold immediately upon completion, in order to add a last-minute date and thus make them look "recent" to later - and not very perspicacious - buyers.

The client of the Egerton 2803 seems to have been a scholar of wide culture, attracted by the expansion of geographical horizons. To substantiate this hypothesis there is the fact that the cartographer wanted to record all his knowledge with the utmost accuracy, ignoring imaginative interpretations. He only recorded what he was absolutely sure about, eventually leaving large empty spaces or dotted lines in order - perhaps - to allow the possible completion of the atlas at a later time.

The planisphere in the Egerton 2803 ms (Plate II, fol. 1b, according to the remake in:
J. Denucé, Les origines de la cartographie portugaise et les cartes des Reinel, (Ghent, 1908)
and the Carta da navigare ("nautcial chart") by Francesco Roselli, 1508, reproduced below

Regarding the authorship of the manuscript, historians of cartography have proposed the names of the Genoese Vesconte Maggiolo or the Florentine Francesco Rosselli. The first appears already in the catalog of the 1895 acquisitions but this attribution, previously favoured by some scholars, does not seem fully sustainable because, even if some similarities are evident between the anonymous atlas and some of the very many charts by Maggiolo, unambiguous identities are never traceable.

For example, comparing the representation of South America in the Egerton 2803 ms with the atlas of Vesconte Maggiolo held in Munchen (1519), the latter shows a definite improvement due to its recording the first explorations to the south: if the Maggiolo atlas reaches a latitude of 28 ° south, the anonymous atlas, at least in the regional table representing South America, stops at approximately 18-20 °.

South America in the ms Egerton 2803 (Plate XV fol.8a)

Regarding the Maggiolo Atlas of 1512 held in Parma - while comparisons between the representations of the New World in the Ms. Egerton 2803 and the Maggiolo Atlas are particularly difficult, because of the short extension of this continent in Maggiolo - it can be noted that in both charts the eastern end of the American continent is located at the same latitude as the mouth of the river Congo, that is, in its correct place. Among the arguments against ascribing the Egerton 2803 ms to Maggiolo, it can be added that Maggiolo usually signed all his works.

The attribution of the Egerton atlas to Francesco Rosselli is based instead on some common characteristics found between the world-map of Ms. Egerton 2803 and two important charts of the Florentine cartographer, i.e. the oval world map of 1507 approx., and the carta da navegare of 1508, both included in the so-called "Atlas of the Countess Bentivoglio" preserved today at Greenwich National Maritime Museum.

There are many most notable similarities between the two. For example, the western Atlantic Ocean is outlined in the shape of a bay like in Juan de la Cosa; a vast bay appears at the mouth of the Amazon and large islands are placed off the South American coast; place-names in North America, which include the names Labrador and Bacalaos although with a different spelling; and the shape of the Scandinavian peninsula.

But the Egerton atlas and the Rosselli atlas also exhibit some differences: the latter shows an Antarctic land which is not found on the Ms. Egerton 2803; the influence of Ptolemaic tradition as concerns South Asia is partially overcome on the Ms. Egerton 2803; but is still strong in Rosselli; the almond-shaped Caspian Sea in the Rosselli atlas, dose not correspond to either the world-map on pl. II. fol. 1b, or in the chart on plate. III. fol. 2a, dedicated to the representation of the Caspian sea; and finally the island Fislanda is missing in Rosselli, while it is represented in the Egerton atlas. Furthermore, the attribution to Rosselli is somewhat unsatisfactory, considering that the production of Rosselli does not include handwritten charts.

The Caspian in ms Egerton 2803 (Plate III, fol. 2a)

To determine other likely sources for the anonymous manuscript atlas, it is not helpful to analyse the tables of the Caspian Sea, of the Baltic and of Asia, which were peripheral regions of the world in early sixteenth century and rather unusual to be found in contemporary atlases.

It is even difficult to attribute the manuscript not to a specific author but, more generally, to a school map of the early sixteenth century - such as the Mediterranean tradition or the Iberianone - because the atlas has characteristics of both: on the one hand, the centrality of the Mediterranean sea, and the purpose of the Mediterranean charts, which were not necessarily intended to be used on board; on the other hand, great care is devoted to producing an updated chart, while there is a careful distinction between actually existing lands and legendary ones, which are typically Iberian features.

To date the atlas it is useful to analyze the plentiful place-names on the tables representing the New World, which is represented - in the Ms. Egerton 2803 - both on the world-map (Plate II. Fol. 7b) and on the regional maps (Plates XIV. Fol. 7b, XV. Fol. 8th, XVI. Fol. 8b, and XVII. Fol. 9a).

Central America in the Egerton MS 2803 (Plate XIV fol.7b)

In the regional tables of the Ms. Egerton 2803, the Atlantic coast of America extends southward to about 18-20 degrees latitude south, and to the north up to the latitude of Cuba, about 23 degrees N, which was the extreme point reached by Vicente Yanez Pinzon and Juan Diaz de Solis in 1509. In those charts there still is no trace of the Pacific Coast.

The overall outline of the Americas on the world map instead represents a continual continental coastline, like on the globe of 1500 by Juan de la Cosa, from 70 ° N to 50-52° S. The planisphere also represents the Pacific coast: a stretch of ink starts from about 52° south latitude, turns to the north, reaches the Central American isthmus and then proceeds to the south-west.

This brief analysis might determine that, when Ms. Egerton 2803 was completed, either the Central American isthmus - from which Balboa had first seen, in 1513, the Mar del Sur - was already known, or that the world map was subsequently corrected.

Also concerning the question of contiguity between the north American coasts and Asia, the opinions of scholars are divided. Some believe that the extent of land to the western edge of the North American map of the world suggests a connection with Asia.

According to others, the two edges of the planisphere are East Asia and America West as not consistent, while the analysis of plate. XXI. fol. 11th, representing the Asian Pacific coast, suggests otherwise. The most convincing answer seems, however, to be that the anonymous author expressely wanted to leave the drawing indefinite, so as not to exclude any solution that might have been proved true in the future.

If, on the cocntrary, we analyze the orientation of the coastline of South America, we see that in both tables of Ms. Egerton 2803 where this region is represented ( the planisphere on pl. II. Fol. 1b and the regional map on pl. XVII. Fol. 9a), it maintains the correct trend NNE / SSW.

In other words, the Egerton manuscript as well as the planispheres of King-Hamy, Pesaro and Fano, do not show what we notice in other contemporary maps, where the coastline was traced from north to south, not an accidental trick, since it implies that the newly discovered lands fall entirely within the area under Portuguese influences, and not beyond the rraya, in the area of Spanish influence.

Tables for the New World are record the voyages of Columbus, Ojeda (1499), Pinzon (1499-1500), Bastidas (1500-1502), and Solis and Pinzon (1508-1509). The date of the Solis-Pinzon voyage and the high number of place-names directly relate to it, also prove that the Ms. Egerton 2803 can not be attributed to the year 1508, as has often been suggested and as is also shown in the catalog of the British Library.

Since the voyage of Solis and Pinzon lasted from July 29, 1508 to the end of August 1509, the Egerton 2803 atlas must be dated to after that date, which indicates a sure reference. Regarding the ante quem date, the year 1513 seems to be indicated, when the Pacific Ocean was discovered by Balboa and Florida was discovered by Ponce de Leon, of which there are no traces in the atlas tables.

To learn more, contact prof. Lorenzo Bagnoli