WWW-Virtual Library: History
Map History / History of Cartography: THE Gateway to the Subject

WWW-VL Main Catalogue WWW-VL
History Central Catalogue
(main menu)


What the
site is
Map History

Portolan charts from the late thirteenth century to 1500

Additions, Corrections, Updates


Volume 1. The History of Cartography (University of Chicago Press, 1987), pp. 371-463 [available online in pdf format since June 2011, with the colour plates separately here (16-17, 23-32)]

[see also the equivalent update page for the Census in Imago Mundi 38 (1986) pp. 67-94]

For the full details of the works mentioned see the Bibliography   |  Portolan Charts Main Menu

Mounted on the web 7 March 2011

This prints out to about 60 pages


'Portolan charts from the late thirteenth century to 1500' (1987)
(Additions, Corrections, Updates)

Explanatory note

These update pages are largely prompted by the major advance for the study of portolan charts represented by the publication in 2007 of Ramon J. Pujades i Bataller, Les cartes portolanes: la representació medieval d'una mar solcada [With an English version of the text entitled, 'Portolan charts: the medieval representation of a ploughed sea', pp.401-526]. (Barcelona: Institut Cartogràfic de Catalunya; Institut d'Estudis Catalans; Institut Europeu de la Mediterrània; Lunwerg, 2007).

The book is richly illustrated, with double-page reproductions of a good selection of charts, and sequences of comparative details drawn from a range of works. Better still, the accompanying DVD features almost all the charts and atlases cited, each of which can be enlarged to high resolution, a section at a time. This is the first time that the early charts have been brought together into a single visual catalogue.

Because this ground-breaking work may not be sufficiently accessible in the English-speaking world - and it is an unusually large and sumptuous volume - those who need it may find it hard to locate. It is also unfortunate that it does not have an index, since it provides a wealth of new detail, particularly about the distribution and use of early charts, and their makers, as well as a wide-ranging analysis of the factors that led to the charts' creation. One of the main purposes of the following notes is an attempt to make Pujades i Bataller's findings more accessible. This is particularly important in relation to the documentary work carried out by his countrymen - especially Gabriel Llabrés, Gabriel Llompart and Jaume Riera - which is little known in the English-speaking world.

Where I have noticed that he has added or corrected passages in my own chapter, references to his magnum opus are given below, and sometimes brief descriptions and comment. It is always the English text that is cited, except where the information, table or illustration is found only in the Catalan original. However, it must be stressed that I have been very selective in my citations from what is a large and detailed study, and I have tended to focus on those aspects that are of particular interest to me.

Pujades i Bataller is an archivist (at the Archivo de la Corona de Aragon in Barcelona) and a palaeographer. This enables him to make informed attributions of likely date and authorship, some of which overturn accepted opinion. The expertise and interests he brings to bear on the subject mean that he covers, sometimes in depth, several topics that my chapter did not address at all. These deal with important aspects of the wider historical context. Those sections are mentioned briefly here, before the page-by-page analysis of my chapter (pp.371-463).

NOTE: there is potential confusion from the fact that our two works cover a similar range of page numbers. Those cited at the head of each section refer to my 1987 'Chapter'.

In 2009, Pujades followed up his initial work with a more specific study, La carta de Gabriel de Vallseca de 1439. Again, this is in Catalan, accompanied by an English version of the text. While the first two chapters offer a general overview of material that had been covered in more detail in the 2007 work, and despite its stated concentration on the single work of one specific chartmaker, there is much new material in the last two chapters, for example on Vallseca's Majorcan antecedents and the context in which his workshop and those of his contemporaries operated.


Chapter 1 of Ramon Pujades Les cartes portolanes (2007, pp. 414-22, in the English- text version) demonstrates three pre-conditions for the appearance of the charts in the early 13th century: vernacular writing; the spread of Aragonese, Genoese and Venetian trading posts throughout the region covered by the charts, and the trading networks that connected those; and the introduction of Arabic numerals, particularly the zero, a pre-condition for effective commercial accounting.

p.425a: he makes the point that literary sources have been well used in the past but not documentary (archival) ones - a tendency his study redresses.

p.425b: he points out, correctly, that research published in minority languages, such as Catalan, has tended to be ignored - a charge I stand guilty of.

pp. 427-8: he demonstrates how silences in the records can be misleading [and elsewhere 'the silence of sources may be as significant as the information they convey through words' (p.443b)]. The majority of surviving inventories were for individuals with higher status than seamen, for example merchants. In addition, inventories of a sailor's possessions in Barcelona might not reflect what he had with him when he died abroad.

p. 428a: introducing his list of those archival documents that mention charts Pujades warns that this 'is no more than a tiny part of the probable total output of charts, because the vast majority of utilitarian works regularly used on board ship disappeared without leaving even the slightest documentary traces'.

pp.428-39: 'References to nautical cartography in medieval documentation'. This sets out the 159 instances so far identified. They are given in the original language, with explanatory comments, but are not translated. Most have already been published (with their source given) - though many are not known to historians of cartography in general - others are 'hitherto unpublished'. These are arranged geographically, as follows, with reference to the 'provenance and/or place of residence of the owner of the chart':

These inventories, the great majority from what is now Spain, are a vital resource and Pujades is owed a large debt for providing this systematic and comprehensive list of what has so far been discovered, documenting 'a total of some three hundred works of portolan cartography prior to 1500' (p.439a).

In a short but important chapter in the volume accompanying the Paris 'L'Age d'Or des Cartes Marines' exhibition of 2012, Ramon Pujades amplified the 2007 listing by drawing attention to recently discovered documents that testify further to shipboard use (2012, pp. 60-7). The first (cited in a forthcoming article by Giovanna Petti Balbi) concerned a Catalan vessel pillaged in Genoa in March 1393. Two of the sailors had their own charts and a third, a Barcelona merchant named Bernat Oliva, was carrying six or seven for sale. The same document referred to another ship, two of whose crew members also carried a chart. Put together with a reference (included in his earlier listing) to another Barcelona merchant Domingo Pujol, this demonstrates that at least 33 navigational charts (probably made in Majorca) were in the possession of just two dealers - from a city not known to be producing its own charts - in the short period of two and a half years (1389-92).

Another recently discovered reference concerns a 'mige carte de naveguar' [translated, intriguingly, as a 'half-chart' for navigation] which was owned in 1458 by a Majorcan fisherman Rafel Garbi.

A further document, dated 1405 and preserved in the Archives of the Crown of Aragon, Barcelona, records how the sailors on a fleet carrying the son of the king of Sicily, when regrouping after a storm, argued about their position in relation to that island, with reference to their charts. {The three paragraphs above added 8 February 2014}

[See further below in the section Introduction, under 'Survival' (p.373) and 'Learned' (p.374)].

Top of page

For the full details of the works mentioned see the Bibliography

Index to the reproductions of portolan charts in the two recent major works by Ramon Pujades

The combined listing that follows provides an index to the illustrations of portoan charts and atlas sheets in:

Ramon J. Pujades i Bataller, Les cartes portolanes: la representació medieval d'una mar solcada. [In Catalan, with English text 'Portolan charts: the medieval representation of a ploughed sea', 401-526]. (Barcelona: Institut Cartogràfic de Catalunya; Institut d'Estudis Catalans; Institut Europeu de la Mediterrània; Lunwerg, 2007). [referred to hereafter as LCP ]

    -- La carta de Gabriel de Vallseca de 1439 (Barcelona: Lumenartis, 2009). [In Catalan and Spanish, with English text, 274-358.] [referred to hereafter as VALL.]

Although there are photographic credits (LCP, p.399 and VALL., p.187) neither volume includes a list of plates.

Given that most of the charts are shown in their entirety, as double-page spreads (i.e. about 30 x 50 cm (LCP) and 25 x 40 cm (VALL.)), these provide a major visual resource for future historians. Almost all of those produced before 1470 are included on the invaluable DVD accompanying LCP, and those can be enlarged to more than their original size, in many cases, preserving legibility. However, the DVD's scans cannot be used for side-by-side comparison on a computer screen. Nor can they be printed out or downloaded. This makes the illustrations of 48 charts and atlases a valuable alternative for the purposes of comparative research.

The references at the end of each entry are, first, to the number given to them by Pujades and used by him on the DVD accompanying LCP and, second, to the number in the 1986 Campbell 'Census', in Imago Mundi volume 38.

The 2009 Vallseca volume includes illustrations of a number of early mappaemundi. These are not listed below, nor are the pages of comparative details, which are mostly found in LCP [for references to those see Stylistic content: 'Illustrations'].

[For an index to the Pujades numbers see at the end of this listing]

Anonymous works

The 'Maghreb Chart' [i.e. a single sheet from an atlas], first half 15th c.? (Milan, Ambrosiana, SP 2, 259). (LCP, p.297 - Pujades C 54; Campbell 89)

Fragment, showing just the rhumb-line network, first half 15th century. (Barcelona, City Archives). (LCP, p.189; VALL., p.51 - Pujades C 55; Campbell E11)

Fragment in the Archives of the Kingdom of Majorca. (LCP, p.189 - Pujades C 76; Campbell E13)

Carte Pisane. (LCP, pp.40-1 - Pujades C 1; Campbell 14)

Cortona chart. (LCP, pp.66-7 - Pujades C 2; Campbell 62)

Chart, second quarter 15th c. (Barcelona, Archives of the Crown of Aragon, MP1). (LCP, pp.298-9 - Pujades C 50; Campbell 127)

Chart, first quarter 14th c. (Florence, Riccard. 3827). (LCP, pp.104-5 - Pujades C 4; Campbell 80)

Atlas, second quarter 14th c. (Paris, BN, Lat. 4850). (LCP, pp.132-5 - Pujades A 9; Campbell 33)

Chart, second quarter 14th c. (Library of Congress). (LCP, pp.140-1 - Pujades C 10; Campbell 152)

Chart, mid-14th c. (Paris, BN, Italien 1704). (LCP, pp.146-7 - Pujades C 11; Campbell 30)

'Corbitis (formerly Combitis) Atlas' (Venice, Marciana, It.VI.213). (LCP, pp.184-5 - Pujades A 11; Campbell 117)

Chart, c.1420. (Milan, Ambrosiana, F.260.Inf). (LCP, pp.210-11 - Pujades C 31 ; Campbell 85)

Chart, second quarter 15th c. (Florence, Archives, CN11). (LCP, pp.292-3 - Pujades C 49; Campbell 72)

Signed or confidently attributed works

Beccari, Francesco. 1403 chart. (LCP, pp.190-1 - Pujades C 25; Campbell 144)

Beccari, Batista. 1426 chart. (VALL., pp.76-7 - Pujades C 36; Campbell 37)

    -- 1435 chart. (LCP, pp.252-3 - Pujades C 39; Campbell 100)

Bertran & Ripoll, 1456 chart. (LCP, pp.318-19 - Pujades C 58; Campbell 41)

Bianco, 1448 chart. (LCP, pp.276-7 - Pujades C 46; Campbell 84)

Carignano chart (destroyed), c.1327. (LCP, pp.328-9 - Pujades C 6; Campbell 65)

Cesanis, Francesco (attrib.), 'Luxoro Atlas', pre-1421. (Genoa, Bib. Berio). (LCP, pp.216-17 - Pujades A 14; Campbell 81)

    -- 1421 chart. (LCP, pp.222-3 - Pujades C 32; Campbell 119)

Cresques (workshop):
    -- Catalan Atlas, c.1375. (LCP, pp.110-11 - Pujades C 16; Campbell 28)

    -- chart, third quarter 14th c. (Venice, Marciana, It.IV.1912). (LCP, pp.152-3 - Pujades C 15; Campbell 115)

    -- chart, last quarter 14th c. (Florence, BNC, Port.22). (VALL., p.72 - Pujades C 18; Campbell 79)

    -- chart, last quarter 14th c. (Naples, BN, D 102). (LCP, pp.164-5 - Pujades C 19; Campbell 97)

    -- chart, end 14th c. (Paris, BN, AA 751). (LCP, pp.178-9 - Pujades C 22; Campbell 12)

Dulceti. 1330 chart. (LCP, pp.114-15 - Pujades C 7; Campbell 166)

    -- 1339 chart. (LCP, pp.120-1 - Pujades C 8; Campbell 13)

    -- chart attrib. to, end of first half 14th c. (LCP, pp.126-7 - Pujades C 9; Campbell 48)

Roselli, 1449 chart. (LCP, pp.286-7 - Pujades C 48; Campbell 36)

    -- 1464 chart. (LCP, pp.324-5 - Pujades C 65; Campbell 38)

Soler, Guillem. 1385 chart. (LCP pp.158-9; VALL., p.63 - Pujades C 17; Campbell 66)

Soler, Rafel, chart attrib. to, second quarter 15th c. (Paris, BN, B 8268). (LCP, pp.306-7; VALL., p.83 - Pujades C 51; Campbell 16)

    -- signed, undated chart (Berlin, Humboldt University). (LCP, pp.312-13 - Pujades C 52; Campbell 34)

Vallseca, 1439 chart. (LCP pp.264-5; VALL., pp.108-9 - Pujades C 40; Campbell 128). A separate facsimile was issued with the VALL. volume

    -- chart attrib. to, c.1440 (Florence, BNC Port.16). (LCP pp.270-1; VALL., pp.116-17 - Pujades C 41; Campbell 78)

    -- 1447 chart. (VALL., pp.110-11 - Pujades C 42; Campbell 17)

    -- chart attrib. to, c. 1447 (Paris, BN, D 3005). (VALL., p.115 - Pujades C 43; Campbell 22)

    -- 1449 chart. (LCP pp.282-3; VALL., pp.112-13 - Pujades C 47; Campbell 73)

Vesconte, Pietro. 1311 chart. (LCP, pp.72-3 - Pujades C 3; Campbell 64)

    -- 1313 atlas. (LCP, pp.80-3 - Pujades A 1; Campbell 25)

    -- 1318 atlas (Venice, Correr). (LCP, pp.92-5 - Pujades A 2; Campbell 120)

    -- c.1321 atlas (Lyons) - the sheet with just the rhumb-line network. (LCP p.322; VALL. p.50 - Pujades A 5; Campbell 5)

Viladesters, Macià de. 1413 chart. (LCP, pp.202-3 - Pujades C 30; Campbell 11)

    -- 1423 chart. (LCP, pp.238-9 - Pujades C 34; Campbell 75)

Viladesters, Joan de. 1428 chart. (VALL., pp.78-81 - Pujades C 38; Campbell 133). Note: it was not possible to include a scan of this on the LCP's DVD.

Virga. 1409 chart. (LCP, pp.196-7 - Pujades C 27; Campbell 23)

Ziroldi (Giroldi). 1426 atlas. (LCP, pp.244-7 - Pujades A 16; Campbell 116)

By Pujades chart or atlas number

LCP = Les cartes portolanes: la representació medieval d'una mar solcada (2007)
VALL. = La carta de Gabriel de Vallseca de 1439 (2009)

C 1 - LCP, pp.40-1
C 2 - LCP, pp.66-7
C 3 - LCP, pp.72-3
C 4 - LCP, pp.104-5
C 6 - LCP, pp.328-9
C 7 - LCP, pp.114-15
C 8 - LCP, pp.120-1
C 9 - LCP, pp.126-7
C 10 - LCP, pp.140-1
C 11 - LCP, pp.146-7
C 15 - LCP, pp.152-3
C 16 - LCP, pp.110-11
C 17 - LCP pp.158-9; VALL., p.63
C 18 - VALL., p.72
C 19 - LCP, pp.164-5
C 22 - LCP, pp.178-9
C 25 - LCP, pp.190-1
C 27 - LCP, pp.196-7
C 30 - LCP, pp.202-3
C 31 - LCP, pp.210-11
C 32 - LCP, pp.222-3
C 34 - LCP, pp.238-9
C 36 - VALL., pp.76-7
C 38 - VALL., pp.78-81
C 39 - LCP, pp.252-3
C 40 - LCP pp.264-5; VALL., pp.108-9
C 41 - LCP pp.270-1; VALL., pp.116-17
C 42 - VALL., pp.110-11
C 43 - VALL., p.115
C 46 - LCP, pp.276-7
C 47 - LCP pp.282-3; VALL., pp.112-13
C 48 - LCP, pp.286-7
C 49 - LCP, pp.292-3
C 50 - LCP, pp.298-9
C 51 - LCP, pp.306-7; VALL., p.83
C 52 - LCP, pp.312-13
C 54 - LCP, p.297
C 55 - LCP, p.189; VALL., p.51
C 58 - LCP, pp.318-19
C 65 - LCP, pp.324-5
C 76 - LCP, p.189
A 1 - LCP, pp.80-3
A 2 - LCP, pp.92-5
A 5 - LCP p.322; VALL. p.50
A 9 - LCP, pp.132-5
A 11 - LCP, pp.184-5
A 14 - LCP, pp.216-17
A 16 - LCP, pp.244-7

Top of page

For the full details of the works mentioned see the Bibliography

'Portolan charts from the late thirteenth century to 1500' (Additions, Corrections, Updates)

Introduction (pp.371-80)

[Note. The entire text of volume 1 of The History of Cartography is now freely available, in searchable form, on the website of the University of Chicago Press, with the colour plates separately here (16-17, 23-32).]

Terminal date for the study (p.371)
Note 1 explains the reasoning behind the terminal date of 1500. Pujades (2007) p.412b likewise extended his study to 1500 but reduced his atlas and chart listing to 1469 (thus reducing the Campbell total of 180 to 120).

The distinctiveness of portolan charts (pp.371-3)
Pujades (2007), p.509b sums up (citing Pascal Arnaud (1996) p.137): 'The new cartography, instead of regarding the sea as an informatively useless space and, by extension, radically simplifiable in its geomorphic traits and reducible in its dimensions (as preceding erudite cartography had done), assigned it the leading role, to the point where it disregarded everything that was not the sea itself and the coastlines that delimited it'.

Survival and lost charts

Survival (pp.373-5)
Chapter 2 in Pujades (2007), 'Survival and possession: the problem of representativity of the preserved works' (pp.423-51) contains much relevant evidence. The 159 inventories he describes referred to 300 'works of portolan cartography prior to 1500' (p.439a). On the next page he noted that the 'majority of these maps lacked any kind of ornamentation'. From this it follows that 'most of the works which have survived to the present day are representative not of standard medieval cartography but rather of sumptuous exceptions to the rule' (p.440a). 'Atlases are only rarely mentioned' - a striking contrast to the fact that one-third of the surviving works are in bound form (p.423b).

An article (The Guardian, 28 August 2009, 'What lies beneath'), about a buried hoard, made two points with interesting parallels for the study of portolan charts. One was that 'gold does not map settlements - high-status coins could be hidden or lost anywhere. But where you have got 100 grots, you have a settlement' (Sam Moorhead, British Museum). The other comment was about the debased coinage turned out in many millions by the Emperor Tetricus 1, of which only four examples survive. If we are interested, as historians, in the charts that were used, those that sustained generations of chartmakers, we must turn our attention away from the sumptuous to the quotidien. On the question of unbalanced survival see further below in the section Chart trade, under 'Production numbers'.

'No extant true book-format Catalan atlases exist, whereas mappaemundi in panels do. Inventories confirm that the genuine atlas never circulated far beyond Venice, which explains why the vast majority of medieval exemplars that have come down to us are Venetian' (Pujades 2007, p.441b). Specific instances of non-Venetian atlases are the early Genoese production (Paris, Lat. 4850, Pujades A 9), the Medici Atlas and the Briaticho atlas of 1430.

As regards survival rates, the few atlases mentioned in the inventories (perhaps 25 compared to 220 charts) were in the libraries of the better off. None was found in a sailor's chest (again, p.441b). This of course confirms the general rule that maps in bound form survive in much larger numbers than those left unprotected. Elsewhere (p.521b) Pujades asserts 'by virtue both of the meticulous way they were made and of the references to their owners, that the purpose of a highly substantial number of atlases produced in Venice between 1400 and 1470 was to serve the intellectual pursuits of proto-renaissance university scholars'.

A useful parallel is provided by the charts prepared for the Dutch East India Company. 'Of the 20,000 charts estimated to have been produced for the VOC in the seventeenth century, and conceivably more than 55,000 for its entire life, some 350 chart that survive are recorded here'. (Review by Sarah Tyacke, in Imago Mundi 63:2 (2001), p.230 (?), of Günter Schilder & Hans Kok, Sailing for the East: History and Catalogue of Manuscript charts on Vellum of the Dutch East India Company (VOC) 1602-1799 (Houten: HES & De Graaf, 2010). That works out at a survival rate of between 0.7 and 1.75%, for the products of a highly centralised organisation that only loaned out its charts, and for a much later period. This supports very low estimates for the early portolan charts.{ This paragraph added 29 April 2011}

[See further below in the section The function of portolan charts]

Lost charts - additional contemporary references, inventories, etc. (p.374)
Considerable work has been done, since 1987, on inventories both of seamen and the learned, to give a better idea of how many charts might have been in circulation:

  • Seamen. Pujades (2007) p.425b explains how ship's inventories (which have been well studied) tend to be misleading because they concern themselves with the ship-owner's property. Charts were usually privately owned, thus their absence from such documents has incorrectly been understood to indicate their absence from the ship. The post- mortem inventory and subsequent public auction of a seaman's effects (found in their caixes de navegar or sailing chests), however, give a different and more relevant picture of chart frequency. [See further below in the section Shipboard use]
  • Merchants. The inventory of the Venetian merchant Marco Bembo, who died in 1445, includes a compass and a sailing chart - which were sold after his death - and a book on navigation, see Susan Connell, 'Books and their Owners in Venice 1345-1480', Journal of the Warburg and Courtauld Institutes 35 (1972): 163-186, especially p.172 [I owe this reference to Catherine Delano-Smith].
              The 1483 inventory of Cardinal Gonzaga included (766) 'La carta da navigare', which was noted as 'probably a portolan map' - see D.S. Chambers, A Renaissance cardinal and his worldly goods: the will and inventory of Francesco Gonzaga (1444-1483) (London: Warburg Institute, 1992), p.172 [I owe this reference to Catherine Delano-Smith].
  • Pilgrims. For an analysis of pilgrim routes and numbers, 11th-15th centuries, see Jacoby (2014). Tucci (1990) pp.15-16 cites the comment of the Irish friar in 1323, when at Crete on his way to the Holy Land. He remarked that the island had a perimeter of 500 miles 'secundum marinarios insulas maris describentes'. See, for pilgrim accounts in general, Hyde (1978). Vagnon (2007b), pp.306-7 and the diagram on p.301, identifies pilgrimage references on four Catalan charts in the Bibliothèque nationale de France. On pilgrims see also:
      Delano-Smith (2004)
      Deluz (1987) Kozlicic (2000)
      Régnier-Bohler (1997)
      Russo (2014)
    See in addition below, under Shipboard experiences of travellers and pilgrims.
  • Learned. See Rodney Thomson, 'Medieval maps at Merton College, Oxford', Imago Mundi 61:1 (2009): 84-90. Page 86 refers to a 1452 list of books of 'philosophy', including two cartae maritimae. One of these was donated in 1374 by William Reed, a former fellow and at that time bishop of Chichester. The author surmises it was a portolan chart evidently from the same period as the source of the Aslake Map (see Barber & Brown, 1992).
              On Dante Alighieri (c. 1265–1321) see Osvaldo Baldacci, Dante lettore di geocarte e portolani (Rome: Accademia nazionale dei Lincei, 2001).
  • Princely collections. See Paviot (2003) for French inventory references to charts.
  • Atlantic and Baltic charts. See Gautier Dalché (2015, p.154) for three examples (late 14th to 15th centuries).
Early Portuguese charts
p.374, 2nd paragraph. Cortesão's claim that there were Portuguese charts before 1400 is refuted by Alfredo Pinheiro Marques, who considers a chart commissioned by Prince Henry in 1443 as the first known to have been produced in Portugal. A large 'map' (14 spans in width = c.280cm?] was seen in Lisbon by H. Münzer in 1494 - see Pinheiro Marques (1987b), 1:29, 40, 41.

Journal of the Institute of Navigation (confused titles)
p.375, note 39. Michael Richey pointed out (private communication 4 March 1989) that I had confused two journals: (1) Navigation: Journal of the Institute of Navigation [American] and (2) Journal of the Institute of Navigation (now Journal of Navigation) [British]. The mistake is repeated on p. 383, note 110.


'Portolan' or 'Portulan' (p.375a)
The strictly incorrect term 'portolan chart' continues to be used by most commentators, including, for convenience, myself. Gautier Dalché (2001), p.10 argues against the usage. As an aside, it is relevant to ask why the term 'portulan chart', rather than 'portolan chart', has been used instead - from the 18th century it seems - particularly in studies written in French, German and Portuguese. Nobody, as far as I know, refers to the related texts as 'portulani'. Contradicting the foregoing, I have now noticed an even earlier instance: the work published in Naples in 1705 by Sebastiano Gorgoglione is entitled Portulano del mare Mediterraneo {Sentence added 14 December 2014; original cited publication date changed from 1726, which proves to be a later edition (12 November 2015)}. That usage can now be pushed back considerably earlier, to a MS described by Kretschmer (1909, p.231) as clearly dated from Civitavechia in 1616, 'Portulano nel quale si descrivono tutte le citta, terre, castelli e porti marittimi di tutta Europa' (Venice, Biblioteca Querini-Stampalia, Cl.III, No.16) {Sentence added 11 November 2015}.

The term 'portolan' was used by Pierre du Val in 1684 on his 'L'Afrique Où sont exactement decrites toutes les Costes de la Mer, suivant les Routiers et Portolans de divers Pilotes' (British Library, Maps 177.b.1.(2.)). Wallis & Robinson (1987) p.12, trace the use of the term 'portolan chart' to the 1890s; Kiss (1988) discusses the term's origin (in Hungarian).

It seems possible that the English-language use of 'portulan' usage can be traced back to the National Maritime Museum, Greenwich, perhaps to D.W. Waters's influential The Art of navigation in England in Elizabethan and early Stuart times (1958). However, the celebrated British Admiralty surveyor, Captain W.H. Smyth, included the term 'portulani' in his The Mediterranean: a Memoir Physical Historical and Nautical (1854, reprinted 2000, p.333). Nevertheless, he used 'portolani' more frequently, on that and the preceding pages, in both cases referring to charts not texts, so this cannot be taken as clear evidence of the settled English use of the 'u' form. {Those two sentences added 18 February 2014}

Call them by some other term if you will, but, if you are following traditional usage, please let it be 'portolan chart' in English!

Contemporary terms for portolan charts (p.375a)
On contemporary terms for what were, or might have been, portolan charts, see Gautier Dalché (2004).

Pujades (2007), p.440a/b lists the various terms used in medieval Catalan nomenclature, interpreting the differences to their inclusion of features that gave added commercial value in the eyes of the notaries. He could find no use of the term 'carta' (p.108) before 1311 (leaving aside the debatable Lull reference of c. 1295). Pujades further deduces 'that specific nomenclature for nautical charts had yet to become widespread and, consequently, that the origins of the new type would have been fairly recent' (p.440b) [contrast that with his support for an origin around 1200, see below in the section 'Liber de existencia riveriarum'.] He also found that inventories sometimes treated the terms 'mappamundi' and 'carta de navegar' as synonymous, which he interpreted as descriptions of portolan charts enriched with inland detail (p.441a).

Characteristics & definition

pp.376-80. See a variant definition for portolan chart given in Pujades (2007) p.411b.

Stefano Bifolco, Mare Nostrum: Cartografia nautica a stampa del Mar Mediterraneo (2020) resorted to defining nautical chart as one having a network of rhumb lines

Vellum (p.376a)
For descriptions of the 'long and complex [production] process that had not changed over the centuries' see the references in Astengo (2007(a)) p.182, note 53.

Rhumb lines ('wind network')
p.376b and note 53. On 'rhumb lines' [another much-used but technically incorrect term] see Wallis & Robinson (1987), pp.201-2, citing Pedro Nuñes as the originator of the term (by at least 1537). See, for a general mathematical analysis, Alexander (2004).
         For a claim that Catalan chartmakers were using a two-circle system as early as 1313 see Gautier Dalché (2006). The latest recorded use of the double network can be seen on the 1447 Ziroldi chart in the Hispanic Society (Pujades C 45) or the unsigned chart in Barcelona (C 50) - see Billion 2013, p.330.

Local scales
p.377a (end of 2nd paragraph). "Unfortunately, no key to the unit of measurement was supplied...". However, the chart sold at Christie's on 21 June 1989 (E.19), and perhaps datable to the second half of the 15th century because of its 32 rhumb lines (see p.396), has, in their words, the 'apparent indication of local scales'. This unusual production consists of six charts on a single sheet, with scales differently identified, e.g. mila davenezia for the Adriatic and mila de larzipelago for the Greek Archipelago. 'We can trace no other chart with local scales such as these, which provide the navigator with specific local information.' This invites further questions, though. If there were indeed local scales in operation throughout the Mediterranean region, how would their respective users have applied them to portolan charts and, indeed, to portolan atlases, which, it is generally agreed, employed a consistent scale [though see further in the section Black Sea, below]?

Debanne (2009 pp.52-3) apparently includes a comparison of units of measure ('millara') in the portolani.

[On scales, see further below in the section Copying & Scales (p.391)].

Orientation of the place-names
p.377b. "The place-names were written inland, at right angles to the shore". An exception (unique?) to that rule can be seen in the unsigned atlas from the early 15th century in the Bodleian Library, Oxford (Pujades A 13). Instead of letting the toponyms' orientation change gradually to match that of the coastline, the scribe has attempted to maintain a constant orientation for as long as possible. See, for example, Sicily on sheet 4, where this awkward arrangement gives a stiff appearance. {This paragraph added October 2013}

Cross on Khios/Chios
p.378, and note 66. My statement that the occasional convention of placing the Genoese cross over Chios "continued until at least the end of the sixteenth century" referred to a revival, not a continuation. For a further comment on the Genoese cross over Chios see The features chosen to illustrate the developing portrayal of islands and estuaries on portolan charts up to 1469.

Opicino de Canistris
p.379, note 71. See Whittington (2014), Morse (2007 and 1996).

Top of page

For the full details of the works mentioned see the Bibliography

The origin and compilation of the portolan charts (pp.380-90)

The main recent sources are Gautier Dalché (1995) and Pujades (2007), pp.506-20. One of Pujades's chapters is wholly concerned with this issue, to which 'as yet, nobody has come up with a satisfactory answer' and to which he 'can produce no irrefutable evidence in support of my theories' (p.506a).

'Mediterranean portolan charts: their origin in the mental maps of medieval sailors, their function and their early development' (an extended essay by Tony Campbell, January 2021)

Ancient or other early origin

Classical origin (p.380)
Pujades (2007) pp.506-7 is dismissive of this theory, arguing that neither the Greek tradition based on geographical coordinates, nor the classical periploi (written sailing directions offering estimated distances and simplified 'wind' directions) could have produced the more sophisticated charts with their distinctive compass deviation (p.507a).

Arabic origin and Arabic charts (p.381a)
The main authority is Soucek (1992). His listing of 'Islamic Maritime Charts' (p.288-) confirms the existence of just three pre-1501 Arabic charts (Census 89, 136, 132), but notes that 'no examples of Ottoman Turkish portolans earlier than 906/1500 are known to have existed'. He cautions against the two extremes: viewing Arabic charts as entirely derived from Western models, or stressing an independent Islamic tradition. 'All we can say with certainty is that chartmakers of North Africa and Ottoman Turkey worked in relative independence of each other, even if their maps were derived from or influenced by the same European sources' (p.263a).

Ducène supports Soucek's position, noting the equivalent of the terms 'compass' and 'rose of the winds' in the earliest identified Arabic reference to a chart (the encyclopedia of the Egyptian, Al-'Umarī, compiled 1330-48). However that describes a chart with what may have been Arabic place-names, which hints at an Islamic contribution that has yet to be fully understood (Ducène, 2013, p.86). {This paragraph added August 2013}

Pujades (2007) pp.508-9 examines the question of possible Arabic antecedence, and firmly dismisses it. He studied the Maghreb chart (Census 89), concluding, on various grounds, that it was copied from a Christian model 'from the first quarter of the fifteenth century' (p.508b) [compare with my toponymic analysis that found no sign of names added after 1339 (Chapter p.416 'Arabic, No.1')]. He also cites (p.509a) late-15th-century documents showing that western charts were being imported into the Maghreb. His note 39 (p.519) contrasts the limited influence of Arabic geographical texts on western thinking compared with those in the fields of medicine and astronomy.

A message to MapHist in April 2009 (see the subject 'islamic roots of European maps') referred to suggestions (claims?) by Fuat Sezgin, in, apparently, volumes 10-12 of his Mathematische Geographie und Kartographie im Islam und ihr Fortleben im Abendland (all published in 2000), 'that the sudden appearance of portolan maps in Europe, circa 1300, is just a continuation of the then centuries old cartographic tradition in the islamic world' (Eric Kirchner). The exchange did not uncover any discussion of that theory by later scholars.

At a London workshop in September 2013 Robin Seignobos pointed out that the 11th-century 'Book of Curiosities' coincided with Sezgin's missing link but was quite different from the portolan charts. However, the abstract of Stefan Schröder's paper at ICHC 2013 in Helsinki (p.81), '"Transitional" or "Transcultural" maps? Latin mapmakers and Arabic knowledge in the first half of the 14th century', identified an Arabic origin in the Mountains of the Moon and the Caspian Sea on the early 14th-century work of Marino Sanudo and others. {This section added 7 October 2013}

Recent Literature on Arabic, Islamic & Ottoman charts, and possible Asian origin:

    Arioli (2015)
    Beshevliev (1988)
    Brentjes (2008)
    Comes Maymó (various)
    Cousins (1984)
    Ducène (2013)
    Herrera Casais (various)
    Historical Atlas of Islam (2001)
    Kahloui (2008, 2017)
    Ledger (2016) - particularly on the Maghreb Chart
    Leitner (1982)
    Lepore, et al. (2011- )
    Nemlioğlu (2014)
    Özdemir (1992)
    Pinna (1996)
    Pinto (2016)
    Rapoport & Savage-Smith (2018)
    Romero-Girón Deleito (2019)
    Savadi & Campbell (2023)
    Sezgin (various)
    Soucek (1992)
    Tibbetts (1992)
    Vernet-Ginés (1962)

See also the separate notes on the three surviving pre-1501 Arabic charts in the update page to the 'Census' (which includes further references):

  • Maghreb Chart (Census 89)
  • Ibrahim al-Tabib al-Mursi (Census 132)
  • Ahmad al-Tanji [formerly known as Kâtibî] (Census 136)
Herrera Casais (2009) p.231 notes that the al-Mursi chart is largely based on a work similar to the Bertran & Ripoll chart of 1456 - the first time a Majorcan model for an Arabic chart has been identified. See also the website of the 2009-2011 project in which she is involved, at the Universitat de Barcelona, Cartografía náutica árabe en el contexto mediterráneo (c.1300-1600). Estudio de detalle de las cartas.

Indian Ocean 'charts' shown to Marco Polo (p.381a)
Tibbetts (1992) concludes (p.262b) 'All in all, the evidence for indigenous maps (pre-1501) in the Indian Ocean that could be used for practical navigation seems to be entirely negative. The so-called charts shown to the Portuguese were much more likely to have been literary compositions similar to those that survive today in Arabic manuscript atlases and geographical works'.

Byzantine origin (p.381b)
Pujades (2007) p.507-8 argues that those who see the Byzantine Empire as the route between the classical and medieval worlds would have to explain why there are eight surviving periploi from Late Antiquity and none from the centuries afterwards. As far as Greek charts are concerned (the expected successors of any Byzantine antecedents) the earliest dates from 1459. Likewise, Tolias (1999) makes no claim in this respect.

Chinese origin
p.381, note 93. For comment on 15th-century Chinese maps produced in connection with Zheng He's expeditions, see a message from Philippe Forêt to MapHist on 10 April 2004 [and other messages in that same thread: 'Waldseemuller 1507 map'].

Top of page

For the full details of the works mentioned see the Bibliography


pp. 381-4. Pujades (2007) devotes the bulk of his final chapter (pp.506-20) to arguments in favour of the charts' medieval origin, centred around 'the fact that they were the products of the written culture of a specific period', and that a historical solution must be sought to what others have tended to treat as a purely geographical-cartographical problem' (p. 506a). He is not afraid to express firm opinions about almost every aspect of this highly controversial subject. Several of the explanations he offers are necessarily hypothetical, and sometimes are themselves built on an earlier hypothesis.
      Pujades (2007) pp.418a, 419b & 420 [talking of (1) numeracy (for business accounting), (2) literacy and (3) the use of written vernacular, states that:] 'throughout the second half of the twelfth and during the initial years of the thirteenth centuries, in the north of Italy the cultural transformations had taken place necessary for the generation and diffusion of a new type of graphic document, namely the portolan chart'; 'such a valuable instrument of written culture as a nautical chart, which would have been practically useless in the hands of an illiterate'; 'the chronology of the advent and dissemination of nautical charts is totally inseparable from that of the rise of vernacular Romance languages to the category of vehicles of written expression'.

Elsewhere, he emphasised other contributions of the 12th century, namely the increase in voyages stimulated by the establishment of merchants' colonies, and the dissemination of the compass and hourglass that permitted more accurate measurement of the direction and distance sailed. He states that 'the question of the wind network thus becomes a fundamental aspect as regards the issue of the origins of nautical charts' (p.511a). [On what Pujades considers to be an important point, namely the adoption of decimal sub-divisions in the charts' scale, see below in the section Stylistic content, under 'Graphic scale bars' (p.393)].

Traditionally, the supposed 'mother' chart was looked for the in period shortly before 1270 (the earliest documented evidence). However, the identification of the 'Liber de existencia riveriarum' (supposed to date from around the year 1200), with its apparent reference to a portolan chart-like object, has fundamentally changed the nature of the debate.


In what amounts to a review of Pujades (2007a) Patrick Gautier Dalché (2012) challenges the Pujades criticisms
of his own interpretation of the 'Liber' and, responding with detailed criticisms of his own, restates his position
{note added September 2013}

The discovery of the 'Liber de existencia riveriarum et forma maris nostri Mediterranei' (Gautier Dalché, 1995) is the most significant development in portolan chart studies over the past two decades. However, its interpretation and significance are not straightforward. Morse (2007) p.37 describes the Liber as 'the first known work to be based in part on what may have been a portolan chart'. Pujades (2007) considers that 'even though there are those who would see in it a more or less remote forerunner to navigational charts rather than, as Gautier Dalché does, a genuine portolan chart - [it] allows us to date the origins of nautical cartography back to the beginning of the thirteenth century' (p.512a). [See also my comments to the Gautier Dalché entries in the Bibliography.]

In other words, Pujades frames his 'Birth' chapter around the assumption that the charts' origin dates from about 60-70 years earlier than any other evidence yet identified. What he refers to as the 'sudden emergence of nautical cartography' (p.515a - my italics) would then have been followed by a long period of silence. This, surely, is the fundamental challenge facing proponents of an early 13th-century origin.

Pujades (2007) note 76 (p. 519) suggests a subtle modification to Gautier Dalché's dating of the Liber (late 12th or early 13th century), arguing instead that the Black Sea details indicate a date after Constantinople fell to the Venetians in 1204 - and, presumably, not immediately after that date. Jacoby (2012, p.72) comments independently that the inclusion of the Black Sea details suggests a likely date in 'the first three decades of the thirteenth century'. This indicates the need for a minor revision to Gautier Dalché's suggestion that the toponymy points to the late 12th century.

Gautier-Dalché himself had suggested that the surviving version of the Liber might date from the first decades of the 13th century rather than 1200. In personal communications (11 August and 9 September 2009) Pujades endorses that, contending that the Liber's compilation and the copying of its extant example probably occurred during the first and third decades of the 13th century respectively. If that dating amendment is accepted it helps to narrow the gap between the writing of the Liber, on the one hand, and the 1270 reference and the Carte Pisane on the other. [Though the claim in Pujades, 2013(b) that the Carte Pisane and Cortona chart should be dated considerably later leaves an enlarged gap between 1270 and 1311, the earliest reliable date for a surviving chart.] The interval might represent not much more than two generations, rather than three. Whatever the time-span, there is (as yet) no physical or documentary evidence of any kind from that intervening period.

Pujades (2007) p.512a notes that the Liber's directions are given in terms of the 16-wind system. Yet in the passage he translates (p.513a) it is the eight winds that are listed, although the 16 winds are indirectly achieved by directions that specify 'between' two winds (p.513b).

Pujades (2007) pp.512-13 discusses the Liber in some detail. In particular, he takes one section of the Prologue and disputes the amendments that Gautier Dalché has made to the text, alterations that radically modify its meaning. He gives both the original Latin and, in this English section of the work, the English translation. These linguistic differences (on which I am not qualified to comment) are certainly significant. 'If we take it literally, the text does not say that in order to make the portolan the author first had to draw a mappamundi, but quite the contrary, that is, that the text had been written as a preliminary source of information on the basis of which to design the Mediterranean and its coasts on a mappamundi' (p.513a).

In Pujades's view, the 'chicken and egg' argument can be ignored. 'The question of whether the final version of the Liber was written before or after the drawing of a nautical mappamundi ["cartula mappe mundi"] is of minor importance. What really matters is that in the first mention of a more or less direct prototype of what would be the nautical charts of the later thirteenth century, both the chart and the portolan appear together and strictly interrelated, which means that both were fruit of the same source of information, the knowledge accumulated through the navigational experience of seamen of the time' (p.513b).

However, the English translation of the Catalan version supplied by Pujades (pp.513a and 315 respectively) includes the following passage about the author's navigational knowledge: 'Reasonably represented are the length, width and the short distance between its two shoreline sides, those of Libya and Europe, in accordance with what I managed to discover and calculate from information provided by ship's officers and their portolans ['gradiens'] and from what I saw myself on my own voyages' [my italics]. The word 'gradiens', previously unknown, has been interpreted by Gautier Dalché (p.81) as referring to a portolan. Pujades's gloss on this is that such a meaning 'would confirm, on the one hand, the existence of portolans during the transition period from the twelfth to the thirteenth centuries and, on the other, that the information the author had compiled to write his portolan and draw his mappamundi came from data on distances and directions amassed by a number of seamen-merchants-military men from the second half of the twelfth century who had recently learnt to read and write' (p.513b).

Pujades (2007) p.516a takes issue with Gautier Dalché about the identity of the Liber's author, seeing him as a cleric rather than a merchant or seaman. Certainly one of the author's non-nautical interests was in identifying Biblical locations. Specifically (p.516b) Pujades suggests that 'in all probability, the author of the Liber was one of the numerous clerics who acted as notaries and chaplains on the naval convoys and as urban notaries'.

Pujades himself stresses the 'archaic nature' of the earliest surviving charts and the remarkable hydrographic refinements that can be discerned over the course of no more than one generation thereafter, as seen in the work of several different practitioners. If we are to speculate about this long gap of silence - and Pujades's claim to have provided 'answers' is too optimistic a term - what do we imagine was happening between, say, 1210 and 1270? Were there rival chartmakers or just a single, two or three-generational family? Were they supplying generally the charts that presumably were becoming indispensable - otherwise why were they being made? - or selling to just a few seafaring customers? And were these proto-charts recognisably portolan charts, and would they have functioned in the same way?

Against my sceptical comments, Pujades (personal communications during 2009) has offered various observations to corroborate an origin date for the portolan charts in the first decades of the 13th century. Elaborating on the vital importance of the use of Arabic numerals - and most of the Liber's distances end with a 5 or zero - he points out that, since the work's author was acting on a request from a cleric belonging to a church in Pisa (p.513a), he would probably have had speedy access to the original Latin text of 1202 by Leonardo Fibonacci, a fellow Pisan. Merchants and others would presumably have had to wait until late in the century for the vernacular version.

Nor, Pujades contends, should we expect documentary evidence before the late 13th century. The literary sources [as cited by himself], which provide our few glimpses into the early history of the charts, are mostly written in the vernacular whose use was becoming widespread only during the latter part of the 13th century. Notarial documents (the other major source) had a parallel development.

Finally Pujades points out, against the generally agreed context that only a minute fraction of the charts produced have survived, that by about 1300 charts were documented as being on Genoese, Catalan and Sicilian vessels [2007, p. 438, no.132, & p.444]. By that time we have just the Carte Pisane and perhaps the Cortona chart as physical evidence.

In my chapter I did not attempt answers where I felt none was justified, at least until further information came to light. I am fully aware that I cannot match the scholarship displayed by Patrick Gautier Dalché and Ramon Pujades, and would not try to do so. But, once hard evidence has been exhausted, all are equally free to speculate. As I stated before (Chapter p.390a) the vibrant and short-lived development discernible in the earliest surviving charts points logically to a relatively recent prototype. The main 'mistakes' on the Carte Pisane, the distorted shape given to Italy and the squeezing of south-west Europe, are significant, but they concerned landmasses and, only to a lesser extent, the seas to either side of them. But, if you look at most of the coastlines on the Carte Pisane, they are remarkably accurate and hence did not need later improvement. That the Carte Pisane might be considered a crude prototype, with uneven writing suggesting the work of an amateur, does not seem to me to devalue its unique worth as the earliest visual record we have. Neither the writing nor the misshapen Italy need have affected the practical usefulness of that chart. The original, not surprisingly, is stained and damaged [did that come from use at sea?]. What we need is a generally accessible redrawing of its coastlines (with any distortion of its vellum corrected with reference to its two hidden circles). That would make immediately clear its outstanding hydrographic merits.

A glimpse of what might mirror, in some respects, the original processes of compilation and subsequent refinement (of which we have no evidence) can perhaps be seen in the way that the Atlantic coasts, and particularly those of the British Isles, were incorporated into the standard portolan chart outline in the early 14th century. What appears on the Carte Pisane, the early Genoese chart in the Riccardiana Library in Florence and, to a lesser extent, the recently discovered Lucca chart (if its claimed early date is corroborated), are pre-surveyed coastal outlines for southern England that must have depended on verbal accounts. By 1313, those theoretical outlines were already being superseded by Vesconte's broadly recognisable picture of a south and south-east coast for England, which was becoming known in some detail. It would be natural to attribute those developments to trading contacts but on that see Chapter pp.407-8.

A steady succession of further improvements followed in the years up to about 1330. This saw the extension of knowledge to Ireland and further north up the British coasts, based now on actual survey. By the time of the late Vesconte and 1330 Dalorto charts, the British Isles had achieved outlines that so adequately served their users' purpose that they would not be significantly improved for centuries. Of course the Mediterranean coastlines would have been generally known before they were charted, and so the parallel is not exact. But transferring practical experience to a chart on the basis of compass directions and estimated distances is likely to have led to significant initial distortions, particularly in terms of long-distance voyages. Identifying and correcting such mistakes would have needed greater skill than that required for a coastal survey of the British islands.

If about twenty years was sufficient to provide broadly realistic outlines for the Atlantic coasts of S.W. Europe and the British Isles, is it likely that it would have taken much of the 13th century to achieve the same for the Mediterranean and Black Sea basins?

If the Liber's supposed chart of 1190, or 1210, had been constructed on the basis of compass bearings would it have looked similar to the Carte Pisane? If yes, why had a century of development not improved it further? If, on the other hand, that hypothetical early prototype was as primitive as an extended period of pre-Pisane development might imply, how could it have been of sufficient use to keep an essentially practical navigational tradition alive through that gap of silence? Pujades's paragraph (p.521b) describing how, after the first (hypothetical) prototypes, 'a rapid process began of their optimisation and of enrichment of their geographical designs and their lists of coastal toponyms', is not justified in the state of present knowledge.

For an analysis of the names on the 'Liber' see the author's extended essay on the Carte Pisane (March 2015)

In a paper delivered at ICHC 2017 in Brazil (and made available online via Academia.edu) Joaquim Alves Gaspar argued, on the basis that the Liber's pelagic courses were taken from a chart, that those "could not have been measured with a marine compass because they are not affected by magnetic declination." From that it followed that "a primitive type of portolan chart, not based on compass directions, was being developed at about 1200." See the fuller elaboration of this argument in his Imago Mundi article in issue 71:1 (2019).

Finally, it is appropriate to mention here the growing body of palaeomagnetic data and its possible evidential value for the question of portolan chart origin. Since this was not corrected, the long understood tilt of the Mediterranean axis is assumed to represent the situation at the time of the construction of the prototype chart. On the basis of data gathered by Korte & Constable (2005), Joaquim Alves Gaspar (personal communication, and see his 2008a, pp.200-01) has pointed out that 'the comparison between the average tilt of the Mediterranean axis in the charts and the variation of the magnetic declination in the area, from 1200 to 1600, suggests that the first prototypes were probably developed between 1200 and 1250'. The difference between those two dates is of considerable importance for the origin argument and we need to wait for further, more precise and well corroborated palaeomagnetic data before basing firm conclusions on this evidence.

[The Liber's ambiguous wording may have analogies in the discussion as to whether Marco Polo was referring to an Indian Ocean chart or text - see the comments above, under "Indian Ocean 'charts' shown to Marco Polo" (p.381a) ].

The Alexiad of Anna Comnena
For a possible early 12th-century reference, see The Alexiad of Anna Comnena, translated by E.R.A. Sewter (Penguin Classics, 1969), Book 13. In 1107, as the Emperor Alexis I realised (in this account by his daughter Anna) that the Roman fleet was stationed in the wrong area 'so that the south winds which blew against him were making the voyage easier for the enemy, he drew him a map of the coasts of Lombardy and Illyricum, with the harbours on either side. This he sent to him, adding written instructions. He advised him where to moor his ships and from what place to set sail if the wind was favourable, in order to attack the Kelts at sea' (pp.414-15). 'If we can take Anna's description at face value, this would be the earliest description of a chart and accompanying portolan'- Edson (2007) p.42.

Lo compasso and portolani

For inclusion of the names from those two early portolani in a geographical sequence alongside the names found on the charts, see the Listing and analysis of portolan chart toponyms along the continuous coastline from Dunkirk to Mogador (early 14th to late 17th century) including the transcribed names from the 'Liber de existencia riveriarum' and 'Lo compasso de navegare' as well as the Carte Pisane and Cortona chart. This is an updated (March 2015) version of an Excel spreadsheet, published in September 2013.

p.382a, n.104. For a long time, the earliest surviving systematic written portolano was considered to be the 13th-century 'Lo compasso de navegare', on which see the new translation and commentary by Alessandra Debanne (2011), and the review article on that by Andrea Bocchi (2011). Pujades (2013(b), p.18) moved the mid-13th-century date previously proposed for it past 1279 when Palamos, which it includes, was founded by Peter II of Aragon. That is on the assumption that the toponym appeared also in the work's putative first version. By contrast, Jacoby (2012, pp.72-3, 76) points out that Clarenza in the Peloponnese, evidently constructed in the 1260s, does not appear in 'Lo compasso' (although it is on the Carte Pisane). [Confusingly, the title of 'Lo compasso' has been differently transcribed: 'Lo compasso da navigare' by Motzo (1947), 'Lo conpasso de navegare' by Gautier Dalché (1995), and, most recently, 'Lo compasso de navegare' by Debanne (2011) - the form now used here.]

The date of the earliest portolan has effectively been pushed back significantly anyway by the discovery of the Liber, which must now be considered the oldest known portolan. [Perhaps further work has been done to date the portolan in the Bib. Naz. Marciana, It.XI, 87 [see n.104]]. Gautier Dalché (1995) compared the Liber and Compasso, noting significant differences.

For a comparative analysis of the texts of the interconnected portolani between 'Lo compasso' and the published 1490 Rizo version, see Thomas A. Warner's Academia.edu page.

Gautier Dalché (2005) describes a late-12th-century periplus attributed to Roger de Hovenden (on which see also Hughes, 2012), and Gautier Dalché's 1992 paper examines the evidence for 12th century portolani .

Perhaps it is time to build on the work done by Jonathan Lanman, when he attempted to construct a chart from the Compasso. Could a coherent chart emerge from the Liber's distances and directions, once its sections were arranged in a logical order? How would such a chart differ from Lanman's? And how much consistency would there be between the descriptions of short coastal hops and those for the oceanic crossings, on those two works and on the earliest surviving charts?

Falchetta (2008, pp.272-4) itemises numerous significant distance errors in one portolan by Michael of Rhodes (Michele da Rodi, i.e. Pietro de Versi), on which work started in 1434. Not only would these have ruled out a work such as that for navigational use - it would have been at best misleading and at worst highly dangerous - but, in his case at least, no attempt was made to correct those mistakes. A few even involved compass directions.

Pujades (2007) p.465a discusses references in the inventories to written portolani. He found just two mentions (1356 and 1453) of llibres de navegar, and concluded that they were 'of only relative use to seamen'. That he encountered many times more references to charts than to written works seemed to confirm that. Elsewhere (note 77 (p.519)) he states: 'Once the nautical chart had come into existence, portolans ceased to provide any information of much interest and became utterly dispensable supplementary instruments.'

That assertion seems to be contradicted by the frequent and detailed hydrographical comments found in 'Lo compasso de navegare'. The 'Glossario' to the 2011 edition by Alessandra Debanne lists numerous words for navigational dangers: for example, for rocks see under scollio, clappa, farillione, planca, for sandbanks and shallows see the variants of secca, arena, asperile, estagno, and so on. The colours and shapes of headlands are described, to help the helmsman's recognition or memory, and there are comments on the sea bottom, e.g. aspereto, as well as measurements of depths taken with a lead (scandallio or segnali). This is not information provided for the possible interest of a traveller but vital knowledge set out systematically for those in charge of the ship's safe navigation. Where are we, how do we get to our next stopping-place, what rocks may hole us or sands cause us to run aground on the way?

It is hard to reconcile the evident absence among the records of sailors' possessions of volumes with so clear a practical function. Particularly as they were copied and revised for several centuries. Is it possible that their contents might have been memorised, as a succession of detailed facts to be added to the visual memories from previous voyages? This is just one of the many outstanding questions. {The two paragraphs above added 7 October 2013}

On the 12th-century portolano of Salento see Uggieri (1994); see Jacoby (2012) for a 15th-century portolano of the Adriatic and Mediterranean, evidently derived from a 13th-century original; on the portolano of Grazia Pauli (second half of the 14th century) see Terrosu Asole & Motzo (1988); and for that by Grazioso Benincasa of the Adriatic, Aegean & Black Sea (1435-45), which is transcribed by Kretschmer (1909, pp.358-420), see also Biondi (1998?). Debanne (2011, pp.20-3) provides a brief summary of the early portolani.

For the Atlantic and northern coasts see Das Seebuch - an online facsimile, mounted by the Deutches Shiffahrtsmuseum, Bremerhaven, with translation into English and German, of sailing directions for the Baltic and the Atlantic coasts of Europe. This was written c. 1470 but is based on an original of the first half of the 14th century. Arranged in a series of geographical sequences, it adds information on tides, depths and descriptions of the seabed to what would be found in a normal portolano. On this see also Albrecht Sauer, Das Seebuch - the Oldest Northwest Sailing Directions: An Important Source for the History of Navigation and its New Online Edition, 2008, and the German Wikipedia under 'Seehandbuch'. {This paragraph added 17 March 2014}

Christie's (Giannalisa Feltrinelli Sale), 3 December 1997 Lot 193, described the following portolano: Italian (perhaps Genoese), 3rd quarter 15th century, covering the whole Mediterranean and the Atlantic coasts to Ostend, with reference to a 1445 shipwreck. Five other versions were referred to. [Was this by Michael of Rhodes - see below?]

p.383b. On Dati, see Hubert Michéa, A propos de "La Sfera" de Dati.

p.384a, note 120. On Pietro de/di Versi [i.e. Michael of Rhodes] see Long (2009). Also 'Michael of Rhodes: a medieval mariner and his manuscript (2005), Conterio (1991), Falchetta (2009b) and Arbel 2014.

See also the sections Function of portolan charts, under 'Shipboard use' and 'Navigational practice', and Toponymic listings.

For an analysis of the names on 'Lo compasso' see the author's extended essay on the Carte Pisane (March 2015), and also Bianca Fadda, 1999.

Top of page

For the full details of the works mentioned see the Bibliography


Recent Literature on Cartometric [Geometric] analysis:
    Aggeev (2013)
    Alexander (2004)
    Balletti (2000)
    Balletti & Boutoura (2001)
    Boutoura (2006)
    Bremner (1984)
    Dias (2018)
    Duken (1988, 1984)
    Fernández García, et al. (2000)
    Ferrar (various)
    Gaspar (various)
    Gómez-Paccard (2006+)
    Hermanski & Nothofer (1986)
    Hessler (2010, see also Rehmeyer)
    Hurtado García (2007)
    Kauffeisen (1987)
    Lepore, Piccardi & Pranzini (various)
    Livieratos (various)
    Loomer (1987, 1986)
    McIntosh (2018)
    Marelić (2019- )
    Mesenburg (various)
    Meyer (2000)
    Nicolai (various)
    Piccardi (2014)
    Rehmeyer (2014)
    Robles Macias (2014)
    Rohen (1987)
    Schober (1987)
    Steger (1896)
    Tobler (various)
    Wagner (1895 - a pioneering study)

Joaquim Alves Gaspar's methodology provides an ideal blueprint for the multi-disciplinary way that such cartometric studies might be carried out in future: 'using the modern tools of digital cartography and numerical modeling, in the light of the historical evidence given by the extant charts and written documentation' (Gaspar, 2010(a), p.2).

The compass and magnetic variation

p.384. Pujades (2007) pp.510-11 reiterates the point made by earlier commentators, namely that there was 'a clear deviation from the Mediterranean axis of between 8 and 10 degrees'. He explains that the 4 degree deviation he had measured was 'in accordance with the more technical system of global calculation'. Later (p.514a) he reminds readers (and illustrates on p. 320) why that deviation (actually to the east) is twisted in a counter-clockwise direction. Astengo (2007(a)) pp.194-9 summarises the arguments relating to the early period and carries it forward to the 16th century.

Lanman (1987) had pointed out that nautical charts started to be corrected for magnetic declination no earlier than the mid-16th century. Gaspar broadly confirmed that finding, stating that 'no significant adjustments were made to the orientation of the Mediterranean axes in the charts from 1300 to 1600' (2008 p.200).

Astengo (2005, p.30) provides a specific example. He describes an anonymous (probably Venetian) atlas in the BnF, produced 'after the mid-16th century', which is notable as representing one of the first attempts to reconcile traditional navigational practice with more scientific methods based on astronomical observation and geographical coordinates. The article describes how each of the two Mediterranean sheets is swung round in relation to their latitude scale so as to correct the traditional 'tilt', even if this means that the continuity between the two is broken. {This paragraph added 7 October 2013}

Recent Literature on the Compass and Magnetic variation:

    Aczel (2001)
    Astengo (1995)
    Cerezo Martínez (1985)
    Gaspar (various)
    Jonkers (2000, 2003)

On this specifically, see Gaspar (various) and Snyder (2009). With reference to my comment on António Barbosa (p.385, paragraph 2), Gaspar points out (personal communication, but see also his 2010(a), pp.34-6) that the contribution by Barbosa (1949) was indeed significant. See also below under Method of compilation.

Latitude scale (p.386a)
I had suggested that the earliest charts with a latitude scale were probably found among the undated charts assigned to the first decade of the 16th century. The Cantino map (c.1502) marks the equator and tropics, i.e. it has the beginnings of a latitude scale. However, Astengo, in his study of post-1500 charts (2007(a)) p.193b, pushes this event even later, stating that the innovation, 'derived from Spanish work, is perhaps to be seen for the first time in two charts of the Atlantic: that by Conte di Ottomanno Freducci (dated around 1514-15) ... and the 1516 chart by Vesconte Maggiolo'.

Nevertheless, Gaspar stated (in a personal communication, and see also his 2010(a) p.129, note 3) that "the earliest known nautical chart with an implicit latitude scale is, if its dating is accepted (there is no consensus on this matter), the La Cosa planisphere of 1500, which depicts the Equator and the Tropic of Cancer (though their positions seem somewhat approximate). To my knowledge no cartometric analysis was ever done to determine if such a scale has any correspondence with the latitudes of the places shown in the chart."

In his 2012 Imago Mundi article Gaspar states that 'Although no latitude scale is shown, the representation of the equator, the tropics and the Arctic Circle indicates that at least some places were charted according to their latitudes. This feature makes the Cantino planisphere [of 1502] the earliest extant example of the new cartographic model that was developing from the introduction of astronomical navigation, in the last quarter of the fifteenth century' (p.182). {This paragraph added 7 October 2013}

A joint article by Gregory McIntosh and Joaquim Alves Gaspar in Imago Mundi 73: 2 (2021) is entitled 'Kunstmann III: the oldest known nautical chart incorporating latitudes'. {This sentence added 3 November 2023}.

Lepore et al. (2011, pp.129-35) argue, on constructional grounds, that the latitude scale of the Beccari chart of 1403 is an original feature, despite being a full century earlier than any others known. I follow previous commentators in interpreting it instead as a much later addition. Among the reasons for this are the following: the accompanying writing is in a different hand, whose amateur nature is evident from the erratically-sized numerals and the occasional corrections to elements of the scale; the labels ‘duitse mylen’ and ‘spa[n]se mylen’ are anachronistic German terms; and, as would be expected if such a feature was being added to an existing chart, attempts were made to avoid disturbing what was already there. This can be seen in the lowering of the number 29 and the slight obscuring of the words immediately above and below it. {This paragraph added 8 September 2016}. Lepore et al. return to his issue in their 2017 Studi Costieri article, supported with detailed illustrations (pp.91-6). Dias (2018) does not come to a conclusion about the dating of the latitude scale. It can be additionally noted that the identical wording for the leagues is present on a chart in the Bibliothèque nationale de France B.119 assigned to the late-16th century and considered to be a Dutch production. {This sentence added 9 November 2020}.

It is worth noting that the first recorded use of a quadrant at sea for determining latitude was by Diogo Gomes in 1461 {Note added July 2020}

I did not discuss medieval coordinate systems, known at least from the time of Roger Bacon (1268) [see Volume 1 of The History of Cartography (p.322).] But it is worth noting the work by the mid-14th-century mathematician and astronomer, Paolo dell'Abbaco, illustrated by Gautier Dalché (2009, Pl.V & pp.127-31).] This is part of the list of coordinates taken, as I understand it, from a portolan chart. The MS, which relates to eclipses, is in Florence, Biblioteca Nazionale Central, Magl. XI, 121, f.94r.

Navigation instruments (p.386b)
Pujades (2007), p.520b points out that inventories of sailors' chests, even in the 15th century, make no mention of astronomical navigation instruments: Jacob's staffs, astrolabes or quadrants. On navigation instruments in medieval Mallorca see Ortega-Villoslada (2010).

Method of compilation

The 'triangulation' theory (p.387a)
For a refutation of this, Pujades (2007) refers to Kelley (1995).

The 'partial maps of the different Mediterranean basins' theory (p.387b)
Pujades (2007) p.511-12 finds no evidence to support this hypothesis, since (a) no such partial maps have survived or are referred to, (b) the earliest references are to works covering the entire area and (c) whereas differences of scale affect the Atlantic coasts and the Black Sea, the Mediterranean is at a single scale [though see above under Local scales]. On the other hand, in his summary of the conclusions found in the 1987 dissertation by Scott Allen Loomer (which I have not seen) Joaquim Alves Gaspar notes Loomer's finding that 'the analysis of the charts by individual basins shows distinctly regional characteristics, which is taken as evidence that they were developed from a series of regional charts tied together' (Gaspar, 2010(a) p.25, referring to Loomer p.169).

Observed directions and distances
In his own doctoral dissertation (p.7), Joaquim Alves Gaspar states that, for his analysis, he used 'a set of distances and directions between places, assumed to be representative of the maritime routes used for constructing the real charts, and the spatial distribution of the magnetic declination at the time the information is supposed to have been collected'. He elaborated (in personal communications) 'that it is no longer adequate to present the various theories regarding the geometric nature of the portolan chart as equally plausible. Two different types of evidence strongly support the idea that charts were made on the basis of observed magnetic directions and estimated distances: the historical sources describing how charts were constructed (the Liber and the texts of the 15th century, already mentioned) and the results of the cartometric analysis and modeling' [Lanman, 1987 and Gaspar's own recent work]. He explained further, that he had already demonstrated 'that the geometry of the portolan charts is well explained by the use of magnetic directions and estimated distances observed by the pilots at sea, and transferred directly onto the plane, as if the Earth were flat.'

Charts' place of origin

p.388. Pujades (2007) is to be congratulated for totally turning his back on the nationalistic stances that have bedevilled much previous discussion about portolan chart origin. While suggesting that the geographical information which they display came out of a shared Mediterranean nautical experience, he, as a Spaniard, has no hesitation in pointing 'towards the littoral arc that stretches between Genoa and Pisa as the birthplace of nautical cartography' (p.515b). He sees the concentration of portolan chart production in Genoa, Venice and Majorca 'as a consequence of the import of the Genoese cartographic and toponymic model into Venice (by the Vescontes) and onto Majorca (by Angelino Dulceti)' (also p.515b).

Top of page

For the full details of the works mentioned see the Bibliography

Drafting (pp.390-2)


On the various aspects of chart drafting see Pujades (2007) pp.476a- and (for the later period) Astengo (2007(a)) pp. 185-9.

Rhumb lines or outlines first?
p.390b (notes 183 & 184). Pujades (2007) p.473b, commenting on the examination carried out in the British Library at my request, states: 'Although this would seem to have settled the debate, the two fragments of unfinished charts I mentioned earlier [E.11 and E.13], combined with Astengo's studies of incomplete charts from the modern period [Astengo (1994), pp.163-5 and Astengo (2007(a)) pp.185-8], indicate that in the case of the wind network not all cartographers proceeded in the same way, which suggests that no general conclusions can be drawn from this particular examination'.
          There may be remaining doubts on this important constructional point but it seems highly probable that all members of a 'school', by virtue of their apprenticeship, would have learnt the same procedure. Perhaps further investigation might be able to use the priority of rhumb lines or outlines as a pointer to the atelier or school to which a given chart belongs.

Pinheiro Marques (1989), p.97, note 16, considered that 'the rhumb-lines were not added after the making of the chart, as a secondary device'.

Pujades (2007, p.474a) makes the interesting observation that on at least some charts the drafting sequence was more complex than previously thought: first the eight main directions (in black or brown), then the coasts, followed by the black place-names and finally the remaining compass directions in green and red. Elsewhere (2009, p.305) he notes - again it is not clear how regularly this procedure was adopted - that space would initially be left for the red names, denoting prominent places, and that those would be added later on. As evidence for this, he detected instances where the red names had been carelessly omitted. Discussing the 1439 Vallseca chart, Pujades (2009, p.335) notes that about a quarter of the place-names are in red.

p.390b (note 183). Pujades (2007) p.322 reproduces the rhumb line network from the Pietro Vesonte atlas of c. 1321 in Lyons (Census 7) and discusses this separately. He follows the suggestion made by Isabelle Raynaud Nguyen (in La Roncière & Mollat du Jourdin, 1984 p.200) that the placing of this bare network in front of the first of the seven charts was intended to emphasise the importance of this element - one that was featured in the supposed portrait of Vesconte on the Correr chart (Census 120). Having disposed of the possibility that this otherwise bare Lyons sheet was an abandoned chart, Pujades continues: 'It is an instrument deliberately intended to reveal to the future owner of the work the nature and characteristics of the wind network he would find at the basis of all the charts on the following pages, while at the same time constituting a kind of exhibition of Vesconte's professional mastery over this fundamental aspect of the art of cartographic production' (p.514b).
          While I cannot explain the presence of such an outline other than as some kind of accident, I am unconvinced by that reasoning. If the outline had been intended to appear in this lavish atlas, why was its border not coloured in? And how could an otherwise ignorant owner understand the purpose of the network - to provide a range of possible compass courses for a navigational voyage - without seeing it in its geographical context, and properly explained?. Finally, rather than the network demonstrating the chartmaker's skill, this would surely have been a routine initial exercise left to the apprentice. The skill came in copying the coastlines, writing the small names legibly and adding the illustrative elements.

On sheets containing the rhumb line network only see also Astengo (2007(a)) p.185b, who refers to the 1548 Vesconte Maggiolo atlas in Florence, Bib. Naz. Cent. These are also interpreted as 'not unfinished charts but explanatory drawings'. This might seem to weaken my scepticism, but just two examples, more than two centuries apart, do not seem to indicate that this represents the chartmakers' consistent response to a real information need.

The fragment discovered by Ramon Pujades in the Barcelona city archives (his C 55, illustrated on p.189; Census E.11) might be thought relevant here because it comprises the rhumb line network without any hydrographic information. However, see my comment to that entry. Another incomplete network might be considered here, namely the 16th-century chart of Greece (Paris, BnF, Rés Ge D 7898), see E.2(A).]

It is perhaps worth mentioning a much later anonymous work, a 'manuscript treatise on the practice and use of the compass rose and the portolan' [France: c.1650], sold at Christie's on 20 May 1998.

On the possible role of rhumb lines in the construction process see below under Copying & scales.

Recent Literature on Construction:

    Astengo (1994)
    Bremner (1984)

p.391a (middle of first paragraph). "there are no signs that chartmakers of the two previous centuries [14th & 15th] had access to an erasable pencil." See also p.392a (2nd paragraph) and n. 196. The Wikipedia entry for graphite states: "Some time before 1565 (some sources say as early as 1500), an enormous deposit of graphite was discovered on the approach to Grey Knotts from the hamlet of Seathwaite near Borrowdale parish, Cumbria, England, which the locals found very useful for marking sheep." However, Pujades (2007) p.483, n.27 cites a contrary opinion from Piero Falchetta (1995, p.73), that pencils were in use from the 11th century. Was this graphite? The point that matters, surely, is whether or not chartmakers were using an erasable pencil before the 16th century, even if 14th-century Italian painters were. Is there direct evidence of this? Or is the clear sign of a lead pencil line on the first sheet of Vesconte Maggiolo's 1512 atlas, the first documented instance of such use (Astengo 2007(a) p.185)? {This sentence added 12 September 2011}

p.392 (note 196). 'Dry-point sketch'. Pujades (personal communication, 16 September 2009) discussed the case of coastal outlines that might have been marked out with a scored line, presumably in the way that the hidden circle of the rhumb line network was created with a compass point. Such guidelines would then be gone over with ink afterwards, though the underlying indentation might remain. Falchetta detected evidence of this technique in the 1421 Cesanis chart (Census 119). The published footnote cited a similar claim for the 1409 Virga chart (Census 23) but Pujades could not see such signs.

Coastline marking
Pujades (2007) 508b, when commenting on the Maghreb chart (Census 89), notes 'the coastline is traced out ... with a fine line of ink reinforced on the inner side with a thicker line in ochre. This latter is a trait characteristic of cartography from the 1420s and 1430s onwards, found in no extant work from the fourteenth century. The first dated work that features any kind of inner line is Giroldi [=Ziroldi]'s atlas from 1426' [Census 116]. Pujades's note 31, p. 518, explains that, on the 1318 Vesconte atlas in Vienna [Census 4], a line is 'drawn not inland from but on top of the coastline, to highlight it in the exceptional context of an ochre-coloured sea ... [and that] both the colouring of the sea and the line that runs along the coast were added during the humanistic period'.

For details of the construction of post-1500 charts see Astengo (2007(a)) pp 188-9 and Astengo (1994), which also has comments on the plants and minerals used for the pigments (pp.167-70). Kriznar, et al (2022) investigates the pigments on the Juan Vespucci chart of 1520. See also the 2011 section on Colour and shape

Copying and Scales

Printing (p.391, note 189)
I had missed the further comment in Nordenskiöld's Periplus: "If, as I believe, Canepa's portolano [i.e. chart] of 1489 and Benincasa's of 1490 were only in part produced by the printing-press, then the two charts of Mattheo Pagano...are the first wholly printed portolanos." (p. 72b). There is, of course, no printing involved in either case. The earliest printed portolan chart is now known to be the woodcut of 1539 by Giovanni Andrea Vavassore, covering the Adriatic, Aegean and eastern Mediterranean, which Pagano would reissue in 1558 ( Periplus Pl.XXVII). See an illustration from the National Maritime Museum, Greenwich [search for 'vavassore'].

Shortly afterwards Nicolas de Nicolay issued his Nova et exquisita description navigationum ad praecipuas mundi partes recens delineata (1544). Copied from a portolan chart – not it seems Agnese but, no doubt, the source could be identified – the much reduced scale and the absence of printed colour meant that the toponymy had to be selective and the highlighting of important places omitted. Although the transfer to print was a tribute to the charts’ unique value, Nicolay’s version is no more than a pale imitation. {This sentence added 10 July 2020}

Master charts or patterns (p.391)
It seems inevitable that chartmakers would have had models to work from (master charts or 'patterns') whose information could be transferred, by some means, to new work. It would not otherwise have been possible for the production of the significant quantity of charts (referred to in surviving documents) by groups of semi-trained apprentices. Indeed, Pujades (2009, p.351) has identified 'documentary records indicating that, in the case of family-operated ateliers, those patterns were passed down to the heir or heirs of the master, and we can deduce ... that the same occurred in the case of cartographers who had learned their trade as apprentices at a cartographic atelier.' He cites the post-mortem inventories of Rafael Monells (1468) [Pujades 2007, p.437, No.121] and Vicenç Pruners (1625) [Pujades, 2007, p.479b]. The Monells document refers to 'un plech de pragamins vells pintats de cartes de navaguar' [a group of old parchments bearing the designs of portolan charts], which Pujades (personal communication 11 August 2009) interprets as patterns, since Monells, who had abandoned chartmaking some time before, would hardly have retained trade goods whose value would only have diminished with time.
          An earlier communication from Ramon Pujades (16 July 2009) talked about the second case, that of Vicenç Pruners's son, Joan, 'artifex cartarum navigandi sive menestral de cartes de navegar', who was said to have inherited 'half of the patterns for making portolan charts' from his father; and that the other half went to his sister. "These two documents show clearly, in my opinion, the importance of patterns in the manufacturing process of portolan charts, being carefully divided among the chartmaker's heirs".
          That no patterns appear to have survived from the medieval period, unless any of the binding fragments can be identified as such, is hardly surprising. We cannot therefore know exactly what they would have looked like or precisely how they would have been used. We should also be hesitant about assuming that copying procedures described by later writers were necessarily in use on the early charts.

See also the Workshops & apprentices section.

Copying (pp.391-)
One obvious way that scale could have been manipulated was via the pantograph, "a mechanical linkage connected in a special manner based on parallelograms so that they move in a fixed relationship to each other". However, the first recorded machine was apparently made in 1603 (Wikipedia, seen 22 December 2010). David I. Bower, in 'Saxton's maps of England and Wales: the accuracy of Anglia and Britannia and their relationship to each other and to the county maps', Imago Mundi 63:2 (2011) p.89 (?) and note 23 (?) cites a reference to an earlier prototype by Hans Lencker in 1571 as well as unsubstantiated claims that Leonard da Vinci has used a similar device. {Last sentence added 25 April 2011}. See also Monsaingeon (2017) on its possible use by Baldasaro da Maïolo Visconte in 1589.
          Pujades (2007) pp.478b-79b, explores in detail the possible ways that charts might have been copied, concluding that the most likely method was 'pouncing' (making holes through the copying intermediary which coincided with significant positions in the original, e.g. changes in coastline direction. Over that, dark powder would be sprinkled so as to fall through and stain the surface of the potential copy beneath, which points would then be joined up. A close examination of the Adriatic coasts on the work of Benincasa seems to support that suggestion. While the overall relationship of the significant headlands seem be to be very consistent, this is not the case with the fine coastal detail between them.

One interesting, though not obviously explicable, detail perhaps illustrates this. The Gulf of Kotor (Cattaro) in present-day Montenegro, has, in reality, an easterly, or south-easterly inclination. On Benincasa's earlier works (and a few later ones too) he showed it as tending to the east. In 1465 this retrogressed into a north-pointing bay. Why it should have changed away from reality and later, perhaps, been corrected is not the point at issue here. What it seems to demonstrate is that, while the bay's entrance would have been fixed by the pounced points, its interior outlines would have been drawn freehand. [It was later realised that the two versions reflected not the period of construction but the chart's scale; in other words, that there were two or more workshop models in use]. {Last sentence added 13 July 2022}.

Pujades suggests, given the inadequacy of paper in the early part of the period (both regarding quality and size), that damask might have been used instead, although that was described by Bartolomeo Crescenzio some centuries later. Corradino Astengo (personal communication, 7 July 2009) considers the Crescenzio passage to refer instead to "sottili aghi damaschini", thin damask needles, as tools to punch through the vellum used as the model, i.e. pouncing. Pujades (personal communication 6 August 2009) wonders if there is evidence of needles being imported from Damascus to Italy.
          Because of different availability of materials, great caution is needed before trying to apply later solutions to the early portolan chart period. Crescenzio was writing about 1600. Another late example is provided by a very unfinished French marine atlas, probably from the late 16th century. This betrays evidence of the workshop models (at different scales) used in its compilation. No doubt, some of its creator's problems were the same as those faced by 14th-century chartmakers; but it would be unwise to assume the solutions were the same (see Tony Campbell, ‘Egerton MS 1513: a remarkable display of cartographical invention’, Imago Mundi 48 (1996): 93-102). The fundamental problem remains that, even if some charts appear (or can be proved) to be direct copies, the majority do not seem to be clones, the product of mass production. On this see the Drafting section in 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition'.

It is fair to say that we still await a definitive answer to all the questions surrounding portolan chart copying, one that could be replicated in a real test. I do not think this has ever been attempted. [On this see Some areas for possible future research into early portolan charts under 'Copying methods'.] [Update:] For a technical reconstruction see Kevin Sheehan, 'The Functions of Portolan Maps: an evaluation of the utility of manuscript nautical cartography from the thirteenth through sixteenth centuries' {PhD 2014], and for a detailed and systematic investigation of the methods definitely or possibly used in portolan-chart copying see Šima Krtalić, ‘Anchoring the Image of the Sea: Copying Coastlines on Manuscript Nautical Charts from the Late Middle Ages and the Early Modern Period’, Imago Mundi, 74:1 (2022): 1-30. {Last sentence added 13 July 2022}. Also the video of a talk given at the Bodleian Library, Oxford, 22 February 2023: 'Cartographers, craftsmen and artists: collaboration and cross-pollination in the creation of nautical charts'.

Much cartometric attention is being directed at attempts to extract the mathematical secrets of the charts' original construction (for selected references see Cartometric analysis). But I am not aware of any GIS studies that have compared one chart with another to test how the coastline details changed over time or between chartmaking centres and individuals. Such analysis might serve to confirm or deny, first, the existence of models used for direct copying and, second, if clear traces of those are found, how widely the models were used and for how long.

p.391 (n.189). On the use of stamps for towns on the Giovanni Vespucci chart of 1520, the Christie's catalogue for 25 April 1990, Lot 94 has further information on that manuscript. Cities are shown 'by a uniform stamp of a turreted town, coloured in green and red'. This chart was acquired by the Archivo de Indias (Seville) and is reproduced in Rosselló i Verger (1995), pl.7.

Rhumb lines and the copying process (p.391)
Pujades (2009) p.305 suggests that "the coastlines were traced from one of the cartographer's chart patterns that also bore the rhumb lines". That implies that the rhumb lines were used to locate the coastline, but elsewhere (2007 p.473b) he agrees that there is no evidence for that. How often are the compass lines and the coastlines found in a precise, repeated position? Indeed, does this ever happen? On page 331 he describes how, on Gabriel de Vallseca's two later charts (1447 & 1449), 'the central horizontal line of the chart network ... goes through quite different places, a feature which suggests they were not traced from the same pattern.' In personal correspondence (11 August 2009) he explained that the relationship of the central east-west rhumb line (his 'diaphragm') on the copying master was used to ensure the correct orientation of the copy, which does not necessarily mean a direct relationship between rhumb network and coastlines.

Chart scales and distances (p.391)
See Pujades (2007) p.477a- , and the table in the Catalan-text section (pp.204-09) 'featuring the scale calculations of all the extant nautical charts I have managed to study from before 1470 which have come down to us sufficiently intact to allow me to make these calculations'. He considered five different crossings (avoiding those that involved islands, since those could be misplaced or mis-scaled). From this he concluded that the great majority were drawn at scales of between 1:5.5 million and 1:6.5 million, which confirmed that the Black Sea and Atlantic 'were initially represented to scales considerably different from the average scale of the Mediterranean basin' (478a). This analysis could be compared to the data set out in the various Richard L. Pflederer catalogues (via the accompanying CD-ROMs), although most of the charts he describes belong to the later period.

The Black Sea
In his later work, Pujades explores the question of the inconsistency of scale between the three different portolan chart regions (2009 pp.298-9, 333-4). The Atlantic understatement was corrected by Francesco Beccari in 1403 and imitated somewhat later by other chartmakers (after 1439 by the Catalans - Pujades 2009, p.334). Conversely, the Black Sea was exaggerated in size (see Chapter, pp.384 & 389). Whereas the sheets of an atlas could be, and were, drawn to difference scales, with varying scale bars to indicate that fact, no separate chart (as far as I know) showed any recognition that the single scale unit provided was not of universal application [though see the comment below about the 1409 Virga chart and above about Local scales]. From the 'Carte Pisane' onwards charts were provided with a single scale measure.

This raises interesting questions that seem not to have been asked, both in terms of the charts' origin and of their subsequent use at sea. 'The Black Sea is considered to have been added later to Lo compasso, though certainly by 1296, and the area is largely obliterated on the Carte Pisane' (Chapter, p.382 n.108). The Cortona chart that is generally considered to be the second oldest survivor, includes an impressively accurate Black Sea, although stretched out east/west. By at least 1311 (Vesconte) the Black Sea had achieved a fully recognisable outline (in no way inferior to that for the Mediterranean), testimony to precise observation and measurement (Chapter, p.371 n.5). Why was the Black Sea presented at a larger scale? Was this a result of the use of a different mile measure in the Black Sea, an indication that it was created by one or more different people at a later stage than the Mediterranean, or that one influential early chart copyist drew it, by mistake, at a larger scale?

A separate question focuses on how these linear inconsistencies might have affected navigational use. It has been too readily accepted, perhaps, that portolan charts were 'drawn to scale', when in fact they were drawn to three different scales, only the middle Mediterranean one - admittedly the version that would have been relevant for the great majority of voyages - being stated, as if it was of general application. Should we not, first and foremost, see those productions as 'compass charts', with the implication that it was that direction-finding instrument, rather than the very imperfect ability to measure distance travelled, that was the primary navigational tool, supplementing of course the all-important experience gained from previous voyages.

On that basis we can perhaps guess that medieval helmsmen learnt to make the necessary mental adjustments, just as, one supposes, they did when they gradually realised their magnetic compasses were (as they would have thought) pointing a little away from north. A man whose gun fired up and to the right would probably just aim lower and to the left, rather than discard it. Likewise, portolan chart users seem not to have found the mismatched scales (at least between the Mediterranean and Black Seas) particularly inconvenient. Otherwise they would surely have demanded charts that reduced the Black Sea to its correct proportions.

A sample check of charts from 1330 to the later 16th century finds instead a ratio between the lengths of the two seas ranging between 2.6 and 2.8. With actual lengths (not necessarily measured in the way I have done from the charts) given today as, respectively, 3,860 km and 1,175 km, that gives a true ratio of 3.3. When was that significant error, if error it be, corrected?

The Black Sea was of considerable importance to the Genoese, and to a lesser extent the Venetians and Catalans, from the mid-13th century onwards. In the 15th century, however, Genoa started to lose its colonies and trading posts to the Ottomans - the ring around the Crimean peninsula, for example, went in 1465 - and by 1484 the Genoese had been pushed out entirely. After that date, even though portolan atlases continued routinely to include a separate chart for the Black Sea (at least until the 17th century), this region was presumably no longer of active interest to Mediterranean seamen.

If, rather than being a mistake, the distortion of scale, though not of course of shape, was intentional, thus providing, for whatever reason, more space for the draftsman, this would have introduced one disadvantage. Because the Black Sea extends further east than the Mediterranean, the Caucusus shoreline defines the chart's right-hand edge. Unless the chartmaker (perhaps a Majorcan) wanted to extend north into the Baltic, or west into the Atlantic, it was the east-west axis that determined the usable space on the skin, and hence the overall size and therefore scale of the chart. By consciously enlarging the Black Sea - a region without any of the place-name concentrations found, for example, where the Mediterranean coastlines turned sharply, making it difficult to fit all the names in - a chartmaker who was making maximum use of his surface would therefore have been forced to make a very slight reduction in the scale of the Mediterranean. This makes so little sense that it is most likely that chartmakers and their customers remained unaware of the inconsistency between the Mediterranean and Black Sea scales, even after Francesco Beccari had identified and corrected the error in the Atlantic, and others had copied him.

Among the random sample of reproductions measured (from La Roncière, & Mollat du Jourdin, 1984) there is one notable exception, the 1409 Virga chart. In this case the ratio was 3.6, against the chart norm of about 2.7 and the correct figure of 3.3. This chart is also notable as being the first to detach the Black Sea from the Mediterranean outlines, so that it could be shifted to the left, thus allowing the eastern Mediterranean to define the chart's right edge rather than the Black Sea. Just moving the Black Sea would have achieved that saving of space. [See Chapter pp.444-5, and note 532 for further comments, e.g. that one or two later chartmakers also moved the Black Sea in from the right-hand edge. The 1421 Cesanis chart was, like Virga's, a Venetian production, rather than a Catalan one, which meant that no interior detail was disturbed by this adjustment.]

It is clear that the reduction in scale by Albertin de Virga was in no sense a correction, even if it accidentally had that effect. Instead, he added an note explaining that the the Black Sea had been drawn to a smaller scale, not reduced in relation to a scale that was consistent across the whole chart. There have been different attempts to read this barely legible inscription:

  • Quiste miga sie di quisto mar Mazor [.] - Ramon Pujades (personal communication 9 September 2009; he confirms that nothing is legible after 'Mazor')

  • quiste miga sia a quiste mar maior, infine alarena - Vagnon (2005) p.16 and note 24

  • Questi mig(lia) sie di questo mar maior - La Roncière & Mollat du Jourdin (1984) p.204

  • Quiste miglie sie a quiste mar Maior - Foncin, Destombes and La Roncière, Catalogue des cartes nautiques sur vélin conservées au Département des Cartes et Plans (Paris, 1963) p.20

[On scales, see further above in the section Introduction, under 'Local scales' (p.377a), and also the Drafting the charts section in 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition' - particularly under 'scales' and 'models'].

Size of the charts
Pujades (2007) includes the sizes of each chart and atlas in his chronological listings (pp. 63-70). In addition, see p.473a: 'the vast majority of complete charts that have survived intact measure between 51 and 66 cm high by between 90 and 102 cm wide. In contrast, the minority of charts over 70 cm high by 110 wide are all decorated works of highly varied dimensions. Lying at the opposite extreme we find a handful of works, all from the fifteenth century, whose dimensions are substantially smaller than normal, varying between 37 and 43 cm high by 68 and 84 cm wide ... And the additional fact that there was no direct proportional relationship between the dimensions of the vellum and the scale of the work constitutes a major aid when it comes to understanding the world of medieval nautical cartography production.'

Top of page

For the full details of the works mentioned see the Bibliography

Stylistic content [iconography] (pp.392-402)

See the following colour plates with grouped details in Pujades (2007):

  • Graphic scale bars (pp.220-1)
  • Calendars (four Vesconte atlases and the Catalan Atlas, pp.226-7)
  • Vignettes of Genoa, Venice and Avignon (pp.228-9); Venice reflected in its lagoon (p.257)
  • Wind circles (p.230)
  • Mansa Musa, the King of Mali (p.232)
  • The camel rider (p.233)
  • The war elephant (p.234)
  • Compass roses (p.235)
  • comparative details (p. 259) from the G. Soler chart (Census 15) and that assigned to his descendant Rafel Soler (Census 16)
  • comparative details (p. 261) from the signed Vallseca charts of 1439 [not 1339 as given in the caption] and 1447 (Census 128 & 17) and elements from unsigned works 'that suggest most strongly that they are by him as well' (Census 78 & 22).
  • comparative details (p. 268) of Antilia [though described as represented horizontally when it is vertical] from the following charts: 1424 Pizzigano (Census 141), 1435 Beccari (Census 100), 1436 Bianco (Census 112), 1455 Pareto (Census 104).
  • the T-O diagrams placed in Asia Minor (p.333) on the 1330 Dulceti and 1367 Pizzigano charts (Census 166 & 99), seen as indicating, in Dulceti's case, access to Latin books.
Astengo (2007(a)) analyses the 'ornamental features' of charts post-1500 (pp.199-203).

Recent Literature on stylistic content:

    Billion (2013, 2008)
    Kupčík (2003)

Discursive notes ('legends') (p.393)
See Pujades (2009, pp.356-8) for the full texts (in Catalan and English) from the 1439 Vallseca chart (also see key map, p.184). The volume also contains a loose, fold-out comparative table of 29 separate descriptive texts on eleven Catalan charts from 1330 to 1439 (for the accompanying note, see p.339).

Ramon Pujades's major work of 2023 - Els mapamundis baixmedievals: del naixement del mapamundi híbrid a l'ocàs del mapamundi portolà [Late medieval world maps: from the birth of the hybrid to the demise of the portolan mappamundi] - supplements significantly his previous transcriptions of the discursive legends; this time with full English translations. See Appendix III: 'Legends with geographic and cultural content in medieval nautical cartography' [Providing full transcriptions from 32 charts ranging from Petro Vesconte in 1313 to Pere Rossell [Petrus Roselli] in the late 15th century] (pp. 727 / 749 - 761 / 783).

Juan Ceva has placed English translations of the Catalan Atlas legends online (2014).

Chet Van Duzer's article, 'Nautical Charts, Texts, and Transmission: The Case of Conte di Ottomano Freducci and Fra Mauro' (BLJ, 2017), includes transcriptions, translations and a comparative study of the legends found on the work of Conte di Ottomanao Freducci and his successors, up to 1541, relating those to the work of Fra Mauro in particular but also Benincasa. See also a rejoinder by Patrizia Licini de Romagnoli (2020).

Graphic scale bars (p.393)
See Pujades (2007) p.481, and for 'Decoration and Additional Geographical and Cultural Information' in general, pp.481-3. His pp.220-1 include illustrations of a range of scale bars. He finds great significance in the fact that the earliest surviving charts already feature a decimal-based scale (also p.514b). He sees this as a link to the dissemination of Arabic numerals throughout Mediterranean Europe, associated with Fibonacci in the early 13th century. 'Consequently, if nautical charts came into being in Western Europe as fruit of the transfer of nautical data compiled in portolans to a graphic format drawn to scale, and if the reduction of distances to a specific scale was carried out using a decimal system, we may be sure that their advent did not take place until the first years of the thirteenth century in the Tuscan-Ligurian littoral zone, or until the end of the century in the culturally more peripheral areas' (p.515a). Though, of course, we do not know what any precursor of the Carte Pisane might have looked like. And elsewhere (p.521a) he notes that 'it was not until the work [Fibonacci's 'Liber Abacci'] was translated and simplified during the last third of the [13th] century that it began rapidly to circulate among Tuscan merchants'.

E.G.R. Taylor had already drawn attention to the pivotal role of Fibonnaci [Leonardo of Pisa] in her 'Mathematics and the navigator in the thirteenth century', Journal of the Institute of Navigation 13,1 (January 1960) [12 pages]. {Note added 20 December 2011}

Ships (p.393)
See the large-scale and well illustrated study by René Tebel (2012). For enlarged details of the ships on the 1413 Viladesters and 1426 Beccari charts and the 1436 Bianco atlas, see Pujades (2007) p.170 (Census 11, 37, 112). For a study focusing mostly on 16th-century shipping see Unger (2010).

Stylistic development

p.395- . To this section could be added Alfredo Pinheiro Marques's observation (in connection with the early-16th-century Dijon chart) that its realistic way of drawing the coast, free from the stylised conventions (e.g. of semi-circles) represents a new form of drawing associated with Portuguese cartography (1987, 1:85).

For an art-historical study of Majorcan charts see various publications by Sandra Sáenz-López Pérez, and also Sabater Rebassa (et al., 2016); and for a specific discussion of the various iconographic features of the 1439 Vallseca chart see Pujades (2009) pp.346-9.

Compass [wind] roses (p.395)
Pujades (2007) illustrates a number of examples (p.235); so too does Winter (1950, p.38). However, Pujades does not reproduce the example on the unsigned (though incomplete) chart formerly with Prince Youssouf Kamal in Cairo (his C 23, my Census 161). Pujades follows the conventional dating of this to the end of the 14th century, which means that its compass rose may be one of the oldest survivors, after the Catalan Atlas and work from the Cresques atelier [depending on the date given to the Corbitis Atlas - Census 117]. Like the Catalan Atlas [and also the simple device over a subsidiary intersection point on the 1330 Dalorto chart - Census 166], as well as two early 15th-century examples that Pujades illustrates, the rose on the Kamal chart is detached from the network rather than being a decorative elaboration of one of its main intersection points. The placement of the compass rose against a rectangular background on the Kamal chart is reminiscent of the treatment on the Islamic chart of 1413 by Ahmad al-Tanji [Kâtibî]. Might this be an Islamic feature?
      Pujades's note 63 (p.519) suggests that the extension of the network from 16 to 32 directions (see my chapter p.396a) 'may be linked to technical improvements which led to the dissemination of the fixed-pivot compass with the wind rose painted underneath, since ... both transformations began to be documented at the same time ... The new instrument made a greater degree of precision possible, which might explain why it was regarded as worthwhile to double the number of wind roses and networks on charts' (citing Kelley, 1995, p.4).

In relation to the doubling of the networks from 32 to 64 lines it is worth noting that Andrea Bianco's 1436 atlas (Pujades A 18), features a 64-line system in the diagram at the beginning, where eight sub-divisions of a wind are labelled. However, the charts in that atlas use the normal 32-line system. {Paragraph added 20 November 2019}

There is a useful summary in Wikipedia under 'Classical compass winds' (seen 25 October 2011) and another under 'Points of the compass' (seen 27 October 2013.

Astengo (2007(a)) p. 192a suggests that 'it seems likely that the charts actually used on board ship never contained compass roses'. See also Sider (1993).

For the symbolism of the compass rose see Faričić, et al. (2023)

Town views (p.397)
Pujades (2007) pp. 228-9 illustrates vignettes of Genoa, Venice and Avignon, and (p.257) Venice reflected in its lagoon. Cebrián (2007) illustrates vignettes of Genoa, Barcelona and Valencia. On town vignettes in general see Sáenz-López Pérez (2007a) I, pp.427-57 and Arnaud (1984); for the representation of Mecca see Sáenz-López Pérez (2007b).

A further example of a realistic cityscape, of which I had not been aware, relates to the depiction of Barcelona on the 1449 Vallseca chart. Sáenz-López Pérez highlights the depiction of Montjuich's de señales, with its distinctive devices for signalling the nature of an approaching vessel (2007a, I, pp.456-7) [see also her 2009(b) article]. Vallseca was a native of Barcelona.

The rudimentary town symbol, formed of a gateway set into a circular wall with one or more towers inside, all viewed from a slight elevation, is typical of Catalan work. The origin of this device can be traced back to the Roman surveyors' texts and is found in mappaemundi from the 10th century onwards (as pointed out by Paul Harvey in a paper at ICHC 2013 in Helsinki (July 2013 - abstract p.26)).

Flags (pp.398-401)
For an overview and classification of medieval armorial devices see Pastoureau (1998), and for a general summary of early flag types, which also notes that the first codification of flags appears to be the Laws of the Partidas, during the reign of Alfonso X, the Wise of Spain (1252-84), see Colin Campbell, Medieval Flags (Edinburgh: Heraldry Society of Scotland, 1983), p.1. {This paragraph added 28 April 2011, with amendments 23 September 2011}

See Pujades (2009) pp.339-45, e.g. p.340, which amplifies the places named in 'Table 19.1: Flags and Chartmakers' Response to Political Change' (my p.400). Durazzo (now Durrës in Albania) shows, after 1336, the two-headed eagle of the Serbs; the flag for Nice (Gules, a cross Argent) marked the moment in 1388 when the commune placed itself under the protection of the Counts of Savoy. Page 341 contains an explanatory list of the 68 flags (see p.126 for the keyed map) and an essay on the flag of the city of Valencia (342-5, but see pp.129-33 for illustrations).

Whereas the Byzantine flag was routinely left over Constantinople after 1453, the Pareto chart of 1455 includes a reference to its conqueror, Mehmed II with his newly minted title, 'Magnus Turcus' - an unusual example of topicality on a portolan chart (Van Duzer, 2017 p.48).

For an earlier general survey see Horstmann (1971) and for flags on Catalan charts see Fernández Gaytán (1987).

Pujades (2007) p.492b comments, about the G. Soler chart in Paris, BnF B 1131 (Census 15) that this work (dated by him to the period 1368-85) 'is the first to feature the new English coat of arms (with the introduction of the two quarters with fleurs de lis) and that of Montpellier (in which the stripes of Aragon are replaced by the fleurs de lis of the royal house of France), while others continued for many years to paint the outdated banners, examples being F. Cesanis and G. Vallseca in their charts from 1421 and 1439 respectively' (Census 119 & 128).

Amaral (1995, pp.175-6) drew attention to the new form of the Naples flag showing the Aragonese-Catalan arms, following the Spanish conquest in 1504. This has potential relevance for the dating of the Reinel chart in Bordeaux (Census 5) and perhaps other works assumed to date from around that period.

p.399. On the 'Libro del conosçimiento' see the Italian version with commentary by Astengo (2000b), and look under that heading in the first section of the Bibliography (comprising Anonymous works) for a 1999 facsimile of the Spanish original with commentary, as well as a 2000 English edition. It seems that the 'concordance tables with the sea charts, pp.70-5' is relevant to flags. On the 'Libro' see also Riquer (1987), Russell (1997) and Marino (1999).

p.401. On the subject of the addition of Spanish flags on the north African coast as a dating aid for the first decade of the 16th century, see the comment to Census update (under A10), about how Walther Ruge had pointed this out in 1911, though it was apparently forgotten thereafter.

Granada (p.401a)
'The Spanish flag seems to have been speedily inserted over Granada after 1492' (my original comment). However, the Sotheby's sale of 24 June 1993, Lot 99 described a Vesconte Maggiolo chart dated 1528, in which 'the Moorish kingdom of Granada in southern Spain is coloured in green even though the Moors had been expelled from Spain in 1492'. This is inherently unlikely and requires corroboration.

This was not discussed in the original chapter but for a discussion on the religious references in the 1413 Viladestes chart see Saenz-Lopez Perez (2012). See also the section 'Toponymy (Religious references)' in the Future research page.

Top of page

For the full details of the works mentioned see the Bibliography

Hydrographic development (pp.402-15)

p.402a. 'Slavish imitation'
"It has always been assumed, no doubt correctly, that portolan chartmakers were accomplished copyists. What is not justified, however, is the commonly expressed extension of this: that they slavishly imitated unchanging models. As we shall show, the portolan charts of the fourteenth and fifteenth centuries display a continuous and wide-ranging development."

I would now wish to modify that statement so as to remove the stigma attached to the word 'copyist'. Between about 1330 - simultaneously the end of the known activity of the Vescontes and the start of Dalorto/Dulceti's - and the introduction of Portuguese discoveries from the latter 15th century onwards, there are indeed few hydrographic (as distinct from toponymic) improvements. Instead, the perceived variations in coastal and island outlines are likely to reflect differences in regional or personal patterns and, as such, should be considered as pointers to lineage rather than date (p.404a).

However, the important point should now be made that the faithful and meticulous imitations made repetitively from the workshop models (patterns) ensured that, while the charts did not 'improve', they crucially did not deteriorate. Had there not been an almost total respect for the authority of the inherited outlines the charts would have corrupted beyond practical use in a short time. For a demonstration of how that would have happened look at some of those from the later centuries. [See further Conservatism and workshop practice.]

British Isles (pp.407-9)
See comparative details (p.309) in Pujades (2007) covering the Atlantic coasts of the British Isles, France and northern Spain, from the Carte Pisane (Census 14), 1313 Pietro Vesconte atlas (Census 25), 1321 Perrino Vesconte Atlas (Census 131), 1330 Dulceti (Census 166). [NB. The second illustration is beneath the first.] Billion (2011, pp.5-6) includes comparative details for southern Britain from the Carte Pisane, Cortona and Lucca charts, and Vesconte.

Atlantic France and Spain
See comparative details in Billion (2011, p.6) from the Carte Pisane, Lucca chart and 1313 Vesconte atlas.

See comparative details (p.314) in Pujades (2007) covering Italy from six works of the first half of the 14th century; and others in Billion (2011, p.6) from the Carte Pisane, Cortona and Lucca charts and the 1313 Vesconte atlas. For Tuscany see Lepore et al. (2011) and Rombai (2007).

For local studies see: Alfieri (1986, 1987 - Focara, Marche); Ventura (1997, 2001 - Puglia, and the South).

Baltic (pp.409-10)
See the comments on the treatment of this region in the works of Grazioso Benincasa and his followers (and predecessors), in the Baltic section of 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition'. See also Gautier Dalché (2013) and Iwanczak (2012), also for the southern coast of the Baltic, Szaniawska (2014). Van Duzer provides some useful references (2017, p.55, note 166).

Atlantic Islands (p.410)
Recent Literature:

    Andrade (2016) - mythical islands north of Canaries
    Dreyer-Eimbcke (1990)
    Fernandes Costa (1979)
    Fernández-Armesto (1996)
    Freitag (2013) - Hy Brasil
    Fuson (1995)
    Johnson (1995)
    Martínez Hernández (1996) - Canaries
    Meliá (1996) - Canaries
    Olshin (1994)
    Pinheiro Marques (1995b) - Azores
    Pinheiro Marques (1990)
    Porro Gutiérrez (2000)
    Pujades (2007) p.268 - comparative details of Antilia from four charts (1424-55)
    Quartapelle (various) - Canaries
    Rodrigues Wittmann (various) - Canaries
    Schmieder (2011)
    Wikipedia entry for Antillia

[For a note on the Cape Verde Islands see 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition'.]

Western coast of Africa (pp.411-14)
Table 19.2. The comment relating to 12.20N Cabo Roxo should have cited the 1465, not 1463 Benincasa atlas. For an examination of the role of Benincasa and Roselli in the transmission of Portuguese names, and a concordance to the toponymy on the charts of Benincasa and his predecessors, see 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition', and also the accompanying Microsoft Word 3 Tables, namely 'Table 2: West Africa and the Cape Verde Islands' and 'Table 5: Comparing West African place names from samotamat to cauo de sancta maria'.

On Portuguese discoveries in general see the various works by Albuquerque and Pinheiro Marques; Damião Peres, Historia dos descobrimentos portugueses, 3rd edition, 1983; Alegria, et al. (1999); J. Cortesão (1997); Serrão & Marques (1998); Relaño (2001a); and Ribeiro (1994). For early 16th-century names for Africa's west coast see McIntosh (2015), and for speculation that the extant charts were copies of others not just older but the products of a more sophisticated (unidentified) environment (Nicolai 2021).

North Atlantic (p.414)
Agnarsdóttir (2000)

Other regions not covered in the 'Chapter'
[For details see the Bibliography]

  • Adriatic - Lodigiani (2005), Marelić (2020)
  • Aegean - Baso & Scarso (2000)
  • Africa (East) - Vagnon (2011)
  • Armenia, Georgia and Azerbaijan - Galichian (2007)
  • Balearic Islands - Mascaró Passarius (2000)
  • Black Sea - Balard (1989), Beshevliev (1988), Džurova (1979), Fomenko (2001, 2007), Gordyeyev (various), Karpov (2000), Zevakin & Pencko (1969)
  • Caucasus - Khotko (2016)
  • Ceuta - Gozalbes Cravioto (1997)
  • Danube - Vagnon (2006)
  • Elbe - Brunner (1994)
  • Euboea (negroponte) - Blackler (2017)
  • Greece (Thessaloniki) - Beshevliev (1994)
  • Hungary - Irás (2007)
  • Malta - Soler & Ganado (2013)
  • Nile - Vagnon (2002), pp.26- , Cooper (2014), Seignobos (2017)
  • Rhône Delta - Oldham (1925)
  • Sardinia - Pinna (1996)
  • Sicily - Allotta (1999 - Agrigento), Borghesi (1986)
  • Turkey - Hild (2004 [Lycia], 2018)

Portuguese 'censorship' (p.414a)
Reference to the edict of 1504. See J.B. Harley, 'Silences and Secrecy: the Hidden Agenda of Cartography in Early Modern Europe', Imago Mundi 40 (1988) p.61, referring to a 1481 censorship measure (citing Jaime Cortesão).

Top of page

For the full details of the works mentioned see the Bibliography

Toponymic development (pp.415-28)

p.416. Table 19.3 'Significant place-name additions from dated works applied to undated atlases and charts' - for comments on, and expansion of, this table see the Microsoft Word table 'Significant Names', which lists all 1,800 recurring names between northern France and west Morocco. This forms the centre-piece of the large section added to these pages in February 2012, accessible via the Toponymy Main Menu.

In September 2013 a replacement table was published, enlarged to 2,800 names with the inclusion of intermittent, rare and unique names'. The Listing and analysis of portolan chart toponyms along the continuous coastline from Dunkirk to Mogador (early 14th to late 17th century) including the transcribed names from the 'Liber de existencia riveriarum' and 'Lo compasso de navegare' as well as the Carte Pisane and Cortona chart is in the form of an Excel spreadsheet. At the same time a detailed analysis of Red names was added. [In March 2015, an expanded and much revised version of the Excel spreadsheet was published HERE, with an updated version of the Explanatory Notes.]

For his general comments on toponymy see Pujades (2007) pp.480-1. Conti (1985) also discussed the charts' toponymic sources and the relationship between the names on the charts and those in the portolani.

p.418 'The undated works and their newly proposed dates'. See the Microsoft Excel spreadsheet A complete chronological listing of works assigned to the period pre-1501 and its Explanation. This includes a concordance between the listings of early portolan charts by Campbell (1986) and Pujades (2007). For comments on specific works, see under their Campbell number in Census of pre-sixteenth-century portolan charts: Corrections & Updates.

Note that the '1325/1330' Dalorto [=Dulceti] chart (Census 166) has now been definitely assigned to 1330 by Pujades, and the date of the 'Pasqualini' atlas of 1408 (Census 1) has been reinterpreted by Falchetta as 1448. The Dalorto had been placed in the chronology after 1327 but the supposed innovations of the 'Pasqualini' (now considered to be by N. Nicolai) have to be reassigned.

p.421. Pujades (2007) pp. 69-70 lists the sizes of pre-1470 atlases. See also comments about small-sized atlases in the context of the six charts on a single sheet (ex Kraus) discussed in the entry for E.19.

p.424. For a comment on how Arabic knowledge of the Maghreb was translated into a denser toponymy for the North African coast, see Herrera Casais (2008b) pp. 296-7. This will need to be tested against the Microsoft Word table 'Significant Names'.

p.425. My own toponymic analysis effectively ceased at about 1430 (because of the lack of usable reproductions) with just a sampling of later work. Pujades's volume on Gabriel de Vallseca and the other 15th-century Majorcan chartmakers (2009, pp.335-7), as well as his comprehensive analysis of the names from east Spain and the Adriatic (2007, pp.350-97) enables him to identify the way that Vallseca in particular absorbs both Beccarian and Venetian names. My own comparison of works by Roselli and Benincasa (both from 1468) had already shown that, by that date at least, the toponymic divergence between Catalan and Italian work had effectively ended. Pujades now pushes that development back by 20 or 30 years.

p.427a. For a corrective to the comment about the first appearance of Livorno, see Innovative Names: Historical time-lag.

'Appendix 19.5 Methodology of the toponymic analysis' (p.461)
Astengo (2007(a)) pp. 203-6 discusses various aspects of toponymy post-1500, e.g. conservativism, corruption, but also the way that the enlargement of a chart could lead to an increase in the number of coastal names.

Toponymic listings

Because my main current interest is portolan chart toponymy, this section has expanded to include relevant works published before 1986, as well as those dealing with post-1500 charts. The unrealisable aim is to be as complete as possible for those works that deal with the toponymy of the whole or part of the continuous coastline between Dunkirk and Mogador (with an exception made for the Mediterranean islands, whose toponymy I have not studied).

Note, though, that some of the works below have not yet been consulted

For the full details of all the publications see the Bibliography

GENERAL TOPONYMIC COVERAGE (i.e. of an entire chart or of numerous works, and general studies)

  • Antichi planisferi e portolani : Modena, Biblioteca estense univesitaria (2004) - the full toponymy of the Catalan Estense world map [cf Milano, 1996]

  • Armignacco (1957) - toponymic comparison (pp.193-222) between the Carte Pisane (Census 14) and Cortona chart (62), with Black Sea coverage from the Carignano map (65) and the 1325 [=1330] Dalorto/Dulceti chart (166)

  • Astengo (2000b) - an alphabetical listing of the places referred to in the mid-14th century 'Libro del conosçimiento' (pp.119-41)

  • Baldacci (1990) - the full name sequences for the 1430 Briaticho chart (Census 109) with a table comparing its toponyms with those of the 1542 Dalolmo, 1553 & 1599 Prunes charts

  • Campbell (1987, 1983) - methodology

  • Capacci (1994) - a general study including an alphabetical listing of medieval names, on about 400 pages

  • Casanova (1894) - Conte O. Freducci?

  • Conti (1985) - comparing charts and portolani

  • Cortesão (1954) - Pizzigano, 1424

  • Cousins (1984) - 16th-century Arabic and Turkish sea-charts, with a concordance to their place-names

  • Debanne (2011) - a transcription of and index to the 13th-century portolano 'Lo Compasso de navegare' [superseding Motzo, 1947]

  • Desimoni & Belgrano (1867) - Luxoro Atlas

  • Falchetta ('Periplus') - the toponymy of the 14th-century chart - Venice, Marciana, Ms. 10057 (Census 115)

  • Falchetta (1994) - toponymic comparison between Venice, Marciana, Ms. 10057 (Census 115), the 1339 Dulceti chart and the Catalan atlas (Census 13 & 28)

  • Gautier Dalché (1995) - a transcription of and index to the 13th-century portolano, 'Liber de existencia riveriarum'

  • Guerreiro (1992) - the complete toponymy of the Aguiar chart, 1492 (Census 146)

  • Kahane & Bremner (1967/1987) - Glossario degli antichi portolani italiani [a detailed investigation of the meaning of the terms used in the early written portolani.]

  • Milano (1996) - the full toponymy of the Catalan Estense world map [cf Antichi planisferi e portolani, 2004]

  • Motzo (1947) - a transcription of the 13th-century portolano 'Lo Compasso de navegare' [superseded by Debanne, 2011]

  • Nordenskiöld (1897) - transcriptions from the Luxoro Atlas, Catalan Atlas, 1426 Ziroldi/Giroldi (Census 81, 28, 116) & 1593 Volcius (pp.25-44).]

  • Pagani (1977) - 1318 (Vienna) Vesconte atlas

  • Pujades (2007) pp.350-97 provides regional lists of the names on 94 charts and atlases [with their modern equivalents], with the Italian and Majorcan works set out in separate lists [see Spain and Adriatic below]. The arrangement is chronological by supposed date of production. On the basis that the changed complement of names and their spelling can be seen as reflecting lineage and dating, Pujades places the undated works in what he considers their most logical place in the sequence. (The value of the presence or absence of particular names for dating purposes was described in Chapter pp.415- ). However, Pujades extends the process chronologically by a further 40 years (to 1469). His thorough analysis of the two chosen areas also allows him to make attributions of authorship on the basis not just of a name's presence or absence but more particularly its spelling. These tables are a very important tool for extending our understanding of the dating, authorship of, and interrelationship between, early charts. See also his general comments on toponymy (2007, pp.480-1).

  • Pujades (2009) pp.151-81 provides a full listing of the 1,854 names on the 1439 Vallseca chart. Proceeding first along the continuous coastline, anti-clockwise from buyetder (Bojador), then dealing with the islands and inland names, with modern equivalents, and the more important places keyed to the feint map (pp.152-3). For English version of the explanation see p.355.

  • [Rizo portolano, 1490]. Portolano per i naviganti ([Venice: Bernardino Rizo de Novaria, 1490] - choose TIFF or PDF format via 'Téléchargement de l'ouvrage' for the facsimile of the portolano text from the Bibliothèque nationale de France's "Gallica"). A full transcription can be found in Kretschmer (1909, pp. 420-552).

  • Rohr (1939) - Atlantic and Mediterranean name lists, covering both French coasts and the Mediterranean islands, from a wide variety of charts.

  • Vernet-Ginés (1962) - the complete toponymy of the arabic Maghreb Chart, which covers the western half of the usual chart area.

for the areas below see also the headings under Hydrographic Development

(in a geographical sequence, moving clockwise from northern France, round the Mediterranean and Black Seas to N.W. Africa)

Caraci (1950 - both coasts)
Caraci (1959 - Mediterranean coast)
Mille (2016 - Mediterranean coast)
Rohr (1939 - both coasts)
   NB. The Mille listing, and privately supplied updates, have been used to update (and partially re-order) the Excel listing (Nos 400-457)

Caraci (1950, 1952)

Spain (Mediterranean coast)
Billion (2011, pp.17-18). Cotlliure-Guardamar - names from the 2007 Pujades listing up to 1367 with the addition of the Lucca chart

Caraci (1959 - south and east coast)

Gozalbes Cravioto (various studies)

González Arévalo (2008 - Granada)

Muñoz Gómez (2013, 2015)

Pujades (2007). Cotlliure-Guardamar [Spain's Mediterranean coast from the French border south to below Alicante]. The modern names are provided and the red names highlighted. Giving names from 61 Italian works in assumed chronological order - from Carte Pisane to the 3rd quarter of the 15th century - in three sequences (pp.386-93), ditto from 25 Majorcan works, starting in 1330 (pp.394-7). See his p.522 for the 'selection and transcription criteria'.

Pujades (2002) - Valencia (superseded by the table in Pujades, 2007 pp. 386-93)

Quaini (1974) - Catalunya

Rosselló i Verger (1995, 1998, 2004) - east coast of Spain

Balearic Islands
Mascaró Passarius (2000)
Rosselló i Verger (2002, 2004)

Ascari (1940-)

Piccardi & Lepore (2012)

Baldacci (1982)
Piloni (1974)

Valerio & Spagnolo (2014)
Bellio (1882); for the islands between Sicily and Africa, see Vittorio Bellio, 'Contributi geografici', Archivio Storico Siciliano N.S. 10 (1885): 16-23

Straits of Messina
Mazzeo (2007)

Caraci (1936) - 16th & 17th centuries
Lepore & Piccardi (2014) - Tuscany
Longhena (1953) - including the 1482 Benincasa & 1494 Giorgio charts
Quaini (1974) - Liguria
Queirazza (1990)

Cassola (2011)

Billion (2011, p.17). Trieste to Ancona - names from the 2007 Pujades listing up to 1367 with the addition of the Lucca chart

Caraci (1936) - 16th & 17th centuries

Emiliani (1937) - Dalmatia

Falchetta, 'Periplus Adriaticus' - comparative list of the place-names along the Adriatic coastlines (islands excluded) from forty-five Italian manuscript portolan charts and atlases of the 14th and 15th Centuries. (Undated web page - about the toponymy from Otranto to Golfo di Larta, noting for each name the variants and their incidences, from 39 pre-1501 atlases & charts, and four portolani. Note that the Otranto-Ancona section is not covered in the Pujades listing).

Faričić (various) - Croatia

Lago (1998, 2009) - Croatia

Marelić (various) - cartometric analysis Milic (1979) - 'Yugoslavia'; comparative toponyms from nine portolan charts: 1318 (Vesconte) to 1480 (Benincasa)

Pujades (2007). Fiume Narenta (Neretva) to Ancona [anti-clockwise from Dalmatia (just south of Split) to central Italy] with selected islands. The modern names are provided and the red names highlighted. The numbering of the settlements demonstrates their sometimes erratic sequence on the charts. Names from 69 Italian works in assumed chronological order - from Carte Pisane to the 3rd quarter of the 15th century - in three sequences (pp.350-73), pp.374-85 (ditto from 25 Majorcan works, starting in 1330). See his p.522 for the 'selection and transcription criteria'. Note that the slightly longer Falchetta listing deals only with Italian works.

Salinari (1952)

Selva (2009) - Croatia

Rovigno (Croatia)

Balkan Peninsula
Banfi (1954)

Bevilacqua (1997) - Crete
Carile (1970) - Morea/Peloponnese
Pikoulas (2001) - Morea/Peloponnese
Soustal (2011) - Thrace

Aegean Islands
Armao (1947)
Gentilli (1936)

Black Sea
Almagià (1944-55)
Armignacco (1957 - including Carignano)
[Atlas. Bulgarian lands in the European cartographic tradition (III-XIX centuries.]
Beshevliev (1988)
Emecen (south)
Fomenko (2007)
Gordyeyev (various)
Koledarov (1970)
Kustova (2006 - Danube mouths)
Nenov (2011 - names in the 'Liber de existencia riveriarum')
Todorova (1975, 1978)

Asia Minor
Almagià (1944-55)
Belke (2016) - Hellespont and Bithnyia
Hild (2019) - Karien
Rebuffat (1999) - south Anatolia

Campbell (1984). [Including a table of the names on 'Lo Compasso' and 43 portolan atlases and charts (up to 1497)]
Livieratos (2010)

Eastern Mediterranean
Almagià (1944-55)
Caraci (1953)
Röhricht (1898 - on Palestine)

Africa (North)
Caraci (1952) - around Algiers
Ducène (2016)
Gozalbes Cravioto (1997) - Ceuta
Kamal (1937)

Africa (North-West)
Caraci (1950, p.29)
Gozalbes Cravioto (1997)

Africa (West)
Campbell (2009) - West Africa [see 'The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition' and the Microsoft Word format Table 5: Comparing West African place names from samotamat to cauo de sancta maria
Kamal (1926-51, Vol. IV, Fasc. IV, 1468) - west coast of Africa from Ceuta
Teixeira da Mota (1950)
Teleki (1906)

COMPLETE (?) TRANSCRIPTIONS FROM SPECIFIC WORKS (all periods) - for details see above

Aguiar, 1492 - Guerreiro (1992)
Anon. Venice, Marciana, Ms. 10057 (Census 115) - Falchetta 'Periplus'; Falchetta (1994)
Briaticho, 1430 - Baldacci (1990)
Caloiro et Oliva, 1639 & 1657 - Conti (1978)
Carte Pisane - Armignacco (1957)
Catalan Atlas - Llompart i Moragues (2005b) and various other transcriptions [search the Bibliography for 'Catalan Atlas']
Cortona chart (Census 62) - Armignacco (1957)
Dulceti, 1339 - Falchetta (1994)
Estense Catalan World Map - Antichi planisferi e portolani, 2004; Milano (1996)
Freducci, Conte O. (16th) - (?) Casanova (1894)
Giroldi see Ziroldi below
'Liber de existencia riveriarum' (13th) - Gautier Dalché (1995)
'Libro del conosçimiento' (14th) - Astengo (2000b)
'Lo compasso de navegare' (13th) - Debanne (2011); Motzo (1947)
Luxoro Atlas (Census 81) - Desimoni & Belgrano (1867); Nordenskiöld (1897) pp.25-44
Maghreb chart [Arabic names for S.W. Europe and N.W. Africa - Kamal (1926-51) IV.iii.1337; Vernet-Ginés (1962)]
Martines, 1591 - Guillén y Tato (1955)
Pizzigano, 1424 - Cortesão (1954)
Rizo portolano, 1490 - (BnF online)
Vallseca, 1439 - Pujades (2009)
Vesconte, 1318 (Vienna) - Pagani (1977)
Volcio, 1593 - Nordenskiöld (1897) pp.25-44
Ziroldi/Giroldi, 1426 - Nordenskiöld (1897) pp.25-44

Top of page

For the full details of the works mentioned see the Bibliography


See the sections above, where references to later charts are interspersed.

In addition, see:

  • Analysis. Astengo (2007(a)) pp. 203-6; (2000a) pp.69-75

  • Census. Astengo (2007(a)), pp. 238-62 and Pflederer (2009)

  • Bibliography. The footnotes to Astengo (2007(a)) pp. 203-6
  • Specific study. Astengo (2003) - on the 1561 & 1567 Maggiolo charts

Published facsimiles of 16th & 17th century charts
(for fuller details of the charts, and links to online scans, see the Microsoft Word 3 listing of Post-1500 charts)

1508. Andrea Benincasa. Sandra Sider. Andrea Benincasa' marine map of 1508. Borg. VIII.. Belser facsimile editions from the Vatican Library. (Belser Inc.: Yorktown Heights, N.Y., 1988). [Earlier version, 1984, with commentary in German by Arthur Dürst?] (x British Library [BL])

1512. Vesconte Maggiolo. [Atlas in the Biblioteca Palatina, Parma]. Der Seeatlas des Vesconte Maggiolo vom Jahre 1512 = Vesconte Maggiolo atlante nautico del 1512 / hrsg. und kommentiert von Georges Grosjean (Dietikon-Zurich: Urs Graf Verlag, 1979). (x BL)

1512. Vesconte Maggiolo. [Chart in the Hispanic Society]. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.III. Also Sider (1992) fig. 7. (BL Maps 147.d.37)

1519. 'Miller Atlas'. Atlas Miller (Barcelona: M. Moleiro Editor, 2003). [Attributed to Jorge Reinel, Lopo Homem and António de Holanda.] (BL Maps 224.c.8.)

1524. Conte Octomanno Freducci. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.IV. (BL Maps 147.d.37)

1537. Conte Octomanno Freducci. E.L. Stevenson. Conte de Ottomaño-Freducci, 1537 Facsimile (New York: Hispanic Society, 1915). (BL Maps 148.d.29)

1542. Battista Agnese. Ingrid Baumgärtner. Der Portulan-Atlas des Battista Agnese. Das Kasseler Prachtexemplar von 1542 (WBG (Wissenschaftliche Buchgesellschaft), 2017).

1544. Battista Agnese. Atlas de Agnese; Atlas de Magallanes (Valencia, 2005) - with commentary by Carmen Líter Mayayo & Fancisca Sanchis Ballester. (x BL)

1546. Battista Agnese. The portolan atlas of 1546 of Battista Agnese from the Russian National Library St. Petersburg. Edited by Arthur Dürst; commentary on the facsimile edition by Tamara P. Voronova (Graz: Akademische Druck- u. Verlagsanstalt, 1993). (BL Maps 217.d.39.)

1552. Bartolomeo Olives. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.IX. (BL Maps 147.d.37)

1553. Battista Agnese. [Venice, Correr]. Atlante nautico di Battista Agnese 1553: riproduzione in facsimile dell'esemplare conservato nel Museo Correr di Venezia. Presentazione di Giandomenico Romanelli; introduzione e commento di Marica Milanesi (Venice: Marsilio Edition, 1990). (Maps 196.c.37)

1554. Battista Agnese. Fac-simile delle carte nautiche di Battista Agnese dell'anno 1554 illustrate da Teobaldo Fischer ... XVII (Venice: Ferd. Ongania, 1881). (x BL)

1557. Battista Testarossa. Testarossa. Battista Testarossa, with a technical commentary by Richard Pflederer; foreword by Francis Herbert (East Sussex: Rediscovery Books Ltd, 2007). (** BL)

1563. Jaume Olives. Ivan Kupčík, Portolánový Atlas Jaume Olivese (1563) ve Vědecké Knihovně Volomouci [Nautical atlas by Jaume Olives (1563) at the Olomouc Research Library] (Olomouc: Univerzita Palackého, 2010).

1566. Jaume Olives. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.X. (BL Maps 147.d.37)

1570. Joan Martines. Atlas Portulano. Joan martines. Añy 1570 (Valencia: Javier Boronat, Editor S.L., 1994). [With commentary by various authors.] (x BL)

1580. Joan Olives (Joan Riezo/Riczo). Atlas de Oliva. Estudio de Maria Luisa Martín-Merás; presentation de Francisco Morales Padrón (Madrid: Testimonio, 1987). [In the Biblioteca del Palacio Real, Madrid; including two charts signed by Baldasaro de Maiolo, 1588 - all reproduced in Rosselló i Verger (1995) pp. 150-67, 196-7.] (x BL)

1582. Joan Martines. Joan Martines en Messina Any 1582. Facsimile with introduction by E.L. Stevenson, etc (New York: Hispanic Society of America, 1915). (BL Maps 148.d.24)

1586. Antonio Millo. [Greece & Aegean only.] Lothar Zögner. World Atlas of Antonio Millo, 1586 (Süssen: Edition Deuschle, 1988). [Facsimile, with commentary of the atlas in the Staatsbibliothek zu Berlin Preußischer Kulturbesitz, MS. Ham.446.] (BL Maps C.18.c.12)

1587. Joan Martines. Atlas de Joan Martines 1587. Introduction by José Ibáñez Cerdá (Madrid: Servico de Publicaciones del Ministerio de Educación y Ciencia, 1973). (Maps 195.e.16)

1588. Baldassare Maggiolo [see under 1580 above].

1600. Vincenzo Volcio. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.XIII. (BL Maps 147.d.37)

1605. Baldassare Maggiolo & Giovanni Antonio Vesconte. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.XIV. Also Sider (1992) fig.24. (BL Maps 147.d.37).

1637. Giovanni Battista Cavallini. E.L. Stevenson, Facsimiles of portolan charts belonging to The Hispanic Society of America (New York, 1916), pl.XVI. (BL Maps 147.d.37)


Balard (1994, 1978)

Carile (1970)

Crowley (2011)

Kahyaoglu (2016 - western Asia Minor)

Moore (n.d.)

Piemontese. Castelli Europei. Paese per Paese

Salmon, (1754)

Stringa (1982)

Zevakin & Pencko (1969)

See also Red names of overseas trading-posts

Top of page

For the full details of the works mentioned see the Bibliography

The business of chartmaking (pp.428-38)

p.428b. Recent Literature on Illuminators:

    Alexander (1992)
    Leduc, Pelletier & Toulouse (1998)
    Llompart i Moragues (1999, 1977-80)

"Establishment of the port" and tides
p.429a, note 390. On this see also David W. Waters, The art of navigation in England in Elizabethan and early Stuart times (London, 1958), p.31; Derek Howse & Michael Sanderson, The Sea Chart (Newton Abbot, 1973) p.47; and Rosselló i Verger (2009 [& English translation 2015]).

Workshops and apprentices

For a discussion of this topic see the 'Introductory notes' to Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops.

Pujades (2007), p.474b states that although 'it has not been possible to establish families of charts whose network centre is in the same geographical position, the picture changes considerably if we focus exclusively on their 'diaphragm', that is, their true horizontal axis'. By this he is referring to the west-east line passing through the centre of the hidden circle. This he traces, in the west Mediterranean, as ranging from south of the 43rd parallel to south of the 37th (from Perpignan southwards). In the Levant, the range is from south of 40 degrees to south of 33. There is a fairly constant 4 degree clockwise tilt, e.g. a line starting at 42 degrees in the west will hit the Turkish coast around 38 degrees. [See above (pp.384-90) under 'Cartometric analysis'.]
      The purpose of Pujades's table of 'complete charts' before 1470 (p.475), is to identify groups with approximately the same central 'diaphragm'. He notes that groups of 15, or perhaps more if marginally different groups were merged, often (but significantly not always) repeat the same chartmakers' names or those of their known apprentices.

See Pujades (2009), chapter 3 on Majorcan ateliers; and Ferro (1992c) on apprenticeship in Genoa. Also see the comments above under Copying & scales.


For a complete, chronological table of 'Cartographers and their ateliers' (up to 1470), including additional, mostly Majorcan, names not previously noted in English-language studies (and some compass-makers who may have also made charts), as well as a comment on those who were seamen - see Pujades (2007) pp. 486-7. He notes for each their known periods of activity, interrelationships (e.g. family and apprenticeship) and relevant biographical details. 'The number of known producers of nautical charts stands at only just over three dozen. Although, needless to say, we should add a few anonymous cartographers (ten or so whose unsigned work has come down to us in part, and others who have left no traces of any kind)' (p.487a). For the reasoning behind the suggested larger figure of 25 unknown chartmakers whose work survives, see Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops and the Tables of signed, attributed, anonymous and workshop productions (a Microsoft Word document). Note also Pujades's comments on individual practitioners in the section, 'The training of medieval cartographers' (pp.489-97 and notes 62-137).

Pujades (2007) pp. 250-1 reproduces the signatures of 28 practitioners.

'Scholars or craftsmen'
Pujades (2007, p.488b-97 examines what is known of the practitioners' 'technical and cultural knowledge', from documentary sources and the charts themselves. He looks for evidence of Latin competency or literary references.

Venetian sailors (p.433)
On the various Venetian practitioners - the Benincasas, Bianco, Cesanis, Fiorino, Nicolai, Soligo (2), Virga, Ziroldi - see Falchetta (1995). Falchetta demonstrated that some of those came from patrician families and had commanded galleys.

Antonio Millo
p.434, note 433. This refers to his signature as 'Armiraglio del Zante'. Millo's 1591 manuscript 'Arte del navicar' (British Library, Add. MS 10365) is signed by him as 'Armiralgio in Candia'.

Recent literature on specific practitioners

For references to specific charts and atlases see under the relevant 'Census' number.

Some of the 15th-century chartmakers or their successors continued working into the next century, e.g. Andrea Benincasa, Conte di Ottomanno Freducci (Van Duzer 2017, Astengo, 2007(a), pp.220-1), Pietro Russo (Astengo, 2007(a), p.225, possibly producing exclusively in the 16th century) and Jehuda ben Zara. [Aloisio Cesani (1574) is not relevant in the context of Alvixe Cesanis.] The surviving output of those practitioners (when in public collections) is therefore noted in the listing found in Astengo (2007(a)), pp. 238-61.

Top of page

For the full details of the works mentioned see the Bibliography

Separate individuals or families

For general studies of Catalan/Majorcan, Genoese and Venetian schools, see Pujades (2007), Chapter 5 'Cartographers and their ateliers' (pp.486-505) and the Bibliography in general. For a detailed study on Catalan chartmaking see Pujades (2009). For Venetian practitioners see the work of Piero Falchetta, and particularly his 1995 essay, 'Marinai, mercanti, cartografi, pittori'; and also Casti Moreschi (1985). For Genoa, see Ferro (1996, 1992c). For early Portuguese charts see Alegria, et al. (2007), Domingues (2009) and Guerreiro (1997).

The attribution of unsigned works to a specific atelier [or the confirmation of earlier attributions] is taken from Pujades (2007) pp. 63-70 (under the 'Autor' column). The attributions are also discussed in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'.

For the very early anonymous works - Carte Pisane, Cortona, Lucca and Riccardiana charts - see A detailed reassessment of the Carte Pisane: a late and inferior copy, or the lone survivor from the portolan charts' formative period? (March 2015) and the separate entries in Brief notes on the main documents discussed.

to follow up the 'Census' numbers see the Update page

Pinheiro Marques (1989a)

Petti Balbi (1986)
See also Innovative Names (Beccari)

The following charts are attributed by Pujades to the Beccari atelier:
(Genoa) (Amedeo Dallai) [Census 171; Pujades C 37]
[United States] Sidney R. Knafel [Census E.15; Pujades C 53]

Benincasa, Grazioso
Astengo (1990c)
Bonasera (1997)
Campbell (2010) - The style and content of Grazioso Benincasa's charts: imitation, innovation and repetition
Falchetta (1995) pp.54-60
Mesenburg (1999)

Falchetta (1993)
Falchetta (1995) pp. 49-52
Pujades (2007) p.496b
See also Bianco's "London" chart of 1448 and Innovative Names (Bianco)

Cola de Briaticho (who signs himself Cholla de Briaticho) was evidently born in the small coastal town of Briatico, between Pizzo and Tropea in the toe of Italy. His three-chart atlas of 1430 does not make that explicit but the inclusion of briatico, in red, at the foot of the Calabrian peninsula where there was already insufficient room for names, can have no other meaning. [The name has otherwise been noted, in black, only on the 1588 Maggiolo.] It seems unlikely, however, that he would have worked there, alone. If he was a professional - and his workmanship is clumsy - it is not clear where he would have practised. {This paragraph added 29 April 2011}

See also Innovative Names (Briaticho)

Ferro (various)
Galliano (1995)

[See the composite version of the pre-War photographs of the Carignano map, engineered by Alberto Quartapelle in August 2022, and available in the Medea database.]
Quartapelle (2023 etc)
Bausi & Chiesa (2019)
Galliano (2008)
see Census [update] 65
See also A detailed reassessment of the Carte Pisane: a late and inferior copy, or the lone survivor from the portolan charts' formative period? (March 2015) and the separate entries in Brief notes on the main documents discussed.

Falchetta (1995) pp. 46-8
Pujades (2007) p.495a (and note 115) confirms my attribution to Cesanis of the 'Luxoro Atlas' (Genoa, Bib. Berio [Census 81; Pujades A 14], see my p.403, fig. 19.10). He dates it slightly earlier than the signed chart of 1421. Pujades (2007) note 116 (p.504) suggests that some of the sheets in the Medici Atlas (Census 76) were based on the work of Cesanis, on which see 'Italian second quarter 15th century'. See also under Cesanis in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'.

Cornaro Atlas
See under Cornaro Atlas in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'

Cortona chart
Capacci (1996)

Cresques (see also under Census 28 - Catalan Atlas, and generally in the Bibliography)
Llompart i Moragues (2005b) (1999-2000)
Pujades (2009) pp.309-10 (on Jafudà Cresques=Jaume Ribes)
Pujades (2007) pp.491-2
Riera & Llompart i Moragues (1984)

A number of charts were attributed by Pujades to a Cresques atelier, on which see under Cresques in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'.

Arnaldo Domenech was a portolan chartmaker (e.g. the chart in the National Maritime Museum, Greenwich with a partially illegible imprint). The Library of Congress has the 1484 guide (co-authored with Piero Roselli, to whom he was apprenticed) to the weights and measures in use in 27 Mediterranean and European ports (Ristow & Skelton 4). This uses a network similar to that on the charts. It can be seen here.

'Dulceto, Dulceti, Dalorto or Dulcert'
See Pujades (2007) pp.490-1. Having examined the inscriptions on the Corsini chart, whose date he confirms as 1330 not 1325 (Census 166) and the BnF 1339 chart (Census 13) he concluded that the supposed 'Dalorto' signature was the result of a mixture of careless conservation and pre-supposition. He read 'de Dulceto' on the Corsini chart and 'Dulceti' on that in Paris. 'It strikes me that our protagonist signed his name indiscriminately as Angellinus de Dulceto or Angellinus Dulceti on Latin legends. Whether this was a Latinisation of the Genoese toponym Dulcedo or that of the Catalan surname Dolcet, is a question I shall leave to those who pursue national glory. As far as cartography is concerned, it makes absolutely no difference where he was born. What interest us is where he trained as a cartographer and where he engaged in his professional career as such, and the toponymy of his charts leaves no room for doubt about the Genoese provenance of his cartographic-toponymic pattern' (p.491b). See also Pujades's note 76 on p.503 about the nationalistic stances adopted in the past over this chartmaker.

For a contrary view of the Dalorto/Dulceti identification see Sáenz-López Pérez (2007a) I, pp.891-2, who cites stylistic differences as an indication that two people were involved.

The following chart is attributed to the Dulceti atelier, on which see under Dulceti in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops': London, British Library Add MS 25691 [Census 48; Pujades C 9]. For new research on Add MS 25691 see Quartapelle (2023 etc)

Falchetta (1995) pp.61-2

See Chet Van Duzer's article, 'Nautical Charts, Texts, and Transmission: The Case of Conte di Ottomano Freducci and Fra Mauro' (BLJ, 2017), for an amended census of his surviving charts, and an analysis of the discursive legends found on them. [See also Patrizia Licini de Romagnoli, 2020.]

Falchetta (1995) p.68

Giroldi see under Ziroldi below

For the chartmaker whose name is now given instead as Ahmad al-Tanji see Census 136

Loret, Rafel
Pujades (2009) pp.312-15

Nicolò Nicolai (author of an atlas, now dated 1448, formerly attributed to 'Pasqualini', 1408, whose son he was)
Falchetta (1995) pp. 62-3
Pujades (2007) p.496b

Pujades (2014)

See Census 110

Falchetta (1995) pp. 66-7

Falchetta (1995) pp.37-9
Pujades (2007) p.494 & 495b

Two charts are attributed by Pujades to the Pizzigani atelier, which see under Venetian (late 14th-early 15th century) in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'

Amaral (1995)
Pinheiro Marques (1989a)

Pujades (2007) p.493 and note 94 (p.504) where he reinterprets the well-known inscription to mean that Roselli had copied from Beccari rather than being his apprentice.

Three charts were attributed by Pujades to a Roselli atelier, on which see under Roselli in 'Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops'. See also the note on the Catalan Estense world map, also attributed to Roselli.

Because of a confusion, the chart in the Hispanic Society, New York, K15 [Census 149; Pujades C 73] was described as the product of a Roselli atelier, on pp. 65 & 206 of his 2007 work. It is included on the DVD, but with the description corrected to that of a Venetian work from the end of the 15th or early 16th century.

See also Innovative Names (Roselli)

Soler (Guillem and Rafel)
Baig i Aleu (2001)
Pujades (2015), on the newly discovered fragment; (2007) pp.492-3, illustrating (p.259) comparative details from the Guillem Soler chart (Census 15) and that assigned to his descendant Rafel (Census 16). Also comparative illustrations (p.83) of the Rafel Soler chart in Berlin and that attibuted to his 'atelier' in Paris, BN, B 8268), on which see R.Soler.
Wikipedia: Guillem Soler

Pujades (2009, pp.312-15) summarises what has recently become known about the Soler dynasty:

  • Guillem Soler, fl. 1368, certainly dead by 1402 (two signed charts, one dated)
  • Joan Soler, Guillem's son, fl.1405-9 (master chartmaker but no charts known)
  • Rafel Loret, Guillem's grandson, fl.1436, possibly died 1442 (master chartmaker but no charts known)
  • Rafel Soler (Joan's son?), fl.1440s, perhaps died in 1446 (two signed charts, neither with legible date)
  • Gabriel Soler (Rafel's son?) fl.1446-74/5 (master chartmaker but no charts known)

Soligo (Cristoforo and Zuan)
Falchetta (1995) pp. 64-6 [referring to 'Zuan', as in the chart signature, rather than the modern form, Giovanni as used in Falchetta's online 'Elenco comparato dei toponimi costieri dell'Adriatico']

Llompart i Moragues (1988)
Pujades (2009) [a major study on Vallseca; for his life see pp.316-27]
Pujades(2007) p.493, illustrating (p.261) comparative details from the signed Vallseca charts of 1439 and 1447 (Census 128 & 17) and elements from unsigned works 'that suggest most strongly' that they are by him as well (Census 78 & 22)
Sevillano (1975)

Two charts are attributed by Pujades to Vallseca, on which see Vallseca. See also The Vallseca contract of 1433.

Pflederer (2019)
De Marchi (2015 on his illuminators)
Edson (2004)
Falchetta (1995) pp. 29-37
Gautier Dalché (2010)
Benedetti (1991)
Pujades (2007) pp.489-90, and, on the question as to whether Pietro and Perrino were one and the same (he concludes that they were father and son) see p.502, note 38.
See enlargement of the illustration of the cartographer at work (Pujades, 2007, p.193)

On the question of works attributed to Vesconte see under Vesconte and for his toponymy see Innovative Names (Vesconte)

Viladesters [note the spelling, corrected from 'Viladestes']
Pujades (2009) pp.310-11

Dürst (1996)
Falchetta (1995) pp. 40-5
Pujades (2007) p.495a

The following chart is attributed to the Virga atelier:
Venice, Civico Museo (Correr) Port. 40 [Census 121a; Pujades C 28], on which see Virga

Falchetta (1995) pp. 45-6
Pujades (2007) pp.495-6, and note 124 (p.504)
New York, Hispanic Soc. K4 - the inscription has now been read as definitely the work of Ziroldi, dated 1447 (Census 147)

Four atlases are attributed by Pujades to the Ziroldi atelier, on which see Ziroldi

For comments on the usigned charts - and these are heavily indebted to the work of Pujades - see the Tables of signed, attributed, anonymous and workshop productions (a Microsoft Word document) and, for the supporting argument, Anonymous works and the question of their attribution to individual chartmakers or to their supposed workshops.

Top of page

For the full details of the works mentioned see the Bibliography

Recent literature focusing on the institution that holds the charts and atlases

(for references to specific charts and atlases see under the relevant Census number)

see Hofmann (2010) for a project to describe and digitise all the charts in French public collections

Germany, Munich
Kupčík (2000)

Great Britain, London, British Library
Herbert (2003)
Pflederer ('Catalogues' 2001)

Great Britain, London, National Maritime Museum
Pflederer ('Catalogues' 2006)

Great Britain, Oxford, Bodleian Library
Pflederer ('Catalogues' 2008)

Italy, Parma
Gorreri (2009)

Italy, Rovigo
Kupčík (2004)

Italy, Siena, Bib. Com.
Baldacci (1990) [i.e. 1430 Briaticho, Census 109]

Italy, Venice, Correr
XIV-XVIII yüzyil portolan ve deniz haritalan
Biadene & Tonini (2000)
Lucchi & Tonini (2001)
Tonini (2009)

Turkey, Istanbul, Topkapi
XIV-XVIII yüzyil portolan ve deniz haritalan
Goodrich (1993)

United States, Chicago, Newberry Library
Pflederer ('Catalogues' 2005)

Unites States, Minneapolis, James Ford Bell Library
Urness (1999)

United States, New York, Hispanic Society of America
Sider (1992)

United States, San Marino, Huntington Library
Pflederer ('Catalogues' 2004)

Top of page

For the full details of the works mentioned see the Bibliography

The chart trade

p.435. The section in Pujades (2007) pp. 497-501, entitled 'The output of cartographers and their ateliers: typology, prices and place of manufacture', adds very considerably to knowledge [at least in the English-speaking world] of the production methods and marketing of charts. Pujades starts by pointing out the recorded existence of many charts and purchasers, but of few known chartmakers or ateliers.

Trade in charts (middlemen) (p.437)
Pujades (2007) pp.459-60 on the trade in charts, with reference to the correspondence of the Datini agents, who 'acted as middlemen in the buying and selling of charts' [see No.104, p.436, also referenced to Nigro (2003)]. For detailed information from the Datini Archive about the supply of charts see Houssaye Michienzi & Vagnon (2019). See also Pujades above, in the addition at the foot, dated 8 February 2014, for further information on chart merchants.

Production numbers
Pujades (2007) p. 498a contrasts the three signed and a few unsigned works emanating from Gabriel Vallseca's atelier with a document referring to the production of 24 charts over six months, i.e. one a week. Multiplied over the 34 years the atelier is known to have been functioning (1433-67) that could mean a notional total of over 1,600 charts. The 1433 document in question is unusual because it was about debt, not a contract, and almost certainly involved the cheaper, less decorated charts, to which few documents normally refer. See further on Vallseca in Pujades (2009) and, for my comments, The Vallseca contract of 1433.

Prices (p.437)
Pujades (2007) pp.279-80 sets out all the references he could find to chart prices (39 in all), converted to a single currency (the Barcelona sous). The records, from the period 1284-1485, are mostly Catalan (Barcelona and Majorca) but also feature Genoa, Naples, Perpignan and Valencia.
          Throughout most of the period, the basic, functional charts were sold at 'approximately two Majorcan or Genoese pounds per chart, which amounted to less than two weeks of a naval officer's average salary' (p.521b). The figures are analysed and it is concluded that 15th-century chartmakers 'continued to reap very little financial benefit from each utilitarian chart they manufactured. For this very reason, they had to strive to make a great number of them ... [therefore] the only effective strategy was to reduce production costs to a minimum. And given the stable prices of prime materials, this meant basically reducing the time invested in making copies' (pp.499-500). This perhaps suggests that the more streamlined ateliers would be more competitive, but we still await a convincing explanation of how production could have been speeded up in practice [see the note above, in the section Drafting, under 'Copying' (p.391) about the continuing uncertainty regarding copying methods.] For further comments see, e.g., Stages in the construction of a chart, and What is the difference between a plain and a de luxe chart?.
          The listing of charts and their prices [Pujades (2007) pp.279-80] includes several, from the early 15th century onwards, that were described as showing signs of use, 'carta usada'.

Centres of production (p.437)
Pujades (2007, p.448, note 48; and p.502, note 37) states that there is no solid evidence that portolan charts were being produced in Barcelona in the late 14th century. Indeed (p.487a) production was essentially confined to Genoa, Majorca (Palma) and Venice. However pre-1501 charts were also signed from Alexandria, Ancona, Barcelona, Galilee, Lisbon, London, Naples, Rethymnom (Crete), Rome, Savona and Tripoli. Apart, perhaps, for Ancona, none of those were centres of chartmaking before 1500 and, in almost every case, just a single work was signed from there. On the Venetian practitioners, and for example the references to the selection of officers for the naval convoys (which sometimes involved chartmakers), see the major study by Falchetta (1995).

Astengo (2007(a)) examines in turn the eight main centres of chart production post-1500 (pp. 206-35). The earlier centres - Palma de Mallorca, Genoa, Venice and Ancona - are joined by Naples, Messina, Livorno and Marseilles.

Top of page

For the full details of the works mentioned see the Bibliography

The function of portolan charts (pp.438-45)

Shipboard use

pp.439-41. Pujades (2007) pp. 444-6 ('References to nautical cartography in literary and technical texts') lists 24 contemporary references before 1501, starting with the de Nangis account of Louis IX's voyage of 1270. Seven of the entries date from before 1400. They are given in the original Latin, Catalan, Italian, etc. Pujades, p.456a: 'Inventories abound with references to old, torn charts, and narrative and poetic sources with allusions to their use on ships'.

My own Preface to the Pujades book roundly endorses one of his clearest findings: 'past statements that portolan charts had no practical purpose and were never taken to sea can no longer be repeated' (p.410b). See Astengo (2000), pp.15-24, who 'cites several examples of decorated charts and atlases from the 16th and early 17th centuries which seem to have been used occasionally on board ship' - via Pujades (2007), p.451, note 189.

As evidence of shipboard use Gaspar (2023, p.9) points out that "A few highly decorated charts are extant, one of them of the end of the fifteenth century, exhibiting marks of navigational use. That is the signed and dated chart of the Portuguese cartographer Jorge de Aguiar (1492)." {This paragraph added 9 November 2023}.

My comment (Chapter p.431a) that 'Occasional clumsiness - for example, several attempts at scraping the hidden circles on sheets of one of the British Library's Grazioso Benincasa atlases (Add.MS.6390) ... - suggests the inexperienced hand of an apprentice' was mistaken and should be interpreted instead as the careless use of a pair of compasses (dividers) by the atlas's owner. If so, that would provide additional evidence of the use of such volumes at sea.

Piero Falchetta points out (personal communications January 2010) that uncorrected errors in the written portolani, both as to distance and direction, meant that they could not have served as reliable navigational aids. On this, see his various publications on the Cotrugli MS of 1464-5, his article on Michael of Rhodes and his 2008 article. Given that sailing in the Mediterranean was often in sight of land, and that portolan charts were drawn at a relatively small scale, he contends that their purpose was primarily for training young sailors and for general orientation. It is certainly the case that the sequence of coastal names is sometimes incorrect. For practical navigation what was more important was the sailor's ability to locate the ship's position by means of soundings and through sensing his surroundings. [See also A deeper understanding of the function of portolan charts.]

Falchetta's comments raise questions about the relationship between the texts and the charts. However, the charts' continually evolving toponymy and the numerous references to their use on board ship, may indicate that their greatest practical value lay in the place-names.

See also under 'Navigational practice' (just below) and also two sections above: Lo Compasso and portolani.

On chart use observed by landsmen, see the section under 'Shipboard experiences of travellers and pilgrims' (p.443 below).

Louis IX in 1270 (p.439)
Gautier Dalché (2004) traces instances of the term 'mappa' (from late Roman times) and 'mappa mundi' (from the 9th century). He asserts that from c.1200 (see the author's 'Carte marine et portulan' (1995)) the term began to be applied to portolan charts ('cartula mappe mundi') (p.190). It is, however, surprising that the next instance that he identifies as referring to a chart dates from almost a century later, in 1294. He places the Guillaume de Nangis reference to Louis IX's voyage of 1270 under 'Avant 1300' [the date of the author's death]. For detailed examinations of this incident see under Gautier Dalché in the Bibliography.

Aragonese ordinance of 1354
p.440a, first paragraph and note 486. Pujades (2007) p.426b, as evidently the first person to read the original source for the claim that the ordinance required each ship to carry two charts, found that there was no mention of charts at all. The relevant document (British Library Add. MS 11568) was incorrectly read by Antoni de Capmany in 1787, and his wording repeated thereafter. For the full detail of Pujades's important discovery see his p.447, note 21.

Navigational practice

See A deeper understanding of the function of portolan charts.

See Duggan (2015) for an argument that "the legal prohibition placed upon winter sailing [was] a Medieval legal fiction, rather than there being in practice a closed Mediterranean".

In 1458 a condottiero, Roberto di Sanseverino, recorded a discussion among the ship’s crew about the position of their vessel. The captain, second-mate, helmsman, pilot, and others, decided to cartezare [i.e., to trace a route or find the ship’s position] but could not agree whether they were within 40 or 50 miles of Crete or had already passed it. “Et dicto di’ el parone dela nave, lo scrivano, lo nochiero, el pedota, l’homo de consiglio et l’altri offitiali de la nave incominciarono ad cartezare et metere per scripto, ciascuno da per sì, lo parire suo, cioè ove se trovasse essere la nave; et niuno di loro fu acordante insieme, perché chi diceva essere apresso ad l’insula di Candia .XL.ta miglia et chi .L.; chi diceva havevamo passato quasi la dicta insula di Candia, et chi in uno modo et chi in un altro”. [I owe to Corradino Astengo this reference and the explanation of the crucial word for which there is no English equivalent.] {7 July 2020}.

For the latest work on Cotrugli see Chet Van Duzer, 'Benedetto Cotrugli's Lost Mappamundi Found - Three Times', Imago Mundi 65:1 (2013): 1-14; and Ribaudo (2016).

There are two surviving MSS of the 'De navigatione' [elsewhere described, from a version without title, as 'Arte de navigare'], 1464-5, by the Ragusan, Benedikt Kotruljevic (known as Benedetto Cotrugli/Cotrulli), each copy made, in the opinion of Piero Falchetta, before 1500, perhaps in the 1480s. As such, it is the oldest surviving medieval treatise on navigation.

The Schoenberg web catalogue entry describes it as 'an account of the whole habitable world at this most interesting moment in the history of exploration. He includes accounts of cartography, topography, oceanography, magnetism, the construction and use of the compass, etc.' Book IV is concerned with cartography, and includes a portolano describing 'the shores of the Mediterranean from Gibraltar to Venice and to Egypt, the Near East and the Greek islands as far as the Adriatic, with distances'. Since this is not identical to the works reproduced by Kretschmer (1909), it is suggested that the text may be based on the author's own experiences.

The MSS are:

  1. Yale University, Beinecke Library, MS 557 (dated 1464, signed and with a title-page)
    First identified in 1990 by Paul Kristeller; later published as: De navigatione. O plovidbi: Priredio i preveo Damir Salopek (Zagreb: Ex Libris, 2005). For a transcription see 'Benedetto Cotrugli, De Navigatione (1464-65). 'Trascrizione del testo del ms. 557 della Beinecke Rare Book and Manuscript Library (Yale University ) a cura di Piero Falchetta'. The original text can be seen via the Beinecke website [just enter 'Cotrugli' in the search box]. See also a repeat of the full transcription in Falchetta (2009a).

  2. Lawrence J. Schoenberg collection, University of Pennsylvania Library MS 473 (dated 1464, but without signature or title-page)
    Acquired in 2004 (and with its provenance set out in the Schoenberg description) this was described by Claudio de Polo Saibanti in a conference paper in 1981, as '"Arte del navigare", manoscritto inedito datato 1464-65', Imago et mensura mundi, Atti del 9. Congresso Internazionale di Storia della Cartografia (Rome: Istituto della Enciclopedia Italiana, 1985), vol. 1, pp. 71-79.
    Piero Falchetta, 'Il trattato "De navigatione" di Benedetto Cotrugli (1464-65). Edizione commentata del ms. Schoenberg 473. Con il testo del ms. 557 di Yale', Studi Veneziani, LVII (2009): 15-335.

[The information above kindly provided by Piero Falchetta, January 2010, whose Studi Veneziani article includes extensive commentary on the two MSS.]

Pujades (2007) p.462- , commenting on the Yale MS, notes that it 'describes the set of operations carried out with the dividers on the portolan chart in order to determine the ship's position at a given moment' (p. 462a - see his p.446 for the original transcription). Pujades records Cotrugli's intriguing comment that 'Where the paths of the dividers crossed was the "point" where the ship was located, which was marked with wax' (p.462b - my italics). There is evidently no explanation of this practice, which is referred to again by Pujades (p.503, note 61).

Pujades (2007) p.466, note 2, provides a translation of part of Cotrugli's text where he explains that the 'four officers at the stern who observed the compass and told the helmsman the route to follow at every moment, were the shipwright, the second boatswain, the caulker and the scribe'.

See also under 'Shipboard use' (just above) and also Lo Compasso and portolani.

Recent Literature on Navigational practice:

    Arnaud (2011, classical texts)
    Arroyo Ruiz-Zorrilla (1997)
    Ash (2007)
    Bochoca & Moal (2019 an edition of Garcie's 'Grand routier')
    Bolumburu & Bochaca (2007, about Garcie and the Bay of Biscay)
    Davis (2009 - ancient)
    Denny (2012)
    Fabre (2009)
    Falchetta (2008, 2009a)
    Fontoura da Costa (1983)
    Gaspar (2008)
    González Marrero & Medina Hernández (2012, classical texts)
    Hyde (1978)
    Kelley (1999) (on 'vigias', nautical warnings of hidden danger)
    Medas (2009, 2004 - ancient)
    Michéa (webpage)
    Obied (2016 - ancient)
    O'Connor (2005)
    Peck (2002)
    Polo Seibanti (1985) - on the 1464/5 'Arte de navigare' [by Cotrugli]
    Porathe & Svensson (2003)
    Richey (1995)
    Rose (2023) - north-west Europe
    Sánchez García (2017) - on lighthouses
    Sheehan (2009) - on lighthouses
    Tangheroni & Vaccari (1992)
    Trethewey (2018) - on ancient lighthouses
    Tucci (1991)
    Waters (1994, 1989)
    See also the references given in 'Literature on Arabic, Islamic & Ottoman charts' above, some of which discuss navigation.

Navigational dangers
Apart from Kelley (1999), on which see above, no discussion on this topic was included in the 1987 chapter. This was belatedly rectified with the following publication: ‘The Representation of Navigational Hazards: the Development of Toponymy and Symbology on Portolan Charts from the 13th Century onwards’ (by Tony Campbell & Captain Michael Barritt RN., The Journal of the Hakluyt Society, December 2020).

Toleta di marteloio (pp.441-2)
See Valerio (2007) and, for a useful summary, the Wikipedia entry for The Rule of Marteloio.

Pujades (2007) pp.463-4 states that 'there is no trace of the presence of the toleta di marteloio among sailors' belongings; furthermore, it does not appear either in late Venetian cases I have managed to document'.

For an illustration of the example in the 1436 Bianco atlas (Census 112) see Pujades (2007) pp.172-3; for a trigonometrical diagram assumed to have been placed there by the owner of the anonymous Majorcan chart in the BnF, Rés.Ge B8286 (Census 16) see his p.176; and see the low resolution scan of the 'Toleta de marteloio' in the Cornaro Atlas.

On the 1390 Genoese inventory reference to a martilogium see Pujades (2007) p.464, where he doubts that it is in fact a reference to the toleta. For more on the 'marteloio' see Masiero (1984) and Hubert Michéa De l'utilisation de martelloios ou roses des vents au Moyen Age.

'Arte del Navigare'
p.443, note 517, see the comments about Cotrugli under 'Navigational Practice' above. Also relevant to navigational matters are two late 15th-century manuscripts: the Cornaro Atlas and one in the Biblioteca Civica in Padua, published by Ornella Pittarello (2006).

Dividers (p.443b)
Pujades (2007) p.442a cites instances of charts found with 'their dividers' sestes, 'evidence of the fact that they constituted an inseparable couple'. Pujades states that these were made of brass (less prone to rust) with tips less sharp than today's steel versions '(which would explain why few signs of holes made by dividers are found on extant charts, not even on the undecorated ones with other signs of deterioration associated with their use, such as rents or stains caused by seawater)' (p.442b). And p.451, note 191: 'one of the arguments most put forward by those who contend that charts were not used daily on board ships is the absence from extant exemplars of the holes caused by dividers'. [Compare that with the Chapter's pp.443-4 and notes 519 & 521.] See Pujades p.116 for a detail of the man with dividers on the 1436 Bianco atlas (Census 112)].

Magnetic compass
Pujades (2007), p.457- discusses the relationship between the magnetic compass and the portolan chart, stating firmly that their histories are 'indissolubly linked', from the time of the first relevant inventory (1294) onwards.

Shipboard experiences of travellers and pilgrims (p.443)
For the many tribulations experienced by Ibn Jubair/Jubayr on his Mecca pilgrimage in 1183-5 see Ronald J.C. Broadhurst, The Travels of Ibn Jubayr: being the chronicle of a mediaeval Spanish Moor concerning his journey to the Egypt of Saladin, the holy cities of Arabia, Baghdad the city of the Caliphs, the Latin kingdom of Jerusalem, and the Norman kingdom of Sicily (London: Cape, 1952). [I owe this reference to Corradino Astengo.]

Recent Literature:

    Delano-Smith (2004)
    Hyde (1978)
    Régnier-Bohler (1997)
[See further above in the section 'Introduction', under 'Survival and lost charts' (p.374).]

Connection with trade

p.444- For a general overview see Pujades (2007), Chapter 1, 'Maritime trade and written culture' (pp.414-22).

Pujades (2007) pp.34-5 has a map (taken from Sezgin, 2004) pinpointing the Aragonese, Genoese and Venetian settlements in the Mediterranean and Black Sea region for the 13th and 14th centuries. Pages 36-7 (again from Sezgin) illustrates the commercial routes of those three powers, c.1330. For trading connections and actual shipping routes see Gluzman (2010), Gambin (2004) and Hocquet (1990).

Gautier Dalché (1995, pp.304-5) includes a general map of the Mediterranean, and an enlargement of the central section, showing the open-sea voyages listed in the 'Liber de existencia riveriarum'. Vagnon (2007b), pp.303-6 and the diagram on p.301, identifies trading and pilgrimage references on four Catalan charts in the Bibliothèque nationale de France.

See Jacoby (1986) for interpretation of a small 13th-century fragment, comprising both a portolano and a merchant's commentary on the places concerned.

On medieval merchants' manuals, pratiche della mercatura, see Stanley (2018).

Coastal or open-sea navigation?
Pujades (2007, pp. 456-7) discusses the merits of the arguments that claimed either inshore or open-seas navigation was prevalent. He concluded that, from fear of the high seas, many voyages were of an intermediate type, involving island hops or other means of restricting the period spent out of sight of land. An important distinction needs to be made between the needs of sailing vessels and galleys. The large number of oarsmen on the latter reduced to a few days the length of time they could travel without putting in for water and supplies. On these issues see also Lane (1983) and Tucci (1991).

Gluzman (2010, pp.264-5) analysed courses taken by over 130 documented voyages, carrying cargo or pilgrims, from Venice to the Levant in the 14th-16th centuries. On pages 270-2 he examined the account of the 1102-03 voyage of Saewulf and challenged the interpretation that this, and most other voyages of the time, involved staying close to the coast. Indeed, for the later period of his main study the pattern was the opposite with open-sea passages of one, two or more weeks being the norm. This is borne out by the description of such direct routes in the early portolani. A direct course would usually be quicker and for that reason it often seems to have been preferred.

Jacoby, in 1997, discussed open-sea voyaging in the 12th century and, (2007(a) p.680), noted that 'Deep water sailing was occasionally practised in the Black Sea by the eleventh century', though, for trading reasons, this was not usual.

Given that the magnetic compass would have been of most use in poor visibility or especially when out of sight of land, the question arises: did the related introduction of the compass and portolan chart, which are assumed to have had benefits for navigation, lead to changes in sailing practices or routes? 'Routes' needs to be distinguished from trading destinations. Even if commercial stops were made on the way this does not rule out that the accepted practice was to take the quickest route commensurate with safety. {This section added 7 October 2013}

Owners (p.444)
On the overlapping identities of merchants and seamen - what he dubs the 'merchant-sailor' - see Pujades (2007) p.453a.

Pujades (2007) pp. 454-5 refers to the 'exiguous presence of ethnical-religious minorities among owners of nautical charts: not one single Moslem and only one Majorcan Jewish convert among all the references I have managed to compile'.

Pujades (2007) p.456a. 'Nautical charts were the property above all of people who were either professional seamen or who, by virtue of their condition as merchants or war fleet officers, often took to the sea themselves or provided others at their service with all the accoutrements they needed.'

Top of page

For the full details of the works mentioned see the Bibliography

Appendixes (pp.446-56)

Whaling or St Brendan? (p.446)
Fig.19.22. It was pointed out by the whaling historian, Klaus Barthelmess (private communication 9 June 1989) that the illustration on the 1413 Macià de Viladesters chart should not be interpreted as a whaling scene (as earlier commentators had done). Instead, it represents the episode in the life of St Brendan when they encounter a great fish, which is mistaken by the pilgrims for an island. He pointed out that there was no evidence of European (Basque) whaling in the western North Atlantic before the 1530s and that the first recorded whaling scene on a manuscript map was on the Pierre Desceliers chart in Manchester University Library. As further evidence, he identified the Jerusalem Cross on the mast top and points out that the person on the quarter-deck wears a crown. In support, he cited works from the 1980s by Selma Huxley Barkham. The reference to whaling occurs on pp.393a and 445a.

However, the original interpretation, that it does refer to whaling, is maintained by Chet Van Duzer (2013, p.50-2). For other whaling illustrations see the same work's index under 'Sea monsters - Whales/whaling'.

Calendars (Appendix 19.1)

p.448a (beginning). "Faced with tables and calendars offering, in almost equal measure, dates that either confirm or contradict those of the works they accompany, we should obviously approach with caution those atlases whose dating has depended entirely on their calendars". I am delighted to acknowledge here, for the first time, the prescient comments of John Holmes, in charge of manuscript maps in the British Museum (now British Library) until his death in service in 1854. He was responsible for most of the work on the three-volume catalogue of the Museum's manuscript maps (1844-61) [whose entries are now accessible via the BL's integrated online catalogue]. His unpublished notes on manuscript maps survive as Add. MSS 20751-3 & 20774. In Add. MS 20752, f.29 (not seen by me until after the chapter was published) he focused on this very point in his discussion of errors made by other commentators (e.g. about the Medici and Pinelli-Walckenaer atlases). These notes might well merit transcribing; they should certainly be looked at by scholars interested in the state of knowledge at that time, and at provenance issues. [It is possible, though perhaps unlikely, that he may refer to charts not listed in the census; if Holmes's notes are trawled for that purpose, the Census's 'Index of Former Owners' (p.90) would need to be used.]

For illustrations of the calendars in the 1373+ Pizzigano and 1468 Benincasa atlases (Census 90, 43) see Pujades (2007) p.181, and for those in various Vesconte atlases (Census 25, 120, 7, 131) and the Catalan Atlas (Census 28) see his pp.226-7.

Vagnon (2005, pp.19-22) discusses and illustrates the calendars on two Benincasa-style charts in Paris, BnF (Italien 1698 and Italien 1710 - Census 29 & 31). She notes differences in the daily dates which indicate that, whereas Ital. 1698 is a direct copy from a Benincasa model (e.g. British Library Add. 6390), Ital. 1710 is not. Where I had stated (Census 31) that the calendar in Ital. 1710 started in '145(4)' she corrects that to 1459.


'Biographical Index to the Atlases and Charts Produced up to 1500' (Appendix 19.2, pp. 449-56). The corrections and updates to that list can be found via Census of pre-sixteenth-century portolan charts: Corrections & Updates and the Microsoft Excel spreadsheet A complete chronological listing of works assigned to the period pre-1501 and its Explanation.

'Atlases and charts known by name' (Appendix 19.3, p.459) The so-called Combitis Atlas (Census 117) is now associated with Corbitis or Corbizzi, and should carry one of those names instead.

'Chronological index of dated and datable atlases and charts produced up to 1500' (Appendix 19.4, p.460). See instead the Microsoft Excel spreadsheet A complete chronological listing of works assigned to the period pre-1501 and its Explanation.

Top of page

HOME (main menu)  |  Tony Campbell articles  |  Portolan charts menu  |  Index   |   Sitemap   |   What this site is about   |   What's New

eXTReMe Tracker