View Diagrams  Download Paper  Download Images 


The Island of Majorca is famed for its Medieval Portolan Charts with their beautiful decoration emanating from the practioners also being originally illustrators and illuminators of manuscripts. Being an Island set centrally in the western Mediterranean Sea between Iberia/France to the west, Italy with Sardinia/Corsica to the east and Africa to the south it was an entrepot, a trade hub, a stopover for navigators who plied their trade on the Seas. Those trade links also gave the Majorcan’s information with which to decorate their charts.

With Northern Europe being ostensibly Christian and N Africa Muslim the meeting point for the different races including the Jews, Majorca was ideally placed. But, although Majorca spawned such erudite scholars as Ramon LLUL, who wrote on most subjects including sailing procedures there is no trace of Chart production pre 1300AD. In fact in the whole area the extant charts we have start in c1290AD and then the Genoese flowering of chart production commences with Petrus Vesconte.

Undoubtedly there were hand drawn charts of coastal areas frequented by particular sailors with notes of hazards, port facilities and sailing directions, but they were in all probability small well used and thus likely to be ruined at sea and jettisoned when useless. They would have existed for several hundreds of years as texts imply regarding the Norman Conquest of Sicily and their fleet, particularly when the “Holy War” meant hundreds of ships and galleys ploughing the Mediterranean Sea.

But we are not discussing such charts, ours are the “Library Editions” used by Ship Owners, Merchants and Politicians, kept on land and practically never taken on board ship unless for transportation. They are the fine art of the Portolan chart world and require a different form of investigation commencing with the original chart.

On the website are already many single subject texts which deal with the Genoese and Majorcan cartographic draughtsmen and I now include the abstract page from each, which if studied first will infill the gaps in this text which does not repeat in full previous research.



Read research papers or books concerning Genoese 14th to 16th Century cartography and there is not one text written which investigates how it all happened and uses simple basic facts such as lifespan, training or raison d’être for the origination. The first extant work provable by attribution is that of Petrus Vesconte and his 1311 chart. But he had to be taught the art and that has never been addressed in the myriad of papers written over the last years. Add to that the Carta Pisane, probably 1290 and the Cortona Chart c1300, both of which have doubtful provenance, but by existing they push the date backwards for the training of their authors. Thus this text sets down clearly known facts and dates and uses the life spans attributable to evaluate the overlap of these cartographers and how their charts also spring from a basic pattern/template which is amply illustrated in the second part of this research.

The text is 20 A4 pages and contains 53 A4 Diagrams (originals are all A3, and some interlink to form the complete chart at large scale.) Link to ChGen/1



Many researchers have written texts which in part discuss the work of A. Dulceto, and imply the finding of magnetic declination, magnetic compass usage to plot the charts and differing projections used to draw them.

I have chosen therefore the same chart, A. Dulceto 1339 (LCP C8) held by the BNF Paris as the base data with which to counter those theories existing in the four texts to be analysed. The chart is first dissected to indicate its fundamental structure and the varying scales used from west to east, Ptolemaic Degrees, Millara, Roman Miles and Marritimo Miglio, as well as the reasons for its skewing. The link from Rome via Claudius Ptolemy to the Millara is clearly shown and thus the origins of the “Portolani” and its accompanying chart are confirmed. Any researcher who has read my previous papers will not be surprised at the findings, quite incontrovertible, as they are merely the data the chart exudes, and thus will understand why I have not written reams of text in explanation of my findings for this 1339 chart. But the intervention of the Ptolemaic degree is surprising and is resolved by the explanation of the various distance measures within a Portolan chart.

Please note that prior to the analysis of the four texts, I have inserted a short but very necessary note regarding the use of Cartometry, Formulae and Computers to analyse a Portolan Chart and thus end up with bland averaging and an incorrect analysis of the chart.

The four texts are, J A Gaspar, ePerimetron, 2008; W R Tobler, Am Geog Soc, 1966 and 2007; J E Kelley jr. Cartographica, 1995; and C Boutoura, ePerimetron, 2006.

The results of this critique are expressed quite directly and thus I would advise caution when future researchers may wish to utilize them; I consider them to be lacking.

NOTE: I am still awaiting a researcher to explain how the magnetic compass was used to plot the courses and then, how the pilot book data led to a Portolan Chart to be drawn. It has been stated again in 2015 that “owing to the navigational methods of the time, which were based on magnetic directions and estimated distances”, that research is so very necessary. Why the “Pilot Books” extant do not use magnetic data requires answering.

The text is 22, A4 pages and contains 23, A4 diagrams. Link to ChDUL/1



The atelier of Cresques Abraham (1325-1387) was undoubtedly a continuation of the atelier of Angelino Dulceto (!name!), who probably arrived in Majorca late 1320’s, having learnt his cartographic trade in Genoa. He was possibly still producing charts in the early 1340’s, as Chart C9 (London, Add.MS 25691) is dated to that period, although some scholars consider it may precede the 1325, C7 Chart. My ChCs/2 text also details this point as a possibility. However the change of date does not preclude the atelier continuing to the 1340’s and being the Cresque base, as the Cresque Chart is undoubtedly a copy of the Dulceto (!) charts. Therefore it is possible to opine that there is a linkage between the two ateliers.

The Cresques chart does however have two unusual features drawn thereon, one is quite unique, the second a copy feature, which are discussed in detail within this text analysis of the chart. It is then compared to the Dulceto (!) charts and other charts following the system developed for the ChCs/1 and ChCs/2 texts.

The text is 7 A4 pages and 11 (16) A4 diagrams. Link to ChCr/1



Commencing with an in depth study of the actual draughtsmanship to explore the hidden features, the text then investigates the donor charts which allowed it to be drawn, much of which has been known for over 100 years, and finally shows that it is the latest manifestation of the “mappa mundi” genre which are usually drawn as circular charts.

The text is 10, A4 pages and contains 21, A4 (A3) diagrams link to ChCATA/1



The 1403AD Portolan Chart of Francesco Beccari has been studied in a historical and a technical paper. Commented upon as unique for its latitudinal scale and what may be considered a rather self serving, apologetic text for past chart failures, it requires researching.

Ignoring the previous texts for the main analysis of the 1403AD chart, the actual evidence paints a different picture. Thus it is then possible to show by comparison to those texts the technical detail that should have been assessed and thus a more correct conclusion arrived at.

Finally the 1435AD Portolan Chart of Battista Beccari is also analysed with the obvious necessity of a comparison to the 1403AD chart.

The text is 12 A4 pages and contains 20 full colour diagram maps. Link to ChFb/1


From those texts it will be obvious that the Majorcan Portolan Chart industry was kick started by the arrival of Angelino Dulceti from Genoa at a time when the very skilled illuminators and illustrators were able to take advantage of a new business venture presented by the Genoese Portolan Chart being available.

The diagram chart sets down the major players in the two hundred years of chart development we are here-in discussing. If you have studied the timeline charts contained in ChGEN/1, the format is the same. For each proponent I have set down the known dates, mostly when charts were drawn, and interpolated their life spans and thus when they were trained in the art. Thus this aids the indicative likely sequential training and thus the passing on of the basic information, the Pattern/Template to complete a Portolan Chart.


We have been informed of some Master/Student relationships and can accurately assume others from the written notes appended to some charts. This diagram chart however is not as well annotated as the chart in ChGEN/1because of the dearth of real data available.

NOTE; the work of Juan Ceva to produce the translations for “The Cresques Project”, using the work of G Llompart is a valuable source of information particularly as the texts document the known facts concerning the Majorcan cartographic draughtsmen.


The time line interlinks Genoa from Angelino Dulcert’s appearance on Majorca c1330AD and then again in c1435AD when Battista Beccario makes the same journey. Both these Genoese cartographic draughtsmen inject the Island with the same basic Portolan Chart.

When Angelino Dulcert arrives it is obvious he trains Cresques Abraham and in all probability Guillem Canterelles and Guillem Soler. Cresques Abraham no doubt trained his son Jafuda Cresques who in turn trains S.Corcos/Mecia de Viladestes and his brother Joan de Viladestes. If you have studied text ChCATA/1, there-in it is stated the Crown of Aragon noted that Guillem Soler was able to produce charts should Cresques Abraham be unable and that Guillem Soler was a “Christian”. We are also informed that there was a Joan Soler and Rafael Soler and thus perhaps a small family atelier linked to the Cresques Abraham atelier.

We also know that Jafuda Cresques departed Majorca for Barcelona to partner Francesco Beccario and try to fulfil a contract for four Portolan Charts which was never completed due to financial implications. However from Jafuda Cresques in training Mecia de Viladestes and his brother Joan de Viladestes, we have some of the finest Portolan charts of the Majorcan genre.

Thus the scene is set for Francesco Beccario to return from Barcelona, stop off in Savona and produce a chart dated 1403and in all probability then return to Genoa to train Battista Beccario in the art as text ChFB/1 discusses.

But on Majorca it appears that only Corcos/Viladestes and perhaps his brother Joan are available to train Gabriel Valscheca, Ga Soler, R Monels. But given the period when Battista Beccario travelled to Majorca this could well have aided the Master/Student relationship.

We are on safer ground when we discuss the next cartographic draughtsman, Petrus Roselli as his training dates coincide very well with the arrival of Battista Beccario and his rather direct attribution on his chart of 1447 clearly indicates thos fact.

“Petrus Roselli composuit hanc cartan de arte Baptista Becarii in civitate maioricana anno domini XXXX VII.

This text is usually translated as “in the style of Battista Beccario” although in Italian Arte = Art or even possibly Skill, but more likely “de Arte” in this context translates as “by the method of “ or “ closely” and thus he is saying he has actually copied BB, used his Pattern/Template as he was his student and this can be clearly shown later.

At the same time Jaume Bertram is placed in the time frame as having been trained by Battista Beccario. However the later proponents of the ARTE appear to have been trained by Petrus Roselli, who being a very prolific cartographic draughtsman has left us charts dated 1447, 1449, 1456, 1462, 1464, 1465, 1466 and 1468. Unfortunately nothing is certain about Petrus Roselli, his life or antecedants. Is he Catalan, Italian and even was he a student of Batista Beccario in Genoa and travelled with him to Majorca as Beccario’s previous student had completed his training? We just do not know!

The next cartographic draughtsman would appear to be Arnaldo Domenech who has produced two extant charts, one being a diagrammatic distance calculator from ports on the Mediterranean coast and is held by the Library of Congress, dated 1484, where-as a normal Portolan Chart dated 1486 is held by the Royal Museum Greenwich and has the intriguing attestation as follows; “ Arnaldo donienech diszipulo petri Rosell composuit anch cartam jn civitates Neapoli year dnj nz cccclxxxvi.”

Comment; that fact raises the possibility of Arnaldo Domenech having a Pattern/Template for Roselli’s work in Naples and finding an apprentice etc or even another cartographic draughtsman whose chart existed in 1508 when Vesconte de Maiollo arrived to produce the Atlas for the Crown of Aragon dated 1511. Maiollo required a full Portolan to draw the atlas especially as one sheet is a Ptolemaic projection and the subsections of Europe and Africa are very individual. I have already questioned in text ChGEN/1 where Maiollo could have obtained such a chart, or the basic data for a chart and shown the obvious Genoese route via Albino De Canepa, but as it is obvious that Majorcan Charts flow from the Genoese charts a definitive answer is not possible, however the later section of this text which compares these charts will assist in that determination.

But now I must introduce Jaime Ferrer of Blanes who it is thought owned a chart in 1493, a Mappa mundi. However the story line is fraught with difficulty.

Jaime Ferrer was born c1445 at Vidreras (Gerona) and moved to Naples in 1466 where it appears he had relatives. He studied astronomy, mathematics and cosmography and in 1469 working for the Court of the Palazzo Real in Naples he became treasurer of a convoy bound for Barcelona. From 1468 to 1474 he was paid by the Neapolitan treasury and travelled widely between 1480 and 1488 when it appears he returned to Catalonia and settle in Blanes.

Then on May 2nd 1493 Pope Alexander VI under pressure from Fernando and Isabella issued a papal bull granting Spain the same rights and privileges in the newly discovered West Indies as had been accorded to Portugal in 1479 over Africa. But the crunch came when on May 3rd the Pope issued another bull stating the famous or infamous line of demarcation. But of course that is determined from the Cape Verde Islands and they had only just started to appear on Portolan Charts, with nothing to the west as yet. Therefore the only information Jaime Ferrer possessed was the fact that the Cape Verde Islands were the furthest west known on the coast of Africa and they were Portuguese. There is one extant chart which clearly indicates the Cape Verde Islands, unattributed and perhaps dated to 1492, held in the BnF Paris, Res. GE AA 562, and referred to as the Columbus Chart!

Thus on the 26th August 1493 Cardinal Mendoza writes to Jaime Ferrer to sort out the mess as follows;A nuestro amigo Jaime Ferrer, el Cardenal de Espana; Jaime Ferrer especial amigo; Nos querriamos fablar con vos algunas cosas que cumplan; por ende rogamos vos que vista esta letra nuestra portans y vengais aqui a Barcelona, y traed con vos el Mapamundi y otros instrumentos si teneis tocantes a cosmografia.

Not being a specialist in Medieval Spanish I asked for help and now include the emails as written;

“I would translate the body of the letter as follows; >We would like to discuss with you several matters that should be pleasing. Therefore, we ask you that, having seen this letter of ours, thus you will both come here to Barcelona and bring with you the mappamundi and other instruments, if you have them, relating to cosmography.< My question; > The only problem I have is the “if you have them”. Does it automatically also refer to the chart as many people refer to it as if it exists, nobody has found it or traces of it, and I suppose I do not believe it exists, but your comments will decide that point for me, hopefully definitively.< Answer> The “if you have them” could qualify both “el mapamundi” and “ostros instrumentos” at a stretch- in that the writer may be unsure whether Ferrer is in possession of both but it is less natural. And, even if the writer is unsure whether Ferrer has “el mapamundi”, he seems certain of the existence of such a thing; otherwise he would have said “un mapamundi” (indefinite rather than definite article).< Response from M J Ferrar> That helps crystallize my thoughts thus; 1) Mendoza assumed that Ferrer had a map as he thought Ferrer was a cosmographer of note; 2) when Ferrer arrived in Barcelona from Blanes and wrote to their Catholic Majesties he included a silly diagram that nobody understood; 3) there could not have been a map to explain the Line of Demarcation in 1493 as nowt was known; 4) Ferrer cobbled his measurements by changing the known Miliaria of 90 per degree which he thought was 720 stades (ie 90 x 8) and reduced it to 700 stades because the text of Strabo was now available; 5) in other words he was a typical medieval scholar and played fast and loose with the measurements; 6) his ideas for solving the 370 leagues were ridiculed by all; 7) when you read Jean Francois Fernel took it on board and made a bigger hash of the measurement, a chart could not have been involved. I have already put pen to paper regarding the Line of Demarcation and debunked it from Ferrers’ work.

I now include the two abstracts from the papers MsFER/1 and MsJFF/1.



In 1481 Pope Sixtus IV confirmed in his Bull, “Aeternis Regis”, the sphere of influence for Portugal and Spain in the Atlantic Ocean. This eventually became known as the Line of Demarcation. In 1495, Jaime Ferrer [1], cosmologist, wrote to Their Majesties of Castile two letters explaining the 370 league distance measure that was incorporated in the Bulls. In 1528, Jean Fernel, a Frenchman, wrote “Cosmotheoria” [2], a text dedicated to King Jaoa II of Portugal and there-in indicated a methodology to determine a degree of latitude measurement. Both however indicated a spurious measurement consisting of 700 stades which was derived from Eratosthenes “Geographia”. There was therefore in that age a complete lack of understanding of the measurements and thus errors in research. However, the greatest problem appears to be the research into the voyage of 1492 and the distance measure of C Columbus described in various terms but never detailed correctly.

The whole text is 17 A4 pages including appendices and 3 diagrams. Link to MsFER/1

Monalosphaerium; Cosmotheoriae & de Proportionibus J F Fernel, Measurements; Accuracy & Fantasy Abound


J F Fernel wrote three texts which indicate his first studies, mathematics and astronomy. Here-in they are evaluated, but only “Monalospharium” and “Cosmotheoriae” are completely investigated, whilst “de Proportionibus” is simply explained. Thus the measurements involved, the length of a degree of latitude and basic geometry, which includes Euclidian methods are all discussed. It is obvious that J F Fernel is using the work of Claudius Ptolemy, he notes the Almagest text, but his greatest usage is of “Geographike Hyphegesis” when he discusses latitude and longitude, although somewhat at variance.

However all three texts as published are full of errors; the figures are awry and the calculations leave a lot to be desired with a lack of explanation of their basic components.

It is obvious from the construction of the three books that they were compiled between c1520 and 1525 thus allowing publication in 1526 and 1528. The errors are twofold being both original and printer/publisher as the errata do not cover the most obvious. Thus it is clear that the three books were written from notes taken over the 5 years, transcribed at a later date and thus errors were made. However the idea and work involved in measuring the land surface to determine the degree of latitude cannot be criticised; it is a pity as will be shown that a modicum of inaccuracy crept into the methodology and J F Fernel was swayed by spurious measurement comparisons. Some never actually existed and are mathematical enigmas. Basically all three texts require to be carefully read before accepted as correct.

The text is 17 A4 pages and contains 15 A4 diagrams. Link to MsJFF/1


The only comment I can make here is that the chart held by BnF already mentioned could be the chart referred to but in that case why was it never mentioned in the correspondence from Ferrer and shown to their Catholic Majesties as part of the explanation?

We should also note that Grazioso Benincasa has Cape Verde shown on his 1467 Atlas, again held by BnF Paris and then following the discovery of the Cape Verde archipelago it started to be shown, but no contemporary Portuguese chart has survived to give us a first chart date. But certainly the 1492 mappamundi has all of the Portuguese discoveries there-on and even states “these islands were discovered by a Genoese Antonio de Noli”. We know that C Columbus called at the Cape Verde Islands on his way to the castle of Elmina in 1482, but the date of the 1492 chart is only estimated from the fact that a Spanish Banner flies over Granada which was captured from the Moors on the 2nd January 1492 and thus the chart must be later than that date. How much later of course is the crunch! Researchers have tried hard to equate this chart with the note written by C Columbus in his copy of Imago Mundi where he refers to “four charts on paper, all of which contain a sphere”, but in all probability that would mean a chart drawn in Lisbon c1485 and either updated or redrawn by A N Other. I believe it all points to Ferrer only owning a chart and not using it in this work.

The final cartographic draughtsman in this section, but by no means the last Majorcan cartographic draughtsman, is Judah Abenzara who has three extant charts dated 1497, 1500 and 1505AD.

The 1497 chart is held in the Vatican Library, is 670 x 925mm, drawn in Alexandria and clearly dated February 8th 1497, “the day when the news came from Cairo ( of the deposition and slaying of the atabek) Qumsuh Khumsumiah. The 1500AD chart is also drawn in Alexandria, with the 1505 chart executed in Safed, Galilee and clearly attributed as follows; Judah abenzara afata la prezente carta jn Safet de galilee en l’anno de dn d v del octobro.

Thus in fact the 1497 chart was drawn in 1496; the 1500 chart is unfortunately degraded to the extent that the full attribution cannot be read and thus we must accept the date. But the three charts offer a varying perspective on the world and the clientele available. Firstly all stem from a basic pattern/template which can be shown to be that of Petrus Roselli. The clear imitation of the roundels containing a typical medieval drawing of a man’s head on the 1497 and 1505 charts is a sufficient guide. But why does the 1500 chart not contain them?

The 1497 chart has not only the roundels but a figure mounted upon a camel, a loaded Elephant and a North African Ostrich, and both the Elephant and Ostrich were by this date nearly extinct in what is clearly Libya and Egypt, but generally these would not be drawn on a chart for a Muslim client. Alexandria was an entrepot and the commission is therefore most likely from a Christian source. Now study the 1500 chart drawn also in Alexandria and note that although it is clearly drawn from the same pattern/template there are no roundels, no people and no animals upon the portolan chart and thus it is quite open to opine that it was either drawn for a Muslim client or Judah Abenzara was bowing to Muslim sensibilities and curtailed the charts decoration rather than fall foul of the Islamic authorities.

Now return to the 1505 chart and apart from its better state of preservation it returns to the Petrus Roselli format with matching roundels also. Thus in Galilee, Judah Abenzara did not feel as constrained by the Islamic diktat, used the same pattern/template and produced some eight years after the first chart an excellent Portolan Chart now held in the Beinecke Rare Books and Manuscripts Library.

Thus I feel confident to opine that Judah Abenzara was of Jewish descent, Judah Ben Zara; was possibly born in Catalonia, but more likely Majorca where he learnt his art. We can thus surmise that as the 1497 chart was drawn in 1496, he would have left for Alexandria by the latest 1495. Whether this was via another country would be pure speculation as by copying the Majorcan/Genoese genre the language and presentation would automatically follow.


In order for the charts to be correctly recognized I have included a selection of the aforementioned cartographic draughtsmen’s work. By studying these charts the so obvious connection between them all will be apparent. I will list them and include the chart with its references for researchers who desire to obtain a copy capable of enlargement for study. They are not necessarily the charts chosen for section 2, which compares the coastal profiles to establish the continuation of the Pattern/Template usage.

1) Angelino Dulcert
This is the 1339 chart drawn in Majorca. It is held by the BnF Paris, Res. Ge. B696.


2) Catalan Atlas
The 1375 atlas was drawn on Majorca by Cresques Abraham and is held by the BnF Paris, Espagnol 30.


3) Anonynous Catalan

Dated 1375/1400 and attributed to the Cresques Abraham atelier it is held by the Biblioteca Nazionale “Vittorio Emanuele III”, ms XII, D 102.


4) Guillem Soler

Dated from 1368 to 1385 it is held by BnF Paris, Res. Ge. B1131


5) Mecia de Viladestes

Dated 1413, it is held in BnF Paris, Res GE AA 566


6) Francesco Beccari

Slightly out of date order but only to allow the next to be shown in order. Drawn in Savona Italy in 1403, it is held by the Beinecke Rare Books and Manuscripts Library,


7) Battista Beccari

Dated 1426 it is held by the Munich Library, Cod, Icon. 30


8) Gabriel Vallseca

Dated 1439 it is held by the Museo Maritimo Barcelona


9) Gabriel Vallseca

Dated 1447 it is held by the BnF Paris Res GE C4607


10) Petrus Roselli

Dated 1462 it is held by the Bnf Paris, Res Ge 5090


11) Petrus Roselli

Atelier of Roselli, held by Yale University, undated.


12) Petrus Roselli

Dated 1466 and held by the JFBL, Minneapolis


13) Grazioso Benincasa

Dated 1482? and held by the Biblioteca Universitaria ,Bolonia


14) Arnaldo Domenech

Dated 1486 and held by the RMG ref G230: 1/9


15) Judah Abenzara

Dated 1497 and held by the Vatican Biblioteca Borgia VII


16) Judah Abenzara

Dated 1500 and held by the Hebrew Union College, Cincinatti


17) Judah Abenzara

Dated 1505 and held by the Beinecke Rare Books and Manuscripts Library


A) 1375 Atlas, Catalan, BnF Paris is drawn in blue and is compared to the 1339/1350 portolan chart attributed to Angelino Dulcert and held in London. It is quite obvious that the Atlas is based upon the Dulcert chart and therefore used the Pattern/template Dulcert must have carried with him from Genoa to Majorca. See texts ChDUL/1 and ChCATA/1.

ChGME/1/D18 & D19

B) The London 1339/1350 portolan chart is compared to the 1385 Guillem Soler portolan chart. They are to all intents and purposes one and the same and thus Soler is also using the Pattern/Template of Dulcert.

ChGME/1/D20 & D21

C) The 1375 atlas is compared with the 1413 Portolan Chart of Mecia de Viladestes, who was trained by Jafuda Cresques and thus we should expect a close copying of the Cresques Atelier work, which is clearly evident.

ChGME/1/D22 & D23

D) Battista Beccari departed Genoa for Majorca and obviously carried the tools of his trade with him. This is the outline of his 1435 portolan Chart held in Parma Italy and is illustrated in LCP as C39.

ChGME/1/D24 & D25

E) Petrus Roselli has described his chart as copying the ARTE of Battista Beccari on this 1447 Portolan chart held in Volterra Italy (see D38).

ChGME/1/D26 & D27

F) the Battista Beccari chart of 1435 is compared to the Petrus Roselli chart of 1447 and as is to be expected the latter is a near perfect copy of the former and thus we can be sure that Petrus Roselli was trained by Battista Beccari and was the recipient of his Pattern/Template.

ChGME/1/D28 & D29

G) In order to establish the later utilisation of the Pattern/Template of Battista Beccari by Petrus Roselli the 1435 chart is compared to the 1462 chart of Roselli. Apart from the obvious creep of the Pattern/Template in its usage the two charts are basically the same.

ChGME/1/D30 & D31

H) The chart of Jaume Bertran and Berenguer Ripoll held by RMG, reference G230:1/7 is compared to the Cresques Atelier chart c1400. They are LCP C58 and C22. The first is dated 1456 and thus over half a century we can readily see very little has changed in the Pattern/Template used, with the exception of a major correction to the latitudinal distance of the Iberian Peninsula. Where the information came from to correct this major error in all existing Portolan Charts is unknown. The original Pattern/Template will change in time as copy after copy is made, but the fundamentals remain. However this correction will be examined in an appendix to the text.

ChGME/1/D32 & D33

I) The first two charts of Judah Abenzara are compared to ensure the same Pattern/Template is in use, which as they were only drawn three years apart and in the same city, Alexandria, that would be expected of them.

ChGME/1/D34 & D35

J) The 1505 chart by Judah Abenzara , drawn in Safed Galilee is compared to the 1462 chart by Petrus Roselli. In part it is a perfect match with the other sections only slightly differing in profile. Thus we can be certain that Judah Abenzara learnt his trade in Majorca and was allowed to copy the Pattern/Template used there by all of the Majorcan cartographic draughtsmen.

ChGME/1/D36 & D37

K) Finally, the 1447 chart of Petrus Roselli which has rather curious wind lines as extensions. This is seen on other Portolan Charts, but not always as finely drawn as here-on. It is a fitting end piece for this investigation.



As already clearly shown in Text ChGEN/1 the same pattern Template pervades the whole cartographic draughting in the City of Genoa. It is now clearly shown that the Pattern template was transferred by both Dulcert and Beccari when they moved from Genoa to Majorca and that all of the Majorcan cartographic draughtsmen used those for over 125 years where-as in Genoa the usage spanned from 1300 to 1550.

This is a tribute to the pioneering work of Petrus Vesconte and his unknown Teachers.

The chart LCP 58 drawn by Jaume Bertran and Berenguer Ripoll suddenly corrects the glaring error in the latitudinal distance of the Iberian Peninsula between the latitudes 37N and 43N usually drawn on Portolan Charts as 6 degrees of 75 Miliaria which is a mistake of the medieval cartographers in believing the Roman Mile and Miliaria were one and the same. That is 75 Roman Miles equals one degree of latitude and 90 Miliaria is the same distance. Hence on a portolan chart the latitude to longitude at 36 North is always 90/72 miliaria, with the Iberian coast scaling 6 x 75 or 450 miliaria instead of 6 x 90 or 540 miliaria.

This is now investigated in the following appendix which discusses all three charts by Jacob Bertran.

RMG G 230; 1/7: 1456 CHART
JAC BERTRAN; CN7, 1482 and Dis Vol Bn 237, 1489

When studying their 1456 Portolan chart, drawn in Barcelona, I was struck by several curiosities. The most obvious is that the Island of Hibernia was missing, although a note in another text states “falta irlanda por deterioro, restaurado actualmente”, and, they have surrounded the chart by a scale bar as if accentuating its importance; there are also rather misplaced ornate Windrose letters for Sirocco and Greco and the written attribution panel with their names Jachobul bartan and Berenguzig zipol slightly awry as well as a different, that is not normal for a portolan charts date definition.

Thus my first task was to actually identify both persons and particularly J Bertran.

The attribution panel appears to be as follows;
Jachobul bartan et berenguzig zipol composines sxanch in civitas barchias anno anativitate dni m.cccc.L sexto—- (1456)


Quite strange as it appears to use a mixture of Catalan, Spanish and Portuguese within and the names of the two cartographic draughtsmen are written using medieval cursive shorthand script rather than plain spelling. The names are normally given by researchers as Jaume Bertran and Berenguer Ripoll. But the first is actually written jachobul bartan; the jacobus being a Hebrew given name, originally “Yaakov”, latinised to Jacobus/Jacob,(which is confirmed by his son on his death, see later) and then translated to the Catalan male given name of Jaume. But Bertran is also an original given name and thus part of the Jewish methodology as we read with Cresques Abraham and his son Jafuda Cresques. However it then becomes a patronymic surname in the far north of Spain, Catalonia and Andorra.

The second person, normally given as Berenguer Ripoll is in fact written Berenguzig zipol (the z being elongated) and in cursive script can be transcribed as a gh or y and printed as the elongated z. But Berenguer as a personal name is the Catalan version of the Germanic Berengar, although it was probably a Norman French transfer here. It was also a surname in 1137, Ramon Berenguer, Count of Barcelona. If we then look at the surname zipol or Ripoll we find it is a town some 80km due north of Barcelona and along with Vallfogona de Ripolles 10km east it is close to the French border of today.

Hence at this juncture it appears both cartographic draughtsmen could be natives of Catalonia and possibly had no contact with Majorca. That opinion is further amplified by the usage of the term “anno anativitate domini” which is both Spanish and Portuguese and translates as “in the year of the Lords birth”, and is in effect a double statement of the same fact. But, the possibility of a Majorcan base is still there.

Rey Pastor and Garcia Camerero in their “Cartografia Mallorquina” text, page 82 make the following point; Collaborating with Berenguer Ripol in his chart of 1456, dated in Barcelona, but Bertran is resident in Majorca, where he worked in 1482 and 1489. One can assume that the Mallorcan teacher was in his youth to teach his art in Barcelona, where the cartographic technique did not bloom much and Ripol being one of his disciples.

I develop this dating and age problem later as it impinges upon many other ideas in texts concerning Jaume or Jacob Bertran as now follow.

In HOC 1, c19, page 433, note 432 the following can be read;
I develop this dating and age problem later as it impinges upon many other ideas in texts concerning Jaume or Jacob Bertran as now follow.
In HOC 1, c19, page 433, note 432 the following can be read;
Cortesao, History of Portuguese Cartography, 2:212 (note 3); An even more tentative identification is suggested by an unsigned entry in the Diccionari Biografic, 4 vols (Barcelona:Albert) 1966-70, vol1, s.v. Jaume Bertran. This Bertran was a mariner who captured a pirate off Majorca in 1453. Another scholar records a Jacme Bertran as patron of the Majorcan Galley both in that same year and in 1455; see Carrere, Barcelona 2;638, 926 N I (note 285). About the middle of the 15th C, Jacme who belonged to a family of converted jews settled in Majorca and Valencia, went to live in Genoa (Carrere, Barcelona 2;584).

In 1456 Jachobus Bertran signed the first of his 3 known charts, Although the surname was a fairly common one at the time, the possibility that the chart maker was a seaman finds an echo in the suggestion by Rodolico that the prefix “Mestra” on his 1489 chart might denote (as with Juan de la Cosa, 1500) the status of a pilot. See Niccolo Rodolico, “Di una carta nautical di Giacomo Bertran, maiorchino” ; Atti del III Congresso Geographico Italiano, Florence 1898, 2 volumes (1899) 2; 544-550, esp 545.

A similar suggestion is made in an unsigned entry in Enciclopedia universal ilustrada Europeo-Americana, 70 vols and annual supplements (Madrid and Barcelona; Esoasa-Calpe, 1907-83), 68:1187, that the Matias Viladestes who commanded a galley belonging to Frances Burges, 1415, should probably be identified with the Macia de Viladestes who signed the 1413 chart. In the same work (66:838-39) Gabriel Valesca is also termed a navigator.

In the following text I clearly indicate the timescale and thus improbable idea that Jaume or Jacob Bertran was a seaman or pilot, and that he should be addressed as Jacob.

To continue, the normal attribution is also given apropos where the chart was drawn, “ sxanch cartam in civitates barchias”, which we may take as their “slang” name for Barcelona and could be a contraction of one of many variant names for the city in the middle ages.

Thus a curious situation arises in that the only knowledge we have of Portolan Charts being drawn in Barcelona is in c1400 when Jafuda Cresques and Francesco Beccari were supposed to draw four charts; only one was completed and both departed, one for Majorca the other for Savona/Genoa. Another text suggests that two charts were in fact completed.

Did the Pattern/Template of one or both remain in Barcelona to be used again in 1456? It should be noted that the Crown of Aragon in 1375 and afterwards used Majorcan based cartographic draughtsmen, Cresques Abraham and Guillem Soler, and not residents of Barcelona for their Charts and Atlases.

However, if we study the “references to nautical cartography in medieval documentation” as set down by Ramon J Pujades, LCP, pp428-431, covering a period from October 6th 1315 to December 4th 1453, basically Catalonia/Barcelona, we read of 73 “charts”, 2 “portolans” and 17 “mappa mundi”. But, what is a chart; basically it could be a single small sheet covering a specific part of the Mediterranean Sea; what is a “portolan” in this context and just what constitutes a “mappa mundi”. The correspondence from June 7th 1399 to May 28th 1400 regarding the Cresques/Beccari work states 7 or more Mappa Mundi and we have the maps of those individuals to judge what they were. Then in 1410 the Royal household has a mappa mundi, a portolan chart and an atlas, thus the possibility of Bertran/Ripoll being able to copy a chart and not rely on the use of a Pattern/Template is there, but is it likely?

Then in 1478 the advent of the Spanish Inquisition begins and the expulsion of the Jewish population begins. We know this expulsion came later to Majorca and Jac Bertran signed two charts, one in 1482 and the other 1489 as drawn in Majorca and thus was probably a “converso”, as with so many other Majorcans, Jewish cartographic draughtspersons.

To investigate this chart fully and then make the necessary comparisons I will first quote page 493 from LCP by Ramon J Pujades;
“The case of Juame Bertam is somewhat different, since he has left us several charts signed during the second half of the 15th century101. Only one of these belongs to the period covered by this study (pre 1470) namely the chart he made in Barcelona in collaboration with Berenguer Ripoll in 1456. Although we cannot be absolutely certain as to the contribution Berenguer Ripoll made, I believe we can attribute the decoration of the chart to him and exclude the toponymical lettering, given on the one hand the disparity of styles between the ornamentation of this work and that of the other two charts signed by Jaume Bertram, and on the other the fact that the lettering of the toponyms is by the same hand as that of the chart signed only by Jaume Bertram in 1489, now preserved in the Biblioteca Marucelliana in Florence. Whatever the case, his cartographic model was obviously Beccarian (although his technique was poor) is revealed by the rest of his extant works”

(101= Greenwich NMM G230 :1/7 (1456): Florence AS CN7 (1482) and Biblioteca Marucelliana s.s. ( 1489))

That text of course raises a considerable dilemma; Battista Beccari was in Genoa and moved to Majorca, producing a chart there in 1435, that is 21 years in advance of the chart under consideration. If Jachobus Bertran was on Majorca and trained there as a cartographic draughtsman he would have of necessity been at least 21 years old in 1456 in order to comply with the ages for childhood and training in the art, but he would also have required several years to establish himself in Barcelona and meet Berenguer Ripoll and thus be at least 25 years old and capable of instructing a “disciple”. That would mean he was at least 60 years of age in 1489 and still drawing! Is that why the comment of poor technique was made? But Jacob Bertran died on the 7th September 1500, and by then would have been c71 years old and this is the typical lifespan of the age as my previous texts have clearly shown and thus he was a youth in Barcelona. If he was the Patron of a galley in 1455, his age and sudden appearance in Barcelona, ready to take on a “disciple” and draw a chart beggar’s belief and thus I conclude given that the Bertram surname is quite common there are several differing personalities in the storyline.

Jacob Bertran was resident in Majorca and had a residence in the Parish of Santa Cruz where he lived with his family who had difficulty with their paternal inheritance as documented in “ARM. Prot. Nic. Costa, n. 4494, f.45”. There-in the son,Johannotus Nicholaus Bertran, filius Jacobi Bertran buxolerii quondam,

discussed the distribution of his father’s goods. I think the son would have described his father correctly and thus conclude that Jacob Bertran learnt his art as I have described as a cartographic draughtsman and thereafter had the time to make acquaint himself with Barcelona and take on a disciple.


The whole chart is as ChGME/1/APPD01, but I have redrawn it and have appended various distance measures and graticular lines there-on which indicate clearly via the scale bar units it is a Miliaria scale.(Diagrams ChGME/1/APPD04 & ChGME/1/APPD05). But firstly Diagram ChGME/1/APPD06 illustrates that the Windrose is a simple expansion of the 92 units and thus we have 14 x 92 or 1288 scale bar units per quadrant. This produces from the 35/30/20/7 subdivisions distances of 490/420/280/98 scale bar units or Miliaria and is so very easy to draw. But was it drawn thus?


On Diagram ChGME/1/APPD06 from the major lines of the wind rose graticule it can be shown that in fact it is a geometric construct based upon the four square Windrose which itself then produces the north/south distance between the scale bars and the east/west extensions through simple alignments. The outer scale bar frame is thus dictated by the inner workings of the Windrose. But it is obvious the vertical centreline was mis-placed and thus a non-symmetrical layout was achieved and by accident no doubt this mis-placing led to the alignments from the centre point to the north-east and south-east corners of the overall diagram being used instead of the actual 45 degree alignment which is the natural wind direction for Sirocco and Greco. Thus to the west we have; NW = Maestro; W = Ponente; SW = Libeccio and of course although not identified, N= Tramontana and S= Ostro, all correctly positioned. But, Greco to the NE and Sirocco to the SE are misplaced wind circles, and that was a very simple draughting error to make.


The basic latitudinal measures vary from 68/72/86/90 Miliaria on the Atlantic coastal area, but the 90 Miliaria measurement continues easterly gradually expanding to 100/110 miliaria with the continual problem on Portolan charts of the Aegean Sea from 36N to 40N always drawn as 5 degrees latitude, that is 4 x 110 Miliaria ( 5×90). But importantly the west coast of the Iberian Peninsula is for once drawn correctly to scale as a proper latitudinal degree.

However when I started to compare this chart to contemporary charts thus possibly establishing both antecedent and descendants, the distance measures of the Iberian Peninsula indicated that at last perhaps a cartographic draughtsman had realised the original error made over a century ago from a mis-interpretation of the Roman Mile latitudinal degree and its Miliaria counterpart.

That change from 75 to 90 scale bar units is amply illustrated on the preceding diagrams ChGME/D32/33 where Charts LCP C58 and LCP C22 are compared. Thus it is necessary to investigate if this correction is nothing more than an aberration or an intentional redrawing of the Peninsula to a correct set of distance measures.

Given that it is suggested that the Bertram/Ripoll chart is aligned to the Battista Beccari chart of 1435 which was drawn on Majorca when Beccari first arrived, we can overlay the two and indicate just how close they are or are not. Diagrams ChGME/1/APPD07 & ChGME/1/APPD08 apply. They certainly are in all probability drawn from the same Pattern/Template with great similarities, but, as the Bertran/Ripoll chart was executed in Barcelona, some 21 years after the Beccari chart, the likelihood of an intervening copy of the Pattern/Template is high.



To indicate just how extraordinary this chart compares with the Iberian Peninsula drawn correctly dimensionally, I have matched the scale bars to a geographical chart of scale 1:10million and used a Conical with two standard parallels geographical plot to mimic the distortion inherent in the Portolan Charts. That distortion actually stems from the Iberian Peninsula having its west coast drawn as 6 x 75 Miliaria latitude degrees instead of the 6 x 90 Miliaria we are now discussing. The slew is no doubt from the original Roman Template with a 12 degree Ecliptic slew. Diagram ChGME/1/APPD09 indicates the resulting overlay.

However, we can only answer the question concerning the change from 75 Miliaria to 90 Miliaria being a deliberate act, a known problem solved, by looking at the later charts drawn in 1482 and 1489 by Jac Bertran to ensure continuity of measurement.

As I was not convinced of the link to Battista Beccaria, even though I could show that the two charts shared a great similarity and thus possibly a joint Pattern/Template it was necessary to test the 1456 chart against another and I randomly chose the chart of Mecia de Viladestes, 1413 chart, BnF AA 566 as the check chart. That chart is shown on Diagrams ChGME/1/APPD10 , ChGME/1/APPD11 and ChGME/1/APPD12. The comparison can be viewed on Diagrams ChGME/1/APPD13 and ChGME/1/APPD14. They show both similarity and the disparity of the Iberian Peninsula having two latitudinal measures, 75 and 90 Miliaria per degree. The residue of the chart indicates clearly a similar Pattern/Template as has been found in these researches.




The collection in the State Archives was formed in the second half of the 19th century after Grand Duke Leopold II of Lorraine had founded it in 1852, for the Central Archives of the State of Florence to assemble in one place and to preserve the historical archives of the Tuscan State. As a consequence, all nautical charts, at the time scattered in several important archive, were gathered here to fulfil the needed conservation.


This chart is reference CN7 and is 1023 x 668mm and is similar to the 1456 chart except that where-as the 1456 chart has a multitude of vignettes (? drawn by Berenguer Ripoll) of cities and no life like figures at all, (which begs the question who it was drawn for), the 1482 chart has various Kingly figures seen on so many Portolan charts and the vignettes of the cities are well drawn. In the years between 1456 and 1482 it is quite open to opine that Jacob Bertran could have become an expert miniaturist, or, that Berenguer Ripol was actually no more than his first apprentice, a disciple if you wish, and assisted enough to be named on the chart, but took complete instruction from “Jachobus Bertran” who carried out most of the drawing work on the chart. I therefore question the previous researcher’s interpretations.

The inscription reads; “Jac Bertran en Malorques al fe …. L’any MCCCCLXXXII.


If we now measure the west coast of the Iberian Peninsula from 37N to 43N the latitudinal degrees are 90 Miliaria which follows through to the 9E longitude and also from 31N to 36N in the eastern Mediterranean Sea. Hence we are looking at a different Pattern/Template to that used in Barcelona 1456AD, but only by addition of the northern Baltic Sea area. But, being 26 years after the 1456 chart as already speculated there could be one or two redraws of the Pattern/Template with the Baltic Sea additions and of course Hibernia included. This information was readily available on Majorca to make the adjustments and produce a full Portolan Chart of the Mediterranean Sea Basin and the North.


A comparison between the 1456 and 1482 charts show that the two charts have the same Pattern/Template as illustrated on Diagrams ChGME/1/APPD17 and ChGME/1/APPD18.


However we now have a serious problem apropos the journey of the 1482 chart from Majorca to Florence and the State Archives.

The whole of the southern Mediterranean coastline of this chart, starting in Africa and then wending its way via the Levant through Turkey is covered with Arabic script no doubt denoting the translation of the Catalan text into an Arabic language. When or how was this chart available to an Islamic/Arab person to endorse it as we now read? The fact that it is covered with non Islamic features, that is the human form which have not been erased is possibly a pointer to an educated person who understood the value of the chart, its accuracy and the fact that few Arab/Ottoman charts, realistic charts were actually available.

There are several known Ottoman charts, Portolan charts that compare with the Genoese and Majorcan genre. They date from 1413, 1461 and 1570, not including the Piri Re’is chart.


The basic structure of this chart is as per the 1456 and 1482 charts in that it is quite obvious the Pattern/Template is virtually the same as the overlay between the two clearly confirms. The actual coastal features being one and the same, but, obviously not identical as each chart is an individual creation even though based upon a single Pattern/Template.

Therefore from 1456 to 1489, Jac Bertran maintained his basic tool in a state that was capable of producing a very good example of a Portolan chart, even though I have surmised it was probably redrawn to maintain that structure.

However it is much more than a chart, it is an expression of Jac Bertran’s skill as a miniaturist with 15 oversized vignettes of cities, all beautifully drawn as well as the Madonna and Child adjacent to his attribution text. A very “Christian” feature for a “Jewish” person, which could be interpreted as a nod towards the client particularly as the only human figure drawn is the Spanish Monarch, King of Castile, complete with Family crest shield and crown.



However this chart is dated 1489 and the historical facts require elucidating. In Spain the union of Castile and Aragon took place in 1469; the Spanish Inquisition and expulsion of the Jews started in 1478 but it was not until 1492, January 2nd that the Reconquista is complete with the conquest of the Muslim emirate of Granada. The Kingdom of Castile held nearly all Spain (except Aragon and its territories), thus we see that Jac Bertran has anticipated the downfall of the Emirate of Granada and with the “Christian” iconography perhaps we can assume this chart was bound for a client in Spain as many of the charts drawn in Majorca were in times past.

I have compared the 1482 and 1489 charts as Diagrams ChGME/1/APPD19 and ChGME/1/APPD20 indicate. Again it clearly shows that Jac Bertran had kept his Patter/Templates even though he amended one for the 1482 chart to show the Baltic Sea area.

NICCOLO RODOLINO, comments upon the 1489 chart, in an 1898 text.

Di una carta nautical di Giacomo Bertran, maiorchino,
This text is similar to most cartographic historians texts and merely describes the content in some detail.
The discussion apropos “Mestra Jac Bertran en Malorque La Feta en Lany MCCCCLXXXVIIII, wobbles between the “Master” being a cartographer and a Ship’s Officer as he compares it to the attribution on the J. de la Cosa chart of 1500 which he states has, “Maestro de hacer cartas”.

He also comments upon the port of Varna on the Black Sea being at the same latitude on the chart as Venice, but states correctly they are 2 degrees apart. This he puts down to the lack of knowledge of the declination of the needle of a compass. He also states that the needle of the two wind rose compasses on the main wind rose centre line are coloured Blue to indicate perhaps the tip of the magnetized needle. Thus we may conclude that He was a man of his age who had not, as so many have not, dissected the chart correctly and even determined the scale bar as Miliaria.


Having produced overlays of the various charts and confirmed they are drawn from the same or very similar following Pattern/Templates, it only remains to discuss the changes in the depictions on each chart as they are perhaps the most intriguing features.

The 1456 chart (APPD01) is very plain containing only vignettes of various cities and in the northern section what may clearly be stated as the same city drawn at least 8 times to infill a rather large void in the north devoid of rivers, mountains etc. Genoa and Venice are by far the most impressive vignettes even though Venice appears to be a luscious land-based city and Genoa is given a magnificent harbour. But look at the buildings; they are all so very similar in design to the extent that even Cordoba/Granada in Spain looks like the northern cities. It is therefore open to opine that perhaps Jachobol Bertran carried out the majority of the drawing work and gave his “disciple” Berenguer Ripol a building design to copy many times over onto the chart to hide the open spaces and hence it could be argued that both persons produced the chart, with the student doing the “donkey” work.

The 1482 chart (APPD02) drawn after a gap of 26 years, a gap in his life story for which we have no knowledge except that he returned to Majorca and started a family. Here we see a blossoming of exotic design, warrior Kings with sword in hand; the Baltic Sea with a dozen vignettes all similarly drawn cities on its southern shore and many other city vignettes scattered over the chart. In fact the chart is so full of cities and people that the northern section is not easy to read at a glance, but requires detailed study.

At this stage in my research I opine that Jacob Bertran was at the peak of his career as a cartographic draughtsman, which at the projected age of 60 years is no mean feat. But, at least being on Majorca he did have a wealth of historical draughtsmen and their excellent charts to follow. I am not enamoured with the comment that this chart is perhaps not very good and find it compares well with the others of the age.

The 1489 chart (APPD03) is more subdued having only major city/places vignettes with flags at the other important cities. There is only one figure, the Spanish Monarch within the chart, but a well drawn Madonna and Child are set adjacent to the attribution panel. The chart is covered in toponyms, probably far more than normal and its state of preservation allows for minute details to be readily seen. But again it is devoid of the northern lands and the continuation of the West African coast south from the strait of Gibraltar. It is certainly meant for a “Christian” and appears to have travelled little in its life time, hence its remarkable condition.


The Al-Furqan Islamic Heritage Foundation was kind enough to assist me in the possible provenance of the over-writing to establish the charts travels from Majorca to its resting place in Florence. This is a copy of those notes;

The Arabic script is Naskh (not Magrebi) which indicates a Middle Eastern provenance of the notes. The language is classical Arabic, and even the correct diacritics are kept as they should (so not much influence of Middle Arabic). The notes could have been most likely made by someone from Syria, Egypt, Lebanon or Palestine. However, mobility of books and travellers was quite a common phenomenon in the area, so based on the script it is difficult to pin point exactly which country this came from, as during the Ottoman period there were no nation –state like boundaries, but provinces of the Empire. For the time-frame, the term “Sultan” for the Sultan of Egypt that is used in the map is an Ottoman term and the Ottoman Sultanate started in 1517 in Egypt. On the chart, Sultan Misr = Sultan of Egypt and Samsun = Samson, as examples.

Naskh is a sans-serif script meaning characters lack “hooks” on the ends of ascending and descending strokes. For example the alif is written as a straight stroke, bending to the lower left. Naskh differentiates various sounds through the use of diacritical points, in the form of 1-3 dots above or below the letter, which makes the script more easily legible. Naskh uses a horizontal base line; in situations where one character starts with the tail of the preceding letter, the base line is broken and raised.

In 16th century Constantinople Seyh Hamdullah (1429-1520) redesigned the structure of Naskh, along with the other “Six Pens”, in order to make the script appear more precise and less heavy.

Thus we may consider our chart travelled from Majorca through the Middle East well after it was drawn to arrive in Italy, Florence by the 19th century.

M J Ferrar, June 2019.