View Diagrams  Download Paper  Download Images 


Within several texts I have endeavoured to introduce my research findings which indicate that the Portolan Chart is not what has been determined over the decades by historians and researchers. I have questioned the charts usage, their design and thus the charts usage of magnetic deviation as determined by previous researchers. I have also shown that the Wind Rose is normally drawn foursquare, NSEW, on each chart and as such is a primary facilitator of its draughtsmanship and measurements, but the “LCN” errors indicated a spurious east which was copied and thus indicates little real knowledge of the geographical situation extant. In other words, the west is geographically correct but the east is distorted.

Within text ChCPC/1 I illustrated conclusively that there was no magnetic deviation in a charts construction, only a distortion caused by the scribal errors within “LCN”, primarily south of Sardinia but also in the eastern Mediterranean Sea basin. These errors actually caused a seismic shift in the drafting of the charts to confuse latter day researchers. This of course persuaded “non technical” researchers of a magnetic bias as the putative historical dates for the discovery and usages of a magnetic compass were supposedly similar.

Nothing was further from the truth, not only was the magnetic compass originally a magnetized needle in a straw floating on water and unfit for any accurate or indeed long term usage, but the basic accuracy even when developed was insufficient for map readings, for sailing Peleio routes, and, the accuracy of distance measures would take centuries to compile for so many Peleio routes to exist in the literature.

This text focuses on three charts, Cortona, Vesconte 1311 and Dulceto 1330 and uses the Riccardiana MS3827 chart as a median to illustrate that an original “LCN” document in its Peleio section provided the basic data for the charts outline construction. The charts are discussed in the order set out in “Les Cartes Portolanes” as C2, C5, C7 and C8.


Being a chart with little history in textual terms it falls to draughtsmanship and mathematical analysis to uncover the basic facts concerning the chart; indeed I have found only one major paper regarding this chart. Although clearly reduced in physical size it would have been a typical Mediterranean Sea basin map with overall scope from Cape St Vincent to the eastern coastline of the Black Sea, some 9W to 42E longitude. In the south there are remnants of the Gulf of Sirte, thus it is latitudinally 30N to the Sea of Azov at 48N. It has two Windrose circles, the junction of which may be fairly described as the 15E longitude. The eastern circle is complete and extends to 35E. Curiously it has a centre line from East to West which could be considered the Gozomalta to Acre alignment. We can thus speculate that the western circle is from 15E to 5W, perhaps the Pillars of Hercules. It is also possible to speculate that the circles were drawn after the chart was prepared and they were thus positioned to suit the “LCN” text, (which I have already noted as a possibility on the Carta Pisane chart), thus following the east/west alignment from Acre to Gozocreti to Gozomalta would be a simple matter, but it is not true east/west, that is San Piero to Grapparola!

There are two possible scale bars set upon the chart. The central circular design is divided into 5×20 and 2×50 millara segments set vertically in the design. The second scale bar is across the north of the chart and a different matter entirely. By measurement it has a relationship to the circular scale such that c17 of the northern units is equivalent to c14 units of the circular scale. Thus we have possibly a simple 6:5 ratio of the scale bars. I have already opined that the millara is 5/6ths of a Roman Mile of 1.47911Km and is thus c1.23 Km nominally. Therefore we can establish from the Cortona scale bars confirmation of that relationship as well as being able to opine that the northern scale bar is none other than a scale of Roman Miles (mpm). It has been suggested that the derivation of the Millara is from the Roman Pes unit of 12 uncia, making the basic unit of measure 10 uncia or 5/6ths of a Pes. I do not subscribe to this theory as the measurement is too small to be of any use at sea and thus requires being a large multiple that can be easily established and measured when sailing.


Firstly, a comparison to “LCN” distance measures will establish if the Cortona Chart is compatible thereto and determine a basic accuracy for the chart. The four diagrams all have drawn thereon “LCN” alignments from the Peleio section and also have “LCN” chart and geographical distances noted.


But from the very commencement of the research it became apparent that the Cortona chart was in one major instance using not only different Peleio distances, but by either deliberate action or serendipitous actions the errors which exist south of Sardinia had been corrected. The correction is partially made by using different geographical distance measures south of Sardinia compared to the extant “LCN”, but also that the overall length of Sardinia is increased by 35 millara to partially mitigate any “LCN” shortfall in the distances. Sardinia itself is N/S 215 millara but the chart has 250 millara, hence we record 690 millara along the 9E longitude where actually it is only 660 millara. But if we tabulate the distances, those given in “LCN”, as drawn on the Chart and the Geographical distances we can observe the accuracy and the chart differences.

San Piero to Boczea; CCCXX 310 (-) 310 (-)
San Piero to Bona; CLX 160 (-) 210 (-50)
San Piero to Galeta; C 125 (+25) 150 (+25)
San Piero to Bizerta; CXL 185 (+45) 200 (-15)
Taulara to Bona; CLX 150 (-10) 193 (-43)
Carbonara to Bizerta; CXX 170 (+50) 160 (-10)
Carbonara to Cape Bon; CLXX 205 (+35) 210 (-5)


Hence the chart plot from Genoa via Corsica/Sardinia, Galeta to N Africa is drawn as 690 millara when the geographical distance is 660 millara and akin to F Beccari trying to correct the same error within his 1403 chart and increasing the distance by 50 millara.

However, if we look at distances to ascertain the overall accuracy of the Cortona Chart it proves very accurate indeed. Thus from Aigues Morte/Menorca/Zaffone, total distance drawn is 600 millara, the geographical distance. I have already indicated the 9E distance as 690 millara on the chart and 660 millara geographically, but from Civita Vecchia to Tripoli in Barbaria is 1050Km, 854 millara; the chart is 850 millara. Also from Castello Roso to Alexandria; chart and geographical distance are equal at 450 millara.





The geographical test is simply that Capo Corso is 43N and thus an excellent locating point, but, more importantly by projecting it westwards the 43N also denotes the Iles D’Hyeres and is the situation on the Cortona Chart. Thus below 38N we have consistent latitudes which vary as they cross the Ionian Sea to the Peloponnese, and below 36N an accurate chart again. The longitudinal lines are however set 6 degrees east to the putative 36N latitude, but are consistent across the chart. A general calculation indicates a 33:40 ratio longitude to latitude which equates to a latitudinal centre line of c35N. But remember, the ratio for 36N has always been thought of as 4:5 in antiquity and that equates to 32:40, so very close.
But as I have already opined, it is the distance measures, probably set out from Genoa, 9 East via the Peleio, which determine the putative ratio which were originally measured on the surface of the globe and thus should reflect the actuality of the geographical ratio.



The basic conclusion regarding the Cortona Chart is therefore that it is a highly accurate chart, which either by luck or judgement has a draughtsman who has either corrected the “LCN” discrepancies which mar the standard Portolan Chart, or had another copy of “LCN” which displayed the more correct distances involved. The distortion is obviated and thus the reasoning for magnetic deviation does not exist on this chart, as with others investigated.
Would that it was a full chart, it may have held even more surprises!


The chart illustrates only the central and eastern sections of the Mediterranean Sea and Black Sea; it is thus c8E to c42E longitudinally with the main area from 8E to 36E, the Mediterranean sea, Mare Nostrum. A calculation of the length of the chart from Gozomalta to the Levante coast along the 36N latitude indicates the chart distance is 1578 millara and the geographical length 1588 millara. Measuring from Gozomalta to Acre we have via the chart 755 + 794 = 1549 millara which is 730 + 850 = 1580 millara geographically. It is worth noting the Riccardiana Chart is 1585 millara. The latitudinal check, Genoa to the North African coast indicates a shortfall of 20 millara, but that hides the fact that Sardinia to N Africa has a shortfall of 40 millara and hence the cause of the chart distortion. In fact as has already been shown previously and is now confirmed on this chart that the distance measures south of Sardinia account for the whole distortion problem.


The method of presentation by Vesconte for this chart affords an ideal opportunity to study its construction from first principles. The chart set upon rectangular vellum terminates in the west primarily at the bordure, whereas in the east the Black Sea penetrates both the east and north bordures, with the linguetta still in place. The horizontal lines of the graticule are parallel to the north and south bordures which also form the limit of the Windrose.


A major indication is the fact that the Windrose has no designation for orientation; the only clue is that it parallels the Gozocreti to Acre alignment which “LCN” indicates as Levante.
“De lo dicto Gozo enn’Acri DCCCXX millara per Levante”. But there is also;
“De lo dicto Gozo de Creti a lo Gozo de Malta DCCC millara per Ponente”.
However, the latter is actually drawn at 775 millara and definitely “Ponente ver lo Garbino poco”, or “Ponente ver lo Libeccio poco”. This is closer to geographical fact.



The background graticule is shown on diagrams ChMAT/1/D10 and D11. It is worth noting here that as with the Cortona Chart, which I believe has had the Wind Rose added, it has used the same alignments for setting out purposes. I therefore opine that these earliest charts were drawn sans Windrose definition until it became apparent they were totally at odds with the actuality of the geographical winds and thence were all drawn correctly NSEW.


Thus in 1318, Pietro Vesconte produced his iconic diagram of the Windrose and its construction ensuring it was set out NSEW, or should I say denoted NSEW to match the correctly known alignment from San Piero to Grapparola.



However, Pietro Vesconte at this point obviously knew that his graticule and thus the Windrose was erroneous being set out to “LCN” east/west as determined by the Gozomalta/Gozocreti/Acre alignment. At the western edge of the chart are two parallel lines, although they do not quite overlap to indicate that fact on the chart. But they are the Geographical North/South alignment and are set out to the correct Windrose distances of 270 and 300 millara. This can only indicate that Pietro Vesconte required acknowledging the true north alignment and thus ensure it was recorded for proper usage. Hence his later diagram!



Thus we may postulate that the “LCN” text was to hand when the chart was conceived and then drawn. But how was it drawn; certainly not from the western edge as the natural North/South alignment from Genoa through Corsica and Sardinia to N Africa, the 9E longitude is twisted approximately 9 degrees to the west. It is however approximately aligned to the putative 36N latitudinal line which can be drawn on the chart.


Thus we may postulate that the “LCN” text was to hand when the chart was conceived and then drawn. But how was it drawn; certainly not from the western edge as the natural North/South alignment from Genoa through Corsica and Sardinia to N Africa, the 9E longitude is twisted approximately 9 degrees to the west. It is however approximately aligned to the putative 36N latitudinal line which can be drawn on the chart.

Therefore I must conclude that the chart was set out from the east, no doubt the Levant shoreline and Acre, and then constructed via some of the “LCN” Peleio distances and directions for the eastern Mediterranean Sea. The diagrams illustrate how these Peleio’s interlink to denote major coastal features and the twisted 9E longitude and 36N latitudinal lines are merely a result of the erroneous distance/directional details given in “LCN”.
It is patently obvious if you align a chart from the east or the west they will differ!
The graticule scales 1854 millara, North/South and thus is in all probability 1840 or 2 x 920 millara sections ensuring the geometric quality of the wind rose based upon those 92 units forming, 35:30:20:7 ratio sections of the 22 ½ degree circular segments.
Thus regardless of orientation the graticule is a standard format Wind Rose as portrayed by Vesconte himself in the 1318 Atlas.


The Pietro Vesconte Chart dated 1311 is drawn using data from an “LCN” text available. That P Vesconte realised there was a problem with the orientation has been clearly shown, and the fact that the distances vary enough to indicate a text variant, but, the wind directions appear constant at this period indicating no revisions of note around 1300AD.


As with the two previous examples investigated, these three diagrams have both the chart distances and geographical distances appended to illustrate the Peleio of “LCN”.
However on this occasion I have also appended the geographical 36N and 9E alignments to illustrate the overall accuracy of the chart and any distortion caused by the erroneous distance and directional data given in “LCN”. Those alignments are used to quantify the overall chart accuracy as follows. Firstly, from Gibraltar to the 9E longitude is 14.33 degrees, of which each degree at 36N is 73 millara, hence a total distance of 1046 millara. The chart scales 1035 millara and is thus quite perfectly drawn in this section.


If we then measure the distance from the 9E longitude to Gozomalta, the chart is 342 millara, but geographically it is 383 millara, a 40 millara shortfall. By studying the diagrams the shortfall already discussed in the previous examples and more importantly within text ChCPC/1, between Sardinia and N Africa of some 50 millara, the reason for the charts distortion becomes so very apparent; 40 shortfall E/W and 50 N/S, and a marvellous slewing of the chart. But certainly nothing to do with a magnetic deviation as other researchers indicated.
The distance from Gozomalta to Gozocreti, geographically is 742 millara and scaled from the chart it is 730 millara, an acceptable difference. Thence from Gozocreti to Acre, geographically it is 850 millara and scaled it is 839 millara, absolutely acceptable, as before.
But if we calculate from Gozomalta, at 14.25E/36N to the Levant shoreline at 36N/36E, geographically it is 1588 millara and chart scales as 1626 millara, c40 millara error. However return to Gibraltar at 5.33W/36N and calculate to Gozomalta at 14.25E/36N, a total of 19.58 degrees at 73 millara per degree we have a total of 1429 millara and the corresponding chart sum is 1035 + 342 = 1377 millara, a 42 millara shortfall as already shown. But calculate the Gibraltar to Levant distance at 36N and it is 41.333 x 73 millara = 3017 millara and the chart scales 1035 + 342 + 1626 = 3003 millara. It is in part accurate!


We may thus dismiss the minor aberrations of measurement within the charts longitudinal distance and accept it as accurate for our purposes.
Latitudinally it is a different picture, but we have several check lines to gauge the charts accuracy north/south. Firstly from Aigues Morte to Maone to Zaffone, we have the following totals; chart = 597; geographical = 576 and “LCN” gives 585 millara; thus acceptable. The most important check line is along the 9E longitude from Genoa to N Africa which is shown on the diagram both segmentally and as a total distance. Overall it is; chart = 613 millara and geographically it is 660 millara and thus we have as already shown a 47 millara basic error which raises the N African coastline and distorts the residue of the chart to the east and to Acre. The next check line is from Civitavecchia to Tripoli in Barbaria; chart = 775 millara but geographically it is 854 millara and the consequential damage of the distance measures south of Sardinia are so very apparent. The next is from Castello Roso to Alexandria and here the accuracy returns as the chart has 455 millara and geographically it is 450 millara.
Thus in the western Mediterranean Sea there is an accuracy of distance measure and direction, none of which is affected by a magnetic deviation, they are geographical. In the eastern Mediterranean Sea area we have the slewed alignment but again the distance measures are generally acceptable.


Thus without discussing the marvellous presentation of the whole chart, it can be shown that the basic Mediterranean Sea basin accords to the “LCN” document for distance and direction data. However the chart has much more than that area and how the data may have been collected and the accuracy thereof will be the subject of a separate paper. I have already opined that the Dulceto Atelier was absorbed into the Cresques Abraham Atelier and then the Catalan Atlas, which bears a remarkable similarity to the Dulceto 1330 chart, produced. Thus we have the opportunity to discuss the ongoing collection of coastal data for these charts and how each may have been actually no more than a copy of an ancient original.


First and foremost it is necessary to deal with the charts scale to enable the distance measures to be gauged accurately. There are four scale bars appended to the chart set as follows, NW/NE/SW/SE, but they are inconsistent in their finer detail and thus it is doubtful that they were actually utilised to draw the map but are a later appendage. However a careful examination of those scale bars elicited a measurement which can be used upon the chart to measure inter-distances. An assessment of the accuracy of that judgement can be made from and by the following text.



The basic construction is a double wind rose with dimensions of 920 millara per radii. Thus the 36 squares which form the wind rose are naturally; 350 x 350; 300 x 300 and 270 x 270 millara. The hidden though semi visible wind rose circles seen on the Les Cartes Portolanes scan are focussed around the central point of the chart, which itself equates to the c14E longitude. They are also visible in various other positions on their circumference, and careful examination indicates a likely continuous circle, although it is not necessary.



The draughtsmanship of the chart is accurate such that it is possible to reproduce the first lines drawn to ensure the charts consistency and dimensional integrity. The chart has been set upon the “skin” in a symmetrical manner; that is the latitudinal centre line is central to the skins breadth, but also the overall length, the longitude is correct. But, was this the original size of the skin as the parallel sides to the north and south appear to have been cut to suit? Was it nailed to a board at some time during its life span and the edges trimmed to remove the obvious signs, but that’s speculation?


Thus we can estimate the longitudinal/latitudinal extent of the chart. The western perimeter is several degrees west of Cape St Vincent (9W/37N) and the eastern limit is several degrees east of the Levant coast and then also several degrees east of the eastern Black Sea coast. I assess that the chart covers basically from 11W to 43E, some 54 degrees of longitude, which is a distance of 3940 millara at 36N. The latitudinal spread is from 30N to 52N, some 22 degrees and thus the chart has a ratio of 54/22 or 27/11 degrees, but 3940/1600 millara or c39/16. The centre line from 11W to 43E is c16E, but the basic framing of the chart allows for some 5.5 degrees east and 2 degrees west which thus offsets the centreline of the chart.
An important position in the Mediterranean Sea is Gozomalta (14.25E) and this denotes the centreline of the chart, and thus we can discuss the NSEW alignments. I have already indicated that the wind rose on a Portolan chart is drawn as if it is NSEW. But both in my ChLCN/1 and ChCPC/1 texts I have shown that the incorrect data within “LCN” used to draw the Portolan charts, negates any use of a magnetic compass and dictates that the direction from Gozomalta to Gozocreti to Acre is an incorrect east west alignment.
Study Ms 3827; Gozocreti to Acre is parallel to a wind rose east/west alignment, in fact the 350 millara south alignment. However Gozomalta is further south than that E/W alignment and as such mimics its geographical position. Also on the chart, Gozomalta is slightly east of Tripoli in Barbaria, not due south as the “LCN” text states.
Thus, is Ms3827 drawn from an earlier version of “LCN” than is extant now and has less scribal errors within it to distort the chart as much as others already discussed?
If we therefore study the Grapparola to San Piero alignment, given in LCN as either East/West or Garbino per Ponente pauco (as previously discussed) we see a slight deviation of the line to the south of west which would be Ponente ver lo Garbino pauco.
But, we must recognize that this East/West alignment is geographically correct and also is correct on these charts. Thus the basic Windrose is correct and the “East” is wrong!
However the right angled triangle, San Piero/Maone/Marseille is drawn correctly at 300 millara per side and 454 millara from San Piero to Marseille. The import of this and other triangles formed by the Peleio routes is fully discussed in text ChCPC/1, where a complete comparison to a geographical map, the “LCN” and MS3827 is carried out.
The diagrams are a simple overlay of geographically plotted “LCN” distances and MS3827 to illustrate the close comparison of the chart and an “LCN” base. To achieve the overlay it was necessary to twist the geographical “LCN” plot thus mimicking the distortion caused by the distance errors south of Sardinia. The comparison is there for all to read, and if required download the diagrams from my website, join them together via the marks and fully understand the overall picture. From such a viewing it will be possible to understand the western Mediterranean Sea accuracy and the eastern Mediterranean Sea distorted axis.

My texts already mentioned carry complete conclusive discussions for the research mentioned here-in. I do not intend repeating myself. But I am appending an essay, ChECP/1 which deals with the actual origination of the Portolan Chart, includes a timescale and my reasoning based upon the complete research package of my papers and their diagrams.

There are numerous papers on this website which investigate the internal workings, the cartography and draughtsmanship required to form a Portolan chart. Some are, ChLCN/1, ChCPS/1, ChCPC/1 and ChMAT/1.
This essay is a simple explanation of conclusions from research and thus my texts.
In the past I have endeavoured to lead researchers through the intricacies of the data required and the physical labour involved to produce a Map or Portolan Chart, for a chart is no more than a coastal map. The research indicated that all early maps were probably based upon a single (or copies of a single) “Alpha” map, and that this map morphed into a coastal chart with the addition of a Windrose, or multiple Windrose to suit the scale of the chart. Thus a basic premise I hold is that there were no Charts just Maps pre the early medieval period.
But from whence came the original map? What data was employed to produce it, and, how was that data amplified over perhaps centuries as the coastal names record, such that both Roman (towns no longer extant) and later medieval sites from the age of the great Italian City Republics or Kingdoms were upon one map? Thus we have a timescale from the end of the Roman period to those Republics circa 1000AD for the production of the “Alpha” map.


Texts describing sailing routes in the Mediterranean Sea basin, the Red Sea and Indian Ocean have been extant for nearly 3000 years. Some commenced as classical texts, others as aide memoires for mariners or as instructions for military purposes. But regardless of their origin they included distance and direction components. One centre which required the production of such data was Rome. We can readily understand that they were able to collect and use such data from a study of the Itinerarium Antonini Augusta and of the Tabula Peutingeriana both c4th Century. Their content illustrates quite dramatically the extent of surveying distance measures that had been undertaken by the Roman State.
Move forward some 600 years and add to that further directional data gathered as trade increased by sea, some of which may well have been previously available, and the origin of medieval texts such as “Liber De Existencia Riveriarum”, “LCN” and “Grazia Pauli”, dating from c1200 to c1300AD is apparent. Thus it is necessary to discuss the most extensive of those texts, “Lo Conpasso de Navegare” and its influence upon Maps and Portolan Charts.


The text we have is actually dated 1296AD but is thought by the original researcher to be a copy of an earlier text probably dated c1250AD. These two dates are actually irrelevant to the history of the Map or Portolan Chart as they are merely the dates of the extant version.
My investigation of LCN indicates it is of two different sections, clearly defined by the text, but mainly by the content and its presentation. The text commences in folio 1;

“Auesto si e lo compaso e la starea de la terra si como se reguarda en quante millara per estarea. Enprimamente da lo capo de San Vicenco a venire de ver Espagne, ver levante”, and continues thereafter to describe coastal sailing, “estarea or starea” , from the Spanish Coast via France to Italy and the Adriatic, and thence Greece to Constantinople, the Levante returning to Egypt, Libya, the Sicilian Channel and North Africa to the Pillars of Hercules at Ceuta/Gibraltar. The Starea section contains 73 Peleio descriptions over 100 millara distance.
We then read at folio 62v, “Lo complimento de volgere tucta la starea.”
“Hor e complito de volgere tucta la starea de la terra, coe a ssavere primaramente la Spagna, e Catalogna, e Provenca, e Principato, e Pullia, e Scravenia, e tucto l’altro mare entro a Saffi.”

The wording here is intriguing, in that the author appears to be stating that not all of that which follows is part of the original text but an addition to compliment the Starea sections. Thus we have 62 folios of Starea, coastal sailing and 44 folios of Peleio, long distance sailing instructions within the text.


I have already written concerning the “Starea” section of “LCN”, deeming it to be unfit for purpose in its present state. That is to say, not only would a mariner have difficulty using the data as the errors are in some places tantamount to being catastrophic for a ship, but that unless the included 73 Peleio details were completely integrated into the data a map could not be drawn there-from. They are not exactly helpful to a mariner when sailing as it is also evident that that within the extant 73 Peleio there are large errors.
However, my researches did find that the main Peleio section, probably in its latest format partly an appendix to the original “LCN” text, could be used to construct a matrix from which a map, a coastal profile could be constructed. That is of course if from the +300 “LCN” Peleio the pertinent routes could be established and the veracity of those routes proven. But one factor of the Peleio section always caused a frisson of doubt in the early research. (Bear in mind this research is spread over at least two decades in a fragmentary manner, as I travelled and worked across the globe.) Initially, I could not grasp why so many minor and some inconsequential islands and capes were the focus of a large percentage of the Peleio data. Reading the “LCN” text provided no clues at all. The only link I could find was the Roman usage of many islands, sometimes for crops if they were fertile, sometimes as purely observation posts to observe shipping, but always likely to have been part of the original Roman Maritime Itinerary as they sat in the main shipping lanes of that period. Thus I could consider that they were original Peleio points and as shown were expanded to the overkill situation we now read.


Thus I decided ( against my better judgement, as the workload would be prodigious), that the only method to solve the conundrum was the recourse of drawing everything as described, and investigate these islands, so prominent in the Peleio section text. What I found was an amazing array of interconnected Peleio routes with some individual islands having from 10 or up to 20 assigned routes. The three included diagrams indicate the western Mediterranean Sea area.




These routes formed an impossibly complicated matrix, and it soon became very clear why they had been chosen from the myriad of islands available, as opposed to major ports and other more well known features. They were actually situated in the most prominent of positions geographically, were therefore perfect Roman lookout posts and could thus provide clear alignments across the various parts of the Mediterranean Sea basin such that a very large interlinked matrix would be formed from them. That matrix actually provides a triangular locator pattern for most of the coastal points denoted and described in the Peleio section. But how they were known and written down was another conundrum to be solved.


At this point in my investigation I became very aware that when I was drawing the Peleio routes it was a complete overkill, an unnecessary expansion of data as I believe the diagrams illustrate. It was as if each Island or Cape featured had as many routes as it was possible to draw from them, drawn. Some are obviously important but many were basically inconsequential in the Medieval Period of these charts. But it was also so very evident that the routes from each island were perfect as they radiated from each like the rays of the sun without encountering a single landmass. This is not possible to achieve without recourse to a visual aid: it is impossible just to write these routes into a text by guesswork.
Thus I concluded that the “LCN” text could not have been written unless an excellent map of the whole Mediterranean Sea basin was available for the author to initiate the alignments, measure the distances involved and note the directions apropos a Windrose. There are only eight named directions used in the text, any intervening direction is noted as 1/8th, ¼, or 1/3rd of a wind of 45 degrees of the wind rose. But they are measured with an accuracy such that the author has had to resort not only to the use of those subdivisions, but when there is a very minor deviation he resorts to a new terminology, “a little more towards xxx”, i.e. 2 or 3 degrees extra deviation. I have shown in a previous text that the author measured the directions with “poco” being up to 4 degrees, but the “little more” is only 2 or 3 degrees, hence the accuracy of direction matches the undoubted accuracy of distance measure, and the usage of a subdivided protractor or similar tool to gauge the direction.
Thus I consider it to be an error by researchers to postulate that the author was using either 16, 32 or up to 132 sub-divisions of the actual wind rose. He is not; only eight are named, sometimes using both forms of the wind name, but no real subdivisions are named only a geometric format for location via the 45 degree wind division. It would have been a simple task to name just the half wind, the 1/16th division, but it does not happen. All we read are 1/8th , ¼ and 1/3rd, they being c5, c11 or 15 degrees, we do not read of a half wind, 22 1/20 in the text, only descriptions like Greco e Levante or similar.


There was in the period c400AD to c1000AD a highly accurate map of the Mediterranean Sea basin which enabled the original compiler of “LCN” to augment the “Roman” Starea and Peleio details with an enhanced Peleio section measured from that map, and it thus became text. Obviously a Peleio section was a feature of an original text as the Island section of the Itinerarium Maritimum indicates, and the 73 Peleio of the Starea section of “LCN” attest. Thus it was possible for an amplification of those routes to be included which no doubt meant the newer ports of the medieval period were now added to the map.
My hypothesis also consists of a timescale and methodology for the transfer of data and thus the twisting of the Portolan Chart as the system developed. It is an appendix to this text.
The author no doubt set out to aid mariners but because of the complete overkill of routes within the text it becomes very difficult to find the normal or major routes easily. But some use of this overkill may have arisen as follows. If a mariner departed Grapparola for San Piero on the easterly course specified (geographically correct) but the winds were unfavourable and he was blown slightly north or south he could read in the text that San Marco was Levante ¼ Greco, i.e. northerly and the Isle of Galeta was Levante 1/3 Silocco, i.e. 15 degrees south of east. But would that really happen as most sailing deviations are considerably more off course such that the “Tavola De Marteloio” would be of more use. The point I make is we can extrapolate easily to theorise unnecessarily when we should not.
Thus I return to the original point that these Peleio routes are the result of an over enthusiastic author who found he could just expand the text from Island to Island, some significant others mere rocks, but as with many medieval texts, rather overwritten, flowery.


But we must now consider the major Peleio Routes as found in the eastern Mediterranean Sea. I have included two diagrams, one being that area from the P. Vesconte 1311 chart with both LCN Peleio data and Chart measures for each route. Basically they are a compatible such that it is obvious they are one and the same. However if the same area is drawn as a geographical plot, as the second diagram illustrates, the mistake made for the Gozocreti to Acre route indicates that there is no constant variation for the errors, and thus it is highly likely the text was written as copy from the maps form. But it is also obvious that the Peleio within the Starea section have been miscopied at some time as the original Folio 53r routes from Rasausem to Gozocreti and Rasausem to Alexandria are corrupt and display the distortion caused by the wrongly copied distances south of Sardinia.


But the Starea section does not include a link to Gozomalta or from N Africa to Sardinia to produce such distortions; it can only have come from an original second section of the Peleio routes following the Starea section.



Originally there would have been a first rate LCN text and an excellent map of the Mediterranean Sea basin coastline. These were undoubtedly paired as the “Starea” section required either intimate knowledge of all coastal features or a guide map as an aid. But gradually they were separated and then the copyist scribes employed in many scriptoria who were trained to copy by rote, made a series of uncorrected errors particularly in the zone south of Sardinia and then in the area of the eastern Mediterranean Sea basin. Thus it was a challenge to produce a new map or maps that could be copied and sold as per normal. That new map was drawn from the” latest” manifestation of the LCN text containing some errors, but not all that we read today. That is obvious from the contradictory distances and directions on the extant maps/charts. These errors are thus probably of the 11th or 12th centuries after the formation of the Maritime States and the usage of the Italian language. If the original texts were Latin, they had to be translated and the distance measures converted at an early period; although that is not entirely necessary as the LCN we read today could be a later translation by a scribe who knew nothing of the actuality of the original text.
The two transcriptions of “LCN” by B R Motzo and A Debanne if studied with a map will indicate the shortcomings of “LCN” to play a major part in the accurate design and construction of later maps. The Medieval cartographers or draughtsmen would have had little real knowledge of such a vast expanse of the Mediterranean Sea basin and took on trust the “LCN” text available. That it is different to the one we read today has been discussed already but even the contemporary text by Grazia Pauli was shown to have differences and thus confirms that hypothesis. Simply put, the Peleio section of “LCN” was partly written by using a map drawn from the original and much earlier corrupt data to detail the distance/direction of further Peleio and thus that new and revised text begat the extant Portolan charts or maps. But, in using it, it distorted the actual form of the Mediterranean Sea basin to the confusion of researchers today. Thus there is no magnetic deviation; the maps are drawn as a single entity; they are generally drawn from west to east; are simply erroneous and therefore it is probably of the utmost importance now to leave the maps alone and trace the original documentation which I believe led to the “LCN”, and then we can draw a map of the Roman or Medieval Mediterranean Sea as originally observed, with natural errors included.


800/900AD; commencing in the 9th century sea trade began to expand and the Mercantile City developed. To sail the Mediterranean Sea basin maps were required and the then only source would have been those closeted in “Church” libraries and they would have been mostly copies of Roman maps and the Itineraries or periplus available. They would in all probability have been localised maps not full “world” maps. At this period the texts would still have been Latin or even Greek and the language problem started to arise with the change from Latin to Tuscan Italian (probably the first) and the problems of determining distance measures at sea. Land surveying was easy but sea travel brought its own particular problems, not only standardisation as research of metrology will show, but the physical work. However the extant Roman Itinerary Maritimum would have produced a basic list of distances from which a new list could be compiled. The necessity for good sailing data was the driving force to enable the maritime mercantile cities to expand and develop.

900/950AD; from the basic itinerary (call it LCN) maps could be constructed but they were nominally the Roman World. Thus they had to be updated along with the LCN. New data flowed into the Maritime Cities and each updated its data and jealously guarded it. But the LCN text could only be updated by actual sailing data. However as it was copied and copied, basically by scribes who were trained for that purpose, copy by rote, gradually LCN became corrupted. Not until full maps of the Mediterranean Sea basin were required by the Maritime powers would this become a major problem. At first each power drew maps to suit their domains or trade routes, not necessarily the whole Mediterranean Sea basin.

950/1000AD; the basic data in LCN gradually becomes insufficient for the Maritime powers as the ships venture further throughout the Mediterranean Sea basin. Thus LCN required updating in its Peleio section to inform the mariners of any and every possible alternative. But hardly any Peleio routes would have been accurately measured at sea for distance and direction and thus another method was required to establish the Peleio. But, it had to be found quickly and be simple to carry out. Thus the full map of the Mediterranean Sea basin is drawn from the extant data available; basic LCN data; 73 Peleio, Starea and updated knowledge for the new routes and then used to detail other routes as required.
A full map thus drawn had some accuracy no doubt gained from the Roman Itinerary measures and their maps, but of course by now the LCN data was corrupt. Thus this new map was skewed (unknown of course) perhaps from Sardinia but most certainly in the Levant and the simple solution adopted was to measure from the map new Peleio routes for distance and direction. Thus a start on the Peleio data was made such that it could be tabulated in the text from Island to Island or Cape to Cape, etc, but it was unfortunately already corrupted.

1000/1100AD; the rise of the mercantile city, sailing hither and thither and establishing new ports led to the further enlargement of the basic LCN Starea text with its small number of Peleio included to be as full a Peleio section as possible. It would be continually updated rather haphazardly at first as each maritime city would jealously guard its data until City State absorbed City State and it could be coordinated. Then the new Republics although all powerful must have allied themselves in text terms to permit a full comprehensive LCN to be written and no doubt used by all. That is of course ignoring the obtaining of data by subterfuge and then the extant LCN not being available to any other City State other than that which produced it at this point in the historical timescale. Each may have had its own errors.

1100/1200AD; the mercantile cities are at their peak, the crusades are in progress and many more maps and texts are required. Thus they are prodigiously copied and with that copying multitudinous errors are introduced yet again. Thus any charts drawn from the new texts are erroneous in many areas and this leads to the erroneous Portolan Charts that are extant today.

1200/1250AD; the LCN text approaches its final state having been fully revised with the addition of the latest towns, the Starea section adjusted accordingly and thus the final Peleio routes could be taken from a map and added to the section. But at all times the data is taken from a corrupt map and text.

1296AD; preceding copies of LCN are now gathered and rewritten, or should it be mostly just copied out with any final adjustments made to accord to the latest data. But as has been shown it is based upon corrupt data in both the Starea and Peleio sections. It is thus obvious the author, or copyist, had no knowledge of the subject and probably little interest in its accuracy as the errors attest.

1300+AD; the latest Portolan charts are drawn using a thoroughly distance accurate but direction corrupt LCN document. It has a good accuracy in the NW Mediterranean Sea area but as has been shown skews to the north east the further east the map is drawn.

1900’s+AD; researchers who have not drawn maps or can even contemplate their construction write about the Portolan Charts and determine from no evidence that there must be a magnetic affect within their draughtsmanship, not realising it is all a textual error, easily spotted if the texts are analysed and not the charts just accepted.
A rewrite of most research texts is probably now the only recourse for accuracy.

PS; I welcome any discussion via my email address and will answer all queries regardless. But please, no assertions, all points to be with proof or excellent circumstantial evidence such that the previous ideas, the “collegiate agreement” by groups, can be put to one side and a new chapter commenced which will see the old information gradually negated.

M J Ferrar January 2016.