The mathematician and geographer al-Khwarizmi entered the service of al-Mamun in the Dar al Hikma or House of Wisdom of the Great Library of Baghdad as a very young man. His most important work, titled Kitab Surat-al-Ard or the ‘Book of the appearance of the Earth', was finished in 833 AD. Al-Khwarizmi studied, revised and completed Ptolemy's work just as the latter had done with that of Marinus of Tyre. The three works basically consist of a list of coordinates of cities and other geographical features following a general introduction.

     One of the corrections which al-Khwarizmi made in Ptolemy's work is the reduction of the latitude of the Mediterranean from 62º to 52º when, in actual fact, it should be only 42º. The Arab opts for the same zero meridian as Ptolemy, that of the Canaries. The amount of inhabited land extends over 180º.

     As far as we know there is now only one existing copy of al-Khwarizmi's most important work, ‘Kitab Surat al Ard', which is kept in the Strasbourg University Library, under the code L. Arab. Cd. Spita 18. The complete title translates as; Book of the appearance of the Earth, with its cities, mountains, seas, all the islands and rivers, written by Abu Jafar Mohamed ben Musa al Kwarizmi, according to the geographical treatise written by Ptolemy the Claudian. In al-Khwarizmi's opinion, ‘the Claudian' indicated that Ptolemy was a descendent of the emperor Claudius.

    The book opens with the list of latitudes and longitudes, in order of ‘weather zones', that is to say in blocks of latitudes and, in each weather zone, by order of longitude. As Paul Gallez points out, this excellent system allows us to deduce many latitudes and longitudes where the only document in our possession is in such a bad condition as to make it practically illegible.

     The Strasbourg manuscript does not include the map of the world, which is also lacking in the Latin translation in Ms.100016 in the Biblioteca Nacional in Madrid. The list of coordinates, which in actual fact makes up the whole book, allows us to reconstruct the missing map. The task of doing precisely this has been undertaken by Hubert Daunicht (Der Osten nach der Erdkarte Al-Huwarizmis: Beiträge zur Historischen Geographie und Gerschochte Asiens, Bonn, Universität 1968).

     Daunicht reads the latitudes and longitudes on the coastal points in the manuscript, or he deduces them from the context when they are not legible. He transfers them onto graph paper and, in sections, joins the points with straight lines. In this way, he obtains an approximation of the coastline as it actually was on the missing map. He then does the same for the rivers and towns.

     The majority of the placenames used by al-Khwarizmi match those of Ptolemy, Martellus and Behaim. The general shape of the coastline is the same between Taprobane and Cattigara. The Atlantic coast of the Dragon's Tail, which does not exist in Ptolemy's map, is traced in very little detail on al-Khwarizmi's map, but is clear and precise on the Martellus map and on the later Behaim version.

     The only real difference is that, on the Ptolemy map, the South American coast to the south of Cattigara curves west to join the African coast, whereas on the other maps, it curves to the east, north-east and north to form the great Dragon's Tail peninsula, that is to say, South America. This peninsula also exists on al-Khwarizmi's map, as we can see most clearly on Daunicht's reconstruction. As Gallez points out, with the identification of rivers and mountains, we have proved that on the Martellus map this peninsula is in actual fact South America, complete with both the Pacific and Atlantic coasts. After studying the Strait of Magellan on the al-Khwarizmi map as well as that of Martellus, we can see that Tierra del Fuego was better known in Baghdad in 833 than in Florence in 1489.


     We have seen how Henricus Martellus drew a map of the world in 1489, based on Ptolemy's work and how, ‘thanks to mysterious but exact and detailed information' as Paul Gallez puts it, he added south and inland South America. We also know that in 833 al-Khwarizmi was in possession of information enabling him to revise Ptolemy's map, by adding the east and south of South America, though to a lesser degree than Martellus as regards the coastlines, and with practically no river system.

     Claudius Ptolemy was probably born in Upper Egypt around 100 AD; he spent the greater part of his life in Alexandria, where he played an important part in the world of science, as a mathematician, astronomer and geographer. He died in nearby Canopus around 178 AD. Among his greatest works, that of most interest to us is his Geographike hyphegesis or Guide to Geography. In the ancient world, geography consisted basically of measuring latitudes and longitudes and situating them on a map. It was really what we now call cartography, though only on a world scale.

     Ptolemy's mapmaking work was intended to be a criticism, correction and updating of Marinus of Tyre's earlier work. His Geographike hyphegesis consists of eight books. The first gives his sources and a general introduction to producing a map of the world. The other books are lengthy tables of latitudes and longitudes for all the cities, capes and other points of importance in the world, using the Canaries' meridian, in the extreme west, as the zero of longitude. The oldest known Greek manuscript dates back to the eighth century and is kept in the Vatopedi Monastery on Mount Athos in Greece. One of the best manuscripts still in existence is the Codex Vaticanus Vindobanesis Parisini, which was probably copied in 1401.

     There has been a lot of discussion as to whether Ptolemy had actually drawn the maps corresponding to his cartographical books. The question seems to have been answered to practically everyone's satisfaction and the general thesis is that the maps that accompany certain manuscripts of his work were actually drawn by ‘Agothdaimon' (the Good Spirit) though there is still considerable difference of opinion as to whether he was a contemporary who collaborated with Ptolemy or if he in fact lived in the next century.

     On Ptolemy's planisphere, ‘India Meridionalis', i.e. South America, has the Pacific coast shown from the 17º N line of latitude (Central America) southwards. Cattigara is placed at 8º30´ south. From there the coast turns to the south, as it in fact actually does, towards the coast of Chile starting from Arica. The Indian and Pacific oceans form a single closed sea and the lands to the south form the Antarctic continent, which is only indicated by the map frame. Although Ptolemy believed that the Earth was spherical, Ptolemy draws the end of his world at 14º or 20º south, depending on the edition of the map, because he believed that the inhabited world could not extend over an area of more that 90º in latitude or 180º in longitude,

     Paul Gallez, who has followed the path of the Pacific from Sanuto's 1574 map and seen it under the name of ‘Sinus Magnus' on Ortelius's map, also dating from 1574, has no difficulty in recognising the Pacific Ocean in Ptolemy's ‘Megas Kolpos'. The Megas Kolpos has been called Sinus Magnus since Chrysollaras's translation of the Geographike hyphegesis from Greek to Latin and its dissemination by Jacob Angelo in 1440. Sinus Magnus means the same as Megas Kolpos; Great Gulf. Paul Gallez mentions the fact that in a nineteenth century Spanish dictionary we come across a freer translation which in fact offers a more precise interpretation; Sinus Magnus – Mar del Sur, or, in English, the Southern Ocean. This is as much as to say that the Sinus Magnus is the Pacific Ocean. Strangely enough, Christopher Columbus was anxiously searching for the Southern sea during his fourth and final voyage of discovery while he was sailing by the Isthmus of Panama (Nito Verdera, Cristóbal Colón originario de Ibiza y criptojudío, pp. 64-65, Ibiza 1999)

     Since the Sinus Magnus is in fact the Pacific Ocean, Ptolemy's map actually includes the whole coast of Peru and part of the coast of Chile. Both here and in al-Khwarizmi's map we find rivers and capes, in particular two especially characteristic capes lying on the Peru-Ecuador coast; the Satyrorum Promontorium or the Cape of Satyrs, which is Punta Aguja, and the Notium Promontorium or Southern Cape which is Punta Pariña, as was shown by Dick Edgar Ibarra Grasso (La Representación de América en mapas romanos de tiempos de Cristo, pp. 41-42, Buenos Aires, 1970).

     As regards the Pacific Ocean, it is useful to see Ptolemy's criticism of Marinus: ‘Marinus does not specify the distance in miles of the crossing from the Golden Chersonese (now the Malay peninsula) to Cattigara. He tells us, however, that, according to Alexandros (we will talk about him later on), from here onwards, the land faces south and if we follow the coast, we will take twenty days to reach the town of Zabai, and that, after leaving Zabai, we sail south and further to the left for several days before reaching Cattigara' (Claudius Ptolemy, Geography I 14 1).

     He says of Alexandros, an excellent historian of Greek mapmaking, that ‘nothing is known except that he is one of Marinus of Tyre's sources' (Hans von Mzik: Klaudios Ptolemaios: Einführung in die darstellende Erdkunde… Wien, Gerold 1938). This lack of information has not prevented scholars from making all kinds of claims about Alexandros; some that Marinus knew him in person, others that he was a Greek trader based in a port on the Red Sea, and yet others that he was a ocean-going captain who had written a ship's log referred to by Marinus of Tyre.

     In Paul Gallez's opinion, Ptolemy had a preconceived view of the world, based on the existence of two orthogonal oceans, and hence the impossibility of there being more than 180º in the total longitude of the inhabited world, which he confuses with the knowable world. Gallez says Ptolemy put on blinkers before writing his Geography; he rejected in advance any information running contrary to his basic premises.

     All Ptolemy's work is based on that of Marinus of Tyre, and it is clear that he has not obtained any new information. He confines himself to copying Marinus, arguing against, trying to refute or ridiculing any claims that might go to show that the inhabited world might take up more than half the sphere, or rather a quarter of the sphere, since Ptolemy keeps to 180º in longitude and 90º in latitude. Ptolemy himself admits to this rather strange way of behaviour: ‘in order to calculate the total area, we have reduced the eastern part of the longitude and the southern part of the latitude to the previously mentioned sizes, 180º and 90º respectively'.

     Ptolemy's admission, the Alexandrian geographer's voluntary error, has been successful for many centuries, for two reasons; because the cultural decadence that followed Marinus was both far-reaching and prolonged and because Marino's work was lost, whereas Ptolemy's has survived, both in the Arab world and the Byzantine Empire, reaching the West in the fifteenth century. In actual fact, as Gallez points out, in his efforts to correct Marinus's work, Ptolemy has only made things worse. However, posterity has honoured, and indeed continues to honour, Ptolemy far more than Marinus.

Nocturlabi / Nocturlabio / Nightlab (Cortesía: Fundació Jaume I, Nadal, 1991)


Little is known of Marinus. He is mentioned by Claudius Ptolemy in his Geography and al-Masudi (Baghdad ca 888 – Cairo 957 AD) in his Kitab al Tanbih wa'l Israf, the former to copy him and the latter to praise him without copying him. Most map historians place Marinus at the end of the first century or around 100 AD. Al-Masudi was known as the Arab Herodotus. He was an historian and geographer who wrote thirty volumes, of which only three still exist today. He places Marinus of Tyre in Nero's reign, that is, between 54 and 68 AD.

     Marinus's only work of which we have direct reference is Diorosis tou geographikon pinokos, to which Ptolemy dedicates fifteen chapters. There were numerous editions, and his basic theories have their roots in Eratosthenes, Hipparchus, and Posidonius in particular. The title of his work literally means ‘Corrections in the map of the world' or ‘Corrections in the map of the inhabited world', which goes to show that Marinus of Tyre wanted to improve and revise one or several works of earlier mapmakers. He appears to wish to amend Posidonius, and that he intended to do so, using Hipparchus's astronomical work and the accounts of several recent voyages.

     Eratosthenes and Posidonius considered that several inhabited worlds must exist on the Earth's spherical surface, separated by uncrossable oceans and by a torrid, uninhabitable belt. Marinus took the liberty of extending the inhabited world to 225º longitude and reached latitude 24º S, leaving no room for other inhabited worlds. In the east, his world ended in a country called Thina or ‘Land of the Chinese'. Marinus seemed to believe that the Land of the Chinese might extend another 45º to the east of the capital (supposedly in the centre of the country), which would give us an inhabited world of 270º in longitude, starting from the Canaries Meridian zero. This arrangement leaves only 90º between these islands and the east coast of China, which is about halfway between Martin Behaim's geographical calculations and those of Christopher Columbus.

     Marinus made use of the measurement of the Earth made by Posidonius, who lived from 135 to 50 BC. While Strabo, who lived between 58 BC and 24 AD, kept Eratosthenes' measurements of 252 thousand stadia for the circumference of the Earth, that is 700 stadia per degree, Marino uses Posidonius's calculations of 180 thousand stadia, with a degree of 500 stadia (Antonio Ballesteros Beretta: Génesis del descubrimiento, vol 3, Barcelona, Salvat 1947). A stadium is an old Greek measurement of length, the equivalent of 600 old Greek feet (192.27m) or 125 paces, which was the exact distance separating the columns in the great amphitheatre of Olympia. The question is as to why Marinus and Posidonius himself adopted Posidonius's measurements instead of those of Eratosthenes. Posidonius's map, which was drawn around 60 BC, was passed on to us by Dionysius Perigetes in about 125 AD. On Posidonius's map the Earth forms a single continent and there is no trace of the Dragon's Tail.

     The work of Marinus of Tyre, which Ptolemy had at his disposal, did not seem to include any actual map. There were only some general instructions on how to make a map of the world and tables of geographical coordinates. These tables consisted of the latitude and longitude of cities and other important places, just as we have seen in the works of both Ptolemy and al-Khwarizmi. Each reader can draw his own map using these tables by joining the points with straight or curved lines, as he deems best, or as his fancy takes him. Paul Gallez includes a new reconstruction of Marinus of Tyre's map on page 137 of The Dragon's Tail. The western half and the northeast quarter were the work of Ernst Honigmann in 1930. The southeast quarter was drawn by Gallez himself.

    Gallez takes for granted that Ptolemy did not reduce the longitudes between the Canaries meridian and that of Taprobane Island, which Ptolemy and almost certainly Marino located at 123º. The real distance from Ceylon to Singapore in the southernmost part of the Golden Chersonese is only 23º. There being no reason for Marinus to have exaggerated this distance, we will therefore locate Singapore at 146º on our reconstruction. From here to the edge of the map at 225º, we are left with a Pacific Ocean measuring 79º, which would more than justify the name of Megas Kolpos or Great Gulf, when we compare it with the Gangeticos Kolpos, which measures 23º, or with any other gulf in the world.

     Gallez, who discovered the South American river system on Martellus's 1489 map, draws the following conclusions after studying Marinus of Tyre's map:

  1. A mysterious sailor called Alexandros gave the earliest known information about South America; information that he probably acquired from Chinese sailors in the Far East.
  2. Alexandros almost certainly reported the correct dimensions of the Pacific, that is 160º.
  3. Marinus of Tyre was the first mapmaker to draw a part of the coast of South America.
  4. Marinus was mistaken when he reduced the width of the Pacific from the 160º in Alexandros's report to 79º of his map.
  5. Ptolemy made an even worse mistake when he reduced Marinus's 79º to a mere 8º.
  6. Ptolemy's mistaken measurements were generally accepted until the sixteenth century.


The geographical location of Cattigara has posed a problem for map historians, to the extent that there are thirteen separate hypotheses that locate it in Asia, in itself a clear proof of their weakness. Paul Gallez maintains that there have been two generally accepted errors; firstly, people did not realise that the Sinus Magnus was in actual fact the Pacific, and secondly that Ptolemy's translation was inaccurate. It so happens that, in his seventh book, Ptolemy refers to it as Cattigara Hormos Sinon – ‘the anchorage of the Chinese'.

     However, we should not be searching in China, but where Marinus of Tyre Tiro situated it, that is on the eastern side of Megas Kolpos, on the east coast of the Pacific Ocean, on the coast of South America. In the early editions of both Ptolemy and Marinus of Tyre the anchorage was given a latitude of 8º30´ N, while in their later editions the latitude was given as 8º30´ S. In some manuscripts it appears at 3º S, so that it all boils down to a confusion between an ‘N' and an ‘S', and a 3 and an 8. In Paul Gallez's opinion, the distortion grid applied to Martellus's 1489 map situates Cattigara to the south of the 40º south parallel and to the north of the Bay of Ancaud (38º S and 42º W), that is, between 40º S and 42º S. In Martellus's map, Cattigara is located opposite the mountains where the Negro and Chubut rivers rise; that is, between 38º S and 42º S. Paul Gallez comments on the curious fact that the natives of the Chilean island of Chiloé, situated at 42º S, have very Chinese looking features.

     There is, however, an area where placenames and archaeological evidence go hand in hand – and within the latitudes given by Marinus of Tyre. Between Chimbote and the mouth of the river Santa situated at 9º S, there is a place, home to an extremely ancient civilisation, called Huaca de los Chinos (Peru), which may well have been the ‘Hornos Sinon' located at the same latitude by both Marinus and Ptolemy. The fact is that there are numerous theories on the location of Cattigara, with clues and arguments in their favour which cannot be totally discarded, though none of them offers clear, indisputable proof. The riddle of Cattigara, with its plethora of clues and lack of concrete proof, is a typical example of proto-history and proto-geography.


Excellent maps of unknown source

     Paul Gallez claims that, in the course of his research, he has studied the whole history of European Renaissance mapmaking, the most interesting examples of Arabic medieval maps and Greco-Egyptian ancient history. He has found the Dragon's Tail on all kinds of maps, either wholly or partially. Without doubt, the most perfect example is that of Martellus in 1489, though there is another, which, while poorer in every way, is nevertheless outstanding for its era – that of al-Khwarizmi. Lastly, the riddle of Cattigara reveals a rather vague but nonetheless real knowledge of an anchorage used by the Chinese on the Pacific coast of South America. Paul Gallez asks himself some intriguing questions. What were Henricus Martellus's sources of information? Where did al-Khwarizmi obtain his data? Who told Marinus of Tyre how to reach Cattigara?

     We can only find documented answers to the third question and, even here, the information is extremely sketchy. In fact, it boils down to a single name, that of Alexandros, of whom we know precisely nothing, except that it was through him that Marinus of Tyre knew how to reach Cattigara. The ‘Alexandros affair' is a typically proto-historic question rather than an historical one.

     The fact is, that in all the maps that Gallez studied, it is in those of Martellus, Al-Khwarizmi and Marinus of Tyre that South America is located at the furthest eastern extremity of Asia. The problem we are faced with, therefore, is that of the trans-Pacific pre-discoveries of South America.

     Al-Khwarizmi had access to the Dar-al-Hikna library, the best in the Arab world, which was the greatest empire of its time and also the most cultured. Baghdad traded with India, China and the Mediterranean and could have inherited knowledge from any of the three. What knowledge? What traditions? asks Paul Gallez.

     Henricus Martellus worked in Florence and Rome in the fifteenth century. Florence and Venice monopolised the spice trade with Alexandria for most of the Middle Ages. The Holy See received regular reports from its Franciscan missions in China during the thirteenth and fourteenth centuries, which were sent from Rubruquis (*) to Montecorvino (**). These reports may also have contained useful information (Jean-Paul Roux: Les explorateurs au moyen-âge. Paris, Seuil 1961).

     So we can formulate three different hypotheses; Chinese, Phoenician and Egyptian.

(*) Rubruquis is the Latin and Spanish name given to Willem Van Ruysbroeck (literally William of the Reed Marsh). He was a Flemish Franciscan friar who was sent to Karakorum, the Mongol capital, by recommendation of King Louis IX of France (Saint Louis).

(**) A famous Franciscan archbishop from Montecorvino Rovella, in the Italian province of Salerno. He donned the Franciscan habit and travelled through Armenia and Persia, reaching Peking in 1294. He lost all contact with his European brothers for over ten years, but in 1305 he found the means of corresponding with Europe and his letters started. He died in 1238 at the age of 81. He had worked in Peking for 34 years (Gran Enciclopedia Rialp).