Mathematical Geography and Cartography in Islam and Their Continuation in the Occident III, Volume of Maps

By Fuat Sezgin

This article is extracted from Fuat Sezgin’s book: Jubilaumsband zum dreiBigjahrigen Bestehen des Institutes fur Geschichte der Arabisch-Islamischen Wissenschaften In europaischen Sprachen erschienene Vorworte zu Publikationen des Institutes aus den Jahren 1984 bis 2011, Institut fur Geschichte der Arabisch-Islamischen Wissenschaften an der Johann Wolfgang Goethe-Universitat, Frankfurt am Main, 2011, pp. 99-106.

The Geography of Ptolemy in a c. 1411 Latin translation by Jacobus Angelus with 27 maps by Claus Swart. Source: wikipedia.

The Geography of Ptolemy in a c. 1411 Latin translation by Jacobus Angelus with 27 maps by Claus Swart. Source: wikipedia.

The volume at hand contains most of the general and sectional maps discussed in the previous volumes I and II (English version to volumes X and XI of my Geschichte des arabischen Schrifttums) on mathematical geography and cartography in the Islamic world and their continuation in the Occident. They are treated in the historical account, because they document the author’s conception of a process of development in the mathematical recording of the old world which began with the Greeks and continued through to the 18th century.

Within this process of development, the Greeks dominated from the 4th century BC through to the 2nd century AD. Between the 2nd and 8th centuries AD, it seems, to all intents and purposes, apparent that no perceivable progress was made as far as the mathematical recording of the Earth’s surface in and around the Mediterranean was concerned. Admittedly, mathematical geography seems to have made not inconsiderable progress in India and Persia in this period, although this apparently remained without consequence for the field of cartographic depiction.

Not only but above all as a result of the translation of Ptolemy’s Geography, which was apparently without maps, and knowledge of the Marinos map of the world (1st half of the 2nd century AD) the field of mathematical geography experienced the beginning of a period of creativity in the first quarter of the 9th century AD in the Islamic world, or to be more precise in Baghdād, the centre of the Abbāsid caliphate, (more or less along the same lines as that experienced in the other sciences which had been adopted from the Greeks). This period continued to the end of the 16th century and saw mathematical geography develop into an independent discipline in the first half of the 11th century.

Since the mid-19th century, Arabic research in Europe has produced a number of writings which revealed quite well the scientific-historical significance of the achievements made in this discipline in the Arab-Islamic world. However, we miss in these studies the answer to that question which mainly interests historians of cartography, namely whether or not and to what extent the co-ordinates produced by astronomers, geographers and navigators of that cultural sphere — whatever the process and whatever the period may have been — are reflected in the cartography. My predecessors, to whom I take this opportunity to express my respect and sincere thanks, were not able to treat this question or excluded it from their studies, because they had no graduated map from the Arab-Islamic world at their disposal.

The epoch-making world map, based on the Ptolemaic Geography and the map of Marinos, on which a large group of astronomers and geographers prepared and worked together for many years on behalf of Caliph al-Maᵓmūn (dec. 833), a son of Hārūn al-Rashīd, only came to light, together with some sectional maps, around 15 years ago (v. Map Nos. la, lc-e). It survived in the author’s copy of an Arabic encyclopaedia written in 1340 as a late version of a (certainly splendid) original. The world map surprises the historian of cartography above all by its globular projection. As does, for that matter, the depiction of a circumnavigable Africa and the island form of the old world, which is enclosed by an encompassing ocean, itself bordered by an (“unnavigable”) ocean of darkness. Furthermore, the map is scaled for the purpose helping users locate specific geographical points which find themselves within the lines of meridian and latitude set at distances of 5 degrees from one another.

It is extremely significant, from the perspective of a history of geography, that in the last quarter of the 19th century (some 3000) co-ordinates were discovered which had been compiled from this world map by an Abū Jacfar al-Khwārizmī for cartographic purposes. Abū Jacfar al-Khwārizmī was probably one of the geographers and astronomers commissioned by Caliph al-Maᵓmūn. The significance of the co-ordinates can only be adequately grasped in the light of the map on which they were based. Together with the other, specifically and purposefully collected details of the book of co-ordinates, they open the way to a complete reconstruction of the original world map. Our reconstruction based on this and on the method of stereographic projection proves to be largely identical with the hand-written manuscript copy, with the exception of some deviations which may very well have arisen as a result of multiple copying in the course of the centuries since the creation of the original in the 2nd quarter of the 9th century.

Traces of this world map and its tables, which were translated into Latin as early as in the 12th century, can be detected in European maps or tables of co-ordinates all the way through to the 17th century. Apart from being copied directly, the Maᵓmūn map seems to have largely exerted its influence by means of the world map by al-Idrīsī (1154 AD, v. Map Nos. 4, 5, 6), which despite minor advances, above all in Northeast Asia, proves to be an ungraduated copy and falls well short of the original Maᵓmūn map.

It was these copies and other imitations of the Maᵓmūn map created on the same basis which led to the caesura with the so-called T-O maps of the Middle Ages in the Europe of the 13th century. The most significant document of this caesura is, as far as we are aware today, the world map (No. 55), which appears in the encyclopaedia produced by Brunetto Latini (around 1265) as somewhat of an alien entity. It has been treated several times in modern research and its appearance has also been connected to the world map by al-Idrīsī, yet it has not yet been seen as a milestone in the history of a European cartography which stood at the threshold to a major revolution.

Upon considering this revolution, attention is overly focused on the map type which appeared around the turn of the 13th to the 14th century and which initially depicted either the Mediterranean alone or together with the Black Sea in an astonishingly developed form. In view of the fact that these maps almost exclusively indicated the coastal settlements and that they were transposed with a network of lines, they are, in the more recent research of cartographic history since the 2nd half of the 19th century, described as Portolan maps and were from the outset considered to be nautical charts. This view is based on the idea that seafarers had been the creators of such maps. In the opinion of some researchers, the linear network was supposed to have served the constructive function of a gratitude, others claim that it represented compass lines. Given such a narrow discussion of these maps, there remains, admittedly, but little room for a consideration of the generally appearing extended depiction of unknown parts of the earth in the world map, which occurred almost simultaneously and in the course of the following century.

It is my opinion, which I sought to substantiate in the two introductory volumes, that various stages of development appear to have taken place on the way to an almost perfect cartographic representation of the Mediterranean. The first such stage, albeit not without contributions by older cultures that lived along the coast of the Mediterranean, such as the Egyptians and Phoenicians, is formed by the achievements of the Greeks and Romans. This stage is mainly known to us from Greek historical works and Ptolemy’s Geography, which largely consisted of a treatment of the world map produced by Marinos. Accordingly, the geographical image of the Mediterranean basin up until the 2nd century AD seems to have been developed to such an extent that most coastal locations and islands appear to have been fixed with more or less acceptable distances. It would seem that the north-south extension had been rather precisely determined on the basis of the latitudes of some major points. By contrast, knowledge of the Mediterranean’s east-west extension was much less complete and the data were undoubtedly not acquired by astronomical means. Based on figures which possibly stem from the Marinos map, Ptolemy cites the length of the great axis of the Mediterranean as 63°, a figure that is indeed 21° too high.

The cartographic depiction of the Mediterranean reached its second stage of development with the execution of the afore-mentioned commission for Caliph al-Maᵓmūn, namely of examining, reviewing and treating Ptolemy’s Geography and creating a new world map as well as a new geographical work. In the course of this first Arabic correction, the longitude between Tangiers and Rome was reduced by 2°45′, between Tangiers and Alexandria by 10°10′ and the east-west axis of the Mediterranean, now fixed at approx. 53°, was reduced (and corrected) by 10° in comparison with the Geography of Ptolemy.
On the basis of our current knowledge, I come to the assumption that a third stage of development took place in the cartographic history of the Mediterranean which lay between the origination of the Maᵓmūn geography (ca. 833) and the production of the Idrīsī Atlas (1154). It has been proven that Arab astronomers and geographers in the first half of the 11th century made a second reduction of between 8° and 9° as a further astonishing correction to the length of the Mediterranean. However, we do not presently know whether this correction subsequently soon found its way into the mapmaking. The ungraduated maps of al-Idrīsī seem indeed to have remained unaffected by this, even though an examination of his geography and his maps does leave the impression that these must have been preceded by not unsubstantial progress in connection with the coastal distances and the depiction of the islands.

The fourth development stage in the cartographic depiction of the Mediterranean seems to have taken place in the 50 to 100 years subsequent to the origination of the Idrīsī maps (1154). This assumption is founded on the fact that the second correction to the length of the Mediterranean, namely to 44° or 45° respectively, which caused a shift in the prime meridian to 28°30′ west of Toledo, is expressed in numerous surviving geographical tables for major coastal towns such as Rome, Constantinople, Alexandria and Iskenderun. As a logical consequence, the Mediterranean already appears in realistic depiction on the oldest so-called European and on the Arab Portolan maps. Since the Arabs and their military power and trading ships were already found all over the Mediterranean from the 8th to the 11th century, they must have been in a position to be able to complement the data gained by astronomical means with further topographical details and, in the case of a (ruler’s) commission, to create a comprehensive map of the Mediterranean. To confirm the assumption that a graticule with largely corrected longitudes based on the prime meridian which had been moved to 28°30′ west of Toledo was not least that underlying the depiction of the Mediterranean, I will cite three of the few surviving graduated Ottoman maps: one map of Egypt (v. Map No. 45), one of Anatolia (v. Map No. 46b) and one of the Black Sea (v. Map No. 155a). Even though these maps originate from a later date, around the 16th century, they do have a high degree of documentary value since they lead us to traces of older Arab forerunners which were based on a graticule whose prime meridian had been set, after correction of the east-west axis of the Mediterranean and of the western section of the Arab-Islamic territory between Baghdād and the Atlantic, at 28°30′ west of Toledo.

In this respect, I have, with the exception of depictions which are part of a world map, so far only come across one single Europeanised graduated map of the Mediterranean. It is a map which has survived on two wall carpets. It was drawn in 1535 by the artist Jan Cornelisz on the occasion of his participation in the Spanish expedition to Tunisia. The map with a southern orientation in the manner of the Arabs, which only covers the western basin of the Mediterranean, reveals the younger Arab correction, even though the longitudinal difference between Tangiers and Rome is given as only approx. 15°30′, i.e. reduced by approx. 2°50′ (v. Vol. II, 136; Map No. 8).

If we assume that Johannes Kepler had known a similar map then it is easier to understand what may have occasioned him to apply the first Arabic reduction of the Mediterranean longitude of 10° of which he had knowledge solely to the western basin and to leave the eastern side, as in Ptolemy, at 10° too high, which resulted in clearly distorting the overall picture (v. Vol. II, 124).

Practically perfect maps of the Mediterranean and translations of a large number of Arab geographical tables whose co-ordinates prove to be identical with the corresponding cardinal points of the great water basin on the maps had been in circulation among geographers and cartographers in Europe since the turn of the 13th to the 14th century. However, these were neither aware of the great correction achieved in the Islamic world nor did the correction find an echo in the world maps produced in Europe until the mid-17th century, although the improved co-ordinates had been adopted by numerous tables compiled in Europe.

What the celebrated Dutch cartographer Willem Jans-zoon Blaeu wrote to his German colleague Wilhelm Schickard on this point is informative and revealing: “Your remarks on the longitude between Alexandria and Rome is what I have always thought on the basis of the observations made by my compatriots at sea, namely that the whole of Europe was indeed depicted as too long” (v. Vol. II, 129). I do not currently know what Schickard had reported to Blaeu with respect to the longitude between Alexandria and Rome; I can only presume that during his work on the book by the Arab geographer Abū l-Fidāᵓ (dec. 1331) he noticed that the differences in longitude between the two above-mentioned cities differed substantially on most European world maps from the data contained in the comparative geographical tables of Abū l-Fidāᵓ. Even Schickard, that great scholar, who had taken it upon himself to record the long-known old world in a mathematically-correct manner (v. Vol. II, 83), based above all on the book of Abū l-Fidāᵓ, did not know that this Arab source, which had only become available to him after a long search, merely compiled Maᵓmūnian and post-Maᵓmūnian co-ordinates by way of comparison, co-ordinates which were long out of date in the Islamic world itself.

On the basis of my present knowledge, the reduction of the great axis of the Mediterranean to approx. 44°-45°, which had been achieved in the Islamic world before the 13th century became noticeable on some western world maps only around two decades after the above-cited letter from Blaeu to Schickard, i.e. as from around 1650 AD.

I will close my observations on the cartography of the Mediterranean with a remark on the above-mentioned networks of lines on the so-called Portolan maps. Arab sources provide us with an unmistakable reference to the fact that these linear networks called qunbas, qunabbas (compass) originally depicted a scheme of radiating lines of wind directions drawn on a separate sheet and used in navigation together with the maps for determining routes. Later the linear network was obviously drawn onto the maps themselves (v. Vol. II, 52 ff.).

Research into the history of geography has been adversely affected by the fact that, after the Brunette Latini-Map (around 1265), the oldest world maps produced in Europe by Giovannni Carignano (around 1310, v. Map No. 68), Marino Sanuto – Petrus Vesconte (around 1320, v. Map No. 56), Angelino Dulcert (1339, v. Map No. 70) as well as the Medici Atlas (1351, Map Nos. 71, 71a-f) have been viewed as Portolan maps due to the linear networks drawn over the maps. This procedure had a disastrous effect on an understanding judgement of the completely new and in some cases very precise depictions of previously unknown parts of the world in these maps. By new parts of the Earth, I mean Anatolia, the Caspian Sea, Lake Urmia, the Indian subcontinent, the Red Sea, and Africa in fairly well-developed form.

Without wishing to repeat the explanations put forward by previous research on the origination of these parts of the Earth on European maps of the 14th and 15th century, I merely refer to the basic concept which I have deduced, namely that all new cartographic elements which approximate with reality and are found in these maps must have been taken from maps created in the Arab-Islamic world. The widespread opinion that a cartographer working in Genoa, Venice or in Catalonia could have depicted these innovations in a correct cartographic manner alone on the basis of the reconnaissance of travellers or on details taken from the book of Marco Polo is essentially nothing more than a stopgap explanation, since practically nothing was known of the mathematical geography and the concomitant cartography of the Islamic world which had been developing continuously since the 9th century and which encompassed this area from the western border of the known world through to southeast Asia.

Witnesses of this mathematical geography and cartography which have become known to me have convinced me that the process of mathematically recording the Earth’s surface experienced a particular advance in the Islamic world in the 13th century. During this phase of advance, that part of the Islamic world which lies to the east of Baghdād itself experienced an upturn, when, in the 1260s, Hulagu Khan established a new observatory at Marāgha, the first capital of the western Mongolian empire. It is of great cartographic-historical significance that some of the astronomers who transferred from the old Baghdād observatory to Marāgha to establish the new astronomical-geographical centre — among them the great and renowned scholar Nāṣīr al-Dīn al-Ṭūsī (dec. 1274) – had the idea to create a new graticule by combining the co-ordinates of the western parts of the known world, which had started out from the prime meridian at 28°30′ west of Toledo since the first half of the 11th century, with the co-ordinates of the eastern parts of the known world, which started out from a prime meridian which ran through Baghdād. Also, the co-ordinates of the eastern part of the known world, which were still based on the data of the Maᵓmūn map, had long become outdated. After the great undertaking of the universal scholar al-Bīrūnī (dec. 1048) to calculate the co-ordinates of around 60 places located between Baghdād and Ghazna as precisely as possible on the basis of his own process of triangulation and application of the rules of spherical trigonometry, the eastern part also received a tight and largely correct network of co-ordinates over the course of time. The errors in the degrees of longitude obtained by al-Bîrûnî are only between 6′ and 40′. The contributions of following generations of which I am aware are of the same quality. After the integration of the co-ordinates of the places located to the west and east of Baghdād, the basis of a graticule arose in which, in comparison with Ptolemy’s Geography, the area of India was corrected by around 30° and the eastern end of the known world by about 40°. In the general and sectional maps which are based on just such a graticule the longitudes of the following towns were: Toledo 28°30′, Rome 45°25′, Constantinople 59°50′, Baghdād 80, Derbent (on the western coast of the Caspian Sea) 83° or 85° and Delhi approx. 114°. Without knowledge of these facts, the graticules of the graduated maps of Asia and the world which appeared in Europe as from 1559 onwards as well as the graticules of those surviving Arab-Islamic maps which had been drafted since the turn of the 13th to the 14th century remain incomprehensible.
With Marāgha and subsequently Tabrīz, to where the capital of the western Mongolian empire later moved, two new points of contact were created between Christian scholars or travellers and Islamic scholars, and a new route was opened via Trapezunt and Constantinople which offered itself for gaining acquaintance with the cartography of the Islamic countries and the import of their maps to Europe, besides the well-known routes via the Mediterranean and the routes of the crusaders.

In the cartographic-historical discussion on the origin of the so-called Portolan maps which appeared in Europe in around 1300, no attention has been given to the fact that the great Qubilai Khan in Peking received the gift of a globe and six astronomical instruments from the new scientific centre in Marāgha. The fact that this globe which was provided with lines of longitude and latitude and was exactly described in the annals of the Yuan-Dynasty by Sung Lien (1310-1381) (v. Vol. I, 311) could have resulted in the creation of a map in China in around 1311-1320 which survived in later versions has been more than merely presumed by researchers (v. ibid., p. 321ff.). It is astonishing that an almost perfect configuration of the Mediterranean and the triangular shape of southern Africa appears on this map. The Sinologist Walter Fuchs, who was the first to make known, publish and examine this map, came to the conviction that the Arabs, as the map proved, had at the close of the 13th century a more or less correct conception of the shape of southern Africa and that their cartographic heritage has apparently been only incompletely handed down to us (v. ibid. pp. 323-4).

At a point in time in which the cartographic depiction of the known world, of the encompassing Ocean and of the minor seas had been more or less well-developed on the basis of maps (not co-ordinate tables!) which had found their way to Europe from the Islamic world by chance and sporadically, the Latin translation of Ptolemy’s Geography was published in the year 1477. The maps which were disseminated with this and whose Ptolemaic origin is doubtful, a point which I would rather not discuss at this juncture, resulted in emulation, above all in Germany and Italy, and led to new world maps being thrown together out of Ptolemaic and non-Ptolemaic elements.

A notable stroke of fate led to abandonment of the Ptolemaic world map being triggered by one person who, as a cartographer, had long been a follower of Ptolemy. It was the Italian Giacomo Gastaldi, who with a depiction of Asia (1559-61, v. Map Nos. 113b-d) which up until that point had been completely unknown in Europe, triggered the break with Ptolemy. Experts found the new map astounding. Abraham Ortelius, one of the celebrated geographers of his day, who published the map in 1567 with some minor deviations in his own projection, considered it his duty to note in the legend that Gastaldi had followed the Oriental geographer Abū l-Fidāᵓ (v. Vol. II, 77). Without addressing the dispute which broke out on the reason for this statement by Ortelius in the second half of our century (v. ibid. p. 106), I would rather refer to my findings, namely that Gastaldi must have based his map on a map which had originated from the Islamic world. The particular, previously unnoticed geographic-historical significance of the Ortelius remark lies in the fact that he considered the draft of a map of Asia, such as that produced by Gastaldi, to only have been possible by means which had been adopted from the Islamic world.

The maps of Asia (1559-61) and the world map (around 1561, v. Map No. 96b) produced by Gastaldi not only mark the beginning of the end of some three quarters of a century of dependence on maps which bore the name of Ptolemy, but also marked the first time that maps had been published in Europe whose graticule was orientated to a prime meridian which ran 28°30′ west of Toledo and therefore reveal models from an advanced stage of development in Arab-Islamic geography. The map of Asia produced by Gerard Mercator (1606, v. Map No. 116), which was based on these maps and presumably included material from a further depiction of northern Asia originating from the Islamic world, served as a basis for numerous imitations, continuing through to the years 1654 to 1669 when the maps of Asia of Nicolas Sanson were published (v. Map Nos. 118b-d), for which he was also able to use an obviously better map which also originated from the Islamic world and whose prime meridian also ran through 28°30′ west of Toledo. Sanson did not fail to repeatedly point out that he had used Arab sources and maps as models for his work. His clearest remark in this respect is made in the legend to his map of Siberia (Description de la Tartarie, v. Map No. 118b). By the way, it should be noted that Sanson was not the first European known to us who stated that he had used maps from the Islamic world. He had at least one predecessor in this respect in the person of Adam Olearius, who as he himself stated had produced the map of Persia (v. Map No. 136a) which bore his name by piecing together regional maps of Persia and eastern Anatolia which he had transposed from the original language during his stay in Shamakhia in 1637. Here again, those model maps had a prime meridian of 28°30′ west of Toledo. It is natural that the co-ordinates of the places marked in Olearius’s map agree with the corresponding values of the Arab-Islamic tables.

Some observations on the cartographic shape of Africa and of the Indian Ocean may now follow. The mathematical geography of the Islamic world shows us Africa from the very beginning in the shape of a peninsula that can be circumnavigated in the south, albeit extending too far east by a longitude of up to 160°. The data known to us at present convey the impression that measurements of degrees of longitude and latitude on the east coast of the great peninsula must have already been underway in the first half of the 11th century. Reports in Arab and European sources allow us to conclude that sea journeys between the Indian Ocean and the Atlantic took place long before the first Portuguese expedition. In this context, it is to be highly regarded that the historian and geographer al-Ya’qūbī, who lived in the second half of the 9th century, speaks of a functioning sea trade between the town of Massa (45 km south of Agadir) and China on ships built in Ubulla (on the River Tigris, near to Basra) as a well-known fact (v. above Vol. II, 384).
The sketch of a world map produced by Nāṣīr al-Dīn al-Ṭūsī (dec. 1274) (v. Map No. 15; Vol. I, 138 f.) is available to us as a basis for the assumption that the cartographic reproduction of the shape of Africa must have already reached a quite well-developed state in the Islamic world as early as in the 13th century. This depiction of Africa gains particular significance when it is brought into connection with the triangular shape in which the great peninsula appears in the above-mentioned Chinese world map. One must bear in mind that Nāṣīr al-Dīn al-Ṭūsī was the director of the observatory at Marāgha from where the globe equipped with longitudinal and latitudinal circles had been sent to Qubilai Khan in China.

In accordance with my basic conception of the general development history of cartography I view the significant depictions of Africa known to us from European maps partly as copies and partly as imaginative imitations of Arab models from varying stages of development. In this respect, I would name, for instance, the maps of Brunetto Latini (ca. 1265, Map No. 55) and Marino Sanuto -Petrus Vesconte (ca. 1320, Map No. 56), the Medici Atlas (1351, Map Nos. 71, 71a), the maps of Albertin de Virga (ca. 1415, Map No. 57), Cristoforo Buondelmonti (after 1420, Map No. 58) and Andréa Bianco (1435, Map No. 59), the Catalan world map (ca. 1450, Map No. 60) and the maps of Andréas Walsperger (1448, Map No. 61), Giovanni Leardo (1452, Map No. 62) and Fra Maura (1459, Map No. 63). Most of these surviving maps provide us with not insubstantial indicators for answering the question as to their relationship with Arab models. An extensive examination of the known maps from this perspective and an examination of the reference points in sources of a historical nature related to this question remain to be undertaken. Some revealing and informative reports from Portuguese sources are already known to me. They document unequivocally the fact that several maps of Africa and the Indian Ocean had found their way to Portugal at the latest from the first half of the 15th century. Some maps which included a clear depiction of the Cape of Good Hope, the island of Madagascar, and the sea routes to India had been brought to Portugal shortly before the first expedition by Vasco da Gama, and it is clearly recorded that they had been received from the Arabs (“Moors”) (v. Vol. II, 380). Furthermore, there are unequivocal references by Vasco da Gama and further early Portuguese travellers to India that they had received or captured maps from Arabs or native seafarers of a world that was new to them and with which they were confronted for the very first time (v. ibid. p. 323 ff.). It must be emphasised that these seafarers had nowhere made an implausible claim of the kind that they had anything to do with the origin of the maps which are viewed as originals of those of Alberto Cantino (Map No. 191) and Nicolo di Canerio (Map No. 192) and which are said to have been produced within a few months of the return of Vasco da Gama. They also give no occasion for the assumption that they had measured distances, determined co-ordinates or done any other preliminary work which would have been required for the production of such a map and which, in any case, would have required several centuries. Their commissions were of an economic, military and political nature to be fulfilled in strange regions which were already, however, known to them by means of largely developed maps which had been made available to them. Thanks to these maps they could reach their target destinations by the shortest routes and with but little coastal contact. Not seldom did they express their admiration for the graduated maps, the nautical instruments and the innovative compass which they were to see in the region of the Indian Ocean and for ships which were larger than their own. We must thank those Portuguese for having made a great contribution to the process of reception for maps from the Arab-Islamic world. Without them, a substantial proportion of these maps would certainly have been lost. I think, in particular, of the Portuguese copies of maps produced in Javanese writing, which reveal a very high level (v. Map Nos. 198 a-z). The question as to how the data on which these exact or almost exact maps were produced came about can be explained with the help of the values recorded in the nautical books of the Islamic world which include the latitudes and lengths measured in zam (1 zam = approx. 24 km) of almost all sailing routes and coastal lines of the Indian Ocean and with the help of the methods of those nautical sailors who measured distances and determined degrees of latitude and longitude while travelling the high seas. It is astonishing that they provide us with transoceanic distances between several positions of the same latitude north and south of the Equator which generally come fairly close to reality. Further it is astonishing to see that the length at the Equator cited here, which is almost identical with the present-day value, only appears on a single map produced by the Portuguese (approx. 1520, v. Map No. 199 b), while arbitrary dimensions are given on the other maps of the world produced in Europe right up until in the 18th century.

European geographers only came to a real preoccupation with extensive differences of longitude towards the end of the 17th century, when the astronomers of the French Academy began to make use of the process of observing the moons of Jupiter with the help of a telescope. In the case of a correction or attempted correction to degrees of longitude of far-distant places, they proceeded in such a way that the data for the maps considered up-to-date were verified by means of the few co-ordinates of prominent points calculated using the new method. The representative of this correction campaign, later described as the reform of cartography, was the French court cartographer Guillaume Delisle, who worked between 1700 and 1726. The degrees of longitude of the maps used by him prove, in comparison with extant maps and geographical tables of the Arab-Islamic world to have been dependent on a prime meridian of 17°30′ west of the Canary Islands, yet in the Mediterranean region they are around between 2° and 3°30′ and in Delhi by between 4° and 4°30′ too high. Delisle seems to have shortened the degrees of longitude of his model maps either proportionally or at will, and he seems to have sweepingly adopted the degrees of longitude of a map after shortening its differences from the prime meridian by some degrees, such as the longitudes of Anatolia and Persia. The process of the revision of the maps of the old world was continued, on a sounder basis than was the case with the predecessor Delisle, by Jean Baptiste Bourguignon d’Anville (1692-1782) and James Rennell (1741-1830). They did not satisfy themselves with using the maps available to them as models, but rather drew on as many Arab-Persian-Turkish tables and other sources as they could tap.

This brief introduction was written without mention of the sources and without stating the reasons for my own views given for the purpose of providing the user of the present volume of maps who does not have at hand the introductory volumes I and II (volumes X and XI to my Geschichte des arabischen Schrifttums) on the mathematical geography and cartography in Islam and its continuation in the Occident a certain impression of their content and of my fundamental conception which I seek to underpin in these volumes.

It is my pleasant duty to thank, in particular, my colleague Eckhard Neubauer for preparing and shaping this volume of maps. Thanks to his conscientiousness alone and his exemplary commitment, this volume has taken its present form.

My thanks for the provision of film material for the reproduction of the maps above all goes to the following libraries and institutions: London: The British Library. Paris: Bibliothèque Nationale de France. Oxford: Bodleian Library, University of Oxford. Göttingen: Niedersachsische Staats- und Universitätsbibliothek. Berlin: Staatsbibliothek zu Berlin, Preu-Bischer Kulturbesitz. Wolfenbüttel: Herzog August Bibliothek. Munich: Bayerische Staatsbibliothek. Florence: Biblioteca Nazionale Centrale and Biblioteca Ambrosiana. Venice: Museo Civico Correr. Milan: Biblioteca Ambrosiana. Vatican: Biblioteca Apostolica Vaticana. Madrid: Palacio Real, Patrimonio Nacional. Seville: Alcazar de Sevilla, Patrimonio Nacional. Washington: Freer Gallery of Art, Smithsonian Institution. Kyoto: Ryukoku University Library. Istanbul: Süleymaniye Kütüphanesi, Universite Kütüphanesi and Topkapı Sarayı Müzesi. In particular, I would like to name Catherine Hofmann (Paris), Nevzat Kaya (Istanbul) and lonathan Potter (London) in person for their assistance.
Manfred Strauss, our reliable printer of many years’ standing, deserves my special thanks for the great personal commitment he showed in accompanying the printing process through all its stages.

Finally, I would like to remark the following:
The volume of maps appears here in an English edition, in accordance with the planned and already commenced English translation of the previous two volumes. The map section with its bilingual captions applies to both editions.

All maps, even when they originally had a southern, eastern or western orientation, have been given a northern orientation here, with but one exception (No. 80).

For details of the origin of the maps and of studies on them, may the user of the volume of maps resort to the numerical map index and the table of contents.

Frankfurt, September 1999