Slide #233

TITLE: Portolano Laurenziano Gaddiano [Laurentian Sea Atlas , or, the Medicean Atlas]
DESCRIPTION: One of the most interesting and problematical maps of the Middle Ages is the Portolano Laurenziano Gaddiano of the Laurentian Library in Florence. This world map forms part of an atlas, commonly called the Medicean Atlas, consisting of eight sheets. The first, an astronomical chart, is of value mainly in the dating of the overall work. The second sheet contains the map which is the subject of this monograph. The succeeding three plates form a normal portolan chart of the Mediterranean region. The last three sheets give special charts of the Aegean, Black, Adriatic and Caspian Seas. Nothing is known of the authorship or the raison d'etre of the work and, beyond the fact that it is of Ligurian provenance, nothing is known of its origin.

The chief source of the map's interest lies in the delineation of the African coasts, a delineation that has given rise to a variety of inferences. At one extreme comes the view of Alexander von Humboldt that this work, in conjunction with others, offers clear proof of a medieval acquaintance with the southern part of Africa. "A long time before Bartholomew Dias and Vasco da Gama we find," he avers, "the triangular extremity of Africa represented in the Portolano della Medico Laurenziana". Nordenskiöld endorses this opinion when he declares that the Africa of the Medicean Atlas is "more correct than the map drawing of Africa.... on Behaim's Globe and on the map by Martellus Germanus". At the other extreme Santarem, the Portuguese savant, declares that this document and all of the same century (i. e., 14th century in his view) "constatent l'indubitable priorite de nos decouvertes et prouvent qu'avant ces decouvertes la cate occidentale de l'Afrique qui s'etend audela du dit Cap (Bojador) etait absolument inconnue aux cosmographes." Between these opposed theories come those of more recent scholars. De la Ronciere, for instance, speaks of the map with its "prescience de la forme reelle de l'Afrique avant le periple Portuguais". Wieser takes a similar view. Beazley is inclined to credit the author with knowledge of the south eastern extension of the continent, but concerning the west coast he says: "We require names of bays and headlands, we look for more definite proof of knowledge of the shoreline .... If its merits (between Sierra Leone and the end of the continent) are not accidental, they are probably due to information of the vast southward extension of Africa brought to Europe by men who had visited the Moslem settlements of the East Coasts or who had conversed with others who came from that region."

A study of the African portion reveals at least three significant facts. First, all details, physical features, pictorial embellishments and legends, stop short of the latitude of the Mos Lune. Second, there is a papable difference of brush work and of color between the areas to the north and south of this latitude. Third, the portion of the map to the south carries two versions of the shape of the continent: in addition to the obvious, colored drawing, there is an ink outline that gives an entirely different conception of Africa; one, indeed, that is more nearly in agreement with medieval map practice, as exhibited in the Sanudo and Fra Mauro planispheres (Slides #228 and #249). It is difficult, from mere inspection, to decide which rendering is prior in point of execution.

Now it is well known that the province of the normal portolan chart of the late 14th and early 15th centuries did not extend farther south than Sierra Leone (= Montes Lunae), a region about which the merchants of Genoa and other Mediterranean seaports were beginning to hear as a result of their growing trade with the Sudan. Is it possible, therefore, that the draughtsman of the Laurentian chart only drew the part of Africa north of the latitude of Sierra Leone and that the southern portion of the continent was drawn subsequently? A comparison of the European and Asiatic portions of the map with the African suggests that such an explanation is at all events feasible. In Europe the draughtsman contented himself with depicting the region Iying to the south of the latitude of central Sweden, that is the region known tolerably well to the medieval geographer. The Arctic is left utterly blank. In Asia there is a north-south break (masked, as in the case of Africa, by a colorwash) occurring approximately in the longitude of the Indus delta, that is, somewhat to the east of the Caspian Sea, which constituted the eastern frontier of the normal portolan chart. Moreover in this vicinity of the map there is unequivocal evidence of emendation: firstly there are two distinct handwritings: there is the neat, upright and almost humanistic script characteristic of the map, and a straggling cursive script confined, it should be noted, to the area east of the break. Among the decipherable words in this hand the following are significant: Mangi, Cipangala (?), Kinsai and Gugut (?). Secondly there are two versions of the south eastern extremity of the continent, one in color and the other in ink: the latter appears to have been superimposed upon the former.

Assuming then that the Laurentian "world" map was originally a portolan chart, a resume of the six regional charts that follow in the Atlas, how can we regard the twin African outlines? Clearly they would seem to be the work of an editor or editors. We can be sure that the Genoese merchants of the latter half of the 14th century were deeply interested in the question of Africa nondum cognita, if only because the Levantine land routes to the East were then occupied by the Turks. One of these outlines, therefore, may represent their solution of that age-long riddle. The other may have been drawn by a cartographer at the Portuguese court in the following century after Don Pedro the King of Portugal's eldest son had returned from his grand tour of Europe. As Galvano tells us, he came home via Italy "from whence he brought a map of the world which had all the parts of the world and earth described.... by which map Don Henry the King's third son was much helped and furthered in his discoveries." This, according to at least one writer, is believed to have been the Laurentian map. On this theory it is possible that the colored and more plausible rendering dates from the time when the true conformation of the African coast was being disclosed. In deciding about the ink outline we are helped by the world map of Albertin de Virga, dated 1415 (Slide #240). A comparison of the two maps suggests that the author (or editor) of the Laurentian map was acquainted with this work, for while the correspondence between the two is only general, it is distinctive and can hardly have been fortuitous as no other medieval maps introduce the Terrestrial Paradise and its river system under exactly the same guise.

It is largely on the ground of these and other characteristics common to the two maps that Wieser's arguments are based. These will be dealt with at some length. In the first place he argues that the accredited date of the Medicean Atlas , c. 1351, is too early. This date, he says, rests solely on the fact that the year 1351 is referred to in the explanatory rubrics of the associated Lunar Calendar. Viewed in isolation this would seem to make the date c. 1351 acceptable, but the fact that the reference to that year is in the past tense and also the fact that the de Virga Calendar starts at the year 1301 suggest that Lunar Calendars started from the first year of a century or half-century and that they do not necessarily bear a definite relation with the date of the map.

Wieser then proceeds to discuss the witness of the maps themselves. He believes that there are several points of evidence for a later appearance of the Atlas, especially its manifold and often close correspondence with the Catalan Atlas of 1375 (Slide #235). Among other common features he points out their delineation of the Caspian Sea which here and there is identical. This concordance he deems to be all the more significant because the form of the Caspian Sea on the Dulcerto chart of 1339 and on the Pizzigani chart of 1367 is not entirely closed. As the drawing of the Caspian basin in the Atlas is better documented and more accurate he concludes that the date of the Medicean Atlas cannot be placed earlier than the last quarter of the 14th century.

Now both the Dulcerto and Pizzigani portolan charts are of the normal type and with one or two exceptions (e. g. the Sudan) confine their province to regions about which they had first hand knowledge. The Caspian to most of the cartographers of this period was not a region on which first hand knowledge was generally available for evidence we have only to look at the maps of the 14th and 15th centuries and, therefore, it may easily be that Dulcerto and Pizzigani preferred not to commit themselves and to leave the eastern and northern portions unmapped. In passing it can be pointed out that as early as 1321 Marino Sanudo represents the Caspian as a closed sea, showing that, because Pizzigani left it open, there is no reason to suppose that the Laurentian map must be later than 1367.

Three further considerations lead Wieser to ascribe the map to a later period. First the verbal similarity between the text of the Lunar Calendar in the Medicean Atlas and the de Virga map. Second the peculiar shape of Africa and its likeness to de Virga's Africa. "The agreement," says Wieser, "goes so far as to make one believe that either one map has been copied from the other or that both have had a common pattern." The neatly executed and clearly defined coast of Africa represents, he believes, the original outline and the figure rendered prominent by the dark coloring a later correction effected by washing out the color already there. Third, the fact that "the world map of Albertin de Virga and the Medicean Atlas both testify to the influence of the Ptolemaic theory on the medieval conception of geography. Only a few years before the construction of de Virga's map the geographical work of the great Alexandrine had been rendered accessible to scholars of the West by the Latin translation of Jacobus Angelus finished in 1409." Because of this Wieser concludes that the Medicean Atlas cannot be dated earlier than de Virga's world map.

The first of these arguments assumes, quite arbitrarily, that the de Virga map is the parent map, but it is equally permissible to assume that the Medicean Atlas is the parent map, so that this is really no argument. The second, namely that the configuration of the southern part of Africa warrants the attribution of a later date, is seen to be poorly grounded when we dismiss, as Wieser does and no doubt rightly, the colored version as a subsequent emendation, for where does the ink outline provide such evidence? In the treatment of the African coast beyond Bojador? Assuredly not. The coast is portrayed with even comparative accuracy only as far as Cape Non and the prominence given to that headland demonstrates quite clearly that close acquaintance with the coast ended at that point. An unnamed and imaginary river having its source in septem montium Pegio et Civitas Tochorum is made to debouch into the Atlantic on the south side of the cape. In this the Medicean Atlas recalls the Pizzigani Chart of 1367 alone of all mid- and late 14th century maps, suggesting that the date is early rather than late. South of Cape Bojador to the latitude of Sierra Leone the coast is depicted as a rhythmic succession of headlands and bays - an obvious confession of ignorance. Here the twin renderings of the coastline begin. Accordingly we must look elsewhere for geographical evidence supporting a later date. It is to be found in the physiographic layout of Saharan Africa? The Mons Attalas [Atlas Mountains] have a greatly exaggerated length and reach almost to Egypt. On their Saharan front they shelter at least five lakes of considerable size, each being the headwaters of a trans-Saharan river. In the heart of the Sahara, there is a mountain complex separating the Nile from what presumably the Nile of the Negroes. The isle of Meroe is placed in close proximity to the headwaters which are supplemented by another river coming down from the northeast extremity of the Atlas range. Whether or not the Nile of the Negroes was intended, as commonly believed, to connect by some subterranean course with the Egyptian Nile cannot be derived from the map. It appears to have its source both in the Atlas Mountains near the region described Hic sunt omines Magni Xll Pedes and in the Mos Lune [The Fouta Djallon Mountains?]. Its course is annotated with a solitary inscription relating to Benicaleb. This river is entirely independent of a third Saharan river-unnamed and taking its source in the Mos Aeris (a westward extension of the Mos Lune ). It flows through a vast lake the Lake Pahlus of later maps e.g., Pizzigani, debouching into the desert near the foothills of the Atlas in Provincia Galla. Curiously enough the only inscription relating to the traffic of the region is unassociated with any of these rivers, being situated near the western extremity of the continent in Provincia Ganuya. It reads "Hic colligitur aureum".

We see from these illustrations that the world map shares the limitations and defects of all 14th century maps. So it can only be the east coast which furnishes evidence of a late date and here Wieser confesses that the southerly trend may be due to the influence of the narratives of Polo and Odoric. "It corresponds," he says, "to the old belief, if instead of giving Africa an eastern prolongation, characteristic of mediaeval geography, the east coast is made to trend south". On this view there is no necessity to postulate a 15th century origin for the Medicean Atlas for there is ample evidence of medieval acquaintance with the east coast as far as Sofala.

The last of Wieser's arguments hinges on the association of Ptolemy's Geographia in the production of the Medicean Atlas and in particular upon the parentage by the gulf which is portrayed in the ink outline. Now if this is "none other than the Sinus Hespericus by Ptolemy" we should expect to find the map bearing other indications of Ptolemaic influence, but such indications are absent.

The river systems of the Sahara conform to the views commonly expressed in medieval geographies, while those of the southern portion of the continent recall the traditional ideas held concerning the Terrestrial Paradise. As for the nomenclature of the continent it cannot possibly be construed as Ptolemaic. What then is the parentage of this gulf? Ptolemy's Sinus Hespericus is described thus: "The great gulf, which is towards the Western Ocean is inhabited by the Icthyophagi Aethiopes and from here to the unknown land are the Southernmost Ethiopians who, in the common tongue, are called Hesperii." Now neither the Laurentian nor the de Virga world maps carry any legend that at all reflects this description. In fact, in neither map is the Gulf so much as named. Furthermore the de Virga map, which Wieser considers to be the parent map, places the Kingdom of Organa on the far side of the gulf near the India of Pres. Joanes - a tacit acknowledgment that that region was not intended to represent terra incognita of Ptolemy.

It would appear, therefore, that the problems raised by Wieser's thesis are greater than those which it solves. In the writer's opinion the unknown author of the outline map and Albertin de Virga had in mind the gulf of which Western Europe was then beginning to hear - the Gulf of Gold. Prototypes of this may be seen in the Carignano, Sanudo and Vesconte maps, none of which, incidentally, can claim Ptolemaic parentage. Moreover the anonymous Spanish Franciscan author of the Conoscimiento, writing about 1350, i.e., near the generally accredited date of the Medicean Atlas, speaks of such a gulf "running inland for fifteen days' journey" and of a river "the river of Guynoa which is very wide and very long .... it is 65 days' journey in length and 40 wide".

Even if we allow this gulf not to be the Gulf of Gold and only an "example of a strangely obstinate convention" we are still under no obligation to accept Wieser's view for other writers of Antiquity besides Ptolemy speak of an Ocean Gulf , e.g., Pomponius Mela according to whose geographical conceptions the corresponding feature in the Genoese world map of 1457 was drawn (Slide #248), and Mela's work was well known throughout medieval times.

To summarize, there would appear to be good reason for retaining 1351 as the approximate date of the original drawing and for regarding the African outlines as the work of later editors: the first dating possibly from the period of the de Virga world map and the second from the period of Portuguese activity along the west African coasts. This view, it is claimed, takes cognizance both of available facts and the prevailing geographical theories. It is in harmony too with palaeographical opinion concerning the period of the script. Apart from some such explanation the Laurentian World Map /Medicean Atlas must remain an enigma, for neither the 14th nor early 15th century holds any record of achievement (in Europe) that can account for so substantial a change in geographical ideas as is exhibited in its colored rendering of the African continent.

LOCATION: Bibilioteca Medicea Laurenziana, Florence, Italy

*Bagrow, L., History of Cartography, plate XXXVI.
*Beazley, C.R., The Dawn of Modern Geography, vol. 3, p. 439.
*Kimble, G.H.T., "The Laurentian world map, with special reference to its portrayal of Africa",
Imago Mundi, vol. I, pp.29-33.


Index of Late Medieval Maps