Taylor & Francis, Ltd. Imago Mundi, Ltd. A Forgotten Ptolemy: Harley Codex 3686 in the British Library Author(s): Marica Milanesi Source: Imago Mundi, Vol. 48 (1996), pp. 43-64 Published by: Imago Mundi, Ltd.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact
[email protected].
Taylor & Francis, Ltd. and Imago Mundi, Ltd. are collaborating with JSTOR to digitize, preserve and extend access to Imago Mundi.
http://www.jstor.org
A Forgotten Ptolemy: Harley Codex 3686 in the British Library MARICA MILANESI ABSTRACT:The text of the British Library's manuscript Harley 3686 is an undated and anonymous Latin version of Ptolemy's Geographiawith an innovative set of eighteen non-Ptolemaic maps of Europe, Asia and Africa. Links with Andrea Bianco's nautical atlas (1436) suggest a Venetian provenance for the manuscript and a date of between 1436 and 1450. Map outlines were derived from portolan charts but inland topographical detail and toponymy appear to have come from the Ptolemaic text. The codex constitutes one of the earliest examples of the synthesis of portolan chart, Ptolemaic map and medieval mappamundi that characterised fifteenth-century cartography and the only example of such a synthesis on a regional scale. Harley 3686 reveals some of the technical and methodological problems that the Geographiamust have presented to the period's cartographers. KEYWORDS:British Library Harley Codex 3686; Ptolemy's Geographia;portolan charts; nautical-regional maps; synthetic mappaemundi; Venetian cartography, fifteenth century.
The great work of classification and description performed by Joseph Fischer, S.J., has meant that fifteenth-century Latin codices of the Geographiaof Claudius Ptolemy have come to be identified almost exclusively with those produced in the scribal workshops at Florence, especially after 1450, and with the printed editions of 1475 onwards.' However, not all the Ptolemaic manuscripts with maps were classified by Fischer. For example, the anonymous sixteenth-century codex of 27 plates without text (now Biblioteca Vallicelliana, Rome, MS R. 1 16) escaped his attention. In addition, although historians of cartography recognise that the translation of the Ptolemaic text into Latin changed the direction of cartography, they have as a rule ignored codices lacking maps. Comprehensive surveys like Fischer's facilitate the studies of those who come after but they also seriously influence them. Fischer's (few) mistakes have become dogma. Above all, whatever he neglected has passed into oblivion. Thus, mapless Ptolemaic codices have been abandoned to historians of mathematics and astronomy, for whom the
Geographiais a minor work in comparison with the Almagest and the Analemmata. Moreover, by considering the Ptolemaic manuscripts only as collections of geographical maps, we have limited our studies almost entirely to codices endowed with a luxurious cartographic apparatus, those originally of princely or wealthy ownership. More humble codices, with or without maps, belonging to students of geography and even perhaps to cartographers, which bear their readers' annotationssigns of their attention and traces of their interests-have been overlooked. Yet in late medieval and early Renaissance Europe only the few could have afforded codices with maps. Paolo dal Pozzo Toscanelli, for instance, never possessed a mappamundi: he borrowed one from a friend in 1459 and kept it until his death in 1484.2 From the great codices studied by Fischer we can derive above all an aesthetic pleasure similar to that experienced by their first purchasers. Moreover, we can learn something about the tastes of the mighty and the criteria they used in choosing vehicles of ostentation and symbols of power.3 Such magnifi
44
cent codices were not for study: rarely does a correction spoil the ordered lines and nobody has dared to put maniculae or references in the margin or to add glosses, write forewords or make indexes. Exceptions to this rule-for example, British Library MS Harley 7182-are few. Nor could any reader have produced a Ptolemaic projection from their geometrical drawings, for in the majority of codices these had not been constructed according to the instructions but were copied and modified until they only appeared to correspond to the text, itself confused and badly translated into Latin by a humanist who knew a little Greek but no mathematics.4 The luxurious and costly Florentine codices were largely 'mass produced' and, apart from prototypes, are repetitive and full of uncorrected errors in both text and diagrams. Their maps were copied from other maps of old design, occasionally with additions. Most of the Florentine illuminated codices thus contain little of interest for historians of cartography and geography. It is no accident that these codices have been studied most thoroughly by palaeographers, humanistic philologists and art historians.5 If we are looking for innovation, we should study codices that belonged to scholars or cartographers. From a study of such working codices, the historian of science Dana B. Durand made an important contribution to the history of geography and fifteenth-century cartography. He showed that the prototypes of the Florentine codices down to Martello in the early 1490s (apart from Lapaccini and Buoninsegni) incorporate innovations originating in Germanic lands.6 Some of the small mapless codices, too, may contain hitherto unknown attempts at cartographic innovation. No less interesting is the variety of provenance of these codices: some were copied not at Florence but at other Italian cultural centres like Ferrara, Bologna and Venice; others are from north of the Alps. I have seen only a modest sample of these working codices, and since the entry 'Ptolemaeus' in the Catalogus translationum has not yet been published, we do not even know their total number. However, Douglas W. Marshall's incomplete list is enough to indicate that the number is very high.7 The few studies undertaken so far suggest that both the mapless and the nonFlorentine codices are a potential mine of new information on the history of fifteenth-century geography and cartography. One of these humbler
codices, a 'forgotten Ptolemy', illustrated with maps which do not correspond to those in any other known Geographia,is the subject of this paper.8 British Library Harley MS 3686 The manuscript in question, now in the Department of Manuscripts of the British Library (London) at the press mark Harley 3686, is listed in the CatalogusManuscriptorumHarleianorum, volume 3, as 'Claud[ii] Ptolomaei Geographia, 8 libris. A Jacopo Anglo traducta, et Alex[andr]o V Pon[tifici] Max[imo] dedicata: cum tabulis geographicis. Codex chartaceus. XV [cent.]'.9 At folio 2 of the Harleian codex, in a sixteenth-century hand, there is an illegible date; at folio 11, in the same hand, a note refers to the volume as belonging to one Bagleius. This could be William Bagley, a Protestant cleric who had lived in Essex and at Cambridge and died in 1540, thus implying that the codex reached England in the first half of the sixteenth century.'0 From 1715 it was in the collection of Robert Harley, first Earl of Oxford (died 1724). In 1753 it was passed on to the British Museum (now British Library). The codex contains the text of the Geographiaof Claudius Ptolemy in Latin, with six explanatory diagrams and 18 regional maps. Blank spaces have been left on some of the pages containing the description of the Greek peninsula and were presumably meant for other maps. The text follows the translation made by Jacopo d'Angelo da Scarperia (Cosmographia, c.1409-1410), and includes the dedicatory letter to Pope Alexander V. Only the list of 'Provinciae seu satrapiae notae' at the end of Book VII is missing. Comparison with fifteen other fifteenth-century Greek and Latin codices, which include the oldest extant dated Latin codices, suggests that the text closest to Harley MS 3686 is a codex in the Vatican (Vat. Lat. 2974) that was made in Florence in 1409. In the Harleian codex, only the geographical order of the provinces in Book m, with the description of Greece following that of Italy instead of Sarmatia, corresponds neither to the original text of the Geographianor to any other codex known to me (nautical atlases included). The order of the other books is unchanged. Many place names have been Italianised and modernised, and some modern names have been added. The six explanatory geometrical diagrams correspond to those in the oldest dated codices that have come down to us." The first projection is drawn on
Table 1: The Maps in Harley MS 3686 Map title
Folio
Ireland Tille, Scotland and England Iberian peninsula France Germany (full page) Italian peninsula, with islands and part of Balkans Corsica Sardinia Sicily Greece Eubea Crete North and west coasts of the Black Sea Region east of the Caspian Seaa Central and east Asia between Sogdiimontesand terraincognita (two full pages)a Strait of Gibraltarand north-west Africab Baltic and Scandinavia, with part of Germany From Poland to Volga River
f.L12 f.L13 fl5 f.20 f.23v ff.28v-29 f.31v f.32v f33v f34v
f.36 f.36v f.41v f.98 ff.98Y-99 f99V
f.100 ff.i00Y-i0i
Size (mm) 154 220 200 205 282 282 45 70 122 245 67 65 200 approx. 282 approx. 282
X 125 X 190 X 210 X 250 X 212 X 242 x 45 X 62 X 80 X 212 X 72 X 75 X 212 X 200 X 415
approx. 282 X 210 282 X 212 282 X 424
aPage reduced by restoration. bWhole page but cut on the left.
an erroneous rectangle (ab < 2ac), the second is correct (ab = 2ac). In the Harleian codex the diagram relating to the 'Circularis spherae cum habitabili terrae descripcio' (Book VII) is missing, which is considered normal in fifteenth-century codices and in their Greek prototypes. However, the British Library's MS Burney 111, a Greek codex with 66 maps of the B recension, contains a reasonable representation of this third Ptolemaic projection (f. 104v); the Bologna printed edition (1477) also attempted to reproduce it. The Harleian codex often contains marginal notations indicating the appropriate clima for localities listed in the text (for example, in folio 62, '14. clima' is added near 'Cimmerium promontorium'). Other annotations and references in the margin of the text, in the copyist's hand, are numerous, but only a few are comprehensible. Evidently the copyist compared his original source with another Ptolemaic manuscript, probably also in Latin since no Greek references occur in the text, and has frequently corrected the coordinates. He has also compared the work with Pliny, whose NaturalisHistoriahe has quoted on folios 29v, 35v and 47.12 On the map of Germany (f.23v), we find the bisurge river, called Ausurgisby Ptolemy, which probably corresponds to Visurgimontesin Pliny (Nat. Hist. IV 100). The Maps in Harley MS 3686 The British Library codex contains 18 nauticalstyle maps in various formats with modern and Ptolemaic chorographic elements. Thirteen maps
are inserted in the text, according to the model of the Greek B recension of Ptolemy's Geographia (Table 1). The map of the Black Sea immediately precedes the description in the text of European Sarmatia. The five maps at the end of the volume follow no recognisable order, probably because the modern binder did not know how they should be collated. Eastern Europe from the Baltic to the Black Sea is the most represented area. The central and eastern regions of Africa are missing, as are those of southern and south-east Asia. We do not know whether these maps were lost or never drawn. Overall, a recognisable picture of the fifteenthcentury inhabitable world is presented, but the maps do not join up to form a single, if incomplete, world map. However, the maps of the British Isles, France, Spain, the Mediterranean, the Black Sea and Africa can be combined, even though-as is usual in portolan charts-the Black Sea map is on a different scale (Fig. 1). The maps belong to the socalled 'transitional' type, in which Ptolemaic toponymy is mixed with modern names, and the western parts of the ecumene are represented according to the current nautical cartography, while the northern and eastern parts have Ptolemaic, schematic or even imaginary shapes.'3 The world picture in the Harleian codex is delimited in the northern and eastern regions by a wavy line, a hypothetical boundary which gives a circular or oval form to the known lands-and to those unknown but mentioned by Ptolemy. An
45
,1~~~~~~~~~~~~~~~~~*
1'
he
46
.
~~~~~~~4
Fig. 1. Area covered by the regional maps in Ptolemy's Geographia(British Library, Harley MS 3686). (Top) The maps of the British Isles, France, Spain, the Mediterranean, the Black Sea and Africa combined. (Centre) The maps of the Baltic, Poland, Russia, the Danube, the Black Sea, and western Turkey combined. (Bottom) Asia east of the Caspian Sea and central and east Asia between the Sogdui montesand terra incognita.The scale bars on the maps of Greece and the Black Sea have different values.
Al
~
~
~ ~
~ ~ ~ ~ ~ ~ ~ ~ ~ ~~zi
i
Z
Fig. 2.
*I Jr~~~4Waire.
~
The map of the Iberian peninsula, folio 15 from Harley MS 3686 (Ptolemy's Geographia),has traced coasts and a freehand interior. (Courtesy of the British Library.)
ocean beyond the confines of the known world is indicated north of 'terra incognita' in Scandinavia (f.100) and north of 'terra incognita de sina intra imaum montem' in Asia (ff.98v-99). Elsewhere, margins are torn and maps may be missing. Thus, we do not know if the Indian Ocean was intended to be open and the whole inhabited world insular, as in most fourteenth- and fifteenth-century world maps, or if the unknown cartographer had followed Ptolemy's outline for the southern parts of the world, as on the Fillastre-Pyrrus de Noha map of 1414.14
All the maps are executed by the same handthat which copied the written text-and are in a consistent style. Their outlines were partly drawn free-hand and partly traced, as is clear from the
verso. On the maps which show Ireland (f. 12), the Iberian peninsula (f.15) (Fig. 2), France (f.20), Italy, the Balkans and Greece (ff.28v-29,
34V), the north
and west coast of the Black Sea (f.41v), the eastern regions of the Caspian (f.98), and Africa (f.99V), the indentation of the stylus used in tracing the drawing on to the page is clearly visible.'5 In contrast, the maps of 'Tille' (Thule), Scotland and England (f.13), Germany (f.23v), Corsica (f.31v), Sardinia (f.32v), Sicily (f.33V), Eubea (f.36), Crete
(f.36v), together with those showing central and east Asia (ff.98v-99), the Baltic and Scandinavia (f.100), Poland and Russia (ff.100v-101), were all drawn free-hand. The geographical features in both series of maps and the diagrams were drawn freehand.
47
The scale of all the tracedmaps is approximately the same, except for the northand west coastof the Black Sea (f.41v) which is on a slightly smaller scale. The general scale is seen on the map of Greece (f.34V) where there is a scale bar, 120 mm
long, dividedin 10 equal units. For the Black Sea the 120 mm scale bar has 12 partitions.'6No other scales are indicated. On the free-hand maps the scales differ from place to place. The same territoriesare represented on different maps in differentproportionsand, occasionally,with different outlines. All maps lack a grid, wind roses, names of coastallocalities,scale (except as above), and decorativeelements. In other respects,all the graphicalconventionsof nauticalcartographywere observed.'7
Fifteenth-centuryItaliannauticalchartsseem to have provided the main source for the maps of Ireland(f.12), the Iberianpeninsula (f.15), France (f.20), and the Balkan peninsula with Greece (f.34V). The inland detail on these maps is quite
different from that of the 'tabulae novae' of the FlorentinePtolemaiccodices.It looks more like that of Genoese portolan charts or of the maps in Andrea Bianco's atlas (1436).18 However, the
Harleianmaps contain considerablygreater detail than the Genoeseor the Biancocharts,especiallyin the case of the maps of the Iberianpeninsula(f.15), where many place-names have Portuguese spelling, and of France(f.20). Anotherunusual feature of the Harleianmapsis the amount of detailfor the Balkan peninsula (f.34V), where Bojana is both
shown and named (bojaneinstead of the usual Lodrin), as is the Vardar River (veria).'9 The red
crossof SaintGeorgemarkedon the islandof Chios also associatesHarleyMS 3686 with some of the Italian-stylenauticalatlasesof the beginningof the fifteenthcentury.20 The Harleian map of Italy (ff.28v~-29) also has a
48
distinctlynauticalaspect. Although the geography is poor, the map contains some unexpected items. The coursesof the Arno and Tiberrivers,which rise in the casentinomo(ns),echo faintlyfeatureson the map of Italy added to CristoforoBuondelmonti's LiberInsularum(NationalLibraryin Berlin, Hamilton codex 108, first half of the fifteenth century) and those shown on the Vatican's anonymous BorgianaV world map (attributedto Fra Mauro, c.1450).2' The Casentinomountains do not appear on any other map of this periodknown to me. The hydrography of northern Italy recalls that of Ptolemy,but the majorlacuson the Harleianmap
Fig. 3. Harley MS 3686: detail from folio 100' showing,
mistakenly,the insulastamariain the lowerDanubeas the easternmost of the three islands. (Courtesy of the British Library.)
can be attributed neither to a nautical nor to a Ptolemaic source. It could be interpreted as misreading of one of the fifteenth-century modern maps of Italy, on which the Navighi (canals) of Milan together with the rivers Adda, Po and Ticino form an 'island' with Milan on its northern shore. A late example of this configuration can be seen on the modem map of Italy in the Geographiaexecuted in Florence in 1469 (Biblioteca Apostolica Vaticana, MS Vat. Lat. 5699).
The free-hand map of the Baltic and Scandinavia (f.100) reveals only indirect links with the prototype established in 1325-1330 by Angelino Dalorto. The Harleian map shows some similarities with the prdesentation on contemporary nautical maps such as Bartolomeo Pareto's (1455), and with the Catalan mappamundi (c.1450) now in the Este collection of Modena. No connection with Claudius Clausson Swart's (Claudius Clavus) map of Scandinavia (1427) or any of its derivatives is apparent. The copyist has made some corrections to place names and has altered the locations of the lakes by reference to Bianco's atlas. The erroneous positioning of the name pumerensisin Scandinavia (f.100 and ff.100v)-eal) may perhaps reflect the fact that Erik of Pomerania was sovereign of Sweden, Norway and Denmark from 1412 to 1439. The coastal delineation of Germany and the relaletion ons of the other regions of eastern central Europe, together with the course of the Danube (ff.23v, 100, 100v-101 ), have an outline similar to, but not identical with, that of the Bianco map, on which boundaries are not marked. The insula sta maria formed by the Danube in the lowest of its course part (f.100vn-dy) might be a copyist's error since, according to early charts, Santa Maria
Fig. 4. The map of Tille [Thulel, Scotland and England reflects the combination of sources of the maps in Harley MS 3686. Coastlines derive from portolan charts but the north-eastward deviation of Scotland points to a Ptolemaic influence. Part of folio 13. (Courtesy of the British Library.)
was the westernmost, not the easternmost, of the three 'islands' of the Danube (Santa Maria, Buda, Sirmio) (Fig. 3). On various nautical maps, how-
ever, including folio 9 of Bianco's atlas, the name of the island Sirmio is abbreviated and can be read hastily as 'st.m.ia' instead of as 'sr.mia'. The copyist responsible for the Harleian MS 3686 tried to identify the ancient names with the modern in Mediterranean and western Europe. He sometimes gave Ptolemaic names to places known to him but not to Ptolemy, for instance, terra incognita de sarmatia asiatica north of the Baltic (f. 100). Elsewhere he introduced the modern equivalent instead of the Ptolemaic name, which is thus reduced to a historical element (for example, on 15, shown in Figure 2, Lusitania nunc Portugallu f.L and on f23V, Germanianunc Alamagna dicitur). The maps also reveal an attempt to reconcile Ptolemaic hydrography and orography with that of contemporary regional cartography. Gaps in geographical knowledge were filled by guesswork; thus almost all seas, rivers, mountains and countries were given a modern name and some sort of outline-in most cases hypothetical. An exception is the Danubian region between Germany and Italy (ff.28v-29), which is Ptolemaic in both names and outline, although neither the river's course nor its source (lacus danubii) is Ptolemaic. The easternmost Danu-
Fig. 5. Harley MS 3686, folios 100v and 101, showing the area from Poland to the Volga river, with European Sarmatia and a small tract of Asiatic Sarmatia enclosed within the western loop of the rahy (Volga). (Courtesy of the British Library.)
49
(Thule), located near the Scottish coast in accordance with the usage introduced by DalortoDulcert cartography, thus became the largest of the 30 Ptolemaic Orkney islands. The island is furnished with a text which adds another Ptolemaic element to the map (the number of the Orkney Islands) and gives modem information on the local climate. To the best of my knowledge, such a text is found elsewhere only in the Bianco atlas (1436).22 In the Harleian codex only European Sarmatia, together with a small tract of Asiatic Sarmatia enclosed within the western loop of the rahy
bian regions, too, like the Balkans and the Black Sea, are substantially modern. In areas described by Ptolemy but unfamiliar to the makers of the nautical charts-in Europe, regions from Scotland to Sarmatia (Russia) and, above all, in the interior of Asia and of AfricaPtolemy's geography prevails. And it is in these regions that the Harleian cartographer's original contribution is most obvious. The coastline of the British Isles (f.13) is nautical, but Scotland shows a perceptible north-eastward deviation of undoubtedly Ptolemaic origin (Fig. 4). The island of tille
,rs" #~~'~~'~~
Flit
~hY8~(~U fett
Aqt4
164v.a
8 It
Un
4e4Zcrnsd,
IerAL~E~ r~
iatu
SItA
'5'?tt~ri1~7is~imsaftI
~
~5j/i )~'El4AfO.
Yrz13RIOf7L7i37TN 50
Fi.6.
arleyMS 366, foio 4v
'ho ingtenotln
wes
irr. Ocasnts
fth lac
e.(oreyo
h
rts
(Volga) (ff.100v-101), is shown; the western Caspian regions, the Caucasus and the course of the Volga itself are not represented (Fig. 5). However, the presence of the loop of the Volga at the same latitude as the Don (both on f. 101) indicates that the map of Asiatic Sarmatia was based on Ptolemy. Folio 101, however, contains the non-Ptolemaic names citrican in sarmatia asiatica (Astrakhan), saray fi. (Saray river) and saray zucco (Saraicik). The well-known Tartar cities of the lower Volga are-most unusually-located on the course of the Sal' and of the western Manyc, northwest of the Caucasus, nearer to the Don than to the Volga. In contrast with this mix of Ptolemaic and modern features, the representation of the territories between the Baltic and the Black seas in the Harleian codex (ff.41v, 100, 100v-101) is nearly all modern (Figs. 5 and 6). The Meotid Marsh has been sufficiently reduced to leave space for the territories east and north of Germany. These lands were poorly known both to Ptolemy and to fifteenthcentury Italians. The Tanay (Don) is labelled as terminusasie de europein both its Ptolemaic and its modern form (ff.100v-101). The region west of the river bears the Ptolemaic name of sarmathia europe but is subdivided into the modern prussia, litogna (Lithuania) and russia. Opposite each other, lanbrechcivitaspolana and guistula civitasrossia indicate the modern political boundary (not marked) between the Grand Principate of Muscovy and the Kingdom of Poland. In accordance with traditional nautical toponymy, sarmathia europe is enclosed to the north and the east by the great arc formed by the rivers tanay (Don), which debouches into the Black Sea, and nu (Neva), which flows into the Baltic. Neither river, however, has its source in a lake but emanates from a group of mountains which at the western end carries the modern name of brancie while the eastern end retains the Ptolemaic/classical riphey (Riphey in the text).23 Evidently the Harleian cartographer's river nu was intended to coincide also with Ptolemy's Chersinus, a river described as flowing into the Sarmatic Ocean at 60?N. North of these mountains and rivers is only 'terra incognita', the words located where the nautical maps usually placed the cities of Russia and the Upper Volga. To understand the Harleian cartographer's representation of the territories north of the Black Sea (ff.41v, 100v-101) requires accepting a complex and sometimes contradictory balance between the
Ptolemaic maps on the one hand and the charts on the other. Ptolemy's Poritus and Licus (Donetz) rivers are given the modern names of alano and cumano,and his Boristhenes (Dnepr), usually called Lussonon the charts and lozo on Bianco's map, here bears the name eltyc. Ptolemaic Tyras (Dnestr), on the charts called Turlo (tarlo on Bianco's map), becomes lussonfluvius nunc turla, thereby acquiring the name that the charts attribute to the Dnepr. On f.41v, the Dnepr originates in mountains bearing the name carmatus (Ptolemy's Carpatus), and the Dnestr emanates from the westernmost sarmatici montes, as in Ptolemy. On ff.100v-101, however, both the Dnepr and the Dnestr are shown flowing from a single group of mountains called carmatus whereas the sarmaticimontesgive rise to the Ragnes (the lerassus of Ptolemy, the Rixa of Bianco, and today the Arges), an affluent of the Danube on the boundary between the vilachia and the territory of the septemcastra nunc veginel (Siebenburgen, Transylvania). The Harleian cartographer thus was identifying the sarmatici montes with the Carpathians.24 Further west, the other great left-bank tributary of the Danube, the Tibisco, bears the modern name of racus and is shown flowing from mountains with the modern name of libulo instead of from the Ptolemaic Carpatus. Thus Ragnes and racus mark, correctly in Ptolemaic terms, the eastern and western boundaries of Dacia. At the same time they identify a number of modern regions. The Harleian cartographer, accustomed to drawing mountains as zones rather than ribbons of upland, has broken the long Ptolemaic chains into a series of mountainous massifs and has moved them eastward in order to make them coincide with his reshaped territories north and west of the Black Sea, whose outlines are based on portolan charts. For the territories east of the Caspian Sea (ff.98, 98v-99),
the
cartographer
made
no attempt
to
follow modern cartography and adhered mainly to Ptolemy; the toponymy is derived directly from his manuscript text and the spelling is the same as in the manuscript, albeit with some Italianisation: genitives, for instance, are not usually recognised as such, and the names transferred to the map are incorrectly left in the genitive case.25 The Syr Dana has both its modern and ancient names: hostia norganzer aut laxartes (f.98) and laxarti (ff.98v-99) (Figs. 7 and 8). The spelling norganzer is in the Venetian style.26 All the regions south of the Caspian Sea are Ptolemaic. On the other hand, the rivers, the delta of the norganzeraut laxartesand
51
Fig. 7.
52
Harley MS 3686, folio 98, with the map of the region east of the Caspian Sea. (Courtesy of the British Library.)
the islands in the Caspian have no equivalent in any known ancient or modern map. For the map on folio 98 of the Harleian codex, the cartographer has used the non-Ptolemaic term caspey in the locale of the Ptolemaic monti Aspisii
(Fig. 7). These mountains are continued across the north-eastern extremity of Asia onto the map on the next page (ff.98v-99) (Fig. 8). Ptolemaic nomenclature reappears only south and east of the gap in the range guarded by the turris lapidea.
Fig. 8.
Harley MS 3686, folios 98' and 99, with the map of central and east Asia between the Sogdii montes and terra incognita.(Courtesy of the British Library.)
The general delineation of rivers and mountains on the second map differs in part from Ptolemy who has one, not three, Sogdian marshes (sogdianorum palus) and no river emodis (in Ptolemy, as in all the Greek and Roman tradition, the Emodi are mountains) or casios (monti Casii in Ptolemy) and, above all, no caspeymountains. The last must have been taken from a modern source although I know of no any other map of classical or nautical origin on which the caspeymountains are so extensive.27 The Harleian codex has only one map of Africa (f.99V) (Fig. 9). Here, even more than in Europe,
ancient and modern cartography was painstakingly interlocked into a new, conjectural representation tingitana28 and of the territory. Between Mauritania Mauritania Cesarensis, Getulia and inner Libya, which are unnamed, the borders of the region are those of Ptolemy. The mountains, too, have Ptolemaic names but their positions have been modified to accommodate the new coastline and the consolidated outline of the Atlas chain, which has three Ptolemaic names (sagapola, athalans, temandrus). Among the inland waters, the river maluo and the nigricespalus have Ptolemaic names and locations. Some unnamed rivers can be identified as Ptolemy's Nigir, Salatus and Thulais;
others are Ptolemaic in name (ballonius,masotiliand stachir) but not in location. Yet others, nameless, correspond to nothing in Ptolemy or on portolan charts. The Nigir and the nigrices palus might be associated with the western Nile system of medieval cartography. The west coast of Africa is represented as far as Cape Bojador. The Canaries Islands have been sketched on the outer margin of the map, and Lanzarote has been given its usual Genoese cross, but the group has no name.29 The stippling on the lower left side of the map indicates, I think, not the Sahara desert but unknown territory where geographical outlines are hypothetical-such as the rivers which rise in the Ptolemaic mountains and thence flow westward or eastward. The torn edge of the page means we cannot tell whether the cartographer was aware of any Atlantic islands besides the Canaries or whether he had any particular ideas about the south-west coast of Africa.
The Originsof HarleyMS 3686 The family likeness of both the text and the geometrical diagrams to the early Florentine editions of the Geographiasuggest that an Italian codex
53
Fig. 9. Harley MS 3686, folio 99", with the Strait of Gibraltarand north-western Africa. (Courtesy of the BritishLibrary.) from the first half of the fifteenth century was used in the preparation of Harley MS 3686. Such a hypothesis does not run counter to the evidence of the watermark (documented at Florence 14241426, Palermo 1425, Innsbruck 1425, Pisa 14271429) which belongs to a type found between c.1370 and c.1490 throughout Europe and in Italy, especially northern Italy. The writer's consistent use of the vertical line in the capital C and of the uncial A, together with the Gothic style of the letters, agrees with the Italianisms of spelling and language.30 The maps themselves, with their nautical basis and their language, also suggest an Italian context. The codex has been written and drawn in
54
Italy. The date of the codex and its maps can only be surmised but I would suggest it was made before the middle of the fifteenth century. The maps contain no hint of the Atlantic discoveries or of any new knowledge of northern and eastern Europe, as can be found on the maps of Claudius Clausson Swart and Nicolaus Cusanus and in Enea Silvio Piccolomini's written work.3' Yet, as is clear from the analysis of the maps, the cartographer had numerous and varied sources at his disposal: at least two different Ptolemaic codices, a text of Pliny, a
map of the Dalorto-Dulcert type, Italian portolan charts or atlases (one very probably drawn by Andrea Bianco) and a quantity of non-cartographical information about Asia and eastern Europe. Such a wealth of source material could only have been available in a major centre for the collection and elaboration of geographical and cartographic material, such as the Papal court, Florence, Venice, Genoa and Naples, or, to a lesser extent, Ferrara.32 The author of the Harleian codex was not totally ignorant of Latin but he wrote it badly with many mistakes. Often he failed to understand it, and he vernacularised it continually. In short, he used the language of the Italian portolan chartmakers. He wrote with a reasonably fluent hand but his transcription of the Ptolemaic text is one of the worst extant. He had, especially at the beginning of his work, particular difficulty in spelling words of Greek origin, which were evidently unfamiliar and not yet well understood.33 At the same time, he was familiar with map drawing. His hand is not that of a painter or illuminator nor is it that of an amateur-at least, not if we compare his work with the clumsy examples of the unknown authors of the Codex Tellerianus-Remensis (Bibliotheque Nationale, MS Lat. 31230) and the Cardinal Bessarione (Marciana, MS gr. 388). The numerous cave and corrigatur and the references in the margin are not sufficient for us to identify the Harleian copyist's nationality or his personal interests. Since the first and seventh books of the Geographialack marginal notes, the geometrical-mathematical and cosmographical aspects of the text may have held no interest for him. On the other hand, he was interested in, and familiar with, problems of scale reduction and with the armillary sphere. A lengthy note on folio 87v, after the indication of the relationship between the central parallel and the meridian in the first plate of Europe (Bk. VII: 'Exposicio omnium summarum quibus continetur in Europa tabula decem provincie quatuor et triginta'), reveals his considerable experience as a geographer and probably as a nautical or chorographical cartographer.34 That the Harleian author's knowledge of the current astronomical and cosmographical literature was much superior to his competence in Latin or Greek35is revealed by a clue that has hitherto been ignored in the study of fifteenth-century Ptolemaic manuscripts: the Latin spelling of the word 'parallel'. The Harleian copyist regularly used the spelling 'paralellus', instead of 'parallelus' (Appendix). This
was no error: 'paralellus' had an equal legitimacy and a longer tradition than the more familiar form: this spelling had been introduced by John of Holywood in his De Sphaera mundi (thirteenth century). But the Vatican Library's Greek manuscript (Urb. gr. 82), the original for the fifteenthcentury Ptolemaic tradition, contains only the likewise, Jacopo d'AnGreek spelling 'rrotpcAX-qXoa; gelo used the spelling 'parallelus' in his translation, either because he was following Pliny's usage (Naturalis Historia, Book VI) or because he did not know or did not wish to take into account the fact that for two centuries previously cosmographers had been using a different form, that of 'paralellus'.36 So most of the Latin codices derived from d'Angelo's work, especially the Florentine and other Italian ones, translate this word faithfully as 'parallelus'. Thus the non-humanistic provenance of Harley MS 3686 is evident both in its extremely bad Latin and in the use of the spelling 'paralellus'. Characteristics of text and writing indicate an Italian author of the first half of the fifteenth century. A plausible date for the Harleian manuscript would be before 1450 but after 1436, the year Bianco produced his atlas. Which of the Italian centres of geographical and cartographic production could have been the unknown map-maker's home? The Ptolemaic Tradition in Florence The language and glosses of the Harleian text together with the maps point to a cultural context that was still strongly allied to the medieval scientific tradition, where the main interest was cartographical and the techniques those of nautical cartography. None of these criteria applies to Florence. The geographical interests of the first generation of Florentine humanists (churchmen and notaries, above all), were not linked to problems of cartographical representation but centred on those of a general nature: the inhabitability of the various parts of the earth, the existence of the Antipodes and the spreading of the Christian faith among the distant peoples. Succeeding generations exhibited a more substantial interest in other aspects of Ptolemy's work. They focused on the identification of place names and the reconstruction of the ancient world and its transformation into their modern world (if they were historians and men of letters), or on the correct determination of astronomical coordinates of places for diagnostic and
therapeutic ends (if they were physicians like Paolo Toscanelli).37 Up to the end of the fifteenth century, however, nobody in Florence-not even Toscanelli-seemed interested either in constructing better Ptolemaic regional maps or in producing modern maps not from Ptolemy's data but according to Ptolemaic (that is, mathematical and astronomical) criteria. However, the possibility of constructing a Ptolemaic-type cartography had fascinated Germanic mathematicians from the first half of the fifteenth century onwards. Indeed, the only effective realisations of Ptolemaic projects were achieved by northerners who were working in Italy on a temporary or long-term basis or who were in touch with Italian scholars. Among such northerners were Claudius Clausson Swart, Magister Reinhardus, Nicolaus Cusanus and, above all, Nicolaus Germanus. Humanism had arrived much later in the Germanic territories than in Italy and the science inherited from the thirteenth century maintained its hegemony within the universities for much longer. On the other hand, the modern tradition of nautical-regional cartography, with its inland detail and portolan chart outlines, which allowed most Italian cartographers to ignore Ptolemaic (that is, astronomical) cartography until the following century, did not exist in the Germanic lands. Unsurprisingly, the authors of the first modern Ptolemaic maps of Italy were of Germanic origin, namely Nicolaus Germanus in Florence and the anonymous author of the Bologna edition of Ptolemy's Geographia(1477).3 Unlike the German mathematicians, native Florentine cartographers used nautical-regional maps to update their Ptolemaic codices with new regional maps. After Lapaccini and Buoninsegni's translations of Ptolemaic place names into Latin and their redrawing of the maps (1420-1430), no further developments were initiated in Florence until the translation into Italian of place names on the maps accompanying Francesco Berlinghieri's Le septe giornate de la Geographia (1482).39 This was a linguistic rather than a cartographical change, however, and the maps themeselves were straight copies of the original. The (modest) cartographical innovations of Pietro del Massaio are chronologically later than those of Nicolaus Germanus and it is therefore impossible to call them original.40 In Genoa, home of innovative fourteenth-century cartographers like Giovanni da Carignano, the fifteenth century appears to have seen no geogra-
55
phical production departing from the nautical model. No study and elaboration of Ptolemy's Geographia appears to have taken place, and it would be difficult to postulate that any Genoese could have undertaken the complex synthesis of the different traditions found in Harley MS 3686. The Aragonese in Naples were the most scientific among Italian cartographers of the fifteenth century, but they were interested in the use of the compass, in a new astronomical calculation of latitude, and in the dimensions of the terrestrial sphere, all of which took a form quite different from the Ptolemaic.4'
A VenetianProvenance?
56
I would suggest that Harley MS 3686 belonged to the geographical culture of Venice rather to than that of Florence or elsewhere in Italy. From the end of the fourteenth century, Venice had a flourishing cartographic school characterised by a strongly innovative attitude both to toponymy and to the structure of atlases and nautical charts.42 In 1436 the Venetian Andrea Bianco produced the first known example of a combination of nautical charts and Ptolemaic maps in the same volume when he inserted a well-drawn Ptolemaic world map and a synthetic world map of the fourteenth-century type, together with the customary charts, into his nautical atlas.43 But Harley MS 3686 is something different. If we consider the individual maps in Harley MS 3686 as a single image (as in Fig. 1), albeit an incomplete one, the result is probably the second-oldest extant fifteenth-century transitional or synthetic mappamundi; the first being the Fillastre-de Noha world map of 1414. The synthetic type of map had its origins in Venice with Pietro Vesconte's elaboration (c.1320) of Fra Paolino's mappamundi, in which the Romano-Christian cartographical tradition was combined with the practice of portolan chartmaking. Vesconte's method of synthesising different cartographic traditions continued to be accepted at Venice as a satisfactory way of representing the world until the end of the fifteenth century; it was used by Albertin de Virga (f7.1390-1428/29), for instance. But in c.1459 Fra Mauro stated on his mappamundi-the last great original effort to place all the known world within the framework of a non-astronomical and non-mathematical cartography-that although Ptolemy provided the measure against which map-making should be judged, it was impossible to integrate Ptolemy's world-picture
with the modern. He also pointed out that Ptolemy himself had recognised that in time better information would become available and should be used.44 Like Fra Mauro, and before he produced his own mappamundi, the author of Harley MS 3686 tried to redraft essentially portolanic outlines to accord with the dictates of Ptolemy and, unlike anybody else, he did it on a regional, almost Ptolemaic, scale. The Harleian's Author's Relationship with Andrea Bianco The author of the Harleian codex would have been, I suggest, a Venetian accustomed to drawing geographical maps, able to read Latin but not to write it correctly. He knew at least the fundamentals of cosmography and had access to a vast quantity of geographical information, including Pliny's text and numerous modern sources, both cartographical and written. He possessed a text of the Geographia,though not necessarily any of the maps, since all the Ptolemaic elements in his maps are derived directly from the text they illustrate and not from an external source. He could have achieved his codex without ever having seen the maps of Lapaccini and Buoninsegni. He could equally well, and more simply, have been looking at the small Ptolemaic (260 X 380 mm) world map, drawn on the first of Ptolemy's projections, which illustrates Bianco's nautical atlas of 1436. The name of Andrea Bianco recurs so frequently in connection with the Harleian codex that an examination of the possible relationship between the unknown author of Harley MS 3686 and the famous Venetian chartmaker is necessary. The Harleian manuscript has, as we have seen, several features in common with the atlas signed by Bianco in 1436: language, geographical names on the maps given in the genitive case instead of in the nominative, the spelling 'paralellus', various notes on the clima in the margins of the Ptolemaic text, place-names in heavily Venetianised spellings, the same outlines for the Atlantic islands and many other lands, and the legend describing the island of tille.45 It is noteworthy, too, that in the Harleian codex the areas of cartographical innovation (Scotland) and those best represented (northern and eastern Europe, from the Baltic to the Black Sea), border, at least in part, areas visited by Bianco: the Black Sea, the by then greatly reduced Byzantine territory, the British Isles and the North Sea, and North Africa.46 Although Andrea Bianco was not the only Venetian to frequent these areas, only he
is known to have drawn a chart in London (1448) and his maps were always strongly innovative. Other features on the maps in Harley MS 3686 also need to be taken into account. Ptolemaic regional boundaries (between taraconensis and hispania belgica nunc castella (f.15), for example), are marked on all maps. Traditionally this practice has been attributed to Nicolaus Germanus. However, boundaries on Germanus' maps are marked with red lines, either continuous or broken; boundaries on the Harleian maps are, on the contrary, marked by red or red and black points, and, since they have been drawn on a nautical rather than on a Ptolemaic basis, they follow a quite different course. Only on one of the Harleian maps, that of the Iberian Peninsula (Fig. 2), are boundaries shown differently; so far from the Ptolemaic world in time and location were the kingdoms of Navarre and Granada that their boundaries are indicated with continuous black lines according to the modern nautical usage.47 However, certain definitive differences between the Harleian codex and Bianco's atlas should be noted. The graphics and the scale of reduction are quite different. Bianco makes no mention of caspey mountains in Asia. And, most important, the hand responsible for the codex is not Bianco's, although similar to it. The two cartographical achievements clearly reflect the same linguistic and cultural environment: that of the Venetian cartographers and their clients.48 It is not possible to say whether the author of the Harleian codex was a member of Bianco's circle (or workshop) or whether he merely knew the 1436 atlas well. We cannot know if the codex was created for a particular owner (the copyist himself?), or if it was intended to be the prototype of a serial production. We can posit, however, that Harley MS 3686 dates from just after 1436. Evidence for Bianco's interest in Ptolemaic geography is found only in his early extant cartographical work, the Marciana nautical atlas (1436), composed before he embarked on the galleys of the Republic bound for the Sea of Azov (1437). Bianco seems to have dedicated himself to 'synthetic' cartography only at the end of the 1450s, in collaboration with Fra Mauro. If Harley MS 3686 really does come from Bianco's circle, it documents the continuity of interest in Ptolemaic geography which has always been assumed but of which little is known in its practical applications. As an attempt to reconcile modern with Ptole-
maic cartography, which failed or at any rate was too broadly hypothetical in an age when cartography was not so freely conjectural as it was to be in the following century, Harley MS 3686 would explain the reasons for Fra Mauro's well-known self-justification: I do not think that I detractfrom Ptolemyif I do not follow his cosmography,because if I had chosen to respect his meridians and parallels and degrees it would have been necessaryin the demonstrationof the areas now known in this latitude to omit many provinceswhich Ptolemydoes not mention, especially in the latitudebetween the south wind and the north wind which he calls'terraincognita',becausein his day it was not known; . . . and although at various times
some have navigated in the southern and northern parts,none the less they have not had time to measure or even to considerthis distance,theirvoyaginghaving been casualand not aimedat this navigationalpurpose. ThereforeI leave to EverlastingGodthe measurement of his works.49 It seems to me that Harley MS 3686 demonstrates that other attempts had been made to follow Ptolemy more closely than Fra Mauro had managed to do and that they had not succeeded.
Further Thoughts Whatever its provenance, the Harleian codex remains a unique testimony to relationships among the various cartographical and geographical traditions of the fifteenth century. It reveals the methods of someone outside the humanistic circles who approached the Geographia of Ptolemy with the practical objective of using it for a better reconstruction of the world-picture on the regional scale. We can conclude by conjecturing some of the problems the author of the Harleian codex may have faced. The first problem, having procured the text, must have been to understand its language. Ptolemy's Geographia contains Latinised Greek words that would have been unfamiliar to anyone without knowledge of Greek. For instance, the difference between 'thorographia/toragraphia' and 'corographia' (f.2) could not have been appreciated. In the process of copying Harley MS 3686, however, the errors in the transcription of the word 'corographia' become rarer and finally disappear. As the copyist proceeded, he learned the meaning (or at least the spelling) of the words he was transcribing. This would suggest that the copyist was a scholar who took an interest in the subject and was possibly copying it for himself. The geometrical part of Ptolemy's treatise must
57
have presented problems. In Books I and VII the procedures for drawing the world map are set out. In Jacopo d'Angelo's translation these instructions are almost incomprehensible, and the related drawings are incorrect. The author of Harley MS 3686 seems to have ignored all the difficulties, transcribing without comment or variant the text and the drawings as he found them in the codex he was copying. Significantly, the only instance of the spelling 'parallelus' in Harley MS 3686 comes in an explanatory geometrical drawing (f.2v), which the copyist failed to correct to conform with his usual spelling. It would look as though from the outset our author passed quickly over the chapters on the ecumene, preferring to devote himself to what he found more intelligible, namely the chorographic, or regional, chapters and their accompanying maps. These posed no particular geometrical problems, the only obstacle being the concept of central parallel and meridian, which the scholar was able to overcome, as he himself noted at the end of the Ptolemaic text, 'experientia circuli ... subtiliter investigando' [with subtle investigation, experimenting on the circle]. From this we may infer that others in his circle had already tackled the question but with less success. Inexpert at creating perspective, our scholar did know how to use the models of the cosmos (spheraartificialis)to interpret the Ptolemaic text correctly. So far as place names and associated coordinates are concerned, our author evidently knew that he was faced with both textual and interpretative problems. He seems to have had access to a second codex, since he was aware of discrepancies between its text and his own. Having no reason to believe more in one codex than in the other, however, and lacking a third with which to compare both, he was unable to correct the spelling of names and the coordinates. Instead he limited himself to adding a warning (with the formulae cave or corrigatur)that the text was doubtful. On the maps he retained the spelling found in his own text, even when it was unusual. On his own initiative, he noted the clima in which certain places were situated (for example, 'Costantinopolis 13. clima', f.40). Occasionally he included other marginal notes, like 'Parisius' beside 'Lucotecia' (f.21v) to link an ancient name with a modern, and 'delphicum appollo' beside 'deli insula civitas' in the text about Greece (f.36v), in an unconvincing flaunting of his classical learning. Notes referring to Pliny's work are found in Hellas 58
and Numidia
(ff.35V, 47); and a new detail on the
Lucanians has been added to the text of Italy ('Locanis numquam peste aut termotu passa sunt secundum Plinium in libro I cap.lo 98', f.29v). Our unknown author paid careful attention to the problem of intelligibility, a major concern in the fifteenth century. As new texts from the classical period became available, the humanists grew aware of the degree of change in the world and its places and peoples. Biondo Flavio's Italia illustrata(c.1450) opened with a comment on the way names and places had changed from classical times. And on behalf of those ignorant of Latin and interested in the modern world rather than the classics, Fra Mauro made the same point, indulging in a little irony over the fact that not even lettered men succeeded in agreeing among themselves on the identification of ancient sites.50 Likewise, the author of Harley MS 3686 tried wherever he could to translate the ancient world of Ptolemy into his own. He added to text and maps the modern equivalents he knew ('celtogalatie situs que Galia dicitur aut francie', f.20; 'Asya minor que nunc Turchia dicitur', f.34"; 'hostia norganzer aut laxartes', f.98, etc.). He went further, however, and gave Ptolemaic names to places on the modern map (the rivers bisurgeand albis, f.23V; Sarmathiaeurope, f.101, etc.), and, occasionally, a modern name to a Ptolemaic place (caspey mountains, ff.98"-99). Above all, on the maps of the Black Sea and the Caspian and of Africa he blended the familiar portolan chart coastal outline with a new Ptolemaic inland geography. Indeed, on the maps of the Black Sea (ff.41v, 100v-101), he offered different solutions for the same problem, the source of the Dnester, perhaps reflecting hypotheses formulated at different stages of his work. Countries like Spain and France presented no problem in reconciling ancient and modern, since the modern maps were sufficiently detailed for the interior. For regions mapped in less detail by modern cartographers, Ptolemy was used to offer a plausible and authoritative alternative. A particular problem is posed by the order of the Harleian text and its maps. The Ptolemaic order of depiction, especially in Europe, did not correspond to modern experience. By placing Greece and its islands, instead of Sarmatia, after Italy, the author of Harley MS 3686 must have wanted to bring the Ptolemaic text closer to the practical experience of his contemporaries who were accustomed, unlike Ptolemy, to consider the Mediterranean as a separate geographical area. Africa, however, con-
tinued to be kept separate from the Mediterranean (Book IV). We know from Fra Mauro that in Venice as elsewhere the cartographers' attention was centred on Ptolemy's Geographia and that they had doubts about both their own and Ptolemy's methods and material. Harley MS 3686 shows us how, at Venice, the problems raised by the Geographia were faced but not solved prior to Fra Mauro's time. It confirms, too, that the Venetians never relinquished the custom of blending the most up-todate information with traditional map-making processes, as Pietro Vesconte had been doing earlier on. Harley MS 3686 documents a forgotten phase in the transition from fourteenth and earlyfifteenth-century synthetic production to a different type of synthesis in the mid-fifteenth century (like Fra Mauro's, for example) in which account had to be taken of Ptolemy's geographical descriptions. The Harleian codex is older than Fra Mauro's world map. Moreover, it is the only surviving witness of how a synthesis of this kind was attempted on a regional scale. Manuscript submitted December 1994. Revised text received October1995. NOTESAND REFERENCES N.B. The errors in the Latin text of B.L. Harley MS 3686 are so numerous that I have not indicated them. I have also refrained from intervening with punctuation and capitalisation, confining myself to a palaeographic transcription which leaves its characteristics in evidence. The quotations from other texts conform with modern practice in regard to punctuation and capitalisation. 1. Joseph Fischer, Claudii Ptolemaei Geographiae Codex Urbinas Graecus 82 (Codices Vaticani Selecti, vol. 19; Leyden, Brill; Leipzig, Harassowitz, 1932, 3 vols.) 2. Sebastiano Gentile, Firenze e la scopertadell'America. Umanesimoegeografianel '400fiorentino,Catalogo dell'esposizione, Firenze, Biblioteca Medicea Laurenziana (19921993) (Florence, Olschki, 1992), 132-34; Sebastiano Gentile, 'Toscanelli, Traversari, Niccoli e la geografia', in Atti del ConvegnoFirenzee il MondoNuovo. Geografiae scoperte fra XV e XVIsecolo,Firenze, 6-8 ottobre 1992, published in RivistaGeograficaItaliana 100 (1993), 113-31. On the cost of illustrated Florentine geographical codices-between 50 and 100 florins-see Gentile, Firenze, 200-1, 209. A 'mappamondo' could be a large map of the world or a codex of Ptolemy's Geographiawith maps. 3. Marica Milanesi, 'Testi geografici antichi in manoscritti miniati del XV secolo', in Relazioni di viaggio e conoscenzadel mondotra MedioEvo e Rinascimento,Atti del V Convegno internazionale dell'Associazione per il Medioevo e l'Umanesimo Latini, Genoa 1991, published in Columbeis5 (Genoa, D.Ar.Fi.Cl.Et., 1993), 341-62. 4. This was Regiomontano's opinion, which was well known from his catalogue of scientific works, Haec opera fient in oppido Nuremberga Germaniae ductu Ioannis de Monteregio (Nuremberg, 1474). Roberto Weiss believed
that Jacopo d'Angelo was not a good translator ('Jacopo d'Angelo da Scarperia [ca. 1360-1410/11]', in Medioevoe Rinascimento. Studi in onore di Bruno Nardi [Florence, Sansoni, 1955, 2 vols.], 2: 801-27; reprinted in Roberto Weiss, Studies in Medieval and Humanist Greek [Padua, Antenore, 1977], 254-77). Weiss considers d'Angelo to have been faithful to the text but inaccurate and describes his Latin style as 'simply atrocious' (p. 821). A similar opinion is more blandly expressed by Johannes Werner in In eundem primum librum GeographiaeCl. Ptolomaei argumenta, printed with the new translation of Book I of Ptolemy's Geographia(Nuremberg, Ioannes Stuchs, 1514). Jacopo d'Angelo da Scarperia had not received a scientific education but, like many Florentine humanists of his generation, he had studied jurisprudence at Bologna University. On scientific studies in the fifteenth-century universities, and on the difference between the juridical and the medico-scientific curricula, see Dana B. Durand, The Vienna-KlosterneuburgMap Corpus of the Fifteenth Century.A Study in the Transitionfrom Medieval to Modern Science (Leyden, Brill, 1952), 32-93; Paul L. Rose, The Italian Renaissanceof Mathematics.Studieson Humanistsand Mathematiciansfrom Petrarch to Galileo (Geneva, Droz, 1975), 46-105; Charles B. Schmitt, 'La cultura scientifica in Italia nel Quattrocento: problemi d'interpretazione', Studifilosofici4 (1980), 55-70. Vladimiro Valerio has observed that the geometrical drawings in the codices of Nicolaus Germanus' Geographia were wrong, but the geographical plates have no mistakes, which confirmed that 'the text and the figures [were not always] checked by the mathematicians and, in any case, not with the same care devoted to the plates' (Vladimiro Valerio, 'Sui planisferi tolemaici. Alcune questioni interpretative e prospettiche', in Esplorazioni geografiche e immaginedel mondo nei secoliXV e XVI, Atti del Convegno, Messina 14-15 ottobre 1993 [Messina, Grafo Editor, 1994], 63-82, quotation 75). 5. The most recent important studies on the Florentine codices of the Geographiainclude those by Albinia de la Mare, 'New research on humanistic scribes in Florence', in Miniatura fiorentina del Rinascimento1440-1525, un primo censimento,ed. Annarosa Garzelli (Florence, 1985), 393600; Sebastiano Gentile, 'Emanuele Crisolora e la "Geographia" di Tolomeo', in Dottibizantinie librigrecinell'Italia del secoloXV, Atti del convegno internazionale, Trento 2223 ottobre 1990, ed. M. Cortesi and E. V. Maltese (Naples, D'Auria, 1992), 291-308; Gentile, Firenze (see note 2); S. Gentile, 'L'ambiente umanistico fiorentino e lo studio della geografia nel secolo XV', in L. Formisano et al., Amerigo Vespucci(Florence, Giunti, 1992), 12-45; Gentile, 'Toscanelli, Traversari, Niccoli' (see note 2); S. Gentile, 'Giorgio Gemisto Pletone e la sua influenza sull'Umanesimo fiorentino', in Firenzee il Conciliodel 1439, Convegno di Studi, Florence 29 novembre - 2 dicembre 1989, ed. Paolo Viti (Florence, Olschki, 1994), 813-32. These scholars have established the documentary foundation for all future study of the subject. 6. Durand, Vienna-KlosterneuburgMap Corpus (see note 4). Innovations continue to be considered generally of Italian origin, as asserted by Roberto Almagia ('II primato di Firenze negli studi geografici durante i secoli XV e XVI', Atti della SocietaItaliana per il progressodelle Scienze,Firenze 28 (1929); published separately, Pavia, Tip. Fusi, 1929; reprinted Florence, Societa di Studi Geografici, 1963). Such an interpretation is based on an erroneous dating (de la Mare, 'New research' [see note 5], 505, 567), and on many prejudices (Marica Milanesi, 'Presentazione della sezione "La cultura geografica e cartografica fiorentina del
59
60
Quattrocento"', in Atti del ConvegnoFirenzee il MondoNuovo (see note 2), 15-32. 7. Catalogustranslationumet commentariorum: Medievaland RenaissanceLatin Translationsand Commentaries,ed. F. E. Cranz, P. 0. Kristeller and V. Brown (Washington D.C. 1960-); Douglas W. Marshall, 'A list of manuscript editions of Ptolemy's Geography', Bulletin of the Geography and Maps Division, Special LibrariesAssociation87 (1982): 17-38. 8. Forgotten in the sense of unstudied, though not unknown to librarians. The secretary of the Hakluyt Society in a review of the principal maps of the British Museum, 1914, observed that the way in which Scotland was drawn suggested a Scottish draughtsman, and that Galway Bay resembled that in the Venetian portolan of 1489, also in the (then) British Museum (J. A. J. de Villiers, 'Famous maps in the British Museum', Geographical Journal (Aug. 1914), 168-88. Tony Campbell, Indexesto Materialof CartographicInterestin the Departmentof MSS and to MS CartographicItems Elsewhere in the British Library (London, British Library, 1992), lists it as 'ante 1450' and unstudied. 9. The codex measures 282 X 212 mm and is composed of folios 1-100 (paper) and 101 (parchment), numbered in red ink at top right in the copyist's hand, with modern numbering in pencil on the pages restored in 1882. No blank sheets. 'Monts' watermarks similar to Briquet 11696 (C. M. Briquet, Les filigranes. Dictionnaire historique des marques du papier (Paris and elsewhere, 1907, vol. 3). 'Cross n. 907, no place, no date, eye-copy from H.3686', in Edward Heawood, Watermarks,mainly of the 17th and 18th Centuries,Monumenta chartae papyraceae historiam illustrantia, 1 (Hilversum, Netherlands, The Paper Publication Society, 1950). All the pages in Harley MS 3686 are protected by reinforcement at the binding, which is modern. Text and maps in the same hand, probably from northern Italy, in Gothic bookhand. Written full page (text) or in two columns (lists of coordinates). In the margin, the copyist has made brief annotations, frequent corrections (cave, corrigatur)and references (place-names) and has added the clima of each province. Titles of chapters are in red with some initials in red and blue, though most of the spaces for initials remain blank. The plates and the maps inserted in the text are unframed. Waterbodies are in pale green; islands and the names of regions are in red; mountains are shown as continuous expanses of upland and coloured light brown; coasts in black are highlighted in yellow; rocks are rendered with black dots; boundaries between Ptolemaic provinces are delineated with red dots, sometimes highlighted in yellow; those separating modern territories (in the Iberian peninsula) are marked with continuous black lines. Small pen drawings represent the columns of Hercules and the Turrislapidea. 10. IndexBiographicumNotorumHominum,part C, vol. 10; John Venn and J. A. Venn, Alumni cantabrigienses(Cambridge, At the University Press, 1922), part 1 (to 1751), see Bagley, William. The Venns list another William Bagley, from Warwickshire, admitted to Cambridge from Eton in 1530. The handwriting of the owner's note is appropriate to the period and the environment in which both William Bagleys lived. See also P. J. Croft, AutographPoetryin the English Language (London, Cassell, 1973, 2 vols.). 11. Biblioteca Apostolica Vaticana (hereafter B.A.V.), Vat. Lat. 2974 (1409), B.A.V. Ott. Lat. 1771 (1411), Laur. XXX.5 (1426). These are all Florentine codices without a map and, therefore, they were not classified by Fischer. No systematic examination of the various types of
explanatory geometrical drawings in 15th-century Greek and Latin Ptolemaic codices has been made, though such analyses would provide a useful key for reconstruction of their stemma (like that which exists for the Greek manuscripts). All we can say here is that the later Florentine codices contain fewer geometrical drawings than the Greek and the older Latin codices which generally contain eight or nine drawings. 12. The quotation on folio 29v is inserted into the text and could have come from the manuscript from which Harley MS 3686 was copied, but the other two are in the margin in the same hand as the text and the maps. 13. David Woodward has given the term 'transitional maps' to world maps like the one under discussion, in which several cultural and technical experiences come together. I would rather define them as 'synthetic maps'. The term transition implies the existence of a unidirectional flow from one condition to another (for instance, from ancient to modern), and thus a progression or regression which is not recognisable in the history of ancient cartography. See David Woodward, 'Medieval mappaemundi', in TheHistoryof Cartography.I. Cartography in Prehistoric,Ancientand MedievalEuropeand the Mediterranean, ed. J. B. Harley and D. Woodward (Chicago, University of Chicago Press, 1987), 286-370, esp. 314; Marica Milanesi, 'La cartografia italiana nel Medio Evo e nel Rinascimento', in La Cartografia italiana. Cicle de conferenciessobre Histbria de la Cartografia (Barcelona, Institut Cartografic de Catalunya, 1993), 15-80, esp. 21. 14. The map by Cardinal Guillaume Fillastre, from Nancy (France), now in the Biblioteca Apostolica Vaticana, Archivio di San Pietro H.31, is the first known world map to show a mixture of Ptolemy and Catalan-style portolan charts. Fillastre was one of the earliest scholars, together with his friend Pierre d'Ailly, to check Ptolemy's geography against that in other classical texts, e.g., Pomponius Mela and Pliny. Pyrrus de Noha, who signed the map and by whose name it is usually known today, was only the scribe. The map accompanies Fillastre's commentary of Pomponius Mela. 15. The method of transfer seems to conform to the description given a century later in Breve compendiode la spheray de la arte de nauegar con nuevosinstrumentosy reglas exemplificandocon muy subtiles demonstraciones:compuesto por Martin Cortes natural de Burjalaroz en el reyno de Aragon y de presente vezino de la ciudad de Cadiz: dirigido al invictissimo Monarcha Carlos quinto Rey de las Hespanas etc. Senor Nuestro (Seville, Anton Alvarez, 1551), Chapter 2. According to Cortes the lines of direction were drawn with ruler and compass on a sheet of parchment (or, here, paper). The coastal outline was traced from a pre-existing portolan chart on to transparent wax paper. The wax paper was then attached to the parchment, and the lines were retraced using a thin but not sharply pointed stylus in such a way as to leave the indentation of the stylus on the surface of the parchment without damaging it. Finally the indentation was filled in with ink. Next the names were written; first the most important ones in colour, and then the secondary ones in black. Thereafter cities, ships, flags and animals were drawn; regions and other notable things were indicated, and all were beautified with colour. Finally the scale and the determination of the degrees were added. On the use of wax paper for copying drawings see also Vladimiro Valerio, Societa Uomini e Istituzioni Cartografiche nel Mezzogiorno d'Italia (Florence, Istituto Geografico Militare, 1993), 82, n. 50, where reference is made to
the treatise on painting by Cennino Cennini. For other systems of copying see Tony Campbell, 'Portolan charts from the late XIIIth century to 1500', in Harley and Woodward, The Historyof Cartography(note 13), 391; and Piero Falchetta, 'Marinai, mercanti, cartografi, pittori. Ricerche sulla cartografia nautica a Venezia (sec. XIVXV)', Ateneo Veneto. Rivista di scienze, lettere ed arti 182 (1995). 16. The Atlantic and Mediterranean countries seem to have been really drawn to the same scale. This would indicate a cartographical context in which the 1403 correction of Francesco Beccari had been accepted. See Campbell, 'Portolan charts' (note 15), 427. 17. See Campbell, 'Portolan charts' (note 15), and Carte da navegar.Portolanie carte nautichedel Museo Correr13181732, Catalogo dell'esposizione, Venice, Museo Correr, 1990, ed. Susanna Biadene (Venice, Marsilio, 1990), note 9. 18. Now in Venice, Marciana Library,MS It. Z 76 (4783). A photographic edition of the atlas has recently been published: Atlante Nautico di Andrea Bianco, with an introduction by Piero Falchetta (Venice, Arsenale, 1993). 19. The Balkans were at this time little known to the Venetians: Alain Ducellier, 'La penisola balcanica vista dall'osservatorio veneziano nei sec. XIV e XV', in Europae Mediterraneotra Medio Evo e prima Eta Moderna,L'osservatorioitaliano, Atti del convegno, Centro di studi sulla civilta del Tardo Medio Evo, San Miniato al Tedesco, 1990 (Pisa, Pacini, 1992), 297-314. 20. These include the maps of Gabriel de Vallseca (Maiorca, 1439 and 1447), with which, however, the Harleian maps have nothing else in common. See Campbell, 'Portolan charts' (note 15), 378. See also Piero Falchetta, 'Manuscript no. 10057 in the Biblioteca Marciana, Venice: a possible source for the Catalan Atlas?' ImagoMundi 46 (1994), 19-28. 21. Roberto Almagia, MonumentaCartographicaVaticana, 1. Planisferi, Carte Nautiche e affini dal secolo XIV al XVII esistenti nella Biblioteca Apostolica Vaticana (Vatican City, Biblioteca Apostolica Vaticana, 1942), 32-41, pl. XIII. Falchetta, 'Marinai, mercanti, cartografi, pittori' (see note 15), esp. note 84, thinks this map is a copy from Fra Mauro by Grazioso Benincasa. 22. The legend on Harley MS 3686 reads: 'insula tille est loco inhabitabilis [ ... margin torn] / yemen est magnam frigilitatem [sic for frigiditatem] et [ . . . margin torn] / caliditatem et sunt insulae triginta'. In Bianco it is: 'ttile est lochus inabitabilis / quam in istate nichil potes / cresere per chalorem et in gieme / propter mag[ni] frigoris chongelacionis'. The myth of the extreme summer heat of Thule had been transmitted from the first half of the 14th century by the Polychroniconof Ranulf Higden. Luigi De Anna, Il mito del Nord. Tradizioniclassichee medievali (Naples, Liguori, 1994), 45, n. 98. 23. Brancie, the city of Brjansk, in the central Russian upland is better placed here, therefore, than the branzica or branchichaon the Dnepr in the nautical cartography, for which see Giuseppe Caraci, Italiani e Catalaninella primitiva cartografia nautica medievale (Rome, Istituto di Scienze geografiche e cartografiche dell'Universita degli Studi, 1959), 50-53. 24. The Danube separates polana in the north from ungariain the south, as in Bianco. Septemcastranunc veginel is a toponym that dates back to the Dalorto-Dulcert maps, on which, however, it has been placed farther east, north of the Black Sea. 25. See, for example, f.98: daitis instead of daix; f.99v: masotilifluvii instead of masotilusfluvius. The error was not
common in either Ptolemaic or nautical cartography. This usage, however, was found frequently in Bianco's 1436 atlas. 26. Norganzeris found on the mappamundi in the Bianco atlas; Norganziis the spelling on the nautical chart of 1367 signed by the Pizigani brothers. The Catalan Atlas (1375) has the form Organci, which is nearer to the original Urgent, on the Syr Darja. 27. Thanks to their classical origin and their association with the Alexander romances, the caspeyor caspicimontes can be found in a large number of medieval mappaemundi. In the 15th century they were still part of general knowledge: on the mappamundi of Fra Mauro, a legend placed in the sea to the north of the populi diti bargu reads 'some believe that these mountains are the monti caspii, but this opinion is not true'. Evidently, the author of Harley MS 3686 was among those who upheld the not true opinion; for he labelled all the mountains that Ptolemy placed in the farthest parts of Asia caspey,except for the aurey mountains. The aurey mountains are probably a combination of the Anarei of Ptolemy with the montes aurei, which Isidore of Seville sited in the remotest parts of northern Asia. 28. The text of Harley MS 3686 has tingitania,whereas in 15th-century Ptolemaic maps, including Bianco's world map of 1436, Tinganicais more often used. Fra Mauro has Tingintanea. 29. See Campbell, Portolancharts (note 15), 378. 30. S. Harrison Thompson, Latin Bookhands of the Later Middle Ages 1100-1500 (Cambridge, University Press, 1969), 70 and passim. 31. Marica Milanesi, 'La rinascita della geografia dell'Europa (1350-1480)', in Europa e Mediterraneotra Medio Evo e prima Eta Moderna (see note 19), 35-59. 32. For a brief evaluation of the cartographic activity at Ferrara in the 15th century see Marica Milanesi, 'Ii commento al Dittamondo di Guglielmo Capello (14351437)', in Alla corte degli Estensi. Filosofia, arte e cultura a Ferrara nei secoli XV e XVI, Atti del Convegno Internazionale, Ferrara, 1992 (in press). For the cartographical production of the Kingdom of Naples see the ample treatment, the fruit of original research, published recently by Valerio, Societa Uomini e Istituzioni (note 15). With the publication of Piero Falchetta's research (see notes 15, 18), an up-to-date evaluation of Venetian chart and map making in the 14th and 15th centuries is beginning. The same cannot be said of the other Italian centres. The attention of historians of cartography and geography continues to be fixed on Florence. Even the interesting 15th-century map production in Genoa remains little known (Gaetano Ferro, 'Cartografi e dinastie di cartografi a Genova', in Due mondi a confronto 1492-1728. CristoforoColomboe l'aperturadegli spazi. Mostra storico-cartografica (Rome, Istituto Poligrafico e Zecca dello Stato, 1992, 2 vols.), 1: 245-61. 33. For instance, on folio 2, 'toragraphia' for 'corographia'. 34. 'Nota cumque dicit medium ipsius pa[ra]lelli proporcionem habere ad meridianum quemadmodum in prima tabula europe quam undecim ad viginti degradibus intellegit tabule linearis pa [ra]lelli qualiter se habet gradus longitudinis ad gradum latitudinis paralellus enim qui per medium ipsius scilicet loci est proporcionem habet ad meridianum latitudinis quam undecim ad viginti eadem modo in reliquis comparandum est et non totam longitudinem ad latitudinem ex quo accidit in quibusdam tabulis maiorem fieri longitudinem latitudinem et aliquibus e converso quasi dicit quod gradus undecim in
61
62
meridiano ipsius tabule prime et ante quantitate quante fuit viginti in paralello quod experiencia circuli dimensurabis subtiliter investigando.' [Note that when [Ptolemy] says that the central parallel is to the meridian as 11 degrees are to 20, as in the first plate of Europe, he means the proportion between the degree of longitude measured on the linear parallel of the plate, and the degree of latitude: in fact, the parallel which passes through the centre of that place has a proportion of 11 to 20 with the meridian of the latitude. With all the other maps it is necessary to compare in the same way [the length of the degrees], not the entire values of longitude and latitude; wherefore [because of this erroneous interpretation] it happens that on certain plates the longitude is greater than the latitude, and on others vice versa, as if [Ptolemy] said that 11 degrees on the meridian of the first plate are equivalent to 20 degrees on the parallel. Something that you will measure, with subtle investigation, experimenting on the circle.] This note is not present in any other Ptolemaic manuscripts known to me. 35. Iacobus Angelus, translator of the Geographia (and also of Plato and Plutarch), was for him jacobus anglus' (f. 1), compatriot (and colleague?) of Robertus Anglus whose writings on the quadrant, the astrolabe and the gnomon were studied in all Italian universities. A glosser of a 1448 Ptolemaic codex of Germanic provenance fell into the same error (B.L. Harley MS 3290). 36. Jacopo d'Angelo did not justify his choice of the spelling 'parallelus', though he did explain his decisions elsewhere (between the terms 'geographia' and 'cosmographia', for instance, and for the use of 'cosmographia'). In this case he was implicitly declaring himself in favour of the word already in use, thus avoiding introducing a useless neologism for a concept already known. 37. For these characteristics see the studies by Gentile cited in note 5. 38. He was perhaps Nicolaus Germanus himself. See Durand, Vienna-Klosterneuburg Map Corpus(note 4), 155. 39. Roberto Almagia, 'Osservazioni sull'opera geografica di Francesco Berlinghieri', Archiviodella Regia Deputazione Romana di Storia Patria 67 (1945): 209-57 (reprinted in Almagia, Scrittigeografici,ed. Lucio Gambi (Milan, Cisalpino, 1961), 497-526); Rossella Bessi, 'Appunti sulla Geographia di Francesco Berlinghieri', in Atti del Convegno Firenzee il MondoNuovo (see note 2), 159-75; Milanesi, 'La rinascita della geografia dell'Europa' (see note 31). 40. Germaine Aujac, 'Le peintre florentin Piero del Massaio, et la Cosmographie de Ptolem&e', Geographia Antiqua, 3-4 (1994-1995): 187-209. 41. Vladimiro Valerio, 'Astronomia e cartografia nella Napoli aragonese', in Atti del ConvegnoFirenze e il Mondo Nuovo (see note 2), 291-303; V. Valerio. Societi Uomini e Istituzioni(see note 15), 73-98. 42. Campbell, 'Portolan charts' (see note 15); Falchetta, 'Marinai, mercanti, cartografi, pittori' (see note 15). 43. An association between Ptolemaic world maps and nautical atlases or general geographical works was unusual at the time. See Falchetta, Atlante Nautico di Andrea Bianco (note 18). Another Ptolemaic world map drawn between 1455 and 1471 by an unknown Venetian cartographer is in the nautical atlas known as Deissmann 47, now in the Topkapi Museum, Istanbul (Falchetta, 'Marinai, mercanti, cartografi, pittori' [see note 15], esp. note 131). A Ptolemaic world map, drawn at Ferrara by a Florentine illuminator, illustrates Guglielmo Capello's commentary (c.1435-1437) on the 14th-century Dittamondoof Fazio degli Uberti; the same work also contains a
mappamundi organised by climatic zones, and one with the seven climata (Milanesi, 'f1 commento' [see note 32]). 44. Fra Mauro wrote: 'This work ... does not possess in itself that completeness which it should, because certainly it is not possible for the human intellect without some demonstration from Above to verify in every respect this cosmography or mappemonde . . . Wherefore if anyone objects to this work because I have not followed Claudius Ptolemy, whether in the form or in his measurements of longitude and latitude [per longefa e per largefa], I do not wish to defend Ptolemy in more detail than he defends himself, when in the first chapter of his second book he says that one can speak correctly of those parts which one practises continually, but that of those less frequented nobody should think that anyone can speak so correctly. Understanding, however, that he could not verify his cosmography in every point, the task being long and difficult, life short and experimentation fallacious, the fact remains that he admits that in the course of time this description may be bettered or surer information may be available than he had. Therefore I say that in my time I have sought to verify his writings by experiment, investigating for many years and frequenting trustworthy persons, who have seen with their own eyes that which here I faithfully demonstrate'. This translation is based on the transcription by Tullia Gasparrini Leporace in II Mappamondo di fra' Mauro, with an introduction by Roberto Almagi& (Rome, Libreria dello Stato, 1956), pl. XL. See also Gunther Hamann, 'Fra' Mauro und seine italienische Kartographie seiner Zeit als Quellen zur friihen Entdeckungsgeschichte', Mitteilungen des Instituts 78 (Hermann Bohlaus fur Osterreichische Geschichtforschung Nachfuhrung, Wien-Koln-Graz, 1970): 358-71; Wojciech Iwanczak, 'Entre l'espace ptolemaique et l'empirie: les cartes de fra' Mauro', Midievales 18 ('Espaces du Moyen Age', printemps 1990): 53-68. 45. The use of the clima instead of the parallel is found in Bianco's version of the Ptolemaic world map. It obviously indicates a knowledge of the Almagestand its derivatives, but climata are also common in other 15th-century codices of the Geographia, all (in so far as I know), Florentine: B.A.V. Vat. 5698 (beginning 15th century, only on the world map); Laur. XXX.2 (Pietro del Massaio, c.1455-1462); Laur. XXX.3 (Nicolaus Germanus, c.1466); Laur. Edili 175 (G. Antonio Vespucci, after 1475). 46. Bianco was described as ammiraglioof the fleet of the Tana (1437), and ammiraglio and uomo di consiglio in Flanders and England (1438, 1443, 1446, 1449, 1450, 1451), Beirut and Alessandria (1440), Romania (1445) and Barbary (1447). See Angela Codazzi, 'Andrea Bianco', Dizionario Biografico degli Italiani 10 (Rome, Istituto Poligrafico dello Stato, 1968): 223-25. See also Falchetta, 'Marinai, mercanti, cartografi, pittori' (note 15); and Falchetta, Atlante Nautico di Andrea Bianco (note 18). The areas represented most often and with the greatest innovations in Harley MS 3686 are on the margins of Europe. In the course of the 15th century the northern countries and those of central and eastern Europe were increasingly depicted in written descriptions and through geographical and nautical charts with inland detail (Milanesi, 'La rinascita della geografia dell'Europa' [see note 31], passim). 47. Regional boundaries-except for Granada-were unusual on 15th-century nautical and world maps. We find them only in the mappamundi of Fra Mauro and in the Borgiana V map in the Biblioteca Apostolica Vaticana, both realised with the collaboration of Andrea Bianco
(Roberto Almagi&, MonumentaCartographicaVaticana [see note 21], 32-40, esp. 36, 38). 48. Fra Mauro was thinking of clients who did not know Latin or did not read the classics when he explained: 'In this work [the mappamundi destined for the King of Portugal] I have of necessity consented to use modern and vernacular terms because truly had I done otherwise few would have understood me except a few lettered men, and it happens that they too cannot reconcile their authors with the practices of today' (Leporace, II
Mappamondo di fra' Mauro [see note 44], pl. XXX). Francesco Berlinghieri in 1482 published his modified text and maps of the Geographiatranslated into the Italian vernacular tongue for the same type of users: Le septe giornate de la Geographia (Florence, Nicolaus Laurentii Alamanus, c.1482). 49. For the transcriptions see Leporace, II Mappamondodi fra' Mauro (note 44), p1s. XLI, XXXIX. 50. See note 48.
APPENDIX: Paralellus/parallelus The first translator into Latin of Ptolemy's Geographia, Jacopo d'Angelo disliked neologisms but used, without explanation, the spelling 'parallelus' instead of the form found in medieval scientific literature, 'paralellus'. D'Angelo probably knew philology better than science, creating his neologism through ignorance of mathematics and astronomy; as Johannes Werner wrote, 'ob summam quadruvii et mathematicarum artium imperitiam' (see text note 4). In short, d'Angelo cannot have read John of Holywood's treatise 'De Sphaera mundi' (1 3th century) whence the spelling 'paralellus' comes. Thanks to Holywood, 'paralellus' became the normal spelling for all medieval astronomical literature: we find it in Pietro d'Abano's Conciliatorcontroversiarumat the beginning of the 14th century, in the Imago Mundi and in the commentaries on Ptolemy's Geographiaby Pierre d'Ailly at the beginning of the 15th century, in the manuscripts and 15thcentury printed editions of the 'Sphaera' itself, and in Ptolemy's Quadripartitusprinted at Venice by Erhard Ratdolt in 1483. The spelling was adopted by Guillaume Fillastre in the maps and text of his 1427 manuscript edition of Ptolemy's Geographia,by Andrea Bianco in the Ptolemaic world map in the atlas of 1436, and occasionally by Fra Mauro on the circular world map of c.1459, where, however, the spelling 'parallelo' appears more often. In mapless Germanic copies of Ptolemy's Geographia, such as the British Library's Harley MS 3290 or the Codex Tellerianus-Remensis (Paris, Bibliotheque Nationale, MS Lat. 3123), too, the form 'paralellus' is used regularly. We find 'paralellus' in the printed map of Germany by Nicolaus Cusanus, dated Eichstatt 1491, whose original drawing dates back to mid-century. In Germany, Jacopo d'Angelo's new spelling 'parallelus' was used only by humanists like Johannes Regiomontanus and, later, Johannes Werner and Willibald Pirckheymer. Otherwise, the traditional 'paralellus' form was the norm for Nicolaus Germanus's manuscript Ptolemaic maps. It was also used (though only on the world map) in Ptolemy's Geographia,printed in Bologna 1477; it appears in text and maps of the Geographiapublished at Ulm in 1482 and 1486, and in the commentary, written in Upper Germany, to the Ulm edition of 1486. In the first half of the 16th century, as 'paralello', it was used by the Portuguese glosser of the Geographia,Pedro Nufies (Petri Nonii Salaciensis opera [Basel, Sebastianus Henricpetri, 1592]). The spelling 'paralellus' thus can be used as a clue in the complicated story of the Latin codices of Ptolemy's Geographia, whose provenance is always so difficult to establish. It is a sign for the traditional scientific culture, based on medieval literature, as opposed to the humanistic scholarship, based on the study of ancient texts. We find the old form 'paralellus' in non-humanistic cultural environments both in Germanic areas and in Italy, mainly outside Florence. Although Nicolaus Germanus worked in Florence, he also used the spelling 'paralellus'. This suggests that he had arrived from Germany with an already established cosmographical culture rather than learning a new one in Italy. Germanus was a fair enough Latinist, but he did not adopt the humanistic spelling 'parallelus' then current in Florence: he seems to have been sure that his version was the more correct. The dubious and sometimes groundless reconstruction of Nicolaus Germanus' life is the weak point of Durand's otherwise fine book The Vienna-Klosterneuburg Map Corpus(see text note 4). His well-documented researches on Germanic cartography in the first half of the 15th century and his implied criticism of the 'Florentine priority' in 15th-century Ptolemaic production, however, remain essential to the evaluation of Italian cartographical production between 1460 and 1480. In 1980, Joseph Babicz ('Donnus Nicolaus Germanus, Probleme seiner Biographie und sein Platz in der Rezeption der ptolemaischen Geographie', in Land- und Seekarten im Mittelalter und in der fruhen Zeit, Cornelius Koeman, ed. [Wolfenbutteler Forschungen. Hrsgg. von der Herzog August Bibliothek, Bd. 7; Munich, Kraus International Publications, 1980], 9-42), rejected Durand's idea that Nicolaus arrived in Italy with his cartographic ideas already formed by studies at Klosterneuburg or in a similar environment (pp. 81-86), but in 1987-1989, after carefully examining new documents, he accepted Durand's thesis ('The celestial and terrestrial globes of the Vatican Library, dating from 1477, and their maker Donnus Nicolaus Germanus (ca.1420-ca.1490)', Der Globusfreund35-37 (1987-1989): 155-66). It seems to me beyond doubt that Nicolaus Germanus arrived in Italy with a rich astronomical, cosmographical and cartographical experience. With his Germanic university heritage, Germanus applied himself in Italy to an already typical Florentine activity, the production of luxury editions of Ptolemy's Geographia.Germanus' innovations are of German origin. The practice of indicating the number of miles per degree and the use of trapezoidal and conic projections for regional maps Map Corpus had already been used in the Trier-Koblenz fragments X and Y dated 1437 (Durand, The Vienna-Klosterneuburg [see text note 4], 84, 145-59). One of Germanus' minor innovations was to use a broken line in red to mark regional boundaries. On the other hand, Ptolemaic regional boundaries on a modern map were, apparently, indicated for the first time in the codex discussed in this paper, B.L. Harley MS 3686. RESUM12: Le texte du MS Harley 3686 de la British Library est un exemplaire Geographia de Ptolemee Afrique.
enrichi d'un ensemble
On peut inferer des rapprochements
latin anonyme
inedit de 18 cartes non ptolemeennes avec l'atlas nautique
d'Andrea
Bianco
et sans date de la d'Europe, (1436)
Asie et
un origine
63
venitienne pour ce manuscrit ainsi qu'une datation entre 1436 et 1450. Le contour des cartes s'inspire des portulans, mais le detail topographique des terres et la toponymie semblent provenir du texte de Ptolemee. Ce manuscrit represente l'un des plus anciens exemples de la synthese entre le portulan, la carte de Ptolemee et la mappamundi du Moyen Age qui ont marque la cartographic du XVe sie&cle,et en meme temps l'unique exemple d'une telle synthese a Le'chelon regional. Le MS Harley 3686 revele certains des problemes techniques et methodologiques que la Geographiadut poser aux cartographes de cette epoque. ZUSAMMENFASSUNG:Der Text der Handschrift Harley 3686 in der British Library ist eine anonyme, nicht datierte lateinische Fassung von Ptolemaus' Geographiamit einem neu eingefuhrten Set von achtzehn nichtptolemaischen Karten von Europa, Asien und Afrika. Verbindungen mit Andrea Bianco's See-Atlas (1436) deuten auf einen venezianischen Ursprung der Handschrift und auf eine Datierung zwischen 1436 und 1450. Die Kartenumrisse stammen von Portolankarten ab, aber die topographischen Details und Toponymen im Binnenland scheinen im ptolemaischen Text ihren Ursprung zu haben. Der Kodex bildet eines der altesten Beispiele einer Synthese von Portolankarte, ptolemdischer Karte und mittelalterlichen mappamundi, welche die Kartographie aus dem fiinfzehnten Jahrhundert kennzeichen, und er ist das einzige Beispiel einer solchen Synthese auf regionalem Niveau. Harley 3686 enthuillt einige der technischen und methodischen Probleme welche die Geographiaan die zeitgenossischen Kartenmacher gestellt hat.
Forthcoming International Conferences and Public Lecture Series of Interest to Historians of Cartography October 3-5 October 6-9
October 24-26
October 31November 3
March 14-16
64
1996 8 Kartographiehistorischen Colloquium, Bern. 15th International Symposium of the International Map Collectors' Society (IMCoS), Riga, Latvia. 12th Kenneth Nebenzahl, Jr., Lectures in the History of Cartography: 'Maps on the move: cartography for transportation and travel'. Society for the History of Discoveries, Portland, Maine. 1997 'Surveying the Record: North American Scientific Exploration to 1900', sponsored by the American Society Philosophical Library, Philadelphia.
'Celebration of Helen Wallis', St Hugh's College, Oxford. June 22-28 18th International Cartographic Conference, Stockholm. 17th International Conference on July 6-10 the History of Cartography,Lisbon. August 14-16 Society for the History of Discoveries, St John's, Newfoundland. September 26-9 16th International Symposium of the International Map Collectors' Society (IMCoS), Budapest. April 12
October
1998 17th International Symposium of the International Map Collectors' Society (IMCoS), Tokyo. 1999
July
18th International Conference on the History of Cartography, Athens.