4 • The Beginnings of a Cartographic Tradition GERALD R. TIBBETTS INTRODUCTION as otherArab literature, from a mixture offolk literature andthetraditionformed byIslamic religiouswriting.The Before the establishment of Islam in the seventh century Qur)an and hadith, besides having a cosmological basis A.D., there is no evidence that the early Arabs accepted that pervades most of the later geographical literature, theideaofrepresentingthelandscapeinasystematicway. also contained a factual geographical element. Telling Practical use ofmaps is documented in only the slightest stories and reciting poetry about desert life augmented manner: even the tribesmen facing in the direction of this. Earlyin Islamichistoryacorpusofgeographicallore their sacred tribal enclosure when praying do not seem musthavedevelopedbasedonakhbar(narratedhistorical to have needed a mnemonic for thepurpose. Arabs seem traditions)andCaja)ib(storiesaboutmarvelouseventsand to have taken up the idea ofmapping onlywhen absorb things), two genres of narrated information that were ing the cultures of the peoples they conquered. written down at an early stage.5 Texts like the CAja)ib al Toward the end of the first century of Islam we do Hind (Wonders of India) and the Akhbar al-Sfn wa-al find a few literary references to military maps. A map Hind (Traditions about China and India) have survived was prepared (ca. 83/702) of the country of Daylam, as some ofthe earliestwritten Arab literary texts, but no south of the Caspian Sea, for al-I:Iajjaj ibn Yasuf(d. 95/ doubt there were other early collections that have been 714), the governor of the eastern part of the empire, so lost.6 that he could understand the military situation from his An example of early lore that has a bearing on carto capital in Iraq (fig. 4.1).1 Similarly, he ordered a plan of graphy is a concept of the world landmass in the shape Bukharasothathecouldacquainthimselfwiththelayout of a bird. This idea may come from early Islamic times, ofthe city while preparingfor its siege (89/707).2 Amap since the information, first given by Ibn cAbd al-I:Iakam of the swamps of al-Bapbah near Basra was also said to (d. 257/871), is mentioned asa tradition emanatingfrom be available in the time of the Abbasid caliph al-Man~ar a prominent Arab of the late seventh century. Ibn cAbd (d. 158/775), seemingly this time because of a dispute al-I:Iakam reports that the head of the bird represented overthe provision offresh water.3Aplanwasalso reput edly drawn in 141/758 for the round city of Baghdad planned for the caliph al-Man~ar.Since this was virtually 1.AbmadibnMubammad ibnal-FaqIhal-HamadhanI, Kitiibal-bul a strongly walled fortress in which only the privileged dan;seeCompendiumlibrikitabal-baldan,ed.MichaelJandeGoeje, BibliothecaGeographorumArabicorum,vol.5(Leiden:E.J.Brill,1885; lived, the drawing was less a cityplan than an architect's reprinted, 1967),283. site plan,4 but since the city had a diameter of two kil 2.AbuJacfarMubammadibnJarIral-TabarI,Ta)rfkh al-rusulwa-al ometers or more, the plan may have been a substantial muluk; see Annales quos scripsit Abu Djafar Mohammed ibn Djarir cartographic attempt. What it was like we have no idea, at-Tabari, ed. MichaelJan de Goeje, 15 vols. in 3 ser. (Leiden: E.J. since nothing of the plan or of the original round city Brill, 1879-1901; reprinted 1964-65),2d ser.,2:1199. 3.AbmadibnYabyaal-BaladhurI,Futu~al-buldan;seeLiberexpug has survived, and literary accounts are so much at vari nationisregionum,ed.MichaelJandeGoeje(Leiden:E.J.Brill,1866), ance that we cannot conjecture the form or extent of 371. the cartography involved. 4. Abmad ibn AbiYaCqubal-YaCqubi, Kitab al-buldan;see Kitab al We have no more information about these obviously boldan, ed. MichaelJan de Goeje, BibliothecaGeographorum Arabi practicalmaps,noristhereanylinktolaterMuslimcarto corum,vol.7(Leiden:E.J.Brill,1892;reprinted,1967),238.Notethat the Arabic word for drawing a plan here is ikhtatta, from which the graphy. Mappingduringthe period Iamdiscussingseems word khiffah, "aplan," is derived. The word usually used in classical to have first appeared and then continued only as an Arabicformap is§urah,"illustration,"which isalso thewordused in adjunct to geographical literature: surviving maps occur the three examplesgiven above. only as illustrations to geographical texts (table 4.1). 5.GeraldR.Tibbetts,AStudyoftheArabicTextsContainingMate rialon South-eastAsia(Leiden: E.J. Brill, 1979),3-4. 6.BuzurgibnShahriyar,KitabCaja)ibai-Hind;seeLivredesmerveilles EARLY GEOGRAPHICAL LITERATURE derInde,ed. PieterAntonievan derLith,trans. L. Marcel Devic(Lei den:E.J.Brill,1883-86);)Al]bara§-$inwa1-Hind:RelationdeLaChine Arabgeographical literaturebeganin much thesameway etdet'Inde, ed. and trans.Jean Sauvaget(Paris: Belles Lettres, 1948). 90 The Beginnings ofa Cartographic Tradition 91 70' SO" TRA~~O~\ANA BUkhara ·sama~and Balkh, KHURASAN - MAKRAN INDIAN o 250 500miles .. o1-1---r,_25.0l....1"50r0--..-75'0.'km ... 30' 40' 50' 7f1' FIG. 4.1. REFERENCE MAP OFTHE MIDDLE EAST. Thisgeneral map ofthe Middle East during the early periodshowsthe approximate borders of the provinces of the Islamic empire. China, the right wing India, the left wing al-Khazar tancesthroughoutthe Islamicworld. Ibn Khurradadhbih, (North Caucasus), and the tail North Africa. The tradi however, does extend the routes through known non tion may be Iranian.? However, the world maps of the Islamicareas;for instance,hegivesthesea routeto China al-BalkhI school (tenth century A.D.) show this bird clearly with Arabia as the head, Asia and Africa as the 7.S.MaqbulAhmad,"Qjughrafiya,"inTheEncyclopaediaofIslam, wings, and Europe as the tail, so the tradition may be a new ed. (Leiden: E.J. Brill, 1960-),2:575-87,esp. 576. Abu al-Qasim much later idea gained from looking at a map of this 'Abd al-Ra~manibn'Abdallah ibn 'Abd al-l:Iakam, KitiJb futul; mi~r; see Lelivredela conquetedel'Egyptedu Magreb etdel'Espagne, ed. sort.8 Henri Masse (Cairo: Imprimerie de l'lnstitut Fran"ais, 1914), 1, ema The earliest nonliterary geographical source materials natingfrom 'Abdallah ibn 'Amribn al-'As. were lists of pilgrimage and post stages throughout the 8. Forthe Balkhischool maps,see below, chapter5. Islamic world, giving the distances between the stages. 9. Abual-Qasim'UbaydAllah ibn'Abdallah ibn Khurradadhbih (or Thoughcompiledforadministrativepurposes,thesewere Khurdadhbih)livedabout204-300/820-911,sohisgeographicalwork was written in his youth if this date (de Goeje) is correct. He was a soon adopted by writers with a geographical bent and governmentadministratorbyprofessionandheldthepositionsofdirec appearinliteraryworks. These listsand the two previous torofpostsandintelligenceintheprovinceofJibalanddirectorgeneral sourcesmakeupthecontentofearlyIslamicgeographical ofthesamein BaghdadandSamarra. His writings ranged overseveral literature. Works of this nature were often titled Kitiib subjects, butthegeographical work wasthe most quoted. Thisimpor al-masiilik wa-al-mamiilik (Book of routes and prov tancewas recognized by both Arabsand modernscholars.Thetextof Ibn Khurradadhbih's Kitllb a/-masillik wa-al-mamalik was edited by inces), and the earliest to survive is that of Ibn Khurra MichaelJan de Goeje, Kitiib a/-masiilik wa'l-mamiilik (Liber viarum dadhbih,writtenabout 231/846.9The bare bonesofthis et regnorum), Bibliotheca Geographorum Arabicorum, vol. 6 (Leiden: book consist of the post or pilgrimage routes and dis- E.J. Brill,1889; reprinted, 1967). TABLE4.1 Islamic Time Chart A.H. A.D. Event A.H. A.D. Event 0 622 The Hijrah; Mubammad migrates 290 903 Ibn al-FaqIh and Ibn Rustah, to Medina geographical writers, flourish 8 630 Mubammad captures Mecca ca. 316 ca. 928 Qudamah, geographical writer, 11 632 Mubammad dies flourishes 11-40 632-61 The four orthodox caliphs 317 929 al-BattanI, astronomer, dies 13 634 Palestine and Iraq taken by the ca. 320 ca. 930 al-]ayhanI, Samanid wazir, flourishes Muslims 322 934 al-BalkhI, scholar and geographer, 13-21 634-44 Umar, caliph; Yezdigird III, shah of dies Persia 334 945 al-HamdanI (Sifatjazirat al-CArab) 14 635 Battle of Qadislya; defeat of dies Sassanid army 331-62 943-73 Ibn J:lawqal's travels 22 642-43 Muslims reach Carthage in Africa; 340 950 Suhrab (Ibn Sarabiyun), Makran and Baluchistan in Asia geographical writer, flourishes 31 652 Muslims take Armenia; enter 340 951 al-I~!akhrI's Kitab al-masalik Khurasan wa-al-mamalik 41-132 661-750 Umayyad dynasty of caliphs based 341-65 952-75 al-Mucizz, Fatimid caliph in Cairo in Damascus 345 956 al-MasclidI, historian, dies 73 692 al-J:Iajjah becomes governor of Iraq ca. 350 ca. 961 al-Khazin, astronomer, dies 76 695 Muslim postal service inaugurated 356 967 Sayfal-Dawlah, Hamdanid sultan 86 704 Qutaybah, governor of Khurasan of Syria, dies (d.714) 365-86 975-96 al-cAzlZ, Fatimid caliph in Cairo 90-93 708-12 Bukhara, Samarkand, and 372 982 l-Judud al-calam; al-MuhallabI, Khwarazm captured; also Sind geographical writer, flourishes 92 711 Muslims enter Spain 386-411 996-1021 al-J:IakIm, Fatimid caliph in Cairo 95 714 al-J:Iajjaj, governor of the eastern 377 987 Ibn al-NadIm's al-Fihrist empire, dies 378 988 Last recension of Ibn J:Iawqal's 132-36 750-54 al-Saffal), first Abbasid caliph Kitab ~urat al-ar4 132 750 al-Fazarl, astronomer, flourishes 386 990 al-QummI, astronomer, dies 136-58 754-75 al-Man~lir, second Abbasid caliph ca. 390 ca. 1000 al-MuqaddasI dies 170-93 786-809 Harlin aI-Rashid, fifth Abbasid 399 1009 Ibn Yiinus, geographerand caliph astronomer, dies 198-218 813-33 al-Ma'mun, seventh Abbasid caliph 428 1037 Strasbourg manuscript of 231 846 Ibn Khurradadhbih's Kitab al-KhwarazmI al-masalik wa-al-mamalik after 442 after 1050 al-BlrunI dies 232-47 847-61 al-Mutawakkil, tenth Abbasid 479 1086 Earliest Ibn J:Iawqal manuscript caliph 480 1087 al-Zarqello, astronomer, flourishes ca. 240 ca. 850 J:Iabash al-J:Iasib, astronomer, (Toledo tables) flourishes 530 1140 al-ZuhrI flourishes 247 861 al-FarghanI, astronomer, flourishes 548 1154 Roger II of Sicily dies 256-79 870-92 al-MuCtamid, fifteenth Abbasid 560 1165 al-IdrIsI dies caliph 569 11.73 Earliest manuscript of al-I~!akhrI 260 874 aI-Kindt, philosopher, dies 590 1190 Abmad al-Tust flourishes 274 887 al-MarwazI dies 626 1229 Yaqiit dies 284 897 al-YaCqubi, geographical writer, dies 685 1286 Ibn SacId dies 286 899 al-Sarakhst dies 732 1331 Abu al-Fida) dies 288 901 Thabit ibn Qurrah dies in detail. He then supplements this material from other This type of work called al-Masalik wa-al-mamalik sources,usuallywithliterarymateriallikethatmentioned becameatradition. Otherauthorswhowrote inasimilar above. That he gives distances between places provides vein and with the same title were al-MarwazI (d. 274/ us with at least one condition that would have encour 887), al-SarakhsI (d. 286/899), and al-JayhanI (tenth cen agedthedevelopmentofsomesortofcartographicresult. tury A.D.). Abu (Abdallah MUQammad ibn Abrnad al- The Beginnings ofa Cartographic Tradition 93 ]ayhanI was a Samanid wazir whose work most writers however, except for general topographical descriptions regarded as a serious contribution to geographical liter and the concept of the division of the inhabited world ature, and he was a forerunner, according to al-Muqad into kishvars.16 dasI, of the later al-BalkhI school. None of these works, ThePersiankishvarsor"regions"wereseven,thesame however, has survived.10 number as the Greek climata, so the Arabs called both Ibn Khurradadhbih is also followed by a series of iqlim(pI. aqalim), causingacertainamountofconfusion. authors who write in a more stylistic way by dropping They consist geographically of six regions encircling a the official framework that held his book together. Such central region that represents the central Iranian area, writers were al-YaCqiibI (d. 284/897), Ibn Rustah (fl. ca. usually called in Islamic times by the name Babil (fig. 290/903), and Qudamah (early tenth century A.D.), who 4.2).17 There is a possibility that the idea was ultimately produced geographical works that have survived and are quite important yet are valueless from a cartographic point of view.11 There are others whose works are lost, 10.AI-MarwazI'sworkwasmentionedonlybytitlebyotherauthors (e.g., Mul)ammad ibn Isl)a:q Ibn al-NadIm, al-Fihrist; see The Fihrist although judging from their titles they must have been ofal-Nadim: ATenth-Century Survey ofMuslim Culture, 2vols.,ed. similarin content. Probablythe culmination ofthis early andtrans.BayardDodge[NewYork:ColumbiaUniversityPress,1970], geographical tradition was the work of al-MascudI (d. 1:329,andShiha:bai-DInAbu'AbdallahYaqutibncAbdallahal-HamawI 345/956).Hismainworkswerehistories,buthe believed al-RomI aI-Baghdadi; see The Irshad al-arib ila macrifat al-adib; or, inprefacinghistorywithadescriptionoftheworldwhere Dictionary ofLearned Men ofYaqut, 7 vols., ed. D. S. Margoliouth the events occurred.12 He was a traveler, an inveterate [Leiden:E.J. Brill,1907-27],2:400). Itwassaidin thelattersourceto betheearliestworkofthisnature.AI-SarakhsI'sworkisjustasobscure, collector of information, and a good critic. Since much thoughal-SarakhsIhimselfisbetterknown. Hewasagenerallitterateur of his information was gained firsthand, he gives us an with otherlost works to his credit. excellent sketch of the physical world as well as good 11.Theworks ofthesewritersarepublishedasvolumes6and 7of criticism of Arab geographical literature. BibliothecaGeographorumArabicorum.ForIbnKhurrada:dbihseenote 9above,andforal-YaCqubIseenote4above. ForAboCAlIAl)madibn 'Umar ibn Rustah's Kitab al-aClaq al-nafisah, see de Goeje's edition, Kitab al-aclak an-naffsa VII, BibliothecaGeographorum Arabicorum, FOREIGN GEOGRAPHICAL INFLUENCE vol. 7 (Leiden: E. J. Brill, 1892; reprinted 1967). Qudamah ibn Ja'far ai-BaghdadI was anothergeographical writerwho was also an admin With the formation of the Abbasid court in Iraq, espe istrativeofficial in Baghdad; his Kitab al-kharaj isin de Goeje'svol. 6 cially under the caliphate of al-Man~ur(r. 136-58/754 (note 9above): Kitab al-Kharaj. 75), literature and science were encouraged and it was 12.AbuaI-HasanCAlIibnal-Husaynal-MascOdI'slongestworkisthe realized thattheconquerednations,Sassanidsand Byzan Murujal-dhahabwa-macadinal-jawhar,publishedasLesprairiesd'or, 9vols., trans. C. Barbierde Meynard and Pavet de Courteille,Societe tines,hadmuchtooffer.Itwassoondiscovered,probably Asiatique,Collection d'Ouvrages Orientaux (Paris: Imprimerie Imperi through Pahlevi scholars and texts, that to the east of ale,1861-1917),esp.vol.1(1861).Alsoimportantishisal-Tanbihwa Islamthe Hindu cultureshadawealth ofknowledge that al-ishraf;see Kitab at-Tanbfh wa)l-ischraf,ed. MichaelJan de Goeje, could be tapped by the new elite of the Islamic world. BibliothecaGeographorumArabicorum,vol.8(Leiden:E.J.Brill,1894; Attempts were made to understand Indian astronomical reprinted, 1967),and the French translation by B. Carra de Vaux, Le livre de I'avertissementetde la revision (Paris: Imprimerie Nationale, texts,Indianscholarswereinvitedto Baghdad,and trans 1896). lations were made into Arabic from the siddhanta lit 13.Therewerevarioussiddhantas,andwhichonesfound theirway erature, a word that became sindhind in Arabic. Several into Arabic is in some dispute. But sincetheywere probably received Arabic works are based on siddhanta texts, but it was throughtheintermediaryofPahlevitexts,theexactprovenanceofthe mainly astronomical concepts that were taken from Arabicworkswasneverclear.ExamplesofArabictextsbasedonIndian them.13 A few geographical concepts were derived from materialsarethe Kitabal-ZijofMubammadibnIbrahImal-FazarIand 2i;al-Sindhindal-$aghirbyAbuJacfarMubammadibnMusa:al-Khwar India, the most important being the concept of the azmi.Theformeriswellknownfromquotedfragments(DavidPingree, Cupolaofthe Earthand theuse ofthemeridianofUjjain "The Fragments of the Works ofal-FazarI,"JournalofNear Eastern (Arln)as theprime meridian,ideasthatcreptfrom Arabic Studies 29 [1970]: 103-23),while the latterexistsas a wholeand was into medieval European literature.14Aprime meridian in translated into Latin by Adelard of Bath in medieval times. See also the extreme east based on a locality called Jamagird or p.97 n.3l. 14.ArlnisamisreadingofUjjainintheArabicscript.Formoredetail, Kangdiz, with longitudes ranging west from it, appears see below,p. 103andp. 175 n.6. in some earlyArab tables oflongitude. The Arabs them 15.See below,p. 103. selves thought this system was of Indian or Chinese ori 16.Nothingis known to bespecificallygeographical, but the astro gin.IS nomicaltablesZi;-iShahriyar,translatedintoArabicastheZijai-Shah, As the Arabs were absorbing Indian scientific infor which had agreat influence on Arab astronomers, may have had geo graphical tables similar to Ptolemy's Handy Tables or those of al mation, they also took in many Persian ideas, as well as BattanIdiscussed below. Greek ideas that had reached Persia. There is little 17.Aclearaccountofthekishvarsystemwithdiagramscanbefound obviousPersian influence inArabgeographicalliterature, in N. Levtzion andJ. F. P. Hopkins, eds. and trans., Corpus ofEarly 94 Islamic Cartography South SECONDCLIMATE Hijaz,Ethiopia, 0' Yemen,Arabia (3.;)\'1>'(\ ~/fS)~eq0/ (()1>~(,. ro °..olq »--~--« FIRSTCLIMATE THIRDCLIMATE India, ~s: EgyptasfarasMorocco, Sind,al-Zabaj (ql") Sudan Dq/6~?/ ro C:3~(,.,1> ~-« CLIMATE}>-~ i o FOURTH East -.:g3~~ 0' FaBrasb,yKlohnu,raIrsaaqn, '""S~ro West ,?-,~e\~~ tVtSIJ.. (()?> UIS L1MAT~.~~ SEVENT.HC.. ~ FIFTBHyzaCnLti.uImM,ATE China,Tibet, Qj Andalusiaand Transoxlana.. lc3f) ~ 1>(\00\ theFranks... Yc3/~ ~?>\\}\ 0/ \,.: SIXTHCLIMATE YajOjandMalOj. Khazars,Turks. Russia.Slavs.. Jl...:..l\ North FIG.4.2. THEPERSIANKISHVARSYSTEM.Thisrepresenta Size of the original: 14.5 X 13 em. From Ahmed Zeki Velidi tion is taken from al-Biriini's Kitab taf;did nihayatal-amakin. Togan, ed., Birunt's Picture of the World, Memoirs of the My translation of al-BlriinI's version of the system is on the Archaeological Survey of India, no. 53 (Delhi, 1941), 61, by right. permission of the Archaeological Survey of India, New Delhi. derived from Babylonian sources, although there are 38.Lotuspetalmapsarediscussedandillustrated below,seefigs. 15.3, resemblances to the Indian cosmographic concept of 16.1, 16.8, 16.14,16.18, 17.20, 17.21,and related text. Mount Meru and the lotus petals.ls 19.CarloAlfonsoNallino,"Al-tIuwarizmieilsuorifacimentodella GeografiadiTolomeo," Attidella R. Accademiade; Lincei: Classedi Greek geography, by contrast, influenced the early Scienze Morali, Storiche eFilologiche, 5thser., 2(1894), pt. 1(Mem Arabs considerably. Greek ideas again came via Pahlevi orie), 3-53; republished in Raccolta di scritti editi e inediti, 6 vols., texts, but scholars acquainted with Syriac and Greek ed. Maria Nallino (Rome: Istitutoperl'Oriente, 1939-48),5:458-532, introduced directly from the former Byzantine provinces esp. 459-63; Ernst Honigmann, Die sieben Klimata und die 7Cbleu; a greater knowledge of Greek geographical material, t7Clatlj.lOI (Heidelberg: Winter, 1929), 112-22; J. H. Kramers, "Qiu@.: including the works of Claudius Ptolemy.19 Ptolemy's rafiya," in The Encyclopaedia of Islam, 1st ed., 4 vols. and suppl. (Leiden: E.J. Brill, 1913-38),suppl., 61-73, esp. 63. Geography was more purely geographical than anything 20. Sezgin, in a recent work I saw after this chapter was written, the Arabs had yet encountered. The lists of latitude and discusses"Ma'municmathematicalgeography"and comparesitfavor longitude values and the mathematical side of Ptolemy ablywith"Ptolemaicgeography";FuatSezgin,TheContributionofthe appealed to them, and they set out with zeal to check Arabic-Islamic Geographers to the Formation of the World Map his results and correct them where they could.20 For (Frankfurt: Institut fur Geschichte der Arabisch-Islamischen Wissen schaften, 1987). What he does is to compare the values given in the instance,oneoftheprojectsofthistimewasremeasuring tables of latitude and longitude of the early Arab geographers and the length of a degree of latitude on the earth's surface astronomers with the values given by Ptolemy, stating that the Arab using the method originally attributed to Eratosthenes.21 valuesareonthewholebetterthanthoseofPtolemy.Infactthelatitude One thing not taken up by Arab scholars was Ptolemy's values are more accurate to some extent, but by no means in every chapter on the construction ofgeographical map projec case. Longitudes are justso muchguesswork. In no case do we know howtheArabsarrivedattheirvaluesexceptbyplayingwiththefigures tions. In view of their interest in projections for celestial of their predecessors. Nor do we know how Ptolemy arrived at his mapping (chap. 2 above) this is surprising, and perhaps it figures. Except for knowledge of the length of a degree of latitude, there can have been little mathematics involved. What it would have requiredwasavastorganizationofmeasurers,andwe have nosignof ArabicSources forWestAfricanHistory(Cambridge:CambridgeUni that. EvenifMa'miinasCommanderoftheFaithful managedtoorga versity Press, 1981),xv-xix. nize this in the Abbasid Empire, a large number ofplaces with coor 18. This Indian concept is probably also derived from the Persian dinates were found outside his domains. kishvarsystem that appears in the Avesta; see part 2 belowon South 21. Carlo Alfonso Nallino has dealt in detail with this in "11 valore Asiancartography.TheconceptisnotoldinIndia;GeorgeRusbyKaye, metrico delgrado di meridiano secondo igeografiarabi," Cosmos 11 Hindu Astronomy: Ancient Science ofthe Hindus, Memoirs of the (1892-93):20-27,50-63,105-21;republishedinRaccoltadiscrittiediti ArchaeologicalSurveyofIndia,no. 18(Calcutta:GovernmentofIndia einediti,6vols.,ed.MariaNallino(Rome:Istitutoperl'Oriente,1939 CentralPublicationsBranch,1924;reprintedNewDelhi:Cosmo,1981), 48),5:408-57;and see below,pp. 178-81. The Beginnings ofa Cartographic Tradition 95 affected all future Arab cartography. The link between clear that a map is understood by the word jaCriifiyah. Ptolemy'smathematicsandactual mapproductionseems Unfortunately al-ZuhrI has not left a map, but only a never to have been made. The impetus Ptolemy's work text that describes the world according to the Persian gave to the Arabs, however, does seem to have aroused kishvar system. interest in map production, as is illustrated bythe impor The actual form ofal-Ma)mun's map is an enigma. AI tance latergeographers gave to the map ofthe caliph al MasciidI'S reference to Marinus and Ptolemy in the same Ma)miin. breathwithal-Ma)munmakesoneimmediatelythinkthat themapmadeforal-Ma)munwasbuiltupfromlongitude and latitude tables on a projection similar to that used THE MAP OF THE CALIPH AL-MA)MON by the Greeks. Both al-MascudI and al-ZuhrI, however (and we must not forget that the latter claims to have Under the caliph al-Ma)miin (r. 198-218/813-33), sci basedhisworkonal-Ma)mun'smap),basetheirsurviving ence flourished at the court, and the caliph surrounded geographical texts on the Persian system of kishvars, himself with scholars. One of the achievements of the which they call aqiilim orclimates,and this confuses the group of scholars working at his instigation was a large issue. It is not known ifthe scholars ofal-Ma)mun really map of the world. AI-Ma)miin's map, however, is likely used mathematical geography to produce their map. If to have been prompted as much by a political motive as theydid,wemightenvisagesomethingresemblingaPtol by a purely scholarly one, since various early rulers (Sas emaic world map or a world map on a rectangular grid sanid, Fatimid, and Sicilian Norman) were said to have such as Marinus was supposed to have used. This form had similar maps constructed to show that they ruled everything in the world that mattered.22 is supported in that al-KhwarazmI, one of the scholars of al-Ma)mun's court, produced tables of longitude and This map has not survived. The only knowledge we latitude of which a manuscript has survived to this day, have of it comes from a number of contradictory ref albeit from the early eleventh century (discussed below). erences in the works of later authors. The earliest and Suhrab, a later scholar who produced another set of probablymostdetailed referenceto itandalsosuggesting tables that has also survived, gives directions for pro acontemporaryinterestincartographyisthewell-known ducinga map on such a rectangulargrid. His tables were passage from al-MascudI that runs, almostcertainlyderivedfrom thoseofal-KhwarazmIand Ihaveseentheseclimatesrepresented[mu~awwarah] are nearly as long. in various colors, without a text, and the best that I If al-ZuhrI used al-Ma)mun's map and that map was have seen has been in the book fughrafiya [Geo based ona Greek model,whydid al-ZuhrI base hiswork graphy]ofMarinusandthecommentarytofughrafiya on the Persian kishvarsystem? Furthermore, al-MascudI, of the divisions of the earth and in al-surah al after claiming that al-Ma)mun's map was so good, also ma)muniyah that al-Ma)miin ordered to· be con based his descriptive geography on this Persian system. structedbyagroupofcontemporaryscholarstorepre It is difficult to read much into the latter's writing, how sent the world with its spheres, stars, land, and seas, ever,for his knowledge ofPtolemyandGreekgeography the inhabitedand uninhabited regions, settlements of peoples,cities, etc.Thiswas betterthananythingthat was very superficial, and he was probably aware of them preceded it, either the Geography of Ptolemy, the onlyat second hand from Arabic material like al-Khwar Geography of Marinus, or any other.23 azmI and al-FarghanI.25 AI-MascudI's statement that all towns in the same Ptolemaic climate have the same lat This reference is vague, but the representation "in var itude is a misinterpretation ofal-FarghanI's climatic lists, ious colors" must mean that he had seen maps reputed and al-MascudI's use of the Greek word "climate" (as to be made by or derived from the Greeks and also that this special map created by al-Ma)mun's scholars was 22. Ardashir Iaccording to several late Arab authors; see Kramers, known to him. The word used for map, ~urah, is not a "lliughrafiya," 64 (note 19). Ibn Yunus and al-Muhallabl prepared a technical term. It means "representation" or "picture" map on silk for the Fatimid caliph al-cAziz (r. 365-86/975-96), see and can refer equally well to a written illustration or below.The Normanwas RogerII, forwhom IdrlsIproduced his map representation. Nevertheless,itis the termusedfor those (chapter7 below). 23. AI-MascudI, al-Tanbih; see the edition by de Goeje, 33, or the maps of Daylam and Bukhara that were mentioned translation by Carra de Vaux, Livrede l'avertissement, 53 (note 12). above, and it came to be regarded as the word for map 24.SeetheeditionofMul)ammadibnAbIBakral-ZuhrI'sjaCrafiyah throughout the geographical literature of the classical in Mal)ammad Hadj~Sadok,"Kitab al-DjaCrafiyya," Bulletin d'Etudes period. Orientales21(1968):7-312,esp.306.AI-FazarIwasanearlyastronomer AI-Ma)mun'smapwaspresumablyseen bythetwelfth (d. ca. 240/800) mainly known as a translatorofIndian astronomical texts. century geographer al-ZuhrI, who states that his own 25.AI-KhwarazmI,Suhrab(IbnSarabiyun),andAbual-cAbbasA\,lmad work (Jacriifiyah) is a copy of a copy (made by al-FazarI) ibn Mul)ammad al-FarghanI will all be discussed in detail in the fol ofthejaCriifiyah al-Ma)mun.24 The description makes it lowingsections. 96 Islamic Cartography iqlim) for the Persian kishvar also shows his misunder he had seen and copied some more detailed list. This standing.26 arrangement also shows possible signs of an eastern Of actual maps (apart from that of al-Ma)mun) that meridian. According to Honigmann, al-FarghanI's astro come from the early period or might have originated nomical tables beara closeresemblance to those ofPtol before the time of al-BalkhI's school, we also have the emy's Prokheira kanones (Handy Tables), which was map said to be made by Abu al-Ijasan cAli ibn cAbd al translated into Arabic by CAyyiib and Simcan as the Zij Ra1)man ibn Yunus for the Fatimid caliph jointly with Ba#amiyus.30This is not so; there seems to be no Greek al-IjasanibnA1)madal-MuhallabLSinceIbnYunuscom originforthearrangementoftownsandclimatesasgiven piled tables of coordinates similar to those ofal-Khwar byal-Farghani,and Ptolemy'stablesarearranged bygeo azml, one might expect his map to differ considerably graphical regions and are similar to the fuller tables in from those ofthe Balkhlschool. Reinaud states that Ibn his Geography. The only Greek connection of al-Far Yiinus compiled a map for a Buyid ruler that was in ghanl is the Greek toponymy he used in areas beyond various colors but had no graticule. The source of this the bounds of Islam. Nor does there seem to be any statement is not given.27 precedentinEasternworksforthisarrangement.Itseems There is also a curious reference in the Fihrist, the to be a simpler form of the arrangement used by al general bibliographiccompilationofIbnal-Nadlm (d. ca. Khwarazml, who was of course al-FarghanI's contem 385/995). Qurrah ibn Qam1ta, a Sabean from Harran porary. However, this form of table never has actual who must have lived in the ninth century A.D., made a coordinates,and in each climate the orderis eastto west description oftheworld(Sifat al-dunya) thatwas copied and not vice versa as in al-Khwar~zmL Thus there was byhis morefamous compatriotAbual-IjasanThabit ibn never sufficient information in this sort of list for the Qurrah al-l:IarranL The author of the Fihrist saw this ~ifat and noted that it was made on DubayqI cloth, "unbleached butwith dyes."28 It ispossible that thispas 26. AI-Mas\idI, al-Tanbfh; see the edition by de Goeje, 25, 32, 44 sage actually refers to a map. (note12);andal-Mas\idi,Murujal-dhahab;seethetranslation byBar bier de Meynard and de Courteille, Les prairies d'or, 1:182-83, 185, 205 (note 12). GEOGRAPHICAL TABLES 27. See Reinaud's edition of Taqwfm al-buldan by Ismacil ibn cAli r Abual-Fida):Geographied'Aboulfeda:Traduitede arabeen franrsais, The Islamic tables of longitude and latitude seem to be 2 vols. in 3 pts. (vol. 1, Introduction generafe d fa geographie des of much more importance in this early period than the Orientaux byJosephToussaint Reinaud; vol. 2, pt. 1,trans. Reinaud, maps themselves. A considerable number of lists ofgeo vol. 2, pt. 2, trans. S. Stanislas Guyard) (Paris: Imprimerie Nationale, graphicalplace-namesexistinArabic literature. Someare 1848-83), I:CCLXIII. See also S. Maqbul Ahmad, "Kharita," in Ency clopaediaofIslam, newed.,4:1077-83,esp.1079,andIbrahimShaw meant to be all-inclusive lists covering the whole of the kat, "Khara)it djughrafiyyi al-cArab al-awwal," Majallat af-Ustadh knownworld,somecoverspecificpartsoftheworld,and (Baghdad)2(1962): 37-68. some give only an unspecified selection of names. Some 28. See the editions of Ibn al-Nadim's al-Fihrist by Dodge, 2:672 listsarefound directlyinoriginalworks,butsomeappear (note 10),and Gustav Fliigel, ed., Kitiib al-Fihrist, 2 vols. (Leipzig: F. only as extracts in the works of other writers. They are C.W. Vogel, 1871-72),1:285.Thisisalso cited byIgnatiyIulianovich Krachkovskiy,Izbrannyesochineniya,vol. 4,Arabskayageografiches generally of two types: those that list places under cli kayaliteratura(Moscow,1957),translatedintoArabic bySalal)aI-Din mates, not always in any order,giving no longitudes and CUthman Hashim,Ta)rikh al-adabal-jughrafial-cArabi, 2vols.(Cairo, only the latitudes of the climate boundaries, and those 1963-65),1:206. that give longitude and latitude values individually for 29.AI-FarghanI'sastronomicalworkbearsseveraldifferenttitles;see each place. Elementa astronomica, arabice et latine, ed. and trans. Jacob Golius (Amsterdam, 1669); the important chapters are 8 and 9, pp. 30-39. The first type are usually included in the works of Honigmann,DiesiebenKlimata, 138(note19).Seealsothearticlesby astronomers, where they appear as a separate chapter. HeinrichSuter(rev.Juan VernetGines),"al-Farghani,"in Encyclopae The bestexampleofthiscategory,and oneoftheearliest diaofIslam,newed.,2:793;byA.I.Sabra,"al-FarghanI,"inDictionary extant,appears in the work ofAbual-cAbbas A1)mad ibn ofScientific Biography, 16vols., ed. Charles Coulston Gillispie (New Mu1)ammad al-Farghanl (fl. 247/861).29 AI-Farghanl's York:CharlesScribner'sSons,1970-80),4:541-45;andbyGeorgeSar ton in his Introduction to the History ofScience, 3 vols. (Baltimore: text gives only the climate boundaries in degrees (mid Williamsand Wilkins, 1927-48), 1:567. climate and maximum latitude). He also gives the length 30.Honigmann,DiesiebenKlimata,116-17,137(note19).Theonly of the longest day in hours and fractions and the length availableedition ofthiswork ofPtolemyis thateditedandtranslated of the gnomon shadow. Having delineated the climate byNicholasB.Halma,8elt)v~'AAeeav~pe~'Ynop,v"p,aet,TOD<;lIToltp,aIOV thus, he lists geographical features inside each climate, npoxelpo~Kavbvar;:CommentairedeTheond'Alexandrie,surlesTables manuellesastronomiquesdePtolemee,3vols.(Paris:Merlin,1822-25), thus giving only a range of latitude where the feature is 1:109-31.Thisshowsthatthetablesoflatitudeandlongitudearevery situated. No longitudes at all are given, but the features muchthesameasthoseinPtolemy'sGeography,althoughconsiderably are listed approximately in order from east to west as if reduced in number ofentries. The Beginnings ofa Cartographic Tradition 97 arranged under several classified heads-for example, towns, mountains, seas, islands, springs, and rivers (fig. 4.3).31 Under each of these heads names are listed in climates, beginning south of the equator and working north. Under each climate entries are given in order of longitude (west to east), so that each place as it occurs is given first longitude and then latitude in degrees and minutes. Thetablesofal-Battanl(who isconsiderablylaterthan al-Khwarazmi) are not so systematically arranged (fig. 4.4).32 Al-Battani's main work was a textbook on astro nomy, and in his work the tables of coordinates for important places were included with his tables for coor dinates of important stars among the general astronom ical tables. First he lists the "centerpoints" ofeach geo graphicalregiontothenumberofninety-four,takenfrom book 8, chapter 29 of Ptolemy's Geography with their respective longitudes and latitudes, and he then lists towns and other miscellaneous features in the same way (180 all together). There is no logical order of places, though the lists drift from west to east so that places in the same region appear together. However, several of these drifts occur successively throughout the total list. A final sequence lists towns and so forth in Spain and North Africa. Other astronomers followed al-Battani's work instead of a geographical listing after the manner of al-Farghani.33 Also, when later geographers selected 31. AI-Khwarazmi (d. ca. 232/847) was well known ro both Arabs and medieval Europeans. His fame was established by his book on algebra, buthisastronomical work,2ijal-Sindhindal-~aghir,wasalso well known. Hisgeographical work was known onlyfrom quotations until a manuscript was discovered in the late ninteenth century. The fullest account of al-Khwarazmi's contribution ro Arab geography FIG.4.3. FOLIO FROMTHESTRASBOURG MANUSCRIPT appears in Nallino'sarticle"AI-l:Iuwiirizmleilsuo rifacimento"(note OFAL-KHWARAZMI'SSDRATAL-'ARl).Thisfoliocontains 19).AlsoseeG.J.Toomer,"al-Khwarizmi,"inDictionaryofScientific the beginning of the section on the mountains, showing the Biography,16vols.,ed.CharlesCoulstonGillispie(NewYork:Charles Scribner's Sons, 1970-80), 7:358-65. The unique manuscript of al mountainsthatliesouthoftheequatorandgivingthelongitudes Khwarazmi's geographical work, Kitilb ~urat ai-art! (Picture of the and latitudes of both ends of the range, their colors, and in which direction ofthe compass their peaks should be. earth),wasdiscovered inCairo by W.Spittaand deposited in the Bib liotheque Nationale et Universitaire, Strasbourg, Cod. 4247. It was Size ofthe original: 33.5 X 20.5 em. By permission ofthe Bib edited by Hans von Mzik, Das Kitilb ~urat ai-art! des Abu Ga'(ar liorhequeNarionaleetUniversitaire,Strasbourg(Cod.4247,fol. MuiJammad ibn MUSil al-ljuwilrizmi, Bibliothek Arabischer Histori lOb). kerund Geographen,vol. 3(Leipzig:Otto Harrassowitz,1926).There are 2,402 entries in this work. Edward S. Kennedy and Mary Helen Kennedy,GeographicalCoordinatesofLocalitiesfrom IslamicSources scientific construction of a map, although information (Frankfurt: Institut fiir Geschichte der Arabisch-Islamischen Wissen derived from this type has filtered through into latergeo schaften, 1987), presents the coordinate lists from some seventy-four graphical works and then into later maps. sources includingal-Khwarazmi. Lists of the second type, which give longitudes and 32.The work ofAbo 'Abdallah MUQammad ibnJabiral-Battanial Sabi', entitled 2ij al-$ilbi', was edited by Carlo Alfonso Nallino, AI latitudes for individual places,areexemplified bythe sets BattilnisiveAlbatenii:Opusastronomicum,3vols.(Milan,1899-1907; oftablesgiven byal-Khwarazmiand byal-Battanias well vols. 1and 2reprinted Frankfurt: Minerva, 1969).The tables occurin as by several later authors. 2:33-54.TheyarealsogivenbyJoachimLelewel,GeographieduMoyen Age,4vols.andepilogue(Brussels:J.Pilliet,1852-57;reprintedAmster dam:Meridian,1966),epilogue,64-93.Therearearotalof273entries. LONGITUDE AND LATITUDE TABLES: Al-Battanlwasan importantastronomer, knownas AlbategniorAlba AL-KHWARAZMI, AL-BATTANI, AND PTOLEMY tegnius to medieval Europeans. He died in 317/929. 33.IbnYiinus(d.399/1009)isthebestexample,withapproximately Al-Khwarazml's tables consist of simple lists of names 290 entries (these are mentioned by Lelewel, Geographie du Moyen 98 Islamic Cartography these two sets of tables frequently vary (fig. 4.5). Al-Bat tanI is very close to Ptolemy, especially in those values he gives for cities, where there is almost always agree ment. The Kitiib al-mallJamah quoted by Yaqiit is also reasonably close to Ptolemy, although about a third of his localitiesare not in Ptolemyatall. Al-KhwarazmI lists many places that are not in Ptolemy, his latitudes vary considerably, and of course his longitudes are mainly about ten degrees less than those of Ptolemy because of his different position for the prime meridian. The dif ferences between these Arab texts suggest that they are derived from different sources, and some explanation ought to be possible. It is therefore necessary to inves tigate the matter in more detail. It is not known exactly when and where Ptolemy's Geography (rEo:rypu<plK1'l iJ<pl'rYTJOl~) reached Arab schol ars. The date usually given for its arrival is during the reign of the caliph al-Ma'miin, and the translation is attributed originally to Abii Yiisuf Ya'qiib ibn Isi)aq al KindI and later to Thabit ibn Qurrah.36 Since al-KindI didnotdie until260/874,hemust have beenquiteyoung if he was translating Ptolemy in the reign ofal-Ma'miin. Thabit ibn Qurrah's dates are 222-88/836-901, so his translation would not have appeared until much later in the century. Both these authors were younger than al KhwarazmI,whowascertainlyworkingunderal-Ma'miin asanastronomerandastrologerandprobablydied about FIG.4.4. FOLIOFROMAMANUSCRIPTOFAL-BATTANI'S Age, 1:165-77[note 32]),also al-Biriini in Kitilb al-qilnun al-Mas'udr Zl] AL-SABI'. This folio shows longitudes and latitudes for fr al-hay'ah wa-al-nu;um (see al-Qanunu'I-Mas'udr(Canon Masudi part of Arabia (left) and part ofSyria (right). cus),3vols.[Hyderabad:OsmaniaOrientalPublicationsBureau,1954 Sizeofthefolio:28 x 19.3em. BypermissionofthePatrimonio 56]),andotherlaterastronomers.Also,al-BiriinI'sreferencesarequoted Naeional, Madrid (cat. no. 908 Derembourg, fol. 175r). bylatergeographerssuchasAbual-Fida'inhisTaqwrm al-buldan;see Geographie d'Aboulfeda, textearabe, ed. and trans.Joseph Toussaint the coordinatevaluesofpreviouswriters, they frequently Reinaud and William MacGuckin de Slane (Paris: Imprimerie Royale, used al-BattanI's values in preference to those of others. 1840), 11,74. Kennedygives a bibliographyofsurvivingastronomers' Another similar list of place-names probably earlier tables, often mentioning where geographical coordinates are given: than al-BattanI's may have been that of the Kitiib al EdwardS.Kennedy,"ASurveyofIslamicAstronomicalTables,"Trans actionsoftheAmericanPhilosophicalSociety,n.s.,46(1956): 123-77. mallJamah (Book of the battle), which was one of the Inoneortwocaseshestateswhichprimemeridianisused. In Edward sources for longitude and latitude ofplaces that the later S. Kennedy and M. H. Regier, "Prime Meridians in Medieval Islamic author Yaqiit (d. 626/1229) used in hisgeographical dic Astronomy," Vistas in Astronomy28 (1985): 29-32,a list isgiven of tionary Mu'jam al-buldiin (Dictionary of countries). astronomers,notingwhethertheyusetheFonunateIslesortheAfrican Because ofthe format ofhiswork, theform ofthe Kitiib coastas a prime meridian (p. 31). 34.Alistoftheseplacesisgivenin Honigmann,Diesieben Klimata, al-mallJamah was completely broken up, and weare left 126-31 (note 19). with a series of sixty-four separate places.34 35. Ptolemy is mentioned in the title of the Strasbourg manuscript Tables of this sort often give variations of the same ofal-Khwarazmi(seevon Miik'sedition, Kitilb~uratai-arc![note31], place-names or even assign different names to the same andNallino,"AI-J:luwiirizmieilsuorifacimento,"477[note19]).Ptol placeaswellas differentcoordinatevalues. Laterauthors emy is mentioned byal-Battani in hissixth chapteron longitudesand latitudes(Nallino'sedition,Opusastronomicum,1:20,textp.28[note tend to pick and choose. They combine or duplicate 32]). Al-Battani also states that his ninety-four regional divisions are material inacompletelyarbitrary manner,so thatas time from Ptolemy: they come, as Ihave already mentioned, from bk. 8, goes on the tables become impossible to use in any sci- chap. 29 ofthe Geography. entific way. . 36. See appendix 1.1,pp. 10-11.The translation ofPtolemy'sgeo The tables of both al-KhwarazmI and al-BattanI graphical work into Arabic is mentioned in the Fihrist;see the Dodge edition, 2:640 (note 10). Translations are also discussed in Kramers, strongly resemble those of Ptolemy, and the name Ptol "!liughrafIya," 63 (note 19); Honigmann, Die sieben Klimata, 112-18 emy (Barlamiyiis in Arabic) is associated with both.35 (note 19);and Nallino, "Al-J:luwiirizmie il suo rifacimento," 459-63 However, the actual values of longitude and latitude in (note 19). The Beginnings ofa Cartographic Tradition 99 Yaqut Ptolemy al-Khwirazmi al-Battini Long. Lat. Climate Long. Lat. Long. Lat. Long. Lat. Adana("Abava) 68° 15' - - 68° 15' 36°50' - - 68° 151 36°50' GreaterArmenia(Khilat) 78° 38°20' V - - 64°50' 39°50' 78° 39°20'[KhilaO 77° 41°[Gr.Armenia] LesserArmenia(Tiflis/Tbilisi) 75°50' 45° - - - - - - - Antioch ('Avnoxna) 69° 35°30' IV 69° 35°30' 61°35' 34° 10' 69° 35°30' Ankara("AyKupa) 58° 49°40' - 62° 42° 58° 43° - - Ahwaz(Lovaa) 84° 35°04' - 84° 35° 15' 75° 32° 83° 34° Bukhara 87° 41° V - - 87°20' 37°50' [88° 34°?] Barda 79°30' 45° VI - - 73° 43° 84° 42° Cyrenaica(Barqa) (BuPKll) 63° 33° 10' III 49° 15' 30°45' 43° 33°45' - - Baalbek('HAtOU1tOAt~) 68°20' - IV 68°40' 33°40' - - 68°201 33° 15' Baghdad 75° 34° IV - - 70° 33°09' 80° 33°09' Balkh(BuK'tpa) 115° 37° V 116° 41° 88° 35' 38°40' 116° 41° Beirut(BllPu't6~) 68°45' 33°20' - 67°30' 33°40' 59°30' 34° 69°30' 33°20' Palmyra(Tadmur) (nUA/lUpa) 71°30' - IV 71°30' 34° 66° 35° 72° 34° Tikrit(Bip8a) 98°40' 37°30' - 78°45' 36°20' - - - - Gurgan('YpKavia) 86°30' 40° V 98°50' 40° 80°45' 38°50' 95° 40° Haran(Ka.ppat) 72°30' 27°30' IV 73° 15' 36° 10' 65° 36°40' 73° 36°40' Aleppo(Halab) (Bepota) 69°30' 35°25' IV 71°20' 35° 63° 34°30' 71° 34°501 Hulwan 71°45' 34° IV - - 71°45' 34° 81° 35° Homs(''E~.l1aaa) 69° 34°45' IV 69°40' 34° 61° 34° 69°05' - Khiva(Khwarazm) ('n~£iava) 117°30' 45° VI 117°30' 44°20' 91°50' 42° 10' - - Raqqa(NtKll<l>optOV) 73°06' 35°20' IV 73°051 35°20' 66° 36° 73° 15' 36° Rome(RumTya) (p05/l11) 35°20' 41°50' V 36°40' 41°40' 35°20' 41°50' 36°40' 41°40' Edessa(Ruha)(''Eb£aaa) 72°30' 37°30' IV 72°30' 37°30' 64° 36°40' 72°50' 37° Rayy('Payata) 85° 37°36' IV 98°20' 34°20' 75° 35°45' 86° 36°30' Zaura 105° 39° V - - - - - - Syracuse(LupaKouaat) 39° 18' 39° V 39°30' 37° - - - - Salamiya 68°20' 37°05' IV - - 62°45' 33°30' 69°50' 34°50' FIG. 4.5. A COMPARISON OF SOME OF THE COORDI MuCjam al-buldan, and the climate where given. This is fol NATES OF AL-KHWARAZMI, AL-BATTANI, AND THE lowed by the coordinates given by the otherauthors. KITAB AL-MALlfAMAH (YAQUT) WITH THOSE OF After Ernst Honigmann, Die sieben Klimata und die no).,elC; PTOLEMY. The first column lists the coordinates from the e1CIGtlf.L01 (Heidelberg: Winter, 1929), 126-27. Kitiib al-mal~amah, as given in the printed edition of Yaqut's 232/847. Ibn Khurradadhbih (fl. 231/846), another Because of the difficulties of transmission, references scholar and the nearest contemporary of al-Khwarazmi, in Arabic literature to Ptolemy and his work that come claims to have made a translation of Ptolemy's descrip from this earlyperiod beforetherewasa complete trans tion of the earth from barbaric (aCjamiyah) into pure lation are short and often erroneous. The orthography (~abilJah) speech.37 What this means is not clear. That ofplace-namessuffers,andfiguresarecorruptedthrough he made a translation into Arabic is dubious, and in any outArabicliteraturebytheinaccuraterenderingofGreek case he may be referring not to tables but to other parts and Syriac letters into the Arabic script. In all these lan of Ptolemy's work. However, it is clear that Ibn Khur guages, numbers are represented by alphabetic signs, and radadhbih's translation was for private use, and it does Semitic alphabets have many letters that are identical not seem to have been used by laterwriters. It therefore except for diacritical points, which are often omitted.38 appears likely that the full translation of Ptolemy's Geo 37.Seede Goeje'seditionofIbnKhurradadhbih's Kitab al-masalik graphy into Arabic did not come until much later in the wa-al-mamalik, 3(note 9); also Nallino, "AI-tIuwarizmie il suo rifa ninth century, certainly after the death of al-Ma)mun. It cimento,"462 (note 19). is thus probable that al-Khwarazmi and Ibn Khurra 38.Thereare many references to thepeculiarities ofSemiticscripts dadhbih had to take the work straight from the Greek, andthemisinterpretationsduetosimilarletters'beingusedfordifferent numbers.WorksIhavebeenconsultingthathaverelevantpassagesare or more likely from a Syriac version, and they may have Lelewel, Geographie du Moyen Age, epilogue, 62-63 (note 32), and had to translate themselves those parts they wished to von Mzik in the introduction to his edition of the Strasbourg manu use. script, Kitab ~urat al-arq" XVI-XXX(note 31).