The Tabula Peutingeriana
and the Madaba Map

by Ekkehard Weber

The documented history of the Tabula Peutingeriana (TP) begins in the year 1507. At this time - or somewhat earlier - Konrad Celtes (Celtis, originally Bickel or Pickel, 1459 - 1508), famous humanist and poet laureate at the Imperial court of Vienna, found the large scroll of pergamene in a monastery of southern Germany, and took it away, evidently without the knowledge of the owners. He wisely never revealed the location to avoid discussions or demands to hand it back. Such behaviour was not uncommon amongst learned people in those days, since monasteries or abbeys seem to have cared little about their antiquities. In his will, dated January 24, 1508, Celtes dedicated the scroll to his friend Konrad Peutinger, minister of Emperor Maximilian I and chancellor of the city of Augsburg. He obliged him to publish the map and to leave it to a public library afterwards. Both obligations were met only considerably later. Nevertheless the map has been known under Peutinger's name ever since.

The first publication was made in 1598 by Marcus Welser, a relative of Konrad Peutinger. In spite of the fact that our map was drawn in the 12th or early 13th century, it was evident from the first moment onwards that it was the copy of an ancient map, clearly showing the Roman road system, together with the long-distance connections to the East, up to India and China. Originally it was one large scroll, approximatly 34 cm wide and about 7 m long, but now the eleven sheets of pergamene of the scroll are kept separately in the Austrian National Library in Vienna. The beginning of the map, containing the extreme west, with parts of Northern Africa, Spain and Britanny, was already missing when our copy was made. So the western parts of the ancient world are lost to us, as well as the title, a sort of headline and probably a dedication, which might have helped towards dating the ancient original more precisely.

Tabula Peutingeriana, sheet IX. On the left-hand side Russia (above) and the Black Sea, the eastern part of Asia Minor, the narrow Mediterranean Sea with the island of Cyprus, and on the bottom the Levantine coast and the Holy Land. On the right side Russia, Syria and the eastern parts of Turkey with the remarkable representation of Antioch.

The map shows the world with some peculiarities. The roads are indicated in red, and strange looking hooks signify the stops or road-stations, the mansiones, often situated in or near cities or villages, where passengers using the cursus publicus could find accomodation and could change horses. More elaborate indications of localities, like the "double-towers" or similar signs, do not refer to a comparative importance of the locality, but to the quality of accomodation to be found there.1 Obviously one could find better accomodation near or in larger cities, so these signs often correspond to them. Additional drawings appear in connection with the capitals of the ancient Roman world, Rome, Constantinople and Antioch, which are represented by human figures, the personifications or tychai of these cities.

Tabula Peutingeriana, sheet IV, detail: Rome.

Rome (segment IV) is represented by a female figure seated on an elevated throne, clad in a richly draped costume of imperial purple, carrying the globe and a long sceptre in her hands. On the figure's head is what seems to be a misinterpeted mural crown. On the right knee is a helmet, a blue shield rests on the left-hand side. This representation is encircled by a yellow ring from which the famous Italian roads, the viae consulares take their beginning; it is comparable to the modern "raccordo anulare". These roads are the only ones bearing their traditional names, and I am convinced that they were already part of the original scheme of the map. We can trace back the first indications of the existence of such a map to the times of Emperor Augustus.2
To the left, outside the above mentioned ring, at a distance of one mile along a road called Via triumphalis, is a building with the legend "ad sanctum Petrum" - the old basilica of St. Peter's, built under Emperor Constantine in the first decades of the fourth century. This is one of the few traces of Christianity in the TP.

Tabula Peutingeriana, sheet IX, detail: Antioch.

The symbol for Constantinople (segment VIII), near a remarkably large Bosphorus, is just a figure seated on a draped throne with a spear in her left hand, and helmet and shield on the hips, like Rome. She wears a purple cloak and a sort of head-band with strange- looking feathers. Here too the draftsman evidently misunderstood the figure's original mural crown. This figure is rendered in a dynamic fashion: her feet are close together, her knees more to the left side, her head turned to the right, and her right arm stretched out in a sort of an adlocutory gesture. She seems to point to a column standing nearby, which is crowned by a figure with globe and spear, obviously a Roman emperor. It is not certain which of the two large columns of Constantinople is meant; it could very well be the "Çemberlitas", the "burned column", set up by Constantine as a landmark for his new capital. Emperor Theodosius II added iron rings to prevent further damage (418 AD; ChronPasch. 573 B). It seems that they are indicated on the drawing. If this is correct, those rings present a clue for dating the ancient original of the tabula to the time of Emperor Theodosius II., i.e. the first half of the fifth century.3 The name of Constantinopolis is written in red letters on either side of the column; the roads in the immediate neighbourhood of Constantinople are missing, evidently not to upset the composition, but the distances are indicated. To the north, on the other side of the "Golden Horn", we find a sanctuary, like in Rome, the church of St. Irene at Pera - Galata (Sycas on the map). At the entrance of the Bosphorus from the Black Sea, at Chrysopolis - Üsküdar, a very interesting lighthouse is shown.4

Fig. D Tabula Peutingeriana, sheet IX, detail: Constantinopolis.

The largest representation is the one of Antioch on the Orontes, the modern Antakya. The city goddess is again seated on a richly draped throne, in a similar outfit as her western counterparts, with a halo around her head. She carries a spear in her right hand; with her left she caresses a nude boy, reclining on the left side of the throne.5 He pours water from a jug and is evidently the symbol of the river god Orontes (Nahr el-Asi). Beneath it is the drawing of what is either a bridge with many stone pillars or, more likely, a portico.6 On the other side of the goddess is a grove with neatly sketched trees and, again, a sanctuary. This seems to resemble the sacred grove of Daphne (bet el-Ma) with its famous temple of Apollo, but this building is more likely a Christian sanctuary, in analogy to the ones indicated in Rome and in Constantinople. The temple of Apollo of Daphne was burned to ground on October 22, 362 and was never rebuilt.7
These three figures have caused a lot of confusion in the attempts to date the map. A century ago Konrad Miller, a scholar who had dedicated most of his life to the TP and other ancient maps,8 was convinced that these figures meant Roman emperors. In search for three emperors who had their residences in Rome, in Constantinople and in Antioch respectively, he came up with Valentinianus in Rome, Valens in Antioch and the usurper Procopius in Constantinople, and thus arrived at the year 365/366. This date cannot be maintained, because the figures described above certainly do not represent Roman emperors, Valentinian did not reside in Rome but in Milan, and moreover no official map would present a short-lived usurper like Procopius for Constantinople. Nevertheless this wrong date is still to be found in handbooks and encyclopedias. Other major urban centres are indicated by town walls, a certain type of circular fortification, sometimes with accurately rendered buildings inside - a type of representation we also find on the Madaba map. On the TP this type is more frequently used in the eastern part of the world. In the west there are only two cities indicated that way, Ravenna and Aquileia. This helps us with the dating of the map. Ravenna gained major importance only after 402, when it became the imperial residence for the Emperor Honorius. Aquileia, on the other hand, was destroyed by Attila's huns in 452. This allows us to date the composition of the ancient original of the TP roughly to the first half of the fifth century.9

The Holy Land on the Tabula Peutingeriana (sheet IX, detail).

Let us now concentrate on parts shown on both the TP and the Madaba map.10 Due to the general layout of the TP, the composer had to accomodate the geographical outlines to a long and narrow band, a task which he achieved with outstanding skill. He made the Mediterranean Sea divide the various regions, so we find the northern parts of Egypt at the bottom of the map, facing up, as now usual. On the other hand the coastline of the Holy Land - together with modern Lebanon, Syria and Jordan - is oriented from left to right. That means that the east is on the bottom of the map, and the west, the Mediterranean Sea, above, exactly reverse to the Madaba map.11
The TP with its smaller scale does not give so many details, and there are only a few names of cities or villages that we find in both the TP and the Madaba map: e.g. Koreous,12 Archelais, Jericho, Ascalon and, of course, Jerusalem. On the other hand the TP gives the main roads which are remarkably not indicated at all in the Madaba map, with a minor exception in the north of Jerusalem. However "the fourth" and "the ninth milestone" to the north - or northwest - of Jerusalem, belonging to the road leading to the Mediterranean coast, are on the Madaba map.13
It is difficult to decide why the Madaba map gives the villages along the roads, watch-towers and even ferries, but not the roads themselves. First of all, we must keep in mind what roads in those "pre-automobile" times looked like. They were mere tracks on the ground, one strolled along wherever one wanted to. Travellers only wanted to avoid swamps, thorny thickets, rocky grounds and the like - and, on the other hand, to avoid trespassing on cultivated fields.
The only exceptions were the Roman military roads. But we may ask whether they still existed, and if so, in what condition ? Probably the indicated short section of a road north of the Damascus Gate (Bab al-'Amud) at Jerusalem was the only paved one, or rather a section of the former military road where the pavement was still intact.14 There might even be a simpler solution: red or white lines, running like a spider's web over the whole area of the Madaba map would have distorted the impression of the picture, disturbing the symbols and written texts or being interrupted by them. Thus the mosaicist wisely left out the roads. A traveller or pilgrim could easily find his way from one point to another, and, above all, the Madaba map was evidently not an itinerarium.

Tabula Peutingeriana, sheet III, detail: the "city-walls" of Aquileia.

Jerusalem on the TP is indicated by the comparatively modest sign of the "double-tower" like Jericho, Neapolis (Nablus), Bostra, Philadelphia (Amman) and others. This sign is found in a similiar form on the Madaba map. Such a sign on the TP indicates the quality of accomodation to be found there and not the importance of a town. It has been pointed out that, together with another sign in the TP (the so-called "square building"), these double-towers indicated a building particularily suitable for the official travellers of the cursus publicus. Here they found not only shelter but were also - not without importance in late antiquity - safe from robbery and from local people, who often enough harboured not too friendly feelings towards soldiers, civil servants and tax collectors. They were strongholds with a large quadrangular courtyard and a well inside. That architectural type lived on in the "caravanserai" in the Orient and in the medieval monasteries in the West, with their cloisters and the fountain inside.

Tabula Peutingeriana, sheet IX, detail: Jerusalem.

The nomenclature of Jerusalem is peculiar: antea dicta H(i)erusalem, nunc Aelia Capitolina - "formerly it was called Hierosolyma/ Jerusalem, but now Aelia Capitolina". This refers to Emperor Hadrian, who destroyed Jerusalem, forbade the Jewish people to settle there again, and founded a new town in honor of the Capitoline Jove, bearing his name (Aelius). The writer in the Notitia dignitatum of the fifth century uses that name, which was familiar to Christian authors too, who often used only the name Aelia (Ailia).
The expression mons oliveti is evidently a concession to Christian pilgrims (note that the mountains were added later), while Bethlehem does not appear in the TP at all. It seems that Christianity is present in the TP, but did not play a major role with its composition.

Tabula Peutingeriana, sheet VIII, detail: Petra, the Sinai Desert and the Nile Delta.

In the west (or, better, in the southwest) we find the name of Petra. The sign was added later in the course of some corrections on our copy. The site itself was indicated right from the beginning in the original.
The Sinai desert is inscribed with two lengthy texts. In red letters:
Desertum ubi quadraginta annis erraverunt filii Israel ducente Moyse
("Desert, where for forty years the children Israel wandered about under the leadership of Moses")
and one beneath, in black letters:
hic legem acceperunt in monte Syna
("Here they received the commandments on Mount Sinai").
A sanctuary there is represented by an architectural sign that is used for Christian churches in the Madaba map as well.15 In this case it does not depict a Christian church or a monastery, but obviously a pagan temple (Ad Dianam - probably near Hal Timn'a). Thus such signs appear on the TP as symbols for both pagan sanctuaries and Christian churches, e.g. the Church of St. Peter in Rome. Therefore it marks any sanctuary, pagan or Christian, where travellers and pilgrims could find accomodation. Monasteries offer it until today, and we know that some pagan temples had this function as well.
Some of the placenames of the TP in the Nile delta correspond with names on the Madaba map, e.g. Pelusium (Tell el Farama), and Niciu - h Nikiou, Athribis and Hermoupolis.16 The rendering of the area makes it clear that the designer of the TP had no knowledge whatsoever of the various rivers and mouths of the Nile. On the other hand we can see even from the few traces extant, that the artist of the Madaba map was acquainted with river estuaries or had better sources for his work. Notwithstanding some coincidences and regardless of its official character, an ancient copy of the Tabula Peutingeriana apparently was not among the immediate sources for the Madaba map.


1. This was stated thirty years ago by Annalina and Mario Levi, Itineraria picta (1967), 79 sqq.; cf. esp. 170: "é stato possibile identificarli con quegli edifici che sorgevano lungo le strade romane e che avevano la funzione di sopperire alle necessità di rifornimento e di riposo dei viaggiatori".

2. Ekkehard Weber, Tabula Peutingeriana - Codex Vindobonensis 324 (1976), 21 sq.; Rudolf Hanslik, RE IX A (1961), 1270 sqq.

3. Wolfgang Müller-Wiener, Bildlexikon zur Topographie Istanbuls (1977) 255-257; the second column of Constantinople is the column of Arcadius (Dikilitas), ibid. 250-253.

4. There was a similar drawing, now hardly visible, of the famous Pharos at Alexandria, el-Iskindiriya, in Egypt. The indication of Alexandria itself is missing, evidently due to an error of the draftsman. There is enough space for a larger drawing, most probably a city-wall with towers, as we have in other cases.

5. In the case of Antioch we can prove that these figures were personifications, indeed, the goddesses of the city. We recognize the Tyche of Antioch by the famous Hellenistic sculptor Eutychides of Sikyon in 296 BC. It is evident that the figure on the map is a sketch of that statue.

6. I am very grateful to Pauline Donceel-Voûte who drew my attention to the fact that these arches are meant to represent a portico, esp. in the case of Antioch, where, as Libanios points out, large porticos were significant for her character as imperial capital (Antiochikos 202 and 196); see her article La Carte de Madaba: cosmographie, anachronisme et propagande, in Revue Biblique 95 (1988), 519-542, esp. 532 sqq.

7. Because he believes in the temple of Apollo, Luciano Bosio is forced to date the TP to before this accident; cf. La Tabula Peutingeriana - una descrizione pittorica del mondo antico (1983),147-159.

8. Strangely enough, he seems to have known nothing about the Madaba Map.

9. We can say more than that. The most striking parallels to the sketches on the Tabula are found on the triumphal arch in Sta. Maria Maggiore in Rome, cf. Heinrich Karpp, Die frühchristlichen und mittelalterlichen Mosaiken in Santa Maria Maggiore zu Rom (1966), fig. 27 and 28. These - early - mosaics can easily be dated, as an inscription in the center reads that Pope Sixtus II (432-440) had made them made for the people: Xystus episcopus plebi dei. In addition to it a rather strange piece of poetry, quoted by the Irish monk Dicuil (Anth. Lat. 724) tells us that during the fifteenth consulate - ter quinis aperit cum fascibus annum - of Theodosius (II, for only this emperor held so many and more consulates), two of his "servants" - famuli in Latin - made a sort of "revised edition" of a map of the world. I cannot believe that many world-maps existed in those days, so I think that this piece of poetry was part of the lost introduction to the map, which came to us in the form of the TP. The fifteenth consulate of Theodosius II was in 435. For the question of the date of the TP see Levi (cited above) and Ekkehard Weber, Zur Datierung der Tabula Peutingeriana, in: Labor omnibus unus (in honor of Gerold Walser, 1989), 113-117.

10. We must note that the area shown on the Madaba Map is on two different sheets of the TP: The Nile and the Sinai Peninsula are on the sheet VIII, and the Holy Land is on sheet IX. This late division is artificial as the original scroll presented the whole region in context.

11. In my opinion this orientation of the Madaba Map is not - or not primarily - due to (otherwise unknown) general rules of Early Byzantine geography. In fact it presents exactly geographically the Holy Land to the visitor of the church like a bird's eye view of it, when he stood in front of the mosaic, looking to the east. Even the Nile, from this point of view, seems to have the correct direction.

12. If identical with the Coabis on the TP.

13. ''To tetarton and to enna(ton), see Herbert Donner, The Mosaic Map of Madaba (1992), 53 nos. 53 and 54.

14. I like the idea of Isaac Roll who has pointed out that this section of the street had been paved only a short time before by order of Justinian to facilitate the transport of the large columns for his church of The Mother of God (Nea Theotokos), built in those days. For the Roman road-system see the article of David Graf.

15. E.g. the church of St. Philipp where the eunuch of Queen Kandake was baptizised; Ain ad-Dirwa near Halhul, according to the identification by Donner (see footnote 12), 60 sq. no. 81.

16. During the congress there was a lively discussion of an enigmatical remark of an early visitor, stating that he saw Rome (?) and Babylon on the Madaba Map. There seems no need to add new ideas about this precarious evidence, probably resulting from an exuberant enthusiasm felt in front of the original, which we can share. But in case he actually did see the mosaic in a better condition than today, or with additional fragments now lost, we may note that at least the second location could well have been Babylonia in Egypt (Qasr esh-Sham, Cairo).

This article was first published in: The Madaba Map Centenary 1897-1997, Jerusalem 1999, 41-46.

logo logo

Created Saturday, December 16, 2000 at 11:32:57
by Eugenio Alliata ofm in collaboration with Stefano de Luca ofm
Webmaster: John Abela ofm - Space by courtesy of Christus Rex
copyright - Studium Biblicum Franciscanum - Jerusalem 2000