9,4:4 Seleukeia

Seleukeia (see DARE) was the seaport of Antioch and the two cities were certainly connected on the ground, but there is no pressing evidence that the Tabula connected them on the chart.

Talbert, who seems to take the view that the drawing of Antioch's protective goddess is a non-original feature of the chart, opines (TPP532) that "a route into Antiochia from Selevcia has been overwhelmed by the former's elaborate symbol". A better argument for reconstructing the line would be that Seleukeia is likely to have been a starting point for itineraries towards Chalcis and Hierapolis, but the issue remains moot.

10,1 Chalcis

The confusing surfeit of roads east of Chalcis has never been conclusively explained. The trend in scholarship has been to see the series Cani - Bersera - Bathna as a duplication of the names Calcida - Berya - Bannis.

Talbert backs this with respect to the first two place-names, but seems agnostic about Bannis (see TPP2379, TPP2435 and TPP2436): "Calcida, Berya and the route to Bathna evidently make a second appearance on the route immediately below (as Cani, Bersera)."

Dussaud (1927) argued that Bannis and Thiltavri were false replications by the chart-maker of two place-labels belonging to an itinerary on the opposite bank of the Euphates, Batnis and Thiar:
Signalons ici une erreur de la Table de Peutinger dans la route d’Antioche à Zeugma. Il faut lire Emma (‛Imm) — Calcis (Qinnesrin) — Beroea (Alep) — Bathnae (ms. Bannis) — Hierapoli (Menbidj) — Zeugma (Balkis), en supprimant Thiltauri et Bathna qui sont empruntés à la route de Zeugma à Edesse (Thiltauri pour Thiar ou Daiara, et Bathna pour Bathnae-Seroudj). 
I have not ventured any emendation in the animated edition. However the abstract shows the Calcida - Berya - Bannis route as duplicated, in both green and brown.

9,4:3 Bacataiali

The place marked Bacataiali appears to be a mutatio on the highway crossing the hills between Laodicea and Antiochia. Miller 803 proposes that the figure of 27 miles is the distance separating it from Daphne, south-west of Antioch. In the Antonine Itinerary the stop at a similar place is named as Baccaias. Miller quotes the spelling Bachaias from the so-called Jerome Maps.

The emendation  restores the line into Antioch.

The alternative of a line to Seleukobelos across the Bargylus Mons and onwards to Apamea by way of the road shown in DARE was considered, but that road seems to be of only local importance.

9,1:2 Rhose

Miller 817 comments:
Nach der [Tabula] müßte man annehmen, daß die Straße von Damaskus bis Hatita westlich von Bosra gelaufen wäre; da aber Chanata an der Strecke über Bosra liegt und ferner die Römerstraße von Damascus - Chanata - Philadelphia zweifellos erhalten ist, und da ferner westlich von einer Straße nichts erkannt ist, so liegt Bosra an dieser Straße, und somit ist die Verbindung nach Philadelphia nicht von Hatita, sondern von Bosra aus zu ziehen. 
My emendation reflects this.

Richard Kiepert questioned whether Rhose, which has no distance figure attached, was a real place, and considered it "verschrieben und identisch mit Bostris, ... also doppelt vorhanden." Miller 818 quotes this, but instead adopts the stance that Rhose must be an unlocated site south of Bosra. On balance it seems best to treat it as some entity in the immediate vicinity of Bosra.

9,1:3 Jericho

Talbert argues for the restoration of a line connecting Jericho and Nablus. This emendation is argued from an appropriate distance figure already being entered on the manuscript here.

Finkelstein 33 notes that the space separating the distance figure into two halves, as "XX XII", suggests the downwards linework must have cut through the script itself, but voices doubt (note 45) that the endpoint of the road is Jericho:
Thomsen and Miller assumed that the number on the map indicates the distance from Neapolis to Jericho. The attempt to identify the line that was omitted from the map with the road from Neapolis to Jericho raises several questions: the number given on the map, 32 miles, is close to the distance between Neapolis and Jericho. Since the road goes down to Phasaelis and not directly to Jericho one would have expected the distance to be given only to the meeting-point with the Jordan valley road (at Phasaelis), about 21-22 miles. The Jordan valley road is, in any case, problematic and Phasaelis itself is missing on it.
The references are to Thomsen, P. 1917. ‘Die römischen Meilensteine der Provinzen Syria, Arabia und Palaestina’, Zeitschrift des Deutschen Palästina-Vereins 40, 20 and Miller 834.

9,1 Thamaro

No emendation is required for the Jerusalem—Thamaro—Rababatora road, but the reader will profit from some elucidation being given. Finkelstein 30-31 notes: "‘Thamaro’ undoubtedly is the Thamara of the Roman-Byzantine sources, and is generally identified with Ain al-Hosb in the Araba."

On DARE, one sees how Thamara is located in the Araba valley, an arid plain south of the Dead Sea located below the Negev plateau. Ain al-Hosb, the name of a British police base there, has been replaced with an Israeli name. More history of the location is given in the Wikipedia article on Ir Ovot. It was reached by proceeding southwards from Jerusalem over a long highway via Hebron, a place which is not marked on the TP. Finkelstein 33 notes:
Obviously enough the place where the Thamara—Transjordan road diverged from the Jerusalem—Elusa road was in or near Hebron. It would seem therefore, that the name and placing of Hebron were omitted from the junction of these two roads.
Thamara is also a crossroads. This junction must have been the end-point of an itinerary used to make up this part of the TP, since the itinerary's only obvious continuation, southwards from Thamara towards the Gulf of Aqaba, is not present on the TP, although this does not prevent Finkelstein 31 from speculating about its likeliest course:
Most probably, the road from Thamara to Aila crossed the Araba to et-Tlah (Toloha of the Roman—Byzantine sources) [DARE] in the east. From there the road went up Naqb (the pass of) Dahal to the vicinity of Bostra (present-day Buseira), where it merged with the Via Nova Traiana. This is the easiest and the shortest way. It is reasonable to assume that the connexion with the Transjordan road should have occurred between the stations Negla and Thornia, and not near Rababatora as indicated in the map.
No such road is indicated in DARE.

Editing the TP I had the option of diverting the Thamara-Rababatora line in the light of Finkelstein's speculation, or even running it straight to Petra, or letting the connection stand where it is drawn, but as a thin black line. I chose the third option, since Finkelstein gathers ample evidence to make its interpretation as a minor road swerving away northward from the junction the most palatable:
Aharoni showed that the distance given on the map from ‘Thamaro’ to the Transjordan road — 68 miles — exactly fits the distance between Ain al-Hosb and er-Rabba in Transjordan north of Karak. Alt, and after him Aharoni, saw in ‘Rababatora’ a combination of two names: Raba (er-Rabba) and Batora = Bettoro, el-Lajjun, some thirteen miles south-east of er-Rabba....
The references are to: Aharoni, Y. 1958. 'Tamar and the Roads to Elath', EI, 5, 129-34 (in Hebrew) and Alt, A. 1935. 'Aus der "Araba" II', Zeitschrift des Deutschen Palästina-Vereins 58, 33.

9,1:4 Amavante

The linework above and to the right of Jerusalem has suffered as a result of a Christian user inserting a gloss about Jerusalem in what is already an overcrowded space. The error leads to a puzzle: a geographically implausible connection from Gofna to a place named ‘Amavante’ — the town of Emmaus. Finkelstein 32 sets out the issue in neutral terms as follows:
Three important roads connected the Jerusalem region with the coastal plain near Lydda during the Roman period. These are (from south to north) the roads:
  1. Jerusalem—Emmaus— Lydda
  2. Jerusalem—Beth Horon—Lydda; and 
  3. Gofna—Thamna—Antipatris.
The appearance of Emmaus midway and the distances given (19 and 12 miles) fit the course of the Jerusalem—Emmaus—Lydda road. Still, other details on the map refuse to fall into place. The road to Emmaus and Lydda went directly west from Jerusalem and reached Lydda itself; whereas the line in the Peutinger map verges westwards at Gofna — far north of Jerusalem — and connects up with [the] ‘inland’ coastal plain road, north of Lydda.
Finkelstein's choice of solution 3 is over-complicated: "On the Gofna—Thamna—Antipatris road [the chart-maker] placed ‘Amavante’ — Emmaus, where the town of Thamna should have been."

The graphic solution is far easier: it recognizes that route-line 1 has been simply miscopied to accommodate the gloss. Miller 835 silently adopts this solution too.

8,4:1 Sinai

At some point in the TP's history, a Christian user has added Mount Sinai to the chart. There was however no space for the mountain icon in the narrow space directly above the Red Sea and below the route-line from Clisma (present-day Suez) to Aila (present-day Aqaba), so the hill drawing had to be placed above the route-line.

This cheat has led to much scholarly confusion, since many have assumed that the route veers south, making it a detour via Mount Sinai, a centre of Christian (and possibly Jewish) pilgrimage in antiquity. An essay by Philip Mayerson, which unfortunately I have not yet been able to read, contends that Phara denotes the well-watered oasis town of Feiran in the southern Sinai.

Finkelstein 30 points out why the line drawn on the Tabula cannot denote the long route through Feiran:
The shortest possible line between these points [Clisma and Aila] crosses Sinai from west to east; indeed, the total of the distances appearing on the map between Clisma and Aila — about 170 miles — closely corresponds to the distance between Suez and Aqaba along the Darb el-Hajj. On the other hand the appearance of 'Phara' on this road is misleading; the road between the heads of these two Red Sea gulfs did not go by way of south Sinai.
The true distance by the direct route, in Finkelstein's calculation, is 265 kilometres or 175 Roman miles. The numbers on the Vienna manuscript add up to 186 Roman miles, the slight excess one would expect to arise from a few twists and turns, yet far short of the arduous journey via Mount Sinai, past Zerbal and Jebel Musa, which would be 50 per cent longer or 400 kilometres in the calculation of Eduard Sachsse.

Before we proceed, it is best to consider what we actually see in the TP. Firstly, I do not read Phara as previous scholars do, but Phira or Phiro. There appears in addition to have been some kind of copyist trouble with the route, which not even Sachsse has correctly described, despite a valiant attempt (he used an edition and did not have access to any photographs of the TP). Starting at Arsinoe at the left, we find:
  1. a rightwards bend but no chicane, the words Clisma XL, no distance figure;
  2. a chicane, a tear in the parchment, the remnant of a word Talbert transcribes as [ - ? - ]++ia (TPPlace449);
  3. a chicane with the distance figure LXXX only
  4. a chicane with the text Phiro L.;
  5. a reverse chicane with the text Haila and a two-tier figure, VI at the top, and X under it. Talbert justifies the false chicane thus: "The chicane at the end of the stretch turns up, not down, to avoid the small inlet with Haila at its end."
If we discount Mayerson's position, what possibilities remain?

The first and easiest solution is to suppose that Phira is an otherwise forgotten name for one of the oases on the direct road. Sachsse proposes it represents the town of Nekhel (Wikipedia) in the central Sinai. Graf opposes this on the grounds that archaeological excavations there show the station to be Islamic and "purely medieval" constructed in the Ayyubid-Mamluk period.

Lipiński 33 argues that Phira matches the Paran (P'rn) desert region mentioned in 1 Kings 11:18, and therefore likely attaches to that region's main fort at at-Tamad, halfway between Nekhel and Aqaba.

The second solution supposes there has been a muddle. Finkelstein 30 argues it thus:
‘Phara’ or ‘Pharan’ is frequently mentioned in Roman and Byzantine sources. There is no doubt about the identification of Pharan with the Feiran oasis in Wadi Feiran in south Sinai. The remains of the town and the ruins of its churches, dating to the time when monasticism flourished in Sinai, are located in Tell Maharrad in the centre of the oasis. It is unlikely that there were two places with similar names in Sinai — Pharan in the south of the peninsula and ‘Phara’ on the Clisma-Aila road. In fact, Christian Pharan is also mentioned in the form of ‘Fara’.
All this points to an interesting error in the Peutinger map. The editor of the map probably knew of an important road existing between Clisma and Aila which crossed the Sinai peninsula; he also had information on stations along the way and the distances between them. In addition, he knew of a place called ‘Phara’ in Sinai, but apparently did not know its exact location, and confused the facts. The line giving the direction of the road, the stations and the distances between them fits in very well with the direct way from Clisma to Aila (by way of present-day Nakhl and Thamad). In place of one of the stations whose name was probably unknown to him, he put ‘Phara’ (of south Sinai) by mistake.
Finkelstein overlooks the valuable clue given by the linework, which suggest a third variant: that the TP chart-maker worked from a chorographic chart with more detailed linework which included an additional road looping south to Pharan. The trace of this may be the reverse chicane. In this case, Phira does not belong in the hook and is only "projected" there by our misreading of the chart.

To sum up, the distance L to the right of Phiro is the distance from some desert station, most probably at-Tamad, to the head of the Gulf of Aqaba. Whether the chicane contains an old name (Paran), a wilfully wrong name (Feiran) or a nothing at all cannot be determined.

Finkelstein, Israel. 1979. “The Holy Land in the Tabula Peutingeriana: A Historical-Geographical Approach.” Palestine Exploration Quarterly 111 (1): 27–34.  DOI: 10.1179/peq.1979.111.1.27.
Graf, D.F. 1996. Map 76 in The Barrington Atlas. Online.
Lipiński, Edward. 2000. The Aramaeans: Their Ancient History, Culture, Religion. Louvain: Peeters.

Mayerson, Philip. 1981. “The Clysma-Phara-Haila Road on the Peutinger Table.” In Coins, Culture and History of the Ancient World: Numismatic and Other Studies in Honor of Bluma L. Trell, Detroit: Wayne State UP. 167–76.

Sachsse, Eduard. 1928. “Die Römische Strasse durch die Sinaihalbinsel nach der Peutingerschen Tafel.” Zeitschrift des Deutschen Palästina-Vereins 51 (4): 265–68. http://www.jstor.org/stable/27929731.

2,5 Provincia

Provincia, in spaced-out black letters, spans columns 2,5 to 4,2 and is laid over the African section of the Tabula. It is often assumed the word prefixes Africa, making a phrase "Provincia Africa", but this seems unlikely. Firstly, Africa is written in red letters, not black. Secondly, most of the regional designations on the Tabula are purely geographical terms without the inclusion of any legal quality.

It seems far likelier that this label has floated away from its proper place, the region that Latin-speakers conventionally called Provincia: Provence. The correct location is marked in the drawing below in the fainter script.
This is a strange error, and it has a counterpart in the peculiarly small regional designation, Gallia Comata, in column 1,3. That region properly includes all the three Gauls north of Lyon, yet the label is shrunken and placed too low on the chart. More research will be required into these two anomalies.

10,2:1 Attas

Proposed by Talbert 2596, this link to Attas is concordant with other reconstructions of the Syrian road system.