• Plate 3 Plan of the eastern
  Plate 3
  
  Plan of the eastern quarter (Nabataean gate, so-called “Trajan’s” palace, large church with a central plan)
  
  Th. Fournet, Ch. March (architects) - A. Ract-Madoux (topographer)
  
  Dentzer et al. (2002)
  quarter from Dentzer et al. (2002)

• Plate 3 Plan of the eastern
  Plate 3
  
  Plan of the eastern quarter (Nabataean gate, so-called “Trajan’s” palace, large church with a central plan)
  
  Th. Fournet, Ch. March (architects) - A. Ract-Madoux (topographer)
  
  Dentzer et al. (2002)
  quarter from Dentzer et al. (2002)

• Plate 4 Elevation and
  Plate 4
  
  Elevation and plan of the west facade
  
  S. Vatteoni (graphic designer) J.-P. Fourdrin (architect) - Ph. Tondon, P.-M. White, S. Vatteoni
  
  Dentzer et al. (2002)
  plan of the west facade from Dentzer et al. (2002)

• Plate 4 Elevation and
  Plate 4
  
  Elevation and plan of the west facade
  
  S. Vatteoni (graphic designer) J.-P. Fourdrin (architect) - Ph. Tondon, P.-M. White, S. Vatteoni
  
  Dentzer et al. (2002)
  plan of the west facade from Dentzer et al. (2002)

• Plate 5 Plan of the
  Plate 5
  
  Plan of the large church with a central plan
  
  P.-M. White, Ch. March, S. Vatteoni
  
  Dentzer et al. (2002)
  large church from Dentzer et al. (2002)

• Plate 5 Plan of the
  Plate 5
  
  Plan of the large church with a central plan
  
  P.-M. White, Ch. March, S. Vatteoni
  
  Dentzer et al. (2002)
  large church from Dentzer et al. (2002)

• Plate 7 Plan of the
  Plate 7
  
  Plan of the southern baths with chronology
  
  H. Broise (architect) - P.-M. Blanc
  
  Dentzer et al. (2002)
  southern baths from Dentzer et al. (2002)

• Plate 7 Plan of the
  Plate 7
  
  Plan of the southern baths with chronology
  
  H. Broise (architect) - P.-M. Blanc
  
  Dentzer et al. (2002)
  southern baths from Dentzer et al. (2002)

• Plate 8 Plan of the
  Plate 8
  
  Simplified plan of the central baths
  
  Th. Fournet
  
  Dentzer et al. (2002)
  central baths from Dentzer et al. (2002)
• Plate 9 Schematic
  Plate 9
  
  Schematic section of the central thermal baths (current and restored state)
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section of the central thermal baths from Dentzer et al. (2002)

• Plate 8 Plan of the
  Plate 8
  
  Simplified plan of the central baths
  
  Th. Fournet
  
  Dentzer et al. (2002)
  central baths from Dentzer et al. (2002)
• Plate 9 Schematic
  Plate 9
  
  Schematic section of the central thermal baths (current and restored state)
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section of the central thermal baths from Dentzer et al. (2002)

• Plate 10 Plan of the
  Plate 10
  
  Plan of the western district (overall plan of the block, with modern and ancient structures)
  
  Th. Fournet - G. Aronica (topographer)
  
  Dentzer et al. (2002)
  western district from Dentzer et al. (2002)

• Plate 10 Plan of the
  Plate 10
  
  Plan of the western district (overall plan of the block, with modern and ancient structures)
  
  Th. Fournet - G. Aronica (topographer)
  
  Dentzer et al. (2002)
  western district from Dentzer et al. (2002)

• Plate 11 Section and
  Plate 11
  
  Section and perspective restored on the cryptoporticus
  
  Th. Fournet
  
  Dentzer et al. (2002)
  perspective restored on the cryptoporticus from Dentzer et al. (2002)

• Plate 11 Section and
  Plate 11
  
  Section and perspective restored on the cryptoporticus
  
  Th. Fournet
  
  Dentzer et al. (2002)
  perspective restored on the cryptoporticus from Dentzer et al. (2002)

• Plate 12 Elevation of
  Plate 12
  
  Elevation of the entrance to the macellum
  
  Th. Fournet
  
  Dentzer et al. (2002)
  the entrance to the macellum from Dentzer et al. (2002)

• Plate 12 Elevation of
  Plate 12
  
  Elevation of the entrance to the macellum
  
  Th. Fournet
  
  Dentzer et al. (2002)
  the entrance to the macellum from Dentzer et al. (2002)

• Plate 13 Plans of the
  Plate 13
  
  Left - Plan of the “north-south street” sector, state 1960, before clearance
  
  Right - Map of the “north-south street” sector, state 2000, after clearances, restoration of the old lines
  
  Th. Fournet - E. Audouy, G. Aronica (topographers).
  
  Dentzer et al. (2002)
  “north-south street” sector before and after clearance and restoration from Dentzer et al. (2002)
• Plate 14 Elevation
  Plate 14
  
  Elevation section on the north-south street at the level of the central thermal baths
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section on the north-south street from Dentzer et al. (2002)
• Plate 15 Elevation
  Plate 15
  
  Elevation section on the north-south street at the level of the Omar mosque
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section on the north-south street at the level of the Omar mosque from Dentzer et al. (2002)

• Plate 13 Plans of the
  Plate 13
  
  Left - Plan of the “north-south street” sector, state 1960, before clearance
  
  Right - Map of the “north-south street” sector, state 2000, after clearances, restoration of the old lines
  
  Th. Fournet - E. Audouy, G. Aronica (topographers).
  
  Dentzer et al. (2002)
  “north-south street” sector before and after clearance and restoration from Dentzer et al. (2002)
• Plate 14 Elevation
  Plate 14
  
  Elevation section on the north-south street at the level of the central thermal baths
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section on the north-south street from Dentzer et al. (2002)
• Plate 15 Elevation
  Plate 15
  
  Elevation section on the north-south street at the level of the Omar mosque
  
  Th. Fournet
  
  Dentzer et al. (2002)
  section on the north-south street at the level of the Omar mosque from Dentzer et al. (2002)

• Plate 16 Plan of the
  Plate 16
  
  Plan of the urban center: from the central crossroads to the praetorium district
  
  Th. Fournet
  
  Dentzer et al. (2002)
  urban center from Dentzer et al. (2002)

• Plate 16 Plan of the
  Plate 16
  
  Plan of the urban center: from the central crossroads to the praetorium district
  
  Th. Fournet
  
  Dentzer et al. (2002)
  urban center from Dentzer et al. (2002)

• Plate 17 Elevation Section
  Plate 17
  
  Elevation section of the exedra (pseudo-nymphaeum) and the nymphaeum (pseudo-kalybe).
  
  Th. Fournet
  
  Dentzer et al. (2002)
  of Exedra (Pseudo-Nymphaeum) And Nymphaeum from Dentzer et al. (2002)

• Plate 17 Elevation Section
  Plate 17
  
  Elevation section of the exedra (pseudo-nymphaeum) and the nymphaeum (pseudo-kalybe).
  
  Th. Fournet
  
  Dentzer et al. (2002)
  of Exedra (Pseudo-Nymphaeum) And Nymphaeum from Dentzer et al. (2002)

• Plate 20 Plan of
  Plate 20
  
  The current sector of the camp with report of the results of the geophysical detection and distribution of the decorated blocks (the size of the pellets is proportional to the density of blocks discovered).
  
  Plan Th. Fournet, geophysical detection A. Kermorvant
  
  Dentzer et al. (2002)
  IIIa Cyrenaica legion camp with geophysics from Dentzer et al. (2002)

• Plate 20 Plan of
  Plate 20
  
  The current sector of the camp with report of the results of the geophysical detection and distribution of the decorated blocks (the size of the pellets is proportional to the density of blocks discovered).
  
  Plan Th. Fournet, geophysical detection A. Kermorvant
  
  Dentzer et al. (2002)
  IIIa Cyrenaica legion camp with geophysics from Dentzer et al. (2002)

• from Dentzer et al. (2002:89)

Stratigraphic studies and excavations carried out at various points below the level of the cathedral at the end of the 5th century led to the identification, alongside minor episodes, of a first monumental restructuring of the sector datable to the course of the 4th century AD.

The new occupation level is located barely ten centimeters above the Nabataean level. This phase witnessed a reversal of the use of the space within the boundaries of the old porticoed courtyard. The old portico became an open space, made of beaten earth. What was the courtyard, however, was covered. Its floor is paved with white mosaics over most of the surface. The variable arrangement of the large tesserae differentiates the space, especially to the south.

In the northern half, there are figurative motifs that form polychrome carpets. Two blocks have also been identified, embedded in this surface, in which cavities have been dug to receive posts to delimit an area with a liturgical function.

Over the entire surface covered with mosaics that could be uncovered, no preserved wall was identified. There must have been one under the water conduit from the end of the 6th century. The organization of this space, which must have been covered, therefore eludes us. We can assume, however, that the building was already a church.

The mosaics, based on basic substructures (simple hedgehogs), bear very clear marks of impacts from blocks falling from parts of high points of the building. It was certainly an earthquake, perhaps the one attested in 447, which caused the violent destruction of the building. The materials used in this construction, already borrowed from earlier buildings of the Nabataean and Roman periods, were clearly carefully set aside to serve in the construction of the large church with a central plan, which can be dated to the end of the 5th century.

Only a few unusable pieces were left in the fill brought in for its construction. It was probably on this occasion that the blocks still integrated into the construction or found in scattered order were numbered. A temporary deposit of building materials, consisting in particular of column drums, was found under a vault supporting the floor of the following phase near the southwest apse.

• from Dentzer et al. (2002:89-92)

It is unknown whether the alterations to the monumental complex in the 4th century, which remained as close as possible to the Nabataean and Roman levels, influenced the organization of the district. However, after the earthquake, larger-scale works remodeled the space for the construction of a large centrally planned church. The land was raised by nearly 2.50 meters, with embankments to the north and vaulted substructures to the south, to create a horizontal foundation for a complex whose footprint reaches 65 by 70 meters (excluding the access staircase), in its current state of clearing. The west facade of this vast monumental base consists of a large wall decorated with niches borrowed from a building from the early imperial period. The center of this monumental program is a square block measuring 45 m on each side, within which is inscribed a circular shape consisting of a back wall, preceded by , preceded by a Corinthian portico that forms a sort of ambulatory.

Four large apses were set into the corners of the square. To the east, along the axis of the monument and more precisely between the two ends of the walls decorated with Nabataean half-columns, a rectangular room, the sanctuary, opened, ending with an apse set within a pentagonal envelope. Its entrance is marked by a threshold closed by a low chancel in a first state, then raised to form probably without a high chancel. The sanctuary was flanked by two adjoining rooms. The southern adjoining room ended externally with a flat chevet in which an apse was inscribed. A door gave direct access to the exterior, towards the east.

The method of covering the central circular space, with a span of more than thirty meters, poses a problem, in the absence of intermediate supports, which exist in the Church of Saints Sergius, Bacchus and Leontius. The latter has a similar plan with smaller dimensions (diameter of about 17 m). Four L-shaped pillars supported a central dome. It must be assumed that an annular vault made of light materials (including basalt scoria) covered the space between the circular portico and the back wall. This vault buttressed a wooden dome as, for example, at Qalaat Semân.

The church was accessed from the west, where a sort of propylaeum was arranged along its axis, the base of a column of which has been found. It is assumed that there was a monumental staircase that bridged the difference in level between the cathedral and the Nabataean arch, the floor of which was at least 4 m lower, even in the early Christian period.

To the central square mass of the building is connected to the south another rectangular building of 12 x 20 m. It is a second church with a basilica plan, with three naves, oriented. To the east, a slightly horseshoe-shaped apse is flanked by two annexed rooms inscribed in a flat chevet. On its southern flank opens, over almost its entire length, a second apse, of much larger dimensions (10 x 6 m). This apse, in a semicircle exceeded on the inside, is part of an envelope whose exterior draws a seven-sided polygon. The west face of this church, in which three doors opened, was undoubtedly preceded by a portico on two floors. The corbels which supported it at its north and south ends are preserved. The portico was located at the level of the church and must therefore have been accessible, from the west, via a stairway.

The northern boundary of this church is tangent to the circular space which constitutes the center of the large church. How were these two spaces articulated? Was the annex church entirely open to the north with simply pillars to support its roof? Two pilasters are preserved on either side at the ends of its northern boundary. In a first state, these two pillars were arranged obliquely relative to the rectangular plan of the annex. In this phase, the northern boundary of the church was therefore not linear. An oblique wall was directed towards two other supports. The annex church then fit into the annular plan of the space surrounding the central colonnade. In a second state, the pillars were offset and a foundation supplement was added to create a straight alignment that intersects the circular plan of the annular space. In this phase, the church, which can then be assumed to be closed, seems to take on a more autonomous character.

One is tempted to attribute these major modifications to a major phase of work that, following the re-examination of the ceramic material by P.-M. Blanc and D. Pieri, can be dated to the end of the 6th century. It is perhaps explained by a new earthquake attested in 551⁴⁹. It is in this phase that we must undoubtedly place significant alterations in the sanctuary: a synthronon was built at the same time as the threshold was modified and the low chancel was replaced by a high chancel. At the same time, a large opus caementicium pipe was installed, which crosses the church and the choir and feeds the fountains on the facade.

In this annex church, the apse which opens towards the south would best be explained as a baptistery. In its first state, it was located in the basement relative to the level of the main church, which rests, on this side, on vaulted substructures. It was accessible by three doors opening directly onto the space surrounding the large church and dominated by it. This lower area must itself have been surrounded by an enclosure and formed part of the church space: it gave, in fact, direct access to the baptistery. In 2000, very limited traces of the baptismal font were found, which had suffered serious damage: its space was, in fact, partially reoccupied by cellars until the modern period.

A major transformation took place, probably in the second phase of work which has just been mentioned: the baptistery was moved upstairs on the main circulation level of the ecclesiastical complex. To this end, corbels were embedded in the existing semicircular wall and two transverse walls were built.

The scale and overall organization of the architectural program, which combines the centrally planned main church with an annex church with an apse to the south, which can be interpreted as a baptistery, and a second room or annex chapel to the north, leads to the interpretation of this ensemble as an episcopal group that constituted one of the monumental centers of the city and dominated this district in the early Christian period.

In this coherent context, one may wonder whether the so-called Trajan's Palace with its triconch hall on floor should not be interpreted as an episcopal palace . A similar plan is found in a large house described by HC Butler to the east of the church of Saints Sergius, Bacchus and Leontius.

• from Dentzer et al. (2002:104-105)

We still do not know what the west-east street looked like at its origins. A survey of a very limited area, carried out in 1993-1994 north of the facade of the southern baths, revealed material from older occupation phases dating back to the Neolithic. Traces of a first level of circulation date back to a period between 1 and 20 of our era in a Nabataean context (unpublished ceramics and coins). For the construction work of the first state of the baths, a terminus a quo was set around 150 AD. For the installation of the stylobate of the first portico of the street, ceramics characteristic of the 3rd century confirm the date suggested by the style of the Ionic capitals. In this phase, the boundary wall of the baths was cut down obliquely, over a length of at least 11.50 m, to make way for the portico, which is approximately four meters wide. The northern facade is thus partially aligned with the current layout of the street. The portico was probably not completed with shops at this stage.

A new stage of transformations has been clearly identified in the 5th century, according to coins. The stylobate was reworked and relaid. A gray-beige limestone floor was laid under the portico and on the foundation of the first stylobate and abuts the timbered wall.

During the 6th century, the northern boundary wall of the baths was finally razed and replaced by a new, recessed wall to gain the space needed to install shops to complement the portico. The beginnings of dividing walls are visible in the northern facing of the wall.

On the other hand, the street paving was re-laid, raised by about thirty centimeters, probably not before the 6th century. The same operation must have been carried out at the neighboring crossroads. The foundations of shops from the Byzantine phase are preserved: remains of fish from the Red Sea have been found there, as well as traces of bone crafts.

In the Ayyubid-Mamluk phase, we observe encroachments of houses on half of the area of the porticoes whose paving had been systematically recovered.

Finally, at a much higher level, modern houses probably occupied the entire surface of the portico in this sector.

• from Dentzer et al. (2002:105)

A fragment of a polychrome mosaic carpet with geometric patterns (Solomon's knot, leaves, V-shaped checkerboard), dated to the mid-4th century, was studied in 2001 in the northern portico of the east-west street, south of the central baths (Khan ed-Dibs). To the north, a second carpet partially covers the first, five centimeters higher. It could mark a shop entrance or thermal baths located right at this location.

The mosaic pavement was certainly replaced, after destruction due to an earthquake in the second half of the 5th century, by the gray limestone tile floor still visible at the eastern end of this portico. This repair is accompanied by a raising of about 25 centimeters, in connection with a new reconstruction of the shops, and probably also of the portico, as was noted during the excavation on the south side of the street.

Marble slabs held together by copper alloy staples decorated the shop fronts. They were abandoned during a reconstruction carried out in the second half of the 5th century.

In medieval times the last limestone floor was removed and reclaimed and new buildings were built on the site of the portico.

• from Dentzer et al. (2002:105,108)

Further west, on the main street, the French archaeological mission in South Syria resumed in 1998-1999 the study of the cryptoporticus and tetrapylon excavated by Suleiman al-Mukdad between 1960 and 1980. The study of this sector was resumed with cleaning and some surveys. The work was extended to the area located to the north of this complex. The results of this research having been published recently, we will limit ourselves here to summarizing the main conclusions.

The cryptoporticus is limited to a simple gallery supporting the northern portico of the main west-east street. 108 m long and 5 m wide, it was terminated at both ends by an apse. The one at the western end, partially destroyed when the cryptoporticus was shortened to make way for the tetrapylon, was discovered and excavated in 1999. While the walls are built in courses of basalt blocks coated with pink mortar to the north, the vault is in opus caementicium using volcanic scoria. To the south, the gallery is lit by thirty-four basement windows that flare inwards. The southern interior wall was also punctuated by niches that are found in the north wall. It is from this side that the three accesses to the cryptoporticus open.

Phases prior to the cryptoporticus: a survey carried out to the west of the wall enclosing the western apse of the cryptoporticus provided a significant sample of residual ceramic material from the Bronze Age to the 2nd century AD.

Construction of the cryptoporticus: additional examinations carried out by P.-M. Blanc in 2001, during cleaning work carried out by the Bosra Antiquities Department, established the date of construction of the cryptoporticus at the end of the 2nd century or the beginning of the 3rd century. In this first phase, the floor was covered by paving or by a coating of sandstone pebbles set in a hydraulic mortar. Steep three-step staircases led down to the circulation level in line with the three access bays. The upper portico and the first row of shops opening onto it were installed in the same program.

Phase of monumental transformations: behind the street portico, a monumental entrance was created with three doors that give access to the space located to the north of the cryptoporticus, but two meters lower. From this phase, a second row of shops seems to have doubled those of the main street. Their split-level arrangement provided a clever solution to the problem of the difference in level between the street and the square to the north.

Installation of ramps: Access to the cryptoporticus, previously provided by three steep staircases, was replaced in the 4th century by the installation of long, gently sloping ramps, which particularly facilitated the handling of goods. During this phase, the cryptoporticus was backfilled to a depth of about sixty centimeters. Does this profound transformation correspond to a change in the status of this space?

Construction of the tetrapylon: at the end of the 3rd or beginning of the 4th century, when the decision was taken to build the circular square of the tetrapylon, the western end of the cryptoporticus was amputated by ten meters and closed by a wall equipped with a door.

• from Dentzer et al. (2002:108)

The tetrapylon, located in the center of a small round square surrounded by a gutter and onto which shops opened, was built during the 4th century, before the development of the square itself. The tetrapylon, or rather the tetrakionion, is currently reduced to four square bases measuring 5.10 m on each side. On each of them were placed four columns crowned by capitals whose cores are preserved. Roughly grooved, they must have been covered with bronze plates, probably gilded.

The circular square, which is distinguished from the street by a pavement of small square slabs, arranged at 45°, is divided into four quadrants by the two intersecting arteries. Each of them presented a facade enlivened by Ionic pilasters which separated access doors to rooms of varying dimensions corresponding to "shops". These facades were preceded by a sidewalk which could ensure the passage of traffic from a street portico to that of the perpendicular street.

The entire tetrapylon with the surrounding square finds a striking parallel in the southern tetrapylon of Gerasa. Kraeling's publication dates the tetrakionia to the 2nd century, but the organization of the square to around the turn of the 3rd to the 4th century. At Bosra, too, the two elements seem to have been built successively and in the same order, but the tetrapylon itself cannot be earlier than the 4th century.

The gilded bronze capitals sought to strike the imagination of visitors and it is undoubtedly to this monument that the text of the Expositio totius mundi et gentium (38) refers: “Bostra… in qua publicum opus tetrapyli mirantur ”

In the 6th -7th century, the tetrapylon seems to have been abandoned with a large part of the neighboring area.

• from Dentzer et al. (2002:115-116)

The architectural analysis revealed a certain number of clues of relative chronology: insertion in the western portico of the propylaea providing access to the baths, repairs to the floors of the porticoes, evolution of construction techniques (in the northern section, ash-grey mortar already identified in late constructions of Bosra), narrowing of the doors, repointing of the mouldings decorating the doors and thresholds, traces of hasty reconstructions such as the courses of door frames put back in place in disorder.

On the north-south porticoed street, two surveys have allowed a closer look at the articulation of the different building elements and have provided some stratigraphic references to the history of the street. By combining the data obtained in these surveys with the observations made by Th. Fournet on the central thermal baths, we obtain the following chronological references:

Phase	Description
1	Occupation of the sector in the 1st–2nd century AD is attested by significant ceramic material, particularly Nabataean, but within the very limited area excavated. No building remains or street layout have been identified for this phase.
2	The first stage of the central baths (Pl. 8, 14), built on a different layout and orientation than surrounding monuments, predates the later development of the north-south street “nymphée-Omar.” Elements of the façade and a long eastern wall with niches have been preserved, along with at least one partition wall.
3	Under the Severans, a porticoed street with shops was installed along the east wall of the baths, slightly reoriented (2.5°) for regularization. A larger southern door led into the baths. Ionic capitals suggest the date, though stratigraphy does not yet confirm it. The street postdates the original spa, and differences in shop depths reflect this misalignment.
4	A second stage of the central baths introduced symmetry, including a northern propylaeum (door 2) and possibly a monumental southern entrance. Two of five consoles built into columns bear Severan inscriptions (late 2nd–early 3rd century).
5	Direct and indirect traces of an earthquake in the 4th century have been found. Could it be the same one that destroyed the large 4th-century mosaic building beneath the church built at the end of the 5th century?
6	A large-scale renovation of public facilities occurred in the 5th century, including carefully laid pavement in gray-beige limestone, a rare material in Bosra.
7	Shops were constructed or reconstructed between the northern entrance to the baths and the northern crossroads, possibly replacing older ones. Doorframes were reused and restored, indicating a phase of architectural recovery.
8	Monumental latrines were added north of the central baths, disrupting earlier layouts. These had direct street access and obscured the original northern insula plan.
9	The colonnaded street was profoundly altered as buildings encroached onto the porticoes and roadway. Reused materials, such as column drums and architraves, formed partition walls. These changes began in the 8th century and continued into the medieval period.
10	Further street transformation occurred north of the “al-Khidr–Fatima” intersection. Roadway narrowing and high sidewalks were installed to manage surface water, likely during the 12th-century mosque renovations. These installations remained in use into the traditional village period.
11	A final major transformation involved leveling and filling the street. An 1837 engraving shows the exedra bases visible before being buried by 2 meters of fill. Recent partial expropriation of houses built atop ancient remains also belongs to this phase.

• from Dentzer et al. (2002:141-142)

Apart from the existence of fortifications and a few traffic routes, we cannot imagine what the city looked like before the Nabataean and Roman periods. We know nothing about the Hellenistic city at the time of the Maccabees. Clearly, the city does not seem to have been created, originally, during a single subdivision operation distributing public and private space in an organized, and therefore geometric, manner, as in Greek or Roman foundations, where streets play a guiding role. Should we not rather suppose, at the origins of the agglomeration of Bosra, a development model of a different type, of which other traces have been found in the Near East, in a series of examples that go from the Bronze Age to the pre-industrial oriental city. These are agglomerations that develop from spatially separated housing cores, but end up joining together in a compact urban fabric, irregularly served by a few passageways and numerous dead ends, with few public spaces apart from a few sanctuaries. Did Bosra present this face?

Analysis of its plan, informed by information provided by excavations, reveals several episodes of urban planning involving limited areas. Is it possible to date them? To which initiatives should they be attributed? What are the driving forces behind this development?

The observations made must make us renounce attributing globally to the Roman era, as has often been done, since the travellers of the 19th century, the essential part of the urban organization still visible in Bosra, without distinguishing either the legacy of earlier phases or modifications that occurred during the RomanoByzantine period, with the exception of large monuments such as churches or later mosques.

• from Dentzer et al. (2002:142-143)

Although lithic and ceramic material attests to the occupation of the site from the Neolithic period to the Hellenistic period, it remains detached from any context, except in the Middle Bronze Age where some remains of buildings have appeared (southern rampart and an installation to the northwest of the city) at the same time as the outline of the tell. The situation has changed recently for the Nabataean phase where buildings are added to a sample of fine and common ceramics characteristic of the end of the 1st century BC to the end of the 1st century AD attest to the occupation of most of the space of the ancient city and link Bosra, in this entire period and in this period only, to the same culture as Petra. The Nabataean mark is also visible in the decoration of the monumental complex of the east and in other buildings of which fragments remain in different parts of the city. The restoration of the southern rampart, at the turn of the gives new weight to the passage of Damascius declaring that Bosra was a place fortified by the Arab kings against the Dionysians of Soada. It indicates not only a strategic project, but also the desire to create or confirm an urban entity. We are faced with a real operation of concerted urban planning with the creation of a new district to the east, centered on what was undoubtedly a large sanctuary, to which led the main street which could serve as a sacred way, and organized on the basis of a geometric module. This operation takes place in the second half of the 1st century AD.

The new contribution of recent years is the discovery of traces of a second urban planning project. In the city center, around from the crossroads of the nymphaeum, a certain number of constructions or walls integrated into recent houses present a common orientation offset from the network of streets (Pl. 16). The most important building is the first state of the southern baths, dated to the second half of the 2nd century, but excavations have revealed, to the west of these rooms and on the same orientation, remains of a construction older than this Roman monument, from the turn of the 1st century to the 2nd century , which could provide a new key to the interpretation of the sector. Dated by Nabataean ceramics and using the same tiles as the monumental complex in the eastern quarter, this building could already have been a thermal establishment, as suggested by the fragments of tubuli different from those of the Roman baths. These baths could be attributed to a Nabataean initiative and their orientation could indicate a desire to reorganize and give a monumental character to a larger area of the city center. It is, in fact, to the same orientation that the facade of the nymphaeum, walls of monumental character integrated into the Sheikh's house, several walls and one of the alleys of the new district opened to the east of the southern baths are linked, here again in a Nabataean context characterized by abundant ceramic material. This coherent orientation implies a project of extensive restructuring of the city center which has seen the beginning of its realization. The whole of this sector saw a strong Nabataean occupation as shown by the ceramics209. The common orientation is not governed by the main west-east street, already traced, since the Nabataean arch concludes its perspective, but perhaps, as in the eastern quarter, by a more important monument which could be located under the Sheikh's house. To the same complex, it is perhaps necessary to attach what could have been a small temple, located on the axis of the north-south street or in the immediate vicinity, of which elements (base blocks or decoration) only appear in reuse A Nabataean cult inscription, recently found in the southern palaestra very close to the central baths, mentions a priest. This program could have been initiated after the development of the eastern quarter, at the very end of the reign of Rabbel II or at the beginning of the provincial period. The decoration of the blocks discovered near the "temple" is clearly distinguished from the style of the monuments in the eastern quarter, as well as from that of the buildings of the second half of the 2nd century.

Th. Fournet found the characteristic orientation of the Nabataean quarter in certain alignments recorded in the modern land register or on the ground, at the eastern limit of the clearings of the praetorium quarter, but also to the east of the Sheikh's house. These observations lead us to question the western limit of this quarter. The Nabataean arch, however, must have marked its western limit at the time of its creation.

In this period of transition between the end of the reign of Rabbel II and the beginning of the provincial period, we still do not know whether this project of restructuring the city center should be attributed to the Nabataean power or to the new Roman authority.

• from Dentzer et al. (2002:143-144)

For G. Gualandi, the monumental urbanization of Bosra dated from the 2nd century and had not undergone any major change in the following centuries nor suffered from direct imperial intervention like Shahba in the middle of the 3rd century. He simply noted a shift in the center of gravity of the city towards the northeast in the late Roman and Byzantine periods with the construction of the church of Saints Sergius, Bacchus and Leontius.

S. Cerulli, who relied on new surveys, in particular of the main west-east street, explained the network of streets that marks the organization of the city, whose center he located at the crossroads of the nymphaeum, as the extension of the network of roads that lead to Bosra. For him, the development of the city accelerated in the Roman era, which favored the north-south axis.

It must be assumed that the authorities of the imperial phase gave up imposing on Bosra, which had become the capital of the new province of Arabia, a global or even partial geometric restructuring, and that they retained the layout of the old west-east circulation axis and probably other north-south routes. In their first monumental state, the southern baths were probably wedged onto an earlier contiguous construction and could be interpreted as an extension of an earlier establishment. In the same period, the cavea of the theater was built, on the edge of the district densely occupied in the Nabataean period and probably outside the rampart.

Urban policy in the Roman period was, on the one hand, directed towards the creation of new types of monuments, particularly for spectacles; on the other hand, it sought to give the city a more monumental character by using formulas developed in the Roman Near East.

The phase of Roman provincial construction apparently began in the time of Hadrian and developed under the Antonines with the construction of the theatre, the two large bath complexes, the exedra and, in the eastern quarter, the temple or sanctuary of Rome and Augustus. These programmes followed, with a certain delay, the creation of the Roman province of Arabia, with Bosra as its capital: the duration of their completion is difficult to estimate.

The period from the Severans to the reign of Philip the Arab represents a key stage in the monumental restructuring of the city. It is in this phase that the transformation of the major arteries of the urban network of Bosra into colonnaded streets takes place. We do not have sufficient inscriptions to measure the progress of the works as was done in Palmyra. At the same time, new monuments were created that are directly linked to them, such as the central arch, Bab al-Qandîl; others, more numerous, were completed or renovated, with the dual perspective of an expansion of the installations and an enrichment of the monumental decor (frons scaenae of the theater, southern and central baths, nymphaeum, exedra). The integration of previous programs into the new arrangement of porticoed streets is at the origin of many works. Furthermore, inscriptions attest to the strengthening of the city's defenses in the 3rd century.

One of the new contributions of the research of the French archaeological mission in South Syria is to show that major urban works continued during the 4th and 5th centuries. Apart from the northeastern quarter around the Church of Saints Sergius, Bacchus and Leontius, several phases of the Byzantine period have been identified on the construction sites in the east and the centre of the city. Irregular extensions of plan increase the capacity of the southern baths: the two palaestra date from this period. A large program of limestone paving has been noted on many points of the porticoes of the north-south and eastwest streets. Large-scale operations are attested by the inscriptions of three persons in charge of the exploitation of limestone quarries south of Qasr el-Hallabat, placed under the authority of a bishop who must be that of Bosra, to supply the and the 7th century. construction site of the 5th century th century also that the large square to the north of the cryptoporticus, probably the forum, was extensively reworked, rather than created.

A new phase of work took place in the 6th century with the construction of churches, but also repairs to the fortifications. The last restoration of the southern rampart can be attributed to this period. It should be noted that in this phase attention was paid to the maintenance and development of the monumental framework of the city. The western facade of the base supporting the cathedral, to the east of the Nabataean arch, is decorated with fountains placed in reused niches that serve to supply the population, but are also part of a spectacular urban setting.

It is currently more difficult to determine the date at which public facilities, including street porticoes, began to be encroached upon, gradually destroying this formula of monumental urban planning, which gave way to private exploitation of space. This development does not seem to have begun in Bosra before the Umayyad period, but became widespread in the medieval period. It should be noted that the encroachments that destroyed the formula of the porticoed street did not affect the commercial activity, which was one of its main functions at the beginning: the commercial use of at least some of the shops was simply extended to the space in front of them. It is likely that part of the porticoes was occupied by stalls or goods during the day from the Roman era. Thus, one passes seamlessly from the portico to the souk. The memory of this activity was preserved until the arrival of the first Western visitors with the name Khan ed-Dibs given to the thermal baths in the center.

• from Dentzer et al. (2002:144-145)

Chronological indications provided by inscriptions or by the analysis of architectural decoration lead to grouping together in the same phase certain large-scale works, monumental programs or urban planning operations. This leads us to question the factors that are at the origin of the changes.

In some privileged cases, the excavation was able to identify, in the succession of a series of states, the impulse which was at the origin of this or that construction – or destruction. Thus, in the eastern quarter, the brutal impact of blocks falling from a height on the mosaics of the first Christian monument which succeeded the Nabataean porticoed courtyard, can only be explained by a violent destruction of this building. It can be attributed without too much hesitation to an earthquake. The very extent of the destruction in the city can provide an argument in this direction, and comparing the sequences on often distant construction sites can be enlightening. The earthquake attested around 447 could be incriminated. Such episodes inevitably provoke, in a more or less short time, a reconstruction operation. One is tempted to explain by the same earthquake the very extensive repair of the floors of the porticos of the north-south street.

Human destruction is less easy to identify at Bosra, except for that, mentioned by inscriptions, of the campaign of Zenobia which destroyed or damaged the temple of Jupiter Hamon. If sensitive threats in the 3rd and 5th-6th centuries justify, as in Adraha, the strengthening of the city's defenses, there is no information on the precise circumstances of the conflict.

We can also assume positive factors pushing for the construction of new monuments: official desire of the imperial, provincial or municipal authority to promote urban development or gesture of a local patron wishing to impose his mark on the city.

Events creating or modifying the political status of the city could explain this type of intervention. While it is impossible to directly link a construction program to the creation of the Roman province of Arabia and the birth of Nea Traianè Bostra, the promotion of the city to the rank of Roman colony under Alexander Severus may have been an opportunity to restructure the appearance of the city by giving it a monumental unity thanks to colonnaded streets. Did the title of metropolis granted by Philip the Arab justify new works around the middle of the 3rd century? Each of these episodes must have resulted in ceremonies and festivals, and more lastingly in monumental constructions or redevelopments.

• from Dentzer et al. (2002:89)

Stratigraphic studies and excavations carried out at various points below the level of the cathedral at the end of the 5th century led to the identification, alongside minor episodes, of a first monumental restructuring of the sector datable to the course of the 4th century AD.

The new occupation level is located barely ten centimeters above the Nabataean level. This phase witnessed a reversal of the use of the space within the boundaries of the old porticoed courtyard. The old portico became an open space, made of beaten earth. What was the courtyard, however, was covered. Its floor is paved with white mosaics over most of the surface. The variable arrangement of the large tesserae differentiates the space, especially to the south.

In the northern half, there are figurative motifs that form polychrome carpets. Two blocks have also been identified, embedded in this surface, in which cavities have been dug to receive posts to delimit an area with a liturgical function.

Over the entire surface covered with mosaics that could be uncovered, no preserved wall was identified. There must have been one under the water conduit from the end of the 6th century. The organization of this space, which must have been covered, therefore eludes us. We can assume, however, that the building was already a church.

The mosaics, based on basic substructures (simple hedgehogs), bear very clear marks of impacts from blocks falling from parts of high points of the building. It was certainly an earthquake, perhaps the one attested in 447, which caused the violent destruction of the building. The materials used in this construction, already borrowed from earlier buildings of the Nabataean and Roman periods, were clearly carefully set aside to serve in the construction of the large church with a central plan, which can be dated to the end of the 5th century.

Only a few unusable pieces were left in the fill brought in for its construction. It was probably on this occasion that the blocks still integrated into the construction or found in scattered order were numbered. A temporary deposit of building materials, consisting in particular of column drums, was found under a vault supporting the floor of the following phase near the southwest apse.

• from Dentzer et al. (2002:89-92)

It is unknown whether the alterations to the monumental complex in the 4th century, which remained as close as possible to the Nabataean and Roman levels, influenced the organization of the district. However, after the earthquake, larger-scale works remodeled the space for the construction of a large centrally planned church. The land was raised by nearly 2.50 meters, with embankments to the north and vaulted substructures to the south, to create a horizontal foundation for a complex whose footprint reaches 65 by 70 meters (excluding the access staircase), in its current state of clearing. The west facade of this vast monumental base consists of a large wall decorated with niches borrowed from a building from the early imperial period. The center of this monumental program is a square block measuring 45 m on each side, within which is inscribed a circular shape consisting of a back wall, preceded by , preceded by a Corinthian portico that forms a sort of ambulatory.

Four large apses were set into the corners of the square. To the east, along the axis of the monument and more precisely between the two ends of the walls decorated with Nabataean half-columns, a rectangular room, the sanctuary, opened, ending with an apse set within a pentagonal envelope. Its entrance is marked by a threshold closed by a low chancel in a first state, then raised to form probably without a high chancel. The sanctuary was flanked by two adjoining rooms. The southern adjoining room ended externally with a flat chevet in which an apse was inscribed. A door gave direct access to the exterior, towards the east.

The method of covering the central circular space, with a span of more than thirty meters, poses a problem, in the absence of intermediate supports, which exist in the Church of Saints Sergius, Bacchus and Leontius. The latter has a similar plan with smaller dimensions (diameter of about 17 m). Four L-shaped pillars supported a central dome. It must be assumed that an annular vault made of light materials (including basalt scoria) covered the space between the circular portico and the back wall. This vault buttressed a wooden dome as, for example, at Qalaat Semân.

The church was accessed from the west, where a sort of propylaeum was arranged along its axis, the base of a column of which has been found. It is assumed that there was a monumental staircase that bridged the difference in level between the cathedral and the Nabataean arch, the floor of which was at least 4 m lower, even in the early Christian period.

To the central square mass of the building is connected to the south another rectangular building of 12 x 20 m. It is a second church with a basilica plan, with three naves, oriented. To the east, a slightly horseshoe-shaped apse is flanked by two annexed rooms inscribed in a flat chevet. On its southern flank opens, over almost its entire length, a second apse, of much larger dimensions (10 x 6 m). This apse, in a semicircle exceeded on the inside, is part of an envelope whose exterior draws a seven-sided polygon. The west face of this church, in which three doors opened, was undoubtedly preceded by a portico on two floors. The corbels which supported it at its north and south ends are preserved. The portico was located at the level of the church and must therefore have been accessible, from the west, via a stairway.

The northern boundary of this church is tangent to the circular space which constitutes the center of the large church. How were these two spaces articulated? Was the annex church entirely open to the north with simply pillars to support its roof? Two pilasters are preserved on either side at the ends of its northern boundary. In a first state, these two pillars were arranged obliquely relative to the rectangular plan of the annex. In this phase, the northern boundary of the church was therefore not linear. An oblique wall was directed towards two other supports. The annex church then fit into the annular plan of the space surrounding the central colonnade. In a second state, the pillars were offset and a foundation supplement was added to create a straight alignment that intersects the circular plan of the annular space. In this phase, the church, which can then be assumed to be closed, seems to take on a more autonomous character.

One is tempted to attribute these major modifications to a major phase of work that, following the re-examination of the ceramic material by P.-M. Blanc and D. Pieri, can be dated to the end of the 6th century. It is perhaps explained by a new earthquake attested in 551⁴⁹. It is in this phase that we must undoubtedly place significant alterations in the sanctuary: a synthronon was built at the same time as the threshold was modified and the low chancel was replaced by a high chancel. At the same time, a large opus caementicium pipe was installed, which crosses the church and the choir and feeds the fountains on the facade.

In this annex church, the apse which opens towards the south would best be explained as a baptistery. In its first state, it was located in the basement relative to the level of the main church, which rests, on this side, on vaulted substructures. It was accessible by three doors opening directly onto the space surrounding the large church and dominated by it. This lower area must itself have been surrounded by an enclosure and formed part of the church space: it gave, in fact, direct access to the baptistery. In 2000, very limited traces of the baptismal font were found, which had suffered serious damage: its space was, in fact, partially reoccupied by cellars until the modern period.

A major transformation took place, probably in the second phase of work which has just been mentioned: the baptistery was moved upstairs on the main circulation level of the ecclesiastical complex. To this end, corbels were embedded in the existing semicircular wall and two transverse walls were built.

The scale and overall organization of the architectural program, which combines the centrally planned main church with an annex church with an apse to the south, which can be interpreted as a baptistery, and a second room or annex chapel to the north, leads to the interpretation of this ensemble as an episcopal group that constituted one of the monumental centers of the city and dominated this district in the early Christian period.

In this coherent context, one may wonder whether the so-called Trajan's Palace with its triconch hall on floor should not be interpreted as an episcopal palace . A similar plan is found in a large house described by HC Butler to the east of the church of Saints Sergius, Bacchus and Leontius.

• from Dentzer et al. (2002:104-105)

We still do not know what the west-east street looked like at its origins. A survey of a very limited area, carried out in 1993-1994 north of the facade of the southern baths, revealed material from older occupation phases dating back to the Neolithic. Traces of a first level of circulation date back to a period between 1 and 20 of our era in a Nabataean context (unpublished ceramics and coins). For the construction work of the first state of the baths, a terminus a quo was set around 150 AD. For the installation of the stylobate of the first portico of the street, ceramics characteristic of the 3rd century confirm the date suggested by the style of the Ionic capitals. In this phase, the boundary wall of the baths was cut down obliquely, over a length of at least 11.50 m, to make way for the portico, which is approximately four meters wide. The northern facade is thus partially aligned with the current layout of the street. The portico was probably not completed with shops at this stage.

A new stage of transformations has been clearly identified in the 5th century, according to coins. The stylobate was reworked and relaid. A gray-beige limestone floor was laid under the portico and on the foundation of the first stylobate and abuts the timbered wall.

During the 6th century, the northern boundary wall of the baths was finally razed and replaced by a new, recessed wall to gain the space needed to install shops to complement the portico. The beginnings of dividing walls are visible in the northern facing of the wall.

On the other hand, the street paving was re-laid, raised by about thirty centimeters, probably not before the 6th century. The same operation must have been carried out at the neighboring crossroads. The foundations of shops from the Byzantine phase are preserved: remains of fish from the Red Sea have been found there, as well as traces of bone crafts.

In the Ayyubid-Mamluk phase, we observe encroachments of houses on half of the area of the porticoes whose paving had been systematically recovered.

Finally, at a much higher level, modern houses probably occupied the entire surface of the portico in this sector.

• from Dentzer et al. (2002:105)

A fragment of a polychrome mosaic carpet with geometric patterns (Solomon's knot, leaves, V-shaped checkerboard), dated to the mid-4th century, was studied in 2001 in the northern portico of the east-west street, south of the central baths (Khan ed-Dibs). To the north, a second carpet partially covers the first, five centimeters higher. It could mark a shop entrance or thermal baths located right at this location.

The mosaic pavement was certainly replaced, after destruction due to an earthquake in the second half of the 5th century, by the gray limestone tile floor still visible at the eastern end of this portico. This repair is accompanied by a raising of about 25 centimeters, in connection with a new reconstruction of the shops, and probably also of the portico, as was noted during the excavation on the south side of the street.

Marble slabs held together by copper alloy staples decorated the shop fronts. They were abandoned during a reconstruction carried out in the second half of the 5th century.

In medieval times the last limestone floor was removed and reclaimed and new buildings were built on the site of the portico.

• from Dentzer et al. (2002:105,108)

Further west, on the main street, the French archaeological mission in South Syria resumed in 1998-1999 the study of the cryptoporticus and tetrapylon excavated by Suleiman al-Mukdad between 1960 and 1980. The study of this sector was resumed with cleaning and some surveys. The work was extended to the area located to the north of this complex. The results of this research having been published recently, we will limit ourselves here to summarizing the main conclusions.

The cryptoporticus is limited to a simple gallery supporting the northern portico of the main west-east street. 108 m long and 5 m wide, it was terminated at both ends by an apse. The one at the western end, partially destroyed when the cryptoporticus was shortened to make way for the tetrapylon, was discovered and excavated in 1999. While the walls are built in courses of basalt blocks coated with pink mortar to the north, the vault is in opus caementicium using volcanic scoria. To the south, the gallery is lit by thirty-four basement windows that flare inwards. The southern interior wall was also punctuated by niches that are found in the north wall. It is from this side that the three accesses to the cryptoporticus open.

Phases prior to the cryptoporticus: a survey carried out to the west of the wall enclosing the western apse of the cryptoporticus provided a significant sample of residual ceramic material from the Bronze Age to the 2nd century AD.

Construction of the cryptoporticus: additional examinations carried out by P.-M. Blanc in 2001, during cleaning work carried out by the Bosra Antiquities Department, established the date of construction of the cryptoporticus at the end of the 2nd century or the beginning of the 3rd century. In this first phase, the floor was covered by paving or by a coating of sandstone pebbles set in a hydraulic mortar. Steep three-step staircases led down to the circulation level in line with the three access bays. The upper portico and the first row of shops opening onto it were installed in the same program.

Phase of monumental transformations: behind the street portico, a monumental entrance was created with three doors that give access to the space located to the north of the cryptoporticus, but two meters lower. From this phase, a second row of shops seems to have doubled those of the main street. Their split-level arrangement provided a clever solution to the problem of the difference in level between the street and the square to the north.

Installation of ramps: Access to the cryptoporticus, previously provided by three steep staircases, was replaced in the 4th century by the installation of long, gently sloping ramps, which particularly facilitated the handling of goods. During this phase, the cryptoporticus was backfilled to a depth of about sixty centimeters. Does this profound transformation correspond to a change in the status of this space?

Construction of the tetrapylon: at the end of the 3rd or beginning of the 4th century, when the decision was taken to build the circular square of the tetrapylon, the western end of the cryptoporticus was amputated by ten meters and closed by a wall equipped with a door.

• from Dentzer et al. (2002:108)

The tetrapylon, located in the center of a small round square surrounded by a gutter and onto which shops opened, was built during the 4th century, before the development of the square itself. The tetrapylon, or rather the tetrakionion, is currently reduced to four square bases measuring 5.10 m on each side. On each of them were placed four columns crowned by capitals whose cores are preserved. Roughly grooved, they must have been covered with bronze plates, probably gilded.

The circular square, which is distinguished from the street by a pavement of small square slabs, arranged at 45°, is divided into four quadrants by the two intersecting arteries. Each of them presented a facade enlivened by Ionic pilasters which separated access doors to rooms of varying dimensions corresponding to "shops". These facades were preceded by a sidewalk which could ensure the passage of traffic from a street portico to that of the perpendicular street.

The entire tetrapylon with the surrounding square finds a striking parallel in the southern tetrapylon of Gerasa. Kraeling's publication dates the tetrakionia to the 2nd century, but the organization of the square to around the turn of the 3rd to the 4th century. At Bosra, too, the two elements seem to have been built successively and in the same order, but the tetrapylon itself cannot be earlier than the 4th century.

The gilded bronze capitals sought to strike the imagination of visitors and it is undoubtedly to this monument that the text of the Expositio totius mundi et gentium (38) refers: “Bostra… in qua publicum opus tetrapyli mirantur ”

In the 6th -7th century, the tetrapylon seems to have been abandoned with a large part of the neighboring area.

• from Dentzer et al. (2002:115-116)

The architectural analysis revealed a certain number of clues of relative chronology: insertion in the western portico of the propylaea providing access to the baths, repairs to the floors of the porticoes, evolution of construction techniques (in the northern section, ash-grey mortar already identified in late constructions of Bosra), narrowing of the doors, repointing of the mouldings decorating the doors and thresholds, traces of hasty reconstructions such as the courses of door frames put back in place in disorder.

On the north-south porticoed street, two surveys have allowed a closer look at the articulation of the different building elements and have provided some stratigraphic references to the history of the street. By combining the data obtained in these surveys with the observations made by Th. Fournet on the central thermal baths, we obtain the following chronological references:

Phase	Description
1	Occupation of the sector in the 1st–2nd century AD is attested by significant ceramic material, particularly Nabataean, but within the very limited area excavated. No building remains or street layout have been identified for this phase.
2	The first stage of the central baths (Pl. 8, 14), built on a different layout and orientation than surrounding monuments, predates the later development of the north-south street “nymphée-Omar.” Elements of the façade and a long eastern wall with niches have been preserved, along with at least one partition wall.
3	Under the Severans, a porticoed street with shops was installed along the east wall of the baths, slightly reoriented (2.5°) for regularization. A larger southern door led into the baths. Ionic capitals suggest the date, though stratigraphy does not yet confirm it. The street postdates the original spa, and differences in shop depths reflect this misalignment.
4	A second stage of the central baths introduced symmetry, including a northern propylaeum (door 2) and possibly a monumental southern entrance. Two of five consoles built into columns bear Severan inscriptions (late 2nd–early 3rd century).
5	Direct and indirect traces of an earthquake in the 4th century have been found. Could it be the same one that destroyed the large 4th-century mosaic building beneath the church built at the end of the 5th century?
6	A large-scale renovation of public facilities occurred in the 5th century, including carefully laid pavement in gray-beige limestone, a rare material in Bosra.
7	Shops were constructed or reconstructed between the northern entrance to the baths and the northern crossroads, possibly replacing older ones. Doorframes were reused and restored, indicating a phase of architectural recovery.
8	Monumental latrines were added north of the central baths, disrupting earlier layouts. These had direct street access and obscured the original northern insula plan.
9	The colonnaded street was profoundly altered as buildings encroached onto the porticoes and roadway. Reused materials, such as column drums and architraves, formed partition walls. These changes began in the 8th century and continued into the medieval period.
10	Further street transformation occurred north of the “al-Khidr–Fatima” intersection. Roadway narrowing and high sidewalks were installed to manage surface water, likely during the 12th-century mosque renovations. These installations remained in use into the traditional village period.
11	A final major transformation involved leveling and filling the street. An 1837 engraving shows the exedra bases visible before being buried by 2 meters of fill. Recent partial expropriation of houses built atop ancient remains also belongs to this phase.

• from Dentzer et al. (2002:141-142)

Apart from the existence of fortifications and a few traffic routes, we cannot imagine what the city looked like before the Nabataean and Roman periods. We know nothing about the Hellenistic city at the time of the Maccabees. Clearly, the city does not seem to have been created, originally, during a single subdivision operation distributing public and private space in an organized, and therefore geometric, manner, as in Greek or Roman foundations, where streets play a guiding role. Should we not rather suppose, at the origins of the agglomeration of Bosra, a development model of a different type, of which other traces have been found in the Near East, in a series of examples that go from the Bronze Age to the pre-industrial oriental city. These are agglomerations that develop from spatially separated housing cores, but end up joining together in a compact urban fabric, irregularly served by a few passageways and numerous dead ends, with few public spaces apart from a few sanctuaries. Did Bosra present this face?

Analysis of its plan, informed by information provided by excavations, reveals several episodes of urban planning involving limited areas. Is it possible to date them? To which initiatives should they be attributed? What are the driving forces behind this development?

The observations made must make us renounce attributing globally to the Roman era, as has often been done, since the travellers of the 19th century, the essential part of the urban organization still visible in Bosra, without distinguishing either the legacy of earlier phases or modifications that occurred during the RomanoByzantine period, with the exception of large monuments such as churches or later mosques.

• from Dentzer et al. (2002:142-143)

Although lithic and ceramic material attests to the occupation of the site from the Neolithic period to the Hellenistic period, it remains detached from any context, except in the Middle Bronze Age where some remains of buildings have appeared (southern rampart and an installation to the northwest of the city) at the same time as the outline of the tell. The situation has changed recently for the Nabataean phase where buildings are added to a sample of fine and common ceramics characteristic of the end of the 1st century BC to the end of the 1st century AD attest to the occupation of most of the space of the ancient city and link Bosra, in this entire period and in this period only, to the same culture as Petra. The Nabataean mark is also visible in the decoration of the monumental complex of the east and in other buildings of which fragments remain in different parts of the city. The restoration of the southern rampart, at the turn of the gives new weight to the passage of Damascius declaring that Bosra was a place fortified by the Arab kings against the Dionysians of Soada. It indicates not only a strategic project, but also the desire to create or confirm an urban entity. We are faced with a real operation of concerted urban planning with the creation of a new district to the east, centered on what was undoubtedly a large sanctuary, to which led the main street which could serve as a sacred way, and organized on the basis of a geometric module. This operation takes place in the second half of the 1st century AD.

The new contribution of recent years is the discovery of traces of a second urban planning project. In the city center, around from the crossroads of the nymphaeum, a certain number of constructions or walls integrated into recent houses present a common orientation offset from the network of streets (Pl. 16). The most important building is the first state of the southern baths, dated to the second half of the 2nd century, but excavations have revealed, to the west of these rooms and on the same orientation, remains of a construction older than this Roman monument, from the turn of the 1st century to the 2nd century , which could provide a new key to the interpretation of the sector. Dated by Nabataean ceramics and using the same tiles as the monumental complex in the eastern quarter, this building could already have been a thermal establishment, as suggested by the fragments of tubuli different from those of the Roman baths. These baths could be attributed to a Nabataean initiative and their orientation could indicate a desire to reorganize and give a monumental character to a larger area of the city center. It is, in fact, to the same orientation that the facade of the nymphaeum, walls of monumental character integrated into the Sheikh's house, several walls and one of the alleys of the new district opened to the east of the southern baths are linked, here again in a Nabataean context characterized by abundant ceramic material. This coherent orientation implies a project of extensive restructuring of the city center which has seen the beginning of its realization. The whole of this sector saw a strong Nabataean occupation as shown by the ceramics209. The common orientation is not governed by the main west-east street, already traced, since the Nabataean arch concludes its perspective, but perhaps, as in the eastern quarter, by a more important monument which could be located under the Sheikh's house. To the same complex, it is perhaps necessary to attach what could have been a small temple, located on the axis of the north-south street or in the immediate vicinity, of which elements (base blocks or decoration) only appear in reuse A Nabataean cult inscription, recently found in the southern palaestra very close to the central baths, mentions a priest. This program could have been initiated after the development of the eastern quarter, at the very end of the reign of Rabbel II or at the beginning of the provincial period. The decoration of the blocks discovered near the "temple" is clearly distinguished from the style of the monuments in the eastern quarter, as well as from that of the buildings of the second half of the 2nd century.

Th. Fournet found the characteristic orientation of the Nabataean quarter in certain alignments recorded in the modern land register or on the ground, at the eastern limit of the clearings of the praetorium quarter, but also to the east of the Sheikh's house. These observations lead us to question the western limit of this quarter. The Nabataean arch, however, must have marked its western limit at the time of its creation.

In this period of transition between the end of the reign of Rabbel II and the beginning of the provincial period, we still do not know whether this project of restructuring the city center should be attributed to the Nabataean power or to the new Roman authority.

• from Dentzer et al. (2002:143-144)

For G. Gualandi, the monumental urbanization of Bosra dated from the 2nd century and had not undergone any major change in the following centuries nor suffered from direct imperial intervention like Shahba in the middle of the 3rd century. He simply noted a shift in the center of gravity of the city towards the northeast in the late Roman and Byzantine periods with the construction of the church of Saints Sergius, Bacchus and Leontius.

S. Cerulli, who relied on new surveys, in particular of the main west-east street, explained the network of streets that marks the organization of the city, whose center he located at the crossroads of the nymphaeum, as the extension of the network of roads that lead to Bosra. For him, the development of the city accelerated in the Roman era, which favored the north-south axis.

It must be assumed that the authorities of the imperial phase gave up imposing on Bosra, which had become the capital of the new province of Arabia, a global or even partial geometric restructuring, and that they retained the layout of the old west-east circulation axis and probably other north-south routes. In their first monumental state, the southern baths were probably wedged onto an earlier contiguous construction and could be interpreted as an extension of an earlier establishment. In the same period, the cavea of the theater was built, on the edge of the district densely occupied in the Nabataean period and probably outside the rampart.

Urban policy in the Roman period was, on the one hand, directed towards the creation of new types of monuments, particularly for spectacles; on the other hand, it sought to give the city a more monumental character by using formulas developed in the Roman Near East.

The phase of Roman provincial construction apparently began in the time of Hadrian and developed under the Antonines with the construction of the theatre, the two large bath complexes, the exedra and, in the eastern quarter, the temple or sanctuary of Rome and Augustus. These programmes followed, with a certain delay, the creation of the Roman province of Arabia, with Bosra as its capital: the duration of their completion is difficult to estimate.

The period from the Severans to the reign of Philip the Arab represents a key stage in the monumental restructuring of the city. It is in this phase that the transformation of the major arteries of the urban network of Bosra into colonnaded streets takes place. We do not have sufficient inscriptions to measure the progress of the works as was done in Palmyra. At the same time, new monuments were created that are directly linked to them, such as the central arch, Bab al-Qandîl; others, more numerous, were completed or renovated, with the dual perspective of an expansion of the installations and an enrichment of the monumental decor (frons scaenae of the theater, southern and central baths, nymphaeum, exedra). The integration of previous programs into the new arrangement of porticoed streets is at the origin of many works. Furthermore, inscriptions attest to the strengthening of the city's defenses in the 3rd century.

One of the new contributions of the research of the French archaeological mission in South Syria is to show that major urban works continued during the 4th and 5th centuries. Apart from the northeastern quarter around the Church of Saints Sergius, Bacchus and Leontius, several phases of the Byzantine period have been identified on the construction sites in the east and the centre of the city. Irregular extensions of plan increase the capacity of the southern baths: the two palaestra date from this period. A large program of limestone paving has been noted on many points of the porticoes of the north-south and eastwest streets. Large-scale operations are attested by the inscriptions of three persons in charge of the exploitation of limestone quarries south of Qasr el-Hallabat, placed under the authority of a bishop who must be that of Bosra, to supply the and the 7th century. construction site of the 5th century th century also that the large square to the north of the cryptoporticus, probably the forum, was extensively reworked, rather than created.

A new phase of work took place in the 6th century with the construction of churches, but also repairs to the fortifications. The last restoration of the southern rampart can be attributed to this period. It should be noted that in this phase attention was paid to the maintenance and development of the monumental framework of the city. The western facade of the base supporting the cathedral, to the east of the Nabataean arch, is decorated with fountains placed in reused niches that serve to supply the population, but are also part of a spectacular urban setting.

It is currently more difficult to determine the date at which public facilities, including street porticoes, began to be encroached upon, gradually destroying this formula of monumental urban planning, which gave way to private exploitation of space. This development does not seem to have begun in Bosra before the Umayyad period, but became widespread in the medieval period. It should be noted that the encroachments that destroyed the formula of the porticoed street did not affect the commercial activity, which was one of its main functions at the beginning: the commercial use of at least some of the shops was simply extended to the space in front of them. It is likely that part of the porticoes was occupied by stalls or goods during the day from the Roman era. Thus, one passes seamlessly from the portico to the souk. The memory of this activity was preserved until the arrival of the first Western visitors with the name Khan ed-Dibs given to the thermal baths in the center.

• from Dentzer et al. (2002:144-145)

Chronological indications provided by inscriptions or by the analysis of architectural decoration lead to grouping together in the same phase certain large-scale works, monumental programs or urban planning operations. This leads us to question the factors that are at the origin of the changes.

In some privileged cases, the excavation was able to identify, in the succession of a series of states, the impulse which was at the origin of this or that construction – or destruction. Thus, in the eastern quarter, the brutal impact of blocks falling from a height on the mosaics of the first Christian monument which succeeded the Nabataean porticoed courtyard, can only be explained by a violent destruction of this building. It can be attributed without too much hesitation to an earthquake. The very extent of the destruction in the city can provide an argument in this direction, and comparing the sequences on often distant construction sites can be enlightening. The earthquake attested around 447 could be incriminated. Such episodes inevitably provoke, in a more or less short time, a reconstruction operation. One is tempted to explain by the same earthquake the very extensive repair of the floors of the porticos of the north-south street.

Human destruction is less easy to identify at Bosra, except for that, mentioned by inscriptions, of the campaign of Zenobia which destroyed or damaged the temple of Jupiter Hamon. If sensitive threats in the 3rd and 5th-6th centuries justify, as in Adraha, the strengthening of the city's defenses, there is no information on the precise circumstances of the conflict.

We can also assume positive factors pushing for the construction of new monuments: official desire of the imperial, provincial or municipal authority to promote urban development or gesture of a local patron wishing to impose his mark on the city.

Events creating or modifying the political status of the city could explain this type of intervention. While it is impossible to directly link a construction program to the creation of the Roman province of Arabia and the birth of Nea Traianè Bostra, the promotion of the city to the rank of Roman colony under Alexander Severus may have been an opportunity to restructure the appearance of the city by giving it a monumental unity thanks to colonnaded streets. Did the title of metropolis granted by Philip the Arab justify new works around the middle of the 3rd century? Each of these episodes must have resulted in ceremonies and festivals, and more lastingly in monumental constructions or redevelopments.