L’analyse de l’enceinte a permis d’achever la première phase d’un projet de recherche visant l’étude du camp légionnaire de Mayence. L’identification d’un plus petit camp antérieur d’environ 34 ha, que l’on peut dater de la phase d’occupation précoce, est l’un des résultats les plus importants. Certains indices plaident pour un début d’occupation en 17 av. J.‑C. La présence de troupes permanentes, sous la forme de deux légions, n’est cependant envisageable qu’à partir de 17 apr. J.‑C. avec l’établissement d’un camp de 37 ha. On peut distinguer au total trois phases dans l’enceinte de terre et bois. L’aménagement en dur n’a sans doute pas été entrepris sous Vespasien, comme on l’avait pensé jusqu’ici, mais après les guerres contre les Chattes, sous Domitien, et dans le cadre de la création d’une province. L’enceinte de pierre présente également trois phases de construction, sans toutefois avoir été reconstruite. Un nouveau mur est adossé contre sa façade probablement dans le deuxième quart du ive s. apr. J.‑C. Enfin, la construction d’une nouvelle enceinte urbaine pour la ville de Mayence vers 369/370-375 apr. J.‑C. marque la fin du camp légionnaire.
New evaluation of the Mainz/Mogontiacum legionary fortress’ architectural defences includes a total of 25 sites covering some 150 years of research history. The investigation emphasizes consideration of the excavations that took place following the publication of D. Baatz’s 1962 monograph on the legionary fortress. This work may be expected to provide new information on the evolution of the defensive systems, which remained unknown for extended periods. Subsequently, and for the first time since 1962, a complete plan of the legionary fortress has been developed. This digital cadastral plan should provide a foundation for future investigations. The plan was constructed using QGIS (a geographic information system) and by geo-referencing all existing excavation plans, as well as digitally redrawing all structures, and combining these with a database. All available data on the structures are included, as well as the catalogue numbers of the features within the present work, alongside the corresponding relevant finds, included as image files. The maps generated herein are based upon this complete cadastre. Site evaluation resulted in the division of the earth-and-timber fortress and the stone fortress into three distinct structural phases for each. The timber phase 1 of the fortress’ defensive circuit consisted of an earth-and-timber rampart fronted by a V-shaped ditch (defensive ditch I). The ground-plan of this first fortress is different and smaller to what was previously known, and the porta praetoria was situated some 60 m south of the later fortress’ corresponding structure. The connection of the back rampart with the long north-western side remains unclear, but its course was identified for the first time for the early fortress. The north-western side (FS 12b), as well as the rear rampart (FS 03, 06) lie within the scope of the later stone fortress. Based on the defensive ditch (FS 04), it was possible to locate the latter structure within the limits of the later retentura. Its position therein means that the connection between the rear rampart and the fortress’ south-eastern side must have had an unusual outline. Future excavations in the area between Helmholtzweg and Langenbeckstraße should help to clarify the progression of the earth-and-timber fortress’ defences. Furthermore, the early buildings in this site likely lie outside the boundaries of the fortress and belong to the Augustan canabae legionis. Based on the new line of the defences on the rear rampart and along the long northern side, the maximum size of the fortress is reduced to 34 ha, compared with the previously estimated 37 ha. Chronologically, the earliest fortress was erected in connection with the Augustan occupation phase, and a date of 13/12 BC was presumed. More recent findings, however, point to an earlier date, around 17/16 BC. Prior to the end of phase 1, several repair efforts on the earth-and-timber rampart were identified. This would have taken place before the defences on the rear rampart (FS 06), as well as possibly on the northern side (FS 12b), were dismantled and advanced by some 30 m to the position of the later stone fortress. Archaeological finds provided dates for these events, placing them during the late-Augustan or early-Tiberian period. These changes fundamentally altered the fortress at Mainz. Perhaps the increase in size coincided with a completely new concept for large parts of the fortress, including the relocation of the porta praetoria to the main gate site of the later stone fortress. As a result, the fortress’ main axis, the via praetoria, had to be moved, which subsequently resulted in changes to the internal plan of the fortress. This radical new alignment was likely connected with the restructuring of the Rhine frontier by Emperor Tiberius in AD 17, which also coincided with a renunciation of the hitherto offensive policy in Germania. De facto, it is only now that we can assume that the two legions XIV Gemina and XVI Gallica were permanently stationed here. During the occupation period, with its flexibly operating units, there is no historical evidence for a double legionary fortress, nor is one to be expected. Such an edifice would have been conceived from the outset for permanently garrisoned units, the likes of which are known from the Tiberian period onwards. Timber phase 2 of the earth-and-timber defensive circuit is the poorest documented of the archaeological features. On the new rear rampart, which had been moved forward, a new earth-and-timber rampart was constructed, the posts of which were also set into parallel running trenches. To date, it was only possible to locate the presence of this new rampart in the north-western section of the rear rampart. The course of the fortress’ northern side during the second phase remains unknown, but was probably also moved forward. On the south-eastern long side, no changes could be discerned, and in this area, the earth-and-timber rampart continued into phase 2. No corresponding ditch (defensive ditch II) could be identified. It was likely destroyed during construction work for the later V-shaped ditches. For timber phase 3, a new ditch (defensive ditch III) was laid out some 20-25 m in front of the earth-and-timber rampart, which was still standing. It cut several features of Claudian date in the canabae legionis, including a drain, the backfill of which is coin-dated to AD 41. The structural alterations are connected with the arrival of legio IV Macedonica and legio XXII Primigenia. It remains unclear whether there was a subsequent, second defensive ditch. No changes to the earth-and-timber rampart can be identified archaeologically during phase 3 and the north-western long side is also the only one for which the course remains unknown for this phase. Clearly, the modifications to the fortification during this period were incorporated into an overall building programme that reveals a comprehensive master plan. For instance, within the canabae legionis numerous groups of open pits were filled in and a ring road was built around the fortress. The restructuring of the road, which left the porta principalis dextra towards the fort located at present-day Mainz-Weisenau, transforming it into a representative funerary street, is also dated to around the mid-1st c. AD, and was likely part of this reorganisation. The same applies to the road leaving the porta principalis sinistra towards Bingen. The end of phase 3 was initiated by extensive building activities. The beginning of stone phase 1 is marked by the backfilling of defensive ditch III, dated, at the earliest, by coins to the late-Vespasianic period, but more likely to the Domitianic period, following the Chattan wars of AD 84/85. The masonry defensive wall (wall I), measuring 1,70-2,30 m wide at its foundation, was fronted by a V-shaped ditch (defensive ditch IV). During the same period the masonry aqueduct was also constructed. With stone phase I, the fortress’ northern side can, for the first time, be identified. It runs along the slope of the Linsenberg and is thus as much as 70 m further north than had been presumed, meaning that the fortress was larger than previously thought, roughly 37 ha, rather than some 35-36 ha. On this side of the fortress no defensive ditch existed at any time during the stone phases. In stone phase 2 the existing defensive ditch was filled in and a new one (Va) was dug. These measures may have occurred in the middle or in the second half of the 2nd c. AD. A reconstruction of the fortress wall did not accompany this work. Consequently, the garrisoning of legio XXII Primigenia pia fidelis in AD 97, which remained there until its end, had no demonstrable effect on the fortress’ defensive circuit. Moreover, the archaeological findings provide no evidence for a reduction in the size of the fortress, which had until then been used by two legions. The defensive ditch excavated at the beginning of stone phase 2 was renewed at least twice. These efforts are well evidenced by two V-shaped bases in the ditch (defensive ditch Vb-c), demonstrating the clearing of sedimented material from within the ditch. They should not be interpreted as separate phases of the defence system. Stone phase 3 is expressed by a wall-facing, measuring up to 1 m wide, fronting the existing rampart and can be linked to a supra-regional construction programme to reinforce the military bases along the Rhine frontier. It is possible that these operations were preceded by the backfilling of defensive ditch Vc. Date attribution using the archaeological material is difficult, but in conjunction with the results of the present study, stone phase 3 can be narrowed down to the second quarter of the 4th c. AD or shortly thereafter. Presumably, a U-shaped ditch was installed in front of the fortress’ reinforced wall, which, to date, could only be identified at the south-western corner of the stronghold. In contrast, along other segments, defensive ditch Vc was infilled to form a hollow and may have remained in use as a defensive obstacle. Much like efforts to date the beginning of stone phase 3, an end date for the legionary fortress remains uncertain. It is likely that legio XXII Primigenia was heavily decimated during the usurpation of Magnentius in AD 350-353, as well as during the Germanic invasions of AD 352 and 355, with only a few units remaining in Mainz. The final abandonment of the legionary fortress likely took place around AD 364/370, i.e. shortly before the construction of the city wall around AD 369/370-375. Presumably the measures that led to the abandonment of the legionary fortress, the dissolution of the remaining units of legio XXII Primigenia and the reduction of the city wall were part of a decision enacted during the rule of Valentinian I, as part of an effort to strengthen the Rhine frontier. The construction of Mainz’s second city wall, which traversed the legionary fortress from southeast to northwest, was preceded by extensive demolition and levelling operations of all fortress structures, in order to obtain building materials, as well as to establish a glacis in front of the city wall. To what extent the area between the city wall and the old praetorial front served as a reduced space for a garrison of limitanei units of the late-4th c. AD remains unclear. (Traduction : Clive Bridger, Xanten.)
