La datation des aqueducs lyonnais constitue un sujet régulièrement abordé par la littérature scientifique. Cependant, les recherches se sont principalement concentrées sur l’aqueduc du Gier, sans trop s’intéresser aux trois autres grands aqueducs romains de la Brévenne, de l’Yzeron et du Mont d’Or. Ces dernières décennies, les opérations d’archéologie préventive, ainsi que les données issues de l’archéométrie (datation radiocarbone, archéomagnétisme et dendrochronologie) ont permis d’avancer dans la compréhension de la chronologie de ces ouvrages. Les découvertes épigraphiques ainsi que la réflexion sur le rapport entre les aqueducs et le développement urbain de la colonie ont également apporté de nouveaux éléments à ce dossier.Cette étude de synthèse permet de proposer des datations fiables pour les aqueducs de la Brévenne (milieu du ier s. apr. J.-C.) et du Gier (première moitié du iie s. apr. J.-C.), ainsi qu’une date de construction possible pour celui de l’Yzeron autour du changement d’ère. La datation du Mont d’Or, dont on a peut-être un terminus post quem, n’est pas encore étayée par l’archéologie.
The dating of the Roman aqueducts which served Colline de Fourvière, site of the Roman colony of Lugdunum, is a subject regularly addressed in the scientific literature. However, the main focus of research has been the Gier aqueduct, with little attention paid to the three other major Roman aqueducts –Brévenne, Yzeron and Mont d'Or. In recent decades, preventive archaeology operations as well as data from archaeometry (radiocarbon dating, archaeomagnetism and dendrochronology) have enabled progress to be made in understanding the chronology of these works. Epigraphic discoveries as well as reflection on the relationship between the aqueducts and the urban development of the colony have also contributed new elements to this field. As a result, a global analysis was needed.For the Mont d’Or aqueduct, a single excavation conducted around 30 years ago near the entry into Lyon (52-54 rue de la Favorite) brought to light a sherd of Drag 15/17 type terra sigillata pottery in the canal foundation trench. The dating of this sherd initially suggested a post 20 BC chronology for the construction of aqueduct. However, a new ceramological study has identified it as a terra sigillata italica dish edge of undetermined origin related to the Haltern 2 form, and dates it between 15 BC and 15/20 AD. This would give a terminus post quem for the construction of the Mont d'Or aqueduct much earlier than that initially envisaged.For the Yzeron aqueduct, beyond the radiocarbon dates of several charcoals which give a very broad chronology, two recent excavations have offered new data. During a study of the elevations of the two piers of the Craponne double siphon, numerous charcoals were found in the masonry: radiocarbon dating places them within the period between the 2nd century BC and the 2nd century AD, with an overlapping range corresponding to the change of era. In 2020, a preventive archaeology excavation at 73-77 avenue du Point-du-Jour (Lyon 5th arrondissement) uncovered two wooden pipes located in a ditch that follows the presumed route of the aqueduct: dendrochronological dating places them in the year 3 AD, which constitutes a reliable chronological benchmark. However, given the fact that the Yzeron canal is a masonry structure –though it has undergone numerous repairs– interpreting this wooden structure is not easy. If it is indeed a pipe linked to the Yzeron, it could indicate an ancient state, a diversion introduced during repair works or a pipe taking water to a few neighbourhoods.The Brévenne aqueduct, whose original name would seem to have been Aqua Claudia Augusta according to the inscription on the Clos du Verbe Incarné fountain, dates from the 40s or even 50s AD. A significant number of preventive archaeology operations support this conclusion. The ceramic finishes associated with the masonry or at levels associated with the implementation of the structure, as well as the radiocarbon dating of charcoal from the canal’s construction phase, also constitute solid arguments. An archaeomagnetic study of the bricks of the leakage tank of the Massues siphon at Tassin-la-Demi-Lune also reveals other important elements, with a chronological range between 23 BC and 88 AD. The dating of the Gier aqueduct, the longest of the Lyon aqueducts and the only one to supply the highest part of Colline de Fourvière, has long been a very controversial subject.Several scientific contributions have addressed the question, either dating it to the reign of Augustus or to between the reigns of Claudius and Hadrian. This study picks up and analyses all the arguments used to support one or the other of these dates: the link between the remains of the hill’s water supply network under Augustus and the altitude of arrival of the aqueduct; pipes stamped with the TI · CL · CAES · brand found in the 18th century on Colline de Fourvière; the aqueduct’s protective boundary stones bearing the name of Emperor Hadrian; the attested stamps on the bipedal bricks of the arch and bridge piers; the opus reticulatum mixtum construction technique; the fountain at the enclosure of the fountain at the Verbe Incarné enclos and its water supply; radiocarbon and archaeomagnetic dating. Thanks to the recent contribution from preventive archaeology, which in 2018 provided a reliable terminus post quem, we can today situate the construction of this structure, or rather its commissioning, within a very restricted chronological range. The dendrochronological analysis of the wooden formwork of the foundation of one of the piers of the Beaunant siphon bridge gives a certain date, 110 AD, as the terminus post quem for the construction of the bridge. The discovery of two boundary stones protecting the canal, better known as “Pierre de Chagnon” and “Pierre de Saint-Joseph”, attests to the existence of the aqueduct during the reign of Hadrian (117-138 AD). This chronology is consistent with the 2nd century BC dating attributed to the attested CLARIANVS stamp on one of the bipedal bricks of the Beaunant siphon bridge.At the end of this effort to collect and collate all the documentation, reliable dating can be advanced for the aqueducts of Brévenne (mid-first century AD) and Gier (first half of the 2nd century AD) as well as a possible construction date for the Yzeron aqueduct around the beginning of the 1st century AD. The dating of the Mont d’Or, for which we perhaps have a terminus post quem, and that of a fifth adduction at Fontanières, are not yet supported by archaeology.
