6 The Aqueducts of Ammaia

It is clear that the Roman town of Ammaia was well provided with drinking water. The monumental nature of its town structures, with public buildings (e.g. bath houses) and paved streets, possibly adorned with some public fountains, from the start needed an efficiently functioning system of aqueducts bringing running water at a constant flow to certain parts of the town. Water provision systems for Roman towns in Lusitania are relatively well known, see e.g. Etienne and Alarcão 1974 (Conimbriga), as are different types of hydraulic systems for rural economic or domestic purposes, see e.g. Quintela et al. 1999. As the Romans usually followed the 'no nonsense – least effort' principle for bringing this water to town, geoarchaeological fieldwork and GIS modelling can be very helpful in reconstructing the routes and ways of distribution. Our on-going work in the field consists partly in tracing the possible places for easy and logical water capture in Roman times, hereby considering aspects of hydrology, topography, archaeological remains and other practicalities. Thus far we have been able to determine and partly study the two main sources for water collection and transport to the Roman town: the springs at 'Olhos de Agua' and a newly discovered aqueduct along a small western tributary of the Rio Sever (Fig. 14).

Figure 14: Cadastral layer with indication of the main features of the local hydrology and recently discovered Roman springs. [View static image]

The western flanks of the hill 'Cabeço do Leão' on the other side of the Rio Sever, east as seen from Ammaia, form part of an area rich in water sources, all to be found at the contact of schists and dolomites. The fractured dolomites act here as aquifer, while the underlying shists are the aquitard. In particular, the site of a 19th-century mill at 'Olhos de Agua', immediately east of the Rio Sever, where the contact between dolomites and schists crops out, is suited for capture – even today this is an important point of capture, bringing water at a constant flow, averaging around 50 litres per second, to the town of Portalegre.

The concentration of springs lies some 8m above the level of the most easterly part of the Roman town, the area in front of the museum at Quinta do Deão. The distance from this spring area to the town wall is only some 300m, which means that this possible aqueduct could have been built with relatively minor effort, although the construction of a bridge was necessary.

During our first campaign of fieldwork (2001) we observed several indirect archaeological indications in this area that could point to the existence of a Roman aqueduct bringing water from this rich spring, via a bridge over the Rio Sever, to the lower parts of the town. They are: a series of ploughed-up Roman building remains near the present-day water source, the remains in situ of a foundation possibly for a Roman bridge near the eastern edge of the river (a bridge existed here until the end of the 16th century, Carvalho 2003, 72) and some larger granite building blocks, which can be interpreted as Roman spoils, in the walls of an abandoned small farm building near the river at 'Olhos de Agua'.

More convincing still were our observations the next summer, when we were fortunate to be able to study this area of natural springs in the best of conditions. Since major building works, to transform the farm complex at 'Olhos de Agua' into a centre to study the surrounding natural park, involved the temporary drying of the artificial pond on top of the springs, we had the opportunity to study several ancient structures for water capture, pre-dating the 19th-century 'industrial' installations. Although exact chronological information remains absent, we consider that at least a couple of the discovered capture points could go back to the Roman era. Several of them are located at the foot of the valley side, and in one case a foundation of a quite regular construction appeared. Partly cut into the quartzite bedrock, partly constructed with irregular boulders, a basin was formed of irregular rectangular shape. This construction of c. 1.5 × 1.3m was probably only the sub-structure of a more monumental spring capture. A few metres from this in situ installation were found a concentration of relocated Roman building blocks of identical imported granite: two rectangular pieces and two fragments of small columns. More fragments of Roman columns (granite and quartzite) were observed at a short distance from these capture points, at least suggesting major Roman interference with the water collection in this area. It is not yet clear, however, whether these Roman installations were purely functional (e.g. basins, water tower, specus construction) or also involved the presence of a sanctuary or nymphaeum.

Further research would be necessary to trace the exact location of the aqueduct bringing the waters of 'Olhos de Agua' to the lower parts of the Roman town. It is very probable that such an aqueduct crossed the Rio Sever near or at the location of the former bridge.


© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions | Last updated: Mon Nov 21 2005