to its waterlogged conditions, Vindolanda has yielded an extraordinary collection of artifacts. Among these is a series of massive wooden water mains that were still carrying spring water into the site. Wooden junction boxes and pipes laid under the road have also been found at Carlisle and Cirencester. These finds are exceptional, but they probably preserve evidence for facilities that would have been found in some shape or form at any major Roman military or civilian settlement.
Managing water on an industrial scale was absolutely fundamental to Roman civilization. The accepted method was to locate a source at a higher level than the settlement and divert the water in, either via an architectural aqueduct or, more commonly, by digging conduits into the ground, carefully following contours to maintain the supply. At Dorchester a reservoir was constructed by building a dam, from which a 11-km (7-mile) long timber-lined conduit brought the water into town. At Greatchesters on Hadrian’s Wall, a channel around a metre wide carried water 10 km (6 miles) along the contours, even though the source was just 4 km (2.25 miles) from the fort. One of the earliest aqueducts in Britain was built in timber to service the Neronian ‘proto-palace’ at Fishbourne.
On arrival at the settlement, the water poured into the castellum divisiorum (‘dividing reservoir’) at the highest point where sediment could settle. Controlled by sluices, wood or lead pipes then carried the water off to wealthy households, public fountains and baths, and various industries. A small-scale version of this can be seen at Housesteads. Water ran down across the fort to its lowest point, where tanks and conduits controlled its passage into latrines. Like most Roman water systems, the principle was that of ‘constant off-take’. Water flowed continuously into facilities, but to reduce the pressure at these outlets the main supply ran first into header tanks on pillars at street corners. The constant flow prevented the header tanks from overflowing and kept the water fresh, while the reduced pressure allowed domestic and bathing taps to work without being blown off.
One of the drains beneath the temple and baths complex at Bath, and still in working order. The Roman management of water was key to draining the site and making it possible to build here.
Additional water supplies could be had from wells, but anything beyond modest requirements required major engineering. At Gresham Street, in London, remains of at least two chain-driven water-bucket systems have been found, one of which was in action by the 60s [ 137]. The wooden buckets were joined together with iron links and shafts, rotating on a massive wooden wheel to lift water from a timber-lined well that was around 6 m (19 ft 6 in) deep. The systems might have been used for light industry or small nearby baths, but neither lasted in use for long and were perhaps abandoned when aqueduct supplies became available.
Supplies at individual houses or villas were more ad hoc. Silchester has produced remains of a hand-operated wooden pump with lead cylinders, capable of shifting 22 litres (39 pints) of water per minute, and possibly used for
Raising water to a domestic cistern. 17 At Chedworth (Gloucestershire) spring water was diverted into a small pool at the site’s highest point, from where it could be piped into the two nearby bath-suites. Stone conduits found near Woodchester (Gloucestershire) probably carried water into the villa complex. Proximity to a river played a large part in choosing a villa site. A series of villas grew up along the Darenth Valley (Kent), all of which lay only a few metres from the river. The Redlands Farm (Northamptonshire) villa was originally built as a watermill, and two stone-lined aqueducts fed the water in from the nearby River Nene. Subsequently the mill was dismantled, and the house extended and converted into a more conventional villa-type structure.
Disposing of waste water was achieved in different ways. York had monumental masonry sewers, large sections of which have survived intact. Silchester, conversely, seems to have depended on open channels in streets. Such facilities were not necessarily bad. At St Albans, the drains around the forum were maintained throughout the period and only became clogged in the fifth century.
A modern reconstruction of the firstcentury water-wheel system. (Museum of London)
World History
