Medieval aqueducts have attracted archaeological attention since at least the mid-19th century, although evidence is scarce. Early descriptions and surveys were usually carried out as part of the descriptions of the religious houses to which they primarily supplied the water, for example that at Christ Church, Canterbury (Willis 1868, 158-183). Other key surveys include the surviving structural remains of the conduit-heads at Bloomsbury, London, which supplied water to the nearby friary (Norman, 1899, 1909 1915), Southampton (Davies 1883) and more recently at Lichfield (Gould 1976) and Canterbury (Bennett, 1988). Very few excavations have been carried out, although survey and excavation in Exeter over the last 50 years have revealed the plan and stages of development of the aqueducts from the late 12th century onwards. Sections of conduits or piping have been found during excavations of other classes of monument such as religious houses and manors.

The overall size of aqueducts varied according to the distance between the source and final distribution of the water; at Lichfield the aqueduct was 2.3km long and that at Scarborough stretched for over 3km.

One of the easiest components to identify, and for which there are sometimes visible surface remains, is the conduit-head; this was the structure in which the water was gathered, near the source, before it passed into the main pipes. This generally comprised a tank housed in a stone superstructure. At Lichfield a cistern was cut into the rock measuring 1.74m x 1.07m and was 2.21m deep; this was housed in a gabled building with a roof of corbelled stones supported internally by a pointed arch, suggesting a 13th-century date (Gould 1976). At Bloomsbury, London, there were two conduit-heads at the source of the water system supplying the friary; the original site comprised a reservoir, 2.7m. x 1.8m, built of Portland stone with an arched roof of white limestone; in one corner was a shallow tank which may have been used for settling the water before it passed to the supply tank (Norman 1909). Later, the water system was extended and another conduit-head was built; this has now been demolished but originally comprised an underground reservoir, c3.1m square, built of blocks of neatly jointed ashlar with a tile floor; the roof was barrel vaulted and had the base of a chimney at one end which would have extended above ground level. The building was approached by a passage containing a flight of steps spanned by three arches (Norman 1899, 1915). The recently surveyed conduit-house supplying water to St Augustine's Abbey, Canterbury was an octagonal structure with four tunnelled openings and three smaller ducts giving on to a central chamber (Bennett 1988); the only visible surface remains of this was a slight mound. The Norman drawing of the water system supplying Christ Church, Canterbury suggests that the conduit-head here was circular.

In some cases the water may be drawn up from a well. At St Sidwell's, Exeter, a square stone well 2.25m deep was situated over a spring and covered with an elm plank; from here water was conveyed in a lead pipe to a further spring and standpipe (Tucker 1858). This system has now been destroyed.

Ponds may also have been part of the medieval aqueduct systems. At Canterbury it is thought that the water may have flowed from the spring into a pond; this may have been linked to a smaller pond, now an overgrown depression, which lies nearby, and formed part of the medieval system (Frere and Bennett 1987).

From the conduit house or pond the water was then carried by direct pressure in pipes, generally of lead; the conduit at Southampton is known to have been fitted with lead pipes, as is that at Bristol. At Exeter a lead pipe, 0.075m in diameter, was found to be set in the base of the stone channel of the cathedral conduit (Fox 1951). A pipe from Waltham Abbey was made from lengths of sheet lead which were bent up and the longitudinal seam sealed with molten lead; the ends were butted and surrounded by a 0.04cm-wide band and the joints were sealed by tooling on molten lead (Huggins 1976, 98). An early pipe at the Manor of the More, Rickmansworth, was wooden; this has since decayed but a void 0.35m in diameter was left (Biddle et al 1959, 154). In some cases the pipes may have been earthenware.

Sometimes the pipes were laid in a trench, as at Waltham Abbey where it was bedded in clay, in other cases the pipes were laid in a conduit. At Exeter this comprised a stone-lined passageway, through which all parts of the pipe could be reached for repairs. At Rickmansworth, the 0.45m-wide conduit comprised brick walls set on a floor of glazed tiles; this may have had a planked roof, although where the water supply was brought through the gateway the roof may have been of brick corbelling. The conduit at Scarborough had a base of solid blocks of sandstone; these were hollowed out to form a half-round channel which was capped with thin flags, some of which show evidence of having been mortared.

Along the course of the aqueduct the water may be passed into settling tanks to clear the water by allowing the sediment to settle. The illustration of the system at Canterbury shows that the water passed through five settling tanks after it had left the conduit-head and before it passed inside the city walls; the tanks were oblong and placed transversely to the course of the pipes.

Along the course of the aqueduct at Lichfield there were a series of plugs; these were sited where the pipe ran close to a stream in order that water could easily escape when the pipe needed to be repaired on the Close-side of the plug.

The illustration of the Canterbury water system also shows that once the water was inside the cathedral precincts there was a series of tanks or cisterns at the various places where the water was required, each of which was at a lower level than the preceding. The water was drawn off by using plugs, spigots or cocks.

Before the final distribution of the water it may have been stored in a water-house; at Exeter, for example, the water was brought in from outside the city to a water-house in the Cathedral Close from where it was distributed evenly between the cathedral, the town and St Nicholas's Priory.

Since there are so few archaeological traces of medieval aqueducts it is difficult to classify them on morphological characteristics; there is, therefore, just one type of medieval aqueduct. Figure 1 shows the courses of the water systems at Lichfield, Bloomsbury, and Canterbury.

Many of the aqueduct systems continued in use during the post-medieval period, although they were frequently repaired and extended. Where they were no longer used, the lead piping was often robbed and the aqueducts fell into disrepair.

Medieval aqueducts were in use from the 12th century onwards to transfer water from its source to where it was required. Initially aqueducts were set up by religious establishments to provide a fresh piped water supply to abbeys, friaries, priories etc. who paid for the high cost of construction, maintenance and organisation of the system. Some outlets were usually granted to the town for general use and occasionally to private houses. It was not until later that the towns took over responsibility for the works; in Southampton, for example, this occurred in 1420.