| Summary : Remains of the Roman aqueduct, circa 12 miles in length, which brought water from the River Frome to Dorchester. Follows the Frome valley as a contour feature, with a total drop of 25 feet.Originally thought to have drawn water from the Frome at Maiden Newton, but excavations have shown there to have been a Roman dam at SY62759426, with a core of crushed flint and puddled chalk. The aqueduct channel consisted of a chalk cut channel ranging from 1.8m to 3.1m wide and from 1.8m to 2.1m deep with sloping sides and an irregular base, with the excavated soil in parts forming an outer bank or form of terrace on the chalk hillside. The chanel may have been clay-lined, and was recut several times. Abandoned in the 5th-6th centuries. |
More information : Dorchester Roman Aqueduct. For full details see linear archive.
This aqueduct supplied water to the Roman town of Dorchester, starting from a point near Frampton and following the contours down the valley of the River Frome to Dorchester.
Major Coates first recognised the aqueduct and described some of its course.
Excavations at several points showed a channel 2ft. 6ins to 2ft. 9ins deep and 6ft. across at the bottom, confirming Coates' identification.
Further details and alternative source for water at Notton suggested.
In 1846 a section was cut for the railway tunnel, and a draining made by a railway surveyor, in which the aqueduct was seen but not identified. A siding near the railway tunnel was excavated 1918 and a good photo taken showing the upcast from the aqueduct thrown into Poundbury Hillfort ditch, above the turf line. Two sections excavated 1939 near hillfort, showing a channel with sloping sides roughly 3ft deep and 5ft 3ins across the bottom. Section to the W of hillfort shows aqueduct cut into natural chalk, but that cut near N rampart shows it cut into the ditch lined with clay to prevent water seepage. Hints also that the corners of sides and bottom were luted with puddled clay.
Traceable as earthworks or in APs for a length of some 12 miles. Has been sectioned at 14 points between 1855 and 1956, and shows a flat bottomed ditch with steep sides, the excavated soil forming an outer bank. No evidence of covering or lining - except for Poundbury Ditch, any clay may be natural. Not much dating evidence - some C1AD and C2AD Samian sherds have been found in the channel and there is no record of later material. Section E of Penn's Plantation, at W shoulder of re-entrant, terrace was 30ft wide in 1955,including slight outer bank, but here ploughing is altering its character. E of the re-entrant traces of the terrain remain as far as Grove House, and it is well marked on both sides of Strap Bottom. Recognisable in Coombe Bottom.
Barely visible along edge of Hill Plantation but well marked running for about 100yds across neck of plantation. Despite renewed ploughing it is impressive in 2 adjoining coombes in Fordington Down. Upcast bank stands 3.6 to 4ft above channel filling. Substantial terrace as it approaches Poundbury.
W end of Quatre Bras section under modern hedge. Scarp of terrace 5m high in this area. Visible stretch on N side of Strap Bottom (not scheduled), but on S side the majority is destroyed, in cultivation with small segment remaining overgrown with bushes. Coombe Bottom section now overlain by modern bank, with small section in plantation still 1m high and remaining. N stretch in Fordinton Bottom area is much reduced in pasture, with the scarp up to 1.7m high on uphill side. S stretch has scarp over 2m high. Stretch running from Whitfield Farm to Poundbury is well visible with scarp heights over 2.5m in places.
Quatre Bras area (area `d' as scheduled) is clearly visible where free of hedge, but it is badly scarred from cattle. E. part of `d' is under tree cover. Area 166/c at Bradford Peverell village does not include the unscheduled N. section which is visible. Much of the S. section here has been destroyed by ploughing, but part remains impenetrable scrub. The Combe Bottom stretch, 166/b has gone except for the short piece within the plantation. Fordington Bottom stretch (166/a), N. part, is unploughed in cereal crop, and marked by posts, but plough is very close to north edge - however the exact position of the channel here is no longer clear. Long double stretch in Fordington Bottom is in grass with cattle, and also has some erosion scars on the uphill slope of both sides of the Combe. The very steep scarp of the section S.E. of Whitfield Farm has areas of considerable erosion and the growth of thorn - much of the eroded material is very loose. Nearer Poundbury there are smaller, less serios erosion scars, probably from rabbits, aggravated by cattle, particularly near the hillfort rampart. Overall condition is fair to good in parts. (1)
Listed. (2)
Full description with plans an photographs. (3) |