There is no clear evidence for the continuity of use of watermills in Britain after the end of the Roman period into early Anglo-Saxon times. It is thought that the economic, political and social upheaval that occurred after the withdrawal of Roman army in the early 5th century AD was greater than in mainland Europe. The subsequent collapse of the market economy, as well as a fall in population, probably led to a decrease both in cultivation and in the demand for cereals, making the survival of watermills less likely.
The earliest archaeological evidence for the presence of mills in the early medieval landscape dates from the late 7th century and there are documentary references from the middle of the 8th century. However, a charter of AD762, in which the Kentish king Æthelberht negotiated an exchange with the minster of St Peter and St Paul, Canterbury, to acquire a half-share of its mill near Chart for the use of his royal vill at Wye, is the first to record a mill as such. During the 9th century charter and place-name evidence becomes more plentiful and the number of references to mills increases through the 10th and into the 11th centuries, presumably reflecting a more stable and settled agriculturally-based society as well as an increase in the survival of documents.
Some evidence of Anglo-Saxon watermills has been found by archaeological excavation.
The remains of timber structures dating from the late 7th century were found at Ebbsfleet, Northfleet, Kent, and Worgret, near Wareham, Dorset and a comparable structure to that at Worgret was excavated in advance of gravel extraction at Wellington, Herefordshire. All these examples appear to have been horizontal-wheeled mills, that at Ebbsfleet being a double mill, with two wheels each driving a pair of millstones. The Ebbsfleet mill was worked by salt water which was impounded in a pond at high tide and released onto the waterwheels through inclined timber troughs when the tide had ebbed low enough for the wheels to turn.
Two of the earliest examples of early medieval mills which have been found in Ireland, where over 150 sites are known, at Nendrum, Co Antrim and Little Island, Co Cork, were also tide mills, both dating from the early 7th century.
The first significant find of a horizontal-wheeled mill in England, however, was that excavated at Tamworth, Staffordshire, in 1971. The timber-built underhouse of the mill and a timber structure forming the end of a mill pond were dated to the middle of the 9th century. Finds included a paddle from a horizontal waterwheel and numerous millstone fragments. The mill appears to have been a double one, similar to that at Ebbsfleet, by comparison with Irish finds and also the number of millstone fragments associated with it. It is possible that each pair of millstones had a different function, such as shelling and grinding, or perhaps one pair was used for grinding meal and the other malt, for brewing. Some of the millstones at Tamworth were from local, Midland sources, while other fragments were from lava stones which were imported from Germany.
Vertical waterwheels were also known in Anglo-Saxon England. A large, somewhat enigmatic structure of which the base timbers were found in the bottom of a substantial man-made watercourse taken off the river Thames at Kingsbury, Old Windsor, Berkshire, in the 1950s, appears to have housed two vertical waterwheels in parallel, with a spillway channel between them. Evidence for the use of vertical waterwheels has also been found at West Cotton, Raunds, Northamptonshire and probably at Corbridge, Northumberland, which both date from the later Anglo-Saxon period.
The survival of both the remains of Anglo-Saxon mills and documentary evidence for them is somewhat random, the main concentration being in the south and east of the country. Domesday Book, the result of the survey carried out for William I some 20 years after his conquest of England in 1066, lists over 6000 mills spread throughout the lowland England, indicating that they were common features of the landscape by that time.
The majority of mills ‒ molendina – recorded by the Domesday survey are thought to be water-powered mills for grinding grain, although references to mill oxen in Huntingdonshire in late 10th century documents and the possible identification of an animal-powered mill at the Anglo-Saxon royal palace at Cheddar, Somerset, does suggest that not all mills were water-powered.
In some areas where the topography was less suited to water-power use, with slow running streams which had little gradient or fall, animal mills would have been invaluable. Windmills, however, do not make an appearance until about a century after the Norman conquest.
At the time of the Domesday survey in 1086 there was on average a mill to every 40 households. Norfolk and Lincolnshire had the highest number of mills and sites, with numbers diminishing in the south-west, particularly west of the river Exe, the Welsh borders and the north-west. No mills were recorded in Lancashire, modern Cumbria or Northumberland and Durham.
The record of mills in Domesday Book is not easy to interpret, however. Some are referred to as fractions, apparently divided between different owners or manors, such as at Coleshill, formerly in Berkshire, where three thirds were held by three different landowners, but the fractional entries do not always add up so neatly. Some mills are listed in groups under one manor or estate, from two to as many as 13, but were they physically grouped together in one area or dispersed throughout several holdings under the same ownership? Both scenarios are possible.
Alongside the mill entries a value is usually given, either in money or in kind. Monetary values vary enormously from as little as 3d (three old pennies) to several pounds, some mills in Huntingdonshire being amongst the most valuable recorded by the survey. Again questions can be asked as to what such values mean: does the given value relate to the available power, which would affect the milling capacity – as indicated above some may have been animal-powered. Or did the values relate to population, the number of customers, or the amount of arable land surrounding them? Is the high value of a mill a reflection of it being long established and so perhaps having good grinding custom? Renders in kind are also recorded, in grain, malt and rye, in honey and also in eels, with several examples of the latter on the Great Ouse in Bedfordshire.
A number of mills were attached to fisheries, presumably in order to make use of the same weirs that directed water to their waterwheels. As well as extant mills, there are references to some having been destroyed, as in York, to make way for the new castle there, to mill sites and to a small number that were newly built since the Conquest. Only eight millers and mill-keepers are recorded, suggesting that in most places milling was not a primary occupation.
It may well be that many surviving watermills stand on or close to the site of those identified under the name of a manor or villa in Domesday Book, but such claims are difficult to prove. The continuity of use of many mill sites is likely to have swept away any evidence through subsequent rebuilding and to date no securely-dated mill remains from the second half of the eleventh century have been found by excavation.
It is also feasible that some surviving artificial watercourses were created to drive watermills in the late Anglo-Saxon period, such as the diversion of the Thames at Abingdon, Oxfordshire, which Abbot Æthelwold had made in c.960 to supply a new mill at the abbey there, although after nearly a millennium of floods and changes in river courses and land-use, surviving evidence may be slight. However, the coincidence of part of a parish boundary with a leat and tailrace, as at Dotton, beside the River Otter in east Devon, suggests a site of some antiquity. Dotton was recorded as a small manor with a mill valued at 5 shillings in Domesday Book and was formerly a small parish with only a short eastern boundary on the river side. The site of Dotton Mill, which stood into the 1960s, may therefore have been occupied for over 900 years.
Darby, H.C. 1977: Domesday England, Cambridge University Press
Hardy, A., Watts, M. and Goodburn, D. 2011: The Mid-Saxon Mill at Northfleet. In P. Andrews, P. et al., Settling the Ebbsfleet Valley. 1. The Sites. Oxford, Wessex Archaeology, 307-49.
Hodgen, M.T. 1939: Domesday Water Mills, Antiquity, 13, 261-79
Rynne, C. 2000: Waterpower in Medieval Ireland. In Squatriti, P. (ed), Working with water in medieval Europe, Leiden, Boston, Köln, Brill, 1-50
Watts, M. forthcoming: Watermills and Waterwheels. In Clegg Hyer, M. and Hooke, D. (eds), Water and the Environment in the Anglo-Saxon World, Liverpool University Press