* staff

* student

* search

Bordesley Abbey tile
Bordesley Abbey Introduction > Watermills

Bordesley Abbey mill: reconstruction of the period 4 mill


The Bordesley Abbey Project was always intended to relate the abbey church to the precinct, and the small-scale excavations referred to at the beginning of this review demonstrated that the valley had been changed in a much more extensive and complex way than was first thought. Even at its simplest, it was necessary to think in terms of three broad landscapes: the pre-monastic, the monastic, and the post-monastic, and to try to characterize each landscape (Astill 1989, 283). The project, therefore, extended its enquiry under the direction of Grenville Astill into the outer areas of the precinct, and concentrated especially on those parts which do not survive very well elsewhere. The focus was on the eastern extremity of the precinct where the earthworks indicated a triangular mill pond and, therefore, a potential site of a watermill and workshops (Aston and Munton 1976, 32). Open area excavation over the mill site was combined with a long, narrow trench across the mill pond and the Arrow valley, and two other cuts across the tail race, in order to set the mill in its context and to recover environmental data, especially macrobotanic material (BAB, BAE, BAH, BAJ, Figure 3, above). Geophysical and geochemical surveys were also conducted over a substantial part of the precinct. Drawing on this fieldwork and using the data from the smaller excavations, an outline development of this part of the Arrow valley can be sketched (Astill 1993, 3—7).

A river valley in the heavy clays and woodland of north Worcestershire has traditionally been regarded as unsuitable for early landuse, with the result that it had been assumed that the Cistercians were the first to exploit the area. Most of the excavations, however, have produced residual finds of unrolled, worked flint of later Mesolithic and Neolithic date (Wright 1976b, 139—40; Watts and Rahtz 1983, 129; Astill 1993, 123) which are sufficient to demonstrate that the valley was, at the least, frequented from the later Mesolithic. Then the valley had a different character from the twelfth century, or today. Excavations show that large amounts of silt were deposited in a valley where the water flowed in an irregular network of channels - anastamosing channels - rather than in a single course. Channels were constantly being cut and occupied by water and then filled with silts and organic material. Such activity is not to be associated with flooding, but reflects the normal condition of a valley in a ‘low energy’ stage of its development, where migrating channels and gravel banks are typical features. The botanic material from some of the earliest excavated channel fills shows a mixed oak and lime woodland which might be compared tentatively with other west midland sequences assigned to the Bronze Age (Carruthers 1993, 209). Some of the fills of the later channels contained large and unabraded sherds of Roman pottery dated to the first and second centuries AD. The botanic material shows a change to an open alder woodland, an indication of growing intensity of use. There was probably Roman occupation in the vicinity, but the wet conditions in the valley bottom might suggest that the area was used as summer pasture by settlements located on the tops or sides of the valley (Carruthers 1993, 210; Astill 1993, 8—13).

The continuity in the the open woodland landscape implies that the landuse in the lower parts of the valley remained the same for a considerable period. The excavation across the boundary bank at the north-western edge of the precinct produced clear evidence of ridge and furrow sealed under the bank (BAC, Figure 3, above). The field was being farmed before c.1200, and it is likely to have pre-dated the foundation of the monastery. This arable may have been incorporated into the precinct as a buffer. The placename ‘Bordesley’ - a clearing where boards were obtained, or Borda’s clearing - implies that the land had been cleared before the abbey was founded, probably in the later Saxon period, and that the land had been acquired from the inhabitants of the settlement; indeed the reference to terra de Bordesleie in the foundation charter might refer to cultivated land (Burrow and Dyer 1976, 131). The adoption of the place name by the abbey might indicate that the settlement, as well as a part of its fields, may have been absorbed, and then deserted, in the way that has been well documented for Cistercian houses in the east midlands; it should be noted, however, that no excavation has produced pre-monastic, medieval material (Barley 1957).

Bordesley Abbey precinct thumbnail

Click here to see the image in full
Figure 12 Bordesley Abbey precinct: a reinterpretation
of the precinct's development.
(Drawn by Brian Williams)
Bordesley Abbey Project

Bordesley Abbey mill pond timber drain
Click here to see the reconstruction in full
Figure 13 Bordesley Abbey mill pond timber drain (later twelfth century, period 3): with plank lid (top) and with lid removed middle), from the south-west; (bottom) reconstruction of the drain structure.
(Drawn by Steven Allen)
Bordesley Abbey Project

Bordesley Abbey timber mill race (period 3)

Figure 14 Bordesley Abbey timber mill race (period 3): from the east, with head race (top), wheel pit (centre) and tail race, with north side missing (bottom); dendrochronological felling dates
of tail race timbers 11745.
Bordesley Abbey Project

Bordesley Abbey timber mill race (period 4)
Click here to see the image in full
Figure 15 Bordesley Abbey timber mill race (period 4): the structure; dendrochronological felling dates of tail race timbers 11868.
(Drawn by Steven Allen)
Bordesley Abbey Project

Bordesley Abbey mill building: plan (period 4)

Click here to see the image in full
Figure 16 Bordesley Abbey mill building: plan (period 4).
(Drawn by Brian Williams)
Bordesley Abbey Project

Bordesley Abbey mill: reconstruction of the period 4 mill

Figure 17 Bordesley Abbey mill: reconstruction of the period 4 mill.
(David A. Walsh)
Bordesley Abbey Project

Bordesley Abbey wooden mill machinery:

Click here to see the image in full
Figure 18 Bordesley Abbey wooden mill machinery:
wheel felloe fragment (OB 169) and reconstruction of trip wheel based on this fragment; paddle blades (OB 282, 165.4); staves from paddles of waterwheel (OB 165.2, 165.3, 191.14).
(Drawn by David A. Walsh (OB 169) and Steven Allen)
Bordesley Abbey Project

Bordesley Abbey mill: copper alloy sword pommel (CA 37) and pommel casings.

Figure 19 Bordesley Abbey mill: copper alloy sword pommel (CA 37) and pommel casings.
(Drawn by David A. Walsh)
Bordesley Abbey Project

The eastern part of the precinct provides dramatic evidence for the impact of the Cistercians on the valley. Here, the building of structures dated to the 1170s demonstrates that virtually all the precinct had been taken into use early in the life of the monastery, and in all probability within the lifetime of the first Bordesley monks. The two-step outward expansion of the precinct proposed after the initial survey of the earthworks (Aston and Munton 1976, 33—5) has thus to be revised (as in Figure 12). The diversion of the river Arrow and the ‘Red Ditch’ represents the first step in the massive drainage operations which were necessary to make the valley permanently habitable. Large culverts were dug to take the water away, a temporary building was constructed, and the alder and hazel trees were felled by slicing off at the roots. The tree stumps were then thrown into the (now redundant) culverts. The lack of significant silting after this operation is a testimony to the abilties of the Cistercians to stabilise and control their environment, but not without the constant maintenance of the drains, culverts and ponds (Astill 1993, 13—16).

The banks of the mill pond were raised soon after; perhaps some of the gravel ridges deposited by previous riverine activity were used as a foundation, which may explain the pond’s unusual triangular shape. At the same time elaborate provision was made for water control and drainage, as shown by a culvert, hollowed out from an oak trunk, which was laid on the old ground surface and subsequently covered by the mill pond bank (Figure 13; Allen 1993a, 95). The drain allowed water to be taken from the pond; it flowed under the bank and debouched into another culvert beyond the mill pond.

A channel was cut at the eastern apex of the mill pond and the upcast dumped to the north to create a platform. In the channel was constructed the mill race, which survived as a substantial oak structure (Figure 14). Nearest to the mill pond a head race and sluices were built, and these controlled and funnelled the water towards the pit which would have contained a vertical water wheel. Downstream of the pit the tail race comprised a timber-lined channel, jointed with crude mortices and tenons. The base timbers had clear signs of being worked with an adze; a base timber and one other timber from the tail race had dendrochronological felling dates of 1174, and two more had dates of 1174—5. The mill building was constructed on the platform, overlooking the wheel pit. It was founded on major uprights, set into post pits, of which one post had been felled in 1175. Inside, the building had a large, high, open space which contained two hearths, with lower lean-tos to the north, east and west. The structure was extended over the mill race as a wheel house.

After a short time - about 12 to 14 years - the building caught fire, and was replaced by another mill. The head race was reused, and a new tail race was superimposed on the first tail race, but on a different alignment. This structure used a different construction technique, where the tail race incorporated baseplates, using ‘through mortices’ with pegs (Figure 15). Three of the baseplates were felled between 1186 and 1188 (Allen 1993b, 74—9). The mill building was rebuilt on virtually the same plan as the first mill, but the structure was founded on padstones and pebble foundations and was not earthfast (Figures 16 and 17). Ceramic tiles were used to cover the roof and to form the two hearths. This mill operated between the late twelfth and early fourteenth centuries. It was then replaced: the head race and building were esentially retained, but another tail race was superimposed on the previous structure. This race was a more sophisticated structure using sawn and smaller timber and was probably prefabricated. The superimposition of the tail races clearly reduced the gradient of the mill race, and must have lowered the efficiency of the water wheel. The raising of the mill pond banks by about a metre should, therefore, be interpreted as an attempt to increase the head of water to maintain the mill’s efficiency. A subsidiary building was also constructed to the east of the mill and may have been an additional workshop.

At some time in the fourteenth century the tail race was again modified by shortening the structure and replacing the side boards. The mill building was partitioned into workshops, and the subsidiary workshop to the east was replaced by a larger and more substantial building which contained an additional forge. The mill went out of use in the late fourteenth or early fifteenth century, when apparently the whole site was covered in flood deposits.

The finds from the silts of the tail races allow a reconstruction of some of the machinery as well as indicating what the mills were producing. Most machinery fragments (Figure 18) came from the first mill tail race, including paddles from a water wheel (of c.3.5m diameter), a wheel fragment with a setting for a cog - interpreted as a trip wheel - and a collection of cogs with distinctive wear patterns. The cogs were of different proportions and this suggested that they were used on three different trip wheels. Stone bearings, showing areas of polished wear and worn seatings for the iron spindles (‘journals’) of shafts, were also recovered. On the basis of these finds, it is argued that water power was harnessed to trip a hammer and power bellows to maintain the temperature of the hearths inside the mill buildings - and these hearths were used for metalworking. This is important evidence for waterpowered metalworking and, with the later twelfth-century date, represents the earliest occurrence of this process in the country so far. The date, however, is very close to the first documentary references to fulling mills which would have probably used similar machinery (Astill 1993, 267—71, 302).

What was being made in these mills? Firstly, the evidence for metalworking exists in the form of incomplete forgings, bar iron and scrap metal, and a small number of tools such as punches, chisels, awls and reamers. Iron seems to have been used to make small items such as nails and tenterhooks, but there were also part-finished copper alloy fragments, and lead offcuts, which were probably used for repairing bronze and ceramic vessels and other artefacts. The same materials were also used to make and repair composite items. A sword pommel, pommel casings (Figure 19) and a dagger chape indicate specialized activity, and the military theme is continued with the arrowheads, a caltrop and the individual riveted sheets from a piece of armour - a coat of plates. The finds point to a mixed and sophisticated metalworking combined with some repair work. The most likely use for the water power would have been to drive hammers to draw down iron bars and to mass produce small items such as nails. The scale of production would have been too large for the community alone, and the mills probably supplied the abbey’s granges and local markets. Indeed, the date the mills were working - between the late twelfth and late fourteenth centuries - coincides with the period of direct farming and intensive marketing in the west midlands; and one reason for the abandonment of this industrial complex may have been a reduction in demand after the general leasing of lands and a decline in the market.

Why did the mills stop production at least 120 years before the dissolution of the abbey? There are local reasons. Excavations through the tail race 50m and 250m downstream of the mill (BAH, BAJ, Figure 3, above) showed that there was a rapid silting in this part of the valley after the construction of the second mill; this would explain the decision to superimpose the tail races rather than to maintain the same level. The raising of the mill pond banks to increase the head of water was a short-term solution, and did not prevent the mills becoming inoperable. The flooding of the mill site was apparently the culmination of a sequence of deteriorating drainage and maintenance of the monastic water system. During the later fourteenth century the supply and overflow channels for the pond were home to plants which grow in still or slow flowing water or marsh, and are common in derelict canals; their presence indicates a lack of maintenance, and in addition the mill pond was silting up. A similar conclusion can be drawn from the evidence for the regeneration of woodland, especially alder and field maple (Carruthers 1993, 210—11).

The lack of diagnostic earthworks and the failure of geophysical prospection to locate a potential successor workshop site in the areas to the west of the mill pond raises questions about what happened to these lower parts of the precinct. The environmental and the archaeological evidence point to the probability that the community was no longer able to maintain the complex water system, and a decision was evidently taken to abandon most of the precinct and to regroup resources around the cloister area which was on higher ground. Such a decision would also have been influenced by the economic and social conditions of the late fourteenth and fifteenth centuries where institutions were desperately trying to reduce expenditure and, in the case of non-mendicant monasteries, this may have meant reducing recruitment; in this they were perhaps helped by a growing unwillingness to follow a monastic vocation (Dyer 1989, 80—108).

Is it possible that the industrial workshops were relocated in or near the cloister, especially as there are cases of other Cistercian houses where buildings in or near the cloister were converted to industrial use in the later middle ages (Astill 1993, 275—6; Coppack 1998, 111)? The project’s most recent excavation (directed by Grenville Astill) was designed to test this hypothesis. A trench that had been dug in 1965—6 by a local society across the south cloister range was reopened. The interim (and only) report mentioned extensive evidence of metalworking, but dated it to the seventeenth or eighteenth centuries, that is post-Dissolution activity. The rarity of post-medieval material from all our excavations, and the abandonment of the mill and workshops lower down the valley in c.1400 raised the possibility that this cloister evidence might be redated. And this indeed was the case, for the latest pottery recovered is now dated only to the fifteenth and sixteenth centuries. The location of workshops in the south cloister range is of immense importance, not only because it completes the sequence of monastic metalworking, but also because it raises important questions about how the various parts of the cloister were being used in the 150 years before the Dissolution - questions which have yet to be answered.










Abbey church

Gateway chapel

Precinct, watermills




Tel: + 44 (0)118 378 8132 * Find Us
Email: archaeology@reading.ac.uk The University of Reading