EXCAVATION REPORT CULVERMEAD 2006.

Culver Archaeological Project, Culver Farm, Church Road, Barcombe, Sussex, BN8 5TR. Email: info@culverproject.co.uk Web: www.culverproject.co.uk EXCAVATION REPORT CULVERMEAD 2006. By Robert Wallace Culver Archaeological Project 2013 1

INTRODUCTION. In August 2006, an archaeological evaluation of Culvermead field was undertaken to assess whether or not any archaeological deposits/features were present. Culvermead is located on Culver Farm Barcombe, East Sussex TQ 425147 (OS Map 198 1:50 000). This evaluation forms part of a larger project. In 1999, a field walking and a geophysical survey were undertaken in Dunstalls field by the Mid Sussex Field Archaeology Team (hereafter MSFAT) under the directorship of Chris Butler, and with the kind permission of Harold Stroude the land owner. They were led to this site by the vast amount of Roman artefacts coming to the surface especially after ploughing. The results of the field walking survey strongly suggested that there was a possible Roman villa on the site, and the subsequent geophysical survey confirmed the likely presence of a winged corridor villa (Fig 1). Figure 1: Geophysical survey results from Dunstalls Field (Butler et el 2006). It was in 2000 that a full open excavation was started, and MSFAT were joined by University College London Field Archaeological Unit (UCLFAU) now known 2

as Archaeology South East (ASE). UCL used the villa site as its training dig for undergraduate and some of its postgraduate students. The villa, when excavated, revealed that there was in fact a number of different occupation episodes with two stone built buildings; the earliest building was a basic rectangular structure (Fig 2) dating to the first century which had been built next to a possible late Iron Age/Romano-British round house. Archaeomagnetic dating suggests that the last roundhouse was contemporary with the rectangular building. Figure 2: The first stone rectangular building contemporary with Late Iron Age/Romano-British Roundhouse (Butler et el 2006) Illustration by Andy Gammon Lewes 2004 and 2006. Furthermore Prehistoric finds where also found during excavation, in particular a stunning Acheulian hand axe dating c500,000-1000,000bc (Butler et el 2006) (Fig 3). 3

Figure 3: Acheulian hand axe (c 500,000-100,000BC) (Butler et el 2006). The second building was a villa of the winged corridor type. The excavations were finished and back filled in 2001; the interim report is available (Barcombe Interim Report 2000-2001UCL Press), the site was extended to the South and excavations commenced on the larger farmstead complex (Butler et el 2006) (Fig 4). Fig 4: An interpretation of what the winged villa and farmstead looked like at Barcombe (Butler et el 2006). Illustration by Andy Gammon Lewes 2004 and 2006. 2003 was my first of three seasons excavating at Barcombe. In 2005 Chris Butler suggested that I investigate the fields around the Barcombe Roman villa 4

in order to increase our understanding of the wider Roman landscape. After consideration, and following my decision to carry on after my BA with a masters degree, I decided to take Chris up on his offer. I have undertaken field walking on the fields around the villa site. One Particular field Culvermead, providing some interesting results from the field walking (Culvermead has been of interest for some time and has been field walked 3 times now; the last was by myself and colleagues from UCL). Because of the previous two field walks we did not grid out, but undertook a surface collection to see if we gained similar results to the previous surveys. The results used here were from the field walking carried out by the members of MSFAT. They carried out the field walking survey using 20 m transects and collection units. Large quantities of Roman building material and Roman pottery were recovered. I also carried out an SMR search and have carried out a geophysical survey at Culvermead with help from UCL students (Fig 5). Figure 5: Geophysical results from the survey carried out in Culvermead. 5

A summary of the excavations and the finds at Culvermead will be discussed below, which will then be drawn together with other sources to give a full interpretation and current thought for the site. EXCAVATIONS AT CULVERMEAD. There were 7 evaluation trenches excavated at Culvermead. These trenches varied in the amount of archaeological material recovered. The trenches were identified by means of alphabetical lettering Trench A and so on (Fig 6). Figure 6: Trench A. B. C. D. E. F. G. 6

Trenches C, E & F were sterile of any Pre c.18 th century archaeological artefacts (this will be discussed later). For a full report on all the trenches see previous work (Wallace 2006 p2-17). Trench A was 13.40m in length, and approximately 1.6m wide (Fig A; Fig A-E can be found at the end of this report). In Trench A we uncovered various artefacts including pottery, Ceramic Building Material (hereafter CBM), iron nails, burnt clay, and Fire Cracked Flint (hereafter FCF). There appeared to be no structural remains such as foundations or post-holes although a possible metalled surface was present. This feature did not cover the entire trench, and a sondage was hand dug to see if the metalled surface had been over cut at the edge of the trench. This was not the case and no evidence of metalling was found in the sondage. Therefore this metalled surface was not part of the Roman road, as the road measures 8m wide, this feature is only 2m wide, therefore it seems unlikely that this was the roman road. Furthermore, example at Boxgrove showed similar metalled surfaces which were interpreted to be internal floors of buildings, a frame would have been made on the floor and a timber frame building built on it, then the internal floors were metalled, hence the lack of post-holes and foundations. Once the full open excavation takes place, we will hopefully have a fuller understanding of this metalled surface (Bedwin and Place 1995 p60). Trench B was 11m length and 1.6m wide (Fig B-C). It was one of the most interesting trenches, a feature approximately 3 meters in length was revealed, the width of which was not ascertained as the feature continued under both baulks. This feature was then half sectioned. There were a number of different 7

layers/deposits throughout the feature; there was a possible cut of a post-hole which was only seen in section (Fig C). Below this was a deposit of gaunt clay within which we found 3 lengths of waterlogged timbers. These displayed a good level of preservation (we stopped excavating at this time as plans had not be made for conservation of waterlogged timber, therefore after recording and taking samples for c14 dating the timbers were covered in heavy duty plastic and backed filled). To the North West of the above feature is a linear cut which also continues under the North baulk, it is most probably a ditch/gulley. Pottery, CBM, iron nail, FCF, and burnt clay were all found in Trench B. Trench C was 11.9m in length and 1.6m wide. As discussed earlier there were no pre c18th century finds in trench C, in fact the only find was a possible metalled surface, which dates to the mid 1840 s (this will be discussed in more detail below). Trench D was 18.5m length approximately and 1.6m wide. This was also an interesting trench; it had two metalled surfaces the older surface has been overlaid in part by the later metalled surface (this will be discussed in more detail below) Pottery, CBM and FCF were all found in Trench D. Trench E was 4.1m in length and 1.6m wide and Trench F was 7.4m in length and 1.6m wide. As mentioned earlier Trench E and F are sterile of pre c18th century artefacts. They both had similar metalled surface to Trench C. Trench G was 6.8m long and 1.6m wide. A large quantity of CBM was found amongst downland flints. Pottery, glass, nails, FCF and burnt clay, iron slag and a coin were also found. The coin is possibly First/Second century and is possibly Trajan (A.D. 97/117), unfortunately the coin in heavily corroded, and David 8

Rudling has initially identified as Trajan, although he will seek a second opinion in the New Year (D. Rudling Pers com 16/12/06). These finds appear to be within a large feature, which extends beyond the evaluation trench in all directions. INTERPRETATION. Any interpretation of the evaluation trenches at Culvermead must remain largely conjectural, although they can benefit future excavations to be carried out in the summer of 2008. Factually though of all the artefacts found in the trenches only 4 were not of the Roman period and these were recovered from the plough soil (Wallace 2006 p41). For the purpose of this interpretation we have to look at the wider picture and as with all good archaeological practice work from the known to the unknown. I will start from the Roman period (there are prehistoric finds and Iron Age buildings). In the mid to late 2 nd century, the first stone building was erected in Dunstalls field, Barcombe; it was a simple rectangular building with narrow sub angular flint foundations. By the early to mid 3 rd century the rectangular building had been demolished and replaced by a much larger winged corridor building. Furthermore a large isled building was built, almost certainly contemporary with the winged villa. The excavations that took place in 2003-2005 also revealed a detached early bath house, which the isled building was then built over. The Villa appears to have gone out of use around 300AD, only a few finds dating to the 4 th century were found at the villa site, and no evidence of occupation was found until the late Saxon period (Butler et el 2006). 9

To the West of the villa site, there is a field called Court House field. In the summer of 2005, the farmer Mark Stroude took me out into Court House field and showed me a large area of a sub angular flint scatter. At the time Mark was growing sweetcorn, in the field, there was however a row which had not grown and Mark gave us kind permission to place an evaluation trench here (Fig 8). Figure 8: Evaluation trench in Courthouse Field showing the barren sweetcorn row. Photo taken by Author. This excavation was carried out by my colleague Rhw Mitcheson, as I was supervising on the main villa site. The excavation went very well and an unknown Roman road was discovered (Fig 9). 10

Figure 9: Part of the metalled road surface of the before unknown Roman road. Photo taken by the author. Ivan Margary traced the Roman road from London to Lewes in the 1930 s and 1940 s. The London to Lewes road passes Culver Farm on the East side of the river Ouse (Margary 1965 p 150). Furthermore Margary states that he believed that there should be a Roman road on the West side of the river linking Offham to the Greensand Way (Margary 1933 p32). It is looking more than probable that the unknown Roman road we have found is the road that Margary suggested would be present. 11

As mentioned earlier the Greensand Way starts at Barcombe Mills and continues West towards Chichester (Margary 1935 p7). At the end of November and into December 2005 I opened another two evaluation trenches in Pond field which is to West of Court House field on the other side of the farm road. Members of MSFAT helped with the excavation, as we attempted to trace the route of the Roman road identified the preceding year. The first trench did not reveal evidence of the road, however we uncovered a ditch which is believed to be a possible Bronze Age field boundary, as a sherd of Bronze Age pottery was found (Pers com C. Butler 10/12/05). In the second trench further west across the field we discovered the base of the Roman road; the metalled surface and the agar having been ploughed away. We also found a ditch on the Northern edge of the road, it has long been debated that Roman roads would have had ditches either side of the road, and the fill of the ditch would have been used to create the agar. Conversely, not all Roman roads had ditches and Davies suggest that if the soils were adequately free-draining then ditches would not have been required (Davies 2002 p72). We had one ditch on the Northern edge of the road, and Roman pottery was found in it. Furthermore we discovered two substantial post- holes along the Southern edge of the road; I believe that these features may be part of a road side settlement (Fig 10). 12

Fig 10: Two post-holes found along the edge of the Roman road in Pond field. Photos taken by the author. Sites such as Sedgefield had timber framed building along the edge of a road, they also had a pottery kiln behind one of the buildings, and so once the pots were ready they could be transported directly onto the road system (Web1) (Figure 11). Figure 11. An artist impression of a road side settlement at Sedgefield (Web 1). 13

To the North West the adjacent field to Pond Field is Culvermead, and I believe the settlement may continue into Culvermead. There are one or two possibilities of what was happening in Culvermead. The first is that the road side settlement continues through Culvermead, and that the settlement is housing the farming work force and their families around the villa complex. The second possibility is that we have a small industrial complex in Culvermead; this is a train of thought that I m pursuing at present and the results from the evaluation trenches allow for the development of a possible scenario at Culvermead. The one possible known feature that should appear in Culvermead is the Roman road and looking at our geophysical survey results (Fig 5) we believed that we had identified the road. It also looked like it had a dogleg, and there were 2 possible reasons for this, the first being that there was some kind of structure present before the road was built and they went around it, or alternatively the ground at the end of the field was not suitable for the road (even today that part of the field is waterlogged and tractors leave large wheel ruts) (Pers com M. Stroude 04/10/05). Our first evaluation trench (trench D) was put across the geophysical anomaly which we thought was the dogleg of the road. Once the trench was opened we immediately saw two types of surface (Fig 12). 14

Figure 12: Trench D showing the two different types of metalled surface, 1 & 2 Photo taken by Rhw Mitcheson. At the North East end of the trench No 1 surface was found which were sub rounded patinated flints in a variety of colours with some CBM in situ, the majority of flints were round beach pebble type flints. The other surface No 2 also had sub rounded flints but they had iron rich clay mineral patination (brown), this surface was sterile of finds. We put a sondage through this surface to see the depth which was 250mm, the river gravels continue for 1.5m although they are not metalled surface (Fig D). Once we had recorded the trench we hand dug out the first surface to see the profile in section. If we indeed did have a Roman road, then we would expect to see a clay agar, and flint foundations; this was not the case, however the trench I excavated in Pound field only had the foundation level of the road, the rest had been ploughed away. The second metalled surface has been identified following research at Lewes district council record office. I viewed the tithe map of 15

Barcombe dated 1842, which showed Culver Meadow (Culvermead s previous name). At this time the field was being used as a meadow and in the South East corner of the field the map shows a pond (Fig 13). Figure 13: Tithe map of 1842 showing a pond in Culver meadow. East Sussex Record Office. The 25 OS map of 1873 then shows the pond as a gravel quarry (Fig 14). 16

Figure 14: Map of 1873 showing a gravel pit East Sussex Record Office Whereas the OS map 6 first edition which was surveyed between 1873-1875 shows a depression where the quarrying had taken place (Fig 15). Figure 15: First Edition 6 OS map showing depression rather than a quarry. East Sussex Record Office. I am now convinced that the second surface was a road surface laid in the 19 th century for quarrying activities, and furthermore it appears to have been in operation for a relatively small time period. Moreover the plans for the railway extension from Culver Junction to Lewes and East Grinstead where put forward in 1873, it is quite possible that the gravel quarrying activities where in fact for use in constructing the railway, they created large banks on which the lines where laid. Also the geophysical anomaly which we now know to be the quarry road leads to the where the railway line was laid (East Sussex Record Office 17

05/11/06). It appears that the anomalies on the geophysical survey to the south of the site are all to do with the quarrying activities. Trenches C, E, F all have similar metalled surfaces, as found in Trench D. And thus it seems likely that relate to C 19 th quarrying activities. In Trench A we also found a possible metalled surface, similar to what I described as the foundations of the Roman road in Trench D, it also contained CBM. I do not believe this is a road surface or foundation, although it is a metalled surface which could be used as a track leading to the road, or it could be part of a court yard. The reason for this conjecture is the other archaeological deposits found in Culvermead. Trench B was opened to the North West of Trench A, we cleaned the surface area, and we found a cut feature measuring approximately 3m in width and continuing under both North and South baulks. When excavating the feature it dropped down steeply and is flanked either side by sandstone. This then descends almost vertically down to gaunt whitish grey clay. At first it was thought that this was natural deposit. However it was decided to excavate further to confirm whether or not the clay was natural. Approximately 150-200mm down we came across a piece of wood. We excavated across the wood which turned out to be 1.524m in length. We left the wood in situ. One problem we incurred was that the trench was filling with water from the bottom upwards (this in fact was a blessing, as it was the underground water supply that was preserving the timber). We continued to excavate around and below the timber when we came across a 2 nd timber and then a 3 rd. It was decided at this point to stop excavating as we had no contingency plans for dealing with waterlogged wood, for example with 18

conservation, and storage etc. Furthermore it was decided after conversations with Peter Drewett, David Rudling, Chris Butler, Luke Barber and James Hale to leave the wood in situ until the full excavation takes place which is planned for August 2008. By this time a conservation and excavation plan will be put in place. Prior to the trench being back filled we took a core sample for dendro dating. This is a process where tree rings are counted and compared to a database of known tree age dates; there are different databases for different countries and areas around the world (Renfrew & Bahn 2001 p 134-135) (Wallace 2006 p19-20). Unfortunately there were not enough rings in our sample to date the wood. 70/80 rings as a minimum are required to date a sample, and ours only contained 30/40 (Pers com M. Bridge UCL 15/09/06). Furthermore we took a sample for C14 dating, this has been sent to Bata Analytic Inc, Miami, USA. The results are still pending. There is a need to date these timbers; although the pottery found in the same context is Roman in date it is possible that the timbers are earlier in date, and through the process of bioturbation have encroached into an earlier feature (Wallace 2006 p19). There could be several possible functions of the feature. The first suggestion was put forward by Peter Drewett who suggested that it could be a wood lined well; these wells date from the Bronze Age onward and through the Roman period (Pers com P. Drewett 24/8/06). Some of the best preserved examples of wood lined wells come from Rhineland and Britain (Hodge 2002 p52). The Roman wood lined well would normally take 1 of 2 forms and in some instances both forms. The first form was a square wooden frame made from 19

planks; these would be made individually and then stacked upon each other. The second being round, which were normally barrels which had the bases taken out then stacked upon each other. Moreover both examples can be used in the same well; an example from Lime Street London has a square construction at the top then changes to the barrel type (Fig 16+17). Figure 16 & 17:16 is from an example from Germany & 17 Is from Lime Street London, the latter has the square top changing to the cylindrical (Hodge 2002 p52). An argument against the preserved timbers in Culvermead being a wood lined well is that they are complete timbers (Fig 18). 20

Figure 18: The waterlogged timbers discovered in Context (314) in Trench B. Photo taken by the Author. Whereas the examples shown in (Fig 16+17) are of plank construction. This could suggest that they are an earlier date. With anticipation of the C14 date result, this dilemma will be answered. Furthermore the timbers at Culvermead do have evidence of being worked. There is definitely a cut in the top of the 1 st timber possibly a mortise joint. Also cut marks were found on the 2 nd piece of timber. Moreover circumstantial evidence within the trench raises further questions. Just to the North West of the main feature we have excavated a terminal of a linear ditch/gulley which continues under the bulk, continuing upslope. Why would they have a ditch/gulley next to a well? (this is presuming that the two features are contemporary, the reason why I believe that they are contemporary is the two 21

fills are very similar and the pottery sherd found in the linear is of the same date as the pottery found in the main feature) The ditch/gulley will be discussed later. If we take the approach that this site is industrial in nature, then we have to ask what was being produced. Pottery has been found field walking, and in the evaluation trenches as has iron slag especially in Trench G. We can look at the industries in the area at the time of the Roman occupation. Iron workings were a major part of Roman life in Regni (now Sussex) (Fig 19). Barcombe Figure 19: Historical OS map of Roman Britain showing iron working sites near Barcombe Crown Copyright 2001. Kilns have been found at road site settlements such as at Sedgefield (Web 1). The fill of Context (314) where the waterlogged timbers were found, is gaunt clay, possibly useable for pottery or CBM production. In the Netherlands part of the clay preparation process was to sour the clay, this process require leaving the clay in a cool damp place. In Britain they normally used clay lined pits to do this, whereas the Dutch used wood lined pits for this process. It could be possible that wood lined pits are an imported process from the Netherlands? Has the timber at Culvermead still got its clay from this process 22

contained in it? (Peacock 1982 p54) (Wallace 2006 p2-3). If this is a wood lined clay pit, then possibly the ditch/gully is part of this process. If we return to this being a wood lined well, I asked the question earlier why would they put a ditch/gulley next to a well up hill. In all fairness we do not know the levels of the ditch/gulley it might well have been running away from the well, maybe there is an industrial area up hill which required large amounts of water. It maybe so they transport the water from the well and flow it to other parts of the site using the ditch/gulley. In Trench G we found what we believe to be a possible collapsed building or structure, a large amount of CBM was found 55 separate pieces weighing a total of 4129g in two contexts (600) (601). CBM in the quantity that we have in Culvermead could indicate the presence of a building or structure (Darvill and McWhirr 1984 p240). As well as the CBM, we found a large amount of sub rounded flints; these could be part of a flint foundation. They may form a possible linear edge to a feature in this trench. In combination with the other anomalies on the geophysical survey which have not yet been excavated, the overall picture looks very interesting (Fig 20). 23

Figure 20: Anomalies still not excavated. CONCLUSION. The excavation at Culvermead was a resounding success, although as is often the case with archaeological evaluations it has raised more questions than answers. That said the questions it has answered have proved invaluable. We now know that there is Roman archaeology present in Culvermead. In August 2007 I will be carrying out an open excavation in Pond field, this will hopefully give us an orientation of the Roman road, and give us an idea of the scale of the possible roadside settlement. With this information we can project the 24

road alignment into Culvermead, this will then hopefully confirm the possible Roman road found in Trench D. We now know that the road is the one that Margary believed existed but was unable to find (Margary 1933 p32). Due to the crops grown in Culvermead, which will be sweet corn in 2007, it is planned that a full open excavation will take place in Culvermead in August 2008, which will allow sufficient recording and interpretation to take place. If the feature in Trench B is a wood lined well, then shoring will have to be considered, or a stepped method of excavation. Furthermore conservation issues will have to be discussed, as well transport and storage problems. Also a full open excavation will answer a number of other questions, such as the nature of the feature in Trench G, what is the metalled surface in Trench A, where does the ditch/gulley go in Trench B, and what are the anomalies which we have not excavated. This work will optimistically give us a better insight to the way of life in Roman Sussex, and importantly we are not just dealing with a stand alone villa complex, because we can now see how the Romano-British were using their environment, and hopefully what they were doing in their landscape. ACKNOWLDGEMENTS. I would like to special thanks to all my colleagues who helped with this project without whose help it would not have been possible.(thanks you guys) Rosie Cummings, Rhw Michelson, Tina Paphitis, Charlie Hammond, Seth Wheeler, Hannah Lewis-Perry, Steve White, Wayne Saunders, Nick Carter, Mr Derek Wise for the use of his shed and garage for finds processing. Mark and Harold Stroude for their kind permission to allow me to dig on their land. Peter 25

Drewett, David Rudling, Luke Barber, Chris Butler, for their guidance during the excavation, Clive Meaton for his proof reading, and guidance. (Wallace 2006 p39). Also Fiona Griffiths for her help in digitising my plans and sections. 26

BIBLIOGRAPHY. Bedwin, O. Place, C. 1995. Late Iron Age and Romano-British Occupation at Ounces Barn, Boxgrove, West Sussex; excavations 1982-83: Sussex Archaeological Collections.133. P45-101. Butler, C. 2000. An Interim Report On The Work Carried Out At Barcombe, East Sussex, In Spring 2000: By The Mid Sussex Field Archaeological Team: Unpublished. Butler, C. Rudling, D. Gammon, A. 2006. Barcombe Roman Villa: Andy Gammon. Lewes. Darvill, T. McWhirr, A. 1984. Brick And Tile In Roman Britain: Models Of Economic Organisation: World Archaeology V15 no 3 pp239-261. Davies, H. 2002. Roads In Roman Britain: Tempus Publishing Ltd: Gloucestershire. Hodge. T. A. 2002. Roman Aqueducts & Water Supply: Duckworth: London. Margary, I. D. 1933. A New Roman Road To The Coast: Sussex Archaeological Collections Vol LXXIV: W. Heffer: Cambridge. P16-43. Margary, I. D. 1935. A Roman Road From Barcombe Mills To The West, Through Streat And Hassocks: Sussex Archaeological Collections Vol LXXVI: Oxford University Press: Oxford. P6-34. Margary, I. D. 1965. Roman Ways in the Weald: Phoenix House: London. Peacock, D. P. S. 1982. Pottery In The Roman World: An Ethnoarchaeological Approach: Longman: London and New York. Renfrew, C. Bahn, P. 2001. (3 rd Edition) Archaeology: Theories Methods And Practice: Thames & Hudson: London. 27

Wallace, R. 2006. Prepare A Fully Illustrated Excavation Report Based ON Your Sample Excavation. Present With Associated Archive: CCE: University of Sussex: Unpublished. WEBSITES. Web1http://www.channel4.com/history/microsites/T/timeteam/2003_sedgefi eld.html Site accessed 3/12/06. MAPS. East Sussex Records Office. Lewes. East Sussex. Ordnance Survey. Romsey Southampton. UK. 28