Mixbury: Site of an Iron Age Hillfort?

October 2020

Context for Mixbury's location

Note: We have included at the end of this Article a skeleton map to help readers identify various locations referred to in it.

1. This Article explores why Mixbury probably developed where it is and examines the case for Beaumont Castle being located on the site of an Iron Age Hillfort. The Summary in the website touches on several factors which indicate that Mixbury existed in an evidenced Iron Age lanscape - in fact, before that. We have referred to the number of "finds" from the Bronze Age through to the Roman period within a five mile radius from the village, and to assemblages of pottery sherds found in the parish from the same period; to aerial photographs showing crop-marks in the parish; and to evidence which points to the probability of an Iron Age track running through it.

Examples of Mixbury Crop Marks

Bronze Age Round Barrow
Undated irregular Banjo Enclosure

In terms of its geography we have concluded that the parish in the Iron Age (of course, it wasn't a parish at that time!) lay close to the border between the Catuvellauni and the Dobunni tribes, sometimes thought to be the Cherwell river; and we have highlighted on the village's western side a line of springs which fed a stream that runs to the west and north of the village. The area's topography shows that the village sat on a higher spur of land that fell away to the west, the north and the east.

Examples of Mixbury pottery sherds

2. Mixbury is not the only parish or village in this area to have one or more of these characteristics and have access to water and game, and the wherewithal for arable farming (with manuring as essential !) But it is unusual in having the combination of all the factors in paragraph 1 above. We think that three stand out in accounting for the siting of Mixbury and why there was fortification here, at least until the early 15th century. Those are

  • The stream now running from the springs to the west of the village. In living memory heavy rains have seen large standing pools of water in the west. The stream, which runs northwards with a line of springs feeding it before it veers to the east, to the north of the fortifications, is thought to account for marshy conditions around it just as it turns east and, in fact, a 2018 OS map has the symbol for this feature at this point. It also powered the village's water mill, referred to in the Domesday Book (1086) and fed a water- meadow, both again before it turns east. And in Medieval times , in its extreme north-west, it ran into fishponds. The man-made features referred to here clearly post-date the Iron Age by quite a margin but, unless the geological conditions have altered dramatically over time, it is reasonable to assume that a stream/small river and marsh ground has always posed an obstacle to carts and livestock approaching what is now Mixbury from the west and north-west.

  • The Iron Age Track. Arthur Beesley, writing in 1841, referred to "the SALTWAY, an ancient road which yet exists under the name at Banbury". (The History of Banbury). In 1955 Ivan Margary (a renowned expert on Roman roads in Britain) identified the course of a Roman road - he numbered it 56a - running , initially, from Droitwich in a south-easterly direction before it proceeded further east , eventually joining what became Watling Street. We think the Iron age track and the Roman road are one and the same. A route for 56a, provided to us by Historic England, projects the road passing directly through Mixbury and, although we have yet to have the precise route confirmed, we have traced it from Evenley to the north-west of Beaumont Castle. We believe it then crossed the "water-hazards" referred to above where the stream is now culverted under the road, just after the sharp right-hand bend to the village after Mixbury Hall. It then proceded east on a terrace up the right-hand side of the slade (a shallow "valley" - which lends its name to Slade Hill) but after that we are not yet sure which path it took.

A crossing point for a track carrying a valuable commodity would have had some significance, and there is no doubting the value of salt. "The Droitwich salt-making industry is exceptional in preserving evidence for salt production from the Iron Age until the 19th century. (From Multi-period salt production remains in Droitwich", Historic England, online. In another website they say"In Britain, organised production dates back to the Bronze Age, about 1200 BC, at least. It developed through the Iron Age into the Romano-British period, where distribution increased geographically". It is reported that whilst salt was moved from Droitwich (also along tracks and roads to other parts of England), the wagons which returned were not empty as they frequently carried wood which would be used in the manufacture of the salt. The Romans are renowned for having built new roads but in the case of 56a, which had no apparent military function, they would have utilised and improved the Iron Age track to provide continuity in the distribution of salt from Droitwich.

  • Topography. The third important factor is topography. The contour map below shows the three critical factors together: it includes the height in metres above sea level. It evident that the fortification which we know today as Beaumont Castle was in a significant position as far as the crossing point is concerned. Historic England (online) observe "The site of a castle was often chosen to control a strategic location, such as a ..... river crossing. Many of the prime sites had been selected for similar reasons, for their prehistoric and Roman predecessors". Although the Castle site was nominally only 6 metres above the crossing, man-made features such as embankments, palisades and watch towers would have cosiderably enhanced its sight lines and defensive capability

Evidence for a Mixbury Hillfort

3. We cannot say unequivocally that there was a hillfort in Mixbury: a future step is to have this work assessed by academia and archaeologists in order to test our work. But, like a jigsaw puzzle, we have pieced together information which already existed in a variety of sources to produce a picture which was not clear before. Several of these items of information were included in the late Adrienne Brunton's thesis for a Postgraduate Certificate in Archaeology in 2012 - "Beaumont Castle, Mixbury, Oxfordshire: a study of its site and earthworks". In this Adrienne wrote "..... by examining the present surrounding landscape and that of the past, as recorded in old maps, I intend to provide evidence for the fact that Beaumont Castle may not have been the only monument on the site". This she did, although she did not go on to assemble the information to reach a firm conclusion, preferring to point out the need for "proper archaeological investigation", which is where we started this paragraph. The pieces of Mixbury's hillfort jigsaw are

  • The context given in paragraphs 1 and 2.

  • The 1880s OS map of Mixbury, and a "modern" photograph of a feature shown on it.

  • A diagram and note prepared in 1974 by an officer of Oxford City and County Museum; Field Department, about what lies under Evenley Road adjacent to the Castle site.

  • Consideration of the verges on either side of Evenley Road, adjacent to the Castle site.

  • A Field Investigation note written in 1986 by Dr I Burrow and J M Steane. (J M Steane was the County Archaeologist for Oxfordshire between 1976 and 1990).

  • A crop mark on a 1984 aerial photograph, and a 2018 satellite image of adjacent field markings.

1880 OS Map

4. The relevant section of the map is set out below. We have added two numbers to flag up key features which we try to interpret below.

We think that 1 is the remains of a "lipped" entrance to the main fortified site: the northernmost of the two lips is shown as intact, but the southern one seems to have been destroyed. An English Heritage website, entitled "How to Spot: a Castle hiding in plain sight", includes the advice "Look for traces of the original entrances into the castle's enclosures - the banks may flare out, creating a longer passageway you'd be forced to travel through". The photograph below, taken looking south towards the feature, shows the northern "lip"; and a bank rising towards the rear of the houses on the west side of Evenley Road.

The following illustration of the entrance to an Iron Age Hillfort at Castell Henllys in Pembrokeshire shows the "longer passageway" to which English Heritage are referring. You can see from it why we have used the analogy of "lips". The area in the blue ring is what we think the photograph above shows, and it equates to the feature marked 101 on the 1880 map, by number 1.

The tip of feature 1 is an estimated 50 to 60 meters to the west of the gap between the northern and southern "courts" shown on the 1880 map which is thought to have been the entrance to Beaumont Castle: feature 1 and the gap line up nearly perfectly. The "longer passageway" occurs where there is more than one ditch around a fortified site. Simply, the entrance has to be beyond the last bank: an approaching road or track cannot run between an inner and outer bank because it would breach the defensive value of the outer bank. Feature 2 on the 1880 map underscores this point. It represents a track from the stream-crossing referred to in paragraph 2 of this Article. It does not approach the fortification's entrance directly but runs a few metres to the west of it.

One thing that we can derive from this analysis is that when the original fortification was in place, its western side had at least two ditches and banks: the technical term for this is bivallate. What we can see today is what we call the moat, and the bank on which the 1972 houses were built, but there was more to the west of this. Evenley Road as we know it today did not exist.

1974 Sewer work

5. Set out below are copies of field notes when observations were made by an Oxfordshire Museums Fieldwork officer, about a trench being dug along Evenley Road in order to lay a new sewer.

The diagram reports the new sewer being laid 3 metres below the road level, but does not say how deep or wide the V-shaped ditch was: neither does it say whether the ditch extended for the whole length of the sewer work. However, given the uniformity of the front garden slopes on both sides down to the road (see below), for the length of the sewer work, it is reasonable to assume that the V-shaped ditch ran for the same length. The majority of the village is located on oolitic limestone which lies close to the surface. We know from householders along the road just to the north and east of the point Mr Hassall flags up, that the bedrock there lies only a foot or so beneath the edge of their front gardens. A V-shaped ditch could be either shallow or deep, but to some extent that does not matter. The important thing is that someone had taken the trouble to cut a ditch at all, through solid stone: there must have been an imperative for doing so. It went down on the east side of the road and up on the west. This supports our observation above that there was a second ditch to the west of the 1972 houses, and this is borne out further by consideration of the east to west profile of the front gardens of the houses on both sides of the road adjacent to the Castle site.

Consideration of front garden slopes

6. Even before we had begun piecing together a picture from the 1880 map and the sewer-work report, we had given some thought to why the verges on either side of Evenley Road at this point are significantly higher than the road itself: and why, after the northern edge of the Castle earthworks, the verge to the east of the road is even higher but almost flat to the west. The photographs here, taken at various points along this stretch of road, illustrate the point.

Looking south from a position south of Church Lane. Relatively flat both sides.
Looking south about half way along the road adjacent to the Castle. Slopes both sides down to the middle of the road.
Looking north along the road adjacent to the Castle. Both sides sloping down.
Looking north from the road to the north of the Castle site. High verge to the east, flat to the west.

We are now fairly sure that the slopes shown in the second and third photographs are the sides of a ditch, the bottom of which was the V-shaped feature under the road. Before that, we had wondered if the slopes were part of a "hollow way" or the course of a long-gone stream which had cut its way into the land. A hollow way was a track worn down by constant usage over a long period of time, or deliberately cut out on both sides to facilitate the control of animals being moved along it. But it seems too wide compared with images of hollow ways we have seen elsewhere, and in the parish; and the verges are not steep enough to stop livestock wandering off the road. The stream idea doesn't really work because there is no spring in the village which would have fed it. The surface water in the village comes from the springs to the west and has done for over 200 years. Of course, 200 years is nothing in the life of the land, but the profiles of the two sides of the road shown in photographs 1 and 4 do not seem to come from a natural erosion process.

1986 Field Investigation

7. The note below was written by Dr I Burrow and J M Steane.

J M Steane was the Oxfordshire County Archaeologist between 1976 and 1990. Without piecing together in writing the information set out in this Article, he and Burrow nevertheless conclude about Beaumont Castle that "There may be a suggestion that there was a pre-Roman Iron Age earthwork on the site which has been adapted by overlying a feudal castle over the most easterly part". J M Steane wrote extensively about Oxfordshire, notably in the 1996 Pimlico County History Guide for the county. There he wrote "The most impressive visual expression of the Iron Age inhabitants in Oxfordshire are without doubt the sixteen hillforts". Is Mixbury the 17th, assuming no other ones have been found in the last 25 years? He was an expert in his field and what we now need is for other experts to come forward and assess this Article.

8. Putting together all the information from paragraphs 1 to 7, we have drawn a cross-section of what the Iron Age fortifications might have looked like to the west of what is now the earthworks of the northern court of Beaumont Castle. We have done so from the southern edge of the north court westwards as far as feature 1 referred to above.

The work includes a certain amount of estimation, but doing it has required us to think about various things. For example, we had to think why the slope either side of the 1st defensive ditch, and that up the 2nd embankment were so shallow compared with many hillforts, such as Maiden Castle or Hod Hill. We came up with two answers. First, because the surrounding land is not much lower than the Castle site it did not facilitate the dramatic tiers of embankment that are the mark of "true" hillforts and which give the defenders the significant advantage of height above the attackers. Second, in the Mixbury situation, a higher outer embankment would have created a blindspot behind which attackers could shelter: the gentler incline was necessary to give defenders on palisade walls a clear line of sight and armed with bows or slings they could still wreak havoc particularly on a densely packed group of men. An online review of "Iron Age hillfort defenses and the tactics of sling warfare" (2016) by Peter Robinson concludes that "With bivallate forts the defenders had an advantage. Indeed, attackers stood little chance of reaching and scaling the inner rampart without having been hit from an equivalent force of defenders - though protection with shields and armour might have mitigated the damage"'

1984 aerial photograph and 2018 satellite image

9. If the bivalate fortifications to the west of the village are part of an Iron Age hillfort, were there similar fortifications to the north, east and south? The first consideration is whether there needed to be. We have referred in paragraph 1 of this Article to the proximity of Mixbury to the probable boundary between the Dobunni and Catuvellauni tribes being the Cherwell river: that would place Mixbury on the east side of the river in Catuvellauni territory. These two tribes were not known to have lived peaceably, with the Catuvellauni often portrayed as the aggressors. An incursion by the Dobunnis westwards along the Iron Age track would bring them to the strategic crossing at Mixbury. Why would they have used the track? And why wouldn't they simply have outflanked the fortification?

We think the answer to these questions lies, at least in part, in what lay around Mixbury. Although to the south and east the land is now relatively clear of trees, that has not always been the case. The Domesday Book recorded 15% of England as woodland, but the Iron Age is 1000 to 1500 years before that when there would have been more woodland still. And woodland, like water, could be an obstacle. Steane writes that " ..... Oxfordshire posesses the largest sub-circular territorial enclosure in the British Isles. This is North Oxfordshire's Grim's Ditch...... A number of gaps in the Grim's Ditch seem deliberate. They probably co-incide with areas of uncleared woodland which would have been considered inpenetrable at the time". As late as 1851 the census for Mixbury records 7 men who were in jobs associated with the timber business - a woodsman, a timber sawyer and 5 carpenters. The woodsman is shown on an 1868 poster as a contact person for the sale of over 100 lots of poles and underwood from "Pits Cover", still wooded, to the south of Mixbury.

Taking a "warband" by existing land routes would have been much easier and faster than slogging through uncleared woodland when the elements of surprise and speed would almost certainly have been lost. So a fortified site overlooking a track crossing a "water hazard" was at a premium. And trying to outflank Mixbury to the south and east would have taken the Dobunnis through other Catuvellauni territory.

The disposition of woodland offers one possible explanation for the apparent lack of 360 degree bivalate fortification of Mixbury. Another explanation is that it is there, but we haven't found it yet, with one exception.

A 1984 aerial photograph shows a curvilinear cropmark in the field to the north of Beaumont Castle: the field is now divided in two by a hedge - the dotted line - towards its eastern side. The open side of the cropmark mostly abuts the northern bank of Beaumont Castle, and the western side runs down to the apparent northern end of the Iron Age fortifications discussed above, where the 1972 houses were built. A 1984 report (now online) by the RCHME - Royal Commission on the Historical Monuments, a predecessor of Historic England - titled "The Classification of Cropmarks in Kent" identified a category of "curvilinear regular enclosed settlements i.e. curvilinear and symetric, but not circular, sub-circular or oval". That is the shape of the feature below.

The RCHME report says that such features over 1400 square metres are from the Iron Age: the area in Mixbury under the curve as far as a line level with the northbank of the Castle site is estimated at over 12,000 square metres. Using the RCHME criteria the feature is Iron Age, although we can not tell at this stage whether it was part of the fortifications or an enclosure for domestic purposes e.g. to contain cattle. Either way, it is significant, not least because it is roughly half the width of one of the 1972 houses, about 5 metres. Interestingly, there is a suggestion that the land between the western edge of the curve and the road was built up. The fourth photograph in paragraph 6 above illustrates this.

10. Three years ago, Google Earth carried the following Satellite image. The area shown is the easternmost part of the field under the curvilinear feature, marked off in the 1984 photograph by the dotted line.

The image shows a series of "Marks" on the field. Most are round, but some are undoubtedly straight lines. In order to disassociate those of a regular shape from "background" marks we have traced over the main elements.

HS2's archaeologist has suggested to us that these marks may be "fairy rings" - a type of fungus that grows in circles on grassland. However, comparison with images of fairy rings elsewhere show fundemental differences from the Mixbury features e.g. there are virtually no linear features in other images; the Mixbury circles show little blurring at the edges, whereas others elsewhere do; and Mixbury circles come in different sizes! Also, in the Mixbury image there is a clear concentration of the marks inside the curve of the curvilinear feature: the field to the east of the curvilinear feature contains virtually none. In fact, the eastern side of the curvilinear feature bends back towards the west, and this curve is replicated by the position of the marks in the satellite image.

So what are the Mixbury circles and straight lines? Wikipedia reports "Round houses were the standard form of housing built in Britain from the Bronze Age throughout the Iron Age, and in some areas well into the [Romano-British] period. ..... They ..... ranged in size from less than 5 metres to over 15 metres". The largest circles in our image are over 5 metres in diameter. From English Heritage we know "The Romans introduced the idea of rectangular plans". ..... "In the countryside rectangular farmhouses began to appear alongside traditional roundhouses". It seems reasonable to interpret the smallest circles and squares are storage pits.


11. Most towns and villages are located where they are for a reason although the further we go back in time the more difficult it is to pinpoint that reason. We think that we have a robust case for concluding that Mixbury's origin as a fortified settlement came about through its location as a bridging point. Its situation as a guarded settlement might also have made it an ideal stop-over site for ox-drawn waggons using the Iron Age track: oxen could only travel at an average speed of 2 mph and so long journeys had to be planned with overnight stops in mind. Ox dung is, apparently, very fertile, and a regular supply of oxen resting at the settlement could have provided an explanation for the "mixen" -Old English "dung heap" - part of Mixbury's (Mixen - burgh) name. The marks on the 2018 satellite image may well show an Iron Age settlement; the Iron Age pottery sherds found just to the east of the Castle site may be evidence of that settlement.

The housing built in Mixbury in the 20th century has undoubtedly destroyed much of the archaeology which could have been used to verify the assertion that the western earthworks of an Iron Age hillfort extended further west from those of Beaumont Castle and beyond the houses on Slade Hill. But we think there is enough documentary and remaining physical evidence to sustain this assertion. Whilst the 2018 satellite photo shows there may have been a large ditch to the north - which would have had some defensive value - we do not know what Iron Age fortifications, if any, existed to the east and south of Mixbury and we must now apply ourselves more vigorously to look for clues. As far as the north is concerned we have a farmer who is interested in history and who has kindly given us permission to undertake geophysics in his part of the field, to the west of the dotted line/hedge. Adrienne would have appreciated that and we leave the last word with her. "It is important for understanding the development of Beaumont Castle and the village that a geophyisical survey of the semi-circular cropmark is done first".

Skeleton map of relevant places referred to in the Article.

Skeleton map of places in the village referred to in this Article.

Help Needed

If you think you can help to answer the questions below please complete the below form.

This question is really for people who live in the village and its aim is to help us find the "missing" south and east fortifications. Do you have any sudden unexplained drops or rises in height in your garden or land, or between you and a neighbour?