The most enduring theory proposes a Trackway, dating to the Neolithic Times, running from Wanborough in Wessex to East Anglia; this theory has its origins in Medieval Literature and the idea of 4 Prehistoric ‘Highways’ that crossed the Country, (Ermine Street, Fosse Way, Watling Street & Icknield Way) 1st included in a fictional work by Henry of Huntingdon in c.1151 The exact definitions of this Track have varied considerably from an ephemeral Route covering a broad area of land to an identifiable linear from which a larger pattern across the Chilterns at least can be discerned. The purpose claimed for such a long distance Trackway has always been for Trade and the free movement of Goods. The Icknield Way has been equally claimed & refuted as a Roman Road and again theories have varied considerably with some proposing it as an earlier Trackway reused by the Romans (Viatores) while others claim it is only Roman in sections. More recent discussions refute the Way as a Roman Road altogether. Other theories have suggested the Way to be Post-Roman at least; its path originating sometime between the 5thC AD and the Medieval Period. The evidence for Trackways will always be problematic and open to debate. Place name evidence for the Icknield Way is limited to just 5 Sites, one in Wiltshire, 3 in Berkshire and at Princes Risborough. Documentary & Map evidence from the Medieval Period can be traced back to the Medieval Literature and so must be treated as problematical. A Medieval Origin for the Way is considered unlikely given that it avoids all major and most minor Medieval Urban Centres making it an anomaly in the Medieval Road Network. Archaeological evidence is similarly scarce and equally open to interpretation. Some evidence for Prehistoric Hollow Ways & Earthwork Traces have been found across the Chilterns at isolated Sites such as Pitstone Hill where the Prehistoric Tracks identified were overlain by a Roman Road and at Whipsnade Downs, Bedfordshire.
These are, however, isolated areas and may represent more localised Tracks rather than a broader Zone of Communication at a National Scale. Recent theories have suggested that the Icknield Way did not comprise a single cohesive Trackway at a National Level but rather Tracks of a Regional or even Local Level that became popularised in the Medieval and later Periods. Significantly, recent excavations at the Aston Clinton Bypass along the supposed Route of the Lower Icknield Way found no evidence of a Prehistoric or Roman Route at this point. The only substantial section of the Icknield Way with any evidence supporting a Pre-medieval Origin then is the Section from Wanborough in Wiltshire to Risborough.
Dorchester to Cuxham
Janet Sharpe & Phil Carter – Introduction
The Roman Road known as the Lower Icknield Way was traced by Margary (1973) from where it branches off from Akeman Street near Aston Clinton in Buckinghamshire as far as the Parish of Pyrton to the Northeast of Brightwell Baldwin. There he lost the Trail but speculated that the Road continued on through Cuxham & Brightwell Baldwin towards Wallingford. His work (which was originally published in 1955 and not updated in the 3rd Edition of his Book) was continued by Morris et al. (1968) who followed the Lower Icknield Way in some detail, mostly from the alignment of existing Tracks & Hedgerows, through Cuxham, Brightwell Baldwin & Berrick Prior to Warborough, where once again the Trail was lost and the Road was assumed to continue into Dorchester. Malpas (1987) added a few more details to the Dorchester end of the Lower Icknield Way and referred to the Agger (the central part of the Road) as being clearly visible where it crosses Green Lane immediately to the West of Warborough. He considered that the Road then presumably continued across the Fields towards Dorchester. The course of the Road through Cuxham is described by Morris et al. (1968) as ‘obscure’ but it was thought to follow the line of the modern Road into Brightwell Baldwin. Here the Authors cite a Saxon Charter of AD 887 which refers to the Mead Land between ‘egsaforda’ (Egsa’s Ford) and ‘strætforda’ (Street Ford). The former Ford has been identified where the Road running North from Cuxham Village crosses a Stream near the Parish Boundary. Morris et al. (1968) speculated that the Street Ford was to the South of this and that the name indicated a Roman Ford (paved) crossing the Stream flowing North through Brightwell Park, where it now forms a narrow Lake, and that the contours suggest that this Ford lay ‘immediately on the Southside of the [modern] Road close to the Rectory’, despite the fact that their Map clearly shows the Roman Road running parallel to and just North of the Modern Road.
Other evidence of Roman activity was discovered in Brightwell Baldwin in the North of the Parish during the 18thC (Clarke, 2001). A possible Burial containing Urns & a square glass jug (now lost) was described in 1705 from the South Bank of the Stream that now forms the North Boundary of the Parish near Cadwell Farm. A Roman Coin Hoard was found close by in 1759. More recently, 11 Roman Coins were included among ‘garden finds’ from Cadwell Farm. It is interesting to note that in this same area 2 Fields adjoining the Parish Boundary on the Chalgrove side are named ‘Stratford Meadow’ and ‘Stratford Furrowlong’, suggesting that the ‘strætforda’ referred to in the Saxon Charter may have been near here, rather than further South. So where did the Roman Road run through Brightwell Baldwin, and was there perhaps more than one Road? As part of the proposed Brightwell Baldwin Landscape Archaeology Project, we tried to find out.
Roman Road Profile -The arrows indicate the position of the dowsing responses
In the Centre a Carriageway was built on a raised Agger after stripping off soft topsoil, using the best local materials, often Sand or Sandy Gravel. The 2 strips of ground between the Agger and the Boundary Ditches were used by Pedestrians and Animals and were sometimes lightly metalled. The Agger was sometimes, but not always, bordered by deep ditches to take Rainwater and keep the Road Structure as dry as possible. The metalling was in 2-layers, a Foundation of medium to large Stones covered by a running surface, often a compacted mixture of smaller flint & gravel. About one ¼ of Road Pavements were “bottomed” with large Stones, mostly in the North & West where Stone was more readily available. Some high-status Roads in Italy were bound together by Volcanic Mortar, and a small minority of excavated Sites in Britain have shown Concrete or Limestone Mortar. Road Surfaces in the Iron-producing Areas of the Weald were made from Iron Slag. The average depth of Metalling over 213 recorded Roads is about 51cm (20ins), with great variation from as little as 10cm (4 in) to up to 4M 13ft) in places, probably built up over Centuries.
We had previously used dowsing to trace the route of the ‘dog-leg’ formed as the Roman Road from Silchester crossed the Sinodun Ridge and there abruptly changed course to cross the Thames downstream of its confluence with the Thame and link up with the Roman Road approaching Dorchester from the Henley direction (Sharpe & Carter, 2003). This Study showed us that whereas surviving Trackways & Hedgerows can indicate the approximate line of a Roman Road, they do not show the exact alignment as both Paths & Hedgerows tend to wander over time. Dowsing, the same Technique that is used by Water Diviners and some Water Board Employees for detecting underground water, can indicate the interfaces between wetter & dryer features in the soil. The exact position of this perceived interface will depend on the water content of the soil and the individual response of the Dowser, but working together as a Team we find that our independent measurements are seldom more than 0.5M apart (our margin of error) and we are usually able to corroborate each other’s results. The classic profile of a Roman Road shows a Central raised Agger flanked by parallel Ditches, usually one but sometimes 2 on each side, and usually separated from the Agger by a flat ‘berm’. Walking a transect across a Roman Road will provoke a dowsing response at the points shown by the Arrows. By flagging the points where the Dowsing Rods cross & uncross, the Road Profile or ‘signature’ can be plotted on the ground and measured. Wherever possible, Dowsing was conducted where a Public Right of Way shown on the Ordnance Survey map crossed the assumed Road line more or less at right-angles.
The Maps used in this study were the OS 1:25,000 Explorer series, No. 3 (1994), Chiltern Hills South, and No. 170 (1999), Abingdon, Wantage and Vale of White Horse.
To survey the Lower Icknield Way between Dorchester and Cuxham, we located and usually measured transects across the Road at 20 Points along its Route:-
Junction with the Dorchester Roman Road Network West of Warborough (1);
Green Lane at Warborough (2);
Warborough High Street (3);
The Green at Warborough (4);
The Footpath to Ladybrook Copse (5);
The Westside of Ladybrook Copse (6);
The 2nd Field Boundary West of Berrick Salome (7);
The Road between Berrick Salome and Berrick Prior (8);
The North curve of the Hollandtide Bottom Track just to the East of Berrick Prior (9);
The North flank of Hollandtide Bottom along a North-south Bridleway (10)
Again on the Road leading to Lonesome Farm (11);
The Bridleway Crossroads at Whitehouse Farm (12);
The Road to Cadwell Farm at Brightwell Baldwin (13);
The Drive in Brightwell Park (14);
The Lane immediately East of Brightwell Churchyard (15);
The Rectory Garden at Brightwell Baldwin (16);
Turner’s Green Lane (17);
The East side of the adjacent Field to the East of this Track (18);
Cuxham Churchyard (19);
The Field Northwest of Chestnut Farm at Cuxham (20).
Each transect was tied into the OS Map by measuring its distance from a marked feature (usually the crossing point of the Right of Way with another Footpath or Road), and the central point of the Road at each transect was plotted on the Map.
Above shows the Route of the Lower Icknield Way between Dorchester & Cuxham as dowsed. We have previously determined the position of a Roman Crossroads in the Field Northeast of the Dorchester Bypass at its Junction with the old Henley Road leading into the Village. A re-examination of this area (1) suggests that a Road coming from the direction of Warborough does indeed join this Crossroads, at or just North of the Crossing Point of the Roads coming from Henley and Silchester. Warborough to Berrick Salome (Points 2-8). We almost fell over the Roman Road where it crosses Green Lane west of Warborough (2) as here the Agger forms a distinct hump across this (presumably) Ancient Right of Way. Malpas (1987) referred to this hump and it is still clearly visible nearly 20 years later, although it appears to have been ploughed out completely in the adjacent Fields. We located the Road again where it crossed the Main Street in Warborough opposite the Church (3). A quick transect at the other (East) end of the Church confirmed that the Agger ran directly beneath it. We picked up the Road again at a measured Transect in the Northeast corner of The Green in Warborough (4), where the tarmac of the modern Road overlaps with the Agger which lies both beneath and immediately to the South of it. The modern Road and the 1st part of the Track to Berrick Salome which forms its continuation appears to approximate the Line of the Roman Road.
A glance at the OS map (Explorer 170) shows that this Track, which is marked as a Bridleway, now has a sharp bend to the Right at about the midway point between The Green and Ladybrook Copse. The road signature was very clear at this Point (5), where the Road appeared to be heading off across the Field in the direction of Ladybrook Copse. The Road was measured here where it crosses the Footpath that follows the West edge of this Copse (6); the South edge of the South Ditch of the Road was 6.0M North of the existing Bridlepath to Berrick Salome. Malpas (1987) stated that the Agger is visible where it crosses the modern Road between Berrick Salome & Berrick Prior. Assuming this to be the case, and assuming that the Roman Road is on a direct alignment, we expected to pick it up where it crossed the 2nd field boundary to the East of the Central Crossroads at Berrick Salome. Working from a photocopy of the OS Map, we initially confused a Parish Boundary with the Footpath and dowsed unsuccessfully for the (non-existent) Road at the East corner of this Field Boundary. We eventually found the Road almost midway between this Parish Boundary and the modern Bridlepath into Berrick Salome (7). The Agger is indeed quite pronounced in the Road between the 2 Berricks (8), and was especially noticeable when cars drove over it. The exceptional perceived width of the Agger at this point (23.65M) suggests that it may have been distorted by the construction of the modern Road across it.
Berrick Salome to Brightwell Baldwin (Points 9-16)
According to Morris et al. (1968), the Roman Road coincides with the Bridlepath leading to Hollandtide Bottom for a distance of ‘about a furlong’ [approx. 200M] to the East of Berrick Prior. We were able to verify this by dowsing (9) and showed that the Agger leaves the Bridlepath to head off across the Fields on course to Brightwell Baldwin at the point where there is a small Copse on the North side of the Bridlepath. Morris et al. (1968) also considered that the Roman Road crossed the North flank of Hollandtide Bottom, on the strength of a scatter of ‘metalling’ in the Plough-soil. By walking along an overgrown Footpath, shown on the OS Map as a Farm Track running North at right-angles from the existing Bridlepath, we were able to confirm this by locating the Road quite high up on the side of the Hill (10). We attempted to confirm the Route again on the ‘minor’ Road leading to Lonesome Farm (11): we thought we had found it but when we plotted the results on the Map they showed a noticeable Southward Dip in the Road alignment at this point and we think that this recording was anomalous. Instead of the quiet Country Lane we expected to find, this Road appears to have become a popular shortcut for Traffic to Chalgrove, and having to jump into the hedge every few minutes is not conducive to accurate dowsing! This finding needs to be confirmed and we suspect that in our haste we may have recorded an old Farm Track instead.
The next clear location that we found was at the Crossroads where 2 Bridlepaths cross at right-angles just South of Whitehouse Farm (12). Here the agger more or less coincides with the Bridlepath leading to Brightwell Baldwin and again heads off across the Fields on a direct Route, whereas the modern Path curves South to the East of this Point. A Double Ditch was recorded on the North side of the Roman Road here. Morris et al. (1968) plotted the Roman Road as passing just to the North of Brightwell Church but we found that, as at Warborough, the Agger passes beneath the Church itself. To confirm the position of the Road, we dowsed a Transect along the lane following the West Boundary of Brightwell Park leading to Cadwell Farm (13) and, at a later date and with the permission of the Landowner, along the Drive within Brightwell Park itself (14). The double North ditch was clearly defined in the Cadwell Farm Lane; we didn’t dowse for it in the Drive. We also traced the Road along the walled Footpath that runs along the East side of the Churchyard (15) and a double North ditch was recorded here as well. These 3 Locations (13-15) show that the Route of the Roman Road approaches the modern Road through Brightwell Baldwin at an angle. The numbers refer to measured transects assumed that the Roman Road ran parallel to it immediately to the North. Instead, by dowsing along the Modern Road, we found that the Northside of the Agger crosses the Road at an oblique angle opposite Glebe Farm, which is the last House in Brightwell Baldwin before the Rectory on the right side heading East, immediately opposite the start of the Footpath to Chalgrove. If the ‘strætforda’ does represent the place where the Roman Road Forded the Stream running North into what is now Brightwell Park, then this Ford must have been to the South of the Modern Road and not to the North. The Owner of the Rectory kindly allowed us to dowse in her back Garden and we found the North Ditch, which appears to be single here, at a distance of 17.95M from the South-facing wall of the Rectory Building (16). The Agger itself straddled a ‘ha-ha‘ which separates the Formal Garden from a Pasture.
Brightwell Baldwin to Cuxham (Points 17-20)
The alignment of the Road suggests that it crosses Turner’s Green Lane just South of where this leaves the Modern Road along the Parish Boundary on the East side of Brightwell Baldwin. We found the North edge of the North Ditch 39.8M from the Southside of the tarmac of the Modern Road (17). At this point, the Road seemed to be heading across the fields straight for Cuxham Church, and once again we found that the Agger ran beneath the Nave of the Church, slightly towards the Southside and that the Squat Tower sat directly over it (19). Morris et al. (1968) considered that the Roman Road probably coincided with the Modern Road between Brightwell Baldwin & Cuxham in order to avoid a presumed Marshy area which is marked as Medieval Fish-ponds on the OS Map on the West side of Manor Farm. Our results suggest instead that the Road runs right through this area and to confirm this we examined the East side of the field immediately to the East of Turner’s Green Lane on the South side of the Modern Road (18). A deep Ditch which may have held Open-water (hidden by vegetation) ran along this edge of the Field, and this made the dowsing difficult by causing the rods to deflect. We picked up Signals at regular intervals all along this field margin, which we interpreted as Field Drains draining into the Ditch, and these responses were superimposed on the Road Signature. However, we did locate the Road along this edge of the field and the centre of the Agger ran 135.25M South along the field boundary from the Southside of the Modern Road, exactly online between our findings at Turner’s Green Lane and Cuxham Church. Our final location in this Survey of the Lower Icknield Way was found just to the Northeast of Chestnut Farm on the East side of Cuxham (20). Margary (1973) traced the Line of this Road along existing Trackways and old Hedgerows to the Northeast of Aston Rowant (where it is clearly marked as ‘Lower Icknield Way (Track)’ on the OS Map) and Lewknor, which today lie either side of the M40, and following the same general alignment on to Pyrton, where it is marked as the ‘Oxfordshire Way’ on the East side of the Modern Road through the Village. Morris et al. (1968) picked up the Thread at this point and assumed that the line of the Road follows the Footpath from Pyrton to Chestnut Farm, although they considered that its ‘course through Cuxham is obscure’. We looked for the Road for a short distance along this Footpath where it heads Northeast from Chestnut Farm, but found no evidence for it. We then walked along the Hedge boundary which runs to the rear of Chestnut Farm Cottages and continues to the Northwest, and we picked up the Road here at an oblique angle. A fragment of Roman(?) Oyster Shell was found here on the line of the North Ditch. This suggests that, rather than the abrupt change of direction postulated by Morris et al. (1968) at Chestnut Farm, the Roman Road curves almost imperceptibly to the North as it leaves Cuxham, presumably to join the existing Footpath some distance to the Northeast of Chestnut Farm.
Results: the Road Dimensions
The maximum Road Width, which was taken as the distance between the North edge of the North Ditch (the inner Ditch where this was double) and the South edge of the South Ditch, and the Agger width were recorded at most of the location points. The mean road width was 30.83M (range 26.25-38.20M) and the mean Agger width was 13.64M (range 9.85-18.90M).
The Lower Icknield Way as Surveyed by Dowsing, in general, follows the Route determined by previous Authors (Morris et al., 1968; Malpas, 1987) from the alignment of existing Footpaths and Hedgerows, substantiated here & there by a raised stretch of Agger and/or exposed Road Metalling. This demonstrates that both methods are valid techniques for establishing the Routes of Roman Roads and complement each other, the extra precision offered by Dowsing enabling some of the bends and corners produced by the meandering of Footpaths over the last 2000 years to be straightened out and also providing information within present-day Settlements, such as Brightwell Baldwin, where Surveying the Road by means of Footpaths & Hedgerows is not a viable option. Morris et al. (1968) substantiated their work by describing 2 sections of the Road excavated on the West side of the Drive at Brightwell Park, which was very close and immediately to the West of our location Point 14, and on the Hollandtide Bottom Track East of Berrick Prior, which was just to the West of our location Point 9. Both these sections of Roman Road which were proven by excavation lie precisely on the line of the Road as determined by Dowsing. The same Authors also referred to ‘faint traces’ of the Road to the West of Ladybrook Copse; this corresponds to our location Point 6 although no surface signs of the Road can be seen today.
Malpas (1987) claimed that the Lne of the Road near Brightwell Baldwin was confirmed in 1976, when a British Gas Pipeline Construction sectioned the Road 600M WSW of Brightwell Church’. He admitted that this location lies some distance South of the direct alignment and suggested that ‘this demonstrates how these minor roads tend to deviate from the direct alignment’. However, Malpas was wrong: this Pipeline Section is now believed to have cut through a modern Farm Track and the record has been deleted from the Oxfordshire SMR (Ian Clarke, pers. comm.). The Roman Road does not deviate from its Course at this point. Nevertheless, apparent deviations can occur. We obtained permission to dowse in the field to the Southwest of Whitehouse Farm, where we picked up the Roman Road Signature on the East side of this Field just above the Hedge-line in the Southeast corner which bounds a dense thicket about 50M wide abutting onto the Hollandtide Bottom Track below. We had previously found that the Roman Road coincides with the Bridlepath at this Point, so why did there appear to be a 2nd Road running parallel to the 1st and about 50M to the North of it? We tried to trace this feature out into the Field but lost it: it may have changed direction at this point, in which case it may have corresponded to a later Farm Track which is recorded on a Map dated 1791 (Ian Clarke, pers. comm.). Alternatively, it could have been a later stretch of Roman Road built at a higher level to replace the original which at this point runs along the bottom of the continuation of Hollandtide Bottom and might conceivably have become muddy & badly rutted. We found the ‘original’ Road again in the Southwest corner of this Field where we would have expected to find it, assuming that our previous recording at the Crossroads was accurate. A glance at the plotted Map will show that the Lower Icknield Way is not dead-straight. It follows a gentle curve between Warborough & Berrick Salome and then with some minor deviations it follows a more or less straight Course to Cuxham, where it begins to curve gently to the North to continue the line of the Road as it follows the edge of the Chilterns up to Aston Rowant & beyond. Although we dowsed the Road from West to East, Margary (1973) traced the Road in the Reverse Direction from its origin at Akeman Street near Tring and subsequent Writers have followed this Direction. That this was indeed the Direction in which the Road was constructed is suggested by the contours on the OS Map, which indicate possible Sighting Points that the Engineers could have used.
Between Lewknor & Ivinghoe the Track diverges into 2 Parallel Courses, known as the ‘Lower’ & the ‘Upper’ Icknield Way. Here a further uncertainty enters the debate:
“The Upper Icknield Way has all the characteristics of a Trackway, following the steep escarpment of the Chiltern Range, but the Lower Way is here peculiar in following a series of true alignments on the flat land in a position unlikely for a Trackway and quite typical of a Roman Road . . . . There can be little doubt that this part of the Icknield Way is a true Roman Road.” – Roman Roads in Britain, I D Margary (1955)
The clear lines of a wide Roman Trackway running across this Aston Clinton Housing Development Site from the South-West to North-East, probably part of the Icknield Way, can be seen running across the Site towards Ivinghoe Beacon. When viewed on Maps of the wider area, the alignment of the Trackway matches the projected course of the Lower Icknield Way, so may have its origins in the Iron Age or earlier.
The line of any Roman Road can usually be divided up into a series of straight stretches of various lengths, each of which is aligned on a different Sighting Point. The Lower Icknield Way heading Southwest from Lewknor appears to be aligned on the 123M Summit of an unnamed Hill. When projected, this alignment Crosses the Thames just South of Wallingford, which perhaps prompted Margary (1973) to think that the Road may have run to Wallingford. However, at Pyrton the alignment changes and the new Sighting Point may be Scald Hill. The alignment changes again at Cuxham and from there the Road maintains a more or less Straight Course as far as Berrick Salome: this stretch appears to be aligned with the right-angled bend of the Thames at Little Wittenham. From Ladybrook Copse to Warborough the Sighting Point was the Summit of Castle Hill and the presumed Ford where the Roman Road from Silchester crossed the Thames was on the same alignment. Morris et al. (1968) thought that the Roman Road coincides with the Modern Road along the North side of The Green at Warborough and that this corresponds to a ‘Streteforlong’ mentioned in early documents. This may be so, but the Roman Road does not make ‘a turn of a few degrees to the South’ at this point as they suggest, and it remains aligned on the Roman River Crossing. At Green Lane to the West of Warborough, the alignment changes abruptly by 23° North to join the Roman Crossroads at the approach to Dorchester. Morris et al. (1968) also envisaged a change of alignment, in this case, to follow ‘an obviously Ancient Track’ opposite the Church that heads across the fields in the direction of Dorchester. This Track may be Ancient but it does not represent the Roman Road, which continues on the Castle Hill alignment until it reaches Green Lane. Judging from its width, the Lower Icknield Way was quite a Major Road, with a mean Agger width of 13.64M (wider than a Motorway directional 3 Lanes & Hard Shoulder) This compares with the Figure of 50-ft [15.24M] quoted by Margary (1973) for Major Roads. It has been suggested that the Romans constructed the Lower Icknield Way as a Parallel alternative Route to the Prehistoric Upper Icknield Way which runs along the Scarp of the Chilterns. Ultimately this Road connected with Watling Street and beyond, and it may have carried a considerable amount of Traffic. It is interesting to note that the Roman Road runs directly beneath the Churches in 3 of the 4 Villages Surveyed: Warborough, Brightwell Baldwin & Cuxham. It appears that Medieval Architects appreciated a solid Roman Agger to support the Foundations of their Churches; we have previously shown that Moulsford Church is Sited over the Agger of the Silchester-Dorchester Road. The Churches at Warborough, Brightwell Baldwin & Cuxham form a (more or less) Straight Line on the Map. Could this be a vindication for Alfred Watkins (1974) who reckoned that alignments of Sacred Sites indicated Ley Lines? If so, perhaps we could reinterpret at least some Ley Lines as Roman Roads: perhaps aligned Medieval Churches could indicate the Routes of previously undiscovered Roman Roads. To return to the Brightwell Baldwin ‘strætforda’, we have established that this either lay to the South of the Modern Road through the Village – or it still remains to be discovered somewhere in the North of the Parish near Cadwell Farm.