The 'Via Devana' - a 'lost road' found! (Pt1)

 The name Via Devana was coined by the 18th century Cambridge antiquarian Dr Charles Mason and was said to be a route which connected Roman Colchester (Camulodunum) to Roman Chester (Deva)^[1]. Ivan Margary^[2] described the route from Godmanchester (Durovigutum) to Colchester as 'Via Devana' and numbered it road 24. Early Ordnance Survey (O.S.) maps show what was the Cambridge Road, then the A1198 and now largely the A1307/A14 Godmanchester to Cambridge road as 'Via Devana'. The route follows a track known as 'Worsted' or 'Wool Street' which runs south from Cambridge as green lanes and tracks. At Worsted Lodge on the modern A11 it crossed a Romanised section of the ancient 'Icknield Way'.

 In 1959 a 10.5-mile (16.8km) length of gas main was laid along Worsted Street, the work being monitored by the archaeologist P.C. Dewhurst^[3]. The trenching machine used produced a clean 3' (0.91m) deep trench, which allowed observation of the Roman road builders' methods and how they changed along the route. 

 The road was more well-made nearer to Cambridge and it appears that its construction began from that end. Dr Cyril Fox of the Cambridge Antiquarian Society reported on a number of sections which were cut across the road in 1921^[4]. They showed that the total width of the agger was c.36' (11m), with the width of the gravel crest c.8' (2.4m). Dewhurst noted that the total height of the road surface above the natural chalk there to be 3' 9" (1.14m).

 At approximately 0.5 mile from Worsted lodge a layer of coal was encountered. In the section it could be seen sandwiched between the make-up layers of the agger. Samples were submitted for analysis, and the results found that the most likely origin for the coal was the Nottinghamshire or South Derbyshire coal fields.

 A pair of roadside ditches were noted at Worsted Lodge. They were 42' (12.7m) apart, while the height of the road surface above the natural chalk was 3' 7" (1.1m). South of Worsted Lodge the road becomes less robust and more sinuous. The route probably continued south through Haverhill, Wixoe and Sible Hedingham in Essex and although not certain may have run along the Colne valley to Colchester. 

Worsted Street at Fulbourn, looking north towards Cambridge. (Photo: C.Coates)

 In the westerly direction from Cambridge the route joined 'Ermine Street' at Godmanchester, Margary's road 2b. After passing through Huntingdon it turned north, later to be followed by the line of the 'Great North Road' which is mirrored by the modern A1. At Alconbury, northwest of Huntingdon, Margary's road 57 took the route westwards to Leicester (Ratae Corieltavorum). Following the 'Gartree Road' it would have entered the town by the south gate.

The Gartree Road/Via Devana approaching Leicester from the southeast.  (Reproduced with the permission of the National Library of Scotland. Licence: CC-BY-NC-SA)

 Margary identified a route leaving Leicester's west gate following King Richard's Road before turning to the south. As probably part of the 'Fosse Way', it joined 'Watling Street' (now the A5) at High Cross (Vernonae), and thence via Wroxeter towards Chester. A second road noted by Alan McWhirr^[5] left the westward road further to the west, turning sharply southwest to join Watling Street near Mancetter (Manduessedum).

Roman Leicester. (After Morris, Buckley & Codd 2011)

A Roman Road Through NW Leicestershire.

 Evidence for an alternative route appeared in 1994 with the publication of the research carried out by Peter Liddle and Fred Hartley of the Museum of Leicester.^[6] They described a probable route leaving Leicester's west gate^[7], crossing the river Soar and running northwest through Leicestershire and Derbyshire.

Roman roads from Leicester to Ryknield St. (After Liddle 1994)

 In the following discussion, numbers in round brackets refer to locations identified by Peter Liddle on his map above. The route is thought to have followed King Richards Road before crossing Western Park and the New Parks estate^[8]. A short length of the Glenfield/Braunstone parish boundary lies on this route. The Mancetter road (14) approaching from the southwest almost certainly joined the Via Devana route at Glenfield, after crossing the Kirby Muxloe and Glenfield golf courses. McWhirr identified crop marks probably representing the Mancetter road, heading towards the M1 from Kirby Muxloe on RAF aerial photographs from 1948. They can also be seen in more recent images (below). The Mancetter Road would have joined the westward road near what was Oakmeadow Spinney to the east of the M1, now covered by part of Glenfield golf course and the Glenfields housing estate.

Faint cropmarks heading towards the M1 at Kirby Muxloe, 2013 (www.getmapping.com)

 Liddle describes two potential routes continuing to the northwest. The more northerly one may have passed through the Coalville area where a hedge alignment and section of metalled track was observed, before proceeding via Ashby-de-la-Zouch. This is the general route of the modern A50, which from Ashby at least follows higher ground almost all the way to a River Trent crossing at Burton.

 The southern route has now been more firmly identified. West of Glenfield, Liddle notes that Mike Ball observed a section of metalled track running north of Kirby Grange (13) on the line of the route out of Leicester. A footpath is shown running towards Kirby Grange from Ratby on early O.S. maps, but is perhaps a little off the line of the proposed road.

 Stamford Street in Ratby (12) represents a change in direction further towards the north. A footpath originally ran from the northwest end of the street to meet a track on the western side of Martinshaw wood. Here a slight hollow way with metalling was observed by Liddle (11). Field boundaries follow the alignment northwest to Whittington Grange where the Loughborough Archaeological Society sectioned a metalled track in 1954 (10). A substantial road 7 feet (2.12m) wide was revealed, which had been remade at some point. Its surface carried wheel ruts of 4"/5" (10-12cm) depth and 4' 8"/4' 9" (1.42-1.44m) apart.^[9]

Whittington Grange (top left) to Ratby.(Reproduced with the permission of the National Library of Scotland. Licence: CC-BY-NC-SA)

 As a brief aside, a fine example of what might be an Iron Age enclosure can be seen c.1km west of Ratby village at Bury Camp.

Bury Camp near Ratby, with the line of the Via Devana running to the northeast through Ratby village.
(Contains crown © copyright and database rights 2019. OGL V3.0)

 To the north of Ibstock (9) a track and footpath lie on the alignment heading southeast to a bend in the Ravenstone Road: here a number of parish boundaries met at a local highpoint in the landscape at 466' (142m). Several properties have been built on the alignment to the southeast of Ravenstone Road, including the former contagious disease hospital. The Roman road may be observed passing through a series of enclosures on an aerial photograph which may have been a small settlement. The alignment continues southeast to a bend in Leicester Road, Ibstock.

Cropmarks north of Ibstock, 2006. (Google Earth)

 Early O.S. maps show an alignment of tracks and footpaths heading northwest towards the former Ravenstone Mill on the River Sence. A Roman settlement was discovered here prior to open casting of coal. A number of pottery and tile kilns were located and a crop mark feature seen on aerial photography appeared to lie close to the line of the road.^[10] The area has since been landscaped as part of the Sence Valley Country Park.

Plot of cropmarks near Ravenstone Mill. (After Lucas, 1981 fig 2)

 Heading northwest towards Normanton-le-Heath, a number of field boundaries and a footpath fall on the alignment. It then passes along the street known as 'The Hollow' (7)(8) before following a footpath and field boundaries further to the northwest. A few further field boundaries are close to the line (5)(6) which if directly straight, would cross close to 'Five Lanes End' at Willesley, another local highpoint in the landscape at 423' (129m).

The line of the route through Normanton-le-Heath to Ravenstone Mill. (Reproduced with the permission of the National Library of Scotland. Licence: CC-BY-NC-SA)

 To the northwest of Willesley the route is marked by a few lengths of parish boundaries. It can next be picked up near Warren Farm at Moira as cropmarks (4), with field and parish boundaries also on the alignment. Slightly further to the northwest, cropmarks on aerial photographs and Lidar imagery show the road crossing the grounds of the Moira Furnace Museum.

Lidar image of Via Devana at Moira passing through the grounds of Moira Furnace Museum (top left), then across fields at Warren Farm (centre). (Contains crown © copyright and database rights 2019. OGL V3.0)

 At the appropriately named Via Devana, just off Bath Lane in Moira, a metalled track was identified during pre-development archaeological excavations.^[11]

Metalled road surface at Bath Lane, Moira. (After Browning 1998, fig 3).

The road had survived in a paddock as a 'flat-topped earthwork bank'. A section cut across it revealed an upper convex layer of slightly compacted pebbles, overlying a compact clay layer containing frequent pieces of sandstone. Below this were further layers of clay. The width of the pebble surface was c.1.55m (4' 3") although it may have been truncated by recent activity.

Section cut across the road at Bath Lane,  Moira. (After Browning 1998, fig 4).

 Further still to the northwest a 1999 aerial photograph of the area which is now the National Forest Camping and Caravan Centre (NFCCC), shows a cropmark feature again on the alignment of the proposed Roman road.

Cropmark crossing the NFCCC Moira, 1999. (www.getmapping.com)

 The route now crosses into Derbyshire. At Overseal, the local memory of a 'pink road', probably recalls the road passing to the north-west of the village (Dr Mark Knight pers. comm.). As Liddle suggests, it may also be the 'curious causeway called Leicester Headland', noted by Mammett as being 'raised throughout with red clear gravel'.^[12] Ironstone is found widely in the Ashby Woulds region and this reddish material could have been used to create a metalled road surface.

 In the adjoining parish at Linton Heath a faint cropmark on Lidar imagery and aerial photographs appears to show the road crossing the sports ground, then across the field opposite to meet a discontinuity in a hedge line.

Faint cropmarks crossing the sports ground and field at Linton Heath, 1999. (www.getmapping.com)

Lidar image of a faint ridge crossing the Linton Heath sports ground.
(Contains crown © copyright and database rights 2019. OGL V3.0)

 From this point the route which the road might have taken is unclear. A direct line would have entailed crossing the Waterfallows gulley. Here a spring feeds a small stream which runs out onto fields below, making for a boggy route. However, early Ordnance Survey maps do show a track running from the bottom of the gulley, across the fields to Linton Manor house to the northwest. An alternative would have been to deviate slightly to the west, to circumvent the gulley and follow the higher ground.

 Opposite the track which leads into what was Linton Manor Farm on Hillside Road is the road to Coton Park. This is a hamlet of former mineworkers' houses who were employed in two nearby collieries. A farm track runs along a ridge behind one of the rows of houses: once again on the straight-line route of the proposed Roman road (3). The modern access onto this track curves away from Coton Park Road to avoid a steeper climb up onto the ridge. If this was part of the Roman route it would not be a surprise, as their roadbuilders often deviated from a dead straight line to avoided difficult terrain or obstacles. This section of lane was referred to as the 'Drakelow' or 'Stoney Way' in C18th and C19th newspaper notices.^[13]

Looking towards the Trent valley from the 'Drakelow Way' at Coton Park. (Photo: C.Coates)

 After crossing an unclassified minor road to Caldwell the alignment continues as a green lane. The line is lost shortly after passing Ashley House Farm as the terrain drops down to the River Trent floodplain. It can be picked up again a little further west, where the Rosliston Road takes a bend and a track and hedge lie on the alignment (2). The edge of Grove Wood then continues the line towards the Drakelow/Walton parish boundary. Visible on Lidar imagery is a feature which curves away from the edge of Grove Wood which might be the remains of a medieval plough headland. However, with the eye of faith it is possible to see a faint feature following a straight line route.

Lidar image of the route approaching the Trent crossing near the former Drakelow Powerstation site.
(Contains crown © copyright and database rights 2019. OGL V3.0)

 The edge of the former Drakelow Power Station is marked by a length of scrubby woodland running to the River Trent; the Drakelow/Walton-on-Trent parish boundary follows it. Near the Trent crossing, Liddle notes that an aerial photograph taken by Jim Pickering shows a cropmark which has the appearance of a Roman fort, perhaps guarding the river crossing.

 On the Staffordshire side of the Trent, the Barton/Dunstall parish boundary is partially marked by a hedge (1) which continues the alignment towards the former Roman Ryknield Street. The old turnpike road and the modern A38 largely follow the Roman road, but at this point they deviate away from a straight line to re-join further to the northeast. This may have occurred due to the land close to the river being liable to flood.

Cropmark to the left of a more recent track, on the line of Ryknield Street at Branston, 2015. (www.getmapping.com)

 Although this area has now been much developed, a cropmark on an aerial photo (above) might mark the remains of the original line of Ryknield Street. The 'T' on the above aerial photograph marks the approximate junction of the Via Devana and Ryknield Street. A section of the agger of Ryknield Street  still survives at Bean's Covert to the northeast and can be clearly seen in the Lidar image below.

Agger of Ryknield Street crossing Bean's Covert (centre) at Branston.
(Contains crown © copyright and database rights 2020. OGL V3.0)

 In the second part of this blog series, I will discuss some options for the westwards continuation of the Via Devana route towards Roman Chester.

Dr Mark Knight is a Cultural Heritage Officer with the Transforming the Trent Valley project.

References:

[1] Via Devana https://en.wikipedia.org/wiki/Via_Devana

[2] Margary, I.D. (1967). Roman Roads in Britain. 

[3] Fox, C. (1923). Excavations in the Cambridgeshire Dykes: I. Preliminary Investigation; Excavations at Worsted Lodge. Proceedings of the Cambridge Antiquarian Society, October 1921 - May 1922. Vol. 24, pp. 21-27. https://www.archaeologydataservice.ac.uk/library

[4] Dewhurst, P.C. (1964). Wool Street, Cambridgeshire. Proceedings of the Cambridge Antiquarian Society, January 1962 - December 1963. Vol. 56, pp. 42-60. https://www.archaeologydataservice.ac.uk/library 

[5] McWhirr, A.D. (1967). The Roman Road from Leicester to Mancetter. Transactions of the Leicestershire Archaeological and Historical Society. Vol:42, pp1-5

[6] Liddle, P. and Hartley, R.F. (1994). A Roman road through North West Leicestershire. Transactions of the Leicestershire Archaeological and Historical Society. Vol:68

[7] Morris, M., Buckley, R. and Codd, M. (2011). Visions of Ancient Leicester. University of Leicester Archaeological Services.

[8] Higgins, T. (1998). Westbridge Place, Duns Lane, Leicester. An Archaeological Watching Brief. University of Leicester Archaeological Services. Report N°: 98/149

[9] LAHS (1958). Excavation of old road at Whittington Grange, nr Markfield, Leicestershire. The Bulletin Of The Loughborough & District Archaeological Society, N°1, 1958, pp. 8-9

[10] Lucas, J. (1981) in Liddle, P.  Archaeology in Leicestershire and Rutland 1981. Transactions of the Leicestershire Archaeological and Historical Society. Vol:56, pp. 104-107

[11] Browning, J. (1998). An Archaeological Watching Brief at Bath Lane, Moira, Leicestershire. University of Leicester Archaeological Services. Report N°: 98/64

[12] Mammett, E. (1852). The History and Description of Ashby-de-la-Zouch; with Excursions in the Neighbourhood, p155. W&J Hextall, London. https://play.google.com/books/reader?id=lxMHAAAAQAAJ&pg=GBS.PA154&hl=en_GB

[13] To Be Sold By Private Contract. Derby Mercury, 10th January 1799, P1.

https://www.britishnewspaperarchive.co.uk