Except where stated otherwise (e.g. when the copyright of photographs is retained), you are free to reproduce any of this work for non-commercial purposes only, provided proper and appropriate attribution, credit and citation is given to the author, any original contributor or source, and the RRRA, in accordance with the terms of this licence.

Creative Commons Licence
This work is licensed under a Creative Commons Attribution-NonCommercial 2.0 UK: England & Wales License

© Mike Haken & RRRA, 2018

Home Gazetteer of Roads Margary's Numbering Itineraries & Sources Glossary/Biography RRRA Website

Margary Number:

Other Numbering System:



28 Miles


North Yorkshire HER MNY23576

Cawthorn Camps - Pastscape

Lease Rigg Pastscape


North Riding of Yorkshire

Ever since it first appeared on John Warburton’s map (Warburton, 1720 - fig 1.), and was subsequently described in Francis Drake’s ‘Eboracum’ (Drake, 1736),  Wade’s Causeway, which runs over Wheeldale Moor, near Goathland, between the Roman sites at Cawthorn and Lease Rigg, has remained one of the best known Roman roads in Britain. In Drake’s day, the feature survived very well between the two sites, although he was not aware of Lease Rigg. For the supposed line south towards Amotherby, we are dependent on his account, although the road was claimed to have been located in the late 8th Century by Robert King, a surveyor from Hunmanby, and traced a mile north to the crossing of the R. Rye, although no evidence for this claim is known (Hinderwell 1811, p19). Drake’s claim that it went northwards to Dunsley Bay seems to have been influenced by Camden in the belief that it is the Δουνον κολπος (Dunum Sinus -fort bay) listed by Ptolemy, but Ptolemy’s co-ordinates suggest a site a little further down the coast, possibly Robin Hood’s Bay or Scarborough (Romaneranames.uk, 2016).

It is well worth taking a look at most of Drake’s description, which gives as good an indication of the striking and unusual nature of the feature as any modern account could. His account begins at the Roman forts and camps at Cawthorn, which survive as well today as they did in Drake’s time.


Codrington, Thomas (1903); Roman Roads in Britain, Society for Promoting Christian Knowledge, London

Davies, H, (2020); Roads in Roman Britain; Tempus, Stroud

Drake, Francis (1736); Eboracum: or the History and Antiquities of the City of York, London.

Frere, S. S. & Fitts, R. L. (2009); Excavations at Bowes and Lease Rigg Roman Forts; Yorkshire Archaeological Report No. 6, YAS, Leeds

Hayes, R. H. & Rutter, J. G. (1964); Wade’s Causeway: A Roman Road in North-East Yorkshire; Scarborough and District Archaeological Society, Scarborough

Hinderwell, Thomas (1811) The History and Antiquities of Scarborough and the Vicinity

Kitson-Clark, Mary (1935); Roman Roads in East Yorkshire in Roman Malton & District Report No. 5

Lyall, J. (2017); Report on a fluxgate gradiometer and topographic survey carried out over Lease Rigg Roman fort, North Yorkshire; unpublished report

Newton, Sir Charles (1847); Map of British and Roman Yorkshire, Archaeological Institute of Great Britan and Ireland, London

Macmahon, K.A. (1964), Roads and Turnpike Trusts in Eastern Yorkshire, East Yorkshire Local History Society, York

Available at http://www.eylhs.org.uk/dl/128/roads-and-turnpike-trusts-in-east-yorkshire accessed 19/7/17

Margary, Ivan D. (1973); Roman Roads in Britain, John Baker, London

Richmond, I. A. (1932); The Four Roman Camps at Cawthorn, in the North Riding of Yorkshire; The Archaeological Journal, Vol. 89, pp. 17-78

Available online at http://archaeologydataservice.ac.uk/archives/view/archjournal/contents.cfm?vol=89 accessed 16/11/17

Romaneranames.uk (2016); Δουνον κολπος ; available at http://www.romaneranames.uk/d/dunokolp.htm; accessed 26/1/18

Toller, H. S. (2015); Pers comms.

Vyner, B. forthcoming; Wade’s Causeway and the Wheeldale Moor linear monument: the end of the Roman road?

Wilson, P.R. (2016); Pers Comms.

Fig. 1 Extract from John Warburton’s New and Correct Map of Yorkshire, 1720, showing Wade’s Causeway marked as a Roman Way, and with apparent certainty.

Click Images to enlarge

Fig. 3  Photo of the causeway, teken in 1981. © M C Bishop 2017, published under a Creative Commons Attribution license

RRRA Forum for RR81b

Fig. 2  Lidar image showing Cawthorn camps and part of Wade’s Causeway still surviving, but unprotected. Outlined in yellow is the short length of extant Roman road running through Camp D (actually a fort) - note the difference in scale between this feature and Wade’s Causeway.


It would seem that almost everyone since Warburton and Drake has accepted that this feature is a Roman road virtually without question, simply because of its connections with the Roman Forts at both Cawthorn and Lease Rigg. Prof. Sir Ian Richmond came close to questioning that assumption, by dismissing the claimed relationship between the forts and the road, but stopped short of dismissing the road itself, despite noting the absence of metalling (Richmond, 1932, pp. 20 & 78). It is certainly true that it appears to run between two Roman sites, the site of four forts/camps at Cawthorn and the fort at Lease Rigg, but is this feature a road between them, or were the forts deliberately sited close to or even on top of a pre-existing feature, and if so was that feature a road at all? In fact, it is atypical on almost every aspect of Roman road construction. Hayes & Rutter in their comprehensive survey and analysis of Wade’s Causeway (Hayes & Rutter 1964) did try to deal with most of the issues by looking at each major characteristic of Roman roads in turn, but all too often avoided or ignored the questions that arose. It is worth a look at the characteristics of construction again, with the benefit of fifty years more research and knowledge.

Alignment and planning

It is certainly true that Wade’s Causway exhibits a similar use of the landscape to Roman roads in upland Britain, and it is also clear that its general course appears to have been set out in a Roman-like fashion, using a series of alignments set out between sighting points (Hayes & Rutter, 1964, p.80). How the deviations from these alignments were determined, and how they compare with known Roman examples in similar territory, is an analysis which has not yet been carried out. The most experienced researcher of Roman roads in upland areas, Hugh Toller, expressed an opinion, albeit a not very scientific one, that there is something about the setting out that just doesn’t feel right (Toller, 2015).

However the road was planned, we would certainly expect to see it constructed as a series of straight sections, with a few curves here and there when the topography demanded. Blaise Vyner has pointed out that the “road” is actually atypical of Roman engineering due to its sinuous route (Frere & Fitts, 2009, p. 277.), and he is unquestionably correct. Wade’s Causeway appears to change direction on occasions when no change is necessary for example about a mile and a half south west of Grosmont (NZ 8135 0385), where it it suddenly swings NNW, and yet the easiest and shortest route towards Lease Rigg is straight on. Whilst most sections appear to have been set out between two sighting points, there is no evidence whatsoever that the lines between the end points were ever set out on the ground, as was always the case with Roman construction. The consequences of this are plain to see. The 1981 photo taken by Mike Bishop (fig. 3) shows that the road fails to keep to the pre-determined straight course and when viewed from above (fig. 4) it becomes absolutely clear that the ’road’ that appears straight on a map is, as Vyner said, quite sinuous. The precise course for the road builders to follow has clearly not been set out on the ground by the surveyors, which would make it the only known Roman military road to display this lack of rigour.

The Embankment or Agger

Until relatively recently, there has been an almost unquestioning acceptance that the remains of the structure seems to have a “decided ‘Roman look’ in terms of their construction” (Frere & Fitts, 2009, p. 277). That notion appears to be based on an assumption that Roman roads often had a base comprising large stones with smaller material on top, and that it is this foundation that can be see at Wade’s Causeway today, the smaller material having disappeared. Certainly the width of the causeway would be quite normal for a Roman road at an average of 5.6m. The quality of construction appears at first sight to be quite variable on the causeway, although much of the variability may be due to damage caused by training activity during the Second World war (Wilson, 2016) and ’reconstruction’ by a former custodian of the site in the early 20th century. It is possible that it may originally have more closely resembled the photo from 1912 (fig. 6).

Extract from Francis Drake’s ‘Eboracum’

My curiosity led me to see it; and coming to the top of a steep hill, the vestiges of the camp (Cawthorn) were easily discernable. At the foot of the hill began the road or causeway, very plain; and I had not gone a hundred paces on it , but I met with a mile-stone of the grit kind, a sort not known in this country. It was placed in the midst of the causeway, but so miserably worn, either by sheep or cattle rubbing against it, or the weather, that I missed of the inscription, which, I own, I ran with great eagerness to find. The causeway is just twelve foot broad, paved with a flint pebble, some of them very large, and in many places it is as firm as it was the first day. A thing more strange, in that not only the distance of time be considered, but the total neglect of repairs, and the boggy rotten moors it goes over. In some places the agger is above three foot raised from the surface. The country people curse it often. For being almost wholly hid in the ling, it frequently overturns their carts laden with turf, as they happen to drive cross it.

It was great pleasure to me to trace this wonderful road, especially when I soon found out, that it pointed to the bay aforesaid (Whitby). I lost it sometimes by the interposition of valleys, rivulets, or the exceeding great quantity of ling growing on these moors. I had then nothing to do but to observe the line, and riding crossways, mu horse’s feet, through the ling, informed me when I was upon it. In short, I traced it several miles, and could have been pleased to go on with it to the sea-side, but my time would not allow me……………..

From the camp (Cawthorn) the road disappears towards York, the agger being either sunk or removed by the country people for their buildings. By taking the line, as exactly as I could, for the city, I went down the hill to Thornton-Risebrow, and had some information from a clergyman, of a kind of camp at a village called vulgarly Barf; but corruptly, no doubt, from Burgh (Barugh). Going to view this place, I was agreeably surprised to fall upon my long lost road again; and here plainly a small intrenchment on it; from whence, as I have hinted elsewhere, the Saxon name Burgh might come. The road is discernable enough, in places, to Newsam-bridge over the river Rye; not far from which is  a mile-stone of grit yet standing. On the other side of the river the stratum, or part of it, appears very plain, being composed of large blue pebble, some of a tun in weight; and directs us to a village called Aimanderby (Amotherby) (Drake, 1736, pp. 35-6)

Fig. 6  Photo of the causeway, taken in 1912.The raised foundation layer beneath the slabs can be clearly seen, although the camber claimed b y Hayes and Rutter (Hayes & Rutter, 1964, p.87) is not so evident.

Extract from George Young’s ‘A History of Whitby’,

“The foundation is usually a stratum of gravel or rubbish, over which is a strong pavement of stones placed with the flattest side uppermost, above these another stratum of gravel or earth to fill up and smooth the surface, the middle higher than the sides, which are secured with a border of flat stones placed edgeways, the elevation was in many places two or three feet….”

(Young, 1817, p. 706)

It is unclear as to whether Young was describing an original layer above the stone slabs, or perhaps a layer of soil built up over time. It is certainly quite different from Drake’s  account of the feature being paved with flint pebble. Known examples of ‘hard bottoming’ usually have a layer of smaller stones above them, with a surface dressing of gravel above that, as described by Drake (although flint seems unlikely as there is none in the vicinity). It should be noted that traces of small stones and gravel have been recorded between the slabs (Hayes & Rutter, 1964, p. 87). In order to make an even surface on the stone substructure, this top layer would have had to have been at least 15cm thick, and if it contained any stones at all, where has it gone? The robbing out of the large stones over much of its length is understandable, but deliberate removal of a top layer and leaving  behind the very stones which would have been most valuable for walling and building makes little sense. It would therefore have to have been eroded away, either by use or by natural means. Even if the feature had been a road after the Roman period, some trace of the top layer would surely have survived outside the main path in the centre, most particularly near the raised kerbs. There is none along the entire length. Given the speed with which vegetation takes hold on the moor which would quickly stabilise the remains of any such layer, such erosion seems improbable.  

This absence has never been properly explained, and might be seen as indicating that the stone slabs were in fact the top layer of the feature. Only 4% of roads in Britain have a claim to being paved (Davies, 2002, p.60), and when each claim is examined closely, the majority are either within forts or settlements or based upon questionable evidence (ibid), leaving just a few possible examples, none of which have been properly verified. When taking into account the uneven-ness of the surface, interpretation of the slabs as paving seems implausible.  

Upstanding stones

Drake’s description refers to what he thought was a milestone in the middle of the causeway. It is no longer there but any such unlikely stone would have impeded passage of wheeled traffic along a road. Other protruding stones do survive, in particular four massive stones at SE 809984, the tallest of which protrudes 17 inches (43cm) above the ‘surface’ of the slabs. To be used as a roadway, they would have needed to be covered with a top layer of stones and gravel at least 45cm thick, and for such a layer to leave no trace lacks credibility . Upstanding stones in the centre of Roman roads are known from a few examples, for instance on the Devil’s Causeway in Northumberland, near Edlingham, but they only protrude a small amount and are assumed to have been level with the original surface, perhaps serving as a guide for the minimum  height of the camber.


The kerbs are also problematic. Kerbs are a rare feature in Britain occurring on just 7% of Roman roads (ibid. p.63), although it is possible that they have been robbed out from some roads. The kerbstones on Wade’s Causeway are a very prominent feature, being mainly thin slabs set on edge and projecting between 75mm and 150mm above the surface of the slab metalling. The use of thin slabs like these for kerbstones is not known in any British Roman road and for that matter the use of kerbs projecting from the slab layer would also be unique. Upstanding kerbs are common enough in southern Europe, but not in northern Europe since with the much higher rainfall they would negate the effect of the camber and prevent proper drainage. It is conceivable that they were intended to retain the edge of the upper layer and that their tops would have been level with the road surface but as that upper layer is entirely absent, this is mere speculation.


Ditches flanking the agger are a standard feature of Roman roads in Britain. They vary in width and in depth and, in upland areas, are generally adjacent to the agger with no gap or ‘shoulder’ in between, although when traversing slopes there is often only one ditch, on the upslope side. A width of around a metre and a depth of perhaps 50cm, would not be unusual. Where ditches have not been found next to Roman roads, this can be explained by there being a substantial shoulder between agger and ditch and the excavation simply hasn’t gone far enough. They can also be difficult to interpret if a road has been widened or rebuilt. If accurate detailed data were available, which currently it isn’t, it may transpire that every Roman road in Britain had at least one ditch. Wade’s Causeway, on the other hand, appears to have none, at least none that resemble Roman roadside ditches. Very shallow ditches some three yards wide (2.7m)  and only 12 - 15 inches deep (30-45cm) were recorded on Flamborough Rigg (Hayes & Rutter 1964, p.87), which may have provided the material for the foundation layer, but these are most unlike any other known Roman roadside ditch.  The gullies Young describes could easily have been like these, or could have been shallow hollw-ways, which are certainly common in the vicinity of the Causway

Crossings of Watercourses

Blaise Vyner observed that the causeway is broken by several watercourses, without any sign of there ever being bridges or fords (Frere & Fitts 2009, p.277). Indeed, the feature seems to take no obvious action to avoid watercourses or ease the gradient in stream valleys by deviating from the planned course by heading upstream slightly, then returning to the course on the other side of the stream valley, as is the usual Roman method of construction. Hayes and Rutter’s claims that a crude pile of stones in the middle of Wheeldale Beck is the remains of footings for a timber bridge pier (Hayes & Rutter, 1964, p.88) lack any credibility especially as there is no evidence of any dressed stone to support the claim.

Relationship to Cawthorne and Lease Rigg

As illustrated in Fig 2. (Shown again opposite), Wade’s causeway appears to survive as a buried feature north of Cawthorn camps. If Hayes and Rutter were correct in their interpretation, the Causeway heads straight up the scarp slope and met the Roman road known to run through Camp D, marked in yellow on the lidar image (fig. 2). It seems, however, that there is no real relationship between the two features, they just happen to cross paths. The characteristics as revealed on lidar seem completely different; whilst the Roman road is straight and relatively narrow, the Causeway is much broader and not as well defined. Whilst they could conceivably join, the impression from the lidar image is that they are two independent features.

At the northern end of Wade’s Causeway near the fort at Lease Rigg, the ‘accepted’ view is that the the Causeway is represented by the modern lane passing through the fort at an angle of about 60°. A road passing through a fort at such an angle would be extremely unusual and even if a road had existed before the fort was built, it would be expected that the road would have been diverted to allow for the usual internal arrangements of the fort. Yet, no sign of a road leading from a fort gateway has ever been observed, and a recent geophysical survey of the site (Lyall, 2017) failed to locate any such road. It is impossible to tell whether the fort was sited on top of a pre-existing feature, or if the feature, ie Wade’s Causeway came later and cut through the fort but it is certainly difficult to see the causeway as a road that could have served the fort.

In both cases, there does not appear to be any relationship between fort and causeway, other than the forts being sited upon the Causeway. Wade’s Causeway does not appear to be a road between the two.

Fig. 2  Lidar image showing Cawthorn camps and part of Wade’s Causeway still surviving. Outlined in yellow is the short length of extant Roman road running through Camp D (actually a fort) - note the difference in scale between this feature and Wade’s Causeway.

Fig. 7  Plan of Lease Rigg Roman Fort, Egton, showing the lack of relationship between the fort and the modern lane, apparently on the line of Wade’s Causeway. After RCHME Survey, 1976.

It would appear that the principles behind the planning of the feature known as Wade’s Causeway are very similar to ones employed by the Roman army, principles not known again in Britain until the late 17th century, or to the best of our knowledge prior to the Roman period. The laying out of the work on the ground, however, falls short of Roman standards. Similarly, the construction is superficially similar to Roman work, but atypical in every aspect. It is almost as if the people who built it were trying to achieve Roman standards, but didn’t know quite how to do it, although that could easily be a false impression. That recurring atypicallity does, however, lead to an inevitable conclusion that it is not Roman.

Indeed, if its course didn’t fall between two Roman forts, it seems highly unlikely that it would still be interpreted as a Roman road by anyone. Given the upstanding stones, we can’t even be certain that the structure is actually a road at all. If not a road, what could it be? Blaise Vyner’s suggestion of it being a prehistoric boundary marker (Vyner forthcoming) seems feasible, except that all known boundary earthworks incorporate large ditches and high banks, and this has neither.

Ascribing a date to the structure is difficult, except to say that it seems unlikely to be post Roman, although that can’t be ruled out. If the accepted line of Wade’s Causeway is correct at Lease Rigg, then it seems most likely that the feature pre-dates the fort. There is no doubt that the Romans sometimes sited forts close to, and some times upon, features which may have had significance to the local population. For example, Newton Kyme fort near Tadcaster is next to a Bronze Age henge, and the Iron Age hill fort at Hod Hill, near Blandford Forum, has a Roman fort built within in. Recent carbon-14 dates obtained from beneath Wade’s Causeway proved not be as conclusive as was hoped, although may indicate a possible late Bronze Age or early Iron Age Date.

The way Wade’s Causeway is constructed might suggest a knowledge of the basics of Roman methods without an understanding of the detail or principles but that cannot be proven.

Whilst the camps survive as well as they did three centuries ago, Wade’s Causeway sadly does not. The enclosure of large areas of the moors in the 18th and 19th centuries created fields which needed walls, and so the feature became in effect a quarry, plundered for its ample supply of stone. Many examples of this are well recorded, such as that at Keys Beck Lodge (Codrington 1903, p. 165), as is its destruction to provide materials for the Stape-Egton road (Hayes & Rutter, 1964, p.11). In the 20th century destruction continued at many points along its length as it was ploughed up during the attempted improvement of agricultural land (ibid. p.12). Today, part of the feature that Drake saw below Cawthorn camps still survives as a buried mound visible on lidar (fig.2), but apart from that it is scarcely visible anywhere except for the one and a quarter mile section in the care of English Heritage. Most of that is now covered with vegetation.  

The only study to ever attempt to quantify how many roads used a foundation of large stones showed that just 26% of Roman roads in Britain utilise what was termed ‘hard bottoming’ (Davies, 2002, p.58).  Where large stones have been used, they are generally graded stones of similar sizes, thus enabling a more regular construction, for example on the Stanegate west of Corbridge is “…built upon a layer of large stones, sometimes 6-12 inches square and well laid” (Margary, 1973, p.446), or on RR712 over the moors above Huddersfield (fig. 5). Wade’s Causeway, on the other hand, incorporates stones that often exceed 30 inches (70cm ) (Hayes & Rutter, 1964, p.85) and in extreme cases are 5ft long. There are no other known examples where stones as large as these have been used routinely in the body of a Roman road construction in Britain.

When such foundation stones were used by Roman engineers, they were usually laid directly upon the subsoil, a fact acknowledge by Hayes and Rutter (ibid, p.86). In the Huddersfield example at Upper Holme in the Colne Valley on RR712 and some others, there was a layer of compacted shale beneath the large stones which was used to both raise the height of the road, and enable the creation of a camber so that water would naturally drain off to either side. This technique was also utilised in Wade’s Causeway where the stones sit above a cambered layer comprising gravel and subsoil sometimes mixed with small stones and the occasional larger slab, this cambered layer being generally between 12 and 15 inches (30 - 40 cm) thick, only a little more than the Huddersfield example.

Above the foundation

Wherever ‘hard bottoming’ was employed on a Roman road, whether or not there was another layer beneath it, there was always a substantial layer above, as can be seen on RR712 (fig. 5). At Wade’s Causeway, this is perhaps the most problematic issue since there is none of this layer surviving. In 1817, the structure was described as follows:

Entry compiled by Mike Haken, last updated 15 November. 2017                   

Fig. 4  Aerial view of Wade’s Causeway on Wheeldale Moor, revealing it’s lack of straightness

Fig. 5  Excavation of RR712 near Huddersfield, showing stone foundation layer

Margary's Roman Roads in Britain, road 81b


Wade’s Causeway


Back to Top Back to Top Back to Top

Wade’s causeway

Historic Counties:

Roman Sites on Route:

Historic Environment Records, HE Pastscape and other records