Home > Resources > Roman Roads












Resources on Old Scottish Roads

Roman Roads

William Roy - The Military Antiquities of the Romans in North Britain

  1. Notes on the "Roman" Roads of the One-inch Ordnance Map of Scotland, The Ayrshire Road, James MacDonald, PSAS, vol.27, (1892-1893), pp. 417-43
  2. Notes on the "Roman" Roads of the One-inch Ordnance Map of Scotland, Preliminary Remarks, James MacDonald, PSAS, vol.28, (1893-94), pp.20-57
  3. Notes on the "Roman" Roads of the One-inch Ordnance Map of Scotland, The Dumfresshire Roads, James MacDonald, PSAS, vol.28, (1893-94), pp.298-320
  4. Notes on the 'Roman Roads' of the One-Inch Ordnance Map of Scotland, The Roxburghshire Roads, James MacDonald, PSAS, Vol. 29, (1894-95), 317-28
    These papers call into question the work of early historians and antiquarians that identified many camp sites and roads as Roman. The papers were important at the time because of MacDonald's standing as one of the foremost archaeologists of the day. In an overview paper (Preliminary Remarks) he argues that much of the earlier work on the Romans in Scotland was speculative and faulty and gives an interesting account of how this state of affairs developed. In the three other papers he selects road systems in Ayrshire, Dumfriesshire and Roxburghshire, and attempts to show that although they might be Roman the evidence for this was very slight. Nowadays, many of the roads he questioned are accepted as Roman.

 Note on the Antiquity of the Wheel Causeway, Haverfield, F, PSAS, Vol. 34 (1899-1900), 129-30
 A short paper that lists references to the Wheel Causeway showing that it was in use in the Middle Ages.

A New Roman Mountain Road in Dumfriesshire and Roxburghshire, I A Richmond, PSAS, Vol 80 (1945-46), pps 104-117

It had long been speculated that the isolated fort at Raeburnfoot (near Eskdalemuir) was linked somehow to the Roman Road network. Although this was thought to have been on a north-south line, no evidence had been found for this. A suggestion by R P Hardie (Roads of Mediaeval Lauderdale) that an old road running across Craik Moor might be Roman led to this route (heading for Trimontium) being examined and accepted as Roman. The paper gives full details of the remains of the road and is illustrated by maps.
- see map (for illustration only)


The Roman Road to Raeburnfoot, A Graham, PSAS, Vol 82, (1947-48), pps 231-234

Referring to the above paper, the author notes work that confirms the road continued south-west of Raeburnfoot, presumably to join the main north-south Roman road north of Lockerbie. He also discusses the possibility of a route down the Esk to Langholm and Netherby (3 miles NNE of Longtown).


Roman communications in the Tweed Valley, Graham, A & Richmond, I A., PSAS, Vol.87, (1952-53), 63-71

An examination of the evidence for a Roman road running westward from Trimontium and Dere Street to link with the Roman roads in the Clyde Valley.


Two Roman Inscribed Stones and Architectural Fragments from Scotland (Ingliston Milestone, Inveresk Altar), G S Maxwell, PSAS, Vol 113, (1983), pps 379-390

The paper contains a discussion on how the Ingliston milestone could provide a clue to the route of Dere Street in this locality.


The Scottish campaigns of Septimus Severus, Nicholas Reed, PSAS, Vol 107, (1975-76), pps 92-102

The author looks at the available evidence to reconstruct the Severan campaigns of 208-210. He argues for a preliminary campaign against the Selgovae, followed by an advance through Fife, and then campaigns up the east coast, supported by the fleet. Two coins of the period depict bridges and he suggests that one showing a bridge of normal construction was built over the Tay near Carpow ( 5 miles ESE of Perth, near Newburgh), and the other, a boat-bridge, was built at Queensferry with a road linking the two locations.


The Roman Road in Annandale, Johnstone, J.T., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, II 9 85

This paper deals with a section of Roman road running north from Moffat to Little Clyde, some 10 miles. The writer notes that the road is more difficult to trace south of Moffat because of cultivation. Details of its construction at three separate locations are given. In view of differences from other Roman roads it is inferred that its main function was the rapid movement of troops rather than carrying heavy and constant traffic.


Notes on Roman Road through Annandale, Robertson, J., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 24 10

This paper gives a detailed account of the course of the road through Dumfriesshire from the River Sark to March Burn north of Moffat, noting where remains of the road can still be seen and discussing the possible route where traces have been lost. Details are also given of an excavation carried out at Holehouselinn.

North of Moffat, the old stagecoach road runs very close to the Roman road and the author discusses the course taken by both roads and gives an account of an excavation of a culvert on the coach road.


Excavations in Dumfriesshire, 1946, St Joseph, J.K., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 24 150

This paper deals with the Roman road running from Raeburnfoot across Craik Muir, probably to Trimontium. The author describes the remains of the road and gives details of three recent excavations. One of a mound just feet from the road at Craik Cross Hill did not give dating evidence but was very likely to have been a signalling post. Another was of the road nearby and the third was at the fort which gave some idea of its structure. Earlier finds suggest an occupation date in the Antonine period.

He then discusses the route the road might have taken to the south-west where it presumably joined the Annandale road.


Upper Nithsdale and Westward in Roman Times, Clarke, J., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 30 111

Recent discoveries had shown that, contrary to previously held opinion, the Romans had had a presence in south-west Scotland. The author lists these discoveries and discusses the extent to which the Romans might have penetrated further into Upper Nithsdale and to the west.

Looking at a possible route to the west from Carronbridge he notes a tradition that a road ran westwards up the Shinnel Water near to Tynron, over to Dalquhair (NX6599) and Polwhat Rig (NS6002). He examined a very old road running via Colt Hill (NX6999) over to the south of Windy Standard (NS6201). He suggests it might have been a patrol track for cavalry rather than a main road which strategic considerations suggest would have ran from Carronbridge to Moniave and then either towards Carsphairn or Dalry to join a north-south road from Glenlochar to the Ayrshire coast.

He then looks at the possibility that the Roman road extended northward from the Carronbridge - Durisdeer area. Field investigations resulted in a possible temporary camp being found near Kirkconnel, and a small fort at Bankhead Farm nearby was also found. Assuming the fortlet implies the presence of a road, he suggests a possible course would have been from Durisdeer up between Dalpeddar Hill and Auchensow Hill then fairly directly past Bankhead towards Corsincon Hill.


Roman Roads in S-W Scotland.

(1) At Glenlochar, Crawford, O.G.S., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 31 22
The author investigated a road leading north out of the fort at Glenlochar (3 miles NW of Castle Douglas) and followed faint traces for just over a mile to the south of Carmichael. With nothing found north of here he notes the possibility that it might have been a branch road leading off a north-south route following an older line of road than the A713 that could have led to the coast. To the north it could have run to the Ayrshire coast though he did not think the road running near Dalmellington was of that date.

He also investigated an old road running west from the Old Military Road near Milton (2 miles south of A75 near Crocketford) to see if it was Roman and could have led to Glenlochar. He describes what can be seen along its course (Glenarm, Barr, Hermitage, Old Bridge of Urr, Clarebrand, Crossmichael and continuing as the Kirk Road probably to Lauriston and Gatehouse) but concludes that although very old it was not Roman.

(2) From Nith to Dee, Inglis, J.A., III 31 25
The author gives an account of a search for two Roman roads in Dumfriesshire. One may have led from Glenlochar through Milton to Dalswinton and the author refers to the previous paper that had traced its course west of Milton, noting that although it had not been thought to be Roman, an air photograph suggested that further investigation might be justified.

East of Milton he suggests a route to Dalswinton leading north past Shawhead and over Dalquhairn Hill. A branch road may have left this road near Loaninghead and run north-east to Carzield.

see map (for illustration only)

(3) The Tweedsmuir Road, Bailey, H. & Blance, B, III 31 27

A discovery of a marble head at Hawkshaw (2 miles east of Moffat to Edinburgh Road and 9 miles north of Moffat) well off the line of the Roman road running north to Crawford suggested there might be a road near Hawkshaw.

The track from nearby Carterhope to Erickstane was examined but although it had features common to Roman roads could not be confirmed as such.

see map (for illustration only)

(4) From Castledykes (Corbiehall) to Crawford, Radford, C.A.R., Reid, R.C., and Truckell, A.E., III 31 30

The authors note that the Roman road runs from Crawford to Inveresk when it might be thought that there should also have been a road to the important Castledykes site.

In searching for a road they followed a short stretch out of the south gate at Castledykes and then over Carmichael Hill to Howgate Farm and Howgate itself noting features suggestive of a road. South of Howgate they found remains of a road close to the minor road leading to Roberton where there may have been a bridge connecting with the Inveresk road on the other side of the river.


Notes on Roman Roads, Reid, R.C., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 32 73

The author gives examples of early charters that have references to early, possibly Roman, roads. The first deals with the Roman road running through Raeburnfoot which must ultimately have led from Newstead to Galloway. A 13th century charter refers to a royal road running from Annandale to Roxburgh which in part at least must be the Roman road and its course might be ascertained if the placenames in the charter could be identified. There is another reference (in the Liber Melros) to a royal road running from the Dunscore area, where the monks had lands, up towards Glencairn.

Another example he gives is a charter of 1558 relating to Wandel in Lanarkshire which refers to the "Merchant Way", which would help determine its crossing point on the Clyde.


Erickstane Brae, Prevost, W.A.J., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 37 111
The author gives early references to the Erickstane, north of Moffat, and then discusses the various roads in this area and historical events associated with them. A useful map is included.


More Notes on Roman Roads, Reid, R.C., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 37 123

This paper details research into the course of the Roman road running across Craik Muir both to the north-east and to the south-east over to a fort near Langholm. It also looks at the possibility of a Roman road running north through Ayrshire.

In view of the distance from Raeburnfoot to Newstead a search was made for a fort near to Roberton, particularly at a place called Africa. Two sites with this name were examined but no Roman evidence was found, nor could the course of the road be identified. The name itself was a Galloway one and may have referred to a woman called Affrica who lived in Nithsdale in the 1200's.

The author then examines the evidence for, and the various proposals that had been put forward for the likely course of a road between Raeburnfoot and the fort at Broomholm near Langholm.

With regard to a continuation of the Nithsdale road into Ayrshire (at the time traced as far as Carronbridge), the author suggests there may have been a fort at Hurlford, near Kilmarnock, if an east-west road from Loudoun Hill passed through there but notes that urban development make it unlikely that it could ever be found. He refers to recent excavations on the potential line south of Hurlford (see Clarke & Wilson below) and then discusses the possibility that the mediaeval toll points for Ayr (Mach, Karnebuth, Lowdun, Croseneton and Laichalpin) may have been sited on Roman roads and so furnish a clue to the course of the possible road.


The Road into Ayrshire. - Fortlet at Bankhead, Kirkconnel Clarke, J.- and Wilson, A., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 37 136

The Road into Ayrshire - A possible Roman Road between the Rivers Cessnock and Ayr: Clarke, J. and Wilson, A., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 37 143
Given that a much-decayed road had been found running past the fortlet at Bankhead near Kirkconnel there was a strong possibility that the road continued into Ayrshire. The paper is an account of excavations undertaken to see if any remains of such a road could be found.

Four roads were found just south of the Cessnock where there was a ford and at least one earlier bridge. One road in particular aligned on the ford, which suggested it had an early date. Two holloways leading to the ford were examined and one in particular with a bed cut out and a roadway laid with stones was suggestive of Roman construction.

They then looked at a point north of Mauchline to see if any early road diverged from later roads leading in to Mauchline. A much damaged road was found which bypassed the town leading to a ford on the River Ayr at Kingencleugh. A holloway associated with this ford showed signs of being cut and laid with cobbles and a parchmark had been reported south of the ford.

The authors recognise that the findings are suggestive only and that proof would only be available if a definite Roman structure was found associated with this line.

The Road into Ayrshire. - Roads in Upper Nithsdale and Beyond Clarke, J. and Wilson, A.Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 37 151

This paper examines the routes a possible Roman road running past the fortlet at Bankhead Farm could have taken.

To the south two routes would have been possible, one running between Dalpeddar Hill and Auchensow Hill and possibly coming from Durisdeer where there were marching camps; the other crossing the Nith at Sanquhar and running down its west bank to Carronbridge where there are also Roman remains.

To the north of the fortlet they suggest two possible routes as far as Corsencon Hill. Given a tradition of a Roman camp on Avisyard Hill and the strategic need for a base in this area. they then look at two possible lines to Avisyard Hill.

They then examine a possible line north of Avisyard with a ford over the Lugar at Cumnock, the fords over the Ayr and Cessnock examined above and the straight line of the modern road to Hurlford.

Again they recognise the need to identify Roman structures on this line if a road is to be proved.


Roman Road near Moffat - Observations during gas pipeline construction, Yates, M.J., Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III 58 95

The opportunity was taken during the construction of a pipeline to examine a Roman road at Muckle Hill near Moffat. The road was 8.5 metres wide and in cross section showed features consistent with Roman construction.


Roman Penetration in West Dumfries and Galloway, Allan Wilson, Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III, 64, 7

This is a thorough review of previous research into the Roman presence in this area, and the author's own findings.

He looks first at an east-west route, tracing it in sections to the Stranraer area through the forts at Glenlochar and Gatehouse, then at a possible route from Drumlanrig to the south west.

The marching camps at Girvan could imply a route to the west and he discusses the alternatives that had been proposed. Near to the Solway, temporary camps suggest a link to the Annandale road and there may have been a road from the coast up towards Drumlanrig and Durisdeer, and then into Ayrshire.

Possible routes along the Dee, Ken and Doon Valleys are disucssed along with a potential route to Girvan by the Cree valley. The paper concludes with an examination for evidence for a route from Stranraer towards Girvan.


Roman Penetration in E Dumfriesshire and Beyond, Allan Wilson, Transactions of the Dumfriesshire & Galloway Natural History and Antiquarian Society, III, 73, 17

Like his paper on west Dumfresshire this is valuable overview of evidence for Roman roads in the area. Evidence is examined for roads along the main valleys such as the Nith, Annan, Tweed, Esk, Ewes and roads that may have linked some of these.


A Roman Frontier Post and Its People: The Fort of Newstead in the Parish of Melrose By James Curle

See section on the Roman Road on pages 7-15


Annals of the Solway Until A.D. 1307 George Neilson, 1899

This is a detailed history of the Solway and surrounding country from Roman times to the death of Edward I on his last incursion into Scotland. It contains much interesting information about the fording points as well as references that will be found useful in reconstructing early routes. Some images of early maps are included.


The Roman Gask Project: Final Report on Roman Road excavations at Parkneuk 


A Possible Roman Road Cutting at Innerpeffray Library Perthshire. D.J.Woolliscroft, with contribution by B.Hoffmann.


A history of the border counties: Roxburgh, Selkirk, Peebles, Sir George Douglas, Edinburgh, Blackwood, 1899

Pages 4-6 on Roman roads and the Wheel Causeway

Roman penetration in Strathclyde south of the Antonine Wall, Part One: The Topographical Framework, Allan Wilson, Glasgow Archaeological Journal Vol.19, 1994-95
Part Two: Romanisation, Allan Wilson, Glasgow Archaeological Journal. Volume 20, Page 1-40, 1996-97


Atlas of Scottish History to 1707, edited by Peter G B McNeill and Hector L MacQueen, Edinburgh: The Scottish Medievalists and Department of Geography, University of Edinburgh, 1996. The Atlas contains maps with explanatory text for many aspects of Scottish history including Roman roads. Now available on Scotlands Places website.

Field Study Group Report, AANHS Collections, 2nd Series, Vol 3, 1955, p30-33

Clarke and Wilson, A Possible Roman Road between the Rivers Cessnock and Ayr, AANHS Collections, 2nd Series, Vol.5, 1959

A Wilson, Roads in Upper Nithsdale and Beyond, AANHS Collections, 2nd Series, Vol.5, 1959

The Roman Occupation of South-Western Scotland, edited S N Miller, Glasgow, 1952

Scotland's Roman Remains, Lawrence Keppie, John Donald Publishers, 1986

Maxwell, G.S., ‘The Evidence from the Roman Period’ in Loads and Roads in Scotland and Beyond, ed. A.Fenton & G.Stell, Edinburgh 1984 (John Donald), 22-48

Margary, Ivan D, Roman Roads in Britain, Vol 2, London 1957 (also 1969 and 1973)

The Romans in Perthshire, David Woolliscroft and Birgitta Hoffman, 2005, Perth & Kinross Heritage Trust and The Roman Gask Project

The Last Frontier: The Roman Invasions of Scotland, Antony Kamm, Tempus Publishing 2004 and Neil Wilson Publishing 2009