Geolocated epigraphy: an introduction

This is an attempt at something a little unusual: providing an account of the inscriptions on stone along Hadrian’s Wall that actually shows where they were found (or, more often, thought to have been found). The data are therefore not only the texts and translations of inscriptions (and even some illustrations of them, although it will generally have to rely on out-of-copyright images such as 19th-century woodcuts or my own photography, where available), but also their geolocation and, moreover, a crude assessment of the reliability of the positional evidence. Thus a piece of an inscription excavated in the last few years will get a green marker as we know its exact position; one said to come from a locality would get a yellow one, indicating a measure of uncertainty; and one just thought to come from the general vicinity would get a red one.

RIB 1638 from Milecastle 38

RIB 1638 from Milecastle 38

Epigraphy, the study of inscriptions, has a long and distinguished history along the Wall, Horsley’s account of the monument in his Britannia Romana, featuring a major study of the then-known examples, most of them illustrated. This was just as well, as more than a few of those he records have since vanished, a warning to all that recovery from the archaeological record is often the worst thing that can happen to an artefact, of whatever size.

The definitive modern account of the stone inscriptions from Roman Britain is Roman Inscriptions of Britain (RIB), a work completed by R. G. Collingwood and R. P. Wright (but which owes much to a study begun by Francis Haverfield). RIB I dealt with the stone inscriptions up to 1955 (1956 in the case of instrumentum domesticum) and was later followed by RIB II dealing with the portable objects, and finally by RIB III which is a supplement to the stone inscriptions, bringing them up to date.

Entries within RIB give locational data, where it is known, so this has been used, together with the 1st edition Ordnance Survey County Series maps to identify the locations as accurately as possible using map regression techniques. Thus RIB 1299 was said to have been found on allotments 300yds to the west of Wallsend fort. These allotments have long since vanished under more recent housing developments, but are clearly shown on older OS maps, so an approximate (yellow) location can be restored for them by comparison with modern maps.

What follows will use the RIB groupings for inscriptions – by fort site and inter-fort lengths of Wall – to organise them, with a blog post for each group. Each blog post will also provide a clickable map showing the locations and the data associated with them (text, translation, and illustration, where one exists). The daily tweets (@perlineamvalli) will report the latest addition to be shown on a Google Map; here the inscriptions are listed in the left-hand margin and clicking on one will reveal the location and information bubble for it.

Inscriptions on Google Maps

Inscriptions on Google Maps

At the end, when we reach Bowness, we shall have a clickable map of all the inscriptions from the Wall which will go onto the Per Lineam Valli website as part of the atlas and which will become part of the Google Earth PLV file for all the components of Hadrian’s Wall.


Wall Mile 1

Wall Mile 1 [HB 141]

The Fossway continues to follow the line of the Wall for some 650m after the proposed site of Milecastle 2 until we reach a roundabout at the junction with Coutts Road and West Farm Road. At this point, the modern road veers off slightly to the south, whilst the curtain wall and ditch carried straight on. We walk for another 475m and then, on our left, we see a sports field changing room at Miller’s Dene, which is located just east of the probable location of Milecastle 1 (which Maclauchlan said was 6 furlongs west of Wallsend fort). The amount by which it is set back from the road shows how much the course of the latter has diverged from that of the Wall.

The Fossway on the line of the Wall ditch

The Fossway on the line of the Wall ditch

Milecastle 1 (Stott’s Pow) [HB 141; haiku]

Milecastle 1 (Stott’s Pow) has, unsurprisingly, never been confirmed, although occupation material has been found. Indeed, like Milecastle 2, it used to be placed further to the east on Ordnance Survey maps. Nevertheless, both Horsley and Maclauchlan reported seeing it here.

The site of Milecastle 1

The site of Milecastle 1


Wall Mile 16

Wall Mile 16 [HB 173]

The Trail leads us downhill towards a crossing of a side road (the usual caution is advised), and then past the lower northern reservoir. Nobody will be surprised to learn that the reservoirs have obliterated all trace of the Wall, even the Vallum, here.

Campbell and Debbeig's piece of curtain wall

Campbell and Debbeig’s piece of curtain wall

When Campbell and Debbeig were conducting their survey in 1749, prior to the construction of the Military Road, they found time to note a portion of curtain wall standing along here to a height of four courses. This of course was soon converted into the raw materials necessary for road construction, but the degree of their antiquarian interest is intriguing. One cannot but wonder at how broadly they interpreted their brief to survey a course for a new road so that it included mapping large parts of the Roman Wall as it then was. Indeed, their detailed survey seems to have been a major (and perhaps the only) source for the map drawn up by Nathaniel Hill for John Warburton’s rehash of the section of Horsley’s Britannia Romana that dealt with the Wall. And they tell us piracy is a modern problem in the publishing industry.

Wall Mile 16

Wall Mile 16

We start to climb steadily, now, towards Harlow Hill, the large stone ballast next to the reservoir making heavy going (it makes it easier if you walk on the grass to either side, where possible) and it soon saps the strength. We are on the line of the ditch and there is indeed a slight depression, but little by way of detail to be seen. It is worth briefly pausing to look back, for the reservoirs are spectacular, even though they are a modern addition to this historic landscape.

Once we reach the top, a short diversion through the edge of some woodland, usually boggy and beset with the plastic reinforcement mats you may already have seen in various places, and we are unceremoniously dumped on a narrow piece of worn verge that eventually turns into an equally narrow pavement. Passing the crest of the hill, we may note a junction opposite and immediately to the east of it is the measured position of Milecastle 16 (Harlow Hill).

Milecastle 16 (Harlow Hill) [HB 173; haiku]

Site of Milecastle 16

Site of Milecastle 16

No trace of this milecastle has ever been noted archaeologically, although antiquarian reports suggest it was once visible. Its prominent location on the hill meant that it would have enjoyed fine views towards Milecastles 17 and 18 to the west and 15 and the fort at Rudchester to the east.

Mapping Hadrian’s Wall

The earliest itineraries of Hadrian’s Wall are the enamelled pans (the Rudge Cup, the Amiens patera, and the Staffordshire Moorlands/Ilam pan) listing the forts at the western end, soon followed by the list of officials ‘per lineam Valli‘ in the Notitia Dignitatum. However, as William Shannon shows in his book Murus Ille Famosus, the earliest proper map is probably an 11th-century Anglo-Saxon one, showing the British Isles with a crude line and a corrupt caption that may once have read ‘murus pictis‘.

Horsley Wall map extract

Horsley Wall map extract

Generalised maps of varying quality proliferated in the 16th century but a significant development came with the publication of John Horsley’s Britannia Romana in 1732, which included a map of the Wall made for him by George Mark. This was the first serious attempt at a survey of the Wall. Soon after, the decision to build the Military Road from Newcastle to Carlisle led to a survey of the proposed line by Campbell and Debbeig in 1749. The two known surviving copies include a detailed record of the remains of the Wall, including what may be the first attempt at an elevation drawing of the curtain wall (from a section west of Harlow Hill which, ironically, would soon be destroyed by the construction of the road itself). Both Horsley’s text and the Military Road survey appear to have ‘informed’ the derivative publication by John Warburton of his Vallum Romanum in 1753.

MacLauchlan's survey of Birdoswald

MacLauchlan’s survey of Birdoswald

In the 19th century, the Duke of Northumberland commissioned Henry MacLauchlan (a former member of the Ordnance Survey who was an advocate of shading over contours to depict topography) to prepare a detailed survey of the Roman Wall, which included several site plans (like that of Birdoswald, seen here). MacLauchlan wrote his Memoir (published in 1865) to accompany the overall plan of the Wall and it was sufficiently highly thought of to be used in several editions of Collingwood Bruce’s Handbook to illustrate the course of the monument and can still be purchased from various outlets as an attractive historic print. Its publication more or less coincided with that of the Ordnance Survey’s First Edition series of one-inch and six-inch sheets.

OS Hadrian's Wall map

OS Hadrian’s Wall map: search Ebay to find used copies

Inexorably, the Survey’s involvement with archaeology led to the production in 1964 of its two-inch Hadrian’s Wall historical map (a copy of which is still on the wall of Housesteads museum). Distinguishing between visible and presumed (or known and not visible) elements superimposed on a clear map base, it also incorporated a simple but effective fold system which made (and still makes) it easy to use in the field. It was updated in 1972 but, in 1989, the Survey produced a completely new map of the Wall that was truly abysmal in a rich and bemusing variety of ways, not least the use of their 1:50,000 map base (photographically enlarged to 1:25,000 for the Central Sector) and a confusing series of conventions. Murophiliacs clung on to their precious copies of the older map. Screeds of irrelevant text and diagrams only served to obfuscate its purpose as a map, rather than an essay-cum-gazetteer.

51851The use of aerial photography for mapping became widespread in the latter part of the 20th century and this resource was exploited by English Heritage’s National Mapping Programme to revise the coverage of the Wall area. This information was ultimately fed into English Heritage’s brand-spanking-new 1:25,000 Archaeological Map of Hadrian’s Wall, first published in 2010. Not without its problems (its map base is greyed out in favour of graduated tones to indicate height – thereby, presumably unwittingly, supporting MacLauchlan’s view – and the awkward folds of the double-sided sheet render it much more cumbersome in the field than the old OS map), it has nevertheless provided up-to-date information and extended coverage down the Cumbrian Coast.

At the same time, the availability of rectified satellite and vertical aerial photography through virtual globes such as Google Earth meant that neogeographical techniques could now be applied to Hadrian’s Wall, and that was the origin of Per Lineam Valli which first went live online in 2007 and is about to see a major upgrade. The combination of separate layers for each of the major components with the ability to hyperlink each site to online resources such as the National Monument Record’s Pastscape database and the UK government’s MAGIC map base, together with plans and photographs of the monument. The rapid development of GPS-enabled smartphones has made it possible to use the Per Lineam Valli file in the field (as a layer in Google Maps Mobile) to tap into these resources.

What is next? The use of lidar is already making aerial survey topographically more accurate and airborne ground penetrating radar, currently being developed for landmine detection, may soon be added to the armoury of the archaeologist. On the ground, you can expect the nascent technology of augmented reality to enable you to see 3D reconstructions of the Wall in situ and in real time using a mobile phone or tablet, allowing a variety of interpretations of, say, a turret to be presented. ‘Paper’ maps (the English Heritage map is already made of waterproof polyethylene, not dead trees) may be supplemented (if not replaced) by flexible e-ink displays that will draw detailed mapping and data access ever closer.

Wall Mile 73

Wall Mile 73 [HB 357–8]

The Turf Wall was eventually replaced by the Stone Wall. It is thought the stretch between Milecastle 54 and Bowness was not constructed in stone until after the abandonment of the Antonine Wall in the AD 160s. This stone curtain wall was built to an average width of 2.45m (and has come to be known as the Intermediate Gauge wall, since it was partway between the two predominant widths of curtain wall in the east, the Broad and Narrow Gauges). It was furnished with new stone milecastles, whilst retaining the old free-standing stone turrets, against which the curtain wall butted.

Although most of Wall Mile 73 crossed Burgh Marsh, a short length of the Wall before Milecastle 73 has been identified where the ground begins to slope up from the marshes.


The edge of Burgh Marsh at Dykesfield

Meanwhile, back on the road, we leave the marsh and begin to climb gently at Dykesfield, after crossing a cattle grid; it is another drumlinoid, this one including the fort at Burgh-by-Sands. Away to the north-east, on the edge of the marsh, there is a Victorian reconstruction of a 17th-century monument commemorating the death of Edward I, just in case we had forgotten that it was feasible for an army to ford the Solway here. Less distant, but still remote from our present course, the line of the Wall runs to the north of the road and, as you might expect by now, there is little to see.

Milecastle 73 (Dykesfield) [HB 356–7; haiku]

The position of Milecastle 73 (Dykesfield) has been tested through geophysical survey and located on the ground sloping up from Burgh Marsh near Watch Hill, where Horsley thought there was ‘a castellum, for at this place they have dug up a larger quantity of stones than the bare thickness of the Wall could well have afforded’.