At several places in Scotland, the archaeological plan contains a tangle of Roman camp boundaries. Some circuits overlap; others share part of a side or occupy neighbouring ground. These concentrations record repeated choices about where to stop, rather than a single settlement laid out at one time.
Rebecca Jones describes important examples near Ardoch, Camelon, Castledykes, Newstead, Glenlochar and Dalswinton as gathering grounds. The term describes the use of a place. It does not identify a special shape of Roman ‘assembly camp’.
Routes and crossings
A suitable camping ground needed space, access to water and a position that worked within the army’s movements. Valleys and river crossings repeatedly offered those advantages. A junction could also allow troops to arrive together and depart in different directions.
At Castledykes, camps occupy a location connected with several routes through southern Scotland. Around Newstead, the River Tweed and the permanent military base formed part of a wider landscape of movement. At Glenlochar, the Dee crossing is central to understanding the complex.
An occupied fort might provide supplies or receive them from troops on the march. Nevertheless, camps beside a fort need not all belong to its occupation: some could precede it and others follow its abandonment.
Camelon and Lochlands
The concentration at Lochlands beside Camelon is particularly complex. Jones’s 2011 synthesis distinguished some fifteen structures representing at least eighteen occupations. Those are the counts in that study, not a claim that all the remains are fully understood or a current total of separate armies.
Its importance lay partly in movement through the Forth–Clyde isthmus. Repeated use of such a position makes sense without requiring every enclosure to belong to the same campaign. The surviving pattern records accumulated military activity.
Three different forms of return
| Evidence | What it may establish |
|---|---|
| Separate camps close together | Repeated or concentrated use of a useful locality; not necessarily simultaneous occupation. |
| One enclosure intersecting another | A possible sequence, best established through excavation at the intersection. |
| A ditch recut along an earlier circuit | Renewal of an existing enclosure; the interval and historical context still need dating evidence. |
These distinctions matter when a plan is coloured by period. An outline reconstructed from cropmarks may be secure enough to identify a camp but not to determine whether it is earlier or later than its neighbour. The smaller enclosure should not automatically be assumed to be the later one.
Ardoch and Dunning
At Ardoch, intersecting camps and their relationships with other military features have made the site important for discussions of chronology. Jones reviews which parts of the proposed sequence are demonstrated and which remain less certain. Even here, relative order and a named campaign are different conclusions.
Dunning illustrates reuse of an existing circuit. The 2014 trenches demonstrated ditch recutting but supplied no secure dates for the two occupations. Earlier finds of second-century pottery in the western titulus support an Antonine connection. A Flavian foundation followed by Antonine reuse is plausible, but the date of each ditch phase remains uncertain.
Why not always reuse the same enclosure?
Returning troops might save effort by adapting surviving earthworks, but an old camp was not necessarily suitable. The size of the force could change, or the ground might have been fouled by an earlier occupation. Ancient military writers warned about unhealthy campsites; Jones considers such practical factors alongside the archaeological evidence.
Those possibilities help explain what to investigate. They do not prove why any particular pair of camps was built. Distinguishing a sensible military explanation from a demonstrated archaeological sequence is essential.
Pennymuir and the changing ground at Beattock
Pennymuir offers a visible example of adaptation: a smaller enclosure uses part of the larger camp’s circuit, and substantial earthworks survive. Jones uses the site to show why the surviving plan can represent more than one occupation. Neither the smaller size nor the height of the earthworks supplies an absolute date. See Rebecca Jones, Roman Camps in Scotland, Callander’s Landscape (17:13 onwards).
At Beattock, the river landscape supplies another part of the sequence. Environmental work showed that alluvium accumulated over an earlier fortlet before camps at Barnhill were constructed. This supports later, Antonine activity there, in contrast with the likely Flavian Stracathro-type camp at Bankend. It is more precise to describe the stratigraphic relationship than to say simply that an entire northern terrace did not exist before the second century. See Rebecca Jones, Roman Camps in Scotland, Callander’s Landscape (44:39 onwards) and the South West Scotland research framework.
The shifting course of the Evan Water also cautions against treating today’s riverbanks as exact boundaries of the Roman landscape. The evidence helps explain different camp positions, without proving why each army chose its particular site.
Reading the map
The site maps are useful for comparing camps with rivers, routes and forts. They should be read as a record accumulated over time. A cluster may show the lasting usefulness of a place even when the dates of its individual enclosures remain unresolved.
Sources
Jones, R. H. (2011; open-access edition 2026), Roman Camps in Scotland. Edinburgh: Society of Antiquaries of Scotland, pp. 29–36 and 87–89. Page numbers refer to the printed pagination.
Campbell, E., Durnan, J. and Driscoll, S. (2014), Excavations at Dunning Roman Camp: Data Structure Report.
Return to Roman Camps in Scotland.
