The area of a Roman camp is often the first figure given in an archaeological description. It is useful, but it can easily acquire a meaning that the evidence cannot support. A large enclosure indicates the scale of the accommodation provided; it does not give us a roll-call of the soldiers who slept there.
Rebecca Jones’s national study allows camps to be compared across Scotland. It also explains why apparently straightforward labels such as ‘30-acre camps’ or ‘legionary camps’ need care. Several different questions lie behind an area measurement.
What exactly has been measured?
A complete surveyed perimeter provides a different level of confidence from two stretches of ditch visible in a cropmark. Where a side is missing, the published area may be a minimum or a reconstruction. A small adjustment to a long boundary can change the total substantially.
Measurements should identify the individual enclosure, particularly at places with several overlapping camps. They should also state whether an attached annexe has been included. Hectares and acres describe the same ground: one hectare is approximately 2.47 acres.
The groups identified by Jones
| Area or proposed series | What the comparison shows |
|---|---|
| About 2–2.5 hectares | A concentration among Antonine Wall construction camps, with examples outside this range. |
| About 10–18 hectares | Numerous camps fall in this broad band. The older proposed ‘30-acre’ series does not form a consistently coherent group. |
| About 25 hectares | A comparatively coherent series, with similarities in proportions, six entrances protected by tituli and frequent annexes. |
| About 45 hectares | A proposed group of large northern camps; membership and the treatment of unusually small examples need scrutiny. |
| About 54 hectares | Another large-camp series, including examples with irregular outlines and incomplete gateway evidence. |
| About 67 hectares | Very large enclosures, including Newstead V; not every proposed member has a securely established complete area. |
These are research groupings, not a Roman classification recovered from an ancient document. Their approximate areas are also not thresholds: a camp just above or below a round number does not automatically belong to a different type.
Why the 25-hectare group matters
The strongest comparisons combine several features. Jones regarded the approximately 25-hectare series as the most homogeneous of the proposed groups. Many share similar proportions, a six-gate arrangement and an attached annexe. Their distribution provides an additional reason to investigate whether they belonged to related operations.
The shared design makes the series worth investigating, but does not settle its date. The road-over-camp sequence at Innerpeffray West challenges the old Severan attribution and supports discussion of an earlier, possibly Antonine context. The date of the road and the attribution of other members still need assessment. See Rebecca Jones, Roman Camps in Scotland, Callander’s Landscape (45:45 onwards) and our dating guide.
Why hectares do not equal soldiers
Camp interiors accommodated roads, command areas, stores, animals and baggage as well as tents. Different forces could bring different proportions of mounted troops and transport animals. Sloping or uneven ground might lie inside the defences without being suitable for accommodation.
The Dunning camp enclosed approximately 47 hectares, but its occupants cannot be counted from that figure. Equally, the very large enclosure at Newstead V should be distinguished from the permanent fort and its garrison. A temporary field army and a barrack-built base require different forms of interpretation.
Space for an emperor’s entourage?
Jones proposes that the exceptional scale of the largest Borders camps could reflect accommodation for Severus, the imperial guard and the administrative staff accompanying the emperor. St Leonards and Newstead V are important comparisons, alongside proposed large reconstructions at Pathhead and Channelkirk. This offers a possible explanation for extra space beyond an ordinary field force, not proof that the emperor slept in every enclosure. See Rebecca Jones, Roman Camps in Scotland, Callander’s Landscape (49:35–50:23 onwards).
The individual measurements still matter. Channelkirk and Pathhead have incomplete boundaries, so a proposed 165–170-acre series must not replace their recorded minimum areas. Equally, Logie Durno’s approximately 58.6 hectares make it much larger than the roughly 45-hectare northern group, even when it is discussed within the same broad campaigning landscape.
The approximately 52–56-hectare camps, including Kair House and Balmakewan, have often been associated with Severan operations. The proposed route and imperial context are interpretations built from several kinds of evidence. A view of the North Sea is not independent proof of a camp’s date or of an ancient writer’s geographical claim.
A useful way to compare camps
A sound comparison records the enclosure’s name, area, measurement uncertainty, shape, entrances, annexe and dating evidence. It then asks which similarities remain when the uncertain measurements are set aside. Estimates of capacity can be discussed as models, provided their assumptions are explained; they should never replace evidence for a particular unit.
For the organisation of accommodation and movement, see Inside a Roman marching camp; for one recurring layout, see Roman marching camps with six gates.
Sources
Jones, R. H. (2011; open-access edition 2026), Roman Camps in Scotland. Edinburgh: Society of Antiquaries of Scotland, pp. 76–79 and 99–108; see also the individual gazetteer entries. Page numbers refer to the printed pagination.
Return to Roman Camps in Scotland.
