The region’s auxiliary forts can be roughly divided by geographical location and association. In the lowlands, Ardoch (MPK665; Woolliscroft and Hoffmann 2006, 90–7 with further references), Strageath (MPK714; Woolliscroft and Hoffmann 2006, 111–4) and Bertha (MPK2048; Woolliscroft and Hoffmann 2006, 144–8) are situated to the west of the River Tay. They sit along the route of the main Roman road passing through the region via the Gask Ridge, often termed the Gask Line. East of the River Tay is Cargill (MPK3570; Woolliscroft and Hoffmann 2006, 150–5). All of these lowland forts along the line of the road have produced evidence of both Flavian and Antonine phases of activity.
Dalginross near Comrie (MPK308; Woolliscroft and Hoffmann 2006, 49–52) and Fendoch, Glen Almond (MPK1479; Woolliscroft and Hoffmann 2006, 53–61) are situated at the mouths of upland glens along the Highland Boundary Fault. As a result, they have been referred to in the past as ‘glen-blockers’ and now more frequently as the ‘Highland line forts’. Like the lowland forts, fieldwalking at Dalginross has produced both Flavian and Antonine evidence whereas only Flavian material is known from Fendoch. Using the example of later, Cromwellian and Hanoverian, invasion forces, Southern (1997) argued that the Highland line forts served as springboards for further campaigns using the local garrison. They suggested that these forts formed part of a wider plan to absorb the whole of Scotland into the Empire which was never fully realised when the army withdrew. Woolliscroft and Hoffmann have also considered the Highland line forts in conjunction with the lowland forts; they interpreted them as part of a single network designed to control both the land in the straths, the entrances into, and the exits from the Highland glens (2006, 231–2). In this scenario the Highland line forts are seen as advanced positions that served as early warning sites, although few have so far shown evidence for a connection to the signalling system spanning the Gask Ridge itself. While the springboard theories are convincing, it has been observed (Hoffmann pers comm) that this is realistically only possible with the legionary fortress at Inchtuthil (MPK3639) due to its position at the end of the main valleys to the north.
All of the auxiliary forts have received some degree of excavation as well as geophysical surveys and fieldwalking in the past with Strageath (Frere and Wilkes 1989) and Fendoch (Richmond and McIntyre 1935; Richmond et al 1939) representing the most extensively excavated. They are characterised by multiple ditches and annexes, with many forts occupying tactically strong positions that control river crossings and the approaches to upland glens. With the exception of Fendoch, the forts are also regularly associated with temporary camps which often date from several periods, overlaying each other or forming clusters. The varied extramural elements present at each of the auxiliary forts is an important feature, which merits further consideration in terms of what types of activity – settlement/manufacturing/temporary accommodation for advancing military units – and the types of people – military/civilian, Roman/non-Roman – they represent. Geophysical survey, fieldwalking and excavations have provided evidence of roundhouses and a souterrain in close juxtaposition with Cargill whereas at Bertha and Strageath there are suggestions of extramural road systems, Roman material and structures. Excavated in the 1980s (Frere and Wilkes 1989), Strageath is also particularly noteworthy for its unusual pair of granary buildings which are approximately twice as large as other auxiliary forts in the region and located opposite each other next to the western gate (see also Woolliscroft and Hoffman 2006, 111–4). A comparison with other fort granaries elsewhere in the Empire, such as Hofheim in Germany on the Taunus-Wetterau Limes, has revealed the Strageath granaries have a larger storage capacity than necessary for the fort garrison. This has led to the interpretation that the fort fulfilled a strategic supply function for forward-operating combat units (Dobat 2009, 41).
With an impressive array of extant banks and ditches, Ardoch has attracted attention since Roy’s survey (1793). It has long been recognised as one of the best-preserved timber forts in Scotland, one of the best survivals of Roman earthwork defences in Britain and indeed the wider Roman Empire (Maxwell 1989, 114; Keppie 2004, 162). Good knowledge of the fort layout and phasing exists from the early excavations, geophysical surveys and subsequent interpretations. Maxwell (1989, 165) describes Christison’s (1898) excavations as pioneering; it was one of the first instances where timber structures were identified through excavation in Scotland. Several attempts to decode the fort’s complex phasing have been made (MacDonald 1918; Crawford 1949; Breeze 1983). The associated tombstone is incredibly rare for Roman forts in Scotland; it offers a unique insight and etail into the first cohort of Spaniards who garrisoned the fort, most likely during the Flavian phase of activity (Keppie 1998, 112–3; Woolliscroft and Hoffman 2006, 94–95). Ardoch occupies a strategic position in the landscape and its importance over time and multiple campaign seasons is evidenced by no fewer than four camps, the largest being 130 acres (54 ha), in addition to an annexe, recorded to the north of the fort (Woolliscroft and Hoffman 2006, 90–6; Jones 2011).