Roman Fort (roman)
- Council East Dunbartonshire
- Parish Cadder (strathkelvin-du)
- Former Region Strathclyde
- Former District Strathkelvin
- Former County Lanarkshire
NS67SW 16 6167 7253
See also NS67SW 20.
(NS 6167 7253) Roman Fort (R) (site of)
OS 6" map, (1971)
The Roman fort at Cadder was first noticed by Horsley (1732); a trial excavation was carried out in 1913, and the fort was examined in detail by Glasgow Archaeol Soc in 1929. By 1945, the last trace of Roman structures had been removed by sand- and gravel-quarrying.
The fort, which commanded a wide view to the N, was protected by moderately steep slopes on all sides except the E, where level ground gave easy access. Almost square on plan it measured 120m by 113m over the rampart, an area of approximately 1.4m ha (3.5 acres). The northern defence was formed by the Antonine Wall and its Ditch, the latter being interrupted by a causeway opposite the N gate. Excavation suggested that the fort was built at the same time as, or very shortly after, the adjacent stretch of the Wall. The rampart was of turf laid on a stone foundation 4.7m wide, and, except on the N, it was thought to have been revetted at the back by an earthen bank.
There were two ditches on the S and E fronts, the outer ditch originally terminating in an external clavicula on the left-hand side of each of the entrances. Clarke (1933) suggested that the claviculae might have belonged to an earlier Flavian ditch-system, which was re-used in the Antonine period, but although this interpretation is possible, claviculae are no longer thought to be exclusiveley Flavian in date; thus the ditch-system to which they belong could also have been dug in the Antonine period, although probably in a very early phase.
An irregular depression encountered immediately outside the outer ditch on the E may have been the remains of a third ditch, which was begun but not completed. The inner ditch was continuous along the W front, but the outer terminated at the SW corner. The principal internal buildings were of stone, but only the foundation trenches survived, packed with clay and cobbles. They comprised the principia (1), which exhibited two structural periods; a number of post-holes which were overlain by the principia evidently belonged to an occupation earlier than the Antonine fort, but their precise date is uncertain. Granaries (2, 3) lay immediately to the N and S. The praetorium (5), had originally been built of stone, but was almost totally demolished at the end of the first period and then replaced by a timber building, which itself exhibited two periods of construction. The internal bath-house (14) was in origin a simple block comprising three rooms, but according to the excavator, it had been twice added to or altered before a final phase in which its bathing installations were destroyed, and the building put to some other, unknown, use.
Of the remaining internal buildings, all of which were timber-built, six (6-9, 12-13) can be identified as barracks; it seems likely that there were only two periods of occupation. A similar sequance was observed in building 11, possibly a storehouse, in block 4, which subsequently became a dumping-ground, possibly in connection with the third-period use of building 10 as a smithy or workshop.
Although excavation showed that a number of buildings had been altered on at least two occasions in the Antonine period, the evidence of the barracks implies that the fort had only one change of garrison. However, although some features (eg. the pits and post-holes beneath both the principia and the guard-house at the E gate, or the construction trenches underlying the via principalis SE of sites 2 and 4) clearly belonged to an occupation that pre-dated the Antonine fort, it is impossible to say whether they were of Agricolan or proto-Antonine date. Flavian samian ware was found on the plateau in 1852, when the garden of Cadder manse was being laid out, but the precise location of the discovery is uncertain; on the other hand, the coarse pottery recovered by Clarke from the 'sleeper building' underlying site 4 was of 2nd century date. It seems reasonable to presume, therefore, that beneath the Antonine fort lay the remains of a proto-Antonine enclosure, probably similar to those underlying the forts at Croy Hill (NS77NW 10) and Bar Hill (NS77NW 8), and possibly, of the claviculae at the S and E gates represent part of its defences, roughly co-extensive with the fort.
The remains of an external bath-house were found immediately S of the Antonine Wall about 80m E of the fort, but no detailed examination was undertaken. As at Balmuildy (NS57SE 12), it probably stood within an annexe, whose precise extent is, however, unknown. Rubbish pits, presumably associated with an extramural settlement, were found outside the E gate, and quarrying disclosed the presence of timber buildings of Antonine date lying more than 90 m SE of the fort, between the fork in the Military Way and the outlying loop-road Y.
From the NW corner of the fort a broad ditch (A) ran southwards for about 250m along the face of the scarp that bordered the fort defences, and then followed the S edge of the plateau eastwards for a further 185m. Its profile varied considerably. Throughout much of its course on the W it exhibited a low mid-rib, but on the S it was initially V-shaped, while in the last 60m it became much narrower and shallower, and was accompanied on its inner site by a length of ditch of similar scale. On the S side of the fort there was also another ditch (B), only 1.5m wide and 0.9m deep, which was traced from E to W in a straight line for a distance of 175m.
The relationship of these ditches to one another and to the fort is uncertain. Clarke (1933) thought that B was the Palisade-trench, and A the accompanying outer ditch, of an Argicolan temporary camp, but the large size of ditch A, and the fact that on the S it does not run parallel to B, rules out this hypothesis. In favour of a 1st century date for A, Clarke noted that where it approached the NW corner of the fort its inner scarp had apparently been covered with upcast from the inner fort ditch, while at the SW corner the outer ditch of the fort stopped short on the edge of the scarp, as if respecting ditch A. On the other hand, if ditch A was already in existence when the Antonine Wall was built, it is curious that it was not packed solidly with turf or other material to prevent subsidence at the point where it was crossed by the Rampart. The nature and extent of the 1st century occupation at Cadder must therefore remain conjectural.
J Horsley 1732; J Buchan 1855; Proc Soc Antiq Scot 1906; G Macdonald 1915; 1934; J Clarke 1933; RCAHMS 1978, visited 1975.
Cadder fort has been completely destroyed by sand quarrying. No trace remains. (See also NS67SW 20.)
Material from this site is in the National Museum of Antiquities of Scotland (NMAS) and the Hunterian Museum, Glasgow.
Visited by OS (D S) 18 January 1957.
No change to the previous information.
Visited by OS (M J F) 24 June 1980.