Once an invading Roman army had secured an area and moved on, the general would often leave behind a small garrison, usually of auxiliaries, housed in semi-permanent structures built of turf and timber, in order to police the recently-subdued natives and also to secure the retreat of the army should it become necessary. Occasionally these would be augmented by the construction of a larger legionary camp or fortress to provide strategic reinforcements, perhaps also by a system of signal-stations or watch-towers to provide communication between these permanent fortifications. There is ample evidence that Agricola instigated the construction of many such fortifications in the wake of his army, also that he planned to complete the conquest of Scotland aided by the construction of several large forts situated at the mouths of the glens leading into the Central Highland Massif. These "Glen" forts would be used as "springboards" to launch his army into the heart of the Highlands, and would be provided with support from the rear by the building of a new legionary fortress at Inchtuthil on the Tay. Although it is possible that these permanent installations were initially laid-out during the tenure of Agricola it is very likely that the forts were actually completed by his successor to the post, Sallustius Lucullus, a governor who was also to fall foul of Domitian, but, unlike Agricola, was recalled to Rome and put to death, no doubt in some grisly manner, as was the modus operandi of this deranged and megalomaniacal emperor.
It is likely that work started on the Inchtuthil fortress no earlier than 85AD, the year after Agricola left Britain. This legionary base was abandoned uncompleted perhaps in the winter of 85/86, certainly by the summer of 86 when the removal from Britain of Legio II Adiutrix for use in Domitian's Dacian wars forced the Twentieth Legion to be withdrawn from Scotland to maintain the garrison at Chester, recently vacated by the Second.
At the same time as the Inchtuthil legionary base was being built, a number of forts were also constructed at the mouth of each glen leading into the Caledonian highlands, to the south-east of Inchtuthil at: Barochan on the Clyde, Drumquhassle at the SE corner of Loch Lomond, Malling at Menteith, Bochastle at Callander, Dalginross at Comrie and Fendoch at the mouth of the Sma' Glen, also to the north-east at: Ardoch, Strageath, Bertha, Cargill, Cardean and Stracathro. There may also have been another fort situated in the anomalously large gap between the latter two forts.
Ardoch may have been garrisoned by an auxiliary unit and a legionary cohort, and excavations at Fendoch and Cardean have demonstrated that both forts were evacuated after only a short occupation period. A number of bronze asses of 86AD have been uncovered at a number of military sites in Scotland; Inchtuthil, Camelon, Strageath, Stracathro, Crawford, Newstead, Cramond, Castledykes and Barochan. The finding of these coins, all in almost unworn condition, proves that all of these sites were garrisoned in that year or shortly afterwards. It is almost certain, given the stong garrisons in the glen forts that the Romans intended these camps to act as a springboard into the Caledonian highlands along the lines of the Glens. These plans were halted by the advent of war in Dacia.
"The abandonment of Inchtuthil was almost certainly accompanied by the withdrawal of the garrisons of the glen forts." (Breeze, p.61)
|ALATA CASTRA (Inchtuthil, Tayside)||NO1239||Legionary Fortress Legio XX Valeria.|
|CORSTOPITVM (Corbridge, Northumberland)||NY9864||Large Fort.|
|TRIMONTIVM (Newstead, Borders)||NT5734||Large Fort.|
|Castledykes, Strathclyde||NS9244||Large Fort.|
|Dalswinton, Dumfries & Galloway||NX9384||Large Fort.|
|CAMELON (Camelon, Central)||NS8680||Large Fort.|
|ALAVNA (Ardoch, Tayside)||NN8309||Large Fort [Glen Fort]|
|Cardean, Tayside||NO2846||Large Fort [Glen Fort]|
|Stracathro, Tayside||NO6165||Large Fort [Glen Fort]|
|Bertha, Tayside||NO0926||Large Fort of uncertain occupation period [Glen Fort?]|
|Broomholm, Dumfries & Galloway||NY3781||Medium Fort in Lowland Scotland.|
|BREMENIVM (High Rochester, Northumberland)||NY8398||Medium Fort in Lowland Scotland.|
|Oakwood, Borders||NT4224||Medium Fort in Lowland Scotland.|
|Milton, Dumfries & Galloway||NT0901||Medium Fort in Lowland Scotland.|
|Easter Happrew, Borders||NT1940||Medium Fort in Lowland Scotland.|
|Loudoun Hill, Strathclyde||NS6037||Medium Fort in Lowland Scotland.|
|Elginhaugh, Lothian||NT3267||Medium Fort in Lowland Scotland.|
|Barochan, Strathclyde||NS4169||Medium Fort on the Forth of Clyde [Glen Fort]|
|Drumquhassle, Central||NS4887||Medium Fort at SE end of Loch Lomond [Glen Fort]|
|Menteith, Central||NN5600||Medium Fort at mouth of Glen Ard [Glen Fort]|
|Bochastle, Central||NN6107||Medium Fort at mouths of Strath Gartney and Strathyre [Glen Fort]|
|Dalginross, Tayside||NN7721||Medium Fort at mouths of Glen Artney, Glen Earn and Glen Lednock [Glen Fort]|
|Fendoch, Tayside||NN9128||Medium Fort at mouth of the Sma' Glen [Glen Fort]|
|Strageath, Tayside||NN8918||Medium Fort at W end of Gask Ridge [Glen Fort]|
|Cargill, Tayside||NO1637||Medium Fort just S of confluence of Rivers Isla and Tay [Glen Fort]|
|Glenlochar, Dumfries & Galloway||NX7364||Medium Fort of uncertain occupation period.|
|Ward Law, Dumfries & Galloway||NY0266||Medium Fort? Fortlet (Lantonside) and Camp? of uncertain occupation period.|
|BLATOBVLGIVM (Birrens, Dumfries & Galloway)||NY2175||Small Fort in southern Scotland.|
|ALAVNA (Learchild, Northumberland)||NU1011||Small Fort; the northernmost Roman fort in England.|
|Cappuck, Borders||NT6921||Small Fort in south-east Scotland.|
|Mollins, Strathclyde||NS7171||Small Fort between the Forth and the Clyde.|
|Chew Green, Northumberland||NT7808||Fortlet/Station on the England/Scotland border.|
|Gatehouse of Fleet, Dumfries & Galloway||NX5957||Fortlet/Station in south-west Scotland.|
|Castle Greg, Lothian||?||Fortlet/Station.|
|Kaims Castle, Tayside||NN8612||Fortlet on the Gask Ridge.|
The war waged in ad83 by Domitian against the Chatti tribe east of the Rhine resulted in the withdrawal of many forces from Britain and a halt to the development of the fortress at Inchtuthil when the Twentieth Legion were forced to pull back to the recently abandoned fortress at Chester following the withdrawal of the Second Legion Adiutrix to the continent. The conclusion of this war was to see the division of Germania into two provinces, Superior and Inferior, along with a marked change of Roman military policy when a series of timber-built watch-towers were erected, spaced between 500-600m apart and fronted by a continuous timber palisade which stretched across the 120 miles (200km) of forested land between the natural barriers of the Rhine and the Danube. The concept that the Roman empire actually possessed limits began to emerge.
This policy was continued in Britain around ad86 when a series of watch-towers were built along the Gask Ridge in Scotland. These small encampments were spaced between 760-1,520m apart, and were seemingly built to replace the Glen forts, most of which were being abandoned at this time, and their garrisons perhaps withdrawn to the Continent. A secondary line of defence to the south-east of the line of the glen forts had been occupied since Agricolan times, and upon the abandonment of these northernmost defences the forts at Ardoch and Strageath to the rear of this line underwent a period of secondary (or perhaps tertiary) occupation, and a (new?) fortlet was built at Kaims Castle. A number of watch-towers were built along the road between these fortifications, the only discernable difference between those built between Ardoch and Kaims Castle (including Westerton) and the others along the Gask Ridge to the east of Strageath is that the former possessed two enclosing ditches and the latter group possess only one.
|ALAVNA (Ardoch)||NN8309||Fort, Watch-Tower W of road|
|Shiehill-1||NN8511||Watch-Tower W of road|
|Shiehill-2||NN8512||Watch-Tower E of road|
|Kaims Castle||NN8612||Fortlet W of road|
|Westerton||NN8714||Watch-Tower E of road|
|Parkneuk||NN9118||Watch-Tower S of road|
|Reith/Raith?||NN9318||Watch-Tower S of road|
|Ardunie||NN9418||Watch-Tower S of road|
|Roundlaw||NN9518||Watch-Tower N of road|
|Kirkhill||NN9618||Watch-Tower S of road|
|Muir o'Fauld||NN9818||Watch-Tower S of road|
|Gask House||NN9919||Watch-Tower S of road|
|Witch Knowe||NN9919||Watch-Tower N of road|
|Moss Side||NO0019||Watch-Tower N of road|
|Thorny Hill||NO0220||Fortlet, Watch-Tower N of road|
|Westmuir||NO0220||Watch-Tower N of road|
Pottery evidence points to all of these fortifications north of the Clyde/Forth being abandoned sometime before ad90.
Work conducted by David Walker on the Sacred Spring at Bath yielded a number of bronze coins of emperor Domitian which were seen to possess a very uneven distribution. The coins dated ad81 to 85 are in quite short supply as are those dated between 88 and 96, whereas the years 86 and 87 are very well represented. Further work on Domitianic bronzes by David Hobley at eleven major sites in Britain showed a similar situation, where the years 86 and 87 heavily outnumber those of the preceeding and following periods.
|The Chronological Distribution of Domitianic Coins in Britain|
|Above table based on data from Hobley (1989).|
It is extremely significant that no coins minted after ad86 appear in any Agricolan fortifications in north-east Scotland.
According to Hobley (1989), the last coins indicating Domitianic occupation at the following forts were all minted in ad86: Stracathro (Tayside), Inchtuthil (Tayside), Dalginross (Comrie, Tayside), Strageath (Tayside), Camelon (Strathclyde), Elginhaugh (Lothian) and Crawford (Strathclyde).
There is evidence of late Flavian occupation at: Newstead (Borders), Cappuck (Borders), High Rochester (Northumberland), Learchild (Northumberland), Broomholm (Dumfries & Galloway), Milton (Dumfries & Galloway), Dalswinton (Dumfries & Galloway) and perhaps Loudoun Hill (Ayrshire/Strathclyde). It should be noted that there are a further seventeen known Flavian forts for which no coinage evidence has been acquired.
Newstead is an extremely interesting case because two 'almost mint' COS•XII (ad86) coins were recovered from the ditch of the Phase I fort, and a COS•XIII (ad87) coin was found in the Phase II fort. This would indicate that at the same time as territories were being abandoned north of the Forth-Clyde, the garrison here in the Scottish Lowlands was being re-assessed, and the fortifications undergoing modification, possibly indicating a change of the garrison unit.
Hobley extended his argument to the movement of legionary forces of Britain. On reviewing the coinage evidence from Wroxeter and Chester he noticed that whereas at the latter site there was an almost equal amount of COS•XII and COS•XIII coins (16 & 18 examples respectively), at the former fortress there was a distinct lack of COS•XIII coins (2 examples) when compared with those of COS•XII (17 examples). This, he argues, lends credence to the belief that when the Twentieth demolished and abandoned their fortress at Inchtuthil and subsequently withdrew from Scotland, upon returning to the Welsh Marches they were not housed back at their old base at Viroconium (Wroxeter, Shropshire), but instead transfered their headquarters to the more favourable site on the Dee estuary at Chester, only recently abandoned by the Second Adiutrix which had been withdrawn from Britain for use in Dacia.
It is thanks to numismatic study that the troop movement in the immediate post-Agricolan era, the abandonment of north-east Scotland, the strengthening of the garrison in Lowland Scotland, and the re-organisation of the legionary forces in Britain, may all be pinpointed to the very narrow period ad86/87.