Combining the evidence of the first excavator Curle (1911); Richmond (1950); and the modern search and rescue excavations of Bradford University (1987; 1989-93; and 1994); it is now possible to suggest a revised phasing for the fort and its annexes. This scheme relates primarily to the site's defences and road network, but an attempt has been made to link this with contemporary buildings and other activities in the interior of the settlement.

Newstead Military Complex
Phase One

The earliest occupation was that of the irregular Agricolan fort (see above) established about 80AD. Curle (1 91 1; and Richmond (1950) recorded this as having been defended by a rampart of earth, built up on a foundation of cobbles, and with two V- shaped ditches in front. The inner ditch measured 2.5 metres wide, and 1.5 metres deep: the outer ditch 2.7 metres wide, and 1 metre deep (according to Richmond's 1950 report p12).

A ditch terminus, encountered by Bradford University in a Curle trench in area RGT 1 0 1 (see Appendix 3, bottom) represents the onlv feature of this phase of the fort recovered during the modern excavation.

No plans of buildings have been recovered from this phase, although Richmond noted sill beam slots and a mass of waterlogged timber building debris lying under the stone foundations of the Domitianic fort (the name given to phase 2). This suggests that phase 1 was indeed a permanent fort, rather than a tented encampment

The earliest extramural development seems to have started on the West side of the fort. The West annexe, an enclosure defended by two ditches, appears to relate to tlus earliest phase of fort defences. In this phase the main approach to the fort (from Carlisle, Luguvalium, via Oakwood Fort near Selkirk), was probably from this side i.e.from the West

Phase Two

The ditches of the Agricolan fort were deliberately sealed, prior to the construction of the next phase of the defences. Amongst the fill in the ditches were unworn coins of AD 86, suggesting a construction date not long afterwards. This phase - phase 2 - is referred to as the Domitianic fort, after the then Emperor, Domitian.

The inner ditch of the preceding phase, phase 1 - appears to have been sealed in a relatively clean condition, indicating that no significant break in occupation took place.

The new defences represent a colossal strengthening of the fort. The rampart was again set on a cobble and rubble base, this time measuring 13.5 metres across, giving it an estimated height of 8.4 metres.

To this can be added the height of the breastwork viz 1.5 metres. (A large depression to the North-East of the fort recently suggested to be an amphitheatre, may have been the clay source for this colossal undertaking).

“ colossal defences”

Immediately in front of the huge rampart was a single huge ditch, 5 to 7 metres across and 2 to 4 metres deep (see Richmond 1950: p8 and section opposite p2; and Curle 1911, section opposite p30). The entrances to this phase appear to have been strengthened by elaborate outworks.

Only a tiny minority of the buildings from the interior of this fort have been identified, none during the modern programme.

Nevertheless, a major change would appear to have occurred from the preceding phase.

Building foundations, though not superstructures, were now in stone, considerably lengthening their potential lifespan.

The position of the main gates, and hence the more important roads, remained basically unchanged, so the approach to the fort probably continued to come from the West.

"first century annexes"

Settlements had now developed on the East and South sides of the complex, and during phase 2 annexe defences were added to protect those communities.

They closely parallel the design of the military defences viz a rampart, fronted by a single massive ditch (6-8 metres wide and 2-3 metres deep), very similar in dimensions to the fort ditch. The arrangement of the ditches suggests that gateways existed between the West and South Annexes and also between the South and East annexes.

The internal occupation of these first century enclosures (or annexes) remains poorly understood. Only in the South Annexe can the modern programme of excavation expect to have encountered evidence from this period; and only a tiny proportion of RGT 33 pottery dates to the late first century (see Appendix 6, top). However, the distribution of Curle's South Annexe deep pits containing early pottery suggests significant occupation, probably following an undetected East-West road, skirting the ditch of the fort. This may be represented by a pronounced scorch mark visible between the first and second century ditches on some aerial photographs.

The character and contents of this group of pits are rather different from the pits of the second century, perhaps indicating a different type of occupation – military rather than civilian.

"mansio” building

The permanent character of the fort's buildings perhaps suggests that this was also the period in which the 'mansio" building was constructed. Certainly the building was no later, the Curle recorded pottery types being confined to those of the first century.

This huge half-timbered construction clearly represented an official institution of some kind. However, RFJ Jones in ‘Military and Civilian in Roman Britain’' has suggested that similar structures at other sites probably functioned as official trading stations (cf Hudson's Bay Coy?), rather than post houses or inns, as the name suggests.

Some sort of bath-house, completely obliterated by its successor, is also suggested on the basis of primitive hypocaust tiles.

Phase Three

Curle and Richmond supported this assertion by noting that “Flavian" (first century) and "Antonine" (second century) pottery types were rarely found together in the same feature.

Richmond also noted that the silted nature of the Domitianic fort's single ditch supported the theory of prolonged disuse.

coin lists

Strong evidence is also provided by examination of the coin lists. Silver, in particular , remained in service for extremely long periods. The two coin hoards used for the comparison were the Castle Bromwich hoard (silver) and the 1905 Croydon hoard (copper), both deposited about 180 AD (see Macdonald 191 1, pp 400, 414). Tables, showing the distribution of coins by Emperor in each hoard as compared with the Tritnontium site, and their consequent "profile", are given on page 3. This technique, applied independently for both silver and copper coinage stronging suggests a hiatus in occupation during the reign of Trajan (98-117 AD).

However, both silver and copper coin numbers suggest significant re-occupation before both the Agricolan and Antonine campaigns.

In the case of the Hadrianic anomaly (more of his coins - 117-138 AD - than might be expected), this may reflect the presence of an outpost fort at Newstead. The coins of Vespasian and earlier, however, cannot be attributed to a fort, as this would contradict well-established historical events, Another mechanism must therefore be operating. The simplest explanation is the need to assemble large quantities of coinage prior to major campaigns. This would have temporarily distorted the proportion of different types of coins in circulation at forts.

This must also be a possibility for the excess Hadrianic coins.

The coin evidence can therefore be said to be strong enough to support the theory of a period of abandonment in the first half of the second century, but not strong enough to assess accurately the length of this period.

Phase Four

Re-occupation of Newstead is traditionally linked with Antonine's reconquest of Southern Scotland about 140 AD.

However, coin evidence leaves open the possibility that it may have occurred a few years earlier. This minor revision of the chronology would be unimportant in itself, were it not for the find that it alters the function of the site from a support centre to the rear of the Antonine Wall to an outpost fort North of Hadrian's Wall.

The character of this phase ie the first fort after the hiatus, has to be substantially revised from the description provided by Richmond. Ironically, Curle's assessment was far closer to the truth.

The Reducing Wall

Richmond believed the stone fort, including the Reducing Wall (or Dividing Wall, as he called it) to have been constructed at the outset of the re-occupation. He envisaged two units; one, legionary; one, cavalry; occupying billets in the two unequal halves of the fort.

However, the stone fort, with its repositioned North and South gateways (East of the original location), can now be seen to have been a secondary development of the second century occupation.

The initial arrangement is associated with an enlarged South Annexe defensive enclosure, pierced by a single North-South road. This road leads to the old Domitianic fort’s South Gate, proving without doubt that a refurbishment occurred predating the curtain wall of the fort. A dump of earth strengthening the rampart, (which Richmond associated with the construction of the curtain wall) probably represents this post-hiatus refurbishment of the rampart.

This re-use of a much-decayed earth rampart fits rather well with the character of the earliest second century ditches. These were of very unusual shape, and were thought by Richmond to have contained an entanglement of branches. This is hardly an arrangement which would be expected, in conjunction with the masonry curtain wall - which was quite strong enough on its own, and was built largely for durability.

Ditches filled with an entanglement of interlocked branches can have been only a short-term expedient, and would seem most appropriate in the context of re-occupation of a broken-down rampart.

Precarious phase

The fort building phases associated with this defensive arrangement remain completely unknown. Failure by Curle and Richmond to identify this important phase strongly suggests a post hole or sill beam construction, without stone footings. This is entirely consistent with what would be expected at the outset of a major period of reconquest.

However, some extramural settlement probably existed even at this precarious phase. Much of the North-South road through the South Annexe was evidently built to an unusually great width (as shown by geophysics) and may have served as a market place (as did the more Easterly North-South road in the next phase).

Some of the buildings excavated in area RGT 33 (see Appendix 6, top) also probably date to this phase, but it is difficult to gauge how soon after the re-occupation of the site reconstruction took off

In contrast, the East Annexe, though enclosing a much larger area from this phase onwards, probably saw no development until phase 5.

an outpost....in strength?

As with the initial first century occupation, phase 4 was probably quite brief. If this was an outpost fort (and I have my doubts) it was certainly an outpost in strength. The unreduced fort had capacity for at least 1,500 troops and occupation in the South Annexe at least had begun, suggesting a significant civilian population.

Phase Five
Following a (short) period as a frontier fort in the true sense, Newstead found itself to the rear of Antonine's Wall, and its fighting force was reduced accordingly. Phase 5 saw a major reorganisation of the fort, which was rebuilt, with durability a key consideration.

Defences in this phase consisted of a two metre thick masonry curtain wall, inserted in front of the existing rampart.

This wall and rampart enclosed a "reduced" fort (of about 4 hectares), leaving the Western fort area outside its circuit. That area became an industrial annex (see appendix 3, bottom viz RGT 125/126 and the 1987 trench, which revealed slag deposits), involved in iron and lead working, the processing of various organic products and probably pottery production.

"out of reach of flaming missiles”

Three relatively small, widely-spaced ditches enclosed both the fort and the Industrial Annex.

The “Reducing Wall” was almost identical to the rest of the Curtain Wall, but was backed by no rampart and fronted by no defensive ditches.

In addition, there does not seem to have been the normal intervallum road down the Western wide of the phase 5 fort. These differences can be explained by the continued maintenance of the existing ramparts round the Industrial Annexe.

An enemy was never likely to threaten the fort interior from this side, so there was no need for ditches to keep opponents at a distance or an intervallum to keep vulnerable thatched roofs out of reach of flaming missiles.

"total re-organisation of space"

The interior of the Reduced Fort represented a total re-organisation of space, with re-positioned gateways (and by implication, the principal streets) and a revised ground plan facing east, probably for the first time. (The fork in the east road leading out of the fort was already in existence, so the main approach, via Dere Street to a bridge East of the fort had probably existed from the preceding phase – phase 4 - at the latest).

Most of the buildings noted by Curle within the 'Reduced Fort" probably originated in this phase - phase 5.

"maximum garrison of 1000"

Twelve barracks in the Eastern part of the fort suggest a minimum garrison of about one thousand.

The central range of buildings includes the principia (the Headquarters building), flanked by two granaries and the commander’s house in the extreme South of the range.

The remains of at least one monumental building existed in the Northern part of the range. To the West were three buildings, only one of which can be identified by function. In Appendix 3, bottom, Building XVII seems to have been built as a fabrica, a workshop, and comprised initially a L shaped building with a covered courtyard open to the adjacent road.

the annexes

In the West annex, the bath house (as excavated by Curle) was probably started in this phase, although this assumption rests on the fact that there was a lot of masonry building throughout the complex at this time

In the Agricolan and Domitianic periods this West annexe had probably been intended to enclose a market place and the larger part of the civilian population.

By Phase 5 however, the orientation of the fort and the main direction of approach had changed, leaving the area as a relative backwater. It is therefore probable that the area suffered its first reduction in size at this time. Closely linked to this change was probably a substantial increase in the population of the South and East Annexes.

The South Annexe saw the construction of a completely new main approach to the fort's South Gate This was some 70 metres East of the existing North-South road, which was not abandoned but was retained as part of an extensive network of side streets. The new road was constructed to a maximum width of 12 metres but narrowed to just 4 metres, where it entered the fort. As with an abnormally wide section of North – South road in the preceding phase - phase 4 - this probably represents provision for market stalls.

"lining the roads"

Lining the roads were strip buildings ranging in size from about 25 square metres to 100 square metres. From the density of excavated examples and known street frontage it can be estimated that the South Annexe comprised 80 to 100 buildings at its height during this phase.

"East Annexe....roadside development"

The East Annexe saw a minor shift in the location of the Eastern road out of the fort. The repositioning was only by about 10-15 metres, the result of the reconstruction of the new masonry East Gate, but it has implications for the level of roadside development in the preceding phase - phase 4.

This is unlikely to have been great, otherwise the existing roadline would have been retained.

During phase 5, however, the first of a series of buildings were constructed 70 metres East of the gateway (see appendix 5). Elsewhere in the East Annexe, further out from the fort, roadside development does not seem to have occurred.

"settlement more limited but substantial"

Settlement was therefore more limited than in the South Annexe, but may still be quite substantial, perhaps stretching for 80 metres, on either side of the road.

Phase Six
Phase Six saw major changes to the arrangements of the fort itself, which possibly related to military difficulties towards the end of the Antonine Wall’s occupation.

Most obvious was the demolition of the Reducing Wall and the extension of the Curtain Wall to include what had been the Industrial Annexe. (The new West Gate and new length of Curtain Wall were carried out to slightly lower specifications than the original construction, but were by no means jerry-built)

"greatly increased garrison"

The Western fort area was now given over to long narrow buildings, with timber sill superstructures, resting on (very) rough stone footings. This would appear to represent the extension of barrack facilities to house a greatly increased garrison.

'front line fighting base"

The very different character of these new buildings suggests a relatively hurried construction.

The cessation of industrial activity in the Western part of the fort was matched by a growth of industrial activity elsewhere.

Therefore, Newstead must have undergone a major change of role, from industrial and distribution centre to front-line fighting base.

“extra-mural population"

Decline in the size of the extra-mural population had very probably begun in this phase, but evidence for the abandonment of large areas dates to the next phase – phase 7.

In the Western Annexe the existing trend towards a reduced defended area was continuing with the construction of a rampart round the bathhouse itself, and the cutting of a ditch between the bathhouse and the former mansio site (see appendix 4).

Phase Seven
Phase Seven probably represents occupation from the final abandonment of the Antonine Wall in the early 160’s AD until the withdrawal of the army. The character of occupation saw drastic changes in almost every area. Large parts of the Western fort area saw the demolition of the existing buildings and their replacement with tiny structures (3-4 metres by 4-5 metres), standing in open space.

The principia (HQ building) was also demolished and replaced by the so-called exercise hall, which overlay the via decumana (street from North Gate to South Gate: see Appendix 3). If the two granaries remained standing, this would have created a large courtyard open to the west. Possibly the ‘hall’ functioned as the HQ building in the late phase of occupation, with yet another re-orientation of direction in which the complex faced.

The barracks in the Eastern part of the fort were presumably still standing, and in fact may have attained their recorded “chalet” form only in this period. If this were the case, it would be tempting to suggest that each chalet represented a soldier and his family, and that a massive reduction in the military strength of the site had occurred.

In fact, however, the coin lists suggest that coin use remained relatively strong. Outside the fort depopulation certainly was dramatic, especially in the South Annexe which, but for RGT 141/15 (see Appendix 6, bottom and top), may be completely deserted.

Field ditches encroaching into the former settlement area prove that the defences and many of the buildings had been abandoned.

The change probably reflects not only the reduced size of the garrison but also the changed security situation.

Decline and Fall

Civilian settlement may survive elsewhere, perhaps even moved inside the fort defences!

The last coin dates to around 180 AD.

An abandonment at this date would be consistent with the latest pottery known from the site.

APPENDICES

Appendix 1

Appendix 2

Appendix 3

Appendix 4

Appendix 5

Appendix 6