CUMBRIA

(1) Birdoswald (Banna) (NY 615 633): two areas were excavated in the eastern extramural settlement outside the Hadrian's Wall fort, the extent of which is known from previous geophysical survey.Footnote ²⁴ Area A measured 25 × 13 m and was positioned to locate and contextualise a building previously exposed briefly by Ian Richmond and interpreted as a signal tower.Footnote ²⁵ This building survived to 2 m in height, and steps led down to a semi-basement. To the north were further structures, part of a large complex identified in geophysical survey. A broad primary wall to a building east of the excavated area was abutted by an apsidal wall, the area between the walls being surfaced by regular flagstones. A stone bench-end was found laid on these flags. The wall of the apsidal structure was cut by a free-standing square stone building, floored with large slabs (fig. 4). This building was associated with a bifurcated water-main; one branch broke through the walls of the apsidal structure, the other ran to the west of and parallel to the square building. The evidence comprised iron pipe-connecting rings (fig. 5) set upright in narrow trenches. A thick and widespread deposit of soot and charcoal, together with box-flue tile fragments and spacer bobbins, suggest the very close proximity of the fort bath-house.

Area B, which measured 14 × 11 m, was situated 35 m east of the principal east gate of the fort. Two stone-built strip-buildings facing onto the Military Way were excavated. One respected the width of the double gate portal, the second respected the width of the gate after its reduction to a single carriageway in the third century. They were built and floored in part with massive stone slabs. A rubble ridge, long recognised as an earthwork, and previously thought to represent spoil from the nineteenth-century excavation of the east wall, was proved to represent the northern walls of the strip buildings. Finds included two domestic portable altars (fig. 6).

FIG. 4. Birdoswald: Stone floor of secondary building in Area A. (© Historic England Archaeological Projects Team)

FIG. 5. Birdoswald: Iron water-pipe connectors, Area A. (© Historic England Archaeological Projects Team)

FIG. 6. Birdoswald: Portable altar in situ in Area B. The inscription was not legible. (© Historic England Archaeological Projects Team)

A third trench, Area C, examined a 20 m stretch of the 4 m wide berm on the north side of Hadrian's Wall in order to establish whether there was evidence here for the obstacle pits found elsewhere on the Wall.Footnote ²⁶ There was no evidence of such features, although drain holes were found in the base of the Wall.Footnote ²⁷ Work will continue in these areas in 2023.

(2) Carlisle Cricket Club, Cavendish Terrace (NY 39971 56717): a community excavation opened three trenches with a combined area of 560 m².Footnote ²⁸ The presence of a large Roman bath-house on the site had been identified in 2017. The discovery was followed the same year by additional investigatory excavation.Footnote ²⁹ These early phases of work successfully engaged volunteer diggers from the community. The discoveries caught the public imagination leading to a National Lottery Heritage Fund bid to fund a community project on a larger scale. The site of the Carlisle Cricket Club occupies what was once an important and well-defended location within the frontier system. It lies on the north bank of the river Eden close to a bridge that connected two Roman forts (fig. 7). The fort on the south side of the river, Luguvalium, was found to have been established under the Flavians. Roman control of this strategic river crossing was consolidated with the establishment of Stanwix fort (Uxelodunum) when Hadrian's Wall was built. This small piece of land, measuring just a few hectares, was protected on its northern side by Hadrian's Wall and on all other sides by a sharp meander in the river Eden. It is likely that this area was occupied by an extra-mural settlement associated with Stanwix fort, close to where Hadrian's Wall crossed the river Eden at its confluence with the river Caldew.

FIG. 7. Roman Carlisle, showing the Cricket Club excavation site. (© Wardell Armstrong LLP)

The remains of three Roman bath-houses that all probably faced onto the main north–south road over the river Eden have been uncovered. First came a Hadrianic bath-house of modest size and similar in floor plan to others associated with forts on Hadrian's Wall. Only exiguous parts of the Hadrianic bath-house and its heated floors survived in areas left between later stone robbing events. Little of the foundations and floors remained. Instead, the evidence for this building was constituted by robber trenches that were large and pit-like close to the modern ground surface and then became linear in form, where they followed the former foundation cuts of the Hadrianic Walls. Only in one area was the wall of the Hadrianic bath-house found to survive, measuring 0.6 m thick. There is evidence to suggest that much of the stone robbing took place in the twelfth century, when Carlisle's Cathedral, Keep and Town Walls were being constructed. Some of the robbing may have taken place during the early third century, when a much larger bath-house complex was constructed under the Severans.

The Severan bath-house was colossal in scale. Its northern, eastern and western limits were not reached within the excavation trenches. Only on its south side was the external wall found and a contemporary east–west aligned street uncovered. The building had walls 1.05 m thick. The stone used in the bath-house buildings is similar to sandstone outcropping locally along River Eden and the River Gelt, where known Roman quarries have been located. Huge pieces of opus signinum and mortared superstructure were found to have fallen through the floors and between the hypocaust pilae stacks. Large quantities of hollow box-flue tile, combed on the exterior, were recovered. Their internal surfaces were often blackened with soot. Several T-shaped iron clamps, used to affix the box tiles to the walls, were also recovered. No roof tiles were found and the evidence of ceramic, nozzled tubes and ‘armchair voussoirs’ suggests the presence of a vaulted roof. A key difference between the stacks of square hypocaust tiles recovered from the Hadrianic and Severan bath-houses is that the Severan hypocaust tiles frequently bore stamps. Twenty-nine tiles with an identical ligatured ‘IMP’ stamp, standing for imperator (emperor) have been recovered (fig. 8). The same stamp has been found of roof tiles from other excavation sites in Carlisle, at Blackfriar's Street and at Annetwell Street. On the latter site there were indications that the tiles were produced at the beginning of the third century, specifically for the construction of the stone fort and its ancillary buildingsFootnote ³⁰. The evidence from the Cricket Club site now demonstrates a larger programme of building.

FIG. 8. Carlisle Cricket Club: Stamped hypocaust tile. (© Wardell Armstrong LLP)

Immediately south of the bath-house were the remains of a Roman street, undisturbed by later stone-robbing events. A 14.2 m stretch of the street was excavated. It displayed a cambered (agger) surface and measured 5 m in width. Levels taken indicated that it had a gradient falling towards the west. Sections of well-preserved, deep, stone-lined drains were recorded. They indicate that the water source feeding the bath-house came from the east, with the waste water exiting the building to its south and west. One of the drains exiting the bath-house on its south side appears to have turned to the west to respect the line of the street. This street may have linked parts of the extra-mural settlement to Stanwix fort and Milecastle 66. Fourth-century pottery sherds retrieved from the street's surface suggest that it was still being used late into the Roman period.

Evidence of a fourth-century rebuild of the bath-house was also recorded. Cutting through the stone wall of the Severan bath-house were the partial remains of a curved wall measuring 2.50 m in length, 0.95 m in height and 0.55 m in width. This was constructed of red sandstone pieces. Reddened, heat-affected ground suggested the location of a stoke-hole nearby. Hypocaust stacks relating to this phase of the bath-house were found built on top of layers of soot, which had built up during the third century. Adjacent to this, on its western side, was a crude wall, possibly the remains of an arch constructed of broken up but sizeable pieces of reused opus signinum.

The finds suggest a date range beginning in the reign of Hadrian and extending into the mid-fourth century and beyond. A total of 410 Roman coins, often in excellent condition, have been recovered from the excavations. The earliest coins are Trajanic (three in total). There was a slight increase in coin loss under the Severans, with most coins dating between the mid-third and mid-fourth centuries. Louise Hird has assessed the pottery assemblage, which is made up largely of domestic food preparation and storage vessels. There is relatively little in the way of table wares, apart from some samian vessels. Unusual vessels include a cheese press and a tazza. The date of the material is second and third century. There is very little, if any, coarse ware of the late first/early second century and, in contrast to the coin assemblage, the fourth century is represented by only a few sherds. XRF analysis of the glass finds by Rebecca Scott demonstrates that most coloured items contain antimony and manganese decolourisers, which indicates recycling and mixing together of glass from Egypt and Syro-Palestine.

Epigraphic evidence from the site also sheds light on the population using the bath-house. For a long time, the main evidence that the Ala Petriana was stationed at Stanwix fort was limited to a reference in the Notitia Dignitatum Occidentis (XL.18) and an inscription found while digging the foundations of Carlisle Journal Office in 1860 (RIB 957), which predated the end of the reign of the emperor Trajan. Further evidence comes in the form of stamped lead seals. The first, found in 1986, confirmed the presence of this unit on the site of Stanwix fort (RIB 2411.84). Further evidence for the presence of the Ala Petriana now comes in the form of a second stamped lead seal found by the Cricket Club excavation's official metal detectorist (fig. 9). Fragments of four stone inscriptions have also been found at the site and examined by Roger Tomlin.Footnote ³¹

FIG. 9. Carlisle Cricket Club: Lead seal. (© Wardell Armstrong LLP)

NORTHUMBERLAND

(1) Vindolanda, the south-western quadrant of the last stone fort (NY 769 663):Footnote ³² in the spring of 2021, after an enforced break due to the pandemic for 2020, work continued with the research project entitled: Understanding Communities and Identities. The area of research was located within the south-western quadrant of the last stone fort at Vindolanda (fig. 10). The work established the extent and density of the fourth-century occupation and revealed a cavalry barrack and a series of associated buildings, which had been partially preserved below two–three levels of stratified sub-Roman and post-Roman occupation. A roughly cobbled yard, like that encountered behind the cavalry barracks in the south-eastern quadrant, extended from the barrack buildings lining the side of the via decumana in the east to the intervallum road to the south and west. The cavalry barrack produced a variety of military equipment, which included several fine examples of lance heads and other associated cavalry equipment. The primary floor surfaces within the barracks were a mixture of cobble stones packed with clay and flagstone set in clay foundations. Although several rooms were extensively damaged by later sub-Roman and post-Roman constructions, an examination of the internal dimensions and appointments of the rooms provided clear evidence for the separation of the living accommodation for soldiers and their horses. The domestic spaces, situated on the western side of the barracks, were furnished with clay ovens. Rooms that functioned as stables faced onto the via decumana, and were finished with rougher cobbled floor surfaces, some of which had sunken into central pits.

FIG. 10. The south-western quadrant excavations at Vindolanda in August 2021. (© The Vindolanda Trust)

A squared building, to the west of the main barracks and north of the cobbled yard, may be interpretated as having been a temple or shrine. A remarkably well-preserved stone carving, SF22528, depicting a naked cavalryman figure and a horse or mule, was recovered from its floor. This carving has been interpreted as a possible representation of the Dioscuri. The Dioscuri were the heavenly twins, Castor and Pollux, and can be associated with the patronage of ‘military endeavours’. The stone carving was discovered facing upwards on the floor surface and appears to have been left undisturbed by later remodelling within the vicinity of the building (fig. 11). A large quantity of Huntcliff and Crambeck ware pottery, including examples of later fourth-century Crambeck painted parchment ware vessels, were recovered from many of the contexts associated with the fourth-century buildings excavated in 2021. The presence of this late pottery is significant as it suggests that the Roman military supply and therefore occupation extended to the very end of the fourth century and possibly into the early fifth century without any obvious interruption.

FIG. 11. The Vindolanda carved figure with mule. (© The Vindolanda Trust)

Excavations continued below the foundations of the fourth-century buildings and revealed the early third-century phasing within the quadrant. This included the remains of a potential schola or officers’ mess, occupied between c. a.d. 212/213 and c. 280. The schola was a long and thin building, standing six courses high on its western side. To construct it, the Roman masons had used the same freshly quarried stone, and building style, as those employed to construct the central range of the fort. This building included a series of rooms that had clearly been used as social spaces with heavily worn flagged floors. Other rooms were focused on cooking and food preparation. A similar configuration, with the emphasis on very large, almost industrial food ovens, was a feature of the period IV schola (c. a.d. 105–120) located by the 2001–02 excavations at the site.

The excavation of the schola produced many fine examples of samian ware, a number of silver coins from the social spaces, along with other examples of portable martial culture that could be regarded as having represented those with a higher wealth and status than items recovered from the excavation of typical third-century barracks within the fort. The 2022 excavations will continue within the same area, completing the plan of the third-century quadrant before dropping down into the Severan, Antonine and Hadrianic levels and buildings below.

TYNE AND WEAR

(1) Newcastle-upon-Tyne, West Road, Fenham (NZ 2139 64889):Footnote ³³ a c. 4.5 m long section of sandstone wall and one much smaller section, both of which are interpreted as part of Hadrian's Wall, were discovered during an archaeological watching brief.Footnote ³⁴ While West Road is known to follow the line of Hadrian's Wall, the two sections of the Wall were previously undocumented.

The site is located to the northwest of Benwell Roman Fort (Condercum). The fort, garrisoned during the reign of Hadrian by a cavalry regiment of 500 troops, appears to have been constructed in tandem with the eastern section of Hadrian's Wall. The vallum at Benwell changes alignment to skirt around the southern limits of Condercum fort, suggesting that the vallum is contemporary with or post-dates the fort's construction.

The archaeological works comprised the monitoring and recording of groundworks during the excavation of two conjoining trenches running for a length of c. 247 m with a width of 0.6 m wide, effectively forming one continuous trench. The trench was excavated to a depth of between 1.2 m and 1.6 m. At the junction with Newminster Road, approximately 155 m from the south-eastern end of the trench, a c. 4.5 m length of sandstone wall was revealed which is interpreted as representing part of the southern face of Hadrian's Wall (fig. 12). The stonework was aligned WNW, at an oblique orientation to the direction of the trench, and had clearly been truncated at its south-eastern end. It is not known how far the stonework continues in a south-westerly direction as the trench was terminated and diverted to the south-west in order to preserve the section of wall. The wall was constructed using large rectangular and square dressed sandstone blocks, with the inclusion of a single large granite rectangular block which had probably been used as a repair. Two courses of stonework were exposed, bonded together with light compacted clay mixed with small, chipped sandstone fragments. The larger blocks were the facing stones of the southern face of the structure, overlying smaller irregular sandstone blocks representing the stepped foundations which were integrated within it, similarly bonded with the light compact clay. The foundations projected c. 0.38 m from the edge of the facing stones. The largest sandstone block measured 0.36 m wide, 0.25 m high and over 0.22 m deep.

FIG. 12. Sandstone part of the southern face of Hadrian's Wall at West Road, Fenham. (© Archaeological Research Services Ltd)

At the extreme northern end of the trench was another section of sandstone wall, which had been truncated sometime during the twentieth century by the insertion of a metal water-pipe. A 0.6 m wide section of the northern face of the wall was visible and was constructed using faced, rectangular sandstone blocks. The wall, which is interpreted as part of the northern face of Hadrian's Wall, continued beyond the limits of the excavation and thus it was unclear how many courses had survived in this location. What is considered to be the rubble core of the structure was encountered to the south of the facing stones, but the limit of the excavation did not extend sufficiently in order to locate the southern face of the structure in this location. The core consisted of small- to medium-sized fragments of sandstone bedded within orange/grey sandy clay.

The largest stone noted in the southern face of Hadrian's Wall investigated at West Road at the junction with Newminster Road measures 0.36 m wide, 0.25 m wide and over 0.22 m deep, which, in comparison with the stones used elsewhere to build the Wall, could be considered as ‘large’, with the majority of the remaining stones being ‘quite large’. This section of the Wall appears to have had a clay and rubble, rather than a mortar, core although not enough of the core was exposed during the watching-brief to determine its exact composition. It has been considered that the use of smaller-sized stones was preferred at a later date to build Hadrian's Wall as they would have involved less effort in constructing the Wall. If this is the case, the larger sizes of the stones used as facing stones within the sections of exposed structure at West Road may indicate they were part of the earliest phases of the Wall's construction.