Welcome Guest [Log In] [Register]

From Aug 2006 - Nov 2013 WeDig provided a live forum for diggers & fans of Vindolanda. It has now been mothballed and will be maintained as a live archive.

Here you will find preserved 7 years of conversation, photos, & knowledge about a site many people love. Vindolanda gets under the skin. (Figuratively and literally as a volunteer excavator!) It's a place you remember, filled with people you remember!

Thanks for 7 great years!

Viewing Single Post From: Excavations of 1967-1969
Member Avatar
[ *  * ]
The Third-Century Fort: Headquarters Building

When the Headquarters Building was excavated (AA4 xiii 218f), no attempt was made to determine whether the third-century structure, which underlies the Diocletianic Headquarters, possessed a sunken strongroom. The western half of the third century Sacellum was available for excavation, without disturbance of the later remains conserved by the Ministry, and an attempt was made, in June 1969, to answer the question. In the event, no structural evidence was found. The water-table within the fort is several feet higher than in the vicus, and that level did not decline, in spite of subsequent spells of fine weather. It was established, however, that there had been a substantial hole in the position where a sunken strongroom could be anticipated, which had later been filled in with fragments of whin-stone, clay and broken tile. At a depth of three feet below the original third century floor-level, the excavation had to be abandoned, because of the water-table. It may be possible to return to this area when the drainage of the field has been improved, but in the meanwhile one can only presume that a sunken strongroom existed.

The first- and second-century forts

In both 1967 and 1968 sections were cut, fifteen yards apart, through a feature to the south of the mansio (site IX below), which both on the ground and on an aerial photograph had the appearance of rampart and ditches. For a variety of reasons, but chiefly because of the depth of the remains and the high water-table, neither section was entirely satisfactory (see Plan B). Some positive evidence did emerge, however, and it will be described in the reverse order to which it was found, that is, starting with the earliest material. In the account which follows, reference is made to the relevant parts of the sections on Plan B by the use of the date (1967 or 1968), and an arabic numeral.

Spoiler: click to toggle
Plan B

On neither occasion was subsoil reached, although both sections were taken down to a depth of 8 feet. At that depth, however, there were remains of timber-work. 1967, 2, was a series of oak posts, some 4 inches square and two feet apart, with a horizontal plank nailed to their southern side. The high water-table and torrential rain prevented a proper examination of this feature, but a deposit of Flavian pottery, glass, leather and scraps of wood around the oak posts gave a firm date. It was hoped that the same feature might be found in the 1968 section, but there the structural evidence proved to be very different, although the rubbish deposit indicated a similar Flavian date. 1968, 10, was a float of interwoven branches (mostly hazel and birch), with upright wattle-hurdles, two feet high and two feet apart, running in the same east-west direction as the timber posts in the 1967 section. The space between the hurdles had been filled with boulders and rubbish. The presence of rampart material above this timber-work in both sections suggests that it had been connected with the construction of a fort rampart. The technique of laying a wooden float in a swampy area to support a structure above it is well-known, being used by George Stephenson in the early days of railway construction (so Mr. Harold Bowes kindly informed me), and by the builders of large houses in the seventeenth century (the Gordon home at Gordonstoun was built in this manner).

Spoiler: click to toggle
Plate XIV, 2

The 1967 section showed that a clay rampart-base had been laid upon the wooden structure, and the 1968 section, in a wet area where the timber float had been necessary, had a heavier stone rampart-base, with clay rampart above it. There were traces of at least two fort ditches to the south of the rampart, and the inner (northern) ditch had been blocked at some stage with rampart material, when a new rampart was constructed, slightly further south than its predecessor. The upper part of this new rampart was of turf, and 1968, 7 showed the post-holes of a timber palisade of more than one period. In neither section, however, were there clear indications of the rampart edges, nor could one be certain of the date of the later rampart, although it must be associated with late second- and early third-century pottery in the outer ditch.

The ditches had been sealed efficiently at some stage. The inner ditch had gone out of use when the second rampart was laid, whilst the outer ditch remained open into the first two decades, or so, of the third century. Above the solid whin-stone boulder filling, a small and irregular building wall was found on the edge of the section (1967, 1), but the building was not investigated, whilst to the south of the 1968 section another crude building had been laid above the boulder fill, but this time huge undressed blocks of sandstone had been used, similar to those found on the period IV structure on site IX (see below). Its position is not shown on the section, and the prevailing swamp allowed only a cursory investigation.

To the north of the ramparts there were several features, not all of which could be satisfactorily explained in the limited area examined. Both sections revealed a rough track at a high level on the northern shoulder of the rampart, presumably dating to the third and fourth centuries, and both produced the southern walls of vicus buildings. In the 1967 section, a trial hole to the north of this wall showed that it overlay a north-south wall of an earlier period, and the improved stonework of this earlier wall and its depth suggested that it was associated with the second period of the rampart, that is with a late-second-century fort. A trial hole in the equivalent position in the 1968 section showed five occupation layers, but no trace of walling. Three of the layers are probably attributable to vicus periods, and the presence of the upper layer (1968, 1), consisting of three worn flag stones in situ, with an unbroken small grey vase at their side, only twelve inches below the turf, reminded us of the difficulties of finding adequate traces of the last period of occupation.

Two further features in the 1968 section call for comment. The stone-lined drain (1968, 6) on the southern edge of the later rampart was certainly Roman, since a Wall period III cooking-pot lay in pieces upon its floor, and the remains of a small clay oven (1968, 9) immediately to the south of the vicus building's southern wall, probably belonged to a vicus period.

The excavators were conscious at the time that the sections were unsatisfactory, but they have been illustrated and described here as a guide to future sections. The high water table and the presence of the huge blocking stones in the ditches demanded much wider trenches than the five feet that were cut, and the necessity of carrying such trenches down to c. ten feet made excavation physically too demanding for many volunteers.

It is clear from the work of Eric Birley in 1930, on the Flavian ditches in the north of the field, and from the 1969 work near the Well (site XII, below), which revealed the western Antonine fort rampart, that it will be comparatively simple to plan the northern and western defences of the earlier forts. At this stage it is only possible to demonstrate that the southern line of ramparts and ditches has been located. Evidence from the section in room VI of the mansio (see below) has warned us that we must be prepared to find two pre-Hadrianic forts on the site, but there is nothing to suggest that more than one second-century fort overlies the first-century remains.

VICUS--outside the Diocletianic west gate

In 1967, vicus buildings outside the Diocletianic west gate, clearly visible in air photographs, were located on the ground by excavation. A trench dug at right-angles to the west fort wall, from the south of the southern guard-chamber, produced evidence for a series of fort ditches, belonging, presumably, to the Antonine, third-century and fourth century forts, of which only the inner (eastern) ditch had remained open. The outer ditches at least two had been carefully filled at some stage, probably at the beginning of the third century, to take vicus buildings above them. The wetness of the season, and the large nature of the filling (often sandstone flags five feet square and five inches thick), prevented a full examination of these features. But the sequence of structures from the fort-wall was as follows: the berm was 12 feet 6 inches wide, and the inner fort ditch 18 feet 6 inches wide. Three feet to the west of this there was a narrow (10 feet) roadway, lying above heavy stone packing which sealed a second ditch. Immediately to the west of this lay the eastern wall of the first vicus building (site IV), whose floor was made up of exceptionally heavy flagstones, which sealed a third ditch. No attempt was made to examine this vicus building, and the position of the neighbouring sites V, VI, and VII was merely noted for future reference. A trench across the roadway which runs from the fort's west gate through the vicus in the WNW direction revealed that it was paved with heavy flagstones, possessed side conduits, and was 18 feet wide.

The vicus buildings in this area outside the fort west gate appear to be similar in design to those outside the south gate at Housesteads, and they are well preserved. The aerial photograph indicated that there was a line of structures front ing a side road which runs the length of the west fort wall perhaps some twelve houses in all and rows of similar structures on either side of the main road through the vicus to its junction with the Stanegate, some 200 yards to the WNW. There are further rows of houses to the south and west of the fort field. It is hoped to publish a provisional plan of all these features in the next report.

VICUS--west end

There had been serious flooding at the west end of the camp field for the past five years, and the principal source appeared to be the Roman well, examined in 1914, and marked with an upright plaque. At Mr. Thomas Harding's request, the Well was reopened, and its spring channelled into a modern 6 inch pipe drain. In the course of laying this drain, six features were examined.

The Well, Site XII (Note: Plan C). Examined but not recorded in 1914 (although the plaque bears the inscription "nothing found"), the Well was stone-built, circular, and with the top three surviving courses on the western side overlapping each other, to narrow the mouth. (Plate VII, 2.) The Well was three feet in diameter below these overlapping stones, and it presumably narrowed to 2 feet 6 inches. No attempt was made to clear it out again. Three flags, set upright in clay and rubble, stood 6 inches to 12 inches from the Well on the Western side, and beyond them, undisturbed by the 1914 excavation, lay, inverted, the broken Well topóa circular carved stone basin 3 feet 9 inches wide. This once capped the Well, and water would be drawn from the basin. The advantage of such a basin top would be to prevent larger items of rubbish being dropped into the water supply (see the note on wood samples from the Well in the HQB, p. 145).

Spoiler: click to toggle
Plan C

Spoiler: click to toggle
Plate VII, 2

The Water-tanks, Site XIII. 12 feet 6 inches WSW of the Well, there lay the northern end of double water-tanks, into which the Well overflow had once been diverted. The northern tank had been largely destroyed, but it had once been 10 feet 6 inches long and 3 feet 3 inches wide, with a flagged floor 2 feet 9 inches below the top of the sides. The southern tank was smalleró9 feet 6 inches by 3 feet 3 inchesóbut it had been deeper, 3 feet 6 inches. A fragment of a Roman milestone had been reused to support one of the side slabs. (Plate VIII, 2.) There were outlet holes at top and bottom of the southern slab, and a grooved stone channel ran from the bottom outlet in the direction of SSE. The tanks produced a few scraps of pottery and a cornelian intaglio of Jupiter Serapis (?) (see Plate XVI, 1 and discussion, p. 146f.). In a depression below the flagged floor of the northern tank there was a deposit of second-century pottery and the sole of a right-footed Roman boot, size 9 1/2.

Spoiler: click to toggle
Plate VIII, 1

Spoiler: click to toggle
Plate VIII, 2

Spoiler: click to toggle
Plate XVI

Antonine Rampart base. The trench for the pipe drain to the south of the outlet hole cut through a rampart base, 3 feet below turf level. Approximately 15 feet wide, with well laid stone kerbs and an interior of cobbles and gravel, the rampart ran parallel to the water-tanks, 12 feet 6 inches away. This presumably represents the western rampart of the Antonine fort, and indicates that the fort was of irregular shape.

Roman drain. The laying of the modern pipe was greatly helped by the discovery of a stone-lined Roman drain, with flagged covers, running beneath the rampart base to a point 10 feet from the western field wall (where a modern 6 inch pipe takes the water into a cattle-trough in the next field). The drain runs in a straight line for 18 yards, and was probably constructed at the same time as the rampart.

Post-Roman causeway. 16 feet east of the Well (site XII), just below the level of the turf, there is a stone built causeway, three feet wide, running approximately N-S across the swampy area produced by the well spring. It is modern, constructed before 1900.

Flagged area. Due west of the water-tanks, and immediately west of the rampart base, there is a broad area, at least thirty feet wide, E-W, covered with flags. This was located during drainage work, but was not investigated further. But such evidence takes the extent of the vicus up to within 10 feet of the field wall at the west end of the station. An aerial photograph suggests that there are perhaps vicus buildings beyond the field wall and, indeed, to the north of the Stanegate as well.

VICUS--Sites XV and XVI

Two small buildings were located immediately to the north-west of site IX (the mansio), but they were not examined in detail. One significant piece of evidence did emerge, however, for below the highest floor level of site XV, sealed by 12 inches of rubble and clay packing lay a coin (no. 23), a little worn URBS ROMA issue of A.D. 330-335. Until Wall period IV pottery is also found on the site, it would be unwise to claim that this represents further evidence of Theodosian activity in the vicus, but it does at least represent vicus reconstruction post-A.D. 330. It will be some time before this site is fully excavated.


The site occupies a prominent position in the centre of the vicus, almost directly opposite the fort bath-house, with the main road through the settlement running between them. An aerial photograph gave the first indication of the structure, and rooms I, II, and III were examined in 1968, before being back-filled. In August and September 1969 the remainder of the structure, excepting parts of the courtyard and a few baulks, was examined and left open.

Summary: In its final form, dating from the early third century, the building was a fifteen-roomed courtyard house, fronting upon the main vicus road (see plans D and E). There were four periods of occupation in parts of the building, together with some minor adjustments to the plan. The history of the structure appears to have been as follows:

Spoiler: click to toggle
Plan D

Spoiler: click to toggle
Plan E

Period 1a, c. A.D. 160. A three-roomed bath-suite in stone (rooms I, II and III), perhaps attached to a timber house. The structure was aligned with the south rampart of the second century (? Antonine) fort, some twenty-five yards to the south, and perhaps should be associated with the praetorium of that fort.

Period 1b, A.D. 160/197. Two stone-built rooms (IV and V) were added to the structure at the northern ends of the building, and the bath-suite was thus enlarged. At this stage, rooms III and IV had hypocausts, and room III also possessed wall-flues.

Period II, c. A.D. 212. A further ten rooms were added to the building, whose northern front was now aligned with the main road through the vicus, opposite the new fort bath house. These rooms were grouped around a courtyard whose greatest measurements were 75 feet x 24 feet. Two of these rooms had hypocausts (VI and VII), and the large room VI was perhaps a dining-room. Room XIV was certainly a latrine. The ground plan of the building, and the regular masonry, suggest that the structure did not belong primarily to the vicus, but was rather a military building, designed to house travelling officials at a convenient point on the Stane gate, almost mid-way between the major Roman sites at Carlisle and Corbridge. It should thus be termed a mansio.

Period III, C. A.D. 300. Complete reconstruction of the whole building, after a fire had certainly consumed the western wing. The remains had been levelled with building debris and clean grey clay before the reconstruction, and the new ground plan contained important modifications. Room VI, the former dining-room, now lost its hypocaust and was converted into a kitchen, with a large stone-built oven in its north-western corner. The adjoining room VII, now deprived of its heat, received a new clay floor above the old concrete one. Rooms XII and XIII were run together by the elimination of their cross-wall, and room XIII's partition wall with the courtyard was abolished, to create a broad doorway. The latrine in room XIV similarly was altered, with the deep channel around its walls filled in with clay and rubble, to create a flat floor throughout the room. It may have been at this stage that the doorway from the courtyard to room II was blocked. Whether or not the building can still be identified as amansio in the early fourth century is uncertain, but no doubt the by now self governing vicus could use the structure for administrative or social purposes if it was not so required. The only evidence which might suggest the presence of women was found in the floor level of this period one small blue bead in Room VII, and part of a jet hairpin in Room XI.

Period IV, Post A.D. 369. The eastern wing of the building (rooms XI to XV) was completely rebuilt once more in the Theodosian period. The old outer east wall was levelled, and huge undressed blocks of sandstone were placed upon the regular masonry, whilst a similar wall of large stones was laid alongside the old courtyard partition wall, just within the courtyard. This created a slightly broader wing than before, and an additional room seems to have been attached to this structure at the southern end of the courtyard, although stone-robbing made its identification difficult. This last period had suffered from stone-robbing, particularly at the northern end, but a patch of flagging (Plate XII, 1) in room XII could be associated with this building's floor, and a similar patch of high flagging at the southern end of the courtyard, badly worn by wheeled traffic, suggested both that carts had been accustomed to backing up to the building and that such activity had gone on for a very long time. It was thus perhaps a store-house of some kind. There was no evidence to suggest that the rest of the former mansio had been reconstructed at this time, although a small clay oven was found in the centre of room VI at a level associated with this period.
Offline Profile Quote Post
Excavations of 1967-1969 · Reports & Papers