Jarlshof Excavation and Survey

Date 2003

Event ID 1039122

Category Project

Type Project

Permalink http://canmore.org.uk/event/1039122

Excavation of the Old Scatness settlement began in 1995 as part of the Old Scatness/Jarlshof Environs Project (see DES 2002, 105-107). Excavation continued in 2003 with the investigation of the buildings W and E of the broch, the broch tower itself, and the post-medieval and Norse features to the E. A geophysical survey was undertaken at Jarlshof (HU 398 095).

The broch tower

Excavation in 2003 showed that the broch was constructed over a midden on the N face, and over a flag raft (or zone of paving) on the NE and NW arcs. The flag raft appears to have been used in order to disperse the weight of the tower. The initial northern wall construction over the midden appears to have been reinforced by large orthostats running in a parallel arc to the broch wall at this point, which contained a rubble fill and acted as a buttress. The flag raft either side of this stretch of walling and concentric arc of orthostats lay over a buried soil to the NE and peat to the NW. The wall of the broch on the NW side was a course wider in diameter at this point and was keyed into the original northern circuit. This increase in diameter appears to have been intentional and may have provided a greater weight distribution over the softer peat beneath. The flag raft was covered with yellow clay, as was a curved wall concentric with the broch. The application of clay seems to have been intended to act as a sealant for the underlying peat. A blocked entrance to the broch within this northern circuit may have been a constructional feature allowing direct access to the cell and staircase; but the destruction which occurred within this zone in 1975 limits interpretation.

The main western entrance to the broch was investigated and excavated to ground-floor level. The in situ pivot and threshold stones provided evidence for the location of an outer doorway to the broch tower. A massive triangular lintel above the entrance may have been intended to provide an impression of monumentality to anyone entering. A structural function for the triangular lintel can also be envisaged, since it would have acted somewhat like an arch, to distribute the weight of the wall above the entrance. The passageway was found to extend at least 1m beyond the outer wall face of the broch, with a slight funnel-like curvature to the northern wall. Excavation of the passageway was necessarily limited, due to the close proximity of later structures, including the corbelled cell (Structure 24).

Structures W of the broch

Excavation continued in the large early roundhouse (Structure 12) to the W of the broch. It can now be seen that this building was constructed on a raft of flags and small rubble above an earlier red midden. Other features which pre-dated this building, including wall fragments, were also exposed in the central area of the building. The original substantial western entrance to this building, with double bar holes on the S side and a pivot stone still in situ, had been blocked in antiquity, and investigation this season showed that the blocking may have been constructed as a stair to an upper storey, or to the lintelled roof of the entrance passage. The first cell to the N of the western entrance contained the remains of a small oven or wall hearth, later dismantled and flagged over.

Also in this cell was a pit dug into the red midden, containing cattle bone; last season it was thought to be contemporary with the building. Further excavation has shown that the pit was earlier, and that the flagging which sealed it formed part of the raft on which the walls of Structure 12 were constructed. Further investigation showed that the early red midden sealed a darker layer, which in turn sealed a mineral sand layer at least 80cm deep. The eastern annexe to this building was also excavated further, and a clay-lined tank surrounded by paving was discovered.

The sequence of alterations to Structure 12 now seems to be clear. At the point at which the aisled piers were lengthened to butt the walls, the southern wall was rebuilt to remove the corbelling in its upper courses. The main western doorway was blocked but another entrance into the new Structure 8 was added next to it, whilst a new external access was created to the SE. The old doorway was converted to a stair leading to an upper storey or a mezzanine. The annexe to the E seems to have been constructed at the same time, and the hearth was moved from the centre of the building to an off-centre position in the SE, perhaps to allow for a central roof post.

Structure 24, to the N of Structure 12, was a virtually complete small corbelled cell. Further excavation this season showed that it actually utilised the S and E walls of a much larger, older building, and that what appeared to be its earliest floor, a compacted clay surface, actually belonged to the earlier structure (Structure 33). This earlier building was itself constructed from the northern exterior wall of Structure 12 and the exterior wall of the Structure 12 annexe. This floor appears to have sealed a yet earlier structure: what at first appeared to be a flagstone on the present surface is actually the top of an orthostat.

To the S of Structure 12, and sharing its southern wall, was Structure 14 which, with its scarcement ledge and long piers replacing aisled ones, seems to have been converted at much the same time. The western doorway of this building was similar in style to that of Structure 12, with evidence for a bolt hole and an in situ pivot, as well as a paved entrance which leads out towards the settlement ditch. The furthest extent of this pavement cannot be known, as it was damaged by machine activity before the site was purchased. Last season, excavation showed a drain or flue running in front of the cells, and this year more evidence for drains in front of the cells was found. There was also a stone-capped drain or flue which led from the 'aisle' behind one of the western cells, under the wall and out towards the ditch.

Between the outer broch wall and S of Structure 12, Structure 22 was also further investigated. This building was originally thought to be an external space or courtyard, but it is now apparent that a set of orthostats for piers surrounded a central hearth similar in construction to the hearths in Structures 12 and 14, formed from a flagged base with a cobble kerb set in white clay. A structure interpreted as a possible ash box was discovered immediately to the N of this hearth, and just to the W the bottom stone of a saddle quern, suggesting that this may have been a food preparation area. A clay-luted tank constructed and capped with stone was also investigated. The NW corner of this structure, where the outer E wall of Structure 14 and the southern wall of Structure 12 met, seemed to be subsiding, and excavation showed again that there were earlier walls underlying these deposits. It seems that most of the subsidence on the western part of the settlement may have been due to the presence of earlier buildings.

Structures E of the broch

The slightly later roundhouse (Structure 21), to the E of the broch, had the greatest diameter of any of the buildings on site - approximately 12m internally. It seems originally to have had short piers, later rebuilt as long thin ones. As excavated last year, the building was seen in its later phases, with a central wall dividing the structure into two and a corn drier built into one half, probably in the early centuries AD. The fill of this building, excavated this year, contained a very large dump of animal bone. To the W, the outer wall of the building produced further evidence that some of the Old Scatness roundhouses had an upper storey; stones bonded into the outer wall and butting against the outer face of the broch wall formed an external stair which may have emerged into the building at approximately the point where a large sandstone slab, c 1.5m across, was found tipped into the building as if it had slipped from a position overlying the pier and wall head. This large slab would presumably have formed part of the upper floor. Finds from Structure 21 include fragments of pale green vessel glass and copper alloy. After the piers in this building started to collapse, it was effectively abandoned; the southern half used only for access, whilst the northern half had a series of later buildings revetted into the midden infill, abandoned in their turn before the construction of the corn drier.

Further investigation in a small sondage on the E flank of the site revealed the large build-up of midden layers above the ditch cut with its stone wall revetting, giving evidence of another part of the ditch circuit.

Area P; post-medieval and Norse

Two small trenches were excavated to the E of the main area, by the side of the airport access road, to further investigate possible Norse contexts identified in an earlier soil sampling pit. Trench P1 contained midden-rich layers dominated by the bones of large gadids (fish of the cod family), some of which were still articulated. These middens sealed a layer with a high proportion of burnt stones, from which several copper-alloy fragments and pieces of steatite vessels were recovered. The only structural element identified was in trench P2: a length of stone wall with a midden infill, running N-S. No floor surface was found in association with this wall. On the western side of the wall a number of windblown sand layers had built up, sealing a shell-rich layer of periwinkles and limpets, whilst on the eastern side the middens contained proportionately more large animal bones than fish remains. On either side of this wall the midden layers sealed soils which could be stratigraphically related to the Iron Age soils identified around Old Scatness. The greatest proportion of finds from both trenches consist of fragments from large steatite vessels. Other finds include worked bone and a number of iron objects, including two possible tanged knives. In both trenches, the Norse contexts were sealed by post-medieval layers, suggesting that the extent of the post-medieval settlement was much greater than formerly supposed. Finds include a 17th-century coin.

Jarlshof

Geophysical surveys were carried out within the terraced zone of the NE corner of the Guardianship area at Jarlshof, using magnetic (fluxgate gradiometry) and resistance survey (using an experimental square array configuration). The surveys allowed investigation of geophysical anomalies associated with both Norse and Early Bronze Age cultural zones of the site.

Investigation by magnetic survey of an area of approximately 1ha within the cultivated field to the N of the Guardianship area indicated two spurs of archaeology running from the main site. These anomalies, one to the W end of the Guardianship area and the other to the E end, were consistent with accumulations of magnetically enhanced material indicative of settlement.

Sponsors: HS, BP Operating Company, EU Special Transitional Programme, SNH, Shetland Amenity Trust, Shetland Enterprise Company, Shetland Islands Council Charitable Trust, University of Bradford.

S J Dockrill, V E Turner and J M Bond 2003