Eday, Green

HY 569 290 This survey was carried out in April 2006 after the landowner, Mr Peter Mason, had alerted Orkney Archaeological Trust to the fact that ploughing had revealed a scatter of material of apparent archaeological interest, including worked flint and a Neolithic polished macehead. The aims of the survey were to define the existence and limits of any possible structure associated with the spread of material on the surface and identify any other anomalies of potential archaeological significance within the area.

Gradiometer survey was undertaken over approximately 1.2ha, and on the basis of the initial results a further small area of 1600m2 of resistance survey was examined to investigate the apparent structures suggested by the gradiometry. The gradiometer survey shows a discrete concentration of magnetically enhanced features of archaeological significance, suggestive of domestic occupation, and identified several outlying features that may be related to this possible settlement. The resistance survey seems to have confirmed the presence of settlement, presumably Neolithic in date, in which at least one structure can tentatively be identified. This structure appears to be roughly rectilinear in shape and measures approximately 13 x 7m, a similar size and form to house 1 at the Knap of Howar on Papa Wray. The anomalies to the SE of this possible house may indicate some kind of work area and/or midden. However an alternative hypothesis is that these anomalies represent the remains of a second structure, which has been damaged by ploughing and lost any clearly definable form.

The survey also highlighted both ridge-and-furrow and modern ploughing, visible as a result of disturbing magnetically enhanced material of archaeological origin. This suggests that the site continues to be actively eroded by agriculture.

Archive lodged with Orkney Geophysics Unit.

Sponsor: Orkney Archaeological Trust and Peter Mason.

J Moore 2006

HY 56971 28987 A fourth season of work at this Neolithic settlement site took place 31 July–4 September 2010. The main aims were to carry out further work on the interior of Structure 1, a Neolithic building and to extend the trench to the N and E to expose the full extent of both Structure 1 and Structure 2. The 2009 trench was reopened and an extension was

stripped of ploughsoil. A total of 84sqm was opened increasing the overall trench area to 193sqm. The new area was planned but not further investigated. Visible features included a probable new building, Structure 3, which is similar to Structure 1 and these may possibly be linked together. Structure 3 also runs parallel to the previously noted gully which continues through the new area to the NE corner of the trench and appears to have had a stone cover. There are other possible structures visible N of Structure 2 and to the SE of the gully. No further work was carried out on Structure 2; however, sampling of the deposits to the S would suggest that this area consists of midden material from the settlement. There are

the footings of at least one other building cutting through the remains of Structure 2. Within Structure 1, the area SE of the hearth was excavated.

The occupation deposits here were taken down in layers and bulk sampled within a 0.5m square grid. These residues and additional soil samples are currently awaiting analysis. The floor by the wall was made up of compacted sand, a layer that continued under the wall. It is not clear whether this sand was in situ prior to the construction of the building; however, it contained a small quantity of cultural material and it is thought to be probably redeposited. Two small scoops had been cut into this layer. Towards the centre of Structure 1, the floor was made up of darker soils interspersed with some flat paving slabs. The previously noted runnel was found to continue along the SE side of the hearth and was partly covered by paving. Further work on the hearth confirmed that it had undergone repairs on at least two occasions, with the side slabs being replaced. The lower edges of the replacement slabs were bevelled and had probably been hammered into place. The two stones that formed the hole over the runnel leading out from the hearth were found to be two halves of the same stone, which had been bevelled and also probably hammered into place, an act which may have caused it to split. Finds from the site consisted of pottery and coarse stone tools together with quantities of worked and unworked flint. The only find of significance was a leaf-shaped flint arrowhead recovered from the ploughsoil over the new area.

Archive: To be arranged

Funder: BEVARS

HY 56971 28987 A fifth season of work at this Neolithic settlement site took place 30 July–3 September 2011. The main aims were to carry out further work on the interior of Structure 1, an Early Neolithic building and to investigate its relationship with the adjacent Structure 3 with which it appears conjoined; to further investigate Structure 2, to establish its relationship with the midden and to establish the extent of the midden.

Three 0.5m test pits were dug a little to the S of the previously opened trench and midden material was found in the base of the most easterly of the three pits. This pit was extended to a 3m N–S slot and revealed that the end of the midden material was c2.3m S of the entrance to Structure 2. Farther to the W a trench edge section had revealed that the midden extended up to 7m from the SW of Structure 2. It would therefore appear that the midden was unevenly spread and it is possible that it was made up of different piles of material. The midden ended at the wall of Structure 2 and would appear to be laid up against it. It was noted that the midden was directly overlying the windblown sand which covers the natural boulder clay outside the structures.

Work on Structure 2 has shown that it only had a slight wall, which suggests that it may be an enclosure rather than a building. The full extent of this structure remains unclear but it has an entrance to the SE. Within this entrance was found a clay-lined sub-rectangular pit measuring 1.50 x 1.15m with the long axis running NW–SE through the entrance. In the pit were three flat slabs sealed in by the clay and, along the long axis, a small linear cairn had been placed over the slabs before the pit was backfilled. Due to time constraints it was decided to complete the excavation of the pit in 2012.

Within Structure1 more of the occupation layers were excavated and sampled, and it is becoming increasingly apparent that the floor of the house has been subjected to substantial activity. There are a number of shallow runnels and small scoops cut into the boulder clay with many of these intercutting. There has also been further evidence of depositional practices within these features but it has not as yet been possible to ascertain if they have any other function. One scoop in particular had an egg-shaped quartzite hammerstone placed on a slightly raised platform at its centre, similar to several other features previously excavated.

The upper hearth in Structure 1 has now been fully excavated. It was found to consist of a single on edge slab that was propped on a pile of small stones in an oval bowl-shaped pit measuring 0.60 x 0.45m. Although the basal fill of the pit contained charcoal there was no evidence of burning anywhere else within this feature. This would suggest that in its final form it did not actually serve as a hearth. The pit was crossed by a runnel and it was here that an upturned quern and rubbing stone had previously been found.

Further work on the lower hearth of Structure 1 has revealed a complex series of events and, although further work is still required, a picture of the sequence is emerging. At the centre of the building is a hollow which may or may not be natural. In the middle of the hollow a bowl-shaped pit c1m in diameter was dug into the natural. A hearth was constructed around this pit, the shape of which is not yet fully apparent, but was clearly not rectangular judging from the positioning of the one surviving side slab and stone holes. At about the same time a V-shaped channel was formed in the base of the pit by the placing of two long flat slabs running NE–SW in the base of the pit. These were held in place by several stones to either side and packed with clay in the base of the channel. Dark deposits began to build up in and around the hearth.

Subsequently, the hearth was modified, all but one of the side slabs were removed and were replaced by a rectangular arrangement of on edge slabs within the pit. It is likely that these were replaced on at least one occasion. This was done by hammering into place stones that had had their bottom edges bevelled to facilitate this. The upper end stone was partly supported at one end by the surviving stone of the earlier arrangement. By this time further deposits had built up.

In its final phase the hearth consisted of this rectangular arrangement and measured 1.00 x 0.75m. The slabs at either end had both been fashioned to have what looks like a mouse hole, both had been broken presumably by being hammered into place. The holes were aligned with the internal V-shaped channel. This was covered with two hearth plate stones, only one of which appeared to have hosted a fire. At the upper, NE end of the hearth, a D-shaped scoop had been created and appeared to have given access to the ‘mouse hole’, however this had been sealed off with a packing of clay. At the other end the ‘mouse hole’ led into a shallow runnel, which in turn led into a larger runnel that ran out through the entrance to Structure 1.

Work at the junction of Structures 1 and 3 suggest that they may be the same building, although the excavation of this area is at an early stage and Structure 3 is much wider than Structure 1. Structure 3 has its own doorway to the SE. Two small flat cobbles, each of which had an incised triangle, were recovered from the vicinity of the junction of Structures 1 and 3. Several large patches of clay were found in the demolition material and it may be that the walls of the buildings had an internal lining of daub.

Other finds recovered from the site have consisted of pottery and coarse stone tools together with quantities of worked and unworked flint. Also recovered were three pieces of Arran pitchstone, two pieces of haematite and some pumice. One piece of worked bone was also recovered.

Archive: BEVARS (currently)

Funder: BEVARS

BEVARS, 2011

HY 56971 28987 A sixth season of work at this Neolithic settlement site took place 21 July – 1 September 2012. The main aims were to carry out further work on the interior of Structure 1, an Early Neolithic building, and to complete the excavation of the hearths and associated cut features in the floor of the building; to investigate its relationship with the adjacent Structure 3 with which it appears conjoined; to further investigate the interior of Structure 2.

In Structure 1 most of the remaining features were excavated down to the natural boulder clay. Final work on the SW hearth confirmed the interpretation from 2011 (DES 2012, 131–132) that it had been constructed in and around a sub-circular bowl-shaped pit between 0.75-0.80m across. Firmly set in the base of this pit were the two stones of the V-shaped channel running under the hearth. When lifted they were found to consist of one single sandstone slab that had been split laterally before being set in place.

Further scoops and runnels were also noted inside Structure 1. One of these runnels came out of Structure 3 and ran down through Structure 1 to where it joined other runnels, some of which still had stone covers. This runnel was truncated from above by a large shallow pit that contained several flat sandstone flags. Originally interpreted as flooring it is now known that this pit cut down through the demolition material overlying much of Structure 1, and it is now thought to be a probable paved pathway leading out from a narrow doorway in Structure 3.

That Structure 3 was in use after Structure 1 was demolished is now clear. The presence of stone holes seen lying on the projected line of Structure 1 has been noted within Structure 3 and this would seem to indicate that Structure 3 was probably constructed later. What is not clear however is whether the two buildings were ever contemporaneous with each other or whether Structure 3 replaced the earlier building. The presence of the covered runnel leading from one to the other would suggest that this might be the case, but further work next year will seek to answer this question.

One other feature of note was a posthole in Structure 1 close to the possible doorway to Structure 3. Measuring c0.5m in diameter this was adjacent to a similarly sized shallow pit noted in 2011 that had contained a deliberately placed hammerstone on a platform at the centre of its base. The posthole however was at 0.35m, much deeper than the pit and judging from the positioning of the packing stones had contained a circular c0.25m diameter post. It is not clear whether the post had rotted in situ or been removed or cut down. The posthole was then covered by a sandstone flag that eventually collapsed into a void below. Finds from this feature included an orange quartzite polisher and several sherds of unabraded pottery.

In Structure 2 the excavation of the pit in the entrance was completed. The three slabs in the pit were removed revealing only boulder clay. The surface of the clay was found to be covered by finger marks which led us to excavate a little further, however nothing further was found and the finger marks would appear to have derived from people dragging wet clay up from the base to line the sides of the pit before the slabs and linear cairn were placed in it. It is difficult to assign function to this feature. The placement of the slabs in a pit could be interpreted as a paved entrance to Structure 2, however the presence of a cairn on top of the paving would suggest otherwise, unless the two events were separated in time.

The excavation of the interior of Structure 2 was begun. The floor is bumpy, not flat and would not have functioned as a usable surface in a working building. This would concur with previous observations that the structure was too flimsy to have been a building and is more likely to have been an enclosure. Within the structure at least three cut features have been noted thus far, one of which, a pit, has been excavated, but which contained nothing of significance. A linear cut feature has been noted which runs under the large sandstone slab at the centre of Structure 2, but this remains to be excavated.

The soils within Structure 2 were slightly lighter and sandier than the midden deposits outside, but these have also proved to be rich in artefacts. These include the usual range of coarse stone tools, flints and pottery that have previously been associated with the site and include the most significant find of the season, a red sandstone roughly triangular polissoir up to 0.40m long. Both sides are deeply bowled and each has two deep parallel grooves ground into their surfaces. Although further research is required, it very much resembles the type of polissoir normally associated with the polishing of stone axes. If so it would add to the repertoire of stone working carried out on this site.

Whilst almost all the material from the site can be demonstrated to be available locally, some other finds recovered this season are clearly exotic. These include more pieces of unworked Arran pitchstone and three flakes of greenstone tuff, with evidence of polished surfaces and probably removed from polished stone axeheads. Visually these are all convincing matches for stone from the Group V1, Langdale series, but this will need to be confirmed by testing.

Archive: BEVARS (currently)

Funder: BEVARS

Mick Miles, BEVARS

2012

HY 56971 28987 A seventh and final season of work at this Neolithic settlement site took place over two periods of six weeks, 19 May – 28 June and 14 July – 25 August 2013. The main aims were to complete the excavation of Structures 1 and 2; to establish the full extent of Structure 3 and to extend the trench to the N to expose the whole of the structure; to investigate the nature and function of Structure 3; to investigate Structure 4; to investigate the gully separating Structures 1 and 3 from Structures 2 and 4; to try and establish cross site stratigraphy between the various structures and the gully; to continue the programme of environmental sampling of all contexts; to try to determine the nature, extent, sequence and dating of the later intrusions into the Neolithic archaeology; to look for evidence of any activity prior to the construction of the Neolithic buildings.

In Structure 1 the point where the two main runnels crossed each other was excavated and it was found that one had been blocked off and replaced by the one leading from Structure 3. This confirmed the presumed sequence and proved that Structures 1 and 3 had been in use simultaneously. A small bowl-shaped scoop in the upper room was also excavated. This was found to contain the remains of a crushed pot, possibly a storage jar.

In Structure 2, a large red sandstone block lying in the interior was lifted. Beneath it was the same mixture of building rubble and midden that was found across the whole interior of this structure. This material was removed and four stoneholes were found, two of these contained the remains of probable stone packing. The stoneholes were set symmetrically in pairs, c1.8m apart on either side of a room c3.2m long x 2.8m wide and could possibly be the remains of bed boxes.

There was no evidence for any further surviving parts of this structure. There was however one stone abutting the stonehole on the NW side that appeared to be a continuation of the wall going off at a slightly different angle, suggesting that the NE pair of stoneholes may have been placed at the pinched midpoint of a two room building. It now seems probable that Structure 2 was a house, although there was no trace of a hearth and no material relating to its use was found. The material overlying Structure 2 has been interpreted as being a mixture of midden and building rubble.

A limited amount of work was carried out on the central part of the gully between Structures 1 and 2. This work confirmed that the gully was a well built stone-lined drainage channel with a cover consisting of large well laid sandstone flags. It is probable that they also served as a pathway in an earlier phase of the site, as the stones were contiguous with other stones forming paving in one of the doorways of Structure 3. This paving continued on the SE side of the gully for a short distance towards Structures 2 and 4 where it was truncated by the supposed quarrying cut. The paved gully cover was seen to continue to the NE corner of the extended trench, separating Structures 3 and 4 and defining the outer face of their walls.

In the part of the gully that was excavated, evidence for part of a small stone-lined and covered drain was noted overlying the gully cover. Only a short section of this feature survived running around the outer wall of Structure 2. Unfortunately this feature and the adjacent area were badly damaged from above by a post-medieval pit. Although not directly linked, it is possible it would have been crossed by the nearby stone pathway leading from the S entrance to Structure 3 through the remains of, and over the surviving wall of the demolished Structure 1. Only a small part of Structure 4 was seen, lying between the gully and the E side of the trench under which it continued. Like Structure 2, it had been damaged by quarrying and had probably also lost an entire room. The mid points of both buildings were clearly established and there was a gap of c2m between them and presumably one structure had followed the other.

The outer face of the wall was formed of large sandstone slabs which also defined the side of the gully. The inside wall was probably built of similar slabs, although only one survived in the trench, together with a contiguous stonehole which formed part of the edge of the truncation. The wall was infilled with sandy clay. There were two stoneholes for on edge stones, one of which retained its packing, set at right angles to the wall. These were 1.8m apart and may have formed the ends of a bed box. The SW stonehole which contained the packing was about 1.2m long. Within the building it turned at right angles and continued for short distance to the NE with packing for another on edge slab.

Adjacent to these stoneholes was a hearth measuring 1.2 x 0.6m internally. The trench was extended 1m to the E and 2m N–S to expose the whole of this feature. It was placed centrally just inside the upper room and had an almost exact length to width ratio of 2:1 as did the hearth in Structure 3. Like the hearth in Structure 3 and possibly the later hearth in Structure 1, it appeared to have had a dual function. At its SW end there was a flat hearth plate with no evidence of burning, whilst the NE end was deeper and had hosted fires, the ashes from which appeared to have spread suggesting it had been regularly raked out. The hearth in Structure 1 had hearth plates at both ends, only one of which had hosted a fire, whilst the hearth in Structure 3 had a series of large deep pits for fires at one end and smaller, shallower pits at the other. Also like the other hearths all the side slabs had their lower edges bevelled to facilitate their placement into the ground. There were layers of surviving occupation material up to 60mm deep across the floor of Structure 4.

The major part of the work in 2013 was the excavation of Structure 3, a large Neolithic stone building. The internal dimensions of the building were 9 x 4.75m and there were two rooms separated by large on edge slabs set into the walls on either side. Apart from some minor disturbance to the W the internal walls of Structure 3 formed a complete circuit of the building. Only the bottom two courses survived but they were well built using mostly large red sandstone slabs up to a metre in length. Many of these had been dressed with facial pecking and in at least one case this pecking had given the facing a neat curvature that matched the layout of the wall.

The newly exposed upper NE room proved to be in better condition than the lower room and it was covered by a substantial pile of stone debris. In the upper part of this material was a setting of two in situ slabs at 90degrees to each other together with a similar but slightly displaced slab on top of the rubble. All three stones were 0.65m long by 0.15m high. Within this setting was a circular pit that had been dug through the rubble and occupation levels into the natural. The pit was empty. It is possible to speculate that this was a robbed out Bronze Age cist built into an existing stone mound.

The rubble consisted of flat thin sandstones, mostly broken and some with evidence of knapping, patches of clay and large red sandstone slabs similar to those in the walls. The thin slabs were spread across the floor and many were embedded in the occupation layers. The clay was similarly spread across the interior whilst the sandstones were found lying in proximity to the walls. This would suggest that the building had a stone roof with clay caulking which had collapsed, bringing down part of the walls with it. The removal of the rubble revealed only small patches of in situ paving. This suggests that the building had been abandoned and stone removed prior to the collapse of the structure.

There were at least two layers of occupation in the upper room and these were extensively sampled. There were also a number of features cut into the natural below. These included several pits, a runnel, stoneholes for a rectangular structure and a hearth.

The hearth was set entirely within the upper room and close to the midpoint of the structure. It consisted of four large grey sandstone slabs forming a 1.5 x 0.75m rectangle. These were set inside a large pit, cut into the natural boulder clay and held in place by clay packing. It became apparent that each end had its own pit and that the SW end of the hearth had been used for fires whilst the NE end had possibly been used as a place to keep things warm. There was some evidence of re-cutting of pits within the hearth and charcoal and ash were also noted in the deposits around it.

There was a rectangular feature was set against the wall to the NW of the hearth. It was formed of two stoneholes for substantial on edge slabs and the central division slab set into the wall. It was c2m long by 1m wide. Both stoneholes retained their packing. Traditionally such a feature would be interpreted as a bed box however this is by no means certain. To the NE of the rectangular feature was a further long cut but no function could be ascribed to it. Around the upper parts of the building were a number of pits and scoops. Immediately adjacent to the NE end of the hearth was a large flat red sandstone boulder sitting in a shallow pit. The upper face of the boulder was pecked all over, as if it was intended to be used as a quern, although there was no sign of it having been utilised.

A runnel which could have served as drainage started at the highest point at the NE end of the interior and ran just inside the SE wall of the upper room. About halfway along it curved in about a metre towards the centre of the room before continuing SW into the lower room. It is likely that the runnel was covered, but only one possible covering stone survived. In the gap between the runnel and the wall were seen several small circular depressions in the floor up to 100mm in diameter and no deeper than 80mm. It is feasible that they may have derived from posts or from the legs of timber structures or furniture.

There were a number of features running across the midpoint of the building between the dividing slabs set into the walls. The arrangement of the features all but separated the two rooms from each other. There would have been no access between the rooms on the NW side of the hearth, whilst to the SE there was just a small gap of c0.5m between the hearth and a square posthole about 100mm across. The runnel passed through this gap into the lower room which stepped down a level.

It was not possible to fully excavate the lower room in the time available. Material from the collapse of the building was similar to that found in the upper room but had been subjected to a great deal of later activity. Several pits were cut in from above and some stone had probably been quarried away. Some of this later activity was probably not prehistoric.

On the SE side of the room there was an irregular clay lined feature formed of thin on edge stones. This feature was also cut into the rubble and was probably later prehistoric. Two flaked stone bars and a red sandstone cobble with a large cup mark were within the feature and three ard points found adjacent to it. One ard point was of particular note. The stone was volcanic, appeared to have been polished and was a very good match for a modified polished stone axe found close by in the ploughsoil in 2010.

The larger of the pits was situated in the centre of the room. It had been cut deeply into the layers below the rubble and through at least one and probably two layers of paving. It had also gone into a stone covered drain that appeared to be coming from the same direction as the runnel in the upper room and heading towards the drain that entered Structure 1 in the S doorway.

A little to the SW of this pit was a feature that was sealed beneath the rubble. This was a clay lined stone feature formed of thin side slabs and the base was formed of a single large grey sandstone flag measuring 1.2 x 0.5m. This stone sloped down significantly towards the centre of the room. The feature could have served to hold water but its function is otherwise unknown. It was also seen to cut into an earlier larger feature which in turn had been cut through paving.

Elsewhere in the lower room some presumed patches of occupation layers were seen and sampled as were some other patches of possible paving but at differing heights. One of the lowest of these patches appeared to be at the same level as the paving in the SE doorway and the main gully cover outside the building. It is probable that there were many phases of activity associated with the interior of Structure 2 only some of which have survived. It is also possible that some of the activity in the lower room may predate Structure 3.

Further evidence of multiple phases is provided by the Structure itself. Unusually the building had four doorways. However, it is improbable that they were all in use at the same time. There was one doorway in the centre of the upper NE end of Structure 3 that appeared to be integral to the original construction of the walls. At some point during the life of the building this doorway was modified. The gap in the inside wall was enlarged and the side slabs removed. The other three doorways were located in the lower room, to the SW, S and SE.

The final piece of evidence for the prolonged use of the site came from the construction cut for the whole of the Structure 3 building. The cut was clearly visible on the NW and SE sides of the building. To the NW the cut was dug into a sandy clay subsoil, whilst to the SE it was cut into the compacted sandy clay infill behind the main wall of the gully and filled with darker midden-like material. Significantly this cut truncated the paving and side wall of the passage leading into the SE doorway, which suggests that the paved passageway originally led into some other structure that had been presumably removed prior to the building of Structure 3. Intriguingly this passageway appears to have been reused in the new building. There were postholes for probable doorposts cut into both the occupation deposits inside the inner doorway and also against part of the passageway that appeared to have been rebuilt but not repaved.

At the landowners request this was the final season of this project. At completion the whole trench was covered over with geotextile to preserve the remains and as far as was possible all upstanding masonry was bolstered with sandbags The site was due to be backfilled by the landowner using imported topsoil.

During the course of the seven seasons the area opened up for excavation totalled c225m2. This represents just 25% of the area of high geophysical responses first identified in 2006. What this project has found undoubtedly represents a palimpsest of Neolithic buildings and activity with many phases, some possibly later prehistoric. A long period of post-excavation work will be now required to analyse and integrate the evidence from c500 contexts, associated drawings, paperwork and photographs, several hundred environmental samples and many thousands of finds. A programme of radiocarbon dating has already been started and we are intending to produce a monograph in due course.

Archive: To be arranged, currently with BEVARS

Funder: BEVARS

Mick Miles, BEVARS, 2013

(Source: DES)