Loch Tay, Oakbank

NN74SW 16 7230 4429

For logboat erroneously reported as having been found at this site, see NN74SE 30.

The crannog for which Dixon suggests a location at NN 726 442 or NN 727 442 is to be identified with that previously noted by the OS at NN 7230 4429; Oakbank Cottage is at NN 722 443. This is probably the more easterly of the 'two great cairns' noted by Gillies (1938).

In 1979 examination revealed that the main mound of the crannog measured up to about 25m in diameter; the highest point was about 1m below the normal level of the loch and the bottom was at a depth of 2m on the landward (NW) side and about 3.6m on the SE. On the SW there was noted

a 'smaller circular feature', possibly the remains of a landing-stage, measuring about 7m in diameter and linked to the main mound by a scatter of boulders. In the angle to the S were noted oak and soft-wood piles and probable softwood planks; at least one oak pile projected from the top. Radiocarbon assay of two piles (one from the top of the crannog and the other from the loch-bed to the S) yielded dates of 595 +/- 55 BC (GU-1323) and 460 +/- 60 BC (GU-1325) respectively. Two rows of piles on the N and a ridge of bottom material were all that remained of a causeway to the shore.

In 1980 trial excavation uncovered an array of soft and hardwood piles embedded in a matrix of bracken, moss, twigs and other organic matter between 1.5m and 2m deep; two of the piles (possibly from the last phase of use) were radiocarbon-dated to 410 +/- 60 BC (GU-1463) and 455 +/- 60 BC (GU-1464). Fragments of iron slag were found with basket-and hurdle-work, numerous small wooden artifacts (including a peg, a whistle and a spindle-whorl), whetstones and other worked stones and vegetable debris which included cereal remains and cultivation weeds.

In 1981 further excavation took place in the centre of the crannog. Removal of the overlying large boulders revealed a layer of small

stones and occasional timbers which probably formed a log floor and a timber wall; the upper floor timbers were removed to expose a second floor beneath. Excavation towards the N edge of the mound revealed a band of large stones overlying a sloping deposit of organic materials, piles and timbers which had been formed by the partial collapse of the structure; three of the timbers were mortised.

Finds included a jet ring, a stone bead, several probable stone net-rings, part of a possible rotary quern; a plate, a dish and a paddle of wood were also found.

T N Dixon 1980; 1981; 1982; 1983; 1984; 1985; I Morrison 1985

A team of ten people continued excavation at this site in a six-week season in 1990. The team was comprised of STUA personell, volunteers, and students from the Department of Archaeology, who had learned to dive specifically to participate in the project.

Oakbank is a multiphase site and there is some evidence for the existence of two distinct structural phases. One perhaps antedates the other, or served as an extension possibly with a different function. Much of the outline of supporting piles for the dwellings has been defined, but substantial areas are still obscured by overlying boulders. The principal aim in 1990 was to remove as many of the overlying boulders as possible inorder to establish the complete outline plan of the structures.

Soon after the project began, it became apparent that all of the boulders could not be removed in the allocated time, even with the help of our custom-made underwater 'dump-truck'. It was decided to determine the diameter of the main dwelling, which entailed one team working in the shallows (1-2 metres) and another team in the deeper water (about 4 metres).Between them the two teams removed more than 20 tons of boulders, revealing worked stone objects, structural timbers and wooden artefacts. In the deeper water, supporting piles were discovered approximately one metre under the loch bed silts and some of them were about two metres out from what has normally been seen as the outer permieter of the main crannog mound. This indicates that the site is larger than previously estimated, and that a considerable period of abandonment took place before part of the site was covered by natural silt.

In the shallow water, near the intersection of the main dwelling and its causeway, several piles were exposed together with hurdles still attached to a 4m long mortised timber. The area was packed with interwoven branches and many wooden knots of twisted hazel were found, in some instances still tied around branches. The hurdles and branches may have been laid as flooring, as the mortised timber resembles a type of sill beam. It is one of the most interesting structural elements discovered on the site so far. One substantial pile had been driven through the middle of a large wooden bowl breaking it into four pieces. The bowl was excavated and after conservation it will be reconstructed. The Oakbank season once again brought up more questions than we were able to answer but the changes in our perception of the size and depth of the site show even greater potential for the future.

This years project was funded by The Munro Fund, the Russel Trust and the Prehistoric Society. Equipment and other support were supplied by the Department of Archaeology and the Scottish Trust for Underwater Archaeology.

T N Dixon 1990.

Excavations at Oakbank Crannog were carried out from 22 June to 4 August 1991 with a team of ten to twelve students and volunteers. The plan was to extend last year's excavation and open another new area adjacent to it. As usual the first task was to remove the boulders that covered the areas to be worked on. The Oakbank "underwater dumper truck" was put into action and altogether 25-30 tons of boulders were removed. Underneath the large boulders was exposed a layer of smaller stones and grit which also contained burnt bone and large amounts of charcoal, suggesting that the hearth was in this part of the house.

The most important discovery was the remains of a substantial floor. Part of the floor was exposed in 1980 but nothing since then, so it was believed that the rest had been destroyed. An even larger area was uncovered this year. Lying across the floor are larger timbers, at right angles to the floor timbers, which may be the uprights of a fallen partition wall that separated two rooms, or functional spaces, within the house.

The floor is situated near the middle of the site and east of it, nearer the edge, was uncovered a substantial number of upright piles with a jumble of branches, stakes and pieces of cut wood surrounding them. These pieces of wood were all produced deliberately by the crannog-dwellers but we can only speculate as to their use.

The most interesting find this year was a small, but finely made, stone pendant. There has been some speculation as to whether it was a small whetstone but it does not show the appropriate wear for that use. The work at Oakbank Crannog will continue in the future.

This years project was sponsored by The Russel Trust, The Carnegie Trust, The Moray Endowment Fund, The Munro Fund and British Gas PLC.

T N Dixon 1991.

Oakbank crannog is situated near the N shore of Loch Tay which is situated in a deep glaciated trough in the Trossachs at an altitude of 107m OD. It lies in front of Oakbank cottage and 190m ESE of Fearnan Post Office; the site-name Fearnan has been attached both to this crannog and to that (otherwise known as Fearnan Hotel) situated some 220m to the W.

Underwater survey has shown that the main crannog-mound measures about 25m in maximum diameter. The highest point is about 1m below the normal level of the loch and the bottom of the mound is at a depth of about 2m on the NW side and about 3.6m on the SE. On the SW there is a projecting 'smaller circular feature' (possibly the remains of a landing-stage) measuring about 7m in diameter and linked to the main mound by a scatter of boulders. The radiocarbon testing of piles and timbers has yielded four dates ranging between 595 +/- 55 bc (GU-1323) and 410 +/- 60 bc (GU-1463), which may be calibrated to between 788 and 697, and about 404 cal BC respectively.

In 1981 a paddle was found during excavation of the organic matrix of the crannog. It is currently undergoing restoration and conservation but is reported to measure about 1.35m in length, being 'shaped very like a modern paddle with a rounded, slightly plano-convex cross-section'. The published photographs show that the blade is roughly rectangular in form and occupies about 52% of the surviving length. The junction between the blade and the handle is protracted and the latter feature (much of which has broken away) is apparently of rounded section. The timber has been identified as oak and bears numerous clear toolmarks.

T N Dixon 1981; [T] N Dixon 1982; T N Dixon 1982; [T] N Dixon 1984; R J C Mowat 1996; information from Dr T N Dixon.

NN 7230 4429 In the summer of 2001, excavations were carried out at Oakbank Crannog for the first time since 1992. The break in examination of the site was due to the efforts made to interpret previous excavation evidence by the construction of a full-size crannog elsewhere in the loch and, as part of that work, the need to create a more secure method of funding the project in the future. The aims of the excavation were to assess the site after nine years of non-disturbance, to choose areas for particular attention in the next few years, and to examine the site in light of discoveries made through experimental archaeology carried out in the process of reconstruction.

The decision was taken to examine area B5 which is at the W end of the main crannog mound, adjacent to the extension area which is apparently separate. Areas to the W and E of B5 were examined in the past and it was hoped that B5 would help to explain the relationship between them. In fact it was not possible to open all of square B5, and a strip about 1m wide was left between B5 and B4. However, it is clear that the material uncovered in B5 shows elements from both B4 and B6.

On the E side of B5, there are large stones embedded in the organic matrix, similar to the situation further to the E. The W side of the trench consisted of organic debris, including bracken and animal droppings, similar to material discovered in B6. The finds from B5 also had similarities with those from the area to the W.

On the W side of B5, a trench 4 x 5m was opened. Large overlying boulders were removed and the underlying organic matrix was excavated. About 40 new piles were exposed along with a similar number of horizontal structural timbers. Most of the piles were alder but there were also a few of oak; they were recorded in situ and sampled.

The organic contexts were typically well preserved, and of similar constitution to those encountered in other areas of the site. The deposits were largely composed of wood debris, bracken, woodchips, animal dung and general comminuted vegetation. In one area the animal dung was particularly concentrated, and appeared to constitute a discrete area of the trench. Several samples of this deposit were taken. There was also a high incidence of burnt bone and charcoal, particularly in the upper levels, and this was complemented by a large number of apparently fire-cracked stones or 'pot boilers' in association.

In the middle of the trench a deposit of fine blue/grey clay was uncovered surrounding a broken pile point. The timber was apparently split by being driven onto a large flat stone which lay underneath the clay. The clay was similar to deposits discovered in previous excavations but its purpose is not clear. Around the clay deposit several pieces of very coarse, poorly fired pottery were recovered, one showing incised decoration. Ceramics have not previously been a common find at Oakbank. A crucible sherd, with a metallic residue adhering to the inside, was also recovered from this area. In previous seasons another fragment of a crucible was uncovered in association with slag. It has not yet been established whether the two sherds are from the same vessel.

It seems likely from the evidence in trench B5 that there were at least two phases of occupation. Two fallen piles were uncovered, overlying and underlying organic occupation deposits, and the relative stratigraphic positions of piles with and without evident tool facets indicated that these had been driven into the site at different stages of construction. Patches of inorganic loch bed silts were uncovered apparently in situ, sealed above and below by organic deposits, suggesting some period of abandonment or inactivity, but this evidence requires further examination as it is very close to the natural edge of the site.

Wooden finds were well represented, by a number of pine tapers, two lengths of twisted hazel rods, a cut and carved wooden point and many piles with clear tool marks on them.

Around the edges of the site in the lower levels of the excavation some animal bones were recovered. The majority are in a very poor state of preservation as is normally the case with bone material from Oakbank. However, this year several examples were well preserved, in particular a fragment of a large animal femur and a small, probably ovicaprid, horn. Many cattle teeth were recovered as well as part of a mandible.

During the removal of the overlying stones and grit, a well-preserved bronze ring was discovered which may be the remains of a small penannular brooch. There is no sign of decoration on the surface. The provenance of the ring, in the overlying grit, means that it is not possible to assign it to a specific period of occupation on the site, although according to the radiocarbon dates it should date to the end of the 4th millennium BC.

The trench was backfilled using excavated loch-bed silts for the protection of the archaeological deposits and it is intended to return to this area for more extensive excavation in 2002.

Sponsors: Scottish Trust for Underwater Archaeology, University of Edinburgh - Dept of Archaeology.

T N Dixon and M G Cavers 2001

NN 726 442 In the summer of 2002, excavations continued at the Early Iron Age site of Oakbank Crannog in Loch Tay. The decision was made to excavate two areas (DES 2001, 79, fig 32, areas A1 and A2) to establish the foundation sequence in shallow deposits where it was possible to excavate to lochbed level. There was also some evidence that erosion was taking place on this part of the site.

The main features discovered in previous excavations in this area were the remains of a large structural timber with associated hurdles and a bundle of thin hazel withies probably used in the construction of hurdles. They lay beside some large piles made of elm, one of which had been driven through the bottom of a large wooden bowl. This situation suggested a period of abandonment in the area and subsequent rebuilding. The underlying deposits would represent the period of occupation before the collapse of the large timber and the hurdles.

Prior to the current excavation, large timbers aligned approximately NE-SW could be seen projecting from underneath the organic deposit; they were overlain with stratified occupation debris. There was a thick layer of bracken interspersed with twigs and small branches; another layer consisted of many small stones embedded in silty gravel. A number of timbers associated with this layer showed signs of burning. Throughout there were cut woodchips with significant concentrations in places indicative of on-site wood cutting.

After removal of the overlying layers there was an alignment of about 15 large timbers lying NE-SW lying on the sand of the lochbed. Underneath these timbers were two very large transverse beams that appear to have supported the others, although there is no evidence of joints to hold them together. They were embedded in the lochbed silts. Two of the NE-SW timbers lying in the centre of the trench had been cut to a point. They seemed to be close to their original positions and to have fallen during occupation of the site. Most of the others seem to have collapsed from the mound outwards and may represent part of the superstructure of the crannog when it was occupied prior to the structural timber and hurdles referred to above. Two very large timbers in the E end of the trench had holes through them. One of these projected from the section of the unexcavated adjacent area.

In previous excavations in and around the excavated areas several interesting artefacts were discovered. They include a canoe paddle, wooden pegs, burnt tapers, a small iron knife blade and other domestic implements.

In 2002, a similar range of artefacts was discovered. Of particular interest is a swan's neck pin, which should date to the earliest Iron Age and fits well with the radiocarbon dates for the site. The pin is very important as it is the first diagnostic artefact found since excavation began in 1980. The pin, like other artefacts from the site, reflects the high status of the people who lived on Oakbank Crannog.

In 2002, a second paddle was discovered but it is very different in shape from the earlier find, being very rounded and relatively short (50 x 25cm) while the earlier one is long and thin (70 x 10.5cm, excluding the handle) - more like an oar. It is possible that the short, wide paddle was for powering a canoe or dugout boat while the other was a steering oar. Both are made of alder, as are most other wooden finds from the site.

The excavated areas were located where the walkway to the shore joined the main site, and at the entrance to the crannog house. The previously excavated large timber associated with hurdles is almost certainly the threshold timber for the entrance-way, and is reminiscent of a similar timber discovered at Milton Loch Crannog in 1950. It is not surprising that many finds have come from here and associated areas, and that there is a mass of confusing timbers.

During habitation of the site, this area would have stood over relatively shallow water, possibly as little as a metre or so in depth. The large number of piles in A2, and A3 to the W, indicates considerable maintenance and repair - as might be expected where the walkway joins the main crannog structure. The large timbers exposed at lochbed level represent mainly the beams and crosspieces that were once supported by the piles. At least two of the timbers are piles that collapsed in antiquity and their length and the exposure of the points suggests they might not have been driven to a sufficient depth into the lochbed.

Excavations continued in the summer of 2003. The decision was made to excavate Area C2 (see DES 2001, 79, fig 32) which had been previously exposed in 1991. The main reason for excavating this area was to expose the foundation structure of an impressive overlying floor and the earlier remains beneath it.

The main features discovered in the earlier excavation in this area were the remains of a substantial timber floor with the remains of fallen uprights lying across the top of it. Associated with the floor to the E was evidence, in the form of burnt timbers and many pieces of charcoal, of the collapsed hearth that had been situated on the floor before its collapse. This area of the site looked like a well-built floor that had been built over the top of a collapsed earlier structure. Whether the earlier collapse was from an earlier floor or some other form of structure was one of the questions that required answering.

The floor, associated overlying timbers and large stones were removed. Beneath the floor was a layer of compacted vegetation, including bracken and a broad range of habitation material such as seeds, nuts, small wooden tapers and fragments of bone. They were all typical of material that would have been crushed down through gaps in the floor during occupation of that part of the site.

Within and under this material were a number of large structural timbers on a different alignment, basically N-S, to the overlying floor timbers. In the N side of the trench a row of piles was uncovered, aligned roughly NW-SE. They were of particular importance as some of them still had the remnants of broken tops attached to them. They clearly represent a phase of occupation that was covered over by the later floor after collapse had taken place. It is unusual at Oakbank to find preserved broken parts of piles as, so far, the upper parts of piles have usually been eroded down to a flat plane, or a conical point in the case of oaks which are much harder than alder. These piles are embedded in the usual organic matrix seen elsewhere.

Close to these piles, two large oak timbers were exposed, aligned E-W. They show signs of erosion as if by use during habitation.

The finds discovered during the 2003 season are typical of the range of objects discovered in most parts of the site, including tapers, points, woodchips and similar material.

The excavations in Area C2 brought to light important evidence for elements of the structure of the crannog from an earlier period than the upper floor. That evidence shows a clear break between one phase of occupation and the next, although it is likely that the builders of the upper floor were the same people who had occupied the structure represented by the underlying sub-floor remains and the broken piles. This is suggested by the lack of any naturally deposited material, such as silt or stone debris, between the two layers. The earlier evidence shows a substantial deposit of stones, particularly in the SE corner of the trench. They may have been sufficient to have made it difficult, or impossible, to effectively drive new piles through them and the collapsed structural elements. Substantial numbers of pile stumps situated 3-5m to the E, in Area C1, suggests that they were driven in further out than the earlier piles to clear the deposit of stones and collapsed timbers, effectively enlarging the crannog area.

The floor remains removed at the start of the 2003 excavation season were built out over the top of the earlier alignment of piles and would have meant occupation of a newer area of platform. It is likely that the stones in the earlier deposits were put there deliberately to help give support to piles driven into the substantial depth of the organic matrix. This would not only have been a common feature in later phases of occupation on crannogs, but would, in some cases at least, have been the reason why sites eventually were abandoned.

Excavation on shore

In spring 2003, geophysical survey was carried out in the field next to Oakbank Crannog to see if there was evidence of structures or features that might have been associated with the crannog occupation. Resistivity survey indicated a circular feature in the NW corner of the field, and a broad linear feature running N-S across the field. A small trench, 2 x 1m, was cut across the edge of the latter feature to establish its structure. A large, flat stone slab edged with smaller rubble was noted in the bottom of the trench. It may represent the edge of a stone causeway or path, but more extensive work is required to verify this.

Archive to be deposited in the NMRS.

Sponsor: Scottish Trust for Underwater Archaeology.

N Dixon 2003

Underwater excavation NN 726 442 Research excavations continued at the Early Iron Age site of Oakbank Crannog (NN74SW 16) off the village of Fearnan, Loch Tay. The 2005 excavation focused on the outer perimeter of the 2600-year-old site, where fresh evidence of the early construction sequence has been uncovered. The site was lived in periodically for perhaps more than 200 years, and several phases of building and repair have been identified (DES 2003, 108-10).

Evidence has been uncovered showing a period of collapse when the eastern part of the house and outer walkway gave way and sagged, breaking up that part of the house. Timbers, once upright, have been found all lying in the same direction, some still preserved with the splintered remains from impact. Discovery of several jointed timbers is also providing further insight into construction techniques used by the ancient crannog builders. Following the collapse, repairs were made and the presence of deep, well-preserved layers of wood chips and debris demonstrate new construction. A ring of new uprights was put in to support that part of the house, and a new section of outer walkway was created. Judging from the depth of plant and animal waste, occupation of the house resumed for many years afterwards.

Most of the artefacts discovered so far relate to woodworking and building, including wooden pegs, twisted hazel hurdle fragments, and lengths of rope made from twisted willow or hazel. Several pine tapers were also found.

Archive to be deposited in NMRS.

N Dixon 2005.