Font Size

100% 150% 200%

Background Colour

Default Contrast
Close Reset


Broch (Iron Age), Souterrain (Prehistoric)

Site Name Bu

Classification Broch (Iron Age), Souterrain (Prehistoric)

Alternative Name(s) Navershaw; Naversough

Canmore ID 1483

Site Number HY20NE 11

NGR HY 26967 09348

Datum OSGB36 - NGR

C14 Radiocarbon Dating


Ordnance Survey licence number 100057073. All rights reserved.
Canmore Disclaimer. © Copyright and database right 2021.

Toggle Aerial | View on large map

Digital Images

First 100 images shown. See the Collections panel (below) for a link to all digital images.

Administrative Areas

  • Council Orkney Islands
  • Parish Stromness
  • Former Region Orkney Islands Area
  • Former District Orkney
  • Former County Orkney

Archaeology Notes

HY20NE 11 2697 0933

There is a cairn, measuring 83ft from N-S by 60ft.from E to W, with a maximum height of 6ft, on the summit of a knoll midway between Navershaw and Bu Point.

RCAHMS 1946. Visited, 5th August, 1929

This grass covered cairn at HY 2697 0933 now measures 28.0 metres E to W by 26.0 metres N to S by 2.0 metres high. The cairn's shape has been greatly altered by a gun emplacement which was built into the top during the 2nd World War. There are no records of any finds associated with the cairn.(Information from Mr J W Rendall, Bu' of Cairston, Cairston, Stromness.)

Surveyed at 1/2500.

Visited by OS(RD) 14 September 1964

Excavation in 1978 revealed that this was not a cairn but a broch with asssociated structures and a souterrain Pre-broch occupation consisted of flatstone rubble and ploughing on a low knoll of boulder clay.

The broch had an internal diamter of 9.1m, the wall being solid based c.5.2m wide and preserved to a height of 1.5m. The entrance was in the ESE, c.3.7m long and c1m wide. Guard cells were not noted. The interior consisted of a central hearth and outer perimeter of rooms divided by socketed upright flagged partitions. In the centre, covered by the hearth and service area, there were post-holes. Middens contemporary with the broch were identified to the E and W; to the E there was a double-faced wall parallel to the broch wall, and to the W a midden was overlain by random flagging.

When the broch was abandoned a souterrain was built into the outside of the broch wall on the E side next to the entrance passage which was restored and blocked off. Seven pillars 1.06m high were erected around the wall of the souterrain to support a stone roof. The entrance was in the S where it joined the modified broch entrance. 2 small buildings were possibly constructed in this phase to the E and N of the broch. In the final phase, collapse of the broch filled the interior and souterrain with rubble; in the latter 2 human skeletons were found.

The mound and structures have been totally removed.

J Hedges and B Smith 1979


Field Visit (14 September 1964)

This grass covered cairn at HY 2697 0933 now measures 28.0 metres E to W by 26.0 metres N to S by 2.0 metres high. The cairn's shape has been greatly altered by a gun emplacement which was built into the top during the 2nd World War. There are no records of any finds associated with the cairn.(Information from Mr J W Rendall, Bu' of Cairston, Cairston, Stromness.)

Surveyed at 1/2500.

Visited by OS(RD) 14 September 1964

Publication Account (2002)

HY20 4 BU ('Navershaw' or 'Navershough')


The site, standing on flat farmland, was diagnosed as a cairn by the Commission [2] but in 1978 rescue excavations at short notice, and with limited time, proved it to be a roundhouse the wall of which was subsequently thickened to give it the appearance of a broch-like building [3, 4].

The main building is categorised as “broch-like” here, like Dun Cuier on Barra (NF60 3), because -- while having the size and proportions of one of those buildings -- there is no evidence that it ever had any of the features of the hollow-walled construction technique so characteristic of the dry stone towers. This is important because of the far-reaching claims that were made, soon after the excavation, about the site being an early Iron Age broch, much older than any other known [4]. A detailed report on the discoveries is available so only a summary is given here [5]. It is worth noting that only the interior of the building was completely excavated; four radial trenches across the roundhouse wall exposed the outer face of the wall, the entrance passage and some external buildings beyond the entrance [5, fig. 1.2].

The structure of the roundhouse.

Several phases in the history of the site were detected, the main building being put up in Phase 2, after a period of unknown length (Phase 1) when the ground had been ploughed. At first this was a circular, solid-walled dry stone building preserved only up to 1.5 m and enclosing an approximately circular area measuring 10.1 m E-W and 9.0 m N-S at ground level. the primary wall. Unfortunately a somewhat misleading impression is given at the start of the account that the building was a single-period structure; the overall diameters corresponding to the internal ones mentioned are given as 20.5 and 19.5 m and it is stated that the wall "is one of the thickest on record" for Orkney brochs [5, 10].

However there are in fact three 'component parts' of the wall the thicknesses of which are given [5, Table 1.1] without much further comment. We have to turn to the fold-out cross sections [5, figs. 1.3 and 1.4] to see in detail how the roundhouse wall was constructed.

The north wall on the N-S section is very revealing in that it shows rather clearly how the most massive part of the rubble-cored wall, with the heaviest facing stones, is the central section with a thickness at this point of just about 3.0 m at its base. The outer face of this has tipped outwards markedly here and the very much looser wall facing outside this -- with a core of small stones instead of massive blocks and an outer face of quite modest slabs (1.9 m thick at base) -- looks as if it has been added to stabilise the main wall. This seems to be confirmed by the fact that the outer face of this is still upright whereas, as noted, that of the central section tilts sharply outwards.

By contrast the relatively thin inner addition of masonry -- some 80 cm thick -- rests against an upright wall face and seems stable (its primary nature is confirmed by evidence from the entrance passage, below). According to the section all three walls rest on the old ground surface, not on any midden material, so the outer two were probably added early in the history of the building. Indeed the sections also show that the paved areas of the internal floor run up to the inner added masonry and not under it, so that the additions may well have been made during construction, or immediately afterwards.

The south wall on this section has hardly collapsed at all but its three components are clearly visible. The E-W section is not so informative in this context, the main wall not having been cut through.

The entrance passage

This was on the ESE side and had been much modified in Phase IIIa, mainly through a contraction of the inner end; the original design is not entirely clear because of incomplete excavation. One must presume that this inner end was at the inside edge of the internal addition of masonry [5, fig. 1.5] and this is what the excavator deduced (below). The inner part of the passage was c. 1.74 m wide and flagged with large paving stones (which seem from the plan to disappear under the contracted south side of the passage at the inner end).

At 2.55 m from that end was a door frame formed of opposed upright slabs set at right angles to the walls; further out the passage was 1.06 m wide. There was what looked like a blocked door to a guard cell immediately inside the door frame on the right (looking in); attempts to trace the plan of the supposed cell to which this led were unsuccessful.

The north face of the passage was the better preserved of the two and no distinction (such as a straight joint) was visible in it between the inner facing and the main wall, suggesting that both belong to the beginning of the building's history. On the other hand the junction with the outer cladding could be seen so the latter clearly was a later addition (as suggested by the partially collapsed part of the main wall described earlier). The original length of the passage would thus have been about 4 m.

The central court

The interior floor was well preserved and thoroughly explored; the building had apparently been abandoned suddenly as “kitchen utensils” were found in situ. Two C-14 dates were obtained for animal bones for the “broch phase” of occupation; they are 490 +/- 65 and 510 +/- 80 b.c. (GU 1152 and 1154) and suggest that Bu was built and occupied in the late Bronze Age or in the early part of the early Iron Age.

A simplified plan of the interior helps to understand the various features [6, fig. 1.10]. A large, D-shaped hearth occupied the centre with a cooking tank sunk into the floor adjoining it; presumably this last was for boiling meat like the larger ones, in special buildings, identified by Hedges on Orkney and dating to the latter part of the Bronze Age (Hedges 197?); however no heat-cracked stone are mentioned here.

Next to the curved or west side of the hearth was what was diagnosed as a curved or crescent-shaped service area with utensils still lying about on the floor. This and the hearth together formed an approximately circular area a little off-centre, and between this and the wall was a broad, ring-shaped paved band divided into irregular radial compartments by stone flags set on edge; there were gaps for doorways between these areas and the one on the north was much larger and occupied about a third of the whole perimeter. This area was paved and may well have been a communal hall.

The central area was separated from the outer “rooms” by a similar row of low flags on edge. The part of this outer area immediately in front of the entrance was interpreted as a sort of vestibule from which one could go either into the middle area, or into the “bent hall” on the right, or into the four smaller radial compartments on the left. The first three of these were mud-floored (in sharp contrast with the rest of the interior) and may have been used for animals.

A number of post-holes were found in those parts of the interior cleared down to subsoil; they were entirely in the central area and were covered by the hearth and by the paving of the “service area”. Their age was not entirely clear but the excavators had the impression that few if any were in use after the stone building had been completed. There is therefore a lack of clear evidence at Bu for a roof and upper storeys supported by timber posts, and this increases the contrast with the hollow-walled brochs.

Many more details of the interior arrangements are in the report [5].

Later occupation

The roundhouse seems to have stood for some time after it was abandoned; a layer of grey silt containing rubble, up to 30 cm thick, was spread across the interior. Rubble from the disintegrating wall fell on that and then, in the following Phase III, another occupation layer accumulated on top. Also at that time a souterrain, or underground chamber, was built into the outside of the wall on the east side, next to the entrance passage which was restored and blocked off [1, 3]; a short passage from the entrance led to the chamber [5, fig. 1.11]. Seven stone pillars supported the stone roof of the oval chamber. No C-14 dates were obtained for objects from Phase III.

The finds are described here in some detail because they are much more numerous than in any other early Iron Age settlement so far discovered in Orkney and, though of limited variety, give quite a detailed picture of the economy and technology of Orkney for the period of about 700-500 BC.

Finds from the pre-roundhouse phases (Ia and Ib)

These were all of stone and included 5 mattocks, 3 rough-outs of ard points, a piece of a saddle quern, 2 pieces of unworked pumice 4 miscellaneous implements and the debris of stone-working.

Finds from the roundhouse (Phase IIa)

Bone objects included 2 pointed implements, 1 made from a sheep ulna, a whale vertebra vessel, a perforated bovine phalanx (sometimes interpreted as a bobbin for winding wool), 2 pieces of perforated whalebone [5, fig. 1.18], a cetacean radius with many cut-marks and the end of a polished and perforated deer metacarpal.

Stone objects included the debris from a manufacturing area inside the house where there were also the tip of a worn ard point, 6 possible mattock fragments, a piece of flagstone pecked to a waist, and several other pieces of chipped flagstone. Elsewhere inside the house there were 3 pieces of saddle querns and 2 rubbers, a broken pivot stone, a mattock blade, a possible rough-out for an ard point, a possible whetstone, 6 pebble pounders or grinders, 9 pieces of pumice (some used as abraders), and a piece of haematite.

Much pottery came from the roundhouse floor and the external midden [5, figs. 1.49-1.51]. The vessels are entirely plain, coil- or ring-built and with no burnishing. The pot shapes are mainly bucket-shaped vessels with near upright rims but many have the shouldered profile a short distance below the rim which is characteristic of early Iron Age pottery at Jarlshof in Shetland (HU00 0) and at other similar early sites in Orkney.

A sliver of iron corrosion was also found.

Finds from post-roundhouse silt (IIb)

The silt was probably sterile so these finds probably belong to Phase IIa or IIIa. There was nothing which does not appear in the other layers apart from two items. One is a curious stone conical object 25 cm in diameter the conical socket in which appears to have been made by grinding [5, fig. 1.35]. The second is a crucible bearing traces of copper.

Finds from Phase IIIa and IIIb [5, fig. 1.14]

Of bone there was a socketed handle made from an antler tine, a fish-tailed long-handled comb (whalebone) and the possible butt of a second one, of antler; a simple pin with a squared-off terminal; 4 pointed implements or awls; a rubbing or polishing implement, and a femur-head button or whorl.

Of stone there was broken pivot stone, part of a saddle quern, and possible gaming counter or blank for a spindle whorl, 2 pounders, 1 chipped stone implement and the butt end of a broken mattock.

Of pottery there was a broken pear-shaped crucible which had contained molten metal including copper [5, fig. 1.47]. There were also a few sherds of pottery which were not dissimilar to those of Phase II; one had a slight shoulder below the rim [5, fig. 1.55].

Metal included 2 lengths of iron corrosion which might be natural.

Animal bones and plant evidence

The largest collection of bones from early Iron Age Orkney has come from Bu but "It is unfortunate and irredeemable that that the bones were identified (by the late Barbara Noddle) without reference to a context and were then destroyed" [5, 89]; however most evidently belonged to the primary phase of the roundhouse and it did prove possible for the excavators to partially reconstruct the proportion of animals found in each phase [5, 123-5].

Sheep, goats and cattle were identified, the latter being in the majority there were a few fish. It was observed that there was a high incidence of death among the new born domestic stock which could be explained in two ways. There may have a been a high natural mortality rate or the young may have been killed (and eaten) so that the mothers' milk would have been available for human consumption. Alternatively fodder for the adults may have been scarce so that the mothers could not support the young. Evidence from South Uist could now mean that the milk hypothesis is most plausible (Craig et al. 2000).

Traces of cultivated cereals such as barley and wheat were identified.

Radiocarbon dates

Four dates were obtained, two for Phase IIa, one for IIb and one for IIIa. Because of the time span into which the site falls the dates cannot be calibrated so precisely as, for example, those from the later broch period.

The first two were 2470 +/- 95 bp (GU 1228, charcoal) and 2460 +/- 80 bp (GU 1154, bone); when calibrated they suggest a time span of about 800-450 BC at a 68% degree of confidence (the first date is slightly more uncertain).

Phase IIb is dated to 2440 +/- 65 bp (GU 1152, bone), or about 770 - 450 at 68% confidence.

Phase IIIa is dated to 2545 +/- 65 bp (GU 1153, bone), or about 860-620 BC at 68%.

Sources: 1. OS card HY 20 NE 11: 2. RCAHMS 1946, 2, no. 927, 325-6: 3. J. Hedges and B. Smith in Discovery and Excavation Scotland, 1979, 25-6: 4. Hedges and Bell 1980: 4. Hedges 1983: 5. Hedges et al. 1987, 1.

E W MacKie 2002

Orkney Smr Note

On the summit of a knoll midway between Naver Shaw and Bu Point, about 250 yards SW of Bu is a cairn measuring 83ft from N to S by 60ft from E to W, with a maximum height of 6ft. [R1]

Site was first seen by Hedges on Thurs 25th may 1978, site was pointed out by Mr Bertie Reid, Howe, Stromness, as a stony bump in a field on the W side of Bay of Ireland. The mound, was described in RC and OS as a cairn. On the basis of its shape, low profile and siting it was suspected to be a burial mound.

Full description and Plans of Structure to be found in J W Hedges book. [R2]

Information from Orkney SMR [n.d.]


MyCanmore Image Contributions

Contribute an Image

MyCanmore Text Contributions