Applecross, Wester Ross
Applecross, Wester Ross
Ref: 59464.01
March 2006
Applecross Broch,
Applecross, Wester Ross, Highlands
Prepared for:
Videotext Communications
49 Goldhawk Road
LONDON
SW1 8QP
by
Wessex Archaeology
Portway House
Old Sarum Park
SALISBURY
SP4 6EB
May 2006
Contents
Summary
Acknowledgements
1 Introduction .........................................................................................................1
1.1 The site.........................................................................................................1
1.2 Archaeological and historical background...................................................2
1.3 Background to brochs ..................................................................................4
2 Methods ................................................................................................................4
2.1 Aims and objectives .....................................................................................4
2.2 Fieldwork methods.......................................................................................5
3 Results...................................................................................................................6
3.1 Introduction..................................................................................................6
3.2 Geophysical survey ......................................................................................6
3.3 Archaeological evaluation............................................................................7
4 Finds....................................................................................................................15
4.2 Pottery ........................................................................................................15
4.3 Ceramic building material and fired clay...................................................15
4.4 Stone...........................................................................................................15
4.5 Glass...........................................................................................................16
4.6 Metalwork ..................................................................................................16
4.7 Slag.............................................................................................................16
4.8 Worked Bone .............................................................................................16
4.9 Animal Bone ..............................................................................................16
4.10 Marine Shell...............................................................................................17
5 Palaeoenvironmental evidence .........................................................................18
5.1 Charred Plant Remains...............................................................................18
5.2 Charcoal .....................................................................................................18
5.3 Mollusca.....................................................................................................19
5.4 Conclusions and Potential..........................................................................19
6 Conclusions ........................................................................................................19
6.1 Introduction................................................................................................19
6.2 Typology ....................................................................................................20
6.3 Chronology.................................................................................................20
6.4 Function of brochs .....................................................................................21
6.5 History of Inhabitation ...............................................................................22
6.6 Conclusions................................................................................................22
7 Recommendations..............................................................................................23
i
8 Archive................................................................................................................24
References....................................................................................................................26
TABLES
FIGURES
PLATES
ii
Applecross Broch,
Applecross, Wester Ross, Highlands
Summary
The main aims of the project were to determine whether the rubble remains situated
on a sandstone outcrop were indeed remnants of an Iron Age broch structure, and if so
to define some of its key characteristics, determine its state of preservation, and date it
more precisely within the Iron Age. Evidence suggesting that this may be a broch site
includes vague documentary references to a stone fort from the 19th century as well as
the presence of a large ‘kerbstone’ which protrudes through the grass on the south-
east of the mound.
Other aims of this project included investigating the broader context of the possible
broch, including the remains of a putative prehistoric stone circle within the campsite,
and traces of walling suggesting the presence of two rectilinear structures, possibly
later buildings, to the north-west of the site.
Eight trenches were opened by hand at various locations across this site. Three
trenches targeted on the possible broch revealed foundations of two concentric
drystone walls with the space between them forming an intra-mural gallery. Evidence
for a flag-stone spiral straircase and internal entrance was also revealed, confirming
that the building is a ground-galleried broch. The internal stairway implies that the
broch had at least two floors. No evidence for remodelling or later alterations to the
broch was identified, but only a small extent of the broch was exposed and extensive
stone robbing may have removed later features.
One of the trenches outside the broch uncovered part of a midden relating to post-
broch activity, suggesting that the site had been subject to reuse in later periods. A
second trench outside the broch provided evidence for a possible causeway that may
have provided extra protection to the broch’s occupants.
Further trenches opened to the north of the broch were targeted upon geophysical
anomalies thought to be indicative of a wider broch settlement, including a possible
wheelhouse. One trench produced only modern disturbance, while a second contained
rubble collapse that may suggest the presence of a structure in the vicinity.
The eighth trench was opened within the putative stone circle, c. 200m to the south-
west of the broch, and was able to prove that the orthostats forming this rough circular
pattern were in fact natural glacial erratics, as no stone holes were identified.
iii
Acknowledgements
Excavation was undertaken by Time Team’s retained archaeologists, Phil Harding (of
Wessex Archaeology), Kerry Ely, Raksha Dave, Brigid Gallagher, Ian Powlesland
and Matt Williams, helped by a number of local archaeologists (Dougie Gordon, Alan
Matthews, Mike McMillan, Jenny Robertson, Harry Robinson, John Welsh and John
Wood) and volunteers (including Nick and Clive Goldthorpe). Catriona Gibson and
Steve Thompson are grateful to all the archaeologists and volunteers who assisted in
the recording of archaeological remains and reinstatement of the trenches.
The owners of the Lochcarron Hotel and Applecross Campsite also made the stay
most pleasant.
iv
Applecross Broch,
Applecross, Wester Ross, Highlands
1 INTRODUCTION
1.1.2 Broch excavations are only rarely carried out; however, several well-known
sites (e.g. Howe, Old Scatness) have provided a glimpse of an increasingly
well-understood type of site. Traditionally, brochs have been seen as rather
mysterious sites and assumed to have defensive role in the Iron Age
landscape. More recent excavations have done much to increase our
understanding of these complex sites (e.g. Armit 1996; 2002; 2003; Harding
1984, 2006).
1.1.4 Applecross was formerly known as Borrowdale, and this name may reference
the broch site, since it is derived from the Norse ‘borger’, a burg or
stronghold, and ‘dalr’, a dale (Watson 1904). The small village is split into
two: the lower part comprises a terrace of houses, shore street, a pub and
shop. The upper part is a campsite and farm complex.
1.1.5 The alleged broch is located on top of the south-east end of a small rocky
sandstone knoll within Applecross campsite, at an elevation of approximately
1
40m OD (centred on NGR 171183 844331). It is likely that this outcrop was
originally a moraine, related to glacial action in this area.
1.1.6 The site is approximately 200m east of the current coast line (OS Explorer
428, 1:25000, 2002), which forms a raised beach in this area. The proximity
to the sea means that winter temperatures are neither too inclement nor
extreme.
1.1.7 The underlying geology of the site is moraine and a massive exposure of
Torridonian Sandstone (BGS Scotland 1975; Geological Survey 1954). This
is described as fine to medium grained pale red sandstones, associated with
small outcrops of Triassic and Jurassic rocks, including limestone around
Applecross village.
1.1.8 The site forms part of a gently sloping landscape, which is presently covered
by grassland, grazed by sheep and cattle. The soil consists of a sandy loam
overlying a sandy clay loam.
1.2.1 Accounts written in 1792 by the parish minister suggest that the name
Applecross came from the presence of apple trees planted in a cross by an
ancient proprietor, but the more likely derivation is from ‘Aber’ meaning
mouth and ‘Crossan’ (the name of the river which means little cross).
1.2.2 Evidence for human occupation in and around Applecross dates back to the
Early Mesolithic period. The Inner Sound is an area with abundant islands
and it offered rich resources to the Mesolithic settlers, including shelter
(caves, rock shelters) and easy access to the sea and marine resources with its
shallow sandy coasts. Test pits excavated at the rock shelter of Sand in
Applecross in 1999 produced flaked stone tools in addition to large quantities
of well-stratified shell midden material (Hardy and Wickham-Jones 2000)
Radiocarbon dates obtained from bone tools demonstrated that this site was
occupied c. 7500 cal. BC. Two further open air sites were discovered in the
vicinity of Applecross bay, both producing diagnostic Mesolithic stone tools.
1.2.3 An alleged stone circle of Early Bronze Age date has been identified in the
garden adjacent to Applecross caravan site. This is described in the SMR as
‘eight large naturally occurring boulders that lie in a rough circle. Some
stones sit close to the surface and others are deeply embedded’. The plan of
this circle is not particularly convincing (Figure 6), but some of the original
stones could have been disturbed.
1.2.4 Iron Age sites in the vicinity include a possible souterrain (NG74SW0061)
identified by the SMR, located c. 150 m north-west of the broch and possibly
contemporary with it. It exists as a hollow with its western boundary formed
by a ruinous and wooded field wall. Souterrains and brochs are quite
commonly found together.
2
1.2.5 The Applecross broch may be the dun first mentioned by Rev. John.
MacQueen of Lochcarron in 1792. He refers to a castle or circular fort built
in Applecross by Mac Beolan during the turbulent times, and it is likely that
he is referring to the broch. From his description, it consisted of a grass-
covered mound crossed by a field dyke measuring c. 18.5 m in diameter and
c. 0.8m in height. Some traces of walling remained visible in places and
some large base stones were extant on the south-east part of the mound. The
site is marked on the 1st edition OS map of 1874 as a narrow wooded
property (now de-wooded) and appears as wall tumble, which could just refer
to a boundary wall. From surveys undertaken in 1968 and again in 1974
(information from SMR) this site was described as having traces of a
possible outwork across the ridge on the north-west. It was accentuated by
the ruinous overgrown remains of a modern wall on its outer edge.
1.2.6 The broch site in Applecross can be placed into a wider Iron Age setting.
Other Atlantic Iron Age sites in the vicinity include Lag an Duin, at the tip of
Loch Kishorn, which is described as a dun or a broch. It is a circular stone
structure that has been heavily robbed but it is a turf-covered stony mound c.
20m in diameter and 1m high.
1.2.8 Applecross was, and still remains, one of the remotest parts of Scotland.
Until the coast road from Shieldaig was completed in 1975, the Peninsula
was split north and south, and access to Applecross village was either by the
Bealach nam Bo (Pass of the Cattle) or by sea. The latter route is an old
winding road, with some spectacular hairpin bends, used historically as a
drove road to take cattle to the market (see Back Cover). It is the highest
pass in Scotland, with a climb from sea level to 626m a.OD (2053 feet). As a
result, during particularly inclement times of the year, the road is quite
impassable, restricting access to and from Applecross.
1.2.9 In 1850 nearly 3,000 people lived in the thriving crofting and fishing
townships scattered along the west and north coast townships of Applecross.
Now there are less than 300. The clearances were responsible for some of the
depopulation, but lack of local opportunities and work also contributed. The
remains of previously thriving communities can be seen in many areas along
the coast. Crofting, fishing and tourism are now the main sources of
employment for the Applecross population.
3
1.2.10 Very little previous formal archaeological work has taken place on the site.
The site is marked on the OS map as ‘remains of broch’ and exploratory
trenches were undertaken by the children of the site owners (the
Goldthorpes) in the early 1970s. However, no archaeological features or
finds were identified.
1.2.11 In 2003 four small trial pits were excavated in the camp field to the west and
south of the broch remains in advance of electricity cabling for new campsite
lodges. Situated parallel to the large bank and walls on the north side of this
field the test pits did not contain any obvious features (SMR No.
74SW0064).
2 METHODS
2.1.1 A project design for the work was compiled by Videotext Communications
(Videotext Communications 2005), providing full details of the
circumstances and methods of the project, as summarised here.
2.1.2 The main aims of this project were to determine whether the mound in the
campsite represented the remains of a broch and, if so, to determine when it
was built, and for how long it was occupied. Additional aims were to
understand more about the function of the putative broch and whether
evidence concerning social, economic and industrial practices could be
revealed.
2.1.3 It was possible that the present mound reflects a much later phase of
occupation and re-use of an abandoned broch site. Many broch sites have
later activity, particularly during Pictish, Norse and early Christian periods.
4
Given the proximity of this site to an important local monastic centre (St.
Maelrubha’s chapel in Applecross bay is clearly visible from the mound), it
was quite possible that the mound also contained the remains of a hermitage
or ‘high status secular occupation site’ (Ian Armit, pers. comm.).
2.1.4 If the remains of a discernible broch were preserved in the mound, it was
proposed to examine and define specific areas of the structure, including the
walls and entrance way, the interior and possible stairwell. In understanding
the structural composition and associated internal sub-divisions, it should be
possible to determine the typology of a broch more accurately - its internal
structure should confirm whether it belongs to the broch tradition of
roundhouses, or rather is a wheelhouse or dun. It was also hoped that datable
finds would be retrieved from good archaeological contexts associated with
the broch. One of the best sources for material culture would be from
external midden deposits associated with the occupation of the broch.
2.1.5 The project also aimed to establish whether there were other external
features, such as an enclosure wall or ditch, or other structures which were
associated with the broch. Frequently broch sites are later elaborated, through
the construction of other buildings, often with later houses abutting earlier
broch walls. Examples of larger broch settlements include Jarlshof, Mousa,
Gurness and Midhowe (Armit 1996, 122). In this context, it was hoped that
the date and function of the mentioned rectilinear stone features immediately
to the north-west of the proposed broch could be determined, and any later
use of the site characterised.
2.1.6 It was also hoped to establish whether the stones in the garden of Cul-an-Dun
formed part of a stone circle or whether they had been naturally deposited.
2.2.1 A geophysical survey of the site was undertaken by GSB Prospection Ltd,
comprising approximately 0.5ha of resistance survey (using a Geoscan
RM15 meter) and 0.15ha of gradiometer survey (using both Bartington grad
601-2 and Geoscan FM256 instruments). These surveys were concentrated
around the broch and their location is shown on Figure 1.
2.2.2 Eight trenches of varying size were excavated over geophysical anomalies or
upstanding features that suggested the presence of walls, structures and other
anomalies across the broch site and beyond it (Figure 1). All trenches were
deturfed and opened by hand, and ceased at the identification of significant
archaeological deposits. All spoil was scanned by metal detector.
2.2.3 All archaeological deposits were excavated and the deposits and structures
were recorded using Wessex Archaeology’s pro forma record sheets, and
drawn at a scale of 1:20 for plans and 1:10 for sections. A photographic
record was kept of the investigations and of individual features and stone
structures. The trenches were located using a GPS survey system, and the
principal contexts were related to Ordnance Survey datum.
5
2.2.4 The work was carried out from 1st-4th June 2005, following which all
trenches were reinstated using the excavated spoil, and the turf re-laid or
replaced. All artefacts were transported to the offices of Wessex Archaeology
at Salisbury, under an authority to borrow unallocated Treasure Trove for
research purposes, where they were processed and assessed.
3 RESULTS
3.1 Introduction
3.1.1 Details of individual excavated contexts and features, the full geophysical
report (GSB 2005/34) and results of artefact and environmental sample
analyses are retained in the archive. A summary of the results is presented
here.
3.2.1 At the time of survey the rain was very heavy causing problems with the
resistance technique. This problem was overcome by re-surveying the area
when the weather cleared a little.
3.2.2 The geology of the region is complex and this factor can influence the
responses recorded by geophysical techniques particularly where bedrock
lies close to the ground surface. Igneous boulders and cobbles contained
within the morainic material are likely to be highly magnetic and produce
anomalies that mask those produced by archaeological features, or else
generate responses that appear to be anthropogenic in nature thus
complicating interpretation of the data.
3.2.3 The geophysical survey data provided a more detailed picture of the stone
mound. An outline of the broch as a circular structure was clearly identified
as a circle of high resistance in the data. This ring was surrounded by a band
of low readings which may indicate the foundations of the structure. (Figure
1). A linear band of high resistance which cuts through the centre of the
broch was caused by the existing modern wall.
3.2.4 Circular high resistance anomalies c. 35m to the north-west of the broch
suggest a circular structure, roughly 13m in diameter; the shape is suggestive
of a wheelhouse. However an interpretation of this kind must remain
cautious as this response was on the slope of the site and
topographical/geological issues may also account for this anomaly.
3.2.5 A few metres to the north of the broch, geophysical survey was positioned
over the rectilinear earthworks, suggestive of house platforms, but no
evidence was found for internal heaths or similar features. An area of low
resistance, when excavated, proved to a midden. The high moisture content
within this area would be the cause of this response.
3.2.6 Within the flat area of the campsite, the survey data was dominated by
ferrous anomalies, and this was probably caused by tent pegs and other iron
objects within the topsoil.
6
3.2.7 High resistance readings, other than those mentioned above, are likely to be
due to topographical features, trees and shrubs. Areas of low resistance have
an uncertain cause but may represent localised changes in the underlying
soils and geology.
3.3.1 Eight trenches were opened, all by hand. Five of these were concentrated on
or just outside the broch structure, and one of the trenches over the broch was
extended to investigate the rectilinear structures to the north-west. Two
trenches were opened over a geophysical anomaly that suggested the
presence of a further structural complex to the north-west of the broch, and
one was opened adjacent to one of the stones within the dubious stone circle.
Two of the trenches lacked archaeological deposits, features and finds.
Archaeological deposits in the remaining six trenches were encountered
almost immediately below the turf.
Trench 1 (Figure 2)
3.3.2 Trench 1 was initially opened as a long narrow trench (9m long and 2m
wide), but later extended in its northern section to a width of 5.5m (Figure 2).
It was placed across the geophysical anomaly thought to represent the
circular broch wall. The trench was originally orientated north-east - south-
west and opened over the south-western quadrant of the structure. It was
hoped that the full extent of the wall (both internal and external walls and
gallery) would be revealed. The first deposit encountered after turf and
topsoil was removed was a relatively thick rubble layer (103 and 106),
relating to broch wall collapse and robbing. Deposit 103 was the number
assigned to wall collapse derived from the internal broch area (including over
the walls) while 106 related to rubble collapse lying outside and downslope
of the broch.
3.3.3 Stone robbing activity was indicated by a rectangular cut (118) filled with
smallish fragments of stone chippings (107). This later disturbance was
against the outer face of the external broch wall 104. It was apparent that
most of the stone that once comprised the broch walls had subsequently been
robbed out since surprisingly little in the way of rubble had to be cleared
away before evidence of wall foundations were revealed.
3.3.4 Evidence for the external broch wall was identified when elements of stone
facing (104) were exposed in the southern half of the trench and, after rubble
in the rest of the trench was removed, further traces of curved stone walls
were revealed. Although robbing blurred clear identification, it was possible
to identify an external and internal broch wall, separated by a gallery.
3.3.5 The external broch wall (Group 121) was formed by a curving wall faced on
both sides with dressed sandstone, with a robbed out rubble wall core (105).
This external wall was c. 1.6-1.7m wide and it was separated from the
internal wall (Group 124) by c. 2.1m which formed the gallery between the
two walls (116). Only two courses of the foundations of the external wall
survived, to a maximum depth of 0.30m.
7
3.3.6 The internal wall was less well defined, with no obvious faced edges defined.
It was probably of a similar construction to wall Group 121, but had been
subject to more robbing. Its external face was poorly defined by 123, while
the interior face (120) was clearer and represented by a line of stones, some
of them dressed. The internal rubble core was represented by 125 and the
overall wall was 1.5m in width, giving a total thickness of 4.8-4.9m to the
double broch wall.
3.3.7 A doorway (Group 117) was identified in the south-western side of this
trench (Figure 2, Plate 1). This represents an internal entrance within the
inner broch wall and suggests an access point leading to the gallery on the
south-western side. The doorway was 1.10m wide and c. 1.5m in length. It
was defined by a stone edging on the southern side (114) and a 0.5m wide
single course of dressed sandstone stones on the northern side (110). Once
the fill of the doorway had been removed an area of flat stones (115) was
exposed, possible representing a paved threshold that had been partially
robbed.
3.3.8 Although also heavily robbed, evidence for an internal staircase (113) was
identified, represented by at least two stone slabs partially overlying each
other within the gallery, implying the presence of at least one upper storey
(Figure 2, Plate 2). These stones were probably originally part of a much
more substantial spiral staircase, a common architectural feature of brochs.
3.3.9 A midden-like deposit (111) filled the interior of the broch. This dark,
organic, shell-rich deposit may have been derived from Iron Age midden
material, or may relate to later reuse of the broch. After brochs fell out of
use, they were frequently infilled with contemporary Iron Age midden
deposits as part of the abandonment process (Ian Armit pers. comm.). Some
of this midden material was mixed up with the broch wall rubble layer
downslope of the structure in this trench and in Trench 3 (see below). This
evidence may imply it was associated directly with Iron Age abandonment,
and washed out gradually as the broch collapsed.
Interpretation
3.3.10 The stone walls exposed in this trench confirmed that the structural remains
were indeed those of a broch, and more precisely a ground-galleried broch.
This type of construction is the most common form for this area of Scotland,
as opposed to the solid-based type of broch which tends to be restricted to the
Outer Hebrides (Harding 1984). Although ground-galleried forms tend to be
more unstable than solid-based types, they are designed to increase stability
on uneven ground. This may well be the case for the Applecross broch, since
it was constructed on top of an irregular rocky knoll.
3.3.11 Two important architectural details were also revealed in this trench. The
first was an internal doorway, implying access into the gallery from the
south-west. However, only the internal broch wall was exposed within the
trench where this access point was identified, which made it imperative to
examine the external wall at this location in order to clarify whether the gap
continued through the structure (see below, Trench 8). The second feature
revealed was the stairway, implying that this broch was not single-storey.
8
3.3.12 It is clear that this broch had suffered heavily from later robbing, and only
the basal foundations remained. It is likely that most of this robbing was of
relatively recent date, since 19th century accounts of Applecross mention a
visible and upstanding ‘dun’ monument (Rev. J. MacQueen, 1792). Much of
the stone may have been reused during the expansion and development of the
Applecross estate, which involved the construction of a large number of field
boundaries and associated estate buildings.
Trench 2 (Figure 3)
3.3.13 Trench 2, measuring 4.5m by 3.4m, was excavated to the east of the actual
line of the broch wall as indicated by the geophysical survey. This trench was
targeted on a large stone jutting out of the turf, thought to resemble a
kerbstone that might have formed part of the broch or an associated structure.
Since the trench was located immediately adjacent to a large tree,
biorturbation was evident in this trench, and extensive root systems were
encountered. On excavation, the possible kerbstone turned out to be a large
boulder, and formed part of a roughly made stone wall or revetment (203;
Figure 3). This was exposed in the southern part of the trench and was at
least 1.2m wide, and 0.6m deep. It was orientated north-west – south-east
through the trench, and some elements of tumble (202) derived from the
broch walls further upslope had later become incorporated within it.
Furthermore, 203 had been subject to some collapse and slippage. Smaller
sandstone chips had been inserted into the voids between the larger boulders
in a drystone fashion (Figure 3, Plate 6).
3.3.14 On the eastern side of the trench a possible linear cut (205) was identified,
thought to represent a ditch or robbed out wall. On excavation, however, it
was clear that this was a natural or geological feature as its edges were
irregular on one side and practically impossible to discern on the other. This
feature did contain limestone fragments (non-local) and a large sawn
fragment of animal bone. Two sherds of modern pottery were also retrieved
from this feature.
Interpretation
3.3.15 This trench was external to the broch and it is possible that the roughly
constructed rubble wall/revetment revealed may originally have formed part
of a raised causeway leading to the broch. Although not actually exposed
during this evaluation, it is likely that the external entrance 1 into the broch
was located on the south-east or eastern side (e.g. Parker Pearson et al. 1996)
and, since this side of the broch was above a steep slope, access may have
been improved through artificial landscaping. It was suggested on site (Neil
Fojut pers. comm.) that the contours and topography of the site would
support the idea that the causeway may have led from the south and then
curved sharply to the west towards the entrance (see Figure 8). This
arrangement would have provided extra security and protection for the
occupants of the broch, in the event of an attack.
1
Note that the entrance identified in Trench 1 was a between-wall entrance providing access to the
internal gallery and the stairwell only. Trench 8 confirmed that this access point did not run all the way
through to the external wall.
9
Trench 3 (Figure 4)
3.3.16 Trench 3 was excavated on the northern side of the broch, and was orientated
north-west – south-east across the broch wall as identified from the
geophysical survey. The trench was also targeted to investigate the rectilinear
structures situated to the north-west of the broch. It was 18.5m long and
varied from 0.7m to 1.8m in width. For spatial and chronological control,
different context numbers were assigned to the various dumps of tumble that
may represent different collapse events. Immediately below the topsoil, a
rubble collapse horizon was revealed (302), which covered the entire trench
and was up to 0.35m deep. The collapse was not deeper at the bottom of the
hillslope, implying that this rubble had also been robbed. Much of the
northern two-thirds of the trench was covered by another lower spread of
rubble (304) in an orange brown silty loam matrix. In the far north-western
part of the trench a separate collapse deposit was identified (306) and this
may have been derived from the rectangular structure (314) to its north. Once
the majority of the different tumble episodes had been removed, it was
possible to discern structural elements relating to the broch in conjunction
with other, possibly later features.
3.3.17 The latest structural element identified relates to a rectilinear structure c. 14m
to the north-west of the broch. This structure still survived as a slightly
visible ‘bump’ under turf. Excavation demonstrated that it was a rather
insubstantial stone structure. Only a single course of rough sandstone aligned
roughly north-south was identified (314) with a maximum width of 0.4m and
depth of 0.3m. These footings had been constructed directly on top of a thick
hillwash layer (320) that had accrued long after the broch had been
abandoned. This same colluvial lens also sealed the collapse deposits derived
from the broch, including broch tumble 318.
3.3.18 The latest archaeological layer exposed relating to broch activity (and partly
excavated) was a greyish brown charcoal-rich silty clay (303) up to 0.18m
thick, covering most of the trench and lying beneath the top layer of rubble,
and sealing the foundations of the broch structure. It appears to represent
later slopewash of midden material used to infill the interior of the broch
after it was abandoned, and was probably contemporary with deposit 111 in
Trench 1. Seven sherds of Iron Age pottery were retrieved from this context,
implying the midden material was related to the immediate abandonment of
the broch.
3.3.19 In the southern part of the trench underneath rubble collapse, a black
charcoal rich lens (315) was identified in association with a fire installation
of some sort (possible kiln/hearth). Only a very small part of this was
excavated and thus it is difficult to know what it represents, but it is possible
that a small part of a flue was revealed, represented by cut 317 (Figure 4,
Plate 8). The feature was filled with a fine ashy deposit that also contained
large quantities of charcoal and slag (316). Two sherds of Iron Age pottery
came from deposit 315, while a further sherd was retrieved from 316. The
feature had been cut into the natural and the natural sandy clay in this part of
the trench was bright pink, implying in-situ burning. A large thin flat stone
10
(321) had been partially placed over this feature and may have been related
to it.
3.3.20 The external broch wall (Group 322) was better preserved in this trench than
Trench 1, and was represented by two separate courses of dressed stones
(307 and 308) defining the curved edges, with a rubble core (309). The wall
had a foundation of small angular stone chippings laid on top of the natural
geology, on top of which a single course of sandstone blocks was preserved
(Figure 4, Plate 7), surviving to a height of 0.6m. Wall 322 was 1.5m thick
and was exposed for a length of 1.85m.
3.3.21 The gallery between this wall and the internal broch wall was 1.5m wide and
had been filled with a dump of dark brown midden material (319), again
presumably after the broch had been abandoned. A further sherd of Iron Age
pottery was retrieved from this deposit. The full extent of the internal broch
wall was not exposed, owing to its proximity to the overhead power cables.
However, it was at least 1.5m wide, and again was represented by an external
edging of dressed stone (311) with rubble core (312).
3.3.22 The dimensions of the broch walls in this trench are very similar to those
defined in Trench 1 and, if its full extent had been exposed, the overall broch
double wall would have measured between 4.8m and 4.9m.
Interpretation
3.3.23 The broch wall was slightly better preserved in this trench. By combining the
evidence from Trench 1 and Trench 3, and working out the curvature of the
broch, it is can be determined that the external diameter of the broch was c.
16.8m, while its internal diameter was c. 11.9m (Figure 8). Furthermore,
datable pottery was retrieved from good contexts in Trench 3 and suggests an
Iron Age date for at least some of the occupation of the broch, as well as
implying that the infilling of the broch with midden material on abandonment
also relates to this period. This trench was also able to show that the broch
structure had been subject to a fairly long history, with later activity
represented in the form of the flimsy rectangular structure at the northern
edge of the trench. This rather insubstantial building probably formed part of
an 18th/19th century outhouse or shed, related to the expansion of the
Applecross estate. The small quantity of slag associated with the kiln/hearth
structure is undiagnostic and may derive from iron smelting or smithing,
possibly of Iron Age date as suggested by the presence of Iron Age pottery
within this feature. However, the larger quantity from layer 303 overlying
the foundations of the broch is very dense and has a flow structure
characteristic of tap slag, indicating iron smelting. A post-Iron Age date is
considered most likely for this material (Phil Andrews pers. comm).
Evidence of grasses and sedges (see below, Charred Plant Remains) and
quantities of wood charcoal (see below, Charcoal) may represent the remains
of fuel used for industrial activities.
Trench 4 (Figure 5)
3.3.24 Trench 4, measuring 2m by 1m, was opened c. 10m to the north-east of the
broch on the steep hillslope. It was targeted in order to investigate a
geophysical area of low resistance that suggested the presence of midden
11
material in this zone. Excavation confirmed the existence of a fairly deeply
stratified external midden (c. 1.25m deep), sloping downwards from west to
east. Four major dumping events were identified within this midden,
although there were also a few discrete, thin, laminated deposits within these
horizons.
3.3.25 As soon as the turf was removed, midden deposits were encountered,
implying some erosion of this slope. The latest midden dump (401)
comprised a dark grey black lens, c. 0.35m deep that contained a quantity of
material culture, including a bone toggle and a bone pin, in association with
animal bone, limpet and scallop shells. Beneath this, 402 was a thinner light
grey ashy deposit (0.15m deep) that was very finely sorted and contained no
stones. This deposit was laminated, with alternate fine ashy and charcoal
lenses and probably represents more than one dumping event. Finds included
animal bone, shells (mussel, scallop, oyster, limpet) and undiagnostic iron
slag.
3.3.26 Deposit 403 was a thicker (0.28m) layer that was less humic and more silty.
Again it was slightly laminated with lenses of scallop, oyster and razor
shells, suggesting some discrete dump events. Burnt stone and a worked bone
point were also recovered from this layer. The earliest dumping event
constituting the midden, 404 was a deep deposit (0.45m) of mid brown sandy
silt that contained less material culture and was even less organic in content
than 403. However, some large chunks of charocoal were present in this
layer, as well as large quantities (c. 45%) of small and medium subangular
stones. Beneath 404, a thin interface deposit (405) was recognised between
the midden and the degraded red sandstone natural below.
Interpretation
3.3.27 Middens contain what people throw away, and this ‘rubbish’ can frequently
provide important insights into societies, particularly in relation to economic
issues. A great deal of almost unidentifiable animal bone came from these
midden deposits, reflecting an intensive processing strategy where all animal
products were fully exploited. Furthermore, evidence of industrial practices
and craft working was identified from iron slag and weaving equipment
(bone points and a pin beater). The animal bone assemblage indicates the
exploitation of a range of different meat resources including domesticated
animals (cattle, and pig) as well as wild resources from both the land (deer)
and the sea (seal and fish). The presence of large numbers of shells suggests
that shellfish (limpet, mussel, oyster, scallop) also contributed to the diet.
Cereal remains were less well represented, although cultivated barley grains
were present.
3.3.28 The main question that this trench raises concerns the date of the midden,
and whether it was contemporary with the occupation of the broch or relates
to later reuse (e.g. Norse etc.) of the broch, which is quite common. Two
sets of evidence may be used to support the idea that it is in fact post-Iron
Age.
3.3.29 The first category of evidence comes from the material culture. Although one
sherd of Iron Age pottery was retrieved from the midden, it came from the
12
latest deposit (401) and may have washed in as residual material from the
broch upslope. The worked bone pins and toggle are not diagnostic of
Atlantic Iron Age types in the north-west coast and western islands of
Scotland (Neil Fojut and Andy Heald pers. comm). This may partly be
because Iron Age broch and duns excavated in this zone tend to be associated
with sandier and hence more acidic soils, which are not conducive to bone
survival. The geology of Applecross, however, had a more neutral pH that
might promote bone survival, and the lack of comparable bone pins and tools
from other broch sites may be an artificial bias. Unlike that from Trench 3,
the slag from the midden in Trench 4 is undiagnostic and does not help with
a chronological determination.
3.3.30 The second set of evidence comes from the nature of the midden deposits
themselves. The earliest midden horizon (404) contained large quantities of
sandstone rubble, concentrated particularly on the western upslope side of the
trench (Figure 5, Plate 9). The excavator argued that these stones were very
loose and included some small chips and fragments that might relate to broch
construction. However, on closer inspection, many of these stones are much
larger than chippings and include some large rectangular sandstone blocks
(up to 0.4m in size). These more probably represent a collapse and
destruction horizon, and therefore imply that all of the midden material that
seals this level post-dates the main Iron Age phase of occupation of the
broch. Only a radiocarbon determination could confirm this suggestion.
Trench 5 (Figure 6)
3.3.31 Trench 5 was excavated c. 200m to the south-west of the broch in the garden
of Cul-an-Dun. It was 1m long x 0.75m wide and was opened adjacent to a
stone upright in order to determine whether this orthostat formed part of a
larger Late Neolithic/Early Bronze Age monument. The stone investigated
was situated on the lawn that had been levelled to some extent. No
archaeological deposits or features were identified in this trench, and it was
clear that no stone hole existed that would support the idea that the monolith
might have originally formed part of a stone circle.
3.3.32 Although at least ten other large stones within a c. 30m radius may form part
of this group, the pattern they depict is not particularly convincing (Figure 6)
They form a very irregular circle and most of the stones are not upright.
None of the stones were shaped, moved or modified in any way and the
irregularity of the stone settings would confirm that this stone forms part of a
spread of glacial erratics in the garden behind Cul-an-Dun.
Trench 6
3.3.33 Trench 6, measuring 2m by 1.9m was excavated in order to investigate a
series of circular high resistance anomalies to the north-west of the broch,
roughly 13m in diameter, that appeared to have a structural form; the shape is
suggestive of a wheelhouse (Figure 1). The small test pit, however, only
revealed modern disturbance and no archaeological features or finds were
recovered (Figure 7). Immediately beneath the topsoil a modern pit was
identified (602) that contained a series of in situ burning horizons. The
characteristic orange-pink degraded sandstone natural was encountered c.
0.2m below the topsoil.
13
3.3.34 After this trench had been excavated, it transpired that Nick Goldthorpe had
excavated this pit as a child in 1974, as part of an ‘archaeological
investigation’ and it was subsequently used as a bonfire pit, probably
accounting for the high magnetic response in this particular spot.
Trench 7 (Figure 7)
3.3.35 Trench 7, measuring 1.65m by 1.20m, was opened to investigate part of the
same extensive geophysical anomaly targeted by Trench 6. A rubble horizon
(702) was encountered below the turf and topsoil, covering the entire trench.
It comprised both small and medium rectangular sandstone fragments within
a brown silty loam matrix that was 0.15m deep. This rubble horizon was
sitting immediately above the degraded orange sandy natural (703), and no
structural elements were identified within this trench.
3.3.36 Interpretation is hampered by the small size of this trench. It is clear that the
rubble encountered is derived from some structure in the vicinity, and it is
unlikely that it is has come from the broch. The broch structure is located
over 50m to the south-east and Trench 7 was situated on a level area, not
downslope from the broch, so stones are unlikely to have tumbled in this
direction. The geophysical survey identified anomalies that were initially
thought to represent a wheel-house. However, the circular anomalies may
indicate the presence of small circular stone structures, possibly associated
with the occupation of the broch. It is not uncommon for broch settlements
to spring up, and for a broch to mark the focal point for a wider settlement
community that expands beyond the broch walls. Without further
investigation, no definite conclusions can be offered, but it is clear that the
rubble encountered in this trench must have come from some nearby stone
building that may be of Iron Age date or later.
Trench 8 (Figure 2)
3.3.37 Trench 8, measuring 1.7m by 1.7m, was opened up adjacent to, and to the
south of Trench 1, in order to confirm whether the entrance way identified in
Trench 1 continued through the gallery and the external broch wall.
3.3.38 As with Trenches 1 and 3, the topsoil was mixed with rubble sandstone
collapse (801), but as soon as this was removed, the foundation of the
external broch wall was revealed. This wall (Group 805) had a total width of
1.1m and comprised external and internal dressed faces (803 & 804) with a
central rubble core (802). Like the other segments of walls revealed, it was
clear that 805 had been heavily robbed, but because it was only exposed and
not excavated to any extent, it is unclear how deep the wall foundations
survive.
3.3.39 Despite the small size of the trench, it was possible to note the extent of
curvature of the broch wall by linking 805 with the external wall (Group 121)
revealed in Trench 1 (Figure 8). Furthermore the wall in Trench 8 was
continuous, and the lack of a break implies that the gap identified in Trench 1
does not relate to an entrance but to an internal access point into the gallery
and the stairwell.
14
3.3.40 This configuration has been recorded at numerous other ground galleried
brochs (e.g. Loch na Berie, Lewis, and and Dun Bharabhat, Cnip; Armit
1996, figs. 7.4 and 7.5), and it is likely as in these examples, the entrance into
the broch would have run at right angles to the staircase. In the case of
Applecross, this would have meant an eastern entrance, perhaps not
coincidentally in the direction where the causeway identified in Trench 2 is
located. From excavations of other broch sites (see Armit 1996, 112-3) we
know that brochs only had a single small and narrow entrance, and these are
generally on the eastern side.
4 FINDS
4.1.1 Finds were recovered from four of the eight trenches excavated. The bulk of
the assemblage is assumed to be of Iron Age date, associated with the use of
the broch, although there were a few modern finds.
4.1.2 All finds have been quantified by material type within each context, and
totals by material type and by trench/site area are presented in Table 1.
Subsequent to quantification, all finds have been at least visually scanned in
order to gain an overall idea of the range of types present, their condition,
and their potential date range. Spot dates have been recorded for pottery. All
finds data are currently held on an Access database.
4.1.3 This section presents an overview of the finds assemblage, on which is based
an assessment of the potential of this assemblage to contribute to an
understanding of the site in its local and regional context, with particular
reference to the use of the broch.
4.2 Pottery
4.2.1 Apart from two modern sherds from Trench 2 (context 205), all of the
pottery is of Iron Age date, deriving mainly from Trench 3, with one sherd
from Trench 4. One sherd is in a coarse fabric tempered with rock fragments;
the remaining 12 sherds are in fine sandy fabrics containing occasional
organic inclusions. One sherd carries an applied strip or cordon (context
316).
4.3.1 Seven pieces of modern ceramic building material (tile) came from Trench 2
(topsoil and context 205).
4.4 Stone
4.4.1 One stone object, a trough quern, was recorded in Trench 3 but not removed
from site. Of the remaining 37 pieces of stone collected from the site, none
are obviously either worked or utilised, although amongst the 17 beach
pebbles present one or two show rather ambiguous signs of wear, perhaps
through use as rubbing or hammer stones. Four joining igneous fragments
from Trench 1 (topsoil) have smooth, but not necessarily worked/utilised,
15
surfaces. One piece of roofing slate came from Trench 2 (context 205), in
association with modern pottery.
4.5 Glass
4.5.1 An annular bead in translucent blue glass came from topsoil in Trench 1.
Blue beads of this type have a lengthy currency in the British Isles, from the
6th century BC through the Iron Age, into the Roman period and even
beyond. One or two pre-Roman examples are known from the west coast of
Scotland, and they also appear on native and Roman sites in Scotland after
the conquest (Guido 1978, 66-8).
4.6 Metalwork
4.6.1 Iron objects comprise a nail (402), a knife blade (111) and two sheet
fragments of unknown function (112).
4.7 Slag
4.7.1 A small amount of ironworking slag was recovered, most of which came
from trench 3 (context 303). The latter group includes pieces of very dense
slag, probably deriving from iron smelting. The surface shows clear traces of
a floor structure and is therefore interpreted as tap slag. On this
technological basis, a post-Iron Age date is considered most likely for this
material. The slag associated with the kiln/hearth structure in Trench 3 is
undiagnostic and may derive from either iron smelting or smithing, as may
be the very small quantities from Trench 4.
4.8.1 Five objects of worked bone or antler were recovered – two points (401,
403), plus a fragment possibly from a third point or gouge (402), the tip of a
possible pinbeater (weaving tool: 401), and a short length of antler tine,
possibly an offcut (111).
4.9.1 Some bone was recovered from samples and is included with the hand-
recovered assemblage here. Iron Age ceramics were present in four of the
contexts that contained animal bone, but no other dating evidence was
available at the time of assessment. The majority of bone was from trench 4
(the midden).
4.9.2 Thirty percent of the total 669 bones were recorded as being poorly
preserved, although the majority of these were tiny calcined fragments from
315. The rest of the bone including all that from trench 4 was in fair
condition. Gnawing incidence was low, at 2%, mainly affecting the bone
from trench 4, suggesting that overall scavenging had not affected the
assemblage to a great degree.
4.9.3 Only 17% of fragments were identified to species, and while this is partly
because a large number of undiagnostic fragments were recovered from
16
samples, it also attests to the fragmentary nature of some of the bones from
larger assemblages in trench 4. Loose teeth comprise 18% of all identified
bone, also probably a result of (pre- or post-depositional) breakage of jaws.
4.9.4 Cattle are the most common species and sheep/goat are also well represented
(Table 2), with smaller numbers of pig. Wild animals were relatively
frequently observed, indicating hunting activity, with land mammals mainly
in bone from trenches 1 and 3, and probable seal and fish particularly seen in
trench 4. Red deer (both shed antler and limb bones) and fish were fairly
common. Fish remains were present in large numbers from midden 407 (see
Table 3; sample finds not included in Table 2), and included two charred
otic bulae that would suggest the burning of fish heads. A single fish vertebra
came from kiln/hearth 317, and given the high numbers of roots could be
intrusive. Other wild animals included possible fox and, in 111, a minimum
of three voles.
4.9.6 Butchery marks were observed on 6% of all bones (wild and domestic), and
included chops for portioning the carcass, cuts from dismemberment and
meat stripping, and fractures made on fresh bone, presumably to extract bone
fat. An antler beam in 103 may have been chopped or sawn, and the means of
formation of a strange rounded fracture across an immature cattle femur in
402 is not known.
4.9.7 Of the 13% of bones that were burnt, almost all were from 315 (which also
contained an unburnt pig tooth) and were tiny calcined fragments, perhaps
from a hearth. Calcined sheep sized animal bone fragments were also seen in
403, but here they were in the minority.
4.9.8 It is notable that the bone from trench 4 contained a very high proportion of
unidentified long bone and rib fragments (for instance only one of 75 hand
recovered fragments in 401 was identified) and this may reflect an intensive
processing strategy where all animal products were fully exploited.
4.10.1 Given the location of the site the presence of a variety of marine shell species
is unsurprising; examples of periwinkle, limpet/mussel, oyster, scallop and
razor shells were identified. A sample taken from midden deposit 111
produced large amounts of degraded mussel (Mytilus edulis) shell, and
degraded mussel shell was also recovered from midden 407 (see Table 3,
sample finds not included in Table 2). No other remains of shellfish were
seen in the samples.
17
5 PALAEOENVIRONMENTAL EVIDENCE
5.1.1 While charred plant remains were not abundant, preservation was very good.
Grains of barley were present in six of the seven midden samples (excepting
context 315). In most cases they could be seen to be of hulled barley, the
grain still tightly enclosed in the palea and lemma. Grains of oats (Avena sp.)
were also present in most of the samples, but could be of the cultivated or
wild variety. They were not present in any great quantity within any of the
samples
5.1.2 Alongside barley the other category that was particularly abundant was
hazelnut (Corylus avellana). Fragments were present in all the midden
samples, but absent from kiln/hearth 317.
5.1.4 Other seeds were generally poorly represented within the samples and
included probable weeds, such as annual meadow grass (Poa annua),
goosefoot (Chenopodium sp.) and ribwort plantain (Plantago lanceolata).
5.1.5 Hulled barley is by far the most commonly recorded crop from Iron Age and
Roman Scotland (Dickson and Dickson 2000). It is also recorded from other
mainland Brochs including Fairly Knowe, Stirlingshire (Boyd 1985). While
other crop remains have been noted from Iron Age Scottish sites, e.g. spelt
wheat (Triticum spelta) from Oakbank Crannog (Clapham and Scaife 1988),
and naked barley from Dun Mor Vaul broch on Tiree (Renfrew 1974), no
evidence though was seen for either at Applecross.
5.1.6 Remains of hazelnut are also recorded from the Scottish Iron Age, e.g.
Sheils, Glasgow (Robinson 1983), although perhaps less commonly than its
representation at Applecross. Given the time and expenditure in the
collection of wild foods, such as hazel, it may be that the inhabitants had
greater access to such resources than upon other less high status Scottish
sites.
5.1.7 The remains of sedges in the hearth, given the small amount of wood
charcoal in this sample, might indicate its use for fuel. It is also possible it
may be waste from material collected for use in thatch or basket making, for
example.
5.2 Charcoal
5.2.1 Charcoal was noted from the flots of the bulk samples and is recorded in
Table 3. Hand picked charcoal fragments were also recovered from midden
contexts 403 (9 pieces) and 404 (11 pieces). These same contexts also
produced the highest quantities of wood charcoal along with that from
18
midden layer 315. Most of the charcoal was relatively large pieces with little
visible roundwood.
5.3 Mollusca
5.3.1 Mollusc remains were generally scarce in the samples. A single shell of
Cepaea sp. a catholic species was recovered from midden deposit 111.
5.4.1 The presence of charred grain and charcoal confirms the presence of burning
and domestic activities on site such as parching, processing storage and
consumption. Marine resources (mussel shells and fish bones: see Section 4)
are typical of other similar sites and conform to a recognised pattern of
exploitation of marine resources. The dumps of mussel shell have created
local microenvironments facilitating preservation of fish bone and land
snails.
5.4.2 The charred plant remains can provide evidence for the diet of the inhabitants
of the broch at Applecross and some indication of the local resources (hazel)
and environment.
5.4.3 The charcoal indicates the selection of wood for fuel. Comparison of the
material from middens 315 and 407 could reveal differences in such selection
associated with possible later occupation on the site. The charcoal can also
provide some information about the local woodland environment, and of any
management of that resource.
5.4.4 No further analytical work is proposed on the flots. Full quantification was
carried out from the flots and so there is no further need for analysis of the
samples. It is suggested that the residues be scanned and that the results of
this assessment be written into the full report.
6 CONCLUSIONS
6.1 Introduction
6.1.2 Broch towers have long dominated romantic perceptions of the rough and
rugged highlands and islands of Scotland: massive stone ruins in an isolated
and bleak rocky landscape. However, it is only in recent decades that
modern excavations have allowed us to understand these monumental
drystone towers properly. In the last 25 years brochs have attracted a great
deal of attention, particularly those in the northern and western Isles. Major
excavations of brochs in the western Isles include Dun Bharabhat, Lewis
19
(Harding and Dixon 2000), Cnip, Lewis (Armit 2003), and Bu and Gurness
in Orkney (Hedges 1987).
6.2 Typology
6.2.1 The results of the evaluation were able to determine that the stone structure
excavated at the Applecross campsite was that of a broch and hence one of
the more complex sub-sets of roundhouses. Although only a small part of the
overall structure was exposed, it was clear that the building had been
constructed with two sets of walls, separated by a central space or gallery.
Furthermore, broch interiors often have paved floors, and evidence for such
paving may have been revealed on the surface by the internal doorway,
although this had been partially robbed. Not enough was excavated to allow
an understanding of the nature of the internal structure, such as whether there
were any intra-mural divisions including cells or hearth structures.
6.2.2 Only the foundations of this broch survived, and it is clear that the structure
had been subjected to heavy stone robbing. Although a good portion of it
was still standing in 1800, during the next decade it was demolished and the
stones were used to build park walls and this may be when the broch was
almost entirely robbed out to its foundations.
6.2.3 The severe robbing of stone has meant that the original form and height of
the broch tower is very difficult to determine. However, it is likely that this
building would have been a fairly tall, visually imposing tower, constructed
from two concentric walls tied together at various intervals by stone slabs to
create a series of superimposed galleries. Certainly the presence of an
internal staircase within the walls indicates that the broch had a second
storey. It was probably several metres in height, with the hollow drystone
wall tapering to the top of the structure. Some of the better preserved brochs
such as Caisteal Grugaig, in Totaig, Dun Beag, Skye (Callender 1921) or
Mousa, Shetland (Ritchie 1988) may provide clues concerning its original
construction. These brochs all had ledges or ‘scarcements’ projecting from
the inner wall which could have supported internal timber floors and a
conical thatched roof.
6.2.4 The external diameter of the broch at Applecross was c. 16.8m, while its
internal diameter was c. 11.9m, and its overall wall thickness was 4.9m thick.
These dimensions compare favourably with other brochs from the West
Coast and Inner Isles. For instance, the mean internal diameter of Skye
brochs is c.10.5m, with a wall thickness of c. 3.5m On the central mainland
brochs tend to be larger with a mean internal diameter of c. 11.74 m internal
diameter, but with thicker walls – c. 5.3m (Martlew 1982, 257). Brochs on
the west coast tend to have thinner walls, larger internal diameters, and
ground galleries, while brochs in the north tend to have thicker walls, smaller
internal diameters and solid bases.
6.3 Chronology
6.3.1 The typology of the broch (ground-galleried rather than solid-based) may
help in determining its chronology. Solid-based forms of brochs, such as Dun
20
Telve and Dun Troddan, are seen as chronologically later developments
(Harding 1984, 215). Prior to the 1970s (e.g. MacKie 1965) it was generally
believed that brochs were built by incomers to Scotland, and that they were a
relatively late development dating from c. 100BC until the 3rd and 4th
centuries AD. The construction of broch towers and wheelhouses is generally
thought to date between 400 and 200 BC with brochs continuing in use in the
period 200BC-AD 200 (Parker Pearson et al. 1996, 61). As Armit has
recently stated (1996, 113-114) brochs form part of the continuum of the
tradition of monumental stone architecture of Atlantic Scotland that begins as
early as the Late Bronze Age or c. 700BC. Brochs are effectively the final
culmination of this complex roundhouse tradition, and Armit (ibid., 116)
argues that ultimately, the search for broch origins is futile, since continuity
in development for over a millennium blurs any sharp typo-chronological
changes in the tradition. It is likely that the Applecross broch was
constructed within the later Atlantic Iron Age tradition, after 400BC,
although only a programme of radiocarbon dates from secure contexts from
the broch structure will be able to provide a more precise chronology.
6.4.1 In recent years, archaeologists have begun to minimise the defensive role
played by brochs, and it is true that many brochs are not sited tactically in the
strongest locations (Harding 2006, 74). A broch would certainly provide
more protection than an open settlement but security may have been only one
of the motivations behind broch construction.
6.4.2 Unlike Iron Age hillforts, however, brochs were not large communal centres
but rather were the product of small and more dispersed communities, and as
Harding suggests (2006, 74-5) may have been the fortified homesteads of the
social or warrior elite. Perhaps the small entrances to the brochs were also
associated with ideas of control of access, perhaps to materials (foodstuffs
etc.) stored within the broch.
6.4.3 However, there is also evidence to support the idea that brochs were not so
much the residences of the warrior aristocracy, but rather functioned as
strongholds and temporary refuges for dispersed populations during times of
insurrection. According to Armit (1990b, 1996), in the Western Isles and
mainland, settlement gradually shifted from the interior to the coastal zone
during the Iron Age, mainly as a result of environmental degradation
(deforestation, soil exhaustion etc.). The adoption of Atlantic roundhouses
therefore might be a reaction to increasing conflict over access to the ever-
decreasing areas of usable land in the later 1st millennium BC (Armit 2003,
24).
6.4.4 Although the evaluation at Applecross was too limited to determine the
various functions of this broch, it is possible that it played a defensive role
within the context of seasonal raiding or limited short-term conflicts in the
wider peninsula (e.g. Blythe 2005). There is indirect evidence to support the
argument that this broch may have borne witness to times of economic stress
and competition for resources. Firstly, it is located in a defensive position, on
an elevated rocky sandstone outcrop, providing commanding views over the
21
coast (the Inner Sound) to the north and west. Secondly, the presence of a
path or causeway, which has been documented at other sites such as at
Gurness (Armit 1990c, 200), may have had defensive connotations. At
Gurness, the path acted as a means to control access into the broch tower,
and it is possible that a similar role was attached to the causeway leading to
the Applecross broch. It may have accentuated the importance of the broch
tower and controlled the movement of people, including approaching
enemies.
6.5.1 The evaluation sought to identify when the broch was established and
whether there was more than one phase of construction or use. The trenches
opened did not suggest that there was much remodelling to the broch, which
should be the case if the site had been subject to periods of reuse after the
Iron Age. However, evidence of later modifications may also have been lost
as a result of the extensive nature of the stone robbing with only the base
foundations left. The nature of the external midden deposits would suggest
that the broch had been reoccupied after the Iron Age, particularly since the
midden layers were sealing elements of broch collapse. Furthermore, the
nature of some of the slag found in Trench 3 to the north-west of the broch
indicates iron-smelting and that the slag was tapped from the furnace. This
technology would suggest a post-Iron Age date.
6.6 Conclusions
22
Evidence for post-broch industrial and textile manufacturing processes was
also indicated by the presence of slag in Trench 3 and weaving tools from the
midden.
7 RECOMMENDATIONS
7.1.1 There is no immediate threat to the broch at present, and it is not affected by
ploughing, animals or development work from the campsite. However any
remedial works to the overhead power cables should bear in mind that they
have been driven through an Iron Age structure. Furthermore, any future
developments to the campsite in the direction of the broch, or its possible
wider settlement, or the insertion of any new services for the campsite,
should not be undertaken without consultation with the regional
archaeologist first to ensure no damage to the structure.
7.1.2 It is likely that the internal midden deposits infilling the broch structure are
contemporary with the Iron Age abandonment, since some of this material
had washed downslope at the same time as the elements of broch wall
collapse. It is proposed that a programme of radiocarbon dating from
material, if suitable, is carried out from both internal midden deposits and
from the external midden as well. These determinations would allow the
dating of the abandonment of the broch (and provide a terminus ante quem
for its occuapation), as well as allowing a more precise chronology for the
later reuse of the broch and whether it relates to Pictish or Viking or later
occupation.
7.1.3 The evaluation has produced a small finds assemblage which includes little
cultural material which can be tied to the use of the broch (a few pottery
sherds, some possible utilised stone, a few pieces of worked bone, and a little
iron smelting slag). Little further useful information can be gained from this
material, and further analysis is not recommended.
7.1.4 Animal bone, however, as the most commonly occurring material type, has
greater potential. This is a small but reasonably well preserved and
potentially interesting faunal assemblage with a relatively wide range of
species, which seems to indicate intensive exploitation of animal resources.
Most of the assemblage appears to have derived from contexts associated
with the use of the broch. Further analysis should focus on the bone from the
broch midden in Trench 4, which should be recorded in more detail,
including fragmentation patterns (size and fracture type, as outlined in
Outram 2002) to better understand the intensity of carcass breakdown. The
fish, small mammal and possible seal bones should be formally identified to
species
7.1.5 A short article, probably between 2000-3000 words with two or three
supporting illustrations, based on the results, finds, discussion and figures in
this assessment report, in an appropriate archaeological journal such as the
Proceedings of the Prehistoric Society of Scotland is suggested as an
adequate level of publication given the results from this project. This would
comprise a brief introduction detailing the circumstances of the project and
23
the aims and objectives; a results section detailing the structural remains
recorded, with finds and environmental information integrated into the text as
appropriate; and a brief discussion of the results, with reference to the
original project aims and objectives.
7.1.6 A copy of this report and the geophysical report will be submitted to the
National Museums Record of Scotland (held by the Royal Commission on
the Ancient and Historical Monuments of Scotland). Copies of this report
will also be submitted to Highland Council Sites and Monuments Record and
Applecross Heritage Centre.
7.1.7 Since this site is on a fairly popular campsite, perhaps some notice boards
displaying the results and interpretations from the Time Team excavations
would be a useful way to inform people about the importance of this site.
Furthermore leaflets providing a summary of the excavations might also be
made available, should any grants/funds be available for such a proposal.
8 ARCHIVE
8.1.1 The archive, which includes all artefacts, written, drawn and photographic
records relating directly to the investigation is undertaken, is currently held at
the offices of Wessex archaeology under the site code APP 05 and Wessex
Archaeology project code 59464. It has been quantified, ordered, indexed
and cross-referenced and the paper archive is contained in one lever arch file.
It is intended that, in accordance of the wishes of the Treasure Trove
Advisory Panel Secretariat, the excavated material and records will in due
course be deposited at Inverness Museum. The paper archive will be curated
by Royal Commission on Ancient and Historical Monuments and
Constructions of Scotland.
8.1.2 The paper archive is contained in a lever arch ring binder file. It includes:
Project Design
Copy of this report
The GSB Prospection geophysical report including a record of all data, plots
of the results, interpretation with detailed comments and conclusions.
24
Finds as described in this report (see Section 4 and Tables 1 & 2)
25
REFERENCES
Anderson, G. & Anderson, P., 1850, Guide to the Highlands and Islands, Edinburgh:
Adam and Charles Black
Armit, I., (ed.)., 1990a, Beyond the Brochs. Changing Perspectives on the Atlantic
Scottish Iron Age, Edinburgh: Edinburgh University Press
Armit, I., 1990b. Brochs and Beyond in the Western Isles. In Armit, I. (ed.). 1990.
Beyond the Brochs. Changing Perspectives on the Atlantic Scottish Iron
Age. Edinburgh University Press, Edinburgh, 41-70
Armit, I., 1990c, ‘Epilogue: The Atlantic Scottish Iron Age’ in Armit 1990a, 194-210
Armit, I., 1996, The Archaeology of Skye and the Western Isles, Edinburgh:
Edinburgh University Press
Armit, I., 2002, ‘Land and freedom. Implications of Atlantic Scottish settlement
patterns for Iron Age land-holding and social organisation; in Ballin Smith,
B. & Banks, I. (eds.), In the Shadow of the Brochs. The Iron Age in Scotland,
Stroud: Tempus, 15-26
Armit, I., 2003, Towers in the North: The brochs of Scotland, Stroud: Tempus
Ballin Smith, B., 1994, Howe: four millennia of Orkney prehistory, Edinburgh: Soc.
Antiq. Monog. 9
Boyd, W.E., 1985, ‘Botanical remains of edible plants from the Iron Age broch at
Fairy Knowe, Buchlyvie, near Stirling’, Forth Nat. and Hist. 7, 77-83
British Geological Survey Scotland, 1975, Loch Torridon Sheet 81E: Drift Edition
Callender, G., 1921, ‘Report on the Excavation of Dun Beag, a broch near Struan,
Skye’. Proc Soc Antiq Scot No 55, 110-131
Clapham, A. and Scaife, R., 1988, ‘A pollen and plant macrofossil investigation of
Oakbank crannog’ in Murphy, P. and French, C. (eds), The Exploitation of
Wetlands, Oxford: Brit. Archaeol. Rep. 186, 293-325
Dickson, C. and Dickson, J.H., 2000, Plants and People in Ancient Scotland, Stroud:
Tempus
Geological Survey, 1954, Geological Map of Applecross, Sheet 81 (1st pub. 1899),
Geological Survey of Great Britain
GSB, 2005, Applecross, Highlands, Geophysical Survey Report, unpub. report for
Time Team, ref. 2005/34
26
Guido, M., 1978, The Glass beads of the Prehistoric and Roman Periods in Britain
and Ireland, Rep. Res. Comm. Soc. Antiq. London 35
Harding, D.W., 1984, ‘The function and classification of brochs and duns’ in Miket,
R. and Burgess, C. (eds.), Between and Beyond the Walls. Essays on the
Prehistory and History of North Britain, Edinburgh: John Donald Publishers
Ltd., 206-20
Harding, D.W., 2006, ‘Redefining the Northern British Iron Age’, Oxford J. Archaeol.
25 (1), 61-82
Harding, D.W. and Dixon, T.N., 2000, Dun Bharabhat, Cnip: An Iron Age Settlement
in West Lewis. Vol. 1: Structures and Material Culture, Edinburgh:
Edinburgh University Dept. of Archaeology, Calanais Research Series 2
Hardy, K. & Wickham-Jones, C., 2000, ‘Inner Sound, Highland, survey and
excavation’, Discovery and Excavation in Scotland 1, 44
Hedges, J.W., 1987, Bu, Gurness and the Brochs of Orkney Part 2, Gurness, Oxford:
Brit. Archaeol. Rep. 164
MacKie, E.W., 1965, The origin and development of the broch and wheelhouse
building cultures of the Scottish Iron Age. Proc. Preh. Soc. 30, 93-146.
MacQueen, Rev. J., 1792, Old Statistical Account of Applecross, from the Parish
Manuscripts
Martlew, R., 1982, ‘The typological study of the structures of the Scottish brochs’,
Proc. Soc. Antiq. Scotland 112, 254-76
Ordnance Survey 1874, First Edition of Raasay and Applecross. 1 inch to mile
Outram, A., 2002, ‘Bone fracture and within-bone nutrients: an experimentally based
method for investigating levels of marrow extraction’ in P. Miracle and N.
Milner (eds.), Consuming Patterns and Patterns of Consumption,
Cambridge, McDonald Institute for Archaeological Research, 51-63
Parker Pearson, M., Sharples, N. and Mulville, J., 1996, ‘Brochs and Iron Age society
– a reappraisal’, Antiquity 70, 57-67
Robinson, D.E., 1983, ‘Pollen and plant macrofossil analysis of deposits from the Iron
Age ditch’ in Jones, M. (ed.), Integrating the Subsistence Economy, Oxford:
Brit. Archaeol. Rep. Int. Series 181, 123-34
Renfrew J., 1974, ‘Cereals’ in MacKie, E.W., Dun Mor Vaul: An Iron Age Broch on
Tiree. Strathclyde: Glasgow University Press, 210-13
27
Thomas, A.C., 1971, The Early Christian archaeology of north Britain. The Hunter
Marshall lectures delivered at the University of Glasgow in 1968, London
28
Table 1: Finds totals by material type and by trench (number / weight in
grammes)
Cattle Sheep/ Pig Canid Deer Small Fish Seal? Uniden Total
Goat (fox?) Mammal -tified
Fragment count 41 29 6 2 16 3 16 4 552 669
29
Table 3. Assessment of the charred plant remains and charcoal
Feature no
Charcoal
Charcoal
Flot size
Other
Volume
Roots%
Context
Sample
Notes
Notes
Notes
Weed
Chaff
Grain
seeds
Res.
?Iron Age
Trench 1
Midden Material
111 4 10 700 0* C - Barley - Mytilus edulis **** C moll-t - -
(C)
Trench 3
Midden Material
303 2 10 700 95 A** hulled C(h) Plantago Hazel frgs. C - - -
Barley x45 x9
oat x2 1x A. elatius ssp.
bulbosum 1x rootlet
315 5 10 500 40 - - 1 bud A* Corylus avellana A** - - -
x40+
Pit/Hearth
317 316 6 8 250 95 A* Barley x7 A* tubers x30 rootlets C fish (C) - -
Avena x2 x15
40+ x Carex sp.
2x Chenopodium
Trench 4
Midden
407 402 1 12 500 5* A - Barley x9 A(h) 1x Avena Corylus (A) A** fish bulae
(A**) x1
403 7 12 300 5* B - Barley x7 A(h) 19x Corylus avellana A** fish (A*) - -
404 3 12 40 10 C - Barley x3 C (h) 1x Avena. Poa sp. x1 C fish (C) bulae -
3x Corylus avellana x1
KEY: A** = exceptional, A* = 30+ items, A = t10 items, B = 9 - 5 items, C = < 5 items, (h) =
hazelnuts, smb = small mammal bones; Moll-t = terrestrial molluscs Moll-f = freshwater molluscs;
Analysis, C = charcoal, P = plant, M = molluscs
30
APPENDIX 1: TRENCH SUMMARIES
31
Trench 3. Northern side of broch and rectangular structure
Max Depth: 0.68m Length: 18.5 m Width: 0.7-1.8 m
No. Type Description Depth
301 Topsoil Topsoil and turf. Dark brown silty sand with frequent angular pebbles 0-0.10m
302 Deposit Rubble collapse associated with hillwash 0.10-0.25m
303 Deposit Dark brown organically rich deposit that is midden derived -
304 Deposit Rubble collapse layer that is distinct from 302 -
305 Natural Orange brown sandy natural comprising degraded sandstone fragments -
306 Deposit Stone tumble at base of hill -
307 Wall External face of external broch wall -
308 Wall Internal face of external broch wall -
309 Deposit Rubble infill between 307 and 308 -
310 Structure Stone rubble demarcating the gallery between the external and internal -
broch walls
311 Wall External face of internal broch wall -
312 Deposit Wall infill to south of wall 311 -
313 Deposit Dark brown silty sand associated with tumble 318 -
314 Wall Single course of N/S aligned drystone wall. Part of a rectangular structure -
315 Deposit Dark brown black organic rich deposit abutting the external face of the -
broch (307)
316 Deposit Fill of 317 -
317 Cut Possible pit feature – not fully excavated. F.W. 316 -
318 Deposit Tumble associated with rectangular structure 314 -
319 Deposit Dark brown midden material dumped between the two broch walls -
320 Deposit Light grey silty clay, sealed by 314. -
321 Stone Flat thin stone overlying 316 and under 303 -
322 Group Group Number for external broch wall comprising 307, 308 and 309
323 Group Group Number for internal broch wall comprising 311 and 312
32
Trench 6. North of Broch. Geophysical Anomaly
Max Depth:0.55m Length: 2m Width: 1.9m
No. Type Description Depth
601 Topsoil Turf and topsoil with rubble. Mid brownish grey silt loam 0-0.15m
602 Cut Cut of modern pit excavated by the children in 1974 0-0.38m
603 Deposit Modern fill of 602 – evidence of burning from bonfires 0-0.38m
604 Natural Orange-pink degraded sandstone natural 0.15m+
33
Geophysical data courtesy of GSB Prospection Ltd
171180
Wall 120
Plate 1: Internal doorway in Trench 1 (Group 117)
Wall 123 (Scales = 0.5m & 1m)
111, Midden
External Broch wall
Group 121
Wall 104
Wall
105, Wall core 110
116,
107, Rubble Gallery
of Broch
Plate 2: Internal stairway in Trench 1 (113)
11 Stairs
113, (Scale = 0.5m)
Internal doorway
to Gallery
106, Wall 804 Group 117
Broch
collapse Wall 114
Trench 1
Wall 803
Plate 3: Post-excavation view of Trench 1 from north east
103, Rubble (Scale = 2m)
802, Wall core
This material is for client report only © Wessex Archaeology. No unauthorised reproduction. Date: 16/03/06 Revision Number: 0
201
202,
Tumble
178.40mOD
204
Cut
205
Bioturbation
Cut
(Roots)
205
Bioturbation
(Roots) 0 1m
Cut [205]
Section
Wall 203,
?Causeway
844320
This material is for client report only © Wessex Archaeology. No unauthorised reproduction. Date: 16/03/06 Revision Number: 0
Plan
continued
below Trench 3
(South)
318, 844340
Tumble
Wall
314
Elevation
External Broch wall
Group 322
304,
Rubble/ Trough quern
Broch collapse fragment SF2
Wall
307
Gallery
310
844345 Trench 3 Wall Internal Broch wall
308 Group 323
(North)
Plan
continued 844335
above
321
Wall
311
Hearth/Kiln
317
Wall core
0 1 2m 312
Trench 3 - Elevation
NORTH EAST SOUTH WEST
40.52mOD
Wall
307
0 1m
This material is for client report only © Wessex Archaeology. No unauthorised reproduction.
Trench 4
404, Section
Broch
collapse
0 1 2m
171200
Trench 4 - Section
EAST WEST
401
402
37.70mOD
403
Stone
404 Shell
Charcoal
405
0 1m
This material is for client report only © Wessex Archaeology. No unauthorised reproduction.
Plate 11: View of 'supposed' standing stone in Trench 5 from north east (Scale = 1m)
This material is for client report only © Wessex Archaeology. No unauthorised reproduction.
Scale: NA Illustrator: MR
Wessex
Archaeology Path: London: Y:\Projects\59464TT\Drawing Office\Report Figures (06-02)\Time Team
844370
Trench 7
703
Trench 6
702,
Rubble/
Collapse
0 1 2m
Stone Rubble/Broch Collapse
(Scale for plans, 1:50 @ A3)
This material is for client report only © Wessex Archaeology. No unauthorised reproduction.
Registered Charity No. 287786. A company with limited liability registered in England No. 1712772.