ArticlePDF Available

Abstract and Figures

The cemetery at Heath Wood, Ingleby, Derbyshire, is the only known Scandinavian cremation cemetery in the British Isles. It comprises fifty-nine barrows, of which about one-third have been excavated on previous occasions, although earlier excavators concluded that some were empty cenotaph mounds. From 1998 to 2000 three barrows were examined. Our investigations have suggested that each of the barrows contained a burial, although not all contain evidence of a pyre. A full report of the 1998-2000 excavations is provided, alongside a summary of the earlier finds. The relationship of Heath Wood to the neighbouring site at Repton is examined, in order to understand its significance for the Scandinavian settlement of the Danelaw. It is concluded that Heath Wood may have been a war cemetery of the Viking Great Army of AD 873-8.
Content may be subject to copyright.
EXCAVATIONS AT THE VIKING BARROW
CEMETERY AT HEATH WOOD, INGLEBY,
DERBYSHIRE
Julian D Richards, FSA, with contributions by Pauline Beswick, FSA, Julie Bond,
Marcus Jecock, Jacqueline McKinley, Stephen Rowland and Fay Worley*
The cemetery at Heath Wood, Ingleby, Derbyshire, is the only known Scandinavian cremation
cemetery in the British Isles. It comprises fifty-nine barrows, of which about one-third have been
excavated on previous occasions, although earlier excavators concluded that some were empty
cenotaph mounds. From
1998
to
2000
three barrows were examined. Our investigations have
suggested that each of the barrows contained a burial, although not all contain evidence of a pyre. A
full report of the
1998
2000
excavations is provided, alongside a summary of the earlier finds. The
relationship of Heath Wood to the neighbouring site at Repton is examined, in order to understand
its significance for the Scandinavian settlement of the Danelaw. It is concluded that Heath Wood
may have been a war cemetery of the Viking Great Army of AD
873
8
.
LOCATION
The Scandinavian barrow cemetery known as Ingleby (SK 342259) comprises fifty-nine
mounds in four clusters. It is set within a small woodland block called Heath Wood on the
western edge of Ingleby parish in Derbyshire. The cemetery has a northern aspect, lying
between 102m and 114m above OD on and below the shoulder of a small bluff on the heavily
dissected southern edge of the Trent Valley which here runs almost due east west (fig 1). The
underlying geology comprises Triassic sandstones of both the Keuper and the Bunter series.1
The soil is acidic and sandy, leading to good drainage but poor preservation. The site is
Scheduled Ancient Monument number ‘Derbyshire 101’.2Heath Wood is currently owned by
the Church Commissioners, although leased to the Forestry Agency.
*Julian D Richards, Department of Archaeology, University of York, King’s Manor, York YO
17
EP, UK.
E-mail: <jdr
1
@york.ac.uk>.
Pauline Beswick,
4
Chapel Row, Froggatt, Calver, Hope Valley, Derbyshire S
32 3
ZA, UK.
E-mail: <Paulinebwick@aol.com>.
Julie Bond, Department of Archaeological Sciences, University of Bradford, Bradford BD
71
DP, UK.
E-mail: <J.M.Bond@bradford.ac.uk>. (JB)
Marcus Jecock, English Heritage,
37
Tanner Row, York YO
16
WP, UK.
E-mail: <Marcus.Jecock@english-heritage.org.uk>.
Jacqueline McKinley, Wessex Archaeology, Portway House, Old Sarum Park, Salisbury SP
46
EB, UK.
E-mail: <j.mckinley@wessexarch.co.uk>. (JMcK)
Stephen Rowland, Department of Archaeology, University of York, King’s Manor, York YO
17
EP, UK.
E-mail: <stephenrowland@yahoo.co.uk>. (SR)
Fay Worley, Department of Archaeological Sciences, University of Bradford, Bradford BD
71
DP, UK.
E-mail: <F.L.Worley@bradford.ac.uk>. (FW)
The Antiquaries Journal,8484,2004,pp23116
Heath Wood comprises a 14ha block of mixed woodland which once formed part of
the parkland setting of Foremark Hall, 1km to the north west. The present hall was built by
Sir Robert Burdett from 1759 to 1761, but is known to have replaced an earlier house first
documented in 1712 but whose existence from at least the middle of the seventeenth century is
Fig 1. Location of barrow cemetery at Heath Wood, Ingleby. Boundaries shown are those
of the modern civil parishes. Contours in metres. Drawing: NMR, Crown Copyright 1995,
based on Ordnance Survey mapping with permission
24 THE ANTIQUARIES JOURNAL
implied by Sir Francys Burdett’s rebuilding in 1662 of the nearby St Saviour’s Church.3The
previous hall was described by Woolley as ‘large and convenient with a large well-wooded park
and coney warren adjoining’ and was assessed at twenty-four hearths in 1662.4However, in the
Burdett papers Heath Wood is still referred to as ‘ye Heath feild’ in 1664, suggesting that at
that stage it remained open heath land. Burdett’s 1791 map shows the area as only lightly
wooded, with several areas of open moor and heath.5From documentary sources, therefore, it
seems probable that Heath Wood was established as part of the landscaping for the present hall
in the late eighteenth century, and before that was open land, possibly even the rabbit warren
mentioned in 1664. It lies en route between Foremark and the second Burdett house-cum-
summerhouse at Knowle Hill and is bisected by a track connecting the two properties. The
track is now in places no more than an overgrown footpath, but is well engineered and, where it
passes uphill through the wood, is also metalled, suggesting it originated as a carriage drive.
The date of its construction is uncertain, but it is also unlikely to be later than the middle of the
eighteenth century when the Burdetts were living at Knowle whilst Foremark was remodelled.
The drive passes right through the cemetery although there is no evidence to suggest that any
barrows were destroyed by its construction; instead, it is possible that the drive was sited so
that the cemetery became a feature to be viewed from it.
Before the plantation of the wood in the eighteenth century, the clusters of barrows
therefore occupied open heath land, and appear to have been respected by agricultural activity,
including medieval ridge and furrow ploughing (see below), probably because of the rocky
and uneven nature of the ground. When originally constructed the barrows would have
commanded impressive views northwards, across the flood plain of the Trent and taking in the
site of the Viking winter encampment of AD 873 4at Repton in the foreground, 4km to the
north west. The spire of St Wystan’s Church is today visible from the edge of the wood, and it
seems likely that the chancel of the Saxon monastic church and the encompassing massive
Danish earthwork, revealed during the 1974 93 excavations by Martin Biddle and Birthe
Kjølbye-Biddle, would have been visible in the late ninth century.6
The cemetery has become known as Ingleby, although Heath Wood is more accurate. The
present village of Ingleby lies 1km to the north east, and is assumed to be Englaby (1009:
Sawyer 1968,922) and Englebi (1086: Domesday). As a label for a Viking cemetery the use of
the name Ingleby has been a source of confusion as it suggests ‘a farm of the Angles’, probably
denoting ‘an isolated survival of English inhabitants amongst a prevailing Scandinavian
population’.7It makes much more sense when it is realized that it denotes not the cemetery but
a neighbouring settlement which was named by a predominant local Scandinavian group:
‘Ingleby is now merely a handful of houses standing in the shelter of steep wooded cliffs and its
site would have been a suitable one to which to retreat in the face of invasion’.8Indeed,
although Derbyshire generally has few Scandinavian names, and only ten place-names
containing the Danish -by, these are clustered south of the Trent in the vicinity of Repton.9
The town of Derby, one of the Five Boroughs of the Danelaw, is 10km north of Heath Wood.10
Apart from Derby itself there is also the village of Bretby, like Ingleby distinctive as the ‘village
or farm of the Britons’. These names testify to the density of Scandinavian dominance in an
area in which Anglo-Saxon settlement was also considerable.11 The assimilation of the Anglo-
Scandinavian population may be reflected in hybrid names such as Ravenstone, recorded
as early as 942, suggesting a rapid fusion of the two groups.12 The name Foremark is also
of Scandinavian derivation. In Domesday Book it is recorded as Fornewerke, or the ‘old
fortification’, the equivalent of the English Aldwark, although the word verk is not recorded in
this sense in the Scandinavian homelands,13 suggesting that the sense developed in England
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 25
under the influence of OE (ge)weorc, ‘fortification’.14 Only later is the second element replaced
by the form mearc.15
PREVIOUS WORK
There have been three previous recorded investigations of the barrows in Heath Wood: those
of Thomas Bateman in 1855, Camden Clarke and William Fraser from 1941 to 1949, and
Merrick Posnansky in 1955. These have resulted in the partial or complete excavation of
twenty mounds.
Excavations by Bateman, 1855
In Ten Years’ Diggings Thomas Bateman describes more than fifty mounds, 21 30ft (6.49m)
in diameter and 23ft (0.60.9m) high, located in a plantation known as ‘The Ferns’, near
Foremark Hall. He opened five, on 22 May 1855, possibly whilst staying at Foremark as a guest
of the Burdett family, and found that in each case:
the mound had been raised over calcined human bones, which lay in the same place on
the natural surface as they occupied when the funeral pile was smothered out by the
casting up of the tumulus. The bones and black ashes of the pyre, reduced by
compression to a layer about an inch thick, generally covered a space about four or five
feet diameter in the centre; above were accumulated stones bearing marks of fire, which
had been first thrown on the glowing embers, and over these earth was heaped to form
the bowl-shaped mound.16
The only finds recorded by Bateman were: ‘two very small fragments of iron, found with two
separate interments, one only having the definite form of a very slender pin, 13
4inches long’.
Neither object is listed in the catalogue for that part of the Bateman Collection later purchased
by Sheffield Museum and it is unlikely that they now survive.17 In the context of the later finds
it may be more likely that the object described by Bateman as a pin was actually an iron nail.
No record of the position of the barrows opened by Bateman survives, although the survey
by the Royal Commission on the Historical Monuments of England (RCHME, see below)
identified one barrow (Mound 16) which, on the basis of its earthwork form, can be inferred to
have been opened by a hole dug in its centre, but which cannot be linked to an otherwise
documented excavation. Modern excavation of a second (Mound 12) produced evidence of
earlier peripheral disturbance but neither of these actual or possible acts of excavation can be
attributed definitively to Bateman.
Bateman was uncertain of the date of the cemetery but, noting the similarity of the mounds,
concluded that ‘no great variation of date, if any, existed as to their age’. From the presence of
iron he further concluded that the mounds were unlikely to be ancient and: ‘would rather seek
to connect them with the eventful period in which tradition affirms the place to have been the
scene of a sanguinary conflict between the Saxons and their Danish enemies’.18
Excavations by Clarke and Fraser, 1941 9
The site was referred to in VCH Derbyshire in conjunction with pagan Anglo-Saxon cemeteries
in the area, but its precise location appears to have been forgotten until rediscovered in 1941 by
26 THE ANTIQUARIES JOURNAL
William Fraser, who also initially ascribed it to the sixth seventh centuries AD.19 Under the
leadership of Camden Clarke and William Fraser members of the Burton-on-Trent Natural
History and Archaeological Society initially excavated six barrows, digging most weekends
throughout the summers of 1941 and 1942.20 The mounds were trenched, usually by a 6ft-
(1.8m-) wide trench, although in some cases the central area was widened in search of a burial
deposit (fig 2). The mounds were made up of sand and stone, but all had been heavily
disturbed by root and animal action and, together with the amateur nature of the excavation
technique, this means that it is now impossible to know whether they possessed any kind of
internal structure. Of the six, two (Mounds 2and 4) appeared to be empty whilst another
(Mound 3) produced several pieces of metalwork and had charcoal and bone distributed
throughout the body of the mound but lacked any identifiable burial. The remaining three
(Mounds 1,5and 6) all produced in situ central ‘cremation-hearth’ deposits and metalwork
(see fig 15 for mound locations). The principal metal finds included a fragmentary iron sword
(from Mound 1), together with iron buckles, an iron strap-slide and a bronze suspension loop.
A survey of the cemetery was conducted by T A Dallman, but would appear to have been
carried out at a small scale and to have been intended as no more than a location plan for the
Fig 2. Mounds 16, showing trench outlines. Drawing: Frances Chaloner, after DAJ 6666
and unpublished plans in Repton School archive
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 27
excavations: the published plan shows a total of sixty-three mounds but only schematically
as open circles.21 The accompanying report contains no detailed discussion either of the form
and inter-relationships of individual barrows or of the significance of the way the barrows
are distributed across the cemetery. Only the six excavated barrows are numbered on this
plan.
Unable to date their discoveries, Clarke and Fraser initially approached T D Kendrick
at the British Museum, who suggested that the burials were Anglo-Saxon but counselled: ‘I
am afraid the proper thing to say to inexperienced diggers is to ask them to postpone the work
until they can get a trained archaeologist to supervise it, for barrow-digging now is big
undertaking.’22
It appears that Kendrick also contacted W F Grimes of the Inspectorate of Ancient
Monuments and Historic Buildings. Grimes visited the site in May 1942 and confirmed that
the burials must be Anglo-Saxon.23 Both Kendrick and Grimes offered to help with
publication but suggested that Clarke and Fraser also contact E T Leeds at the Ashmolean
Museum for assistance with the finds. Leeds was approached in 1944 with a description of the
work and immediately questioned the early Anglo-Saxon dating.24 On full examination of the
finds he wrote:
I am convinced the finds are not Saxon though of the late Saxon period. Bateman after
all was not very far from the truth ... I feel that meagre as in some respects they are,
your finds illuminate some of the sepulchral darkness of late Saxon times.25
A short description of the mound structure, and where recorded the skeletal remains and
artefacts from each mound, is presented below. A similar consistent format is also followed for
the 1955 interventions. The descriptions are derived from the published reports, supplemented
with additional information from archival correspondence and personal observation, in some
cases involving some reinterpretation of the finds. Concordances are provided to the original
reports.26
Mound
1
STRUCTURE
Diameter: 9.75m; height at centre: 0.86m. A trench 6ft (1.78m) wide was dug from just beyond
the southern edge of the mound to 4ft (1.22m) beyond the centre; Dallman’s plan and a
photograph of the mound under excavation looking north indicate that this was extended into
an 8ft (2.44m) square in the centre, and that the south-east quadrant of the mound was also
partially removed. The report claims that the mound make-up was stratified according to
vertical bands of marl and gravel with little sandstone alternating with bands with larger pieces
of sandstone; however, this seems unlikely and it is more probable that the trench simply
encountered variable densities of stone reflecting differences in the horizontal distribution. In
the centre of the mound there was a roughly circular hearth of charcoal, c1.75m in diameter
and c50mm thick, surrounded by a ring of burnt sandstones. Fragments of burnt bone were
embedded in the charcoal surface; several burnt artefacts were also recovered from this layer. It
is noted that there were also several large pieces of burnt bone outside the hearth, and that
there was clean sand to a depth of 0.9m below the hearth. Major disturbance by root action and
burrowing animals is recorded; this may have been responsible for the displacement of some of
the burnt bone into the mound make-up.
28 THE ANTIQUARIES JOURNAL
Fig 3. Mound 1: sword, surviving fragments. Drawing:
Frances Chaloner
Fig 4.Mound1: sword, reconstruction.
Drawing: Sven Schroeder, redrawn after
DAJ 6666,fig5
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 29
SKELETAL REMAINS
Incomplete remains of human skeleton; age and sex unknown. Four fragments of burnt animal
bone, three tentatively identified as: fragmentary shaft of sheep tibia; sheep vertebra; carnivore
pre-maxilla with sockets for two teeth, perhaps of a young dog; also unburnt animal remains
comprising seven teeth and skull fragment of a cow. DAJ 6666 (1946), 202.
ARTEFACTS
1. Iron sword, incomplete, comprising end of channelled blade and fragments of hilt and
pommel (figs 3and 4). L385mm, W50mm. DAJ 6666 (1946), 6,1011,1416,figure5.
In addition to the blade, four other fragments are labelled as ‘Blade (1)’ in an inventory
dated 1985 held in Derby Museum and are grouped with the sword blade. They were used
in the 1946 reconstruction drawing (fig 4) and identified in the report as comprising a
portion of the pommel, a small piece of the hilt, the upper guard and half the lower guard.
Leeds considered the sword, with its curved upper and straight lower guards, to be a hybrid
of various forms, particularly Petersen Norwegian Types L and R, whilst noting that Type
R was introduced into Scandinavia from western Europe, and dated it to ‘the latter years
of the ninth century’.27 Shetelig, on the basis of the reconstruction drawing, believed ‘the
pommel to be of the Wallingford type, Jan Petersen Type L, and the guard narrow and flat,
with a slightly humped outline, perhaps suggesting a form of the guard like Jan Petersen
Type O’; he dated the sword to cAD 900.28 Swords of this period were generally over
0.9m in length, including the blade and tang, implying that at least half the blade is missing
and may have been removed from the pyre. The X-ray plate shows no trace of pattern
welding but this was, in any case, in decline by this period. The purpose of the broad
shallow groove, or ‘fuller’, down the centre of the blade was to reduce the weight of the
sword.29
2. Bronze suspension loop (fig 5,no.2). L42mm, W16mm. DAJ 6666 (1946), 6,16 17,
figure 6. Leeds noted that it was ‘stoutly made with one end looped, on which is a large
nipple, and expanding to the open end; two rivets, one above the other on the median line.
The jaws of the loop are set sufficiently apart to have taken a doubled strap. It might
conceivably have served to suspend the scabbard from the belt.’30 Theloopmaybeclosely
paralleled by those on the strap distributor found in the Viking burial at Ballateare, on the
Isle of Man, which Wilson suggests would have functioned as part of a shoulder sling for the
sword. If so then it was originally one of a set of three. Wilson also notes parallels from Meols
in Cheshire, Bledlow, Buckinghamshire, and St Mary’s Abbey, York; he considers that the
mounts were the product of an Anglo-Saxon workshop and concludes that since similar
mounts are not known in association with scabbards in Scandinavian contexts this was an
Anglo-Saxon method of attachment.31
3. Iron strap slide, oblong with an elongated oval opening (fig 5, no. 3). L26mm, W17mm.
DAJ 6666 (1946), 6,1617, figure 8.
4. Iron buckle, round in section, missing tang (fig 5, no. 4). L49mm, W30mm. DAJ 6666 (1946),
figure 7.
5. Bronze fragments. Two unidentifiable bronze fragments are also noted in DAJ 6666 (1946)
but are documented as missing in the 1985 inventory; one is described as ‘a portion of a
rounded bar, the other a thin rod, expanding and flattening in one direction, in a way
reminiscent of an Anglo-Saxon toilet implement’.32
30 THE ANTIQUARIES JOURNAL
Fig 5. Mound 1: objects bronze suspension loop, iron buckle, iron strap slide. Drawing:
Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 31
Mound
2
STRUCTURE
Diameter: 8.5m; height at centre: 0.75m. A 6ft- (1.78m-) wide trench was dug from the south
for a distance of 21ft (6.4m); the central area was increased to a rectangle 8ft (2.44m) across. No
traces of a hearth or burial were found. Irregularly distributed sandstone blocks were present
in large quantity; in the central area the mound was dug to below natural ground level and
revealed undisturbed conditions.
Mound
3
STRUCTURE
Diameter: 8.5m; height at centre: 0.75m. A 6ft- (1.78m-) wide trench was dug from the south
for a distance of 18ft (5.5m). No hearth was found, but traces of bone and charcoal were
unevenly distributed throughout the excavated area. There was evidence for extensive
disturbance from burrowing animals.
SKELETAL REMAINS
Incomplete remains of human skeleton; age and sex unknown.
ARTEFACTS
1. Iron buckle (fig 6, no. 1); rectangular loop with laterally projecting ends; folded backplate.
L46,W20mm. DAJ 6666 (1946), 16, figure 9.
2. Two iron nails, of different sizes, with spherical heads (fig 6, nos 2i and 2ii). (i) Large nail:
L30mm, not identifiable in 1985; (ii) small nail head: D10mm; shank: L18mm. DAJ 6666 (1946),
16, figure 10.
3. Fragments of thin sheet bronze, recorded as missing in 1985. (i) Narrow half-tubular
piece of curved rim; (ii) a double piece, rounded and riveted at one end: L12mm, W9mm.
Described by Leeds as possibly belonging ‘to a stoup or bucket’, 3in. (76mm) in diameter.
DAJ 6666 (1946), 18.
4. Fragment of bronze rod, slender and bent at one end, also noted as missing in 1985.
Lc32mm. DAJ 6666 (1946), 18.
Mound
4
STRUCTURE
Diameter: 13.4m; height at centre: 1.37m. A trench 6ft (1.78m) wide was dug from south to
north across the mound; in the central area this was expanded to a square 10ft (3m) across,
excavated to a depth of 2ft (0.6m) below the natural ground surface; the trench was widened at
the north. The report notes the presence of sandstone slabs and blocks of considerable size,
dispersed both across the surface of the mound and throughout the trench. No trace of an
interment was encountered.
32 THE ANTIQUARIES JOURNAL
Mound
5
STRUCTURE
Diameter: 8.5m; height at centre: 0.91m. A trench 6ft (1.78m) wide was dug from the east for
a distance of 19ft (5.6m). Small sandstones were dispersed throughout. Traces of a charcoal
hearth were revealed in the central area; trench expansion showed that this extended for 5ft
(1.5m) beyond the original southern trench edge.
SKELETAL REMAINS
Considerable portion of a human adult skeleton, probably female. Also eleven fragments
of burnt animal bones, tentatively identified as: parts of astragalus of cow or horse; caudal
vertebra of cow; magnum of cow; atlas of cow; fragment of shank-bone of ?horse; fragment of
lumbar vertebra of cow; sesamoid bone from behind the metacarpo-phalangeal articulation of
cow. DAJ 6666 (1946), 21 2.
ARTEFACTS
1. Large iron nail with discoid head (fig 7, no. 1). Head: D30mm; shank: D7mm, L62mm.
Fig 6. Mound 3: objects iron buckle, two iron nails. Drawing: Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 33
This is a substantial nail that must have been derived from something structural; it may have
been a clench nail but the headless end is broken. DAJ 6666 (1946), 18, figure 11.
2. Small iron nail (fig 7, no. 2). Head: D8mm; shank: L14mm. DAJ 6666 (1946), 18.
Mound
6
STRUCTURE
Diameter: 9.1m; height at centre: 0.84m. A 6ft- (1.78m-) wide trench was dug from the
southern to the northern edge of the mound, its western edge following the central north
south line of the mound. A very thin charcoal layer was found in the centre, lying almost
entirely in the northern half. Very few pieces of sandstone were encountered.
SKELETAL REMAINS
Incomplete remains of adult skeleton, possibly female. Also several badly preserved burnt
animal bones: fragment of cow or sheep tooth; ?pig astragalus; cow ?skull; also unburnt cow
metacarpal and toe-bones. DAJ 6666 (1946), 20 1.
ARTEFACTS
1. Iron buckle, with strap slide secured between backplate and two terminal rivets (fig 8,
no. 1). L52mm, W40mm. DAJ 6666 (1946), 18 19, figure 12.
Fig 7. Mound 5: objects – two iron nails. Drawing: Frances Chaloner
34 THE ANTIQUARIES JOURNAL
Fig 8. Mound 6: objects two iron buckles, six nails, bronze strap-end and two other
bronze fragments. Drawing: Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 35
2. Iron buckle, with two terminal rivets (fig 8, no. 2). L50mm, W39mm. Leeds suggests that
buckles with swollen tongues can be paralleled to Scandinavia and cites an example from an
inhumation grave at Stengade, Tullebolle, Svendborg, in the island of Langeland.33 DAJ 6666
(1946)18 19, figure 13.
3. Six iron nails, recorded as unidentifiable in 1985 (fig 8, no. 3). L c25mm with flat circular
heads. DAJ 6666 (1946), 18 19, figure 14.
4. Bronze strap-end, with two rivets; no trace of decoration (fig 8, no. 4). L38mm, W10mm.
DAJ 6666 (1946), 1819, figure 15.
5. Bronze ?ornament. L c25mm. Described by Leeds as perhaps a small brooch with divided
bow, but badly crumpled by fire; missing in 1985.DAJ 6666 (1946), 18.
6. Three bronze fragments, only two identifiable in 1985 (fig 8, no. 6). (i) Formless fragment:
L24mm, W6mm; (ii) piece of thin rod with a looped end rod: L20mm; looped end: D6mm,
L10mm. DAJ 6666 (1946), 18.
Clarke and Fraser excavated a seventh barrow in the autumn of 1948, and issued a rather
summary report on it in the following year.34 The excavation uncovered a central cremation-
hearth deposit and various pieces of metalwork including a fragment of a second sword, again
dated by Leeds to the ninth=tenth centuries AD, although it has never been published. In the
report this barrow (Mound 7) is described as lying ‘a few yards to the west of mound 6’.35 No
site plan was published, but a copy of Dallman’s survey reproduced in a later article by
Posnansky shows Mound 7as lying a short distance east of Mound 6.36 It seems likely that it is
Clarke’s account that is in error since no barrow exists on plan in the location he describes, and
Posnansky’s siting of it coincides with a barrow whose earthwork form preserves visible signs
of excavation.
Mound
7
STRUCTURE
Diameter: 8.5m; height at centre: 0.9m. A trench 7ft (2.1m) wide was dug from the south west
towards the centre of the mound. The report notes that the first 12ft (3.65m) was mainly marl
and gravel, but at this point larger pieces of sandstone were encountered, and here a very thin
layer of charcoal, interspersed with burnt bone fragments, was located. The trench was extended
to reveal the whole of the charcoal hearth which was found to be well off-centre, only 2ft (0.6m)
within the north-west edge of the mound. The mound was considerably disturbed by animal
and root action. DAJ 6969 (1949), 78 9.
SKELETAL REMAINS
Unrecorded.
ARTEFACTS
1. Iron sword blade, incomplete, comprising end of blade, unchannelled (fig 9, no. 1).
L263mm, W50mm. As in the case of Mound 1, less than half the sword blade is present. Again,
the X-ray plate of the blade shows no evidence of pattern welding. DAJ 6969 (1949), 79 81.
2. Iron hilt guard (fig 9, no. 2). L135mm, W15mm. Found 2.5ft (0.76m) from the blade;
described by Leeds as the lower guard of a sword. DAJ 6969 (1949), 79 81. Not identified in
36 THE ANTIQUARIES JOURNAL
Fig 9. Mound 7: sword and hilt guard. Drawing: Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 37
1985, but two fragments were labelled as iron arrowheads which were said to have ‘come with
Viking swords’. Inspection of these fragments in 2003 demonstrated that they were conjoining
and actually formed a misshapen hilt guard, as illustrated in figure 9. Leeds considered that
the sword resembled Petersen Type H, which has a guard of similar shape with a triangular
pommel, and was dated by Petersen to AD 800 950. Type H is the commonest Viking sword
found in Norway, and it is of note that well over a quarter of Irish Viking Age swords (twenty-
five out of ninety) can be classified as belonging to Petersen’s Type H.37
3. Approximately 24 small iron nails (figs 10a and b). Heads: D c716mm; shanks: L15
30mm. Many described as having the shank clenched over, and one had a smaller rivet
attachment. Leeds suggests that the length of the nail and the depth of the turnover indicate
that they were studs riveted through the leather and wood of a shield. DAJ 6969 (1949), 79 81.
4. Iron ?spur (fig 11). L c80mm, W c60mm. Several larger pieces of iron are also described as
having been found on the hearth; a fragment of an iron object, accessioned in the same system
as other Heath Wood finds, and possibly a spur, may be one of these. Its form is similar to an
example from Balladoole, on the Isle of Man.38 DAJ 6969 (1949), 79.
Mound
8
STRUCTURE
Posnansky’s 1956 plan also indicates the position of an eighth barrow which Clarke and Fraser
excavated in late autumn 1949.39 Although Clarke never reported on this barrow and nowhere
else is the excavation recorded, Posnansky had received a letter from F W Munslow – one of
Clarke’s collaborators with whom he was in touch.40 The letter includes an extract from
notes made in November or December 1949 which record: ‘An eighth mound was opened but
results were most disappointing, especially after the numerous finds in M7. The mound had
been extensively burrowed by rabbit and fox and the hearth was thin and scattered, the metal
finds were negligible, and very little bone was found’. The surviving form of Mound 8as
recorded by the RCHME survey makes it clear that it had indeed been opened at some time.
Excavations by Posnansky, 1955
Following the publication of Mounds 17,in1951 the cemetery was designated a Scheduled
Ancient Monument by the then Ministry of Public Buildings and Works (MoPBW).
Nonetheless, the mounds remained under threat from forestry operations and in 1954 FT
Wainwright, who had retained an interest in the site, wrote to the Inspectorate to express his
concern over reports of damage to the mounds. He was reassured by a visit to the site by the
then Assistant Inspector, Stuart Rigold, and it appears that the Inspectorate was aware of
the problem and contemplating further excavation.41 In 1955 the Forestry Commission gave
the MoPBW notice that they wished to fell and replant Heath Wood, and an agreement was
reached whereby the Forestry Commission would clear but leave unplanted the four principal
barrow concentrations in the wood defined on the basis of Dallman’s survey, whilst the
MoPBW excavated seven ‘outlying’ mounds adjudged to be those most at risk from casual
damage during forestry operations. These seven were excavated over a three-and-a-half-week
period by Merrick Posnansky, and the results subsequently published (fig 12), although it was
really only Arbman’s inclusion of a plan of the cemetery in his 1961 synthesis that rescued the
38 THE ANTIQUARIES JOURNAL
Fig 10a. Mound 7: iron nails. Drawing: Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 39
Fig 10b. Mound 7: iron nails (continued). Drawing: Frances Chaloner
Fig 11. Mound 7: iron ?spur. Drawing: Frances Chaloner
40 THE ANTIQUARIES JOURNAL
Fig 12. Mounds 915: plans showing excavated areas and cross-sections. Drawing: Frances Chaloner, after DAJ 7676, figs 3
and 4
41
site from complete obscurity.42 Clarke and Fraser’s barrow numbering system was retained
and extended to cover the new excavations (Mounds 915). One mound (Mound 14) proved
on excavation to be natural. Of the other six, only one (Mound 11) contained a cremation
hearth and metalwork, the others being apparently empty and described as cenotaph burials.
All the mounds were heavily disturbed by animal and root action. In Mound 11 the false cairn
covered a layer of charcoal and burnt human and animal bone some 23in. (50 75mm) thick.
The metalwork was fragmentary, but included at least one piece of ?silver-wire embroidery,
compared by Elisabeth Crowfoot to parallels in ninth- and tenth-century AD Scandinavian
contexts.43 Following excavation the upstanding portions of the mounds were supposedly
levelled by machine, although the RCHME survey found four at least in part still standing.44
Mound
9
STRUCTURE
Diameter: 4.56m; height at centre: 0.75m. North-west, north-east and south-east quadrants
excavated; sandstone cairn overlying a mound of sandy soil, resting upon natural sand, 9in.
(225mm) deep. No surrounding ditch, but possible stone kerb on south-east side. No trace of
any hearth or interment. Mound overlay a V-shaped ditch.
Mound
10
STRUCTURE
Diameter: 24.5m; height at centre: 0.4m. Trenched along north-north-west to south-south-
east and west-south-west to east-north-east axes across centre; sandstone cairn, including some
very large blocks. No traces of a ditch but the report notes that a kerb was identified. No
indication of an interment.
Mound
11
STRUCTURE
Diameter: 4.56m; height at centre: 0.4m. North-east and south-west quadrants excavated,
extended to form 15ft (6m) square in centre. Dispersed stones generally less than 0.3m across,
with a well-preserved ‘kerb’ also observed in south-east quadrant, although on the basis of later
excavations this could have been bedrock appearing at the surface on the edge of a surrounding
ditch. Mound comprised 0.3m of loose sand and stone, above hearth of charcoal and burnt
bone, 5075mm thick, 4m in length by 22.75m wide. Hearth overlay 50 75mm of compact
reddish sand, resting on more than 0.3m of compact brown sandy subsoil.
SKELETAL REMAINS
Incomplete adult skeleton, sex unknown, including small pieces from the shafts of various long
bones, a number of small pieces of cranial vault, some rib and vertebral fragments and a distal
femoral fragment. Also fragments of burnt animal bone, possibly including a small dog and a
sheep. DAJ 7676 (1956), 55.
A small quantity of bone from Mound 11 surviving in Derby Museum was submitted to
Jacqueline McKinley for re-analysis. She noted that it was well preserved, clearly having been
42 THE ANTIQUARIES JOURNAL
in a very different burial environment from Mounds 50 and 56, but was unable to determine
the sex. The surviving sample comprised 44.9g of a human adult and 11.5g of animal bone.
Examination of the cremated animal bone by Julie Bond and Fay Worley provided more detail
about the animals that were cremated and whether they were present as whole animals or as
joints of meat, and about the taphonomic history of the material.
The twenty surviving fragments of burnt animal bone contained the remains of a horse, a dog
and a sheep or goat. The dog remains may all come from the left hind foot of the animal with
the identification of a left astragalus and calcaneum and an un-sided metapodial fragment. The
few pieces of bone remaining might suggest a partial offering, perhaps a skin, but the uncertain
history of this assemblage makes this a dangerous assumption; it is more likely to be the result of
partial retrieval or preservation bias. Experimental studies have shown that these elements are
likely to be found in close association in the pyre debris and possibly separated from abdominal
and proximal limb elements of the skeleton.45 They could therefore be a distinct group in the
pyre debris and missed during collection of the rest of the material for disposal elsewhere. A
fragment of glenoid cavity from the right scapula of a sheep or goat was also identified. This may
be the remains of a food offering in the form of a joint of meat, though the possibility that the
whole animal was originally present cannot be totally dismissed. A further fragment of cranial
bone was only identifiable to the level of sheep- or dog-sized mammal. The occurrence of
this fragment does not clarify the form of inclusion for either the sheep or the dog. Horse was
identified from a possible fragment of a first phalanx and a scaphoid carpal fragment. There
was one further fragment, which could only be identified as from a large mammal cranium, and
Fig 13. Mound 11: wire-wool embroidery. Drawing: Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 43
which might even have been human. There were a further twelve fragments which could not be
identified to even an ‘animal size’ taxonomic class or skeletal element.
ARTEFACTS
1. Spade iron, reported as unidentifiable or missing in 1985.DAJ 7676 (1956), 45.
2. Silver-wire embroidery, two fragments (fig 13). Traces of carbonized fibres were found
between the strands of wire, indicating that it had been in contact with, and had perhaps been
attached to, a piece of textile. L28mm, W8mm. DAJ 7676 (1956), 45,52 3. Identified by
Elisabeth Crowfoot as O
¨senstich, of which there are ten fragments from Viking graves at
Birka where they are found as headbands (on women) or caps (on men).46 There are also six
other fragments from ninth- and tenth-century burials in Gotland and Sma˚ land.47 Silver- and
gold-wire decorative costume embellishments are also known from other possible high-status
Scandinavian contexts. At Peel Castle, St Patrick’s Isle, Isle of Man, two of the seven
accompanied graves contained small balls of silver wire interpreted as decorative tassels. The
graves are dated to the tenth century and are interpreted as belonging to leading members of
Fig 14. Mound 11: three iron nails. Drawing: Frances
Chaloner
44 THE ANTIQUARIES JOURNAL
the Viking community on Man.48 A gold-wire-work toggle was found in a grave in Cathedral
Green, Carlisle, suggested as being that of a high-status Norse incomer, dated 900 50.49 A
silver-wire applique
´was found in excavations around the church of St Mary Bishophill Senior,
York, and has been interpreted as coming from a burial.50 Finally, some of the individuals
buried on top of and around the charnel deposit at the nearby site at Repton had silver or gold
embroidery in their costume.51
3. Several pieces of twisted and corroded metalwork, possibly part of 2.DAJ 7676 (1956)45,53.
4. Three small iron nails (fig 14). (i) Head: D15mm; shank: L20mm; (ii) head: D15mm;
shank: L20mm; (iii) head: D10mm; no shank. DAJ 7676 (1956), 45 6.
Mound
12
STRUCTURE
Diameter: 68m; height at centre: 0.3m. North-west, south-west and south-east quadrants
excavated; discontinuous cairn, overlying earth mound no more than 0.23m above natural
subsoil. Shallow ditch, comprising series of irregular hollows on east, with traces of a kerb.
Two irregular disturbances dug down from top of mound. No trace of any interment, but V-
shaped ditch located below south-west quadrant.
Mound
13
STRUCTURE
Diameter: 1.53m; height at centre: 0.7m. Trench 5ft (1.5m) wide dug north south across
mound, with 5ft- (1.5m-) wide extension dug out to east. Steep and imposing cairn contained
large stones; no trace of a kerb. Posnansky suggested it was surrounded by ‘a discontinuous
ditch’, but this may relate to the presence of a causeway. No trace of any interment.
Mound
14
STRUCTURE
No details recorded; found to be natural.
Mound
15
STRUCTURE
Diameter: 6.7m; height at centre: 1.4m. North-west half excavated, with extension, 10ft (3m)
wide, in centre, and small trench in south-east quadrant shown on plan. Sparse covering
of stones; surrounded by a very shallow ditch, but no trace of a kerb. No indication of any
interment or hearth.
The 1955 excavations also revealed that Mounds 9and 12 overlay a substantial linear ditch
of V-section, up to c1m deep and c23m wide. Some twenty pieces of hand-made pottery
were found in the ditch fill under Mound 9, including two rim sherds which were drawn and
published by Posnansky as well as a number of fragmentary body sherds.52 Both rims appear to
represent bowls with slightly everted rims; re-examination of the sherds suggests that one has a
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 45
groove running round the outside of the rim c15mm below the lip. The fabric is gritty, with
inclusions of coarse angular quartz and occasional voids. The sherds are poorly fired, dark
grey-brown on the inside and buff to pinkish-brown on the outside. The finish is smooth to
slightly burnished. These sherds were examined in 1955 by Gerald Dunning, David Wilson
and J R C Hamilton but could not be dated.53 A recent assessment of the evidence for Anglo-
Saxon Derbyshire concludes that the sherds are unlike any Anglo-Saxon pottery from the
region and are probably prehistoric.54 However, the pottery also includes an out-turned wheel-
thrown Roman rim in a cream gritty fabric with micaceous flecks. Unfortunately, since this
sherd was not published in 1956, there must be some doubt about its association with the other
pottery.
The RCHME Survey, 1993
In 1992 English Heritage asked the RCHME to conduct an earthwork survey of the cemetery
in advance of proposed tree-thinning operations by the Forestry Commission. The survey was
carried out at a scale of 1:1,000 between February and March 1993. Its purposes were twofold:
to produce a detailed plan of the cemetery to replace that made by Dallman in the 1940s, and to
attempt to define the cemetery’s overall extent. The survey therefore took in the whole area of
Heath Wood in order to identify evidence of past land-use that might have affected the present
survival and distribution of barrows.
Before the survey only the fifteen mounds excavated in the twentieth century had been
numbered, leaving forty-eight of those recorded by Dallman without a unique identifier. The
existing numbering sequence was therefore extended to cover all the mounds shown on the
1946 plan plus others identified for the first time during the survey (fig 15). Of the sixty-three
barrows claimed by Dallman, one (Mound 14) had already been proven by excavation to be
natural. The survey threw considerable doubt on the correct identification of a further four
(Mounds 29,30,38 and 51) and failed to locate another two (Mounds 26 and 63) whose former
existence must also be doubted. The number of barrows correctly identified by Dallman,
therefore, is probably fifty-six. To these can be added three barrows (Mounds 28,42 and 45)
identified for the first time in 1993; some doubt must attach to the correct identification of a
fourth newly recorded mound (22). The most likely total of barrows in the cemetery is thus
fifty-nine. Of these, two (Mounds 9and 10) seem to have been totally destroyed following
excavation. Detailed descriptions of all barrows and discredited mounds are included in the
archive site report.55 Full details of the survey were published in 1995; therefore only an
outline summary is provided here.56
The fifty-nine barrows clearly cluster into four distinct spatial groups, with only a few
isolated barrows lying between these groups (see fig 15). The survey indicated that the current
barrow distribution is real and found no evidence for ploughed-out barrows or quarrying
close to the cemetery. The survey also appeared to show that some barrows in the cemetery
were constructed with an encircling ditch and some without. In the 1995 publication it was
suggested that the absence of a ditch was correlated with there being a burial in the barrow. Of
the eight empty mounds recorded through excavation, six had or possibly had ditches, whilst
of the six barrows producing evidence of interments none seems to have had a ditch. Posnansky
had suggested that the empty mounds represented cenotaph burials and this interpretation was
taken up in the 1995 paper, with the proposal that the cenotaphs might represent members
of the Viking Great Army who had been Christianized and were given inhumation burial
at Repton. However, at that stage this idea could only be advanced as a hypothesis. When the
46 THE ANTIQUARIES JOURNAL
opportunity for further excavation in Heath Wood arose in 1998, the issue of the cenotaph
mounds was therefore one of the questions chosen for examination.
EXCAVATIONS 1998 2000
Aims
Scheduled Monument Consent was granted to enable an evaluation of the site through the
excavation of three trenches: (i) 10 20m complete open area excavation of Mounds 50 and 56
in the main group (fig 16), to establish their stratigraphic relationship and to test the hypothesis
that cenotaph and cremation mounds could be distinguished on the basis of their earthwork
form; (ii) 220m slit trench to examine the V-shaped ditch recorded by Posnansky and
recover pottery and=or radiocarbon samples which would establish whether it was prehistoric
or Viking Age; (iii) c510m trench across Mound 21 in the north-eastern group to establish
if this group was contemporary with the other sections of the cemetery or if these were
prehistoric ‘foundation’ barrows.
In addition, it was hoped that the excavation would throw light on the degree of animal and
root disturbance of the barrows, and their general state of preservation.
Methodology
Excavation took place over three seasons: 624 July 1998 (site code HW98), 424 July 1999
(site code HW99) and 10 29 July 2000 (site code HW00), under the auspices of the
Department of Archaeology, University of York. It was directed by Julian D Richards and
Marcus Jecock, with funding from the Derbyshire Archaeological Advisory Committee, the
Society of Antiquaries of London and the University of York.
All sites were excavated using a combination of light and heavy tools, with sample sieving of
mound make-up to 10mm. A standard context recording pro-forma was completed for each
layer or cut feature. Context numbers were allocated in blocks to each area of the site, with the
first digit(s) of the context number also reflecting the area number. Plans were drawn at 1:20
and sections at 1:10. A full black-and-white and colour photographic record was also maintained.
Over the first two seasons all four quadrants of Mound 50 were excavated, designated
Areas 14. Baulks were initially maintained between each quadrant, leaving excavated areas
which measured c77m, 6.57m, 56.5m and 5.57m respectively. Areas 3and 4were
coterminous with Areas 5and 6of Mound 56, providing a continuous section line running
north east=south west across the centres of both mounds. Within the mound the cremation
hearth was split into sub-contexts, each 0.5m square, in order to facilitate sieving and analysis.
Three quadrants of Mound 56 were examined over three seasons, designated Areas 57.
Again, baulks were maintained, giving excavated areas which measured c44m, 55.5m and
45m respectively; the fourth quadrant (Area 8) was left undisturbed, not least because of the
presence of a tree growing on it.
The V-shaped ditch was examined over two seasons. In May 1998 a geophysical line survey
was conducted across the area between Mounds 9and 12 in an effort to pinpoint the ditch.57
This failed to reveal any linear features, probably because of a combination of the effects of
tree-root systems, light sandy soils and (it later transpired) the depth of the feature. In July
1998 a trench 1m wide c20m in length was excavated on an east west alignment across the
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 47
Fig 15. Extract from RCHME site survey of Heath Wood, Ingleby, including complete
barrow numbering system. Drawing: NMR, Crown Copyright 1995
48 THE ANTIQUARIES JOURNAL
Fig 16. Plan showing excavated areas: Mounds 21,50 and 56 and Areas 9and 10.Drawing: Marcus Jecock, based on NMR,
Crown Copyright 1995
49
area between the approximate positions of Mounds 9and 12. This was designated as Area 9.It
was extended to be 2m wide along a 5m length once the location of the ditch, running
perpendicular across the trench, was established. In 2000 an adjacent area, 68m, was
excavated as Area 10. Posnansky’s 1955 trench, c1.53m, was located within this area and
was re-excavated until the ditch was bottomed; two further sections, each 2m wide, were also
cut across the ditch within Area 10.
In the 2000 season, Mound 21 was examined by means of two trenches, 27.25m and
26m, located so as to provide a section 13.25m long running south east to north west across
the approximate centre of the mound. The north-west trench was designated Area 11; that
to the south east was Area 14 (leaving Areas 12 and 13 for any future investigation of the
other quadrants). A number of areas were also cleared of forest debris and topsoil in order to
investigate the line of earlier field boundaries and clarify their influence on the development of
the cemetery (Areas 14E, 15 and 16).
Mound
21
(figs
17
19
)
In both trenches a compact undulating surface of natural weathered sandstone with patches of
reddish sand (1111,1410) was encountered at a depth of up to 0.50.6m below the present
ground surface. Some shattered sandstone fragments rested on this surface, possibly the result
of root activity and natural frost action. Beyond the edge of the mound there was a spread of
clean natural yellowish sand (1407). At the south-east edge of the mound there was a hollow in
the underlying bedrock filled by a silting of grey-brown silty sand with a few charcoal
fragments and small pebbles at its base (1409). The charcoal flecks suggest burning in the area
at the time of deposition. The hollow lay to the south east of a jumble of large sandstone
fragments, which might represent the edge of the barrow; in which case the hollow could be a
remnant of the quarried ring-ditch.
The natural surface was overlain by a c0.5m thick deposit of orange-brown sand (1104,
1108,1405,1408). This layer was heavily disturbed by rabbit burrowing and substantial tree
roots (fig 18). It contained many pebbles and small sandstone fragments, and occasional
charcoal flecks, as well as intrusive post-medieval pottery from the overlying layers. Although
separate context numbers were allocated to this layer according to whether it was on or off the
raised mound area, it was impossible to distinguish between these layers in plan or section, and
there was no discernible barrow edge or surface.
The upper 0.1m below the present ground surface was consistently darker and more mixed
than the underlying sand and contained occasional small pebbles and small sandstone fragments
(1102,1107,1403,1404), although the horizon between these layers and the underlying material
was very indistinct. This layer had been badly disturbed and may represent an ancient
ploughsoil. In Area 11, three Romano-British sherds were recovered from this layer. Two are
in a grey ware with an iron-rich fabric; the third is a body sherd in classic Derbyshire ware.
An overall date range in the mid- to late second century to mid-third century AD has been
suggested recently for kilns at Lumb Brook, south Derbyshire, producing similar Derbyshire
ware.58 These sherds were abraded, however, and are undoubtedly residual. In Area 14, the
equivalent horizon contained a number of potsherds in a fairly fine orange fabric with an over-
fired semi-opaque oatmeal-coloured glaze, probably all from the same vessel. The potsherds
are no later than the seventeenth century and must pre-date the plantation of Heath Wood.
Beyond the south-east end of the mound there was a layer of larger sandstone blocks (1406).
These stones may relate to a bank defining a northsouth field boundary immediately to the
50 THE ANTIQUARIES JOURNAL
Fig 17. Mound 21 at end of excavation, looking north west towards centre of mound with
Area 14 in foreground and Area 11 and Mounds 16 25 in the background. Photograph:
author
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 51
south east of Mound 21. This was planned to the north but was not seen this far south in the
RCHME survey, although a surface exploration in Area 14E found traces of it. If so, it would
run approximately parallel to, and c100m to the east of, the bank and ditch excavated in Area
10, dated to the Iron Age (see below).
Overlying all of the excavated area was a thin layer, c50mm thick, of black humic soil
(1101,1401). This was concentrated in depressions around the mound but it was also present
over the entire mound. It represents decaying vegetable matter derived from bracken, leaves
and other forest vegetation, which has built up in the recent past and has not yet rotted down
into the underlying sand. At the south-east end of Area 14 a small spread of disturbed orange-
brown sand (1402) rested directly upon 1401 and appears to have been formed as upcast from a
nearby animal burrow.
Mound 21 was unusual in that no clear mound edge was visible in section, although the
surface topography indicated a clear mound, and there are traces of this, and of a possible
quarry ditch, in the profile of natural. The homogenous nature of the overlying layers, and the
even depth of subsoil within and beyond the mound, suggests that it had been subject to
ploughing. This must have taken place before Heath Wood was planted and the late medieval
potsherds found within it may be linked to this activity. There are traces of ridge and furrow in
the wood to the north and this may have extended southwards, encompassing all of Mounds
1625 although Mound 21, on the edge of the group, would have been particularly vulnerable
to encroachment by ploughing. Disturbance by ploughing may explain the absence of any sign
of a cremation hearth, although the two trenches would also have failed to detect any off-centre
Fig 18. Mound 21 illustrating extent of tree-root disturbance in Area 14.Photograph:
author
52 THE ANTIQUARIES JOURNAL
Fig 19. Mound 21: NW SE section across mound. A– A marks the junction of two trenches at the centre of the mound. Drawing:
Sven Schroeder
53
cremation deposit, as was found in Mound 56 (see below). The significant tree and animal
damage might also have removed traces of a burial deposit. What is clear is that there was no
prehistoric cremation vessel buried in the ground beneath the centre of the mound, and that
Mound 21 is therefore unlikely to be Bronze Age. If we accept that Mound 16 was one of those
tackled by Bateman that yielded iron objects (and it seems reasonable that he would investigate
those mounds adjacent to the carriage ride rather than those further into the wood) then we can
conclude that Mounds 16 25 belong to the Scandinavian cemetery, rather than representing
primary prehistoric barrows. It is less clear, because of later ploughing, whether the mounds
originally contained cremation hearths.
Mound
50
(figs
20
31
)
Beneath Mound 50 bedrock comprised keuper marl-bedded sandstone, sloping downwards
from north west to south east at an angle of 10 20 degrees from horizontal (122,218,408,418)
(figs 23,24 and 25); in some places it was intermixed with clean yellow-brown sand (417);
elsewhere there was a compacted layer of natural orange-red sandy clay and gravel (121,217).
On the south-east side of the mound this was overlain by a layer of brownish-grey silty sand
with a high humic content (219,409,413), perhaps representing a buried soil. In the area
around the mound the natural subsoil had been quarried to form a ring-ditch, in some places
cutting into bedrock itself (fig 20).
These deposits were overlain by a substantial deposit of orange-red sandy clay subsoil,
some 0.30.75m thick, with abundant rounded small and medium pebbles (120,406,414).
This appears to have been landscaped in such a way as to provide an artificial platform,
accounting for its varying thickness. North west of the mound and in the ring-ditch there were
equivalent layers of orange-red brown sandy clay subsoil (123,309) which had been retained in
elevated spines running towards Mounds 6and 7and appear to have been landscaped in order
to form two causeways leading on to Mound 50. In the centre of the mound a clean deposit
of reddish-orange sandy clay (116,215,412), up to 0.15m in depth, appears to have been
deliberately laid as part of the preparation for the construction of the pyre, providing a smooth
and level surface. This was reddened beneath the pyre, possibly as a result of heat action.
The cremation deposit (119,214,308,411) comprised an oval area, c34m, of blackened
sand and charcoal, in which fragments of burnt bone were embedded. This layer was up to 50mm
thick, and sat directly upon the clean sand levelling deposit. It was excavated in blocks of 0.5sq m
which were boxed in plastic tubs and then wet-sieved under laboratory conditions to separate the
bone and charcoal (fig 21). The cremation hearth was buried under a thick deposit of sandy clay,
up to 0.7mthickonthecentreofthemound(105=106,207=213,303=304,403=410). As well as
rounded pebbles and gravel, this layer contained a largely random tumble of large angular
sandstone blocks, concentrated towards the centre and base of the mound. The stones sloped
inwards towards the mound centre, suggesting dumping from the outside inwards, or earlier
disturbance in the mound centre, although there was no obvious trace of a cut to indicate the
latter. There were voids between the stones, and numerous animal burrows. Root disturbance
and contamination was found throughout the layer. In the north corner of Area 1an equivalent
layer of orange-red brown sandy clay make-up (115) for Mound 7was observed in section.
Within Area 1remains of a cow skull were recovered, close to the base of the mound, and
lying immediately above the level of the cremation hearth, but offset to one side (fig 22). A
horse tooth was recovered from the mound in Area 4but carbon dating yielded a calibrated
date of AD 1487 1640 (at 95.4per cent probability) (OxA 13077), suggesting it was intrusive.
54 THE ANTIQUARIES JOURNAL
A number of redeposited prehistoric flints were found in the mound make-up, as well as a
single small body sherd of quartz-tempered prehistoric pottery (sf101).
Around the base of the mound, and continuing up its sides at an angle of c45 degrees, there
was a series of interleaved layers of black humic material and reddish-brown sandy clay.
The basal layer generally comprised a black humic silty loam, up to 0.15m thick (118,306). It
contained many small rounded pebbles, which were concentrated around the base of the
mound, probably representing initial erosion and run-off of smaller stones from the mound
surface. The layer as a whole has been interpreted as an initial build-up of turf and topsoil over
the mound, although it is difficult to say whether this developed naturally or was deliberately
laid. To the north west it continued into the ditch of Mound 6(307). The layer was overlain
by dark brown sandy loam (117,209,305,405) which followed the same contours, and
may represent a buried soil and=or phase of erosion of the upper mound surface. This layer
was noticeably thicker around the immediate foot of the mound. At its base in two areas it
incorporated concentrations of sandstone fragments (216,407) which may have been eroded
Fig 20. Mound 50, with Mound 56 in left background. Excavation of Mound 50 is at the
level immediately above the cremation hearth and the stones in the mound make-up can
be seen clearly. Bedrock is protruding in the base of the ring-ditch in the foreground. To
the right, one of the causeways leading on to Mound 50 is visible; at the far right is the
spoil heap from the 1948 excavation of Mound 7.Photograph: author
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 55
Fig 21. Mound 50: plan showing cremation hearth with density of cremated bone.
Drawing: Marcus Jecock
Fig 22. Mound 50: on initial exposure of burial deposit with cow jaw. Photograph: author
56 THE ANTIQUARIES JOURNAL
from the mound at an early stage in its construction, and may even represent a trample of
construction debris in specific areas. It was in turn overlain by another discontinuous layer of
black humic silty loam, generally less than 50mm thick (111,208,302,404), representing an
upper turf layer which originally covered the mound and surrounding ditch, but had been
disturbed by animal and root action. These deposits did not extend over the top of the mound
from where they had presumably been eroded. A similar dark layer (114) was observed in
section overlying Mound 7.
In Area 1the original mound surface had been disturbed by a number of holes (108,110,
113). Although circular in plan, on excavation they invariably had a number of channels
running off them and are interpreted as the holes left by rotten tree trunks and their associated
root systems. The fills were generally reddish-brown sandy clay with small pebbles (107,109,
112).
The tree-root holes were generally sealed by a substantial layer of reddish-brown mixed
sandy clay (104,202,301,402), up to 0.2m thick, which covered the buried turf lines around
the lower slopes of the mound in all areas. This deposit represents erosion from the crown of
the mound mixed with some humic material, which had collected around the lower slopes and
filled up the ditch. A number of large unabraded sherds of post-medieval pottery of a late type
of Midlands Purple ware were incorporated in this layer, resting upon the mound surface,
including at least two vessels: a pancheon and a jar. The fabric is very high-fired and is nearly
stoneware. All the glazed sherds have glaze on the inside only, either a treacly dark brown, or
yellow and brown. The pottery is likely to be a local product that would not have travelled far.
The vessels are utilitarian storage or kitchen wares, probably dating from the late seventeenth
century, or more likely the eighteenth century. This would immediately pre-date the planting
of Heath Wood in the mid- to late eighteenth century, suggesting that the mounds were extant
to that date and then eroded fairly rapidly.
To the north of Areas 1and 2there were a number of layers which appear to have been
associated with the excavation of Mound 7in 1948 (fig 24). These included a layer of reddish-
brown sand and pebbles, up to 0.1m thick (102=212) which sealed a layer of black humic
material with matted roots (101). The first is interpreted as spoil from the 1948 excavation
trench, sealing the humic material that had been cleared from Mound 7. A pair of parallel
grooves (204,206), approximately 1.5m apart, ran from the Mound 7trench. They were filled
with black humic silty material (103,203,205), perhaps trample or wash-off from the excavation,
which has filled hollows left by vehicle ruts. They terminated in a subrectangular feature (211),
filled with orange-brown silty sand and stones (210), which cut into the underlying turf line.
The feature may represent the setting for a large boulder, possibly removed during the 1948
excavation. The fill may have come from spoil from the excavation of Mound 7. In Area 2an
area of black humic silt (201) in the depression round the mound had been cut by the wheel
ruts and so must represent more ancient erosion and run-off.
Finally, the whole area was covered by a general layer of loose dark brown humic material
containing large amounts of matted tree roots and bracken (100,200,300,400=401). This layer
was thin on top of the mound but up to 0.15m thick around its base. A line of wooden posts was
set into this layer, running northsouth across the mound and representing the remains of a
pheasant pen. There were also several tree stumps within the layer.
NON-FERROUS METAL OBJECTS
1. Sword hilt grip (fig 26). Fragment of decorated silver strip. The outer surface is decorated
with the simplified, severe, geometric forms that characterize the Northern Trewhiddle style.
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 57
Fig 23. Mound 50: E W section. Drawing: Sven Schroeder
58
Fig 24. Mound 50: N S section. Drawing: Sven Schroeder
59
Fig 25. Mound 50: plan showing ditch and causeway and Mound 7material to north. Drawing: Sven Schroeder
60
There is a Maltese-style cross in a circle at one end, a St Andrew’s-style cross in the centre
(with a small central cross) joined to make a diamond at the other end, with a small diamond in
the centre. The edges have fine notches along their length. The reverse surface is very uneven,
with a raised pointed ‘mound’ of metal in the middle. This uneven surface, if intentional, may
have helped it to key into an organic surface if attached like a mount, but could also be due to it
having started to melt under high temperature. The grainy X-ray image, the black shiny
surface and the uneven raised surface on the reverse all point towards this object having
undergone damage in a high temperature. Unfortunately, both ends are broken and missing
so it is not possible to say how long it was. There are also no surviving forms of fixing (for
example, rivet holes). The strip is curved at one end but it is unclear whether this is its original
form or whether it is damage that occurred subsequently, either during cremation or burial.
It is also bent out of shape along one edge and to a lesser degree at the other end. L27mm,
W8mm, T1mm. HW99,411,sf165.
This appears to be a decorative mount from a sword hilt, perhaps similar to that from the
River Witham, at Fiskerton near Lincoln, or that recovered more recently from the stream bed
at Gilling West, North Yorkshire.59 The silver strips on the hilts of both swords were inlaid
with niello, a black-coloured copper or silver sulphide. Unfortunately, no visible remains
of black inlay survive on this fragment. Webster dates the heyday of Trewhiddle style to the
middle third of the ninth century, up to c875.60 The use of Trewhiddle-style decoration might
suggest that this was an Anglo-Saxon sword but Hall suggests that the production of
Trewhiddle-style objects represents a continuation of pre-Viking styles among an artistically
conservative group of craftsmen in Anglo-Scandinavian York seeking to redefine a
Northumbrian identity.61 Its presence at Heath Wood is another indication of the wide
cultural links of those buried in the cemetery. Such swords also found their way back to
Scandinavia; Anglo-Saxon swords from burial mounds at Dolven, Grønneberg and Hegge in
Norway have similar lozenge-shaped Trewhiddle-style mounts on their pommels.62 Webster
suggests that these finds are likely to represent Norse Viking activity in the north of England,
rather than Danes in the south.63
2. Tiny stud with square cross-section and a small domed head (fig 27); tip broken and
missing; appears to have been plated. Head: D3mm; shank: L5mm. HW99,411,sf168.
Fig 26. Mound 50: sword hilt grip. Drawing: Frances
Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 61
3. Tiny stud with square cross-section and a small flat head (fig 27); appears to have been
plated with silver; similar to sf168. Head: D3mm; shank: L6mm. HW99,411,sf271. Both this
and the previous stud are too small to have held anything together but they may have been used
to attach cloth or leather to a wooden object. Silver and copper-alloy tacks were found in some
numbers in Winchester, occasionally in burials and mainly in late Anglo-Saxon contexts.64
4. Flat folded copper-alloy object which has been bent (fig 27). A round white shape with a
glassy appearance is encrusted in one corner. In a matrix of sand, charcoal and burnt bone.
L25mm, W14mm. HW99,308,sf128.
5. Fragment of curled sheet, heavily mineralized and covered in green copper corrosion
products with mauve staining (fig 27). L9mm, W6mm. HW99,308,sf166.
6. Solidified fragment of molten object (fig 27). Visual inspection suggests presence of silver
chloride, silver sulphide, copper and lead. XRF analysis confirmed high silver and tin peaks
with traces of copper and lead. L17mm, W12mm. HW99,411,sf131.
7. Solidified fragment of molten object (fig 27). Visual inspection suggests presence of lead
sulphide and silver. XRF analysis confirmed large silver peaks with traces of copper. D10mm.
HW99,411,sf141.
8. Solidified fragment of molten object (fig 27). Visual inspection suggests presence of copper
and silver chloride and possibly lead. XRF analysis confirmed a strong silver peak with traces
of copper and lead. L12mm, W7mm. HW99,308,sf154.
9. Solidified fragment of molten object; thin and brittle flat section connected with an
elongated round mass (fig 27). Visual inspection suggests this was a silver object, with traces of
silver chloride, silver sulphide and copper as a component of the silver alloy. XRF analysis
confirmed presence of large silver peaks with traces of tin and lead. L13mm, W6mm. HW99,
411,sf156.
10. Solidified fragment of molten object, possibly silver (fig 27). D7mm. HW99,411,sf160.
11. Solidified circular fragment of molten object, possibly silver (not illustrated). D6mm.
HW99,411,sf159.
IRON OBJECTS
Where possible measurements have been taken from the actual object as revealed on the X-ray,
ignoring corrosion products.
1. Hinge pivot (fig 28). Shank and guide arm both of rectangular section. Surface blistering.
L63mm, W9mm, T6mm; guide arm: L36mm. HW99,410,sf110. Hinge pivots were generally
used to hang doors or shutters, the shank being driven into the wall or jamb, and the guide arm
bring slotted into the hinge eye. This example is very similar to one from Fishergate, York.65
2. Small knife (fig 29). Broken; complete tang and the beginning of the blade covered by
organic material, perhaps horn. Blistering, especially of the tang. L48mm, W11mm, T5mm.
HW99,410,sf113. Too little of the blade survives to enable the identification of form. The
length is incomplete but in any case falls within the major concentration of lengths (4585mm)
measured from the sample of 128 Anglo-Scandinavian knife blades from Coppergate, York.66
It is unlikely to have been a weapon or hunting knife, but would have been suitable for a wide
variety of domestic or craft activities.
62 THE ANTIQUARIES JOURNAL
Fig 27. Mound 50: other non-ferrous objects. Drawing: Sven Schroeder
Fig 28. Mound 50: hinge pivot. Drawing: Sven Schroeder
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 63
3. Large nail with square head tapering down into the shank (fig 30). Dimensions are
approximate because of the poor conservation state of the surface. Head: W11mm, T9mm,
H7mm; shank: L20mm, W5mm, T4mm. HW99,209,sf112.
4. Large nail in two fragments: a head and a shank fitting to it (fig 30). Head is probably
square, tapering down into a shank of square section. Same type of nail as sf112. Head:
W11mm, T9mm, H8mm; shank: L28mm, W4mm, T4mm. HW99,118,sf114.
5. Probable nail head, extremely corroded (fig 30). W11mm, T10mm. HW99,308,sf149.
6. Small nail, heavily oxidized (fig 30). Head: D8mm; shank: L16mm. HW99,304,sf104.
7. Small nail, shank bent, very oxidized (fig 30). Head: D8mm; shank: L18mm. HW99,410,
sf105.
8. Small nail with round flat head and square shank (fig 30). Head is plated. XRF analysis
confirmed the presence of a large amount of tin and a trace of lead. Head: D8mm; shank:
L20mm. HW99,304,sf120.
9. Small nail, very oxidized; original surface of head lifted up (fig 30). Head: D8mm; shank:
L18mm. HW99,308,sf127.
10. Small nail, bent giving a circular shape; very oxidized, cracks and original surface lifted up
(fig 30). Head: D10mm; shank: 0.25mm. HW99,411,sf129.
11. Small nail, very oxidized; seems to be plated with non-ferrous metal (fig 30). Head:
D13mm; shank: L20mm. HW99,411,sf130.
12. Small nail, end of shank missing; very oxidized, original surface lifted up; possible plating
on head (fig 30). Head: D10mm; shank: L13mm. HW99,308,sf132.
Fig 29. Mound 50: knife blade. Drawing: Sven
Schroeder
64 THE ANTIQUARIES JOURNAL
Fig 30. Mound 50: nails nos 317.Drawing: Sven Schroeder
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 65
13. Small nail, with square shank and round head, flat or slightly domed; completely oxidized;
plated with non-ferrous metal (fig 30). XRF analysis revealed the presence of tin with a tiny
trace of copper. Head: D10mm; shank: L18mm. HW99,411,sf134.
14. Small nail, very oxidized; plated with non-ferrous metal (fig 30). Head: D11mm; shank:
L18mm. HW99,411,sf142.
15. Small nail, with incomplete broken shank; round flat head and square shank; completely
oxidized; plated with non-ferrous metal (fig 30). XRF analysis confirmed the presence of tin.
Head: D12mm. HW99,411,sf143.
16. Small nail with domed head, very oxidized (fig 30). Head: D8mm; shank: L17mm. HW99,
308,sf145.
17. Small nail with domed head and square shank, completely oxidized; plated with non-
ferrous metal (fig 30). XRF analysis confirmed the presence of tin traces. Head: D10mm;
shank: L22mm. HW99,308,sf147.
18. Small nail, very oxidized; probably plated (fig 31). Head: D11mm; shank: L21mm. HW99,
308,sf148.
19. Small nail, end of shank bent; very oxidized (fig 31). Head: D8mm; shank: L20mm.
HW99,411,sf158.
20. Small nail with square cross-section with tip broken off; small flat head, with non-ferrous
plating (fig 31). XRF analysis confirmed the presence of a large amount of tin. Head: D9mm;
shank: L11mm. HW99,411,sf162.
21. Possible small nail head, heavily corroded (fig 31). D10mm. HW99,410,sf118.
22. Small nail fragment, with square cross-sectioned shank and circular flat head; completely
oxidized (fig 31). Head: D11mm. HW99,415,sf126.
23. Small nail, incomplete, shank broken; very oxidized, head completely mineralized (fig 31).
D9mm. HW99,411,sf157.
24. Part of nail shank, completely oxidized (fig 31). L16mm. HW99,308,sf146.
25. Fragment of nail shank with tip, very oxidized (fig 31). L14mm. HW99,308,sf151.
26. Fragment of nail shank, very oxidized (fig 31). L11mm. HW99,308,sf152.
27. Fragment of nail shank with tip, very oxidized (fig 31). L12mm. HW99,308,sf153.
28. Small and completely oxidized iron fragment (fig 31). HW99,308,sf155.
29. Fragment of ?nail shank, square cross-section, narrowing to a point; heavily corroded
(fig 31). L16mm. HW99,411,sf163.
Like the nails from Mound 7, the majority of the Mound 50 nails are flat-headed, although
the large examples numbers 35appear to have large solid heads, and others such as 11,14
and 15 appear to be dome-headed. The presence of tinning and possibly soldering on some
nails suggests a decorative role. Nearly 2,200 nails and tacks were recovered from Anglo-
Scandinavian contexts from Coppergate, York, including 1,300 with flat heads, but only 44
with tin-plated heads. Of the complete nails recovered from Coppergate, 65 per cent are
66 THE ANTIQUARIES JOURNAL
Fig 31.Mound50: nails nos 1829 and iron clamps nos 30 34.Drawing: Sven Schroeder
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 67
3065mm long, whereas the Heath Wood nails tend to be smaller on average.67 Ordinary flat-
headed nails were in common usage in the ninth century, but were probably used for a wide
variety of purposes. Nails were not seen joining the timbers in any of the Anglo-Scandinavian
buildings excavated in York, and the majority were probably used in furniture. Surviving
Viking Age chests and boxes were usually jointed together but their lock plates, hinge
fittings, corner brackets and bindings were nailed on.68 Examples of nailed chests of the
period include two from the Oseberg ship as well as the Ma
¨stermyr tool chest.69 Coffin nails
of 1520mm in length were found in the cemetery associated with the Viking fort at Fyrkat.70
An adult grave from St Patrick’s Isle, Peel, was placed in a wooden chest or coffin
represented by seventeen nails. These were c50mm in length but six smaller nails secured
the sides to the base of a box that formed a coffin in a child’s grave.71 Tinned nails were used
to attach the hinges and lock of a small box from one of the tenth-century Fyrkat graves, and
were found on a casket from a contemporary grave at Sønder Onsild.72 The function of these
nails is discussed further below but they are clearly derived from several objects, including
a possible chest that may have been used as a coffin container, as well as decorative tacks
possibly used in the construction of shields. Some may simply have been attached to wood
used as fuel for the pyre.
30. Iron clamp with two rivets (fig 31). The cleaned rivet shows a round head (D4mm) which
was probably decorative, while the shank situated between the clamp is functional and of round
shape. The surface reveals some parallel lines, which could be decorative or keying for plating.
L22mm, W20mm, T8mm. HW99,308,sf135.
31. Several fragments of iron clamp in very bad condition, with one rivet (fig 31). L24mm,
W15mm, T8mm. HW99,308, sfs136,137,138,139,140.
32. Two fragments of iron clamp, each section with a rivet; very oxidized (fig 31). L23mm,
W20mm, T8mm. HW99,411,sf144.
33. Iron clamp, with a single rivet, oxidized (fig 31). L21mm, W8mm, T7mm. HW99,308,
sf150.
34. Small fragment of iron clamp, with traces of a single rivet (fig 31), possibly related to
sf135.L10mm, W8mm. HW99,308,sf167.
A minimum of four iron clamps were present in the pyre. They appear to represent metal
clamps from a shield rim. Continuous gutter-shaped metal-edge bindings were obsolete by the
Viking Age but small clamps are occasionally found in graves. There are twelve cases from
Birka and a single example from Lindholm Høje.73 These may be made of iron or bronze and
are generally c20 25mm square. They are sometimes distributed evenly around the shield rim
to fasten a leather edge binding, but may also have fastened the joins between the planks. The
shield itself is likely to have been hemispherical and up to 1m across; it would have been made
of wooden boards, probably covered in leather which could have been painted. A central iron
boss would have been attached to the shield by nails, and would have protected the hand grip at
the rear. Bosses were usually attached by broad-headed nails, the points of which were either
bent over or flattened on the reverse of the shield. At Birka four is the most common number;
there are occasionally six (as in the case of Gokstad), and five were sometimes used, as at Cronk
Mooar.74 Several of the nails recovered are of a type appropriate to having performed this
function. The absence of the shield boss from the pyre suggests that, like the sword blade, it
was collected after the cremation.
68 THE ANTIQUARIES JOURNAL
Mound
56
(figs
32
38
)
Bedrock (613,718) was encountered in two quadrants, where it comprised alternating layers of
sandstone and keuper marl sloping gently from north west to south east. Even at this depth
there was evidence of root activity along the bedding plane. Lying directly on the bedrock was
a sand and gravel natural subsoil (607=612,717,722), although this had been quarried away in
the area around the mound.
In each of the three excavated quadrants the natural subsoil was overlain by a layer of
orange sand, up to 0.5m thick, containing small sandstone or quartzite pebbles (509,614,719).
This clean deposit appears to have been a deliberately dumped layer overlying natural,
presumably to provide a level surface for the cremation (figs 32 and 33).
On the eastern edge of Area 5this was overlain by a thin spread of blackened silty loam
(508), c1.11.2m, and 10 50mm thick, containing a few fragments of burnt bone and a
ringed pin (sf245); the latter within a small scoop, c0.1m across by 0.2m deep, containing
blackened sand (510). The sides were not crisply defined and it sloped at an angle, suggesting
the cremation had been disturbed by animal or root activity. The cremation deposit (615) was
also just visible in section in Area 6, as a lens 3mm thick.
Fig 32. Mound 56 from the north, showing ditch and causeway, and location of cremated
bone on mound surface at lower left. Photograph: author
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 69
Fig 33. Mound 56: plan after excavation showing area of cremation. Drawing: Sven Schroeder
70
The mound above the cremation deposit comprised a layer of orange-brown sandy loam
(507,602,704) surviving up to 0.5m thick at the centre of the mound. This material contained
angular sandstone slabs and large quartz pebbles, with very occasional charcoal flecks. On the
western edge of Area 7there was an equivalent layer (713) representing the make-up of the
adjacent Mound 55 (fig 34).
In the ditches encircling the mound there was a succession of interleaved layers of dark
orange-brown sandy silt (506,606,609,611,710,711,715,716) and black humic silty loam
(503,505,605,608,709,712,714). These are believed to represent initial erosion of the
surfaces of Mounds 55 and 56 into the surrounding ditch and alternating turf lines, prior to
their consolidation (fig 35). The turf layers were very discontinuous, having been disturbed by
roots and animal burrows, and having been eroded away completely from the upper surfaces of
the mound. To the north-west corner of Mound 56 there was evidence for a causeway, where
an area of higher ground had been left across the ditch in order to provide access to the mound.
A clay pipe bowl (sf 119) was found resting on the uppermost of these layers, confirming that
after consolidation the mounds were then relatively stable until the post-medieval period.
At some stage there had been substantial erosion of the upper mound that filled the
surrounding ditch with a reddish-brown sandy silty loam (501,603=604,707)containingquartz
Fig 34. Mound 56 from the south west, with Area 7in foreground after exposure of
original mound surface on lower slopes. Photograph: author
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 71
pebbles and occasional larger angular sandstone blocks. This layer also incorporated sherds of
the same type of Midlands Purple ware as those found in the equivalent layer over Mound 50,
again suggesting that erosion of the mounds only took place after the creation of the plantation in
the eighteenth century. In Area 5there were two irregular scoops in the mound surface filled
with dark brown humic material (502,504), representing tree holes or rabbit burrows. In Area 7
the erosion layer (720,721) was darker on the lower slopes. It also merged imperceptibly into the
equivalent layer of dark reddish-brown sand (706) which had eroded from Mound 55.
A spread of fragmented sandstone blocks and occasional quartzite pebbles (703=708) was
found around the western side of Mound 56. A layer of fine dark brown sandy silt loam (705)
lay between and under the stones, incorporating further sherds of Midlands Purple type,
dating the stone capping to the late seventeenth or eighteenth centuries. Rather than being
related directly to Mound 56, therefore, it appears that the stone capping may reflect field
clearance from pre-plantation ploughing adjacent to the main group of mounds. Mound 56 lies
at the edge of this group and stones disturbed by ploughing may have been conveniently
thrown on to its surface.
Above the stones there was a general spread of brownish-grey sandy loam (702), along the
edge of Area 7. This may represent a pre-forestry soil, possibly an agricultural ploughsoil at
Fig 35. Mound 56, showing profile and build-up of eroded material downslope in north-
facing section in Area 6.Photograph: author
72 THE ANTIQUARIES JOURNAL
Fig 36. Mound 56: E W section. Drawing: Sven Schroeder
73
Fig 37. Mound 56: N S section. Drawing: Sven Schroeder
74
the edge of the barrow cemetery. Finally, the whole area was covered by a general layer of loose
dark brown humic material containing large amounts of matted tree roots and bracken (500,
600=601,700=701). This layer was thin on top of the mound but up to 0.15m thick around its
base.
COPPER-ALLOY OBJECTS
Plain-ringed, loop-headed pin (fig 38). The shank is of circular cross-section with three
decorative vertical grooves, just below the head. The head has been curled around to form a
loop for the ring. The ring itself is of flat oval cross-section. It is decorated with four bands of
parallel triple grooves radiating outwards. The pin is in poor condition; it was found in two
pieces lying a short distance apart and as part of the surface is missing it is impossible to rejoin
the whole pin or to be sure of its original length. The X-ray shows that extensive damage has
Fig 38. Mound 56: ringed pin. Drawing:
Frances Chaloner
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 75
occurred on the shank, with internal cracking and disruption of the surface below the loop.
One broken edge has blobs of copper metal spilling over the edge on to the outside surface. It is
unclear whether this is redeposited metal resulting from corrosion processes or molten metal
through exposure to high temperatures. Ring: 22mm; shank: L100mm. HW00,508,sf245.
The type appears to have been adopted by Norse settlers during the ninth century and is
known from burial and settlement sites.75 There are two examples from tenth-century
graves on St Patrick’s Isle, Peel, on the Isle of Man, where they are interpreted as closing
shrouds.76 Three are known from the coastal trading site at Meols, and it is assumed that
they relate to the Hiberno-Norse phase in the early tenth century.77 One is known from
the church at Brigham, Cumbria,78 andanirononewasexcavatedfromAllSaints
Pavement, York.79 There are a further sixteen from Scotland, seven of which are from
burials. Sixty examples are known from Viking Age levels in Dublin, the majority dated
AD 92575.80 There are numerous examples from Scandinavia itself, including Denmark
and Sweden, but especially from Viking graves in Norway,81 where they are thought to
represent the adoption of an Irish fashion in dress pins, probably post-840 when the
permanent settlements had been established in Dublin. Fanning suggests that a grooved
looped-over pin-head is characteristic of Scandinavian workmanship.82 According to Petersen,
the plain-ringed loop-headed form of ringed pin occurs in both male and female Viking graves
in Norway.83 The latter belong mainly to the ninth century, whilst the pins from male graves
are datable to the ninth and tenth centuries. It has been demonstrated that ringed pins would
have functioned as dress-fasteners, fastening the cloak at the shoulder or across the chest.84
Cremated Human Bone
Cremated bone from eight contexts from Mounds 50 and 56 was received for analysis. The
main cremation deposit contexts were subdivided into 0.5m squares and these subdivisions
were maintained within the osteological analysis to allow more detail of the distribution of
the cremated bone to be ascertained. Osteological analysis followed the writer’s standard
procedure for the examination of cremated bone.85 Age was assessed from the stage of skeletal
and tooth development86 and the general degree of age-related changes to the bone.87 Sex was
ascertained from the sexually dimorphic traits of the skeleton.88
Disturbance and Condition
The excavated deposits may have been disturbed by animal and root activity (see above),
although there is unlikely to have been much loss of bone from the deposits by these mechanisms.
The condition of the bone was generally rather poor. The surface morphology was often ill
defined as a result of erosion reflecting the microenvironment chiefly influenced by the
groundwater and sediment chemistry.89 The condition of the bone from the different contexts
within Mound 50 varied slightly, that carried by worm or root action into the pre-pyre
levelling and mound make-up being moderately heavily eroded, whilst that from the main
deposit of cremated material was only slightly moderately affected. The latter was additionally
obscured by an adhering substance that appears to comprise mineralized fuel ash and soil
forming a hard concretion on and around some of the bone. The poor quality of the surface
morphology had a detrimental effect on the quantity of bone fragments it was possible to
identify to skeletal element, a maximum of 30 per cent of the assemblage from context 410
being identified in detail. The bone from Mound 56 was moderately eroded.
76 THE ANTIQUARIES JOURNAL
A further observation is that very little trabecular bone survived, a total of only 29.9gof
axial skeletal elements being identified, and there were also very few fragments of articular
surface. It is known that bone porosity has a major effect on its survival and it has been
demonstrated that trabecular bone is the first to be lost in acidic soil conditions, often
crumbling to dust as it is excavated.90
Demographic Data
The remains of a minimum of two individuals from Mound 50 and one individual from Mound
56 were identified from amongst the cremated bone (table 1). Poor bone survival limited the
demographic detail it was possible to obtain, there being insufficient evidence to suggest the
sex of the remains from Mound 56, whilst those of the adult from Mound 50 are probably
female. Comparative data from other British Viking cremation burials are relatively scarce, but
both male and female adults were identified amongst the remains from the other Heath Wood
mounds excavated in the 1940s and 1950s, though there is no record of any immature
individuals being recovered either here or in any of the other rare cremation burials of this
date.91 It is not impossible, given the very poor condition of the bone at Heath Wood, that
immature bone may have suffered preferential destruction because of the burial environment,
or have passed unrecognized owing to the poor morphological definition of the bone. No
pathological lesions were observed.
Aspects of Pyre Technology and Ritual
All the surviving bone was the white colour indicative of full oxidation of the organic
components of the bone.92 It should be noted, however, that trabecular bone is often amongst
the last to be oxidized (being internal and often with high soft tissue coverage) and as very little
of this survived we may have a biased representation.
The weights of bone recovered varied greatly between Mounds 50 and 56.The198.1gof
bone (predominantly human) from Mound 56 represents a maximum of c20 per cent of the
total weight of bone one would expect to find at the end of an adult cremation.93 The 3,459.5g
of bone from Mound 50, the vast majority of which (98.6per cent) was from the in situ pyre
site, represents a substantial quantity of bone, probably in excess of that expected from the
cremation of the two human individuals identified. A considerable quantity (table 1) of the
bone from this mound was identified as animal in osteological analysis and it is not unlikely
that more amongst the unidentifiable material may be of animal origin rather than human. In
this instance, it is likely that the pyre site also formed the place of burial for the human remains
Table 1. Summary of results from osteological analysis
Mound Deposit types
Total bone
weight Age and sex Animal bone
50 Pre-pyre
levelling; ?in situ
pyre and burial;
mound material
3,459.5g1) Adult c18 45 yrs,
?female; 2)
?infant=juvenile
544 fragments (min
284.4g): horse, dog, pig,
sheep=goat, cattle
56 ?In situ pyre site
and burial
198.1g Adult c18 40 yrs,
unsexed
1fragment (0.2g): pig
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 77
and that of their pyre goods (the cremated animal bone). There is no published record of
the weights of bone recovered from the other Ingleby mounds, nor from the few other
contemporaneous British burials.
The majority of bone from most contexts was recovered from the 5mm sieve fraction,
4366 per cent, with most from the 10mm sieve fraction in only two cases (Mound 50, contexts
304 and 410 at 60 per cent and 48 per cent respectively). The maximum fragment sizes were
also relatively low, with a maximum of 49mm. A number of factors may affect the size of
cremated bone fragments, the majority of which are exclusive of any deliberate human action
other than that of cremation itself.94 In this instance, the burial environment clearly
detrimental to bone survival will have increased the level of fragmentation along the lines of
the dehydration fissures formed during cremation. However, whilst there is no conclusive
evidence for deliberate fragmentation of bone, and much to suggest the opposite, it cannot be
totally dismissed. There are no extant comparative data from the other excavated deposits
within the cemetery or from contemporaneous burials elsewhere in Britain.
The quantity of different skeletal elements recovered is heavily biased by differential bone
survival (see above). Poor surface morphology also makes it difficult to distinguish individual
long bones and frustrates the normal ease with which skull fragments may be recognized.95
Meaningful comment is not really feasible, other than to observe no unexpected absences of
skeletal elements.
The distribution of bone from across the main deposit of cremated material (i.e. the pyre
site) in Mound 50 was assessed to try and ascertain the formation process of the deposit (see
fig 21). The majority of the bone (59 per cent) was recovered from Area 4(context 411), a
substantial proportion (37 per cent) being recovered from Area 3(context 308)andonly
c2per cent from Area 2on the north-east margins. Skeletal elements from all areas of
the adult were represented across all parts of the pyre site, though there were apparent
concentrations within a 11.5m spread in the western half of the deposit and a 10.5m
spread in the east. The identifiable fragments of the immature individual were confined to a
11m spread in the western half of the deposit, overlapping with the concentration in the
adult bone. The animal remains were similarly spread across the whole of the pyre site, with
apparent concentrations in a 11.5m area in the eastern half. The general impression is of a
relatively random distribution of both individuals and skeletal elements across the entirety of
the main deposit of cremated material. Given the way in which a pyre burns and collapses,96
the deceased and their pyre goods generally maintain their horizontal position on the pyre
relative to one another. If this deposit represented the site of a pyre that had been allowed to
burn down and then had a mound raised over it to form the burial, the various remains
should have shown an ordered anatomical and spatial distribution. An example of the latter
may be seen in the contemporaneous Vendel warrior burial from Vallentuna.97 That this is
not the case in the Mound 50 pyre site demonstrates some deliberate human manipulation of
the remains subsequent to cremation and prior to burial. The implied mixing of the cremated
remains may have resulted from a thorough raking-over of the pyre site to break up any
incompletely cremated soft tissues and facilitate their full oxidation, an action which would
also have led to increased fragmentation of the bone itself. Though it appears that the
manipulation involved a substantial movement and mixing of bone fragments from across the
pyre, the relatively limited distribution of the immature remains suggeststhisindividualwas
placed on the western portion. Analysis of the animal remains confirmed that the several
species were distributed randomly across the pyre site, again suggesting disturbance of the
remains. (JMcK)
78 THE ANTIQUARIES JOURNAL
Cremated Animal Bone
The cremated animal bone was also highly fragmented, the majority of pieces being in the
1020mm size range, although many fragments were under 5mm. The largest fragments were
c30mm long. This report covers all the fragments thought to be non-human in origin; due to the
nature of the material, a small number of fragments could not be definitely assigned to either
human or non-human categories by McKinley. The material was extremely hard to identify even
by the standards of most cremated bone, a fact reflected in the various levels of uncertainty
recorded in the identifications. Nevertheless, a range of animals was identified; horse, dog and
sheep or goat in Mound 11, pig in Mound 56, and horse, dog, pig, sheep or goat, and cattle in
Mound 50 (table 1). It is also possible to say something about whether these animals were whole
carcasses or were simply joints of meat or token offerings; this is discussed further below.
Most of the fragments were fully oxidized, light grey or white in colour, with the exception
of a few fragments, notably dog phalanges. Experimental studies have shown that these distal
limb elements can become detached from the carcass fairly early in the cremation.98 They may
then fall into the pyre debris away from the main concentration of heat, resulting in less
complete combustion and in some cases better preservation. Such preservation, therefore, is
not indicative of the animals, or particular elements of them, being put on the pyre at a late
stage, or that they were necessarily particularly near to the edge of the pyre.
The fragments were carefully scrutinized for butchery marks and any pathological features,
such as have previously been found in other cremations. A horse axis vertebra from the Anglo-
Saxon cemetery at Sancton, Humberside, showed marks of decapitation whilst other bone
from cremation cemeteries has produced evidence of the dismemberment and skinning of
animals and the presence of animals with significant pathologies being used as offerings.99 No
butchery marks or pathologies were found on the Heath Wood bones.
The identified fragments were weighed for each context. It is obvious from these weights
that only a small fraction of the original bone was present, and that even the largest assemblage
of animal bone, from Mound 50, is only a fraction of the bone that might be expected to survive
from this number of animals.
Identifications
Mound 50 produced the greatest mass of animal bone from the assemblage: 544 fragments
were found, weighing a total mass of 284.4g. The mound contained the cremated remains of a
horse, a dog, a pig, a sheep or goat and possibly an ox; there was no evidence for more than one
of each animal. The majority of the fragments came from the horse and the dog.
The horse was represented by fragments of lower limb bones and the cranium, with
elements from the right and left sides and from the fore and hind quarters of the body.
Positively identified fragments came from the occipital condyle region of the cranium, the
petrous bone from the cranium, left radius=ulna (fused) and tibia, carpals (three fragments:
scaphoid, navicular and cuneiform), astragalus (two fragments, one left-hand side), proximal
and distal sesamoids (five fragments), metapodia (three fragments, one from a right metatarsal), a
lateral metatarsal and the first, second and third phalanges (five fragments of the first phalanx).
There were also fragments of longbone diaphysis, tooth socket, cranial and indeterminate
horse-=cattle-sized bone. This pattern of identified fragments, with elements from both sides
of the body, the head and limbs, suggests that all of the horse was present on the pyre. There
are no butchery marks to indicate that the head or other limbs were removed before burning, as
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 79
in the case of the Anglo-Saxon horse at Sancton, or that the animal was divided into pieces
before being placed on the pyre, as in Ibn Fadlan’s famous account of a boat burial amongst the
Rus.100 The fused radius=ulna fragment suggests that the animal was over three years old at
death. No estimation could be made of the sex of the animal or its stature.
The identified dog bones from Mound 50 come from all regions of the skeleton and from
the left and right sides of the body. The left third and fourth premolar region and the right
third molar region of the maxilla and a fragment of mandible were present. All teeth had
erupted, suggesting an age at death of over six months. The axis, atlas and a caudal vertebra
could also be identified. From the forelimbs it was possible to recognize a fragment of scapula
blade and right humerus, both right and left ulnae, right radius, the left and right third
metacarpal and a fragment of third or fourth unsided metacarpal. From the hind limbs a
fragment of acetabulum from the right pelvis, left tibia, astragalus and calcaneum were
identified. The assemblage also contained a further metapodial fragment, a second phalanx,
five first phalanges and two further phalanges, first or second, which may have come from hind
or fore limbs. The generic morphology of some skeletal elements resulted in several fragments
being distinguished only as coming from a medium-sized mammal although it is likely that
they belong to the same dog. These include caudal and thoracic vertebrae, rib fragments, tooth
roots, cranial, petrous and mandible fragments, ossified rib cartilage, cranial fragments,
humerus fragments and indeterminate longbone diaphysis fragments. Fusion of the distal
radius suggests that the dog was over eleven months old at its death. Although unpredictable
shrinkage due to cremation precludes metric analysis of these fragments, the dog appears to
have been roughly of medium size, the bones being a little smaller than a reference specimen,
which in life stood c0.55m tall at the shoulder.
Mound 50 also contained the remains of a pig. This animal was represented by only four
fragments but these were all identified as coming from the left-hand side of the animal,
introducing the possibility that only the left side was originally present as a food offering. The
elements were a left astragalus and acetabulum from the hind limb and left ulna and humerus
from the fore limb.
Two fragments of burnt bone were probably from a sheep or goat (a left distal tibia and
lumbar vertebrae) and two fragments were probably cattle (a metatarsal and mandible). They
are both from skeletally diverse areas of the skeleton so it is unlikely that they represent a
partial carcass animal offering. It is possible, given that only two fragments of each animal bone
were identified, that they were unintentional inclusions in the assemblage, incorporated into
the assemblage from surrounding deposits, but it may simply be that the rest of the cremated
bone from these animals has been removed.
Mound 50 also contained three very small bone fragments that cannot be distinguished
morphologically as either bird or mammal. These fragments could be from the dog but there
remains the slight possibility that they could indicate the presence of a bird on the pyre.
Mound 56 produced only one fragment (0.2g) of burnt animal bone; a pig left fifth
metatarsal. The specimen was fully oxidized and its size suggested that it came from a small
individual. Nonetheless, the fused proximal epiphysis indicates that the animal was over
21
4years old at death.
Interpretation
The presence of several different animals in Mound 50 confirms earlier studies and extends the
number of identified animals. Even allowing for the very comminuted nature of the cremated
80 THE ANTIQUARIES JOURNAL
bone, which made identification difficult, it was possible to show, in the case of Mound 50, that
a whole dog and horse had probably been present, along with at least one side of pork and
possibly parts of a sheep or goat and an ox. Even if the unidentifiable material is included, the
weight of bone present is insufficient to account for the rest of these carcasses, there being only
c300g in total. It is suggested elsewhere in this report that human bone and artefacts have been
removed from the pyre sites and the animal bone evidence implies that the animal remains
were also intentionally gathered up.
Animal bone was also identified amongst four of the five pyre sites previously excavated in
Heath Wood, including sheep, ox and possibly dog from Mound 1, ox and possibly horse from
Mound 5, and ox and possibly sheep and pig from Mound 6. The inclusion of animals on
cremation pyres, either as entire cadavers or jointed pieces, has formed a consistent element of
the early medieval cremation rite, the quantity and range of species tending to increase over
time.101 A similar pattern has been observed in Scandinavia, including Sweden, culminating
in c80 per cent of Viking burials containing cremated animal remains. These included up to
ten species per burial amongst which dog, horse and sheep or goat were the most commonly
occurring species.102 There are few parallels from Viking graves in Britain, the most famous
being perhaps the burials at Ballateare and Balladoole, Isle of Man.103 The crucial difference is
that the Manx mounds covered human inhumations, and the cremated bone, which seems to
have been entirely animal (including horse, dog, ox, sheep and cat), was deposited as a layer
towards the top of the mound.
The number of dogs and horses in the Heath Wood cremations is an interesting feature.
Horse and dog inhumations and cremations are known from all over Europe, from the Iron Age
to the Viking period, and there is often a strong correlation between the two animals.104 Perhaps
the earliest account of a similar practice is in Tacitus’s Germania, where he describes how a dead
man’s horse was sometimes cast into the flames of the funeral pyre.105 In Ibn Fadlan’s account of
a cremation among the Rus (see above), a dog, two horses, two cows, a rooster and a hen are
included among the offerings. In Swedish cremation graves from the Vendel and Viking periods,
horses are common and though mainly occurring in men’s graves they are also found with
women.106 Many horse cremations are now known from Anglo-Saxon England and, contrary to
expectation, they are found in roughly equal numbers in male and female graves.107 In contrast
to the situation in Viking Age Sweden and at Heath Wood, dogs are curiously rare in Anglo-
Saxon cremations, though present at both Spong Hill and Sutton Hoo.108
All the animals in the Heath Wood mounds are domestic, in contrast with some of the
Swedish Viking cremations where parts of wild animals may also be present; bear (possibly
in the form of bear skins), lynx and deer have been reported. Anglo-Saxon cremations have
also produced evidence of wild animals, including bear, red deer and fox.109 In view of the
fragmented nature of the Heath Wood material and the relatively few Viking cremation graves
known and investigated in Britain, it may be unwise to regard this fact as of any great
significance. What is certain is that the identification of a range of animals in these mounds, and
the observation that the animal offerings have been removed from the pyre sites along with the
human bone and other funeral goods, adds another strand to our understanding of the ritual
involved. (JB and FW)
Unburnt Animal Bone
Two instances of unburnt animal bones were recovered from Mound 50. (i) An upper molar,
very heavily worn to leave a well of dentine surrounded by a sleeve of enamel. This is an equid
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 81
tooth, but is quite small, perhaps from a pony. HW98,403,sf31. (ii) Fragments of skull of
adult cow. HW98,106, sfs32,46 54,5659. The recovered bones were as follows: most of the
right maxilla, containing the fourth premolar (P4) and the first and second permanent molars
(M1and M2); most of the right mandible, including the lower ramus, P4,M1,M2and M3;an
upper right M3; right mandibular ramus in the area of the condyle; area of temporal bone,
internal surface of the auditory region; fragments of occipital bone around the foramen
magnum.
Although most of the cow skull was not preserved, there are representative bones from
most parts, enough to suggest that a whole head was deposited. It appears that the skull was
buried on its left side and was rapidly buried before much weathering could occur. No
butchery marks were observed, nor was there any evidence of cooking. Neither were any horn
cores present, making it impossible to suggest the sex of the animal. There were, however,
enough teeth to allow an examination of the wear patterns and suggest a rough age. The
material was examined using both Grant and O’Connor’s methodologies and on the basis of
O’Connor the animal may have been about six years old.110 (SR)
The Prehistoric Ditch (figs 3943)
Natural in this area comprised a clean and loose yellow sand (902=1020=1022=1024) with
occasional red mottling (possibly from chemical weathering of sandstone inclusions). It was
apparently of considerable depth, at least 0.8m, and sloped from south to north, following the
natural contours of the site.
A substantial linear ditch (905=1025) had been cut into natural, and ran north south across
Area 9and for the full length of Area 10 (figs 39 42). It is presumably the same feature
encountered by Posnansky beneath Mounds 9and 12, and also observed by him in a trench cut
between Mounds 9and 12 (see below). It sloped northwards, following the natural contours,
and was generally 0.70.8m in depth, by c2.5m wide at the top, and narrowing to c0.3m wide
at the bottom, where it generally had a flattened base (figs 39,40 and 41). On the west side a
substantial spread of medium to large angular sandstone blocks (1004), c3m wide, extended
for the full length of Area 10 (but had not been observed in Area 9). There was a layer of loose
orange sandy loam (1003) in and around the stones. This is interpreted as the remains of a
bank, constructed from stones removed when the ditch was dug. The base of the bank indicates
both the original ground surface, approximately 0.6m below the present one, and also that
there had been little truncation of the top of the ditch, which had been buried under centuries
of slippage of sand downslope, followed by agricultural and then forestry activity (fig 42). The
whole bank and ditch feature is regarded as being a land boundary rather than having
a defensive role, given that it cuts across, rather than follows, the contours. Pottery sherds
support a late prehistoric date.
There were varying degrees of primary silting along the excavated sections of the ditch. In
some places there was a layer of mottled red and purple sand (906,1010,1013,1016,1017,
1019) which provided a hardened crust or lining on both sides of the ditch. In some sections
there was a primary fill of softer yellow or orange sand (903,1009,1012,1014,1015) beneath it,
often filling the base of the ditch, and sometimes extending up the sides.
These two main distinctions in primary fills appear to represent immediate slumpage and
weathering of the ditch sides followed by a phase of consolidation providing the more compact
and weathered stained surface. Along the western slope of the ditch there was a spread of large
angular sandstone blocks, mixed with loose light yellow sand (907,1018=1021=1023). This is
82 THE ANTIQUARIES JOURNAL
Fig 40. Area 10: ditch section. Photograph: author
Fig 39. Area 10 from the south, showing bank and three sections across ditch.
Photograph: author
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 83
Fig 41. Area 10: plan after excavation, showing location of ditch sections in fig 42.
Drawing: Sven Schroeder
84 THE ANTIQUARIES JOURNAL
Fig 42. Area 10: ditch sections, from south to north, as indicated on fig 41.Drawing:
Sven Schroeder
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 85
interpreted as tumble from the adjacent bank, possibly deliberately backfilled into the ditch, or
displaced by animals.
The ditch had been backfilled with a homogenous layer of orange-brown sandy loam (904,
1005,1006,1008,1011), up to 0.8m deep, with occasional sandstone blocks and quartzite
pebbles. There was no stratification visible within the ditch fill and the backfilling may have
been a single event. A small number of plain body or base potsherds were recovered from
excavation and sieving of the ditch backfill. The sherds are small to medium in size (weights
18g) and abraded in varying degrees, which suggests they are residual. Their fabric is a coarse
quartz-tempered ware, characterized by sparse, milky white, angular and poorly sorted quartz
inclusions, some up to 5mm in size, in a fine sandy, iron-rich clay matrix. Dating is difficult,
but on the basis of an absence of decoration and the similarity of the fabric to quartz-tempered
wares from elsewhere in the Trent Valley, a first millennium BC date would appear to be
likely.111 At nearby Swarkestone Lowes, on the north side of the Trent, a vessel rim in a coarse
quartz-tempered ware, from an entrance post-hole of a post-ring round house, was dated on
typological grounds to the Late Bronze Age=Early Iron Age, in the date range of the ninth to
fifth=fourth centuries BC.112 Nothing of later date was recovered from any of the ditch fills,
suggesting that the ditch had been backfilled by the Roman period.
Fig 43. Area 10, showing Posnansky’s trench re-excavated. Photograph: author
86 THE ANTIQUARIES JOURNAL
The whole site was covered by a substantial layer of orange-brown sandy loam (901,1001)
up to 0.35m in depth, indistinguishable from the ditch fills which it overlay. The layer
contained occasional quartz pebbles and sandstone blocks, as well as evidence for animal
burrows and extensive root disturbance. The layer appears to be a subsoil, possibly created by
erosion and soil creep from land further up the hill. It is notable that such a thick sandy loam
layer was not found in the area of the barrows.
At the north end of Area 10 a regular rectangular cut (1007), 1.42.7m, was visible on
removal of the subsoil. It was filled with a loose-mixed orange-brown sandy loam (1002), with
grey and brown staining. The base of this layer coincided with the sides of the ditch as also
revealed in the two new cross-sections. The backfill contained numerous sherds of modern
pottery, including a porcelain teacup, and charcoal flecks. Its location allows it to be identified
as Posnansky’s 1955 backfilled trench, dug to investigate the V-shaped ditch (fig 43).
The whole site was overlain with a layer of accumulated and decomposed brown humic
debris (900,1000), up to 0.15m deep. This sloped northwards, following the natural contours,
but was thicker in the immediate vicinity of tree stumps and roots.
DISCUSSION
Prehistoric Activity
What can be concluded, therefore, about the archaeology of Heath Wood and its barrow
cemetery? The recent limited excavations do not allow much to be said about the pre-cemetery
use of the site, although the hill was certainly part of a prehistoric field system in the Late
Bronze Age and Iron Age. It appears to have been entirely coincidental that Posnansky found
one of the field boundary ditches running under Mounds 9and 12. The feature is clearly not
associated with Scandinavian activity and is not part of a fortification. Rather, it was probably
part of a rectilinear field system, such as those identified by Derrick Riley from aerial
photographic evidence across the Lower Trent Valley.113 Unfortunately, excavation of a 6m
length of the ditch fill in the area between Mounds 9and 12 yielded little new dating evidence
to add to the potsherds recovered in 1955. The pottery recovered supports a late prehistoric
date, and the excavation demonstrated that the ditch was associated with a parallel linear
bank. Fieldwalking in the fields adjacent to Heath Wood by members of the Derbyshire
Archaeological Society has also recovered evidence of prehistoric activity, including several
flint clusters and sherds of similar abraded black hand-made pottery.114 A few residual
prehistoric flints were also found in several of the mound make-ups and must have been part of
the general soil matrix across the site. The excavated section of ditch appears to have been
backfilled at quite an early date and this ditch must have been completely level by the time the
site was used as a cemetery. Elsewhere in Heath Wood, however, the survival of parallel linear
banks identified in the RCHME survey may indicate that other sections of the field system
were preserved, and some may even have influenced the development of the cemetery. Mounds
1625 in particular might respect one such boundary, running north south to the east of the
barrows. Later ridge and furrow ploughing north of the cemetery certainly followed the same
alignment. Possible traces of buried surfaces underneath Mounds 50 and 56 may indicate that
the area was still farmed or grazed in the early medieval period, but soil conditions meant that
no analysis was possible.
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 87
The Barrow Cemetery
By the time it was used as a barrow cemetery, the area of Heath Wood appears to have been
open agricultural land overlooking the flood plain of the Trent. There is no evidence for any
prehistoric barrows on the hilltop, but some ancient land boundaries may have survived and
influenced the development of the cemetery.
There is no evidence for the chronological development of the site. As noted above, the
RCHME survey suggests that the clustering of the mounds is real, but apart from the double
row of Mounds 16 21 alongside the later carriage ride, they appear to be fairly randomly
arranged. Apart from the isolated outliers, they are quite tightly packed within each group,
with little clear space between mounds. There is considerable variation in the size of the
mounds and it might be that the larger examples are the primary ‘founder’ burials with the
smaller mounds filling in the gaps in between. However, the groups may equally have
developed from one side to another, or from the centre outwards. Where adjacent mounds
have been excavated there is no stratigraphic relationship visible between the construction of
the mounds. Nor is there any build-up of deposits such as to suggest any great passage of time
between the construction of neighbouring mounds. In the main cluster it was possible to
observe the intersection of Mounds 6,7,50,55 and 56. Layers of mound material eroding from
the upper slopes of adjacent mounds were interleaved, but construction of the mounds appears
to have started from a cleared level surface, with no overlap. The possible significance of the
different clusters will be revisited below.
The construction of the cemetery would have been a very arduous process. In 1998 a team
of a dozen people equipped with shovels and wheelbarrows took a full day to backfill two
quadrants of a mound. Unless considerably more labour was available, construction of a
barrow by spade or shovel with baskets or buckets to transport the soil must have taken several
days of intense work.
It appears that each mound was built by first roughly marking the area to be used and then
clearing the immediately surrounding ground down to bedrock to form a ring-ditch. In some
cases this involved cutting into the bedding plane and levering up massive stones to be used in
the mound. Excavation of Mounds 50 and 56 has proved that the superficial presence or
absence of a ring-ditch is purely a product of differential erosion, survival and recovery. From
the RCHME survey it appeared that Mound 50 had an encircling ditch, but this was not
observable around Mound 56. However, excavation has demonstrated that soil had been
quarried away to bedrock in the area between the two barrows, so as effectively to create a ring
around each mound. This was no doubt a practical and economical way of creating mounds,
rather than serving any other purpose. The subsoil was left in situ in the centre of the mound,
and spines of subsoil were left to form causeways on to the mound: Mound 50 had two such
causeways; in Mound 56 one causeway was observed, adjacent to the burial deposit. These
causeways were wide enough for a single person and would have provided access during the
cremation and subsequent mound building. Mounds 50 and 56 were also both built up with a
dump of orange sandy clay, presumably made up of soil removed from the ditch. This levelled
any unevenness in the mound and provided a level platform for the burial (fig 44). From this
point onwards the practice varied between each mound, apparently reflecting different burial
rites practised within the cemetery as a whole.
The excavations of 1998 to 2000 have revealed that in some cases the mound marks the site
of a pyre which extended across the platform, whilst in others it appears that the cremation
took place elsewhere and that only a small proportion of the cremated material was placed on
88 THE ANTIQUARIES JOURNAL
the mound platform. Mound 50 contained the remains of a cremation hearth, which would
have been discovered even if no more than a narrow trench had been excavated across the
barrow. Mound 56, on the other hand, contained only a few grams of cremated bone,
associated with the ringed pin. This material had been buried towards the edge of the mound
and was only discovered when a third quadrant was examined in the final season of excavation.
Because there had been no cremation hearth, there was no trace of the burial in the other
quadrants, not even the charcoal flecks that had characterized the lower levels of Mound 50.
All previous excavations in Heath Wood had employed trenching and had only investigated
part of the mound interior. It is clear that these excavations could easily have missed such
peripheral and transient traces of cremation as were encountered in Mound 56. All previous
interpretations based upon the presence of cenotaphs or empty mounds are therefore now in
doubt, and whilst it is accepted that these apparently empty mounds represent a different rite,
it is more appropriate to see them as burials of those cremated elsewhere and marked at Heath
Wood by a token burial deposit.
Cremation hearths were reported in Mounds 1,5,6,7and 11, plus the five mounds
excavated by Bateman, although it is unclear whether he really found hearths in each and he
certainly found few objects. Traces of a cremation, but no hearth, were found in Mound 3, but
Mounds 2,4,8,9,10,12,13 and 15 were described as empty and interpreted as cenotaphs.
These should now be reinterpreted as potentially containing small cremated offerings, as in
Fig 44. Profiles of Mounds 50 and 56 showing cremation deposits. Drawing: Sven
Schroeder
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 89
Mounds 3and 56. If Mounds 50 and 56 are included, but Mound 14 is omitted as being
natural, Mound 21 as difficult to determine and the five Bateman mounds as being uncertain,
this gives a total of six definite cremation hearths and ten possible token cremation burials. The
proportion of excavated burials with cremation hearths is therefore 38 per cent, compared with
64 per cent token burials. If this ratio were to be reflected throughout the fifty-nine barrows of
the complete cemetery, twenty-two cremation hearths and thirty-seven token cremations
would be anticipated. The significance of this difference will be discussed further below.
Mound 50 was used as a platform for a cremation pyre. Before the pyre was built a layer of
clean yellow sand was spread in the centre of the mound. This was dissimilar to the subsoil
immediately adjacent to the mound, but could have been brought from elsewhere in the area
(and was similar, for example, to the very sandy soil excavated over the ditch between Mounds
9and 12). The pyre was built upon this layer, which became reddened from the heat.
McKinley has demonstrated that bodies are likely to have been placed within the pyre but
towards the top, rather than resting upon the ground.115
The effectiveness of the cremation has left a meagre record of those objects originally placed
with the body but it is clear that these included weaponry and other objects placed on the pyre
and subjected to its heat. Of the six recorded mounds with cremation hearths, three were
accompanied by swords or fragments of swords. In two cases (Mounds 1and 7) parts of the
blade and hilt were present; in Mound 50 there was just a mount from the hilt. The swords
must have been deliberately mutilated prior to cremation; it is less clear if the complete sword
was then cremated with the body but only partially recovered, or whether only a portion of it
was put on the pyre. The intentional ‘killing’ of weapons also occurs in Viking graves at the
nearby site at Repton, Derbyshire, in the cremation burial at Hesket-in-the-Forest, Cumbria,
at Ballateare on the Isle of Man, at Islandbridge, Dublin, and in Scandinavia.116
It seems likely that at least two of the burials were also accompanied by shields. Mound 50
contained a number of shield rim clamps as well as nails which might have held a boss in place;
it is likely that Leeds was also correct in interpreting the twenty-four small tacks recovered
from Mound 7as representing the nails holding the leather binding to the rim of a circular
shield. The remaining objects recovered from the cremation hearths comprise buckles and
other strap fittings or sword attachments from a belt. The tip of a small knife blade recovered
from Mound 50 is also from a type of object that may originally have been suspended from a
belt. The silver-wire embroidery from Mound 11 and the fragments of burnt bronze and silver
from Mounds 1,6and 50 may represent small personal ornaments or dress fittings.
The various nails recovered from Mounds 5,6,11 and 50 may have come from other
objects, but could also have come from timbers used for fuel. Mound 50, excavated completely
under modern conditions, probably has the most representative collection of nails, and given
that many are plated it seems likely that they originated from objects in which they had a
decorative as well as a functional purpose. The two small tacks from Mound 50 are too small to
have held anything together but they could have been used for attaching mounts to other
objects, such as a box.
In 1995 we suggested that some of the nails recovered in earlier excavations might have
come from coffins or chests which held the body, or from biers on which it was placed.117 The
hinge pivot now recovered from Mound 50 supports the idea of the body being in a chest. We
also speculated whether some of the biers might have been constructed from sections of ships’
planking. However, closer examination of the nails indicates that, with the possible exception
of the large example from Mound 5, they are not of the clench nail type, and we are now
inclined to agree with Graham-Campbell that this interpretation is unlikely.118 The use of
90 THE ANTIQUARIES JOURNAL
chests as coffins or sections of planking as biers, however, remains likely. Throughout the
Viking Age in Scandinavia chests were used primarily as furniture for storage or for travelling.
They also had a secondary function as coffins. Because they possessed lids and were usually
fastened by iron hinges, chests could be locked, perhaps thus combining connotations of a
voyage with a theme of private property and the sanctity of personal ownership. One end of the
chest found at Lejre, Sjælland, had been broken in order to facilitate the insertion of a fully
extended adult corpse.119 The Fyrkat chest had probably undergone the same alteration, whilst
the body in the Forlev chest was laid out with bent knees.120 Similar types and numbers of nails
are known from the cremation graves at Birka.121 At York, at least four instances of Viking
burial in wooden, domestic storage chests have been detected.122
The cremation hearths apparently contained the bones of both sexes. Indeed, where it has
been possible positively to determine sex, in Mounds 5,6and 50, the subjects have all been
probable females. Mounds 5and 6did not contain any grave goods thought to be diagnostic of
gender, but the molten remains of dress fittings and ornaments could be appropriate to a
female cremation. The recovery of the sword hilt guard and shield rim clamps from Mound 50
also requires comment. Whilst it is possible that the female has been misidentified, it would be
unwise to dismiss the skeletal sexing simply because it conflicts with the grave-good evidence.
It is also possible that the objects have been misidentified, or even that the woman was
cremated with weapons as a mark of status, although this would be highly unusual.123 It should
be noted too that the bones of a child or juvenile were also identified in Mound 50, and that
children are more likely to be buried with females. It could then be that the weapons were
markers of the status of the child or juvenile, rather than the adult. Ha
¨rke has drawn attention
to the fact that Anglo-Saxon boys of all ages were buried with weapons, although shields were
generally given to boys only when they reached puberty.124
Where two individuals are cremated together it is also possible that one was sacrificed to
accompany the other, and that a slave might have been intended to accompany his or her
master to Valhalla as an item of property. The practice is documented amongst the Rus, and
appears to be reflected in the pagan Viking burial at Ballateare on the Isle of Man.125 The
presence of the remains of more than one individual on the pyre of Mound 50 also reminds us
that more of the cremation hearths might represent double burials. Holck has noted the
presence of a number of double graves in Norwegian and Danish cremation burials,
comprising forty-eight (4.4per cent) of the total from Norway and twenty-one (1.3per cent) of
those from Denmark. Holck concludes that it is unrealistic to believe that such a large number
of double graves could result from two people dying simultaneously. Whilst they may not
necessarily have been cremated at the same time, Holck has also suggested the possibility of
human sacrifice. Where the double graves are of adults and children he suggests an alternative
and more practical explanation: because of their lower body fat ratio children take much longer
to cremate than adults and it may be that the complete cremation of the child was only possible
with an adult body beside as a ‘heat producer’: ‘the child needed to be burnt in the flames from
the adult’.126
Each of the hearth deposits also included the cremated remains of animals. The records
from previous excavations are likely to be incomplete and to under-represent smaller animals.
Nonetheless, sheep, cow and a possible dog were identified in Mound 1, cow and a possible
horse from Mound 5, cow plus possible sheep and pig from Mound 6and a possible small dog
and sheep from Mound 11. Re-examination of the bones from Mound 11 has confirmed the
presence of a dog and sheep or goat, and added a horse. In Mound 50 a variety of animal
species was identified, probably reflecting the use of flotation and improvements in the
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 91
recognition and identification of cremated animal bone. These comprised horse, dog and pig,
probably cow and sheep, and possibly bird. Given the variety of the species represented it is
unlikely that all these animals should simply be regarded as the remains of funeral feasts,
although given the weights and range of bones represented it does not seem as if complete
carcasses of all animals were placed on the pyre. It is more likely that some sacrificial animals
were cremated on the pyre alongside the human corpse, whilst in other cases some joints of
meat were cremated and the other parts were consumed by the mourners. From the bones
represented in Mound 50 it appears that the remains of a complete horse and a large hunting
dog were cremated, along with joints of pork, mutton and beef. Whilst the horse and dog were
to accompany their owner on the journey to Valhalla, the sheep, cow and pig were to provide
food along the way.
It is impossible to speculate about the ceremonies that might have accompanied the
cremation, but the pyres burning on the skyline above the River Trent must have been visible
for miles, particularly against a night sky. Once the fires had died down and the hearths cooled
it seems as if they were raked (on the basis of the evidence of Mound 50; see McKinley above).
This would explain the degree of fragmentation of the bone and the apparent haphazard
distribution of body parts across the hearth. It may also have been at this stage that some of the
more obvious surviving fragments of grave goods were collected, which would explain the
rather random assortment of surviving nails, as well as the missing parts of the three swords
and the absence of shield bosses.
An alternative cremation rite is represented in Mound 56. In this case the equivalent of
a shovelful of charcoal and burnt bone was placed on the edge of the mound, on the side
adjacent to Mound 50, possibly suggesting some satellite relationship to it. This must
represent only a small part of the cremated body. Care had been taken, however, to include
the ringed pin as a personal ornament. This had also been burnt on the pyre, but not so
completely as to destroy it. Remains of personal dress items, including a buckle and some
bronze fragments, were also recovered from Mound 3, which again appears to represent a
token cremation burial. In neither case was it possible to determine sex, but both were single
adult burials. In Mound 3no animal bones were identified; in Mound 56 a small fragment
of pig bone was recovered, suggesting a food offering rather than an animal sacrifice. The
location of the actual cremation pyre is unknown in both cases, but this issue will be returned
to below.
The final stage of both cremation rites involved building up the mound to give it a domed
shape. There does not appear to have been any specific method to this and no evidence for
mound structure. There were no kerbs in either of Mounds 50 or 56, although the stone
bedrock from the ring-ditch around the mounds at first gave the impression of a kerb, a feature
that was also observed in the earlier excavations. The make-up of the mounds appears to have
been determined by the material that was most easily available; in some cases this included a
higher proportion of large rocks and boulders; in other cases the mound was largely composed
of sand and gravel. In Mound 50 the section indicated that the mound had been built up from
the outside and soil and stones tipped into the centre. Soil probably had to be brought from
elsewhere in the vicinity; there is evidence of quarrying within Heath Wood.
In three of the cremation hearth burials, from Mounds 1,6and 50, unburnt animal bones
were incorporated in the mound make-up. In Mounds 1and 50 unburnt cow skulls were
placed above the hearth and buried in the mound; unburnt cow bones were also found in
Mound 6. These may also represent offerings, but could perhaps be the residue of meals eaten
by those constructing the mounds.
92 THE ANTIQUARIES JOURNAL
Once the mound profiles were complete it seems as if they were fairly rapidly consolidated.
There was initial run-off and silting of the ring-ditch but the mounds appear to have been
quickly grassed over, certainly within a year. The excavated profiles of Mounds 50 and 56 may
indicate that they were deliberately turfed to prevent erosion. They then appear to have
retained their original form for several centuries. The main mound clusters would have been
resistant to cultivation, although medieval ploughing may have cut into and started to level the
cluster of Mounds 16 25 on the edge of the hill. The mounds in the main group, however,
appear to have retained their form until the mid-eighteenth century, although they may have
provided a welcome habitat for rabbits and other burrowing animals, as they do today. At that
stage the landscaping of the grounds of Foremark Hall and the plantation of the wood appears
to have led to substantial erosion of the mounds until the wood itself was consolidated,
shrouding the barrow cemetery in trees and protecting it from the gaze of possible robbers,
albeit still leaving it subject to periodic bouts of subsequent forestry activity.
The Wider Context
The age of the Heath Wood cemetery has been debated but the recent excavations confirm that
Bateman, and then Leeds, were correct in ascribing it to the period of Scandinavian raiding
and settlement of England, that is, to the late ninth or early tenth centuries AD. A number of
finds are particularly diagnostic, namely the two sword blades and hilts from Mounds 1and 6,
the hilt guard from Mound 50, the silver-wire embroidery from Mound 11 and the ringed pin
from Mound 56. Each of these confirms a late ninth to early tenth century date for these burials
but do not, unfortunately, allow the date to be tied down any more precisely. Samples of
cremated human bone from Mounds 11,50 and 56 were submitted to Oxford University
Radiocarbon Accelerator Unit. These yielded uncalibrated dates of 1247 27 BP (OxA–
12698), 1191 26 (OxA12700) and 1163 26 (OxA12699) respectively (figs 45a– c). Once
calibrated the determinations have a 95.4per cent probability of falling between AD 680 880,
AD 770950 and AD 770980. For the bone recovered from the 19982000 excavations, the
sample from Mound 50 has a 68.2per cent probability of falling between AD 780 890, and that
from Mound 56 the highest probability (42.3per cent) of falling between AD 810900. Whilst
these dates do not allow us to date the cemetery with greater precision, they are at least
consistent with the period of Scandinavian activity in the ninth and tenth centuries.
The burial rite itself is also diagnostic. By the early eighth century cremation had ceased to
be employed in England and was condemned by the Church. By the ninth century most native
English burials were unfurnished inhumations, often, but not exclusively, in churchyards.127
Cremation burials within mounds are found in Late Iron Age and Viking Age Scandinavia,
where they clearly represent a pre-Christian burial rite. The inclusion of weaponry on the pyre
and the cremation of animals are practices also known from Viking Age Scandinavia.
Scandinavia
In Scandinavia cremation seems generally to have been more common in Sweden and Norway
than in Denmark. In most Scandinavian cremation burials the body was clothed and adorned
with jewellery and fastenings and incinerated on a pyre. The cremated bones and molten
objects were then treated in a variety of ways, according to local custom and belief. In central
Sweden, for instance, the burnt remains were usually separated from the ash and charcoal and
placed in a pottery vessel, which was then buried in a pit; in parts of Finland they were
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 93
(a) OxA12698 HW56=11 Mound 11, cremated bone, human
(b) OxA12700 HW99=50=411 Mound 50, cremated bone, human
94 THE ANTIQUARIES JOURNAL
scattered on the ground. The remains were then either covered by a mound or marked with
arrangements of stones, sometimes in a ship setting.
In Denmark cremation cemeteries are mainly found in northern Jutland, at sites such as
A
˚sted, Kirkelund, Ris Fattigga˚ rd, Hørby, Lindholm Høje, Øsløs and Højstrup.128 These
generally consist of upwards of around thirty barrows, but may contain some empty mounds, as
well as ship settings. The grave goods include iron objects, including rivets and nails, as well as
glass beads and potsherds, but weaponry is less common. The absence of prestigious, status-
conferring objects from this kind of grave in Denmark has led to the view that they belong to the
lower levels of Scandinavian society. However, as Roesdahl has stressed, we know relatively little
of pre-Christian burial rites.129 With 700 graves, Lindholm Høje is the largest burial ground in
Denmark, although the majority of the graves date to the Germanic Iron Age. Oval and ship-
shaped settings predominate, but other shapes include triangles, squares and circles. Burial pits,
small mounds and cremation patches also occur. Cremation seems to have taken place on the
spot; occasionally unburnt animal bones are found. The later male graves often contain buckles
and rivets, and some weapons occur in the Viking period burials. It has been suggested that
the cremation temperatures at Lindholm were fairly high, often more than 1000C, whereas
Swedish pyres reached 8001000C, and that this explains the poor artefact survival.130
(c) OxA12699 HW00=56=508 Mound 56, cremated bone, human
Figs 45a–c. Plots of radiocarbon results showing: (i) the radiocarbon determination on
the left-hand axis; (ii) measurements on known age material as the uneven double line;
(iii) the likelihood of different possible ages of the samples shown as the solid black
distribution. Graphs: Oxcal computer program (v3.8), Oxford University Radiocarbon
Accelerator Unit
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 95
Shetelig felt that the resemblance between Heath Wood and the north Jutland sites was so
striking that those cremated at Heath Wood must have come from north of the Limfjord.131
This view was accepted in our 1995 assessment of the site132 but there are also parallels from
both southern Sweden and Norway, where cremation was as popular as inhumation. In rural
communities in Norway and eastern central Sweden cremation was the most common form of
Viking Age burial. Cremation graves under mounds cluster around Viking Age farms, usually
on rocky outcrops. At Birka there are about 1,600 burial mounds in Hemlanden and 400 south
of the hillfort, mainly cremations. Stolpe excavated almost 560 cremations and about 550
inhumations in coffins and chambers. The cremation graves contain various fragments of
molten jewellery and iron nails but traces of weaponry are also common in Birka whereas they
are unusual in the north Jutland cremations.133 Crowfoot suggested that the distinctive
Scandinavian style of wire embroidery, known as O
¨senstitch, is also a southern Swedish trait,
and it is certainly found in Birka.134 The only similar example known from England is of a
related needle-binding method, or na
˚lebinding, on a woollen sock recovered from a tenth-
century context at Coppergate, York,135 although silver-wire embroidery has been found in
other Scandinavian-type burials in England (see above).
The ideological significance of cremation has tended to be neglected by early medieval
archaeologists. Most have followed the general view that the flames release the ‘soul’ or identity
of the deceased for the journey to an afterlife, and that this ascends to the sky on the smoke
from the funeral pyre. Grave goods are provided for the use of the deceased in the next world
and will therefore reflect the status to be assumed in that world. Literary support is found in
Heimskringla, written in Iceland by Snorri Sturluson, c1230. It was said that Odin:
ordered that all the dead were to be burned on a pyre together with their possessions,
saying that everyone would arrive in Valhalla with such wealth as he had with him on
his pyre and that he would also enjoy the use of what he himself had hidden in the
ground. His ashes were to be carried out to sea or buried in the ground. For notable men
burial mounds were to be thrown up.136
According to Snorri Sturluson cremations were the oldest form of burials and related
specifically to Odin worship whereas inhumations were a new fashion introduced by worship of
Freyr. The simultaneous occurrence of each rite in the same area does not permit us to follow
his chronological distinction, but the difference may relate to cult preferences.137
In many societies fire has a cosmological significance as an agent of transformation. It
can ward off evil spirits and is entrusted with the task of presenting offerings to the gods. Fire
changes a substance from one state to another and therefore helps the dead travel across
boundaries.138 Williams notes that in the story of Baldr’s death, the ship, the horse and
transformation by fire all serve as metaphors for changing the physical and spiritual status of
the deceased.139 In some circumstances cremation can also be seen as a way of preventing the
dead from walking again. A common theme in the later Sagas is the use of cremation to get rid
of dangerous corpses, those individuals who have been polluted by a ‘bad death’ and must be
dispersed by an extreme form of disintegration of the body.140 The ritual mutilation of swords
and other weapons, seen at Heath Wood and also in several inhumation burials, might be
viewed in the context of preventing a known violent individual from continuing this behaviour
in the next world, or might be to protect these objects from robbers. The deposition of
mutilated weaponry in Scandinavian bogs, going back to the Germanic Iron Age, is seen as a
means by which thanks are given to the gods for victory in battle. Maybe the mutilation of
weapons in graves performs a similar act of dedication of the weapons to the world of the gods.
96 THE ANTIQUARIES JOURNAL
The cremation of animals is also a distinctive feature of the Heath Wood cemetery, and is
paralleled in other Viking Age burials.141 Cremated animal offerings were recorded in the
Scandinavian mound burial at Hesket-in-the-Forest (see below) and from mound burials with
inhumed bodies at Ballateare, Balladoole and Cronk Mooar, on the Isle of Man.142 Unburnt
animal offerings are also known from other Viking graves, including inhumation burials at
Knoc-y-Doonee,143 on the Isle of Man, and from Dublin. Two inhumation graves found in
Islandbridge cemetery in the mid-1930s were accompanied by animal remains. One of the
skeletons was accompanied by a cattle jaw bone, while a cow tooth and a horse tooth were
found in the second grave.144 Williams has argued that the sacrifice of animals as part of
mortuary ritual should be seen as part of the rite of passage: ‘the killing of an animal is more
than the disposal of the deceased’s property, it is instead a key social, economic and
cosmological statement of the relationship between the dead and the soul’s journey to a
perceived afterlife’.145
Even animals that do not bear the dead can act as guides, and no distinction is necessary
between eating an animal at a funerary feast and its use in journeying with the soul of the dead:
‘animals may be more than symbols of identity and wealth; in some cases they merge with, or
become integral to, the changing identity of the deceased’.146
The British Isles
Viking burials, as defined by the presence of grave goods, are uncommon in the British Isles
and particularly rare from lowland England.147 For the period AD 800 1000 there are fewer
than twenty-five burial sites in England which have been described as Scandinavian, and the
majority of those are single burials. Halsall seeks to reduce the body count still further by
arguing that many of those identified as Scandinavian simply on the basis of objects from the
grave may have been misidentified and could well have been Anglo-Saxon.148 Repton and
Heath Wood are the only sites with identified Scandinavian burials in Derbyshire, despite the
recorded settlement of a section of the Great Army in Mercia (see below). It has been
suggested that a ring-ditch excavated at Foston, some 10km west of Derby, may have been a
Viking burial mound but this is based solely on two radiocarbon dates from the ditch fill.149 It
is evident that the majority of Scandinavian settlers must have adopted new burial rites, such
as unaccompanied burial with hogback grave markers, becoming assimilated with the native
population through the adoption of churchyard burial.150 Hadley suggests that in general
Scandinavian leaders were very adept at adopting indigenous practices and forms of lordship
and that in their desire to take control of parts of northern England it was inappropriate to
display any sign of ‘otherness’. Indeed, she argues that they may have encouraged the fashion
for local churchyard burial and for local displays of status, rather than burial in Anglo-Saxon
minsters whose prestige would have been undermined by attacks on their resources and the
loss of their relics.151
Cremation burials are particularly infrequent in the British Isles, and Heath Wood is the
only known Viking cremation cemetery, although a small number of single cremation burials
have been recorded, mainly in antiquarian accounts. In common with other accompanied
Scandinavian burials these tend to cluster around the Norse fringes.
In Orkney ‘burnt bones’ were recovered from a mound at Lyking, Sandwick, on Mainland
in the nineteenth century, together with an iron spearhead, a buckle and an unburnt comb.152
One of two burials from Lamba Ness, Sanday, was a cremation: ‘a deposit of burnt bones was
found about the centre of a mound’,153 associated with an unburnt pair of oval brooches, a
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 97
lignite arm-ring, a bronze ringed pin and an amber bead.154 In both cases the cremation was
under a mound although the presence of unburnt objects distinguishes these burials from those
at Heath Wood. The artefacts must have been added to the burial after cremation as grave-
good offerings. There is a third possible Orkney cremation from Knowe of Moan, Harray,
where a small cist, eighteen inches square, was found associated with a Hiberno-Saxon mount
and sixty-four beads, although no burnt bone was recorded.155
From Skye, in the Hebrides, a prehistoric cairn at Tote, Skeabost, contained a secondary
Viking burial, probably a cremation, with an axe, bronze pin and ivory bead.156 At Ca
`rn
a’Bharraich, on Oronsay, a bed of charcoal and boat rivets was discovered in the centre of a
substantial mound; pieces of bronze and a stone sinker were recovered from the mound, which
also incorporated a number of inhumation burials, possibly secondary.157 At King’s Cross
Point, on Arran, there was the burial of a high-status woman, whose cremated remains were
within a stone setting under a mound. An iron lock plate and clasp may represent a wooden
chest; four rivets, an iron nail and the burnt remains of a whalebone plaque were also
recovered. A styca of Wigmund (837 54) may date the burial to the mid=late ninth century.158
At Blackerne, Kirkcudbrightshire, an eighteenth-century account notes that a ‘parcel of human
bones, along with several teeth, were found in the heart of a cairn’.159 Two finds of Viking
objects from mounds, but with no record of a skeleton or any bones, may also have been
cremations. At Boiden, Loch Lomond, a bent sword and damaged shield boss were found in a
mound160 and at Millhill on Arran a sword and shield were recovered from a gravel mound.161
In general, however, evidence for cremation is scarce in Scotland, despite its popularity in
Norway.
The absence of Viking cremation in Ireland is even more surprising. Of an estimated
seventy or eighty male burials and possibly about ten female burials in the Dublin area, there
are no certain cremations. In the great cemetery at Kilmainham-Islandbridge there is no
evidence for mounds and although the presence of some bent swords and spearheads has
sometimes been taken as indirect evidence for cremation the bending is unlikely to have
resulted from heat damage and is more likely to represent the deliberate mutilation of
weaponry accompanying inhumation burials. The absence of cremation in Ireland might be
attributed to the influence of Christianity.162
Cremation is not known from the Isle of Man, although Kristin Bornholdt Collins163 has
noted that Clay’s near contemporary account of the Kirk Andreas hoard (c970) hints at the
possibility, for with the coins were found ‘ ... some horse teeth, bits of charred wood, and black
earth, which indicated burning’.164 On the basis of this report Dolley suggested a ‘doubtful
association with one or more cremation burials possibly Viking in date’165 although Graham-
Campbell rejected the idea as cremations are otherwise unknown on Man in this period and
‘such material must surely represent no more than general occupation debris’.166 If it is a
cremation it is an isolated and very late example.
In England there are one probable and three possible cremation burials from the north
west. At Hesket-in-the-Forest, Cumbria, a layer of charcoal, cremated bones and ashes with
several grave goods was found in 1822, lying on a bed of sand under a cairn, 22ft (6.7m) in
diameter.167 The burial at the centre was covered with large stones, in turn covered by smaller
ones. Shetelig argued that this was a cremation after the Norwegian fashion but Cowen
disputed this, noting that all the bones were of animals and that no human skeleton was
recovered.168 Nonetheless, the burial was accompanied by extensive Viking Age grave goods,
comprising a sword, two spears, a shield, an axe, a horse bit, a pair of iron spurs, an iron sickle
blade, a whetstone, two small iron buckles and an antler comb and case.169 The sword, shield
98 THE ANTIQUARIES JOURNAL
boss and bridle-bit were all burnt, suggesting that the weapons and a horse may have been
placed on the cremation pyre. The sword and spears were bent; the shield had been broken in
two; the sword had been deliberately bent back twice on itself, by heating and hammering,
rendering it useless.
Another mound burial was also discovered in 1822 at Claughton Hall, Garstang,
Lancashire, when a small sand mound was cut through in the course of road building. No
skeleton was found but the objects comprised a pair of gilt copper-alloy oval brooches,
apparently wrapped up back to back in cloth and encasing two beads and a molar tooth, a
Carolingian silver mount reused as a brooch and various iron objects, including a sword,
spear, axe and hammer. This may have been a double burial of male and female, but it is more
likely that the burial was male and the brooches enclosed a ritual deposit of various amulets or
keepsakes.170 There may have been a wooden chamber below the surface but the finds also
included a Bronze Age axe hammer and a pot containing a cremation, now lost, so the finds
from Claughton may represent another Scandinavian cremation, or secondary usage of a
prehistoric barrow in which traces of the body comprising the secondary inhumation burial
had disappeared. Edwards also suggests that an urned cremation found with a Viking sword
at Inskip (SD 4438), only five miles from Claughton, might represent a Viking cremation.171
Finally, there may be a fourth Viking cremation from north-west England at Bents Hill,
Crosby Garrett, Cumbria (NY 7006). In 1873 Greenwell excavated a secondary cremation in a
Bronze Age round cairn, comprising the burnt remains of a juvenile male inserted close to the
apex of the mound. The associated finds comprised an iron knife, an iron buckle, iron shears
with an ornamented handle and an iron bridle. Greenwell ascribed it to the Anglian period but
the finds make it more likely to be Scandinavian.172
In summary, the rarity of distinctive Scandinavian burials from England, and the particular
shortage of comparable cremation burials from the British Isles as a whole, makes Heath Wood
stand out as being of particular significance. In order to understand why there should be
a Scandinavian cremation cemetery of some fifty-nine mounds in Derbyshire we need to
consider the local context.
Heath Wood and Repton
By AD 800 Lindsey and the other Anglo-Saxon kingdoms of the Midlands had come under the
control of Mercia. Nonetheless, this powerful overlord provided inadequate protection against
Viking raids. In 865 6a highly mobile Viking force landed in East Anglia. It is generally
assumed that it came to England direct from Scandinavia, but Simon Keynes has suggested
that it was a composite force, with elements drawn from groups that had been active in Ireland
and on the Continent. It is difficult to get a clear understanding of the size of this army, beyond
the fact that the compilers of the Anglo-Saxon Chronicle considered it to be ‘Great’ and that
it was powerful enough regularly to over-winter in England.173 On the basis of its success and
coherence over several years, some historians have been tempted to suppose it comprised
perhaps two or three thousand men, but others have suggested it may have numbered as few as
three hundred.174 The leaders included Ivar the Boneless and his brother Halfdan, both sons of
Ragnar Lothbrok, as well as another king called Bagsecg, and several ‘earls’. If Ivar is the same
person as the I
´mar who is recorded as raiding in Ireland in the 850s and 860s then it must be
assumed that they met up in England and assumed joint leadership.
After the conquest of Northumbria in 866, and East Anglia in 869, it is believed that the
Great Army was reinforced by what the Anglo-Saxon Chronicle describes as a ‘Summer Army’,
EXCAVATIONS AT THE VIKING BARROW CEMETERY AT HEATH WOOD, INGLEBY 99
under Guthrum, before it forced Wessex to make peace in 871.175 The army then appears to
have turned its attention to Mercia, over-wintering in London in 871 2and Torksey in 872 3.
In the autumn of 873 the Great Army arrived at Repton and took over the monastic complex,
driving the Mercian king Burgred into exile in Rome, and placing one of his thegns, Ceolwulf,
on the throne as its puppet king. The Anglo-Saxon Chronicle records that the army took
wintersetl, or winter quarters, there. Its choice had a symbolic as well as a tactical significance.
By the late ninth century Repton was one of the principal ecclesiastical centres of England, and
was closely associated with the power of the Mercian royal family. A monastery had been
established here in the seventh century. It appears to have been a double house for men and
women ruled by an abbess of noble, possibly royal, rank.176 Several of the kings of the Mercian
house were buried at Repton, including Aethelbald after his murder at Seckington in 757.In
849 Wigstan (Wystan) was brought to Repton after his murder in a struggle over the succession
to the Mercian throne. Wigstan was buried in the mausoleum of his grandfather, Wiglaf (827
40). This is almost certainly the crypt, which survives beneath the chancel. Miracles took place
at the tomb and the entrances to the crypt were lengthened to deal with the flow of pilgrims.177
Before the end of the ninth century, Wigstan had come to be regarded as a saint. By this stage
the church had a north and south porticus, and a chancel over the mausoleum, with burials to
the south and east. The seizure of one of the holy places of Christendom could not have gone