Added June 29, 2001. Updated April 25, 2006, 10:50
hours.
- Under Construction!
- Only some
illustrations have been added.
- This page
uses *.PNG graphics, which are visible in Explorer 5.5 and later. If you
use a different browser, they may not be visible.
A SPATIAL ANALYSIS OF MEGALITHIC
TOMBS
13. MOUNDS AND ENCLOSURES
By
Maximilian O. Baldia
1993, 1995,
1999-2001©
All rights reserved
13.1 Introduction
Many, but not all chambers are located
in burial mounds. Following Daniel, one suspects that all chambers were
intended to be covered by mounds, since they are an integral part of chamber
construction. They waterproofed the chambers and provided structural support by
keeping the stones from shifting and collapsing (Fig.13.1). However, some
chambers, such as Heveskesklooster (Bakker 1992) (Bakker 1992), may never have
been completely covered.
Figure 13.1. The complex, watertight seal of the
Jordehøj, Møn passage-grave (Hansen 1993a:60 Fig. 97). Elaborate layers of
crushed flint, clay, stones, and soil cover the chamber.
Not counting 880 unclassifiable
megalithic tombs, 2411 structures exhibit evidence of mounds, while 2217
chambers are not known to have had mounds or enclosures. Thus 52.0959% of the
chambers were covered by mounds. Considering the enormous destruction of the
tombs, this may be a fairly strong indication that most, if not all chambers,
must have had a mound cover.
|
TABLE 13.1
Classifiable chambers without traces
of mounds or enclosures
CHAMBER TYPE
NO MOUNDS TOTAL NO
MOUND
UD
61 272 22.42%
ED 67 261 25.67
PD 28 61 45.90
GD 64
494 12.96
GG 604 1143 52.84
GA 46 58 79.31
|
|
UD =
urdolmen, PD = polygonal-dolmen, ED = extended-dolmen, GD = grand-dolmen, GG
= passage-grave, GA = gallery-grave.
|
Figure 13.2. Complex four layer stratigraphy, Kong Svends Høj, Lolland
(Dehn et al. 1992:85 Fig. 7).
Polygonal dolmen, passage-graves, and gallery-graves
are the least likely to show evidence of a mound-cover (Table 13.1).
Bakker suspects that the smaller passage-graves in Netherland never had mounds,
but similar size mounds elsewhere do.
There is a dearth of interregional information on TRB
mound composition. Schuldt distinguishes
cairns from other (undefined)
tumuli. Excavation reports, such as those by Hansen (1993a:20,24-25
Fig. 20, 31), provide good insights into the construction of the mound
(Fig. 13.2). But it is difficult to form any meaningful interregional
comparisons, due to the complexity of the stratigraphy resulting from continued
rebuilding and expansion that often lasted right into the Bronze Age, and
considerable destruction (ibid.).
The 2411 burial mounds
have traditionally been dichotomized into long-mounds and tumuli.
Table 13.2 demonstrates that the long-mounds outnumber tumuli, which is
surprising, since Schuldt (1972) found that tumuli outnumbered long-mounds in
Mecklenburg-Vorpommern. There is also a small number of transitional mounds.
The term transitional merely refers to a transitional mound form which is
neither exactly a long-mound nor a tumulus.
Figure 13.3. Tomb evolution at
Bygholm Nørremark (after P. Rønne 1979).
|
TABLE 13.2
Number of traceable mounds by type
Long-mounds 1261
Transitional
(Dutch) 95
Tumuli 812
Unclassifiable 243
TOTAL 2411
This
traditional classification has no chronological implications. A small number
of these mounds are complex mounds. These include unique types, such as
Katterø, Svendborg Amt, Lindhöft 110, and Tannenhausen 817.
|
|
TABLE 13.3
Dimensions of Polish long-mounds
proposed to be among the oldest tombs by Midgley (1992:52)
ID.
NO. TOWN FORM µL WIDTH 1 WIDTH 2 µW AREA
8228.0 Sarnowo 1 TR 76.25 11.07m 5.36m
6.19 471.99
8237.0 Leíniczowo TR 72.42 8.71 1.61
5.16 373.69
8245.0 Obalki TR 32.24 5.52 3.58
3.23 104.14
8253.0 Wietrzychowice TR 11.76
The mounds show relatively little overall consistency
in their dimensions, but their orientations range from east northeast to
east. L = length, W = width.
|
In spite of the large number of mounds
and hundreds of years of their investigation, there are no helpful guidelines
for analyzing mounds. In fact the tables given here are the first actual
tabulations of mounds covering the whole TRB culture area. Consequently, it
seems that no one has as yet determined significant regional or temporal trends
in mound dimensions beyond a few speculations already mentioned. Even Midgley
only dealt with a small class of mounds, the
"earthen long-barrows". She hypothesized that they were the
earliest tombs (Table 13.3). Others have argued that rectangular or round
mounds were first.
Laux (e.g. 1979a, 1991) suggests that
mounds increase in height through time. Schuldt's illustrations also seem to
indicate an increase in mound height concomitant with an increase in the size
of chambers and a likely trend to construct chambers closer to the surface. In
most cases though, the reporting of mound height is unstandardized and erratic,
largely due to rebuilding, destruction etc., that the available data, although
recorded in the data base, does not lend itself to meaningful analysis.
13.2 Mound and enclosure origin and evolution.
Even if one takes the Polish
radiocarbon dates into consideration, one can presently only justify an approximate
date of 3250±200 b.c. The first mounds probably were a natural byproduct
of digging (Mesolithic?) burial pits and then using the soil as back fill to
cover the interment. In time the resulting mound would have received greater
architectural elaboration which no doubt increased in ritual significance,
especially when several ancestors came to rest in close proximity under mounds
that gradually coalesced into a single more or less unified structure.
In Denmark
the Lindebjerg long-mound with its archaic Svaleklindt Group EN A/B
pottery (3060±100 b.c./3814±109 B.C.) may serve as example of this
process. Its seems to have grown by
accretion as new burials were added and covered. The final building stage
included all common megalithic features. Barkæer demonstrates a similar
evolution (Liversage 1992) and the excavation of Bygholm Nørremark
suggests a comparable process.
At Bygholm Nørremark Rønne
envisioned several successive stages of construction in which mortuary houses
and graves were enclosed by a stockade (Fig. 13.3). The enclosed area was
later covered with a trapezoidal mound surrounded by a stone enclosure
which retained the same form as the stockade. At an even later stage the mound
was enlarged and surrounded by a rectangular enclosure while a
primeval passage-grave was being constructed within the mound. Thus Danish
long-mounds are thought to have arisen as EN mortuary houses. However,
T. Madsen cautions that: Without totally rejecting the possibility of large mortuary houses, I
would suggest that we may be dealing with sequentially constructed long
barrows (Madsen 1979:315). This seems to be confirmed by the side by side and
linear arrangements of urdolmen in rectangular to slightly trapezoidal
long-mounds, as well as by the occurrence of multiple passage-graves in
long-mounds. Thus the evolution of Bygholm Nørremark shows that the
stockade, rather than the houses, provided the outline for the mound and the
stone enclosure. The stone enclosure is thus functionally equivalent to the stockade.
In this sense it is not just a retaining wall, but also a (symbolic?) barrier
that cordons off a long established sacred space.
Sequential mound construction must
have been the norm in the beginning throughout the TRB. The EN C
long-mound Flintbek LA 3, Schleswig-Holstein, shows a tiny oblong mound,
covering a central Konens Høj type grave. The mound was expanded in both
directions with three additional graves (Phase 1-4). A small rectangular mound with enclosure was
added, containing the perpendicular oriented Grave F and Extended-dolmen I-III, followed by a larger,
all-encompassing Phase 6 rectangular, fully megalithic enclosure, plus an
even larger northeastern rectangular mound/enclosure addition (Phase 7)
with large perpendicular
Extended-dolmen IV. The apparently succeeding nearby Long-mound
LA 4 exhibits a similar development from central dolmen with stone-circle
to axially expanded rectangular mound/enclosure (Zich 1992, 1994a, c, 1995a).
In Central Germany
sequential mound construction has been shown at Halle-Dölauer Heide
Mound 6 (Behrens and Schröter 1980 Fig. 36-42). The tomb is located
inside the famous causeway. The first Salzmünde burials (Graves I
and II) were said to be covered by village debris. This was followed by the
Salzmünde Grave 1 covered by the first "core mound." It was
followed by another grave covered by more mound fill. The next major phase was
a second core mound with grave. After this several additional graves and more
mound fill were added in conjunction with the construction of the trapezoidal
mound by the Salzmünde Group and Corded Ware culture. The final mound
elevation occurred in yet another Corded Ware burial phase, while still
another Corded Ware burial and a Bronze Age grave were inserted later.
The final grave was placed in the mound during the Late Bronze Age.
Similarly, the Baalberge primary grave of the Pohlsberg,
Latdorf, Central Germany (ID 6604) was a stone-cist
akin to the below surface urdolmen. Along with several additional chambers and
other graves, the mound later received a trapezoidal megalithic enclosure. But
sequential mound construction is not confined to long-mounds alone. Tumuli,
such as Schortewitz, also underwent such changes.
Figure 13.4. Danish urdolmen with EN C
pottery by mound shape. Of 36 dolmen
which contained identifiable Early Neolithic flasks 18 are in
long-mounds, 6 in tumuli, and 12 had no clearly defined mound.
All this should dispel the ever popular Long-house Theory
which derives the long-mounds from long-houses. But just in case it doesn't, it
should be noted that this naive theory also fails to explain the coexistence of
long-mounds and tumuli in the EN (Fig. 13.4).
|
TABLE 13.4
THE
NUMBER OF MOUNDS AND THEIR RESPECTIVE FORMS
LONG-MOUNDS 1261
Rectangular 490
Trapezoidal 472
Rhomboidal 1
Unknown 238
TRANSITIONAL 95
TUMULI
812
Oval 19
Round
76
Stone-circle 101
None/unknown 616
UNCLASSIFIABLE 243
Enclosures 10
Enclosure+mound 233
TOTAL 2411
The number of mounds and
their respective forms. It should be noted that out of 243 unclassifiable
mounds only 10 (4.12%) had no trace of a mound. This means that evidence of
an enclosure is highly likely to imply the existence of a mound, even when
there is no trace of one. The nearly square enclosure of Gaarzerhof
ID 5004 is counted here as a short long-mound.
|
The available data show only two possible
cases where two multi-side-stone dolmen might have existed in a single tumulus.
It is almost as rare to find more than one primeval passage-grave in a
tumulus. The early coexistence of
long-mounds and tumuli may thus be the result of a functional difference, where
long-mounds covered one or more chambers and other graves and the earliest
tumuli encompassed a single chamber. If additional burials had to take place in
the vicinity of a tumulus, a tumulus would be incorporated in a long-mound,
creating a complex mound as described below.
Regarding mound shape, it is logical
that as long as the prehistoric architects built small chambers, the form of
the mound did not rally matter. This may be yet another reason for the
coexistence of round and long-mounds. But as chamber size or passage-length
increased mound form became more important and advantages and disadvantages of
specific mound shapes became more marked.
Tumuli are especially efficient in
containing the large polygonal-dolmen. But some grand-dolmen in Mecklenburg
actually forced the architects to stretch the round enclosure into an oval one.
A similar process in the West Group stretched nearly round enclosures with
a kidney-shape into ever more elongated forms as the passage-graves increased
in length. Furthermore, polygonal-dolmen and oval primeval passage-graves are
most efficiently created in tumuli, where passage length can readily increase.
In fact, such an increase in length is virtually invited by round mounds.
The advantage of the long-mound is
that it accommodates the increase in chamber length of parallel chambers, such
as passage-graves. But long-mounds have disadvantages for the construction of
larger axial dolmen. Thus the advent of grand-dolmen caused a dilemma. The only
choice was to increase the mound's width or to rotate the chamber. This
probably was a major factor leading to the abandonment of the autochthonous
grand-dolmen in favor of passage-graves in Mecklenburg-Vorpommern. But the advantage of lengthening the chamber
sacrifices the easy lengthening of the passage.
In the TRB West Group, where
short passages are common, round mounds, possibly under the influence of
coexisting long-mounds, evolved into ever longer transitional mounds. In the
largest chambers the mounds and enclosures seem to loose their early
significance as a separate unit. The mound itself sometimes barely covers the
chamber which is almost immediately surrounded by one or two enclosures. Thus
at Thuine 874 the inner enclosure practically touches the chamber. This is
a clear indication that the "expense" of building ever larger
chambers forced an economic decision to reduce the mound and enclosure size in
proportion to increased chamber size in the West Group. This means that
the Dutch type mounds/enclosures together with their passage-graves were
designed as a whole unit, as Bakker has argued.
The same evolution may have occurred
among the gallery-graves, where the larger chambers occasionally provide
evidence of an enclosure close to the chamber. For example at Etteln the
suspected enclosure remnant was only about 70 cm from the side wall of the
chamber (Günther 1978:233 Fig. 4).
13.3 Enclosure shapes
Although mounds are more successfully
classifiable by shape than chambers, it remains unclear which, if any form, is
the oldest. Furthermore, the mound's shape is usually determined by the
enclosure. A likely exception is the grand-dolmen with nearly square enclosure
(mean dimensions 9.34:8.40 m) of Gaarzerhof (B46, ID 5004). Schuldt's
plan (1972:166 Fig.166) the enclosure covered with a round tumulus. But when
mound contours are carefully mapped tumuli, such as Klekkendehøj and
Gundsølille show a curious angularity (Fig. 14.1, Hansen 1993a:101
Fig. 101), remotely akin to pyramidal North American Woodland period
Mounds.
The number of mounds and their
respective enclosure forms are shown in Table 13.4. As a rule the
enclosure consists of border stones that form what is sometimes called a
curb. It is constructed with the same kind of megaliths used to construct the
chambers, even if no megalithic chamber is present within the mound. The
resulting enclosure classification is
given in Table 13.5.
The enclosure is seen to function as a
retaining wall designed to keep the mound from being eroded
(e.g. Schlicht 1979a). Of course, the relationship between the stockade
and the megalithic enclosures at Bygholm Nørremerk add a likely sacred
function to the megalithic enclosure, no doubt separating a hallowed ground of
the other world from its mundane surroundings, setting into motion an evolution
of elaborate enclosure designs. Thus the enclosure for the passage-grave Jordhøj, at Mariager is reconstructed as a
platform for offerings (Hansen 1993a:19 Fig. 22). Kong Swendshøj's
enclosure is restored as supporting an actual catwalk (Fig. 9.3), and the
small enclosure stones together with a second ring of even smaller border
stones are reconstructed as outlining a
"paved" catwalk that surrounds the tumulus at the level of the
elevated passage entrances of Klekkendehøj on Møn (ibid. p. 14 Fig. 12).
|
TABLE 13.5
Classification of enclosures
Rectilinear
Rectangular
Trapezoidal
Rhomboidal
Circular
Round
Oval
Transitional
Dutch (Oval/kidney-shaped)
Complex enclosures
Side joints
End joints
Divisions
Secondary enclosures
Double enclosures
Rectilinear
Double Rectangular
Double Trapezoidal
Transitional
Dutch
Circular
Unlike
the chamber classification, this classification does not follow a specific
chronological scheme, although some forms occur only in the EN and some
others can be shown to exist only in the MN.
|
Wooden enclosures are very rare.
Trapezoidal ones, presumably similar to Bygholm Nørremark include
Stengade I, Langeland, which contains Konens Høj-type graves and
EN C pottery. This mildly trapezoidal structure's mean dimensions were
approximately µ30.65:2.84 m or 87.5 m2. Half the size of Stengade I is
the trapezoidal bedding trench of Teglværksgåde (µ=14.00:2.96 m,
41.44 m2). In function, form and size they resemble
megalithic enclosures and are, therefore, treated in the analysis in the same
manner as their megalithic counterparts. Surprisingly, the literature does not
mention circular wooden mound enclosures for the TRB.
|
TABLE 13.6
Frequencies of mounds and enclosures
in relation to megalithic chambers
TYPE RE TR
RO LN RD
OV SC TU
DU MN ??
EN TOTAL
|