Further Investigation of the Acheulian Site at Cuxton
FURTHER INVESTIGATION OF THE ACHEULIAN
SITE AT CUXTON
R.J. CRUSE
With contributions from
D.R. Bridgland, P. Callow, A.P. Currant, R.N.L.B. Hubbard,
N.C. Debenham and S.G.E. Bowman
1. SYNOPSIS
Rescue excavations in 1984 at the well-known Palaeolithic site at
Cuxton showed that the deposits were more complex than those
accessible in the 1962-63 excavations. Under the hand-axe-rich
gravels another assemblage with retouched flake artefacts but lacking
hand-axes was found. Both assemblages were redeposited and had
been incorporated in gravels tentatively linked to the Kempton Park
gravels of the Middle Thames. The gravel was overlaid with redeposited
loam, whose fine silt component has been dated by thermoluminescence
to greater than 100,000 years bp. Fauna! and pollen
evidence from this important site has also been examined.
2. INTRODUCTION
In February and March 1984, the Maidstone Area Archaeological
Group carried out a limited excavation in advance of the construction
of a drive at no. 15 Rochester Road, Cuxton (N.G.R. TQ 71126655).
The original objective was to seek further evidence of Roman burials
associated with the inhumation found i n the adjacent garden in 19621
and also to investigate whether the nearby Acheulian site2 extended
eastwards from the Rectory across the A228 road.
1 P.J. Tester, 'A decapitated Burial at Cuxton', Arch. Cant., lxxviii (1963), 182.
2 P.J. Tester, 'An Acheulian Site at Cuxton', Arch. Cant., lxxx (1965), 30.
39
--
'\ RECTORY
\SITE
\
\
\
\
R.J. CRUSE
'
'\
\
\
Fi g. 1. Location Maps.
BTre;"'h
y
fnet,
A
,,,,.- -
*Roman
Burial
,UC
10 88
An area of 7 sq. m. was examined without finding evidence of
further Roman activity. However, the second objective was amply
realised when a hand-axe was found in a stratified context in the first
day's excavation. At this point, the present writer was invited to
progress the investigation of the palaeolithic deposits. From a brief
review of the literature, it was clear that Peter Tester's 1962-63
excavations had firmly established the importance of the site to
palaeolithic studies and that the major objective should be to obtain
as much independent evidence as possible on the formation and
characteristics of the gravel deposit and its associated artefacts. 3
3 D.A. Roe, The Lower and Middle Palaeolithic Periods in Britain, London, 1970.
40
......
0 m•trN
No
15
INVESTIGATION OF THE ACHEULIAN SITE AT CUXTON
3. EXCAVATION
The area investigated is shown in Fig. 1. Trench 1 extended from the
roadside path to the garden hedge (area BXY). When the topsoil was
removed, it was apparent that the gravel deposits were best preserved
in the south-east comer, where they were furthest from the garden
path and from the disturbance associated with the 1921 widening of
the main road. 4 A succession of gravel layers was encountered, which
were distinguished by changes in their physical and component
char:;tcteristics (Fig. 2.1 and Table 1). Each gravel layer was carefully
trowelled and all potentially worked flint was collected for postexcavation
study. The position of each hand-axe and of organic
fragments was recorded. Sample areas of the finer gravels were
sieved, without yielding any additional flintwork. As trench 1
deepened, its width was rationalised to 0.80 m., following the
southern section. A sondage at the western limit established the
position of the underlying frost-shattered chalk.
Trench 2 was 2 m. wide and continued from the hedge into the
garden (Area AXY). It was soon apparent that considerable quantities
of the upper gravels had been removed. Layers 10 to 14 only
survived in the west-north-west corner. Layer 9 was better preserved
in the northern section, but was increasingly severely disturbed
towards the south and east. Excavation was, therefore, restricted to a
1.50 m. trench in the northern area against the hedge (XY) and this
continued down to the chalky debris. On the final day, the baulk
(XY) was excavated down to Layer 8.
The exposed profiles were examined by Dr D.R. Bridgland.
Drawings of the main section AXB (Fig. 2.1) and of the baulk
(Fig. 2.2) were prepared with his assistance. The trenches were then
backfilled and the drive was constructed. In demolishing his workshop
to construct a garage, Mr Cogger found an Acheulian hand-axe
built into the wall ( confirming modem gravel extraction from the
garden). A further hand-axe was found c. 1.50 m. below the ground
surface when Mr Cogger dug a garage pit. Both these axes, together
with four unstratified hand-axes found by Mr Cogger in 1983 when
planting rose bushes in the front garden, had the characteristic 'hard
limey deposit' previously observed upon the 1962-63 artefacts and
upon some of the axes from the upper gravels in Trench 1. 5
The 1984 excavations yielded c. 20 cu. m. of stratified gravel
4 D. Church, Cuxton, a Kentish Village, Sheerness, 1976, 96.
5 As they appear to be derived from similar contexts, these unstratified axes are
labelled 12 ? in Table 3.1.
41
A
KEY
+>-
N ., C> .. Gravel
Q Flint
666 6 il. Chalk
... Sand
~ Clay
•• Bone
• Axe
.....
-,. Fault
RJC
84
Projected 2
1
Not
Excavated
1m
Fig. 2.1. Section AXB (South-west Face).
B
ft m
17
16
5
15
14
45
OD
INVESTIGATION OF THE ACHEULIAN SITE AT CUXTON
deposits, which contained 220 artefacts, including 9 hand-axes and 23
flake tools. The hand-axes were restricted to the upper gravel levels,
which contained between 5 and 25 artefacts/cu. m. This is modest by
1962-63 excavation standards, where the gravel yielded over 200
artefacts/cu.m. The highest stratified concentration found in 1984 was
in the lower gravel layers 3 and 4 which yielded c. 75 artefacts/cu.m.
(Table 3.1).
TABLE I
Description of Layers
15 Top soil/disturbed.
14 Redeposited eolian sand with some silt.
13 Unbedded, grey sandy gravel.
12 Bedded, fine chalky gravel.
11 Unbedded, very coarse dark gravel.
10 Unbedded, coarse gravel.
9 Planar bedded, sandy gravel with medium coarse chalky
layers and fine sandy lenses.
8 Cross stratified yellow sand.
7 Planar bedded gravel, partially clast-supported.
6 Cross bedded, gravelly sand with foresets.
5 Dark fine gravel, partially clast-supported.
4 Orange-yellow, very coarse, slightly clayey gravel.
3 Grey very clayed sand.
2 Clean yellow sand with some clay lenses.
1 Chalk rubble (presumed upper surface of chalk).
4. SPECIALIST INVESTIGATIONS
4.1 Analysis of the Gravel Deposits
Gravel Characteristics (Figs. 2.1 and 2.2)
(D.R. Bridgland)
Although the limited extent of the excavation did not permit a
comprehensive assessment of the sediinents, the available sections
indicated a considerable preponderance of the gravel, often extremely
coarse, over sand. They gave every indication of similarity to the
assemblage of sedimentary types of Shakespeare Farm Pit,
Allhallows,6 and thus can probably also be attributed to a braided
6 D.R. Bridgland and P. Harding, 'Palaeolithic Artifacts from the Gravels of the
Hoo Peninsula', Arch. Cant., ci (1984), 41.
43
R.J. CRUSE
X
1m
Fig. 2.2. Section XY (from Trench 2).
44
y
17m
16m
15m
RJC
84
INVESTIGATION OF THE ACHEULIAN SITE AT CUXTON
fluvial environment, laid down during a Pleistocene periglacial
episode.
It is likely that both the clast-supported and matrix-supported
gravels seen at Cuxton have resulted from deposition on longitudinal
bars and as channel 'lags'. Most of the small, thin sand lenses
were either cross-stratified or ripple-laminated. The thick sand bed
(layer 2) near the base of the sequence, showed little sign of
stratification other than a few clay partings and lenses of very clayey
sand, which may represent a waning flood event. At its base this sand
clearly overlies a 'lag' deposit (layer 1).
Palaeocurrent measurements were possible from the two main
cross-bedded sand layers 6 and 8. The former indicated flow towards
043° (mean of 2 readings) and the latter toward 346° (mean of
3 readings). Given the very low number of measurements, these
palaeocurrent records are in keeping with deposition by the main
river. A single fault was observed in the section, with a downthrow
towards the south-east, which may suggest a partial collapse of the
deposits towards the centre of the contemporary Medway channel.
On chalk, however, such structures are commonly caused by solution
of the underlying bedrock surface.
The length of time represented by the depositional sequence is
clearly important. Although each of the gravel units within the
sequence probably represents a single flood event, each was truncated
by later flood events with varying amounts of erosion of its
upper levels. The erosion of layer 6, prior to the deposition of layer 7,
was very clear. It is impossible to quantify the extent to which these
units have been eroded and reworked by later floods, or the length of
any periods of quiescence between flood events, for which there is no
record in the sedimentary sequence. The time interval represented by
the aggradation of the Cuxton deposits cannot, therefore, be estimated.
If the amounts of erosion were modest, it is possible that the
deposits could have accumulated in a few years.
Clast Lithology
The composition of the Cuxton gravels has been analysed and
compared with samples from elsewhere in the Medway basin: from
Stoke (N.G.R. TQ 822748) on the Hoo peninsula, from Aylesford
(N.G.R. TQ 726597), where Medway gravel overlies Folkestone
Beds and from Lower Hayesden, near Tonbridge (N.G.R.
TQ 563459) (Table 2.1). The last of these sites overlies Weald Clay,
but is only 2 km. downstream of the Hasting Beds outcrop. As the
Aylesford and Cuxton samples contain a high proportion of nondurable
calcareous clasts (predominantly entish Ragstone and
45
TABLE -1
Terrace grai·el composition in i·arious pans of the Medway ,,alley
total sample excluding calcarceous
Lower Hayesden Aylesford Cuxton Stoke Aylesford Cuxton
unit unit unit unit
Provenance Lithology 1 2 l 2 9 7 IA lB 1 2 9 7
Central Weak H.B. sst. 32.2 37.8 0.5 3.1 5.9 0.2 0.5 0.7 7.7 7.0
H.B. sit. 20.2 18.3 1.0 0.3 2.7 0.2 1.4 0.9 3.2
H.B. iron. 23.9 25.4 1.0 1.7 0.3 4.8 0.6 0.5 2.0 2.5 0.9 5.6
clay/iron 0.2 0.7 0.4 0.8
TOTAL 76.3 81.8 1.0 3.4 4.5 14.1 0.9 0.9 2.0 5.0 11.1 16.6
Lower dense cht. 2.7 2.1 12.7 13.3 6.3 10.7 11.8 8.9 24.7 19.7 15.4 12.6
Greensand porous cbt. 3.1 4.4 7.7 5.2 6.3 15.6 8.2 5.6 14.9 7.7 15.4 18.5
wea. chert 4.4 0.8 7.7 8.1 1.4 2.7 5.7 3.1 14.9 12.0 3.4 3.2
sst/chert 0.6 0.3 0.2 0.9
sil. sst. 0.2 0.3 0.2 1.1 0.2 0.2 0.4 1.3
ironstone 0.7 0.2 0.8
ragsrone 45.2 26.6 1.0 0.9
wea. rag. 1.0 4.8 0.5
ca/c. sst. 1.7 1.2 0.7
TOTAL 11.0 7.5 76.0 59.4 16.1 32.2 25.9 18. l 16.l 32.2 35.0 36.6
Chalk n. Tert. fit. 7.3
escarpment Tert. flint 5.2
dipslope TOTAL FLINT 12.5
Chalk
Exotic Pal. chert
metaqutzt.
Non Secific ironstone
TOTAL COUNT 481
Notes: 16-32mm size range counted.
3.9 21.0 31.8
6.9 0.2 2.7
10.8 21.2 34.4
0.7
0.2 0.2
1.0 2.6
389 575 421
18.2 35.4 42.6 53.8 41.0 47.2 44.4 41.9
3.8 4.1 30.0 27.2 0.3 3.9 9.4 4.8
22.0 39.5 72.6 81.0 41.4 51.1 53.8 46.8
57.3 14.3
0.3 0.4
0.2
0.7 0.2 2.0 3.9 1.7
286 441 636 426 295 284 117 372
Abbreviations - H.B. = Hastings Beds; sst. = sandstone; sit. = siltstone; iron. = ironstone; cht. = chert; wea. = weathered;
sil. = siliceous; rag. = ragstone; calc. = clacareous; n. Tert. = non-Tertiary; flt. = flint; Tert. = Tertiary; Pal. = Palaeozoic;
metaqutzl. = metaquartzite. sst./chert = cherty sandstone/sandy chert
R.J. CRUSE
Chalk), material which rapidly disappears from the gravels downstream
of its source outcrops, data are also presented for these sites
which exclude this non-durable material. The presence of substantial
quantities of material from the Central Weald and the Lower
Greensand in the Cuxton samples provides a strong argument in
favour of a mainstream Medway origin. The upstream component of
the gravels would be expected to become progressively less common
with distance downstream as a result of dilution by rock types from
the lower reaches of the Medway valley. This is evident from the
Lower Greensand component, which decreases between Aylesford
and Cuxton. In the durable fraction, however, this decrease is rather
less than might be expected, since considerable quantities of flint
ought to have been added to the river's bed load as it passed through
the gap in the North Downs. A probable reason for this is that,
although it is upstream of the North Downs Chalk outcrop, the
Aylesford gravel already reflects the input of flint from tributary
valleys flowing in the Gault Vale, along the southern side of the
escarpment. The increase in Hastings Beds content between Aylesford
and Cuxton is somewhat surprising, as this material is typically
soft and friable and would be expected to dwindle rapidly downstream.
This may be a consequence of qifferent erosion and transport
regimes in the Medway when the two types of gravel were deposited,
as exact contemporaneity between the three sites has not been
established. The substantial incorporation of chalk in the Cuxton
gravel is noteworthy, as this material rarely survives more than a few
kilometres downstream from its source.7 Not surprisingly, there is a
marked decrease in the amount of Lower Greensand and especially,
Hastings Beds material between Cuxton and Stoke, particularly when
the non-durable are excluded from the Cuxton data.
Discussion
The composition of the Cuxton gravel in comparison with deposits
both upstream and downstream in the Medway valley, supported by
the evidence from palaeocurrent measurements and the thickening at
the deposits towards the modern river channel, strongly suggests that
the main river was responsible for its deposition. The suggestion by
Zeuner& that the deposits were indicative of a small tributary stream
cannot therefore be supported.
A most important consideration is the relationship of the Cuxton
7 D.R. Bridgland, 'The rudaceous Components of the East Sussex Gravels',
Quaternary Studies, 2 (1986), 34-44.
8 Tester, op. cit. in n. 2, 34.
48
Rectory
19 ste
16
15
14
13
12
11
mOD
Q 1,0
1s9it84e
- - ..... ...... -...,.
'
'
'
•
CHALK
V
GRAVEL
2p 3,0m
'
'
vA v Handaxes
•: Other artefacts
'
SANb, -,;: '
' ....
'
'
' ' \.-Original
. Profile?
....
\
\
\
Raily
Cuttin
RJC
86
Fig. 2.3. North-west-outh-east Section across Terrace.
----
'
-- -
--. -
'
R.J. CRUSE
gravel, with its Acheulian implements, to the Pleistocene succession
in the Medway basin and its links with the Thames sequence. The
occurrence of this impressive Acheulian industry at a lower (and
therefore younger) position in the Medway valley than that occupied
by the comparable site at Swanscombe in the Thames valley has been
a problem since the Cuxton site was first discovered. The discovery at
Shakespeare Farm Pit, Allhallows, of comparable pointed hand-axes
in the higher (and therefore presumably older) Shakespeare gravel9
underlines the difficulty with the Cuxton occurrence (Fig. 3.1).
The most recent work upon the correlation of the lower Medway
terraces with those of the Thamesll>--1:l has traced the terrace
aggradations downstream in each valley to the confluence area
between the Hoo Peninsula and Southend. These elevation studies
indicate that the Cuxton deposits are most likely to be an upstream
extension of the Binney Gravel (Fig. 3.1), which is correlated with
the Kempton Park Gravel in the Thames valley (Fig. 3.2). A
mid-Devensian age is, therefore, indicated.
The occurrence of well-preserved and abundant Acheulian artefacts
in gravel low down in the Medway terrace sequence at Cuxton
remains a problem, as hand-axes are not abundant in the Kempton
Park Gravel and when found are normally rolled and very obviously
derived. Tester suggested13 that Cuxton represented a low sea-level
phase within the '100 ft.' aggradation peri•od (comparable with that
cited by King and Oakley to explain the Clactonian gravel at Little
Thurrock), 10 as he wished to retain a close chronological link
between the typologically similar Middle Acheulian industries. There
is little geological evidence for such a drastic and short-lived rejuvenation
interval and palaeolith typology would rarely be claimed
these days to be sufficient grounds for proposing such disruption to
conventional terrace stratigraphy.
Perhaps the most plausible explanation is that the Cuxton artefacts
were formerly contained in earlier fine-grained deposits, which
preserved them in near mint condition and they were then incorporated
by reworking into the Binney Gravel. This hypothetical early
deposit must have been situated very close to the present Cuxton site,
as most of the material has experienced very little transport. It may
well have been entirely eroded away.
9 Bridgland and Harding, op. ciJ., 52.
10 W.B.R. King and K.P. Oakley, 'The Pleistocene Succession in the lower Parts of
the Thames Valley', PPS, i (1936), 52-76.
11 Bridgland and Harding, op. ciJ., 52 and Fig. 3.
12 D.R. Bridgland, The quaternary fluvial Deposits of north Kent and east Essex,
ungublished Ph.D. thesis, City of London Polytechnic, 1983.
3 Tester, op. cit. in n. 2, 43.
50
150
140
130
120
110
100
90
80
...... E70
60
50
40
30
20
10
0.0.
ui
Q)
.c
0
0
a:
···C·o··b·h·a·m Park
Wouldham
rw:1 · · co;,;;a;,; PBr!c Gravel
Lodge Hill
·Windmill Inn High Northward Hm
•••·•••Halstow. ? m. ,Hlgh Hi,i$,0W Grav6t
Clinch s· ... .. ;;;:· .... _Chnch Street
treet Grave/ ;;.•. • •• ••
Tower Hill ..... .., .... •. St Mary-Hoo
I Newfd·;;,,-rm .. ........ Oegenhamfarm ?
Fnndsbury O O .Gillingham O 00Hooooooo Cold Arbour Turke"y Hall ·:;.;•: Dagenham Farm Gravel
%:? wJ
T,le Barn .,._.
•
• m···········;fflt.'1.r:,, ···· ····s,;.;,l-;,,·······•· .. ····•· ... {f. Fnfilii2ury ... ···• •··•• • •• NewhaJJ _-,:J:;.! Slough '/UJBre Gra11el
---w;;;,_:1f ::::::::::::::;;i;;;;:;,;;::'."····
0
km
5
Fig. 3.1. Long Profile of Lower Medway Terraces (modified from note 6, Fig. 3).
INVESTIGATION OF nm ACHEULIAN SITE AT CUXTON
SMOIJB41f\f
MOlSleH 461H
Ii
S I•
5
Ii' g,
lt
R.J. CRUSE
4.2 Analysis of the 'Sandy Loam' (R.J. Cruse)
Particle size analysis was carried out by Dr J.A. Catt upon samples
from layer 14 and from the loam exposed in the rectory garden (Table
2.2). He considered layer 14 to be a very well sorted fine sand, with
only traces of clay, silt or coarse sand and interpreted it to be a
windblown sand. Although the analysis of the rectory sample has a
peak at the same particle dimensions as layer 14, it is less well sorted,
with more silt, clay and coarse sand. Dr Catt suggests that this may be
the re$ult of an eolian sand being mixed with some loess. As layer 14
was only vestigial, the rectory sample is probably more representative.
In his 1962-63 excavations, Mr Tester found the loam to contain
fragments of water-worn flint and a few 'unrolled' artefacts. There is
thus little doubt that the loam is a redeposited mixture of materials.
Dr P.A. Gibbard's recent review of the loams and brickearths in the
Middle Thames valley (his 'Langley Silt complex'), has confirmed
that they often overlie Taplow or younger gravels and that they have
given thermoluminescence (TL) dates ranging from the late Wolstonian
to the late Devensian. 14
Size (μ,m)
1,000 - 2,000 )
500 -1,000 )
250 -500 ) Sand
125 - 250 )
63 -123 )
31 -63 )
16 -31 )
8 -16 ) Silt
4-8 )
2-4 )
2 Clay
TABLE 2.2
Particle Size Analysis
Cuxton Cuxton
layer 14 Rectory
'loam' 'loam'
1.4 2.3
3.3 5.4
17.4 28.7
68.4 26.5
2.0 10.1
1.4 7.0
0.7 3.4
0.6 4.3
0.7 2.7
0.6 2.2
3.5 7.4
Middle Thames
'Langley Silt'•
1.3
2.1
6.4
7.3
8.7
19.0
14.1
7.7
4.4
3.4
25.6
• Average of nine sediment samples (from note 14, Table 1)
+ from note 15, Table 1.
Typical
Loess+
0.0
0.0
0.0
0.1
0.8
27.2
30.9
11.7
4.7
1.9
22.7
14 P.A. Gibbard, The Pleistocene History of the Middle Thames Valley, Cambridge,
1985, 56, 139.
52
INVESTIGATION OF THE ACHEULIAN SITE AT CUXTON
4.3 Section Across the Terrace (R.J. Cruse)
Investigation of the Cuxton terrace profile by Mr A. Daniels established
the section shown in Fig. 2.3 by augering at locations halfway
down the kitchen garden and in its extreme south-east corner. Once
the topsoil and disturbed ground were penetrated, he encountered a
sandy loam with a few flints which continuously capped a series of
gravel deposits with variable amounts of chalk in them. At the base of
this gravel was a narrow band of stained flints in a darker coarse
gravel (as was observed in 1962-63) and then chalky debris was
encountered. This profile demonstrated that the chalk bench and the
terrace deposits slope gradually (1 in 24) towards the south-east.
Having established that the sandy loam overlies the gravel, a
convenient location for a loam sample for thermoluminescence (TL)
dating was sought. The exposed loam in the garage cutting of no. 11
Rochester Road, some 30 m. to the north, was selected and the TL
determination gave a minimum age of 100,000 years bp (Appendix
1).
"'
"'
"'
"'
100
1
AaQ,tMM!