Knowlton Landscape Project summary

Summary of field work 2002-3


By John Gale, Paul Cheetham, Joanna Laver and Claire Randall



A second season of field evaluation took place between 4th August and 29th August 2003 based at High Lea Farm in the parish of Hinton Martell in East Dorset.  Investigations into the nature and extent of a largely levelled barrow cemetery, consisting of geophysical and topographic survey as well as targeted excavation were undertaken, in addition preliminary examination of a probable Late Iron Age/Romano British rural settlement was also undertaken. The following interim report outlines the findings of this season’s investigation alongside those of the previous summer’s work.


Fig 1 Area excavation of Barrows 1 and 2 photographed from the air in September 2003. © John Gale



As part of the research design brief for the Knowlton Project (Gale 2002, Gale 2003) concerning the evaluation and quantification of prehistoric settlement and associated activity within the Allen valley of East Dorset, attention has been partially focussed on the extensive barrow cemeteries that dominate the valley.  Several concentrations of round barrows (mostly surviving only as groups of ring ditches in an archaeological rich prehistoric landscape that has been devastated by intensive arable farming), occur over a 5km stretch of the Allen Valley between the Knowlton complex to the north, and the High Lea farm Group to the south.  Very little is known about any of these funerary monuments beyond those general assumptions drawn by analogy with examples studied elsewhere in Wessex and further a field.  The only published material concerning any of the barrows found within the thirteen parishes that broadly constitute the Allen valley (Wimborne St Giles, Woodlands, Gussage All Saints, Horton, Chalbury, Moor Crichel, Hinton Martell, Witchampton, Hinton Parva, Pamphill, Holt, Colehill and Wimborne) are largely descriptive, (RCHME 1975; Grinsell, 1959 1982) and are now quantitatively out of date.  The majority of the defined barrows contained within the RCHM(E) (Royal Commission of Historical Monuments England – now contained within English Heritage) and Grinsell listings for the above parishes are found within the three parishes of Wimborne St Giles, Woodlands and Gussage All Saints and the greater number of these in the parishes of Wimborne St Giles and Gussage All Saints.  Within these latter parishes are the large cemetery groupings of Oakley and Wyke Downs respectively which are strictly speaking outside of the Allen Valley and subsequently are not considered within this study.

The clear foci for the funerary monuments within the
Allen Valley seems to be an association with the river itself and perhaps secondarily with the chronologically earlier henge complex at Knowlton.  Interestingly all of the barrows within the Allen valley (so far defined) are located on the east bank of the river.  Whilst the east bank of the Allen Valley contains long shallow slopes and greater space for the location of barrow groups the geography of the west bank rarely precludes space for such construction.  It would seem therefore, that a conscious choice was made to avoid the western bank for the construction of funerary monuments within the prehistoric period.  At this stage it is unclear as to why the west bank was avoided but it is possibly a factor related to the contemporaneous belief systems of the resident community or possibly a more mundane issue concerning territorial boundaries i.e. the river itself and rights over land adjacent to it.

Whilst the greater concentration of barrows within the Allen valley appear to be located around the Knowlton Henge complex their distribution would seem to extend from here to the south, mainly in linear groupings more or less following the line of the river Allen.  The end of these concentrations would appear to be found at the High Lea Farm Barrow group between the villages of Witchampton and Hinton Martell. 

As with the majority of barrows in the Allen valley the High Lea Farm group is extremely poorly preserved with only two of the group surviving as upstanding earthworks.  However, both of the surviving barrows owe their partial conservation to their location within the headlands of the field known as Kings Close where the plough is frequently lifted and turned and therefore consequently is subject to a lesser degree of damage. Elsewhere in the field the presence of numerous ring ditches are attested through crop marks visible on aerial photographs.  One such photograph (NMR 4437/04, SU 0006/8) taken obliquely when the field contained a crop of pea in 1989, illustrates a dense concentration of ring ditches and other maculae which appears to represent a definable cemetery group of ploughed out barrows, consisting of at least three distinct linear groupings.

Whilst such evidence attests to the presence and location of an extensive barrow group it provides little effective information as to the sequential chronology, or nature of the barrow clustering.  In addition, whilst the barrows have had much (if not all) of their above ground features destroyed, the survival of below ground features (grave cuts, earth fast cremation urns etc.) remains undetermined. With these factors in mind it was decided to initiate a two step evaluation of the cemetery contained within Kings Close over two successive seasons commencing in 2002.

During the examination of the aerial photographic evidence available at the National Monuments Record centre at Swindon during the winter of 2002, the presence of an undated crop mark enclosure was noted in a nearby field (Great Higher) within 400m of Kings Close to the south.  Because of its close proximity to the barrow group the research design for the field programme of 2003 was extended to include exploratory investigations into the extent, state of preservation and initial dating of this site. 


Back to top


Kings Close evaluation  - step 1 (August – September 2002)


During a four-week period a programme of fieldwork was instigated to explore the location and survival of archaeological features associated with the known barrow group at High Lea Farm. The 15ha field known as Kings Close was subjected to an extensive programme of Geophysical analysis using a range of techniques and equipment to both locate and explore the surviving features of the barrow group (see below).  In the wake of the geophysics programme a sample area of the field was defined to assess the extent of archaeological material within the current plough soil and to examine if the barrows were still being actively disturbed by the arable regime.

Test-pitting 2002

A total of 67 (1m x 1m) test pits were hand excavated over an area of 25940 sq. m located in the north of Kings Close.  The sampled area was chosen to maximise the range of features evidenced by both the geophysics results (see below) and those contained within the aerial photograph (NMR 4437/04, SU 0006/8) taken in 1989.  The test-pits were placed systematically within a grid and were spaced at 20m intervals.

After marking out each test-pit was hand excavated and the plough soil was removed and screened through a 6mm-mesh sieve.  Soil layers and or features below the plough soil were not excavated but when present were contextualised and recorded.  With few exceptions the plough soil (defined here as the last cultivation horizon resulting from the 2001 ploughing of the field), lay on top of a further soil layer largely differentiated from the uppermost soil by a larger quantity of chalk within its matrix.  In all likelihood this soil is probably an earlier plough soil, the result of an episode/s of deeper ploughing in recent times.

Material recovered from the active plough soil throughout the test-pitting exercise consisted of minimal quantities of cultural material mostly comprising of relatively modern ceramic material (brick, tile and small quantities of earthen-ware, which are probably of Verwood types.  With the exception of a singular flint tool (an awl), material dateable to the late Neolithic or the early-mid Bronze Age consisted of a few struck flint flakes with no evidence of prehistoric pottery having been recovered.

The excavation of the test-pits failed to find any clear evidence that the barrows were being actively destroyed at the present time and certainly no material was recovered within the plough soil to indicate that funerary deposits were being actively destroyed.

Test trenching 2002 

Alongside the systematic sampling strategy employed to evaluate the surviving archaeology at the site, limited trial trenching was applied in targeted areas where specific features were identified as having survived evidenced by the results of the geophysical survey (fluxgate gradiometry).  Two adjacent barrows were chosen for further investigation and prior to excavation both barrows were subjected to further detailed geophysical investigation (see Geophysics below).  To confirm the presence of the ring-ditches of both the chosen barrows and to tentatively explore the area bounded by the ring ditches two small excavation trenches were aligned to cross each of the barrows ditches. The trial trench located on the smaller of the two barrows was subsequently extending to its centre.  At this stage the intention was to confirm the presence of archaeological deposits only, with the option to excavate being part of a later phase of investigation (Step 2 below).

Trench 1

Trench 1 was positioned on the larger of the two barrows and was 1m wide and 8m long.  The modern plough soil was removed from the trench by hand and was screened through a 6mm sieve.  At the base of the plough soil only two contexts were recorded, the natural bedrock and the fill of the barrows ditch.  The barrow ditch fill was approximately 1.3m in width and was bounded on either side by weathered chalk bedrock in which the scoring marks of the plough could be clearly seen.  Materials recovered from the plough soil were small in number and consistent with those from the test-pits – there was no evidence of funerary deposits either in situ or disturbed.

Trench 2

Trench 2 was positioned on the smaller of the two barrows and consisted of an ‘L’ shaped trench. The longer primary axis of the trench was 10m in length crossing the barrow from outside of the ring ditch ending at its approximate centre where a 3m extension was excavated perpendicular to the main axis. As with Trench 1 the plough soil was removed by hand and screened through a 6mm-mesh sieve.  Material recovered from the plough soil was similarly consistent with previous examination of the plough soil elsewhere on the site, with no evidence for funerary deposits or material.  As with Trench 1 the presence of a soil fill that bisected the excavation trench, is entirely consistent with the ditch predicted by the geophysics results. This ditch fill was approximately 0.75m wide and was bounded internally and externally by plough damaged natural chalk.

Trench 3

In addition to the trial trenching of a sample of the known barrow population within the group a third trial trench was instigated to examine the anomalous results provided by aerial photography and geophysical investigation in the north of Kings Close.  Whilst aerial photography determined the presence of a number of maculae in the north of the field which appear as crop marks consisting of ‘filled’ circles they did not similarly register on the magnetometer survey of the same area.  Whilst it could not be discounted that these features were possibly pond barrows (the aerial photograph identifies at least five similar anomalies) it seemed that they were much more likely to be geomorphological in origin.  To further explore these features Trench 3 was located over one of them. The trench was 10m long by 1m wide. 

As with previous trenches all of the plough soil was removed by hand and screened through a 6mm-mesh sieve.  Below the plough soil a well-sorted humic soil was encountered over the whole of the trench, which was systematically removed by hand and screened.  The depth of this deposit was nowhere greater than 10cms and it appears to be an earlier agricultural horizon that had survived recent plough activity because of the greater soil depth in this part of the field. Below this layer a further continuous layer of soil was found which was remarkably chalk free.  A sondage was excavated in the north-western corner of the trench to explore this soil and 0.6m of it was removed without encountering any change in the deposit.  It would appear that the deposit is a colluvial soil fill, part of a much deeper deposit that is probably natural in origin and is almost certainly a doline/sinkhole, not an uncommon feature in areas where the cretaceous chalk beds are closely associated with Tertiary sand and gravel deposits (for a discussion on the formation of Dolines in Dorset see House 1991).  The colluvial deposit contained, as one might expect, a chronological mix of cultural materials with pottery being the most numerous material recovered.  The fabrics noted consist of later earthen wares (of Verwood types) and also fragments of later prehistoric material albeit in relatively small quantities.  The trench was not excavated beyond a depth of 1.2m below the modern ground surface.


Back to top



Kings Close evaluation  - step 2 (August – September 2003)


The programme of archaeological investigation instigated in 2002 had clearly established the presence and location of surviving in-situ archaeological deposits within the High Lea Barrow Group but the extent of survival was as yet weakly defined.  Step 2 of the evaluation process, instigated during the late summer of 2003, was targeted to investigate a sample of those known deposits via the area excavation of the two ring-ditches (barrows) that had been the subject of the trial trenching in the previous season (fig.,1, 2 and 3).  In addition, further geophysical investigation was undertaken within Kings Close, to complete the area survey and to apply complimentary techniques to defined ring-ditches in an attempt to draw as much information as possible from the surviving archaeological deposits (see below).

Area excavation of Ring –ditches 

An area of 1043 sq. m was marked out for excavation to contain the two adjacent barrows sampled in the previous season’s fieldwork.  Having previously determined the archaeological characteristics of the plough soil it was subsequently removed by machine, down to its interface with underlying deposits.  Over the western half of the excavation trench the plough soil overlay natural chalk with the exception of the soil fill of the surviving ring ditch.  Whilst the western quarter of the larger barrow was similarly exposed by the arc of its soil filled ring ditch the eastern two thirds was covered by a thin layer of soil.  This soil layer, which clearly lay over the ring-ditch fill was probably a remnant of an earlier agricultural horizon falling below the plough depth of the most recent agricultural activity within the field.  The layer was similar to that encountered within some of the test pits of the previous year, and was identical to a similar soil found in Trench 3 (see above). The greater depth of soils on this part of the site is most certainly due to natural fluctuations in the topography of the field, with some dips and troughs in the bedrock containing slightly greater depths of soil. Such an occurrence is if course partially masked by the topographical smoothing effect at the surface caused by such agricultural processes as ploughing and horrowing.

Ring ditch – Barrow 1

Following the removal of the overburden of plough soil by machine stripping the residual plough soil was trowelled back to reveal a heavily plough scarred natural chalk surface.  Also revealed were the surviving remains of the cut and fill of a circular ring ditch forming an almost perfect circle 17m in diameter.  The width of the remaining ditch was on average 0.76m.  It was decided to excavate a 20% sample of the remaining ditch achieved through the placement of eight box-sections aligned on the cardinal and first level directional points of the compass (north, north-east, east, south-east etc.). All eight box-sections revealed a similar sequence of deposits within a heavily truncated ditch that was never greater than 0.3m deep.  The ditch cut was flat-bottomed only marginally narrower at the base than at its truncated top.  The ditch fills were uniformly similar, consisting of a primary chalk slumping occurring at both sides possibly suggesting that the original barrow may have had an outer bank as well as an internal mound. Alternatively this initial chalk slumping may have been the result of the collapsing of the upper sides of the ditch walls soon after construction, especially so, if as is likely the ditch walls were cut to a shear angle. The secondary fills of the ditch appeared to the product of the slow filling of the ditch with colluvial material that contained a relatively high proportion of chalk. The fills were generally devoid of artefactual material although the secondary fills did produce a small number of heavily gritted but largely non-diagnostic pot-sherds which may be Bronze Age in date. 

It is clear from the filling of the ditch that it has almost certainly been heavily truncated by farming activity over a considerable period of time.  It is a matter of conjecture as to how much of the ditch has been destroyed through agricultural activity (and consequently the barrow and the original ground surface), but the shallow nature of the extant ditch might suggest that in this instance as much as 0.4-0.5m has been eroded away since the monument was first erected.


© John Gale   

© John Gale
Fig 2 (above) Barrows 1 and 2 partially excavated  (2003) Fig 3 (right) Ditch section of Barrow 1 – partially excavated.

Further evidence for extent of localised erosion and degradation of the monument may be found in the only surviving funerary deposit discovered within the barrow.  On the eastern side of the barrow less than 2m from the ditch edge lay the base of a small pit, in which was found the base of a food vessel which contained the cremated remains of a juvenile (possibly neonatal).  The surviving pot, which is less than 10cm in height, had been placed on the floor of a pit that had been cut into the natural chalk bedrock.  At least 20cm of the upper parts of the food vessel has been destroyed by plough activity but this figure does not account for the depth that the funerary pit was sunk into the barrow overburden and underlying chalk bedrock. The combination of a truncated ditch, and burial pit with attendant funerary urn would seem to suggest that in excess of 0.2m of original ground surface has eroded away since the monument was first erected and that the true loss of soils between then and now is likely to be twice that figure.

Ring ditch – Barrow 2

Following the removal of the overburden of plough soil by machine stripping the residual plough soil was trowelled back to reveal a composite surface of plough-disturbed natural on the western half of the barrow with remnants of an earlier soil on the eastern half of the excavated area. Approximately two thirds of the barrow ditch was observable as an arc of soil against a largely plough disturbed chalk bedrock.  The soil filled ditch scribes an arc of approximately 22m diameter with the width of the ditch being never less than 1.25m.

As the field was to be brought back into production it was not possible to proceed further with the excavation of the barrow that is scheduled to be completed in the summer of 2004.  However, it is clear that no funerary deposits remain within the western half of the interior surface of the barrow and its close proximity to Barrow 1 suggests that it will have suffered a similar degree of erosion.  The ring ditch does have a causeway across it in its south-western quadrant which may suggest that it is Later Neolithic in origin. The nearest geographical and excavated example of this form of monument was investigated by Pitt-Rivers within the Worbarrow complex near Sixpenny Handley towards the end of the 19th century (Barrett et al 1991; Gale 2003a). 


Back to top

Geophysical Surveys at Kings Close

During the 2002 and 2003 fieldwork campaigns Bournemouth University undergraduate students surveyed an area of over 13.5ha of King’s Close with fluxgate gradiometers and twin-electrode earth resistivity, both employed at a 1x1m reading interval. The results of the gradiometry (Fig., 4) clarify several uncertainties in the interpretation of the aerial photography. Most immediately evident on the plot is the oval ditch of a conjoined pair of barrows that appear more as a figure of eight on the aerial photograph, so revealing an apparent unifying element in the construction of these monuments. Perhaps more importantly, a number of rather amorphous features (the maculae noted earlier), lying mainly in the north-west of the field, which were suggested to possibly be pond barrows, are shown to be unlikely to be so, and do seem to represent the fill of natural hollows, some of which may represent dolines/ sinkholes. All the other ring ditches shown on the plot have been seen on aerial photographs, but a number of additional anomalies have been brought to light by the survey.



Fig 4  Fluxgate gradiometer survey of Kings Close 2002-3             © Bournemouth University

Firstly, there is a length of ditch, showing as a dark (positive) linear anomaly running south along inside the eastern boundary of the survey and parallel to the Wimborne to Cranborne road.  Additional short linear anomalies that may represent ditches are to be seen entering the survey area in the south, running north-east near the smaller single ring ditch, and one entering from the north-west margin of the survey area.  The latter is associated with an area of magnetic noise that corresponds to a surface spread of post-mediaeval building material noted during survey. Earth resistivity survey revealed the possible remains of a building in this area of higher magnetic noise, however, other than this, taken across the area as a whole, the earth resistivity survey provided little additional information. High-resolution (0.25x0.5m) caesium gradiometry was undertaken over the group of three ring ditches to the south west of the oval ditch. This work confirmed the presence of an apparent causeway in the south west quadrant of the central ring ditch and provided evidence of ongoing plough damage, suggesting, as excavation confirmed, that these feature have been severely truncated in this part of the site.

Electromagnetic survey was undertaken over the scheduled double ditch barrow in the south-west of the survey area, the oval ditch, and the other surviving barrow abutting the top north-west edge of the survey area which is the bounded by
Mill Lane. The detailed results of this work will be published elsewhere, but this technique, giving simultaneous inphase and quadrature responses, although sometimes difficult to interpret with confidence, clearly gives more information on the nature and survival of the archaeological remains than can be obtained with either resistivity or magnetometry, or even when both are combined. However, this is notwithstanding the role and effectiveness of gradiometry as the principle technique for initial locating and mapping of these monuments.


Geophysical Survey at Great Higher

In 2003 only fluxgate gradiometry was undertaken in Great Higher, although further techniques will be incorporated in future seasons. The results so far (Fig 5), unlike those from King’s Close, have considerably improved on the information currently available from aerial photography. Compared to the murky outline of a possible enclosure identified on the aerial photographs, the magnetic survey has revealed a complex palimpsest of ditch systems and enclosures that largely defies interpretation and phasing from this initial survey. Along with these linear features is a scatter of what are apparently pits (one confirmed as such by excavation) spreading across the whole central area of the survey.  These pits presumably continue to both the east and west, but are not detected in any numbers to the south and north of the central area that is defined and bounded by the curvilinear ditches. Running north-west to south-east through the survey area is the tell-tale positive and negative beaded response of a steel pipeline that has cut right through the site, presumably without the presence of the remains being noted.  The discovery of building foundations during excavation, invisible to gradiometry, requires that earth resistivity, electromagnetics, and possibly ground penetrating radar, will be employed during the 2004 season in an attempt to map these foundations prior to any further excavation.


Fig 5 Flugate gradiometer survey of Great Higher. © Bournemouth University


Back to top

Trial trenching at Great Higher 2003

As a result of the initial analysis of the data from the aerial photography and geophysical surveys, two trial trenches were instigated during 2003.  The larger of the trial trenches (Trench A) consisted of an ‘L’ shaped trench 50 m long by 2m wide located across six boundary/enclosure ditches at the western extent of the field.  This trench was thus located to investigate the nature, extent and possible date of a maximum number of possibly inter-related boundary ditches, three of which were aligned parallel with each other and closely spaced. The plough soil from this trench was hand excavated with every fifth barrow of soil screened through a 6mm sieve to sample any cultural material contained within it.  As with Kings Close the base of the plough soil overlay for the most part natural (plough damaged) chalk bedrock except in those areas where the upper fill of six ditches were recovered. The resultant ditch sections, with one exception, contained within their secondary fills quantities of domestic pottery largely comprising 1st – 3rd century AD Black burnished wares and likely regional variants of similar coarse wares.  Occasionally this pottery was mixed with other domestic debris such as occasional tesserae and other ceramic materials, as well as oyster shells, all of which appeared to be indicative of settlement debris that had fallen into the ditch rather than having been thrown in, or otherwise deliberately placed within them.  It is likely that the material has washed-in perhaps after the settlement was abandoned and the land was brought back into production at as a yet undefined date.

A smaller trench  (Trench B -8m x 1m) was located within the centre of the complex of ditches, placed over a section of an enclosure ditch which seems to demarcate a ‘D’ shaped enclosure of approximately 0.4ha and also over what appeared to be one of many ‘pit’ like features in its interior. This trench was only partially excavated but an enclosure ditch was exposed as well as a short length of walling material (un-coursed flint rubble) within which were found two bow brooches that are possibly 1st –3rd century AD in date.  The presence of unexpected in situ walling material in this trench necessitates that further geophysical work be undertaken on this site before further excavation takes place. 

Future work

In the light of the preliminary findings from two seasons of work summarised here plans are now in place to complete the excavation of Barrow 1 in Kings Close and to further explore the likely Romano-British settlement complex at Great Higher during the summer of 2004.  


As this project currently resides in part within the research led teaching initiative of the archaeological provision at the School of Conservation Sciences at Bournemouth University, much thanks is owed to over 60 students (undergraduate and post-graduate) who have thus far taken part in the two seasons of fieldwork. In particular, gratitude is owed to the student supervisors Phillip Dunn, Martin Blake and Tracey Minall.  To our colleagues Dr Ellen Hambleton, Dr Kate Welham, Dr Helen Smith, Jeff Chartrand, Damian Evans and Bronwen Russell, thank you for your continued help and support. 

Any field project is indebted to the support of the landowners that its investigations affect, and in this instance we are fortunate to have the support and assistance at High Lea of Sir Richard Glyn and particularly the estates manager Mr John Maidment and his staff.


Back to top


This page was written and compiled by John Gale,
School of Conservation Sciences, Bournemouth University.

Last revised: Date 10th November 2004