Visit date: 5th May 2012

Weather: Cold and windy, bright but cloudy


The field trip was conducted by Land Rover and was a discontinuous selection of sites, rather than a progression around a landscape. For this reason, the sites visited will be listed rather than the perambulation described.

  • Snail Down Barrow Cemetery
  • Fittleton Long Barrow
  • Lidbury
  • Chisenbury Warren
  • East Chisenbury Midden


Chisenbury Warren

Grid Reference: SU 1785 5380

Site Overview

Chisenbury Warren presents as a series of earthworks 500m long (Fulford et al. 2006 : 54) covering approximately six hectares (Short 2006 :78), with a trackway leading into, and across the length of the earthworks (McOmish et al. 2002 : 100).  It is on the side of a gently-sloping, south-east-facing hill at the eastern end of Rainbow Bottom on Salisbury Plain (McOmish et al. 2002 : 98).  Despite the presence of a medieval rabbit warren in the woods behind the settlement, there is no evidence of extensive damage to the site, and no record of previous excavation, just the occasional surface find (Fulford et al. 2006 : 54).


Figure 1: Chisenbury Warren from the south-east April 2011. Copyright K Bragg.

Figure 2: Trackway leading into Chisenbury Warren from the south west April 2011. Copyright K Bragg.

Investigation History

Year Investigation type Investigation Details
1980 Accidental Discovery Fulford et al. reported that in the cutting of an infantry trench in the 1980s, an adult female was found, and an individual burial. It was not clear if this represented the location of a cemetery or just a single burial event (Fulford et al. 2006 : 54).
1993 Excavation Reading University excavated and a geophysical survey was done to confirm the earthworks, this also revealed pits and ditches. The object of the excavation was to establish whether what was visible as surface remains represented an entire village (i.e. all the builds were contemporary with each other) or whether what the earthworks represented was in fact a drift in settlement over time. Another objective was to determine if all the platforms represented dwellings or if a mix of uses was represented. As this is one of the best-preserved examples of its kind, it was also important to establish the state of the below-ground remains to inform the conservation of other sites (Fulford et al. 2006 : 54).


Chronology and Current Interpretation

Excavation showed that the earliest settlement on the site potentially dated from the Late Iron Age (Fulford et al. 2006 : 73) with evidence for continuation of the settlement into the late Roman period (McOmish et al. 2002 : 98).

The finds from the site are consistent with activities of subsistence, industrial and domestic natures and of a rapid expansion in the early Roman period from a smaller site (Fulford et al. 2006 : 73-74)

Figure 3: Chisenbury Warren settlement and fields, Wiltshire. Romano-British village. (Source Wilson 2011 Fig 1)

Chisenbury Warren is a ‘ladder-style’ nucleated settlement – so-called because its main axis is a single lane or street and the buildings are arranged around it (Wilson 2011 : 2).


East Chisenbury ‘Midden’

Grid Reference:  SU14605323

Site Overview

The site occupies a false-crested position on a spur overlooking the River Avon (McOmish et al. 2010 :37), just north-west of the village of East Chisenbury. From the site a good view in most directions is possible.


Figure 4: View of and from East Chisenbury Midden, looking east-ish May 2012. Copyright K Bragg.

The site itself consists of a mound of deposited material covering approximately five hectares and up to two metres deep (Wilts SMR SU 15 SW 154).


Investigation History

Year Investigation type Investigation Details
Late 1700s Visit Colt Hoare visited the site and was made aware of the fertility of the land, and made finds of ‘rude British pottery’ but did not, it seems, recognize the mound as being other than the natural slope of the hill (McOmish et al. 2010 : 37-38).
1945 Excavation Walls and Bray and then Bray alone excavated at a location probably on the northern edge of the mound, and found much pottery and bone (McOmish et al. 2010 : 38-39).
1992-1993 Rediscovery of site, then excavation As the site was under threat from the development of a routeway, and prior to the information from the 1945 excavation being unearthed, an augur survey and test pit strategy was proposed to establish the nature of the site. Some of the augur probes did not reach the bottom of the deposit, implying the ground level was not level beneath the mound (McOmish et al. 2010 :43).In both test pits a complex sequence of deposits was found, with similar content in each test pit. The excavators felt they could tell separate deposition events apart only when the materials were of different composition, but some events seem to have been ‘capped off’ with a layer of compacted chalk to form a surface (McOmish et al. 2010 :50).Beneath the mound, the excavators noticed a layer thought to be a buried land surface and possibly plough soil, with what looked to be evidence for settlement beneath this.



Figure 5: RCHME earthwork survey plan. (Source McOmish et al. 2010 Fig. 3)

Chronology and Current Interpretation

The mound at East Chisenbury contains a complicated sequence of deposits of varying compositions, including ashy material, waste products and LBA/EIA pottery. It is surely difficult to generalize from such a small sample from such a large site, but McOmish et al. consider the excavations to show that the mound consists of material collected and potentially stored elsewhere (little weathering is observed) and then transported to the site and deposited, a suggestion reinforced by the mixture of both late and early pottery forms in one deposit. The animal waste and bedding does not appear to have been produced in situ as the edges of the deposits are not consistent with trampling having occurred (McOmish et al. 2010 : 86-87).

The ‘settlement’ beneath the midden also contains pottery of the All Cannings Type but appears to have been short-lived and the area returned to agricultural production before the formation of the midden (McOmish et al. 2010 : 88-89).

The compacted chalk layers do not seem to be ‘pavements’ as suggested for the similar features discovered by Maud Cunnington at All Cannings, and particularly the upper layer of compacted chalk is interesting as it seems less localized and may have covered the entire mound. McOmish et al. therefore suggest that it is consistent with a ‘capping’ event that would have left the deposit white and very visible (McOmish et al. 2010 : 87). This idea seems similar to how modern humans dispose of their rubbish, it is collected up in a single place and then covered over with concrete. In the case of East Chisenbury, however, the midden deposits continued above this layer (McOmish et al. 2010 : 88), so it was not a process termination indicator. Tubb (2011b : 40) suggests that the chalk layers may be viewed as a deposit in themselves. It is pointed out that a chemical reaction between the midden deposits and the chalk would produce a fungicidal chemical (McOmish et al. 2010 : 88). It would be very interesting to see if there were evidence for exploitation of this in the form of extraction holes, or perhaps gaps in the chalk layer where the reacting mass was removed entire.

Tubb (2011) broadly agreed with the deposition mechanisms as posited by McOmish et al., namely that material was potentially transported to the site having been curated elsewhere; material was generated onsite as a result of specialised activity; or that the site was in fact both settlement and midden: a form of tell (McOmish et al. 2010 : 84-86). Where Tubb takes exception is with the classification of the midden material itself as an unwanted product; he would rather it be seen as part of a complicated process of social reproduction in a time of change and transition. He suggests that the primary purpose of All Cannings Ware is to do with feasting and display, and that the deposition of examples of this material, plus the byproducts and waste products of that feasting process are of importance to society. Instead of the municipal landfill site, this is instead a record of a society’s reaction to a change in how relationships are formed and renegotiated, and a visual statement of that process in a prominent landscape position.


Fulford, M.G. et al., 2006. Iron Age and Romano-British Settlements and Landscapes of Salisbury Plain, Salisbury: Wessex Archaeology Report 20.

McOmish, D. et al., 2002. The Field Archaeology of the Salisbury Plain Training Area First Edit., English Heritage.

McOmish, D., Field, D. & Brown, G., 2010. The Bronze Age and Early Iron Age Midden Site at East Chisenbury, Wiltshire. The Wiltshire Archaeological and Natural History Magazine, 103, pp.35-101.

Short, B., 2006. England’s Landscape: The South East, Collins/English Heritage.

Tubb, P.C., 2011a. Late Bronze Age / Early Iron Age transition sites in the Vale of Pewsey : the East Chisenbury midden in its regional context. Wiltshire Archaeological and Natural History Society Magazine, 104, pp.44-61.

Tubb, P.C., 2011b. The LBA/EIA Transition in the Vale of Pewsey, Wiltshire, Oxford: British Archaeological Reports (British Series 543).

Wilson, P., 2011. Introduction to Heritage Assets: Roman Settlements, Swindon: English Heritage. Available at:



Visit date: 4th March 2012

Weather: Initially driving rain, then blizzard


This field trip took us on a walk past Withy Copse and out onto the area within the enclosure. We did some random surface collection to demonstrate how much material was visible on the surface of a ‘black earth’ site. It then began to snow – really heavily –  and we retreated to Devizes Museum to look at the fabulous exhibits there (and thaw out).

Figure 1: Blizzard on Martinsell 4th March 2012. Copyright K Bragg.

Withy Copse

Grid Reference: SU17216429

Site Overview

The site itself is a low mound within a wooded area, on a slope to the north of Martinsell enclosure and was described by Maud Cunnington (1909 :125) as being no more than 0.75m  height above ground level, and covering just over 19m long and a maximum width of just over 13m. The mound is at a distance of about 91m from the rampart of the enclosure (Cunnington 1909a : 18) and is in a south-west to north-east orientation (1909 :125).

Investigation History

Year Investigation type Investigation Details
1907-8 Excavation Maud and Ben Cunnington excavated the mound, which they described as consisting of “a fine black mould” (1909 :125). Within this were found animal bone (mostly sheep, pig and ox); large quantities of pot sherds (more numerous towards the surface) and a fibulae.  (1909 :127). Also discovered was a filled-in ditch that Cunnington speculated might have been associated with a structure on the site but did not excavate due to the extension this would cause to the current excavation and the vegetation cover (1909 :125).
1975 Finds Evaluation The Savernake Ware collected by the Cunningtons was reassessed by Swan and the interpretation of the site as a kiln was put forward again on the basis that whilst there were no vastly distorted vessels that Mrs Cunnington might have considered diagnostic of ‘wasters’, there were indications of firing irregularities, and of items that may have been kiln furniture (Swan 1975 :38). Swan also proposed a later, and post-conquest date for the pottery, but this has subsequently been proven uncertain by Timby demonstrating the origins of Savernake Ware pre-dated the conquest (Timby 2001: 73-84 cited in Tubb 2011: 101).


Chronology and Current Interpretation

Cunnington mentions that it had been suggested that the mound was the remains of a kiln site, as so much pottery was discovered, but she dismissed this on the grounds that no malformed ‘waster’ pots were discernable (Cunnington 1909a :18) and her previous excavations had included kiln sites at Milton Lilbourne (Tubb 2011 :100) so she may be assumed to have felt able to recognize such when seen. Her interpretation, based on the types of pottery found, and the date of the fibulae was that this was a Late Pre-Roman Iron Age midden (Tubb 2011 :100). She stopped short of associating it with the hilltop enclosure, as she pointed out that the date of this was unknown (Cunnington 1909a :18). Cunnington was especially careful to point out that the absence of Gaulish Samian ware (when other sites locally show a presence), coupled with the imported Arratine Ware giving a range of second century BC to early first century AD indicate that the site was earlier than the conquest and did not continue beyond (Cunnington 1909a :19-21). Having been vindicated in the dating by Timby’s work Cunnington’s assessment of the site as a rubbish heap rests on the validity of Swan’s assessment of whether or not there was evidence for a kiln, and this isn’t clear without further investigation.


Martinsell Hilltop Enclosure

Grid Reference:  SU 1766 6395

Site Overview

The hill on which the enclosure is sited is one of the highest points in southern England and the third highest point in Wiltshire (the other two occurring further west in the same range) (Tubb 2011 :99). The enclosure itself covers approximately 13 hectares (NMR SU 16 SE 6) and follows the contours of the hill (Tubb 2011 :118). The enclosure is formed of a single ditch and bank, with the ditch on the outside and where the topography provides less natural defence on the western side, the rampart is more pronounced (Corney & Payne 2006).

Investigation History

Year Investigation type Investigation Details
1821 (pub’d) Survey and excavation Colt-Hoare excavated but without finding any evidence for habitation (Goddard 1913 :307).
1970s Surface Collection Late Bronze Age – Early Iron Age pottery found on surface of interior after ploughing (Tubb 2011 :116)
1996 Geophysical Survey The Wessex Hillforts Project surveyed the interior of the enclosure with a magnetometer, which showed a few anomalies, but the results were such that no evidence for occupation could really be proposed (Corney & Payne 2006 :120).


Chronology and Current Interpretation

Cunliffe (2005 :422) gives Martinsell as an example of an early hilltop enclosure associated with an extensive system of linears, which he interprets as to do with cattle management (2005 :424). Tubb ( 2011 :118) rejects Cunliffe’s comparison of Martinsell with other Iron Age sites such as Balksbury and Walbury and suggests that a more reasonable comparison would be Liddington Castle, visible from Martinsell itself.

Despite the reported finds of LBA/EIA pottery from within the enclosure, Tubb considers them to be residual and unconnected with the construction of the ramparts, an activity he attributes to the Early Iron Age. He also notes an absence of evidence for Middle Iron Age activity on the hilltop, a gap that continues until the end of the Iron Age, stating that any activity may have been less archaeologically perceptible, rather than absent (Tubb 2011 :122).

The results of the Wessex Hillforts Project’s geophysical assessment lead Payne et al. to suggest that Martinsell was unlike many of the hillforts in the region and had perhaps been used as a temporary camp or for seasonal gatherings (Corney & Payne 2006 :120). The question as to what activity may have occurred inside hillforts has been the subject of much discussion, with Hill roundly rejecting Cunliffe’s view of Iron Age society, which he claims is based on Irish Medieval Society and the opinions of classical writers, and which requires hillforts to be central places supported by a strict hierarchical society (Hill 1996). Hill points out that much of what Cunliffe claimed made Danebury a central place, could be found in other Iron Age settlements (Hill 1996 : 96-99). Perhaps because hillforts are highly visible to archaeologists, their importance has been overstated and a greater knowledge of unenclosed settlement patterns may redress the balance.

It may be that the landscape position and the nature of the hill-form itself determined that an enclosure would serve a purpose; Tubb suggests that the visibility of Martinsell from quite a wide area around would make this a prominent place. He also points out that the LBA/EIA activity, in the form of Black-earth sites in the locality would have already marked out the place as a location of importance in the local “landscape mythology”. Therefore meaning that the transition in activity between that which required/created the Black-earth sites and the creation of the hill-top enclosure, may be seen as a continuity (Tubb 2011 :122).


Corney, M. & Payne, A., 2006. The Monuments and Their Setting. In A. Payne, M. Corney, & B. Cunliffe, eds. The Wessex Hillforts Project: extensive survey of hillfort interiors in central southern England. London: English Heritage, pp. 39-130.

Cunliffe, B., 2005. Iron Age Communities in Britain: an account of England, Scotland and Wales 4th Editio., Abingdon: Routledge.

Cunnington, M.E., 1909a. 11. Notes on a Late Celtic Rubbish Heap Near Oare, Wiltshire. Man, 9, pp.18-21.

Cunnington, M.E., 1909b. Notes on a Late Celtic Rubbish Heap Near Oare, Wiltshire. Wiltshire Archaeological and Natural History Magazine, XXXVI(CXI), pp.125-139.

Goddard, E.H., 1913. List of Prehistoric, Roman, and Pagan Saxon Antiquities. The Wiltshire Archaeological and Natural History Magazine, 38, pp.153-378.

Hill, J.D., 1996. Hill-forts and the Iron Age of Wessex. In T. C. Champion & J. R. Collis, eds. The Iron Age in Britain and Ireland: Recent Trends. J. R. Collis Publications, pp. 95-116.

Swan, V.G., 1975. Oare Reconsidered and the Origins of Savernake Ware in Wiltshire. Britannia, 6, pp.36-61.

Timby, J., 2001. A reappraisal of Savernake Ware. In P. Ellis, ed. Roman Wiltshire and After: papers in honour of Ken Annable. Devizes: Wiltshire Archaeological And Natural History Society.

Tubb, P.C., 2011. The LBA/EIA Transition in the Vale of Pewsey, Wiltshire, Oxford: British Archaeological Reports (British Series 543).


archaeo.log is proudly powered by WordPress and the Theme Adventure by Eric Schwarz
Entries (RSS) and Comments (RSS).


Notes from a field