Land North of Culmstock Road Hemyock Devon

Land north of Culmstock Road Hemyock

Devon

Archaeological Evaluation

for Waddeton Park Ltd

CA Project: 880160 CA Report: 16643

July 2017

Land north of Culmstock Road Hemyock Devon

Archaeological Evaluation

CA Project: 880160 CA Report: 16643

Document Control Grid Revision Date Author Checked by Status Reasons for Approved revision by A 18 Jonathan Derek Evans Internal – Derek December Orellana review Evans 2016 B 24 July Jonathan Derek Evans Internal Revised to Derek 2017 Orellana review include results of Evans further specialist analysis

This report is confidential to the client. Cotswold Archaeology accepts no responsibility or liability to any third party to whom this report, or any part of it, is made known. Any such party relies upon this report entirely at their own risk. No part of this report may be reproduced by any means without permission.

CONTENTS

SUMMARY ...... 2

1. INTRODUCTION ...... 3

2. ARCHAEOLOGICAL BACKGROUND ...... 4

3. AIMS AND OBJECTIVES ...... 5

4. METHODOLOGY ...... 5

5. RESULTS ...... 6

6. THE FINDS ...... 8

7. THE PALAEOENVIRONMENTAL EVIDENCE ...... 8

8. RADIOCARBON DATING ...... 10

9. ARCHAEOMETALLURGICAL RESIDUES ...... 10

10. DISCUSSION ...... 11

11. CA PROJECT TEAM ...... 12

12. REFERENCES ...... 13

APPENDIX A: CONTEXT DESCRIPTIONS ...... 14 APPENDIX B: THE FINDS ...... 15 APPENDIX C: THE PALAEOENVIRONMENTAL EVIDENCE ...... 16 APPENDIX D: OASIS REPORT FORM ...... 17 APPENDIX E: ARCHAEOMETALLURGICAL RESIDUES REPORT ...... 18

LIST OF ILLUSTRATIONS Fig. 1 Site location plan (1:25,000) Fig. 2 Trench location plan showing archaeological features and geophysical survey results (1:1,250) Fig. 3 Trench 4: plan, section and photograph (1:200 & 1:20) Fig. 4 Trench 8: plan, section and photograph (1:200 & 1:20) Fig. 5 Trench 9: plan, section and photograph (1:200 & 1:20) Fig. 6 Trench 10: plan, section and photograph (1:200 & 1:20)

1 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

SUMMARY

Project Name: Land north of Culmstock Road Location: Hemyock, Devon NGR: ST 13482 13567 Type: Evaluation Date: 7–9 November 2016 Location of Archive: To be deposited with the Royal Albert Memorial Museum, Exeter (RAMM), and the Archaeology Data Service (ADS) RAMM reference: RAMM 16/48 Site Code: CULM 16

In November 2016, Cotswold Archaeology carried out an archaeological evaluation on land north of Culmstock Road, Hemyock, Devon. A total of twelve trenches was excavated within the site.

The evaluation recorded a probable ironworking platform. Deposits associated with this feature were radiocarbon dated to the early medieval period. Retrieved metallurgical residues were consistent with an origin within an early medieval non-slag tapping iron smelting furnace.

The evaluation also recorded two ditches and a gravel surface, all of which were undated.

2 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

1. INTRODUCTION

1.1 In November 2016, Cotswold Archaeology (CA) carried out an archaeological evaluation for Waddeton Park Ltd on land north of Culmstock Road, Hemyock, Devon (centred on NGR: ST 13482 13567; Fig. 1). The evaluation was undertaken to inform a planning application for residential development of the site, which is to be made to Mid Devon District Council (MDDC).

1.2 The scope of the evaluation was defined in discussions with Stephen Reed, Senior Historic Environment Officer, Devon County Council Historic Environment Team (DCCHET; the archaeological advisors to MDDC). It was carried out in accordance with a subsequent detailed Written Scheme of Investigation (WSI) produced by CA (2016) and approved by Stephen Reed. The evaluation fieldwork was also in line with Standard and guidance for archaeological field evaluation (CIfA 2014), Management of Research Projects in the Historic Environment (MoRPHE) PPN 3: Archaeological Excavation (Historic England 2015) and Management of Research Projects in the Historic Environment (MoRPHE): Project Manager’s Guide (Historic England 2015).

1.3 The evaluation fieldwork was monitored by Stephen Reed, including a site visit on 9 November 2016.

The site 1.4 The proposed development site covers approximately 6.6ha and comprises four agricultural fields outside the north-western edge of Hemyock village. Culmstock Road (the B3391) runs along part of the southern site boundary. Houses front the road outside the south-eastern and south-western corners of the site and a cemetery fronts the road outside the south-central site boundary. A small stream runs along the eastern site boundary, with residential estates beyond. Further fields lie to the north and west of the site.

1.5 The site occupies a north-facing slope and falls from c.140m AOD at its southern boundary to less than 130m AOD at its northern boundary.

1.6 The underlying bedrock geology of the area is mapped as Mercia Mudstone Group of the Triassic Period. This is overlain by colluvium deposits in the northern and

3 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

western parts of the site. No superficial deposits are recorded in the remainder of the site (BGS 2016).

2. ARCHAEOLOGICAL BACKGROUND

2.1 The site has been the subject of a desk-based heritage assessment (CA 2015) and a geophysical survey (PCG 2016). The following section is summarised from these sources.

2.2 The settlement at Hemyock is recorded in the Domesday survey (1086) and is thought to have originated in the early medieval period.

2.3 Hemyock Castle (Scheduled Monument 1004583) lies c. 80m south-east of the site.

2.4 Archaeological works immediately outside of the site’s south-eastern corner identified building remains potentially associated with a small medieval farmstead, as well as a linear bank which may represent a dump of excavated material from the castle moat or a stockpile of raw material for the pottery or iron industries. Quantities of ironworking slag were also recovered.

2.5 Previous archaeological work on the southern side of Culmstock Road (c. 40m south-west of the proposed development site) recorded ironworking debris in the vicinity of medieval ditches.

2.6 The site itself likely formed part of the agricultural hinterland to Hemyock from the medieval period onwards.

2.7 The geophysical survey of the site (PCG 2016) identified a number of zones of magnetic variation that were considered to potentially relate to iron smelting (red- coloured anomalies marked 1–4 on Figure 2 of this report), as well as a limited number of potential pits. A linear anomaly in the south-eastern part of the site was interpreted as potentially indicative of a former track.

4 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

3. AIMS AND OBJECTIVES

3.1 As defined in the WSI (CA 2016), the objectives of the archaeological evaluation were to provide further information on the likely archaeological resource at the proposed development site. This information will enable MDDC to identify and assess the significance of any heritage assets at the site, consider the impact of the proposed development upon that significance and, if appropriate, develop strategies to avoid or minimise conflict between heritage conservation and the proposed development, in line with the National Planning Policy Framework (DCLG 2012).

4. METHODOLOGY

4.1 The fieldwork comprised the excavation of 12 trenches in the locations shown on Figure 2. Four trenches were 10m long; four trenches were 20m long; four trenches were 30m long. All trenches were 1.8m wide. The trenches were located to sample geophysical anomalies, as well as to test a representative sample of geophysically “blank” areas.

4.2 There were some variations to the trench plan specified in the WSI (CA 2016): T5 was moved to avoid a concrete well (which accounted for the geophysical anomaly in this area of the site – anomaly labelled 4 on Fig. 2); T3, T6 and T7 were moved slightly to avoid overhead power lines. Trenches were set out on OS National Grid (NGR) co-ordinates using Leica GPS and surveyed in accordance with CA Technical Manual 4: Survey Manual.

4.3 All trenches were excavated by a mechanical excavator equipped with a toothless grading bucket. All machine excavation was undertaken under constant archaeological supervision to the top of the natural substrate. Where archaeological deposits were encountered, they were excavated by hand in accordance with CA Technical Manual 1: Fieldwork Recording Manual.

4.4 Deposits were assessed for their palaeoenvironmental potential and samples were taken and processed in accordance with CA Technical Manual 2: The Taking and Processing of Environmental and Other Samples from Archaeological Sites. All artefacts recovered during the evaluation were processed in accordance with CA Technical Manual 3: Treatment of Finds Immediately after Excavation.

5 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

4.5 The project archive and artefacts from the evaluation are currently held by CA. The artefacts (subject to the agreement of the legal landowner) and the project archive will be deposited with the Royal Albert Memorial Museum, Exeter (RAMM) (RAMM reference: RAMM 16/48). A digital archive will be prepared and deposited with the Archaeology Data Service (ADS).

4.6 A summary of information from this project, as set out in Appendix D, will be entered onto the OASIS online database of archaeological projects in Britain.

5. RESULTS

5.1 This section provides an overview of the evaluation results. Figure 2 shows the trenches and the recorded archaeological features overlain on the geophysical survey results. Detailed summaries of the recorded contexts can be found in Appendix A. Details of the artefactual material recovered during the evaluation can be found in Section 6 and Appendix B. Details of the biological/palaeoenvironmental evidence recovered during the evaluation can be found in Section 7 and Appendix C.

5.2 The natural substrate comprised greyish yellow and red clays with outcrops of small and medium-sized stones. It was encountered at a depth of 0.35m–0.5m below the present ground level. The natural substrate was typically sealed directly by the modern topsoil, although three trenches (T9, T10 and T11) featured colluvial layers between the natural substrate and the topsoil.

5.3 Archaeological features were identified in T4, T5, T8, T9 and T10. All of these features were cut into/overlay the natural substrate and were sealed by the topsoil or colluvial layers (where present). These trenches are discussed in more detail below. T5 exposed a modern ditch; this feature contained modern artefactual material and corresponded to a modern path noted by the geophysical survey (PCG 2016). All other trenches were archaeologically blank.

Trench 4 (Fig. 3) 5.4 North-east/south-west orientated ditch 402 (Fig. 3, Sec. AA) was 1.85m wide and 0.34m deep. It had an undated single silty fill, 403, from which a single piece of ironworking slag was recovered. Ditch 402 corresponded to a linear depression

6 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

visible in the surface of the field, and was in the location of a linear anomaly interpreted as a potential former trackway in the geophysical survey report (PCG 2016).

Trench 8 (Fig. 4) 5.5 North/south aligned ditch 803 (Fig. 4, Sec. BB) had been truncated by modern stone field drain 807, but survived to 0.81m in width and 0.2m in depth. It contained a sequence of three undated fills (804, 805 and 806), the uppermost of which (806) yielded one piece of ironworking slag. A palaeoenvironmental sample (Sample 1) was recovered from charcoal-rich lower fill 804 (see Section 7). Ditch 803 was in the location of a pit-like geophysical anomaly.

Trench 9 (Fig. 5) 5.6 Substantial cut feature 904 was identified in the north-western half of T9. This feature consisted of a flat-based, 0.78m-deep cut (Fig. 5, Sec. CC). The base of the cut was covered by 0.04m-thick compact stone surface 905. This surface was partially overlain by 0.05m-thick light grey silty clay layer 907, which contained charcoal, ashes and ironworking slag. An environmental sample (Sample 3) was taken from this material (see Section 7).

5.7 Layer 907 was partially covered by 0.33m-thick dark grey silty clay deposit 906, which contained a substantial amount of ironworking slag. An environmental sample (Sample 2) was taken from this material (see Section 7). Layers 906 and 907 were almost 100% hand-excavated, but no artefactual material other than slag was recovered.

5.8 Layer 906 was sealed by two colluvial layers (902 and 901), which had a combined thickness of up to 0.7m.

5.9 Feature 904 corresponded in location to a zone of strong magnetic variation identified by the geophysical survey.

Trench 10 (Fig. 6) 5.10 Surface 1003 (Fig 5, Sec. DD) was identified in the south-eastern end of the trench. It comprised a compacted gravel layer measuring up to 0.08m in thickness. No associated dating evidence was recovered. This surface was in the broad location of a pit-like geophysical anomaly.

7 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

6. THE FINDS

6.1 The artefactual material recovered during the evaluation is listed in Appendix B and is discussed further below.

Metallurgical residues 6.2 Ironworking slag amounting to a total of 6,047g in weight was hand-recovered from four deposits; an additional 6,904g of ironworking slag was recovered from palaeoenvironmental samples taken from deposits 906 and 907 (associated with potential ironworking platform 904; T9). This material is discussed in detail in Section 9.

Other finds 6.3 Three small and featureless fragments of fired clay (24g) were recovered from layer 906 (T9). All are of a similar ‘fabric,’ soft with common quartz sand inclusions. None feature the vitrified/slaggy surfaces which might suggest relation to metallurgical processes (e.g. hearth/furnace linings) and the date and function of this material is unclear.

6.4 Two flakes of flint and greensand chert (4g) were recovered from deposit 806 (fill of ditch 803; T8) and are suggestive of some earlier prehistoric activity in the area. Both exhibit edge damage/breakage, which suggests that they are redeposited in a later context.

7. THE PALAEOENVIRONMENTAL EVIDENCE

7.1 A total of three palaeoenvironmental samples (67 litres of soil) was retrieved and processed with the aim of evaluating the preservation of palaeoenvironmental remains and recovering palaeoenvironmental evidence of domestic or industrial activity on the site. It was also hoped that the environmental assemblages might provide dating evidence. The samples were processed by standard flotation procedures (CA Technical Manual 2: The Taking and Processing of Environmental and Other Samples from Archaeological Sites).

7.2 Preliminary identifications of plant macrofossils are noted in Table C1 (Appendix C), following standard nomenclature (Stace 1997).

8 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

7.3 The flots were large, with varying amounts of rooty material and uncharred seeds. The charred material displayed varying levels of preservation. Some of the charcoal showed signs of vitrification and iron coating.

Trench 8 7.4 A large quantity of charcoal fragments greater than 2mm was noted within fill 804 (ditch 803; Sample 1). This assemblage included mature wood fragments and a tuber fragment. There is no indication of the date of this dumped material.

Trench 9 7.5 Deposits 906 (Sample 2) and 907 (Sample 3), both of which were associated with potential ironworking platform 904, contained large amounts of charcoal fragments, including both mature and round wood pieces – these are discussed in more detail below. No charred plant remains were recovered from this feature. These samples also produced slag fragments indicative of ironworking (see Section 9).

7.6 Several uncharred remains were recovered from context 907 (Sample 3), including seeds of elder (Sambucus nigra), brambles (Rubus sp.), buttercup (Ranunculus sp.) and hemp-nettle (Galeopsis), as well as alder (Alnus sp.) cones. This appears to be indicative of a mainly scrub/wasteland damp environment in the vicinity.

Charcoal analysis 7.7 Charcoal identifications are noted in Table C2 (Appendix C).

7.8 The charcoal from deposits 906 (Sample 2) and 907 (Sample 3) was abundant but mostly poorly preserved due to silt impregnation within the pores of the charcoal, which in some cases prevented full identification. Some of the charcoal within deposit 907 was only partially charred. The charcoal identified was dominated by alder/hazel (Alnus glutinosa/Corylus avellana) roundwood, with that from deposit 907 also containing a high proportion of oak (Quercus) fragments. Other species identified included birch (Betula), ash (Fraxinus excelsior) and willow/poplar (Salix/Populus).

7.9 Typically, charcoal assemblages associated with metalworking activity are dominated by oak, as it is considered to be one of the few species which burns at a high enough temperature for smelting to take place. These assemblages are unusual as they contain a high proportion of alder/hazel charcoal, and much of the

9 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

charcoal within deposit 907 is only partially charred. Alder/hazel wood was known to have been converted into charcoal (which is more calorific than wood) and used within metalworking activities; however, the dominance of alder/hazel together with the partially charred nature of the material within deposit 907 suggests that the material within these deposits is an accumulation of not only waste associated with iron smelting, but also discarded firing debris from other domestic/industrial activities.

8. RADIOCARBON DATING

8.1 Radiocarbon dating was undertaken in order to confirm the date of waste deposits 906 and 907 within ironworking platform 904. The samples were analysed during May 2017 at Scottish Universities Environmental Research Centre (SUERC), Rankine Avenue, Scottish Enterprise Technology Park, East Kilbride, Glasgow, G75 0QF.

8.2 The uncalibrated dates are conventional radiocarbon ages. The radiocarbon ages were calibrated using the University of Oxford Radiocarbon Accelerator Unit calibration programme OxCal v4.3.2 (2017) (Bronk Ramsey 2009) using the IntCal13 curve (Reimer et al. 2013).

Table 1: Radiocarbon dating results

Feature Lab No. Material δ 13C Radiocarbon Calibrated radiocarbon Calibrated radiocarbon age age 95.4% probability age 68.2% probability Context 906 SUERC Charcoal -26.8‰ 1170 ± 29yr BP 771–903 cal AD (91.8%) 777–793 cal AD (11.4%) Ironworking -73284 Corylus avellana 919–963 cal AD (13.6%) 801–893 cal AD (56.8%) platform 904 (hazel) Context 907 SUERC Charcoal -26.2‰ 1277 ± 29 yr BP 663–775 cal AD (95.4%) 684–721 cal AD (39.8%) Ironworking -73283 Alnus glutinosa 741–767 cal AD (28.4%) platform 904 (Alder)

9. ARCHAEOMETALLURGICAL RESIDUES

9.1 A full report on the assessment of the archaeometallurgical residues from the site (GeoArch 2017) is included as Appendix E of this report. The following section presents a brief summary of this information.

9.2 Ironworking slag amounting to a total of 6,047g in weight was hand-recovered from four deposits (Appendix B), all of which were undated. An additional 6,904g of

10 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

ironworking slag was recovered from palaeoenvironmental samples taken from deposits 906 and 907 (associated with potential ironworking platform 904; T9).

9.3 The residues were all compatible with an origin within a non-slag tapping iron smelting furnace.

9.4 The slags were mainly flow slags. Some pieces showed amalgamation both against the wall of the pit (where there were indications of reaction with the material of the pit wall) and around large moulds of the wood pit-packing. Others were in form of small individual flow slag prills. Several pieces showed the formation of a slag ‘puddle’ in the base of the pit. Some pieces were large fragments, rich in finer charcoal, that typically form what is known as the ‘furnace bottom’ slag in such furnaces.

9.5 Non-tapping furnaces were employed both in the Iron Age and early medieval periods. The present material shows closer affinity to early medieval assemblages.

10. DISCUSSION

10.1 The evaluation identified a small number of archaeological features at the site. These were mainly concentrated towards the central area of the western site boundary, although there was also a ditch in the south-eastern part of the site.

10.2 There was a variable correspondence to the geophysical survey results (PCG 2016). T4, T5, T8, T9 and T10 all contained archaeological features corresponding either exactly or broadly to geophysical anomalies. There were, however, no archaeological features corresponding to anomalies tested by T1, T3 and T11.

10.3 Two flakes of worked flint/chert (4g) were recovered as residual/redeposited material from ditch 803 (T8). These are suggestive of low-level background earlier prehistoric activity in the area.

10.4 Feature 904 (T9) comprised a substantial, flat-bottomed cut with a stone lining at its base, partially infilled with two deposits of burnt material and ironworking slag. It is possible that this represents an ironworking platform. Associated deposits 906 and 907 were radiocarbon dated to the early medieval period. The retrieved metallurgical

11 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

residues were consistent with an origin within an early medieval non-slag tapping iron smelting furnace.

10.5 Previous archaeological works (see Section 2) have recovered quantities of ironworking slag immediately outside of the evaluation site’s south-eastern corner and on the southern side of Culmstock Road (c. 40m south-west of the present evaluation site). Although this slag was not dated definitively, it was in association with medieval features. The ironworking platform within the present evaluation site must therefore either pre-date the previously-recorded activity, or the inferred date of that activity is incorrect.

10.6 Typically, charcoal assemblages associated with metalworking activity are dominated by oak. The assemblages from ironworking platform 904 were dominated by alder/hazel; this, together with the partially charred nature of the material, suggests that the charcoal within these deposits is an accumulation of not only waste associated with iron smelting, but also discarded firing debris from other domestic/industrial activities. The charcoal assemblage appears to be indicative of a damp scrub/wasteland environment.

10.7 Ditches 402 (T4) and 803 (T8) were undated artefactually, although both contained single fragments of ironworking waste. The function of these features is uncertain; they may have been minor drainage or boundary features. Alternatively, ditch 402 was in the location of a linear anomaly interpreted as potentially indicative of a former track (PCG 2016), so may represent a minor channel worn by passage.

10.8 Gravel surface 1003 was of unknown date and function.

11. CA PROJECT TEAM

11.1 Fieldwork was undertaken by Jonathan Orellana, assisted by Edoardo Vigo and Parris Stubbings. This report was written by Jonathan Orellana. The finds report was written by Katie Marsden. The palaeoenvironmental evidence report was written by Sarah Wyles and Sarah Cobain. The archaeometallurgical residues report was written Dr Tim Young. The evaluation report illustrations were prepared by Sam O’Leary. The project archive has been compiled and prepared for deposition by Jessica Cook. The project was managed for CA by Derek Evans.

12 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

12. REFERENCES

BGS (British Geological Survey) 2016 Geology of Britain Viewer http://www.bgs.ac.uk/discoveringGeology/geologyOfBritain/viewer.html Accessed 1 November 2016

Bronk Ramsay, C. 2009 ‘Bayesian analysis of radiocarbon dates’, Radiocarbon 51, 337–360

CA (Cotswold Archaeology) 2015 Land North West of Hemyock, Devon: Heritage Desk- Based Assessment CA typescript report 15098

CA (Cotswold Archaeology) 2016 Land north of Culmstock Road, Hemyock, Devon: Written Scheme of Investigation for an Archaeological Evaluation

DCLG (Department of Communities and Local Government) 2012 National Planning Policy Framework

GeoArch 2017 Assessment of archaeometallurgical residues from N of Culmstock Road, Hemyock (880160) GeoArch Report 2017/03

Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Grootes, P.M., Guilderson, T.P., Haflidason, H., Hajdas, I., HattŽ, C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., & van der Plicht, J. 2013 ‘IntCal13 and Marine13 Radiocarbon Age Calibration Curves 0–50,000 Years cal BP’, Radiocarbon 55, 1869–1887

Stace, C. 1997 New Flora of the British Isles Cambridge: Cambridge University Press

PCG (Pre-Construct Geophysics) 2016 Archaeological Geophysical Survey: Land to the North of Culmstock Road, Hemyock, Devon

13 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

APPENDIX A: CONTEXT DESCRIPTIONS

Trench Context Type Fill of Context Description L W D No. No. interpretation (m) (m) (m) 1 100 Layer topsoil mid brownish grey sandy silt 0.4 1 101 Layer natural substrate small and medium sized stones within a light grey clay 2 200 Layer topsoil mid brownish grey sandy silt 0.4 2 201 Layer natural substrate small and medium sized stones within a light grey clay 3 300 Layer topsoil mid brownish grey sandy silt 0.5 3 301 Layer natural substrate small and medium sized stones within a light grey clay 4 400 Layer topsoil mid brownish grey sandy silt 0.35 4 401 Layer natural substrate small and medium sized stones within a red clay 4 402 Cut ditch NE/SW orientated, with moderate 1.85 0.34 sloping sides and flat base 4 403 Fill 402 single fill of ditch light pinkish grey silty clay 1.85 0.34 5 500 Layer topsoil mid brownish grey sandy silt 0.4 5 501 Layer natural substrate compact light brown silty clay 6 600 Layer topsoil mid greyish brown sandy silt 0.4 6 601 Layer natural substrate compact light brown silty clay with frequent small stones 7 700 Layer topsoil mid greyish brown sandy silt 0.4 7 701 Layer natural substrate small and medium sized stones within a light grey clay 8 800 Layer topsoil dark greyish brown sandy silt 0.45 8 801 VOID 8 802 Layer natural substrate light yellowish brown sandy clay 8 803 Cut ditch N/S aligned, with steep sides and 1.5 0.42 flat base 8 804 Fill 803 1st fill of ditch dark blackish grey silty clay with 1.08 0.13 frequent charcoal flecks 8 805 Fill 803 2nd fill of ditch mid grey silty clay 1.3 0.25 8 806 Fill 803 3rd fill of ditch mid brown silty sand 1.5 0.21 8 807 Cut field drain NE/SW aligned, V-shaped profile, 1.1 >0.4 base not reached 8 808 Fill fill of field drain mid brown sandy clay 1.1 >0.4

9 900 Layer topsoil dark greyish brown sandy silt 0.35 9 901 Layer colluvium light pinkish brown silty clay 0.4 9 902 Layer colluvium mid bluish grey silty clay 0.25 9 903 Layer natural substrate light yellowish grey clay with small stones 9 904 Cut platform moderate sloping SE side and flat >6.1 >2 0.8 base 9 905 Layer stone surface small stones sub-angular stones and >2.9 >2 0.05 occasional gravel 9 906 Layer deposit dark blackish grey silty clay with 4.4 >2 0.25 abundant fragments of iron slag and charcoal flecks 9 907 Layer deposit mid grey silty clay with iron slag 1.35 >2 0.06 10 1000 Layer topsoil dark brown sandy silt 0.3 10 1001 Layer colluvium firm mid yellowish clay 0.18 10 1002 Layer natural substrate mixed of small and medium sized stones and yellowish grey clay 10 1003 Layer surface firm and compact mid light grey clay 0.08 with small stones and gravel 10 1004 Layer colluvium firm mid grey clay 0.15 11 1100 Layer topsoil mid brownish grey sandy silt 0.25

14 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

Trench Context Type Fill of Context Description L W D No. No. interpretation (m) (m) (m) 11 1101 Layer colluvium light grey clay with occasional 0.25 fragments of iron slag 11 1102 Layer natural substrate small and medium sized stones and yellowish grey clay 12 1200 Layer topsoil mid brownish grey sandy silt 0.4 12 1201 Layer natural substrate small and medium stones within a orangey grey clay

APPENDIX B: THE FINDS

Table B1: finds concordance Context Class Description Ct. Wt.(g) 403 Metallurgical residues tap slag 1 71 806 Metallurgical residues tap slag 1 343 Flint/chert flakes 2 4

906 Metallurgical residues Indeterminate (dense) ironworking slag 16 5300 Fired clay 3 24 1101 Metallurgical residues Indeterminate (dense) ironworking slag 5 333

Table B2: finds from samples Context Sampl. no. Class Description Ct. Wt.(g) 804 1 Burnt stone 1 7 907 3 Metallurgical residues Indeterminate ironworking slag 400 805 906 2 Metallurgical residues Indeterminate ironworking slag 602 6099

15 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

APPENDIX C: THE PALAEOENVIRONMENTAL EVIDENCE

Table C1: Assessment table of the palaeoenvironmental remains Vol Flot size Roots Charred Charcoal > Feature Context Sample (L) (ml) % Grain Chaff Other Notes for Table 4/2mm Other Trench 8 Ditch 803 804 1 21 300 10 - - * tuber frag *****/***** - Trench 9 Ironworking platform 904 906 2 32 150 10 - - - - ****/** - (uncharred (****), slag Sambucus, Alnus, 904 907 3 14 575 n/a - - - ****/***** material Rubus, Galeopsis, (*) Ranunculus) Key: * = 1–4 items; ** = 5–19 items; *** = 20–49 items; **** = 50–99 items; ***** = >100 items

Table C2: Charcoal identifications Context number 906 907 Feature number 904 904 Sample number (SS) 2 3 Flot volume (ml) 150 575 Sample volume processed (l) 32 14 Soil remaining (l) 0 0 Period EMED EMED Charcoal quantity >2mm **** ***** Charcoal preservation Poor Poor Family Species Common Name Betulaceae Alnus glutinosa (L.) Gaertn. Alder r/w 17 4 Alnus glutinosa (L.) Gaertn./ Alder/Hazel r/w 58 47 Corylus avellana L. Betula L. Birches 1 Corylus avellana L. r/w Hazel r/w 10 3 Quercus petraea (Matt.) Liebl./ Sessile Oak/ Fagaceae 12 43 Quercus robur L. Pedunculate Oak Oleaceae Fraxinus excelsior L. Ash 1 3 Salicaceae Salix L./Populus L. Willows/Poplars 1 Total 100 100 Key: * = 1–4 items; ** = 5–19 items; *** = 20–49 items; **** = 50–99 items; ***** = >100 items

16 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

APPENDIX D: OASIS REPORT FORM

PROJECT DETAILS Project name Land north of Culmstock Road, Hemyock, Devon Short description In November 2016, Cotswold Archaeology carried out an archaeological evaluation on land north of Culmstock Road, Hemyock, Devon. A total of twelve trenches was excavated within the site.

The evaluation also recorded two ditches and a gravel surface, all of which were undated. Project dates 7-9 November 2016 Project type Evaluation Previous work Desk-Based Assessment (CA 2015) Geophysical Survey (PCG 2016) Future work Unknown PROJECT LOCATION Site Location Culmstock Road, Hemyock, Devon Study area (m2/ha) c. 6.6ha Site co-ordinates ST 1342 1356 PROJECT CREATORS Name of organisation Cotswold Archaeology Project Brief originator N/A Project Design (WSI) originator Cotswold Archaeology Project Manager Derek Evans Project Supervisor Jonathan Orellana MONUMENT TYPE None SIGNIFICANT FINDS None PROJECT ARCHIVES Intended final location of archive Content

Physical Royal Albert Memorial Museum Iron slag, flint RAMM 16/48 Paper Royal Albert Memorial Museum Context sheets, trench RAMM 16/48 sheets, section drawings Digital Archaeology Data Service (ADS) Born-digital data; scans of primary site archive BIBLIOGRAPHY CA (Cotswold Archaeology) 2016 Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation CA typescript report 16643

17 1122 1133 1144 1155

1166

1155

1144

1133

1122

N Cirencester 01285 771022 Milton Keynes 01908 564660 Cotswold Andover 01264 347630 Archaeology w www.cotswoldarchaeology.co.uk e [email protected]

PROJECT TITLE DEVON Land north of Culmstock Road, Hemyock Devon FIGURE TITLE Site location plan

0 1km

FIGURE NO. Reproduced from the Ordnance Survey digital mapping with DRAWN BY SO PROJECT NO. 880160 the permission of Ordnance Survey on behalf of The Controller CHECKED BY DJB DATE 17/11/2016 of Her Majesty's Stationery Office c Crown copyright Cotswold Archaeology Ltd 100002109 APPROVED BY DE SCALE@A4 1:25,000 1 134 136 N

site boundary evaluation trench archaeological feature layer/deposit modern field drain 136 B B section location

Geophysical Survey Key (Pre-Construct Geophysics 2016)

0 50m

Reproduced from the Ordnance Survey digital mapping with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationery Office &URZQFRS\ULJKW&RWVZROG$UFKDHRORJ\/WG

Andover 01264 347630 Cirencester 01285 771022 Exeter 01392 826185 Milton Keynes 01908 564660 w www.cotswoldarchaeology.co.uk Black Down e [email protected]

Cottage Cemetery 7 PROJECT TITLE Land north of Culmstock Road, Hemyock Bean Close Devon

Butts FIGURE TITLE 134 Close Trench location plan showing Sherwood Lulworth archaeological features and geophysical survey results

CULMSTOCK ROAD DRAWN BY SO PROJECT NO. 880160 FIGURE NO. CHECKED BY DJB DATE 17/11/2016 ST APPROVED BY DE SCALE@A3 1:1,250 2 SIMCOE PLACE Dene 136.0m AOD N 135.7m AOD T4 A A

134.0m AOD archaeological feature ditch 402 134.4m AOD limit of excavation

0 1:200 10m

Section AA NW SE 134.8m AOD

400 topsoil

401 natural 403

ditch 402

011:20 m

Cirencester 01285 771022 Milton Keynes 01908 564660 Cotswold Andover 01264 347630 Archaeology w www.cotswoldarchaeology.co.uk e [email protected]

PROJECT TITLE Land north of Culmstock Road, Hemyock Devon FIGURE TITLE Trench 4: plan, section and photograph

Ditch 402, looking north-east (scale 1m) DRAWN BY SO PROJECT NO. 880160 FIGURE NO. CHECKED BY DJB DATE 17/11/2016 APPROVED BY DE SCALE@A4 1:200 & 1:20 3 ditch 803 B B T8 136.2m AOD N 136.0m AOD 136.1m AOD 136.4m AOD field drain 807

archaeological feature field drain limit of excavation 0 1:200 10m

Section BB

WE 136.1m AOD 800 topsoil 808 806 806 805 805 ditch 803 ditch 803 804 804 802 natural field drain 807 field drain 807

stone field drain

011:20 m

Cirencester 01285 771022 Milton Keynes 01908 564660 Cotswold Andover 01264 347630 Archaeology w www.cotswoldarchaeology.co.uk e [email protected]

PROJECT TITLE Land north of Culmstock Road, Hemyock Devon FIGURE TITLE Trench 8: plan, section and photograph Ditch 803 truncated by stone field drain 807, looking north (scale 1m)

DRAWN BY SO PROJECT NO. 880160 FIGURE NO. CHECKED BY DJB DATE 17/11/2016 APPROVED BY DE SCALE@A4 1:200 & 1:20 4 Section CC

SE NW 133.4m AOD

900 topsoil

901 colluvium

902 colluvium natural 903

iron-working platform 906 iron-working waste 904

907 iron working waste 905 stone surface

011:20 m

133.2m AOD T9

132.8m AOD iron-working platform 904 N C

133.5m AOD

C 133.8m AOD

archaeological feature layer/deposit field drain Cirencester 01285 771022 limit of excavation Milton Keynes 01908 564660 Cotswold Andover 01264 347630 Archaeology w www.cotswoldarchaeology.co.uk e [email protected] 0 1:200 10m PROJECT TITLE Land north of Culmstock Road, Hemyock Iron-working platform 904, looking south-west (scale 1m) Devon FIGURE TITLE Trench 9: plan, section and photograph

DRAWN BY SO PROJECT NO. 880160 FIGURE NO. CHECKED BY DJB DATE 17/11/2016 APPROVED BY DE SCALE@A4 1:200 & 1:20 5 131.9m AOD

131.5m AOD

T10

N D

131.6m AOD surface 1003 132.0m AOD

layer/deposit limit of excavation 0 1:200 10m

Section DD

NW SE 131.3m AOD 1000 topsoil

1001 colluvium

1004 colluvium surface 1003 natural 1002

011:20 m

Cirencester 01285 771022 Milton Keynes 01908 564660 Cotswold Andover 01264 347630 Archaeology w www.cotswoldarchaeology.co.uk e [email protected]

PROJECT TITLE Land north of Culmstock Road, Hemyock Devon FIGURE TITLE Trench 10: plan, section and photograph

Surface 1003, looking north-east (scale 1m) DRAWN BY SO PROJECT NO. 880160 FIGURE NO. CHECKED BY DJB DATE 17/11/2016 APPROVED BY DE SCALE@A4 1:200 & 1:20 6 © Cotswold Archaeology Land north of Culmstock Road, Hemyock, Devon: Archaeological Evaluation

APPENDIX E: ARCHAEOMETALLURGICAL RESIDUES REPORT

18 GeoArch Report 2017/03

Assessment of archaeometallurgical residues from N of Culmstock Road, Hemyock (880160)

Dr Tim Young 15th February 2017 Assessment of archaeometallurgical residues from N of Culmstock Road, Hemyock (880160)

Dr T.P. Young Abstract

This small collection (approximately 14.8kg of macroscopic residues and 6.9kg of sieved residues from flotation) from a field evaluation included abundant slag from the basal pit on a non-slag tapping iron smelting furnace. The residues were all compatible with an origin within this single technology.

In detail, the slags were mainly flow slags Some pieces showed amalgamation both against the wall of the pit (where there were indications of reaction with the material of the pit wall) and around large moulds of the wood pit-packing. Others were in form of small individual flow slag prills. Several pieces showed the formation of a slag ‘puddle’ in the base of the pit. Some pieces were large fragments, rich in finer charcoal, that typically form what is known as the ‘furnace bottom’ slag in such furnaces.

Magnetic separation of subsamples of the supplied sieved residues showed much fine-grained slag debris, fired clay, burnt stone and possible ore. No hammerscale was observed.

Non-tapping furnaces were employed both in the Iron Age and early medieval periods. The present material shows closer affinity to early medieval assemblages. It can be compared with the two nearby excavated early medieval smelting sites, Churchills Farm and to the south of Culmstock Road. The latter contained furnaces with rather deep pits, with wood pit-packing and straight blowing walls, the former had (alongside slag-tapping furnaces) non-tapping furnaces with medium sized sub-circular pits, with basal puddles and cereal packing. The present material, whilst in the same broad technology, is thus different from both.

The assemblage (and therefore the site) is highly significant, both for its potential to inform of the development of iron smelting in early medieval Hemyock, but also for its potential to provide further evidence for development of iron smelting in Wessex.

Contents

Abstract ...... 1 Interpretation ...... 2

Methods ...... 2 Summary ...... 3

Results References ...... 4 Description of Residues ...... 2 Ironworking residues ...... 2 Table 1: ...... 5 Distribution of residues ...... 2

GeoArch Report 2017/03: residues from Hemyock

Methods Sample <3> (from context (907)) included a small proportion of haematite-rich material, some containing chert, that was probably roasted iron ore. All materials were examined visually with a low- The macroscopic material also included a small powered binocular microscope where required. As an proportion of fired ceramic, including reduced-fired evaluation, the materials were not subjected to any material from context (906) and a piece of bloated high-magnification optical inspection, not to any form of cermaic from context (1101). It is possible that these instrumental analysis. The identifications of materials pieces may include heat altered material from the pit in this report are therefore necessarily limited and must margin, rather than necessarily being from the be regarded as provisional. constructed furnace superstructure. No large pieces of oxidised fabric from the upper structure were The examined materials are listed in Table 1. observed.

This project was commissioned by Katie Marsden, of Cotswold Archaelogy. Distribution of residues Small quantities of residues were recovered from undated ditches 402 (T4) and 803 (T8), as well as from the colluvium in T11. The majority of the recovered materials, however, was derived from the deposits Results overlying the stone surface in T9.

Description of residues The submitted materials amounted to an overall total of approximately 14.8kg of macroscopic residues and 6.9kg of sieved residues from flotation. Interpretation The residues are indicative of iron smelting in a non- The identifiable materials are all from iron smelting in a tapping bloomery furnaces. Non-tapping furnaces were non-tapping furnace. typically employed in the Iron Age and early medieval periods, whereas slag tapping (in which the slag is The catalogue of the residues is presented in Table 1. encouraged to largely flow out of the furnace) was practiced in the Roman period, in some areas in the early medieval and more widely in Britain from the late Ironworking residues 9th century onwards until bloomeries were replaced by the production of iron in blast furnaces in the early Much of the assemblage comprises varieties of flow post-medieval period. slag. These range up from small-scale individual prills to coarse flow slag amalgamations surrounding wood Certain discrimination of the residues from non-tapping moulds and to the charcoal-rich amalgamated material furnaces of these different periods is not currently typical of so-called ‘furnace bottoms’ (FBs). FBs tend possible, but the present assemblage more closely to be very dense near the blowing wall and some of resembles early medieval examples than collections of the largest blocks (e.g. the 3.5kg block from context Iron Age date. (906)) show the interaction with the pit wall as is common in the part of the furnace. This whole range of In detail, the presence of well-fluxed slag puddles variation is exhibited by the material from context provides a close comparison with the non-tapping (906). furnaces of Churchills Farm, Hemyock (Young 2015b; Smart et al . in prep.). These furnaces had relatively In addition to these flow slags, pieces from both shallow circular basal pits (0.4 to 0.5m diameter and contexts (806) and (906) show evidence for the mostly less than 300mm deep, one example being formation of ‘slag puddles’ in the base of the furnace 450mm deep), which were packed with cereal steams pits. These puddles show elongate fayalite crystals prior to smelting. These furnaces show close similarity extending across the entire thickness of the puddle (up with examples in the Forest of Dean at Clearwell to almost 25mm), suggesting a well-fluxed composition Quarry (Pine et al. 2009) and Yorkley (Young 2015c). at least as siliceous as fayalite. These occurrences are all probably mid-8th -mid-10 th century in age. It is possible that the cereal packing Other slag materials include vesicular slag with a may indicate an Irish influence on the technology lobate/prilly basal surface (context (1101)) and some (Young 2011, 2012). low-density and charcoal-rich slags from context (906); both are probably variants of the flow slag textures The present material does not show evidence for mentioned above, although neither has criteria cereal packing, but rather the more usual wood enabling definite attribution to an origin in smelting. packing (as was also probably present at Yorkley).

The fine-grained debris in the sieved sample <2> from Wood packing was normal on a second group of sites, context (906) included small prills and angular with larger basal pits to the furnaces, including those fragments of slag, alongside particles of fired ceramic on the S side of Culmstock Road (Young 2014; and concretion. The slag fragments in samples <1> Rainbird & Young in press) and at Burlescombe (from context (804)) and <3> (from context (907)) were (Devon) (Reed et al. 2006), as well possibly as those rather more rounded and therefore presumably at Ramsbury (Wiltshire) (Haslam 1980) and Millbook reworked. The identification of the more worn very fine- (Sussex) (Tebbutt 1982). These furnaces may include grained material in these two samples is less certain, a few examples at around 0.5m diameter, but most are with the distinction between concretionary material, larger, ranging up to 1m diameter and have a broadly slag and fired clay difficult to make on the basis of splayed prolife and a planar blowing wall. These visual inspection. examples also have given 8 th -10 th century dates, but are possibly just slightly earlier than the cereal packed examples.

2 GeoArch Report 2017/03: residues from Hemyock

At broadly the same period there are also examples of slag-tapping furnaces. These occur late in the sequence at Ramsbury, where the furnace has a splayed profile and a planar blowing wall, and contemporaneously with the cereal-packed non- tapping furnaces at Churchills Farm, where they are formed at one end of a small but elongate pit, as seen on 11 th -13 th century examples.

The history of iron-smelting technology through the 8 th to 11 th centuries in Wessex is clearly very complex. The present material does not precisely resemble the material from any known site, but fits within the range of variation of 8 th -10 th century examples.

The location of the site adds to understanding of the topography of early ironmaking in Hemyock. The productive area of T9 lies immediately NE of an area in the same field flagged by Smart (in prep) as a potential clay extraction pit (medieval clay extraction occurred alongside early medieval iron production at Churchills Farm). The 7th -9th century site S of Culmstock Road (just west of the SW corner of the present development area) was also of limited geographical extent. The 9th - 10 th century site of Churchills Farm lay to the SE of the castle. Waste disposal into medieval clay pits also suggested that a later (11 th -12 th century?) iron smelting site also lay close to Churchills Farm.

The samples produced no evidence for hammerscale. The site therefore parallels the two known early medieval sites in Hemyock in showing a remarkable lack of evidence for smithing. The raw iron produced must have been worked into usable iron elsewhere. A model of dispersed iron smelting (perhaps dependent on woodland resources) with centralised bloomsmithing seems likely. No centralised smithing facility has yet, however, been discovered in Hemyock,

Such a model is perhaps particularly significant given the likelihood that Hemyock was a royal estate in the late Saxon period. The abundant evidence for iron production, employing different and at times innovative smelting technology may hint at a specialised role for this estate. Although the evidence for iron smelting across the Blackdown Hills, drawing on the rich iron resources of the clay-wth-flints (Reed 1997; Young 2015a) is well established, this focus of early medieval activity at Hemyock appears unusual. Elsewhere in Wessex, there is evidence for iron production on other royal estates, for instance at Ramsbury (Haslam 1980) and Pucklechurch (Young & Young 2013).

Summary

This material provides evidence for an additional focus of early medieval iron smelting on the margins of Hemyock. Detailed further investigation of this site would be highly desirable, as would scientific dating of the deposits.

The collection of material is not tied to structural evidence for iron smelting and is not currently independently dated. Indeed, if the suggestion of clay extraction from immediately to the SW of T8 and T9 is correct, there would be possibility that the slag occurrences have been reworked during overburden removal. Further scientific investigation of the residues is therefore not advised at this time.

Nonetheless, the materials are of great importance and should be deposited with the site archive in the appropriate receiving institution.

3 GeoArch Report 2017/03: residues from Hemyock

References

HASLAM, J. 1980. A middle Saxon iron smelting site at Ramsbury, Wiltshire. Medieval Archaeology , 24 , 1-68.

PINE, J., ALLEN, J.R.L. & CHALLINOR, D. 2009. Saxon iron smelting at Clearwell Quarry, St Briavels, Lydney, Gloucestershire. Archaeology in the Severn Estuary , 20 , 9-40.

RAINBIRD, P. & YOUNG, T. in press. Saxon iron smelting furnaces at Culmstock Road, Hemyock, Devon. Proceedings of the Devon Archaeological Society , forthcoming.

REED, S.J. 1997. Blackdown Hills Ironworking Project, Archaeological recording of an iron extraction pit, Broadhembury, Devon. Exeter Archaeology, Unpublished Report 97.38.

REED, S.J., JULEFF, G. & BAYER, O.J. 2006. Three late Saxon iron-smelting furnaces at Burlescombe, Devon. Proceedings of the Devon Archaeological Society , 64 , 71-122.

SMART, C. YOUNG, T., ALLAN, J. DAWSON, D., TAYLOR, R.T., ANDERSEN, J., ROLLINSON, G., CHALLINOR, D., JONES, J., MARSHALL, P., TOMPKINS, M., COLLINGS, T., KEEN, L., LANGHAM, G., KENT, O, BRONK RAMSEY, C., DUNBAR., E. & REIMER, P. (in prep.). Industry and the making of a rural landscape: iron and pottery production at Churchill’s Farm, Hemyock, Devon .

TEBBUTT, C.F.1982, A Middle-Saxon ironsmelting site at Millbrook, Ashdown Forest, Sussex, Sussex Archaeological Collection s, 120 , 19–36.

YOUNG, A. & YOUNG, D.E.Y, 2013. The Bristol and Bath Science Park (Spark), Emersons Green, South Gloucestershire. Archaeological Excavation (Excavation Area 1) and Recording Project 2010. Unpublished Report, Avon Archaeological Unit.

YOUNG, T.P., 2011. Archaeometallurgical residues from the N7 Castletown to Nenagh scheme, Camlin 3 (E3580), Co. Tipperary. GeoArch Report 2011/23. 62 pp.

YOUNG, T.P. 2012. Appendix 3 -. Exploiting the bog -: iron production and metalworking, pp. A3.1 – A3.60 In: P. Stevens & J. Channing, Settlement and Community in the Fir Tulach Kingdom. National Roads Authority and Westmeath County Council.

YOUNG, T.P. 2014. Archaeometallurgical residues from Culmstock Road, Hemyock, Devon (ACD591). GeoArch Report 2014-11. 43pp.

YOUNG, T.P. 2015a. Analyses of ores and residues from Dunkeswell, Devon. GeoArch Report 2014-31.

YOUNG, T.P. 2015b. Archaeometallurgical residues from Churchills Farm, Hemyock, Devon. GeoArch Report 2015-31. 77pp.

YOUNG, T.P. 2015c. Assessment of archaeometallurgical residues from Yorkley, Gloucestershire. GeoArch Report 2015-24. 4pp.

Table 1: summary catalogue.

c s c wt I wt I no notes

403 70 70 1 broken slightly lobate dense slag, probably flow lobed

804 1 >1 496 assm dominated by coarse sand grains, slag is little rounded

804 1 >0.5 52 assm stone-rich assemblage, even in magnetic fraction there is no hammerscale

806 344 344 1 fragment of basal puddle with prilly base, lobed at one end where flowed under strange wood fragment with scaly surface

906 3055 16 fragments of dense FB and related material; two pieces show basal puddle, or perhaps burr, with lots of fine included charcoal higher in the sections; largest block 1315g has possible basal flow with interaction with ceramic and probable clay clasts; one of the smaller basal puddle fragments shows crystals extending across whole thickness 20-25mm (close to eutectic?).

906 4205 718 6 small fragments of FB-like material - including one probably conjoining a piece in the sample above; includes unusual basal flow wood-like surface texture 3470 1 weathered and rounded large FB block,

906 2 bag 1 3655 1098 48 dense slags ranging from large coarse flow slag block to small individual prills 2565 190 low-density and charcoal-rich slags

906 2 bag 2 3130 1044 72 dense slags 2033 215 low-density vesicular and charcoal-rich slag 58 14 reduced fired clay and slagged similar material

906 2 >0.5 80 assm fine slag ceramic and charcoal debris, no observed hammerscale in magnetic fraction 906 2 >1 898 898 assm dominated by angular slag fragments, some ceramic, mostly magnetic 906 2 >2 1795 assm dominated by angular slag fragments, some ceramic, mostly magnetic 906 2 >5 2090 assm mixed chert gravel, slag frags, prills, ashy concretions, dark charcoal bearing concretions, rust

907 3 >0.5 23 assm some fine slag, charcoal, quartz/chert, orange ceramic etc, much lower proportion of slag than <2> 907 3 >1mm 236 assm stone-dominated - with wide variety of lithologies, lots of charcoal, some slag, some red fragments that might be roasted ore 907 3 >2 380 assm stone rich; plenty of charcoal, red particles possibly ore - soft haematite attached to hard chert? 907 3 >5mm 814 814 assm assemblage of mainly small dense slag fragments, very little low density material

c s c wt I wt I no notes

1101 332 22 1 bloated ceramic with lightly slagged surface 286 3 vesicular slag with some charcoal, resting on dense basal layer with slightly reddened lobate base 16 1 dense thin slag sheet, with wood mould contact, passing into highly vesicular material

GeoArch

geoarchaeological, archaeometallurgical & geophysical investigations

Unit 6, Western Industrial Estate, Caerphilly, CF83 1BQ

Office: 029 20881431 Mobile: 07802 413704

E-Mail: [email protected] Web: www.GeoArch.co.uk