Settling the Ebbsfleet Valley High Speed 1 Excavations at Springhead and Northfleet, Kent The Late Iron Age, Roman, Saxon, and Medieval Landscape Volume 2: Late Iron Age to Roman Finds Reports by Edward Biddulph, Rachael Seager Smith, and Jörn Schuster with contributions from Leigh Allen, Phil Andrews, Elina Brook, Kayt Marter Brown, Nicholas Cooke, B M Dickinson, Damian Goodburn, Martin Henig, David Holman, Lorraine Mepham, J M Mills, Quita Mould, Cynthia Poole, Ian Scott, Ruth Shaffrey, and Jacqui Watson Principal illustrator Rob Goller Other illustrations by Elina Brook, Tom Goskar, Nick Griffiths, S E James, Sophie Lamb, Sarah Lucas, Lucy Martin, Rob Read, Rachael Seager Smith, Georgina Slater, and Elaine A Wakefield Oxford Wessex Archaeology 2011 292 Settling the Ebbsfleet Valley Blank Page Chapter 1 Introduction 189 Chapter 3 Springhead Metalwork by Jörn Schuster with contributions by Elina Brook, Jacqui Watson, and Martin Henig This report describes all illustrated late Iron Age and Roman small finds from the excavations at Springhead (Saxon objects can be found in Vol 4, Chap 3), but an attempt has been made – as far as possible – also to mention those finds only described in the finds database. Where necessary, the latter are referred to by their original object/small find number (SF …), which allows them to be found in the database available online at http:// owarch.co.uk/hs1/springhead-northfleet/. Apart from basic descriptions and all measurements for both catalogued and non-catalogued finds, this database also contains digital photos of most non-ferrous metal objects. Grave finds are only referred to by their small find number, and these are also used on the grave plans. Most metal objects were x-radiographed and conservation was carried out on selected objects by Wiltshire Conservation Service, Salisbury (now Chippenham). If no material type is mentioned at the end of a catalogue description and no XRF-analysis was carried out, the material is assumed to be copper alloy. In cases where an XRF-analysis of an alleged copper alloy object returned an inconclusive result, the object’s material is mentioned as copper alloy. The alloy names and the compositions they represent follow those established by Bayley and Butcher (2004, 4 table 5–6). Table 46 gives an overview of the 2633 recorded metal small finds from the various Springhead sites, identified by the HS1 event codes. The figure of 2633 includes finds from all periods, eg, the 65 recorded metal small finds from the early/mid-Saxon cemetery on the eastern side of ARC SPH00, a small number of medieval and later objects (Catalogued in Vol 4, Chap 3) as well as one copper alloy stud from a late Bronze Age ditch. Additionally, c 3000 nails or nail fragments were recorded as bulk finds from the entire site. This large number of metal small finds is a result of the consistent use of metal-detectors during the excavation. Site 51742 was not part of the HS1 excavation contract but as it was located at Springhead nursery immediately adjacent to the ARC SHN02 site, it was conveniently integrated in this report; objects from site 51742 have negative small find numbers, eg SF -523, which is also how they can be found in the online database. All metal small finds in this report and in the database have been grouped according to functional categories following Crummy (1983, 5–6). The following list gives an overview of the range of small find numbers used for the various Springhead sites: 51742 ARC 342E02 ARC ERC01 ARC SHN02 (Settlement) ARC SPH00 (Sanctuary) ARC WCY02 SFs -565 – -500; SFs 50900–98; SFs 52001–20; SFs 15001–999, 18000–999, 20000–552; SFs 2–1900, 1959, 6001–4, 9000–467 SFs 53001–5 A breakdown by functional category of all recorded metal small finds of the late Iron Age and Roman phases is shown in Table 47. Roman objects found residually in Saxon or later layers have not been included in this table (4 in Saxon, 2 in medieval, and 12 in modern layers), neither have objects of unknown date from such layers; however, the Roman objects have been considered in the typological discussions, and they are also included in the various distribution maps (Figs 82, 85, 86, 88, 90, 93, 96, 101–2, 104). Table 46 Springhead metal finds: material type per site of recorded metal small finds 51724 ARC 342E02 ARC ERC01 ARC SHN02 ARC SPH00 ARC WCY02 Total Copper alloy 15 14 5 597 773 1 1405 Iron 12 7 3 426 265 1 714 Lead alloy 6 16 1 337 126 1 487 Silver 1 4 15 20 Gold 1 3 4 Other metal 3 Material type Total 34 37 9 1368 3 1182 3 2633 Settling the Ebbsfleet Valley 190 Table 47 Springhead metal finds: recorded small finds of late Iron Age and Roman phases by functional category Function Group Total Personal Toilet/Medical Textile Household Metrology Script Transport Construction Tools Fittings Agricultural Militaria Votive Metalworking Unknown Total 491 56 18 64 41 18 25 3 70 582 9 33 11 268 723 2412 Personal Adornment or Dress Brooches The database contains 317 late Iron Age and Roman brooches or brooch-related entries like springs or pin fragments. This catalogue lists the 135 illustrated brooches but those only mentioned in the database have been referred to in the typological sequence. The catalogue loosely follows, where feasible, the sequence adapted for the Richborough assemblage by Bayley and Butcher (2004). The type numbers used there are an adaptation of Hull’s typological sequence (ibid, 52 and appendix 2), which is also used in this study where possible, the type definition given as ‘T00’. However, as it was not possible to check illustrations for all of Hull’s types, other typologies are referred to as necessary, the most frequently used are those by Riha (1979; 1994), Feugère (1985), and Böhme (1972). One-piece sprung brooches catchplate. Pin missing. One spring coil remaining. SF 9360, Context 6682, Intervention 6682 (Artefact). Early Roman. 4. Complete. Flat bow decorated with wavy rocker line down length. Triangular catchplate. Four rectangularsectioned spring coils. SF 1817, Context 6448, Intervention 6448 (Layer), SG 300271 (Post-holes). Early Roman. 5. Incomplete. Dotted decoration on flat bow runs in grooves down sides and in wavy line down centre. Rectangular to oval-sectioned pin. Four spring coils. SF 1527, Context 3391, Intervention 3391 (-). 6. Incomplete. Two fragments. Rectangular-sectioned flat bow, sharp angle at head, tapers to foot. Punched dotted wavy line decoration down bow, possibly grooves down edges. Triangular catchplate. Spring broken, pin missing. Two spring coils. SF 1815, Context 6447, Intervention 1001 (Spring), SG 300012 (Watercourse). Early Roman. 7. Complete. Flat bow, wavy dotted decoration down length, two grooves along edges. Small triangular catchplate. Bow similar to SF 1841. Four spring coils with internal chord. (Leaded) bronze. SF 1848, Context 6445, Intervention 1000 (Spring). See also SFs 1504 (similar small catch plate at West Thurrock, cf Schuster 2009, fig 8, 19507), 1841, 18322 and possibly 673, 15713 and 15890 (parallels for SF 15713 eg, at Baldock (Stead 1986, 110 fig 41, 29) and Colchester (Crummy 1983, 7 fig 2, 6)). Rod or wire bows 8. Complete. Sharp angle at head; straight, circularsectioned bow tapers to foot. Two groups of transverse lines which continue around the back of the bow: upper set at midpoint of bow has five lines, lower set just above foot has four lines. Triangular catchplate. Four rectangular-sectioned spring coils. (Leaded) brass. SF 15723, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. 9. Incomplete, T 12. Slightly arched bow without reverse curve, continues into foot without break. Outer edge of catch-plate with notch above pin rest. Iron. SF 15964, Context 16776, Intervention 16776 (-). Watling Street. Early Roman. See also SF 666. Nauheim-derivative brooches T11 Flat bow (Fig 81) 1. Incomplete, flat, narrow rectangular-sectioned bow, sharp angle at head, tapers to foot. Spring and pin missing. Large triangular catchplate. SF 1567, Context 3232, Intervention 3231 (Ditch), SG 300030 (Ditched enclosure). Late Iron Age. 2. Incomplete. Flat bow, central groove decoration. Foot and subtriangular catchplate bent upwards. Beginning of spring survives. SF 9175, Context 6447, Intervention 1001 (Spring), SG 300012 (Watercourse). Early Roman. 3. Incomplete. Flat bow, short longitudinal groove down centre, filled with punched transverse lines. Triangular Variants 10. Incomplete. Narrow rectangular-sectioned bow, parallel sided. Sharp bend between bow and foot which continued at right angles. Lower part of foot, catchplate and pin missing. Four spring coils with internal chord. Copper alloy. SF 15725, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. See also SF 668. This is possibly the fragment of a Nauheimor Nauheim-derivative brooch; it is distinguished from the brooches discussed above by the flared head of the brooch which almost looks like the buds of little wings similar to the fragment of an iron Nauheim brooch from Vienne, France (Feugère 1985, pl 53, 758). Chapter 3 Springhead Metalwork 191 3 1 2 Corrosion by-products 4 6 5 7 8 Drawn from X-ray 11 10 0 9 Figure 81 Springhead: copper alloy and iron brooches brooches 1–11 50mm Settling the Ebbsfleet Valley Nauheim-derivative Brooches 561900 Rhône and a few further upstream (Feugère 1985, 219 fig 15), but also on the northern fringes of the Keltiké, eg, in the oppidum on the Dünsberg, Germany (Schlott 1999, 15, Taf 1, 9.11). The sharp angle between bow and foot of the variant Cat No 10 has parallels at Colchester (Hawkes and Hull 1947, pl 92, 64), Neuss, Germany (Simpson 2000, pl 5, 5), and some Almgren 16brooches eg, at ‘De Hoorden’ near Wijk bij Duurstede, Netherlands (van der Roest 1988, Taf 7, 182.194–5.199) or Maurik, Netherlands (Haalebos 1986, 93 fig 43, 129.134.136), but, unlike at Springhead, in most of those instances the flat bow is more swollen and has a line of punched dots in the middle. Only three brooches have a rod or wire bow and one of these, Cat No 9, is the only iron brooch found in a Roman context at Springhead. Of the two copper alloy brooches, Cat No 8 has a straight decorated bow with two groups of transverse lines while SF 666 has a plain, slightly arched bow with a short groove on the inside. Nauheim-derivatives do already occur in early 1st century AD contexts in Britain, eg, at Skeleton Green (Mackreth 1981, 131) or Verulamium, King Harry Lane Site (KHL) graves 270 and 317 (Stead and Rigby 1989, 89); they become more common after the Conquest. At Colchester those with a flat bow are found in Claudian to Neronian contexts (Hawkes and Hull 1947, 312), and at Baldock a distinction between flat- and the simpler 561700 Hull distinguishes brooches with the reverse curve (T10) of the bow from those with a simple curve (T11). At Springhead only the latter type was found, although it might be that Cat No 2 originally had a reverse curve, but this can no longer be verfied as its foot is distorted. T11 can be sub-divided into variants depending on the treatment of the bow: 12 have a flat bow which can be plain (Cat No 1), decorated with a single longitudinal groove (Cat Nos 2 and 3) or with lateral lines filled with punched dots (SF 1504), lateral lines and a wavy line (Cat Nos 4–7, SF 1841) or punched dots (SF 18322) down the middle of the bow. While the flat bow variants are common on numerous sites in southern Britain (Bayley and Butcher 2004, 147; 192 fig 168), the wavyline decoration – mainly consisting of punched dots but on Cat No 4 it is a rocker line – has few parallels and may be a regional variation: one is found in a period I context (50 BC–AD 40) at Harlow, Essex (Gobel 1985, 72 fig 39,1) and in that report two further examples are mentioned from the Darenth Valley in Kent and Fringinghoe in Essex (ibid, 75); the head of a fourth parallel was recently found at Horton, Berkshire (WA 2006a, 21; image at www.flickr.com/photos/ wessexarchaeology/68984200/). It should be noted that the same ornament already occurs on Nauheim brooches of Feugère’s types 5a45–49 which are mainly found in the south of France near the mouth of the 561800 192 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 A2 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 82 Springhead: distribution of Nauheim-derivative brooches Chapter 3 Springhead Metalwork ‘poor man’s’ variants showed that the former were predominantly pre-Flavian while the latter continued to the end of the 1st century (Stead 1986, 123–4 fig 51), a distinction which cannot be demonstrated in the Springhead assemblage. While most of the Nauheimderivatives at Springhead (Fig 82) are probably postConquest, found in context with pottery dates ranging between AD 50 and 160, a number may be slightly earlier, including SF 666 which was found in the base of Viewing platform 2 and Cat No 1 from the enclosure ditch 300030. One-piece sprung brooch with sharp angle at head of bow (Fig 81) 11. Incomplete. Part of rectangular-sectioned, sharply bent bow and spring only. Two spring coils. Leaded brass. SF 18321, Context 17425, Intervention 17425 (-). The sharp angle of the bend of the head makes this an unusual brooch which may have had a simple spring construction of originally probably only three coils. The latter, if true, would link it to a group of probably postConquest Nauheim-derivatives in south-east England (Mackreth 1995, 968–70, fig 406, 54–61). In fact, Mackreth’s no 54 exhibits a similar sharp angle at the head of the bow which is rectangular, not flat like Cat No 11. The closest parallel so far seems to be a brooch from Augst, Switzerland, where it was found in a context dated to the 1st half of the 1st century AD (Riha 1994, 61–2; Taf 4,1933; Riha type 1.12). It should be noted that Riha assumes this brooch to have four spring coils, but the spring is incomplete. One-piece sprung brooches with chord held by hook Feugère type 9b (Fig 83) 12. Incomplete. Spring of eight spring coils intact, axial rod visible, chord held by small rectangular hook. Pin broken. Neck of bow extended slightly but no wings as such. Flat bow, single ridge decoration down entire length, tapers slightly towards foot. Perforated catchplate, partly broken. (Leaded) brass. SF 1505, Context 3390, Intervention 3390 (Layer). This brooch is of similar construction as the ‘Simple Gallic’ and the Colchester brooch but with a very flat bow typical of Feugère’s type 9b which is clearly distinguished from the former because of the more strongly developed wings in type 14a (Feugère 1985, 242). In Gaul at least, both types also seem to have clearly distinct distributions: type 9b in the Languedoc region (ibid, fig 22), while 14a occurs further north and east, in Switzerland and along the Rhine where it is found, for instance, at Hofheim (Ritterling 1905; type Hofheim Ic) or Asciburgium (Bechert 1973, Taf, 23–6, but also including two Feugère type 9b in 23, 208 and 26, 234). Stead (1986, 111 fig 42, 52) has related a very 193 similar brooch from Baldock to some Hawkes and Hull (1947) type III brooches from Colchester which are more clearly the typical ‘Simple Gallic’ brooch T89, Feugère type 14a or Riha 2.2. Feugère’s type 9b ranges between the end of the 1st century BC and the middle of the 1st century AD (Feugère 1985, 241); the Baldock brooch belongs early in the 1st century AD (Stead 1986, 123–4). Simple Gallic brooches T89 13. Incomplete. Half of spring remains. Bow undecorated, diamond-shaped section, tapers slightly towards round foot end. Chord is broken, held by triangular hook. Triangular catchplate with subrectangular hole. Four spring coils. Brass. SF 323, Context 2221, Intervention 2221 (Artefact). 14. Incomplete. Small wings with two transverse grooves. Narrow rectangular-sectioned bow, tapers slightly towards foot. Part of perforated catchplate remains. Pin broken. Six spring coils. SF 389, Context 2230, Intervention 2227 (Pit), SG 300073 (Pits). Early Roman. See also SFs 1288, 1529 and 9345. Cat No 13 clearly has a reverse curve of the foot; however, the diamond-shaped section of the bow is less often found in these brooches, but a good parallel exists at the Titelberg, Luxemburg (Metzler 1995, 299 Abb 149, 2) and others with slightly less profiled sections are known, for instance from Baldock (Stead 1986, 111 fig 42, 54), Neuss, Germany (Simpson 2000, pl 2, 17), or a Hofheim Ic brooch from Asciburgium, Germany (Bechert 1973, Taf 24, 214). It is a feature similar to that found in Almgren 19a brooches with a steeper neck which are widely distributed in both the German provinces as well as outside the limes (Haalebos 1986, 32 fig 11; Völling 1994, Beilage 7; 1998, 45 Abb 2). SF 1529 shares the very pronounced reverse curve of the foot but is slightly unusual in the treatment of the head above the wings in that this is formed like a rectangular block, a feature also present but less marked in Cat No 14. The very pronounced reverse curves of Cat No 13 and SF 1529 suggests a Continental origin, while SFs 9345 and 1288, with completely straight bows, and Cat No 14, with a gentle convex arch of the rectangular bow and profiled wings, lead on to the Colchester brooch which is the British variation of the type (Bayley and Butcher 2004, 148–9). At Verulamium, KHL, Simple Gallic brooches (Type Bb), mostly with straight or even slightly arched bows were found in graves of phases 1 and 2, covering the period from the beginning of the 1st century AD to c AD 55 (Stead in Stead and Rigby 1989, 89; but note the subsequent shift in the dating of all cemetery phases c 10–20 years earlier, summarised, for instance, in Haselgrove and Millett 1997, 291–2). Unfortunately, most of the brooches of this type from Springhead are metal-detector finds from the sub-soil, but SF 1288 was found in pit 3363 within Late Iron Age enclosure 300037, while Cat No 14 lay in early Roman pit 2227 north of the portico structure. Settling the Ebbsfleet Valley 194 13 12 14 15 16 18 0 50mm 17 Figure 83 Springhead: copper alloy brooches 12–18 Chapter 3 Springhead Metalwork 195 20 19 22 21 23 24 26 27 Tinning Stone embedded in corrosion products 25 0 50mm 28 Figure 84 Springhead: copper alloy brooches 19–28 196 Settling the Ebbsfleet Valley Colchester brooches T90–91 With plain bow and multiple perforated catchplate (usually three triangular holes) 15. Complete. Long hook at head with flat broad end decorated with three ring-and-dot stamps. Ovalsectioned bow. Open catchplate with zigzag decoration along inner edge and below foot. Iron corrosion around head, probably from axis bar. Brass. SF 1506, Context 3447, Intervention 3446 (Pit), SG 300243 (Pits). 16. Complete. Hexagonal-sectioned bow. Long hook, slightly tapering towards its straight end, above the hook is a punched double concentric circle with double transverse notches either side. Remainder of bow plain, tapers slightly to foot. Large triple perforated triangular catchplate. Six spring coils. Brass. SF 15741, Context 0, Intervention 0. 17. Incomplete. Plain wings. D-shape sectioned bow with pronounced arch. Catchplate originally with triangular and large trapezoidal perforation. Five spring coils. Brass. SF 1608, Context 3722, Intervention 3720 (Ditch), SG 300045 (Trackway). Early Roman. See also SFs 361, 909, 1235, 1477, 1665, 9372, 18015 and 18339, the latter, which is badly distorted, could well belong to the preceding type. With single groove on bow 18. Incomplete. Long hook gets thinner after chord, rounded end with circular punch decoration. Sharp angle at head, hexagonal-sectioned bow, tapers slightly to foot. Broad central groove down length, on reverse of upper part of bow Two parallel grooves faintly visible. Perforated catchplate, half missing. SF 18335, Context 17439, Intervention 17439 (Layer), SG 300599 (Layers), Property 3. Mid-Roman. See also SFs 9342 and 18744. Small Colchester brooches (41 mm and shorter) (Fig 84) 19. Almost complete. Short side wings cover only half of spring on left, broken after second inner coil on right. Sharp angle at head of bow; chord hook reaches up as high as this bend. Oval-sectioned bow, tapers slightly to foot. Transverse groove visible at foot end, zigzag decoration along junction of catchplate and bow. Catchplate is subrectangular, almost triangular. Tip of pin missing. Eight spring coils. Leaded brass. SF 20266, Context 16039, Intervention 16039 (Layer), SG 300485 (Layers), Property 10. Mid-Roman. 20. Complete. Two transverse grooves decorate both wings. D-shaped sectioned bow, tapers to foot. Triangular catchplate, punched decoration along junction to bow. Seven spring coils. Brass. SF 15213, Context 10664, Intervention 10664 (Layer), SG 300407 (Layers), Property 11. 21. Almost complete. Wings have stepped decoration. Dshaped sectioned bow, tapers slightly to foot, central rib down length with groove decoration from head to mid point of bow. Triangular catchplate. Tip of pin missing. Eight spring coils. (Leaded) brass. SF 907, Context 5414, Intervention 5414 (Layer), SG 300148 (Deposits). Early Roman. See also SFs 253, 9152 and 18809. Variant with flat bow 22. Incomplete. Hook, part of wings, and part of spring coil remain at head. Flat cross-sectioned bow, tapers slightly, tip of foot missing. Lower part of catchplate missing. Badly corroded. SF 437, Context 2230, Intervention 2227 (Pit), SG 300073 (Pits). Early Roman. As mentioned above, the Colchester brooches are the British variation of the ‘Simple Gallic’ brooch based on the LaTène III-scheme. Colchester brooches are widely distributed in south-eastern Britain. They were made in Britain from around the birth of Christ and continued in use into the 2nd half of the 1st century (Bayley and Butcher 2004, 148–9). At Verulamium, KHL, they are found in graves of phases 1–3, spanning the 1st half of the century (Stead in Stead and Rigby 1989, 100–1; for the earlier dates of the KHL phases see Haselgrove and Millett 1997, 291), at Colchester the majority was found in post-Conquest but pre-Boudican contexts (Hawkes and Hull 1947, 309). On the basis of a comparison of earlier types and Colchester brooches from Colchester and Canterbury, Marlowe Car Park, Mackreth (1995, 957) discussed the possibility of a gap in the occupation of both sites, since, as he argues, none of the brooches of the main run of the type nor those with later traits, like developing flanges at the heads, needs to have been deposited before AD 40–5. At Springhead, brooches with earlier traits like a sharp angle at the head similar to type Ca at KHL include Cat No 15, SF 361, 909, 1477 with plain bows and Cat No 18 with a bow with single groove decoration. However, none of the larger Colchester brooches from Springhead has the decorated wings found in the typologically earlier types, all have the plain wings found in the typical Colchester brooch Type Cd at KHL and most of those found at Colchester; decorated wings are entirely confined to the smaller variant (Cat No 20–1, SF 253). The flat, highly arched bow of Cat No 22 is unusual, but considering the relatively sharp bend behind the head and the small rectangular hook it is probably relatively early in the sequence. It was found to the north of the portico structure in pit 2227 which contained pottery ranging AD 50–160. Of the seven Colchester brooches subjected to XRFanalysis it is interesting to note that all are made of brass apart from Cat Nos 19 and 21 – typologically among the latest Colchesters from the site leading on to the derivatives – which have been determined as leaded- and (leaded) brass respectively; this corresponds well with the results of Bayley’s study (Bayley and Butcher 2004, 148–9 figs 110–1). A very corroded Colchester brooch (SF 553) was found in the fill of Saxon grave 2827, and it is thus likely to have been an accidental inclusion rather than a deliberate deposition of a curated object. Chapter 3 Springhead Metalwork The Colchester brooches are distributed relatively evenly between the two sides of the settlement at Springhead (Fig 85). 26. Broad, rounded head, edges of bow tapering in a gentle curve to foot 23. Incomplete. Spring cover complete, pin missing. Reeded bow quite narrow, tapers slightly to foot. Most of originally perforated catchplate missing. SF 943, Context 6035, Intervention 6035 (Layer), SG 300087 (Deposits). Early Roman. See also SFs 163, 765 and 9344. 561700 Broad, rounded head, edges of bow waisted in the middle, foot as wide as head 24. Incomplete. Bow has flat cross-section with beaded rib decoration down length. Spring cover has three incised grooves at edge and radial lines on front face (only visible on the right due to corrosion and mud on other side). Sub-rectangular catchplate with triangular opening. Pin broken Brass. SF 632. 25. Incomplete. Spring cover is complete, pin missing. Longitudinal ribs decorate bow which flares out very Colchester Brooches Straight head with sharp angle, edges of bow waisted in the middle, foot as wide as head 27. Complete. Tips of cylindrical cover slightly damaged. Bow of rectangular shape flaring out slightly towards foot. Decoration of three longitudinal grooves, outer two contain greyish/brown coloured material (trace of white metal coating), central groove with beading/zigzag decoration on ridges either side. Subrectangular catchplate. Brass. SF 379, Context 2220, Intervention 2220 (Artefact). Straight head with sharp angle to straight-sided bow 28. Incomplete. Flat-sectioned bow with reeded decoration, tapers slightly, small part of open catchplate 561900 Langton Down brooches T21 slightly towards foot. Large sub-rectangular perforated catchplate. Stone adheres to front of bow. SF 953, Context 400106, Spring. Almost complete. Cylindrical spring cover with groove running along outer edge and sides, squashed and part of inner side missing. Rectangular-sectioned bow, flares slightly towards foot end, decorated with longitudinal grooves. Triangularly perforated sub-rectangular catchplate. (Leaded) brass. SF 20165, Context 19462, Intervention 19544 (Ditch), SG 300627 (Ditch), Roadside ditch 3. Early Roman. 561800 Brooches with spring in cylindrical cover 197 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 85 Springhead: distribution of Colchester brooches Settling the Ebbsfleet Valley 561700 The 15 Langton Down brooches found at Springhead can be sub-divided into four different variants depending on the shape of the head above the cylindrical spring cover. The development of the type is subsumed in Feugère’s type 14 which includes Simple Gallic brooches (type 14a) whose wings eventually developed into the cylindrical cover found in the Langton Down and Nertomarus types (his types 14b1b and 14b2; Feugère 1985, 264–6). Langton Down brooches are fairly widely distributed in Gaul, especially middle and eastern Gaul and western Switzerland, the Rhineland, and southern Britain (Feugère 1985, 265; Riha 1979, 98; 1994, 87; Bayley and Butcher 2004, 150). The type developed in the Augustan period and, on the basis of the Augst evidence (Riha 1979, 99; 1998, 88 table 103) Feugère (1985, 266) suggested that the type with rounded head and waisted bow (Riha type 4.4.1) is earlier than that with straight head and bow (Riha type 4.4.4), which starts in Langton Down Brooches the late Augustan–Tiberian period. In Britain this suggested development from curved to straight is supported by the evidence from Verulamium, KHL (Stead in Stead and Rigby 1989, and less clearly so Colchester (Hawkes and Hull 1947, 318). At Springhead, Langton Down brooches are more commonly found on the Sanctuary site (ARC SPH00), with only two from the Roadside settlement to the west of the Ebbsfleet (Fig 86). Brooch with flat bow with twisted silver wire inlay (Fig 87) 29. Incomplete. Spring in cylindrical cover, half missing. Straight flat-sectioned bow tapers slightly, decorated with one lateral longitudinal groove either side and a central groove with inlay of twisted wire. The wire consisted of two strands, one of which is probably silver, the other very corroded, powdery, light green residues of copper alloy. Catch plate missing apart from small ridge near middle of bow. Three spring coils. Brass. SF 393, Context 2342, Intervention 2342 (Layer). Early Roman. A rather similar bow to that of Cat No 29 is found on a brooch with a simple four-coil spring with inner chord but without silver inlay from Camulodunum (Hawkes and Hull 1947, 318; pl 104, 85). While Hawkes’ and Hull’s 561900 remains, 19.5mm of pin remains. SF 168, Context 3996, Intervention 3223 (Ditch). SG 300030 (Ditched enclosure). Late Iron Age. See also SFs 320 and 1819. Fragments of Langton Down brooches, various variants: SFs 1530, 1789, 1807 (although this could also be the head of a rosette brooch like Cat No 32) and 18730. 561800 198 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 86 Springhead: distribution of Langton Down brooches Chapter 3 Springhead Metalwork 199 Silver ? 30 29 32 31 34 33 35 38 36 Surviving tinning 0 40mm 37 Figure 87 Springhead: copper alloy brooches 29–39 39 200 Settling the Ebbsfleet Valley claim seems reasonable that the Camulodunum brooch represents a prototype for the Langton Down type, mainly because of the spring construction, the Springhead brooch is clearly a more developed example. A small number of Langton Down brooches from Augst with silver inlays (Riha types 4.4.5–6) are technically similar but both the shape of their bows, which tend to be straight or flared towards the foot, as well as the sharp angle at the head and the wider cylindrical covers suggest (Riha 1979, Taf 19, 522–4; 1994, Taf 14, 2131) that they are typologically slightly later than the Springhead brooch; however, this cannot be verified by the contexts they were found in which only date rather generally to the 1st two-thirds of the 1st century AD. Brooches with a circular or rhomboid disc in the middle of the bow (rosette or thistle brooches) Rosette brooch with separate disc T26A 30. Feugère type 19d1. Incomplete. Right side of spring remains, pin broken, lower right side of spring cover missing. Head decorated with longitudinal grooves. Central lozenge-shaped plate (higher zinc brass than body) with triple lines and punched triangle-and-dot border decoration has a small slot pointing towards the foot end (cf Stead and Rigby 1989, 333 fig 133, 3 or 337 fig 136, 3), originally used to facilitate fitting of plate onto the bow. This plate is held in place by another crown-like ring (the remains of the rosette/thistle plate) with triangular indentations on its upper edge outside the bow and a folded flange below the bow where it is held in place by an iron rod with small copper alloy terminals. A further lunate disc with small, radiating lines is set in the space between the bow and the crown-like ring. Foot worn but longitudinal grooves still visible. Sub-rectangular-shaped catchplate on reverse. Five spring coils. Brass. SF 1568, Context 3232, Intervention 3231 (Ditch), SG 300030 (Ditched enclosure). Early Roman. Rosette brooches with disc cast in one with the bow T26B 31. Feugère type 16a2. Incomplete. Part of cylindrical cover missing, transverse groove visible along top. Longitudinal grooves visible beneath corrosion product on bow, part of central disc missing. On the back of the central disc two lines which continue the line of the foot. Sub-rectangular perforated catchplate. Part of pin retained separately. Seven spring coils. SF 9142, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. 32. Feugère type 16a2. Incomplete. Spring in cylindrical, slightly box-like cover with closed sides, partly broken, sides have incised cross decoration. Longitudinal ribs decorate upper part of bow, large disc cast in one with the bow continues into damaged foot, concentric groove decoration just visible but preservation is poor. Fragments of small sub-rectangular catchplate. Striations visible on reverse of brooch, which continue the outline of the foot on the disc. Brass. SF 18693, Context 19026, Intervention 19026 (Other). 33. 34. Feugère type 19a2. Incomplete. Spring cover decorated with double line border around angled grooves just above head. Longitudinal grooves on flat, profiled bow probably contain traces of tinning. Circular ‘rosette’ around junction of bow and foot. Foot flares out slightly, one corner slightly damaged. Spring missing. Catchplate with one round hole. SF 1541, Context 3391, Intervention 3391 (-). Feugère type 19d2. Incomplete. Spring cover decorated with longitudinal and transverse incised lines. Flat bow, upper part circular with grooved arched crest at centre and a lozenge-shaped disc which is cast in one with the bow and the spring cover, flared foot with longitudinal grooved line decoration. Perforated catchplate, pin rest missing but probably right sided. On reverse two incised lines radiate from below head. Pin missing. Leaded brass. SF 169, Context 3996, Intervention 3223 (Ditch), SG 300030 (Ditched enclosure). Late Iron Age. Thistle brooch with leontomorph bow and foot riveted to back of bow Riha type 4.5.7/Feugère type 19f2 35. Incomplete. Cylindrical spring cover. Reverse of bow is flat, upper part of front of bow with three transverse mouldings reminiscent of stylised lion. On the inside of the outer moulding is a short subrectangular stud/rivet. The foot, which would have been riveted to the bow, is missing. Seven spring coils (two left, five right). SF 50988, Context 200wb, Intervention 200wb. Mid-Roman. Simple rosette brooch T27/Feugère type 20c/Riha type 4.7.1 36. Almost complete. Half of spring survives in cylindrical cover. Flat-sectioned bow of disc and flaring foot, bent, no decoration visible. Subrectangular catchplate, tip of pin missing. Five spring coils. SF 927, Context 5600, Intervention 5600 (Layer), SG 300009 (Deposit). Mid-Roman. Hinged rosette brooches Feugère type 20d1 37. Incomplete. Hinged pin missing, pin was held by axis set in tube at top of bow formed by rolling forward the bow’s head. White metal coating. Disc on upper part of bow has circular perforation at centre with remains of iron rivet in place. Lower part of bow has lines along the edges and band of zigzag decoration down centre. Catchplate on reverse is incomplete. Brass. SF 18724, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. 38. Almost complete. Miniature brooch, of rosette or thistle type. Hinged pin held by axial rod set in outward-turned hinge, still free moving, tip of pin broken. Upper part of bow is oval-shaped plate with rivet protruding from centre; differential corrosion of plate might derive from rosette- or thistle-shaped foil. Lower part of bow is flat with moulded/scalloped edges and three wide longitudinal grooves. Sub-rectangular catchplate. Brass. SF 18881, Context 17803, Intervention 17802, 1, SG 300687 (Layers), Property 3. Early Roman. See also SF 316. Chapter 3 Springhead Metalwork Rosette brooch with double-lugged hinge T238/Feugère type 20e1/Riha type 7.10.2 39. Incomplete. Plate type, disc-shaped bow with flaring foot. Small perforation at centre for attaching decorated disc. Two small lugs extending from edge of disc, mark where third would have been on the left side. Hinged pin survives; small trapezoidal catchplate. (Leaded) brass. SF 15535, Context 16022, Intervention 16022 (Layer). A close parallel for this brooch with similar side lugs was found at Augst (Riha 1979, Taf 59, 1583). See also SF 15063 which is possibly the flared foot of such a rosette brooch while this is less certain with the small foot fragment SF 1897. The eleven brooches with a circular or rhomboid disc in the middle of the bow found at Springhead present an interesting sequence of the developed and later typological stages of this group of brooches. Those of Hull’s rosette or thistle type T26 are here further distinguished by the shape of the disc which can be separate from the bow, as in Cat No 30, or cast in one with it, as in Cat Nos 31–4. At Springhead, the latter can be further distinguished by the decoration of the arched bow and the foot which in Cat Nos 31 and 32 exhibits the rather plainer treatment found in Feugère’s type 16, while Cat Nos 30, 33, and 34 have the longitudinally grooved bows and feet of his type 19, already familiar from the Langton Down brooches. The lines (Cat No 31) and striations (Cat No 32) on the undersides of the discs of two of these brooches are nice examples of a typological rudimentary feature, in this instance reminiscent of the once separate disc which sat above the foot as in Cat No 30 or brooches of Feugère types 15, 16a1, 19a1, or 19d1. It has been suggested that incised diagonal crosses on brooches and other objects, like that found on the side of the spring cover of Cat No 32, probably have a religious significance (Henig 1984, 149; Butcher 1993, 157). Butcher has mentioned about 50 brooches from Britain, of which 20 were from religious contexts. Brooch types with such marks include penannular brooches, southwestern bow brooches, headstud brooches, and some crossbow brooches. An X-mark on a rosette similar but perhaps slightly earlier than Cat No 32 was found at Neuss, Germany (Simpson 2000, pl 1, 9; p 11; Simpson mentions a similar brooch from Hurstborne Tarrant, Hants, but does not mention whether it also had a diagonal cross. For the use of X-marks on pottery cf Seager Smith, Marter Brown and Biddulph, Chap 1, Fig 58). While in Gaul and southern Germany the plainer brooches of Feugère type 16 are already found in the last decades of the 1st century BC (Feugère 1985, 269), the more ornate type 19 sets in a decade or two later, in the late Augustan period (Hawkes and Hull 1947, 314), but both types continue well up to and after the middle of the 1st century AD. A fragment of a Feugère-type 16 brooch from Fishbourne (Cunliffe 1971, 103 fig 38, 22) with a disc smaller than that on Cat No 32 but already 201 attached to the spring cover, was found in a context dated earlier than AD 75. On account of its open spring cover, typologically slightly earlier than Cat No 32, is a brooch from Canterbury, Marlowe Car Park, for which Mackreth (1995, 972–3 fig 407, 83) suggests a date in the latter part of the 1st century BC. At Verulamium, KHL, where no Feugère-type 16 brooches were found, those of type 19 with the central disc cast in one with the bow are predominantly found in phase 2 and 3 graves (Tibero–Claudian and post-Conquest; but see also Haselgrove and Millett 1997, 291–2) with only one in a phase 1 grave (Stead and Rigby 1989, 93–4; type Fa and b), while the variants with separate discs, like Cat No 30 but all with circular discs (type Fd, and also the larger Fe), occur predominantly in the pre-Claudian phase 1 graves. At Springhead a pre-Conquest date is feasible for Cat Nos 30 and 34, both of which were found in the fills of late Iron Age ditched enclosure 300030 which had all but silted up by the mid-Roman period. Cat No 35, with its leontomorph bow, belongs to Riha’s type 4.5.7 which is mainly distributed in Gaul (but rare in the south, cf Feugère 1985, 291) and the Rhine provinces (Riha 1979, 105; 1994, 92–3). British finds are known, for instance from Colchester (Hawkes and Hull 1947, pl 93, 76, dated Claudian/Neronian and post-Boudican) and Verulamium, KHL Site graves 188, 218, and 306 (Stead and Rigby 1989, 94). A set of two such brooches was recently found associated with a third rosette brooch of Riha type 4.7.2 in grave 8273 of the small cremation cemetery south of Coldswood Road, Manston, Kent. The grave, which also contained a terra nigra cup Cam 56, two platters Cam 7/8/Thompson 1982, G1–6, and a whiteware butt beaker Cam 113, is likely not to date much after the Conquest (G Jones 2009, 155–6, fig 2.38). While Hawkes and Hull (1947, 315) still claimed that none of the British parallels need to be pre-Conquest, KHL Site grave 218 belongs to phase 2, giving a Tiberian to very early Neronian date (or late Augustan to Claudian, cf Haselgrove and Millett 1997, 292). The later development of the rosette brooches simplifies the construction of the bow, and the disc becomes attached to the spring cover without the upper part of the bow arching above it; Cat No 36 is the only example of this stage. Originally, it would have had a cover of sheet metal on the bow and foot. The funerary contents of the fight against chthonic forces of the scenes depicted on some brooches with preserved metal sheets can, unfortunately, not be verified because of the condition of the Springhead brooch (cf Feugère 1985, 294–5, fig 36), but it is interesting to note in this context that the brooch was found in the colluvial deposits of the spring area. The type is mainly Claudian in date (ibid, 297; Hawkes and Hull 1947, 316; Mackreth 1995, 972–3, fig 407, 84), but in Augst, Switzerland, where 19 examples have been found so far, context dates range from late Augustan to Claudian, with continuation of use possibly extending to the beginning of the 2nd century (Riha 1994, 94). The type is well represented in an arc reaching from western Switzerland along the Settling the Ebbsfleet Valley Early hinged brooches Aucissa brooches T51 Almost complete. Head with outward-turned tube hinge is very corroded, but notch in the middle of either edge still visible. Profiled bow with broad grooved rib down length tapers towards angled triangular catchplate. Lower part of bow has four transverse ridges above collared foot knob. Tip of pin is broken. Brass. SF 15968, Context 16825, Intervention 16825 (Layer), SG 300576 (Layers), Property 4. Mid-Roman. See also SFs 612 and probably 1860. Middle rib as high as side ribs Riha type 5.2.2 42. Incomplete. Inward-turned tube hinge, steeply arched bow is terminated by transverse cross mouldings at top and bottom. Central longitudinal rib of same height as sides. Lower part of bow plain, tapers to foot knob. Turn of triangular catchplate and pin missing. Brass. SF 206, Context 2100, Intervention 2100 (Layer). 43. Incomplete. Outward-turned hinge, pin missing. Rectangular-sectioned bow, gentle curve (most likely due to intentional flattening of original higher arched bow), three longitudinal grooves, central one contains zigzag decoration, transverse groove at base. Small triangular catchplate on thin foot with protrusion which carried now missing foot knob. Brass. SF 1875, Context 6444, Intervention 1002 (Spring). 44. Incomplete. Outward-turned tube hinge, axial rod held by large moulded knobs of which the left one survives. Wide, strongly curved bow with two deep longitudinal 561700 Middle rib higher than side ribs Riha type 5.2.1 (Fig 89) 40. Incomplete. Head with outward-turned tube hinge has central hole and two smaller ones either side of it (visible in x-ray). D-shaped section bow, tapers to foot, longitudinal grooved decoration, possible beading at foot. Small part of catchplate remains. Part of hinge and pin missing. SF 1843, Context 6445, Intervention 1000 (Spring). 41. Rosette/Thistle Brooches ( = later) 561900 Rhône to the Languedoc, with some in northern France and a fair number in south-eastern Britain (Feugère 1985, 296 fig 37). The next stages in the development of the rosette brooch affect the spring cover which is exchanged for a tubular hinge, as seen on Cat Nos 37–8 and SF 316, and is eventually dropped completely in favour of a double lugged hinge at the back of the lozenge-shaped or, as in the case of Cat No 39, circular plate. Both types should belong to the middle and the 2nd half of the 1st century (Feugère 1985, 297; Riha 1994, 93–4; 158; Bayley and Butcher 2004, 154–5). The distribution of rosette and thistle brooches at Springhead shows a prevalence of the earlier variants in the area to the east of the Ebbsfleet (Fig 88). 561800 202 er Riv Eb t lee bsf 172800 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 88 Springhead: distribution of Rosette and Thistle brooches Chapter 3 Springhead Metalwork 203 42 41 40 0 40mm Surviving tinning 45 Corrosion by products 44 46 48 43 49 47 50 Figure 89 Springhead: copper alloy brooches 40–51 51 204 Settling the Ebbsfleet Valley grooves. Bow tapers slightly to foot. Large (oversized) moulded knob attached to foot with lower side decorated with cut-out triangles, knob is split along one side. Leaded bronze. SF 1731, Context 6379, Intervention 1001 (Spring), SG 300015 (Deposits). Early Roman. A Hod Hill derivative brooch from Richborough has a very similar pronounced foot knob when seen from the top, which is, however, cast in one with the rest of the brooch (Bayley and Butcher 2004, 80, fig 60, 156). Even more ostentatious is a brooch from Grenoble, France, with five large knobs and a studded bow (Feugère 1985, pl 136, 1707). Bow with wide central groove Riha type 5.2.3 45. Incomplete. Head with inward-turned tube hinge, pin missing. Beaded line separates neck of bow from head plate with ring-dot punch either side. Flat-sectioned bow with two raised ridges down length, transverse grooves decorate bow towards foot. Badly bent. SF 1798, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. 46. Incomplete. Beaded decoration on head with inwardturned tube hinge. Flat-sectioned bow with longitudinal central groove, beaded decoration on either side. Moulded footknob. Bow bent so that head almost touches catchplate. Part of pin missing. SF 9256, Context 400104, Channel fills. Brailsford 1962, 8; Bayley and Butcher 2004, 151) and, on the basis of the Springhead evidence, there is nothing to refute that. Of the ten brooches found at Springhead (Fig 90) only Cat No 41 was found on the Roadside settlement site; of the others, five were found in the spring or channel fills (Cat Nos 40, 43, 44, SFs 612 and 1860). Aucissa brooches are widely distributed in the Roman Empire and especially common on, but not confined to, military sites of Augustan to Claudian date (A Böhme 1972, 11; Böhme-Schönberger 1998, 354; Feugère 1985, 319–20), in Britain, eg, at Richborough (Bayley and Butcher 2004, 66–9 figs 50–1; 151, map p. 189, fig 166) or Camulodumum (Hawkes and Hull 1947, 321–3, pl 96–7, 125–38). No brooch with a definite maker’s mark was found at Springhead; but Cat No 47 is too corroded to rule out the possibility of a stamp on its head plate. ‘AVCISSA’, from which the group takes its name, is only one among several others, albeit the most common, and, considering the small overall number of Aucissa brooches, disproportionately well represented in Britain (further discussion of the names and distribution map of AVCISSA stamped brooches in Feugère 1985 (321–2, fig 46). A more recent map and in-depth discussion of the possible later stage represented by brooches with the AVCISSA stamp is presented by Böhme-Schönberger 1998, 353–9, Abb 1). Bagendon brooches T52 T52A/Riha type 5.4. Incomplete. Brooch broken into two pieces which join (head broken off in antiquity). Head with outward-turned tube hinge, short stub of hinged pin remains and is still free moving. Panel of four transverse moulded ribs at top of head, central two are beaded. The highly arched, P-shaped bow is divided into four ribs pierced by four transverse iron axial rods, all bearing three copper alloy beads in the gaps between the ribs. A further panel of three moulded ribs, also beaded, define break to foot which is plain and tapering to a separate large moulded footknob. Remains of white metal coating which originally covered the entire surface. Triangular catchplate. Brass. SF 18742, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. See also SF 252 which has an undivided bow with fragments of iron rods pierced through the bow like T52B (Riha type 5.3). 48. Thin, strip-like bow Riha type 5.2.4 47. Very fragmentary. Probably closed tube hinge (certainly not turned outward); pin held by iron axial bar with side knobs. Transverse grooves decorate ends of head tube, part of flat head plate with side notches remains, perhaps with lettering on it, but it is too corroded to be certain. No catchplate survives, pin broken. Smaller fragment possibly lower part of bow. Brass. SF 820, Context 0, Intervention 0. The sub-division of the catalogue for the Aucissa brooches adapts that devised by Riha for the assemblage at Augst (Riha 1979, 114–21; 1994, 101–7). Feugère based his sub-division of the type predominantly on the construction of the tubular hinge and called brooches with inward-turned hinge type 22b1 (Cat Nos 42, 45–6, SF 1860), those with outward-turned hinge – the classic Aucissa type – are his type 22b2, (Cat Nos 40–1, 43–4 [variant], SF 612). It has been suggested that the variants with wider, flatter bows are somewhat older than those with thinner, wire-like bows (eg, A Böhme 1972, 11), but this need not be the case according to Riha (1979, 114) considering the evidence from Augst. Based on his typology, Feugère suggested a range between 20/10 BC to the beginning of Tiberius’ reign for type 22b1, and a similar or perhaps slightly later start for type 22b2, and an end of production by the end of the Claudian period (Feugère 1985, 323–4). In Britain, Aucissa brooches arrived in recognisable numbers only after the Conquest (Hawkes and Hull 1947, 322; The Bagendon brooches are variants of the Aucissa type and, as such, fall broadly into the same Augustan–Claudian chronological span, with the variant with undivided bows starting and finishing slightly later, staying in use into the Neronian period (Bayley and Butcher 2004, 151; Riha 1994, 107–8); at Augst on the upper Rhine there is evidence for both types occasionally continuing to the end of the 1st century (ibid) while, on the lower Rhine, the variant with divided bow is no longer in use by the Claudio–Neronian period (Haalebos 1986, 43; van der Roest 1988, 161). Generally not as frequent as the Aucissa brooch proper, Chapter 3 Springhead Metalwork 561700 a) With multiple cross mouldings on bow, flat panel on foot and half-profile foot knob T74/Riha type 5.6 49. Incomplete, pin missing. Tinned. Three beaded cross ribs on upper bow, lower bow has triangular shape with incised lines down sides and flat cross section. Small footknob. Triangular catchplate. SF 560, Context 2831, Intervention 2831 (Layer), SG 300186 (Deposits). Early Roman. 50. Incomplete. Upper part of bow rectangular-sectioned, longitudinal (possibly beaded) ridges down edge with three pairs of transverse knobs at sides. Lower part of bow is flat, triangular, tapers to small collared knob at foot. Triangular catchplate. SF 1840, Context 6445, Intervention 1000 (Spring). See also SF 645. Aucissa & Bagendon Brooches 561900 Hod Hill brooches T60–79 b) With lateral lugs at the top of the grooved bow T 63/Riha type 5.7.3 51. Almost complete. Hinged, axis bar missing so pin present but separate. Transverse mouldings on wings at top of upper part of bow, upper panel of bow also moulded with beaded decoration. Lower part of bow made up of two panels, edges concave, tapers to moulded foot. Subrectangular catchplate. Brass. SF 708, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. (Fig 91) 52. Incomplete. Hinged, iron axial rod still survives, right terminal knob missing. Upper half of bow decorated with double moulded knobs at top, niello (copper (I) sulphide) Y-shaped decoration repeated down central field, raised longitudinal mouldings all finely knurled. At junction between bow and foot three transverse rectangular cross mouldings, the central one with niello lines and knobbed terminals both sides; beaded transverse cross mouldings above and below central panel. Lower part of bow tapers to separate large, moulded, collared footknob, four circular perforations down either external edge probably held now lost riveted knobs (one appears to have rivet in place). Triangular catchplate with circular perforation. Pin bent backwards, tip missing. Brass. SF 15398, Context 10525, Intervention 10524 (Gully), SG 300389 (Gully), Property 11. Early Roman. 561800 it is interesting to note that, in Britain, the variant with divided bow T52A is mainly found on native sites like Bagendon or Maiden Castle but is lacking from military sites like Camulodunum, Richborough, or Hod Hill, thus suggesting that those found in this country are not linked to the invading army, while the later variants with undivided bows could have reached Britain both before and around the time of the Conquest (Simpson 2000, 38). Cat No 48, which was found in the channel fills in the waterfront area of property 4, can unfortunately not add anything to this question (Fig 90). 205 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 Unstratified 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 90 Springhead: distribution of Aucissa and Bagendon brooches Settling the Ebbsfleet Valley 206 54 Niello 53 52 55 57 58 56 61 59 60 0 Figure 91 Springhead: copper alloy brooches 52–61 50mm Surviving tinning Chapter 3 Springhead Metalwork Similar Y-shaped niello decoration on the side panels of a comparable brooch from Hod Hill (Brailsford 1962, fig 9, C69). See also SF 254. c) With lateral lugs at the bottom of the grooved bow, more or less flaring towards the narrower, usually profiled foot T61/Riha type 5.7.4 53. Incomplete. Flat rectangular-sectioned bow. Upper half with wide grooves either side of beaded ridges with two lines of ?tin/niello inlay, transverse cross moulding at bottom. Lower half of bow tapers to foot, with series of transverse cross mouldings. Catchplate and pin missing. Severely corroded. Brass. SF 1866, Context 6444, Intervention 1002 (Spring). 54. Complete. Upper, wider half of bow separated into two concave panels, beaded decoration on ridges in between. At mid point of bow two mouldings protrude. Lower part of bow has series of transverse cross mouldings, two of which are beaded. Foot tapers slightly to moulded foot knob. Large patches of, once probably complete, white metal coating on bow. Subrectangular catchplate. Brass. SF 15333, Context 10405, Intervention 10405 (Layer), Property 11. MidRoman. (Pl 3) 55. Incomplete. Trapezoidal panel on upper half of bow with two ridges dividing it into one wider central and two smaller lateral fields; side wings at bottom of panel. Narrow central cross mouldings define break to lower, flat part of bow and foot, tapering to moulded foot knob. White metal coating on bow. Subrectangular catchplate. Pin is broken but retained separately, rectangular cross-section. Brass. SF 18278, Context 16294, Intervention 16294 (Layer), SG 300484 (Road), Watling Street. See also SFs 784, 9358 and 18287. d) With lateral lugs set in the middle of the bow T62B/Riha type 5.7.6 None illustrated, see SFs 594, 1768 and 1801. SF 594 is extremely corroded and may possibly have transverse mouldings like Riha type 5.7.7. e) With transverse bar set in the middle of the bow with transverse moulding T62/Riha type 5.8 Not illustrated, see SF 212. Niello inlay, often found on this distinct type, is not visible on this specimen due to corrosion. f) With D-shaped bow, which can be decorated, and flat foot similar to Riha type 5.12.4 56. Incomplete. Hinged. Upper part of bow has D-shaped section, lower part below cross moulding is flat, tapers slightly to moulded foot knob. Transverse cross moulding at top of head, upper part of bow decorated with five transverse dotted lines and incised short transverse lines lengthwise (very faint/worn), lower part undecorated. Sub-rectangular catchplate. Two fragments of possibly circular-sectioned pin remain. Leaded bronze. SF 549. Contex 0. 207 Plate 3 Hod Hill brooch (Cat No 54), brass L 38 mm Photo: E Brook 57. Incomplete. Cross bar with iron corrosion around head possibly from remnants of axial rod. Transverse moulding at neck of bow, D-shape sectioned bow with central groove and flat beaded ribs either side, two further transverse mouldings define change to flat foot, tapers to half-profiled footknob. Triangular catchplate. Bronze/gunmetal. SF 715, Context 5414, Intervention 5414 (Layer), SG 300148 (Deposits). Early Roman. See also SF 9147. SF 1294 may also belong to Riha type 5.12, but is too corroded to be certain about the variant. g) With one or two transverse mouldings at head and bow tapering without break to foot with rudimentary knob Riha type 5.10/Bayley and Butcher 2004, Hod Hill d) 58. Incomplete. D-shape sectioned bow with two pronounced transverse mouldings at bend on upper part; bow tapers to narrow moulded foot. Triangular catchplate. Pin missing. Leaded brass. SF 15907, Context 16022, Intervention 16022 (Layer). 59. Complete. D-shape sectioned bow with two grooves besides a central rib, tapering towards foot which ends in half-profiled knob. Cross moulding at head. Perforated triangular catchplate. Brass. SF 315, Context 2221, Intervention 2221 (Artefact). See also SFs 548 and 901. h) With rhomboid bow T77/Riha type 5.10 60. Incomplete. Hinged pin, broken. Bow has sharp angle at top with double transverse moulding. Main part of bow is flat and of elongated rhomboid shape. Beaded rib decoration down centre of lozenge; grooved border on upper edges, diagonal notches decorate edges of lower part. Transverse double moulding at foot. Subrectangular catchplate. Brass. SF 1289, Context 3312, Intervention 3311 (Pit), SG 300214 (Pits). Early Roman. Bayley and Butcher (2004, 249) list four further examples from south-eastern Britain. A similar treatment of the foot, but on a different variant, is found on another brooch from Richborough (ibid, 78 fig 58, 135). i) Hod Hill derivative brooches with small round settings for inlays Riha type 5.16 61. Riha type 5.16.2. Incomplete. Hinged, pin missing. Beaded transverse cross moulding at head, upper part of bow rectangular in shape and section – top and bottom edges beaded, two recessed discs with perforation at centre possibly contained decorative inlay, now missing. Second beaded transverse cross moulding separates lower part of bow which is of 208 Settling the Ebbsfleet Valley D-shaped section, punched transverse lines centre, tapering to moulded rudimentary foot Triangular catchplate. (Leaded) bronze. SF Context 6444, Intervention 1002 (Spring). Well preserved inlays were found on a pair of brooches same type from Wederath-Belginum, Germany (Abegg 301 Abb. g and h; 306 Abb. 6). down knob. 1857, of the 1989, (Fig 92) 62. Riha type 5.16.4. Fragment. Rectangular crosssectioned bow with three punched dot-and-eye decorations poss held inlay or enamel, bow tapers to small moulded footknob (now twisted). Front of bow and hinge missing. Leaded brass. SF 15332, Context 10405, Intervention 10405 (Layer), Property 11. Mid-Roman. j) Uncertain Hod Hill variants 63. Incomplete. Broken, but fragments join. Hinge formed by rolling forward head of bow, iron axial rod still in place, pin missing. Rectangular-sectioned bow with sharp angle at head, tapers to foot. Corrosion obscures most decoration, but there may be traces of lines along edge near foot end. Hook of catchplate missing. Copper alloy. SF 20265, Context 16039, Intervention 16039 (Layer), SG 300485 (Layers), Property 10. Mid-Roman. SFs 657 and 15955 very probably also belong to the Hod Hill series. The sequence adopted here for the Hod Hill and Hod Hill derivative brooches loosely follows that of Riha for her types 5.6–5.16, not all of which are represented at Springhead. The series is generally accepted to have developed out of the Aucissa brooches, although it has been suggested that the varieties with lateral lugs similar to Springhead groups b–e could have been developed out of the Kragenfibeln (Feugère 1985, 247; 333; Haalebos 1986, 46; 92 fig 42, 21). The Continental series cover most of the 1st century AD, with a start in the Tiberian, possibly even the late Augustan, period (Riha 1994, 112–4 Tab 141, Variante 5.7.9) and some variants like groups f–h continuing into the 2nd century, a small number even reaching the 3rd (Feugère 1985, 335). While the low numbers of Hod Hill brooches at, for example, the Saalburg and Zugmantel forts in Germany (Böhme 1972, 12, Taf 2, 28–38) suggest they were already going out of use by Domitianic times, numbers only start to increase in the lower Rhine area in the years after the Batavian rebellion of AD 69–70 (Haalebos 1986, 47). In Britain, where they are mainly distributed south-east of the Fosse Way, their main period of use is in the years between the Conquest and about AD 70 (Bayley and Butcher 2004, 153; 191 fig 167). Two brooches from Baldock similar to those of group c from Springhead have been found in contexts ascribed a pre-Conquest date in the 1st and 2nd quarter of the 1st century AD respectively (Stead 1986, 124; 118 fig 47, 112 and 114); this has been called into question by Mackreth (1995, 975) based on the fact that at least the earlier brooch would be earlier than the varieties from which it descends. Considering Feugère’s suggestion mentioned earlier, that it is not only the Aucissa brooches but also the Kragenfibeln which may be counted among the predecessors, this early date may not seem so spurious after all. However, brooch 112 from Baldock is the only object from pit 121 with a date in the early 1st century AD; the pottery is dated to AD 70–120 (Stead and Rigby 1986, 418) and, therefore, the early date of the brooch is not compelling. The typologically early details such as the separate foot knob (Aucissa) and the remains of riveted knobs on the side of the foot (Bagendon) of Cat No 52 suggest that this brooch is likely to have been produced in the middle of the 1st half of the 1st century AD. Similar details can be found on an early Hod Hill brooch from Chichester (Mackreth 1978, 281 fig 10.27, 40; 285), and the subsequent development, where the foot knob becomes a moulded part of the foot, is demonstrated by examples from Hod Hill (Brailsford 1962, 9; fig 8, C59; fig 9, C66 and C80). Unfortunately, the pottery date ranges of the groups identified at Springhead do not allow to distinguish chronological differences between the different variants; 21 brooches of this type were found on the Springhead Sanctuary site and only nine on the Roadside settlement (Fig 93). Colchester derivative brooches Two-piece Colchester brooches a) Central rib (or groove) down whole length of bow As the sub-varieties take into account the length of the brooch and the shape of the foot end, the following heads of brooches can be ascribed to group a) only: SFs 961, 981, 1561, 9255, 9432, 15786, and 20015. ai) Large brooches (over 45 mm) 64. Incomplete. Two-thirds of spring intact with some of pin. Simple curved bow, D-shaped cross section, tapers towards foot, central crest on upper part but lateral groves continue as thin lines to foot end. Triangular catchplate with triangular hole. Four spring coils on the right remaining. Leaded gunmetal. SF 500, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. 65. Incomplete. Transverse groove decoration on wings. Crest continues as ridge, longitudinal grooves on ridge and outer edges of bow. Large perforated triangular catchplate. Eight spring coils. SF 740, Context 2948, Intervention 2948 (Layer), SG 300156 (Deposits). Mid-Roman. 66. Incomplete. Wings have four transverse grooves each side. Pronounced crest on upper part of bow. Ridge down centre decorated with two parallel grooves. Part of pin remains but is separate. Catchplate perforated with circular and triangular holes. Eleven spring coils. (Leaded) bronze/gunmetal. SF 814, Context 5745, Intervention 5745 (Layer). Chapter 3 Springhead Metalwork 209 Surviving tinning 62 65 63 66 64 67 68 69 0 50mm Figure 92 Springhead: copper alloy brooches 62–9 Settling the Ebbsfleet Valley Incomplete. Pin missing. Possible groove decoration on wings. Plain crest continues as rib down centre of bow with zigzag decoration, tapers slightly to foot. Perforated triangular catchplate. Casting flash visible on reverse. Nine spring coils. SF 1279, Context 3324, Intervention 3324 (Layer). 68. Incomplete. Spring intact, pin missing. Plano-convex sectioned bow, central ridge down length. Large triangular catchplate. Ten spring coils. Leaded copper alloy. SF 15187, Context 10806, Intervention 10806 (Surface), SG 300438 (Layer), Property 11. See also SFs 629, 962, 1518, 1839, 1845, 1864, 9361, 15217, 15344, 18374, 18887, 20162, and foot 15043. 561700 aii) Small brooches (less than 45 mm long) 69. Incomplete. Pin missing. ?Tinned. Possible ringed decoration visible on tip of right wing. D-shaped sectioned bow, tapers to foot, pronounced ridge down centre. Sub-rectangular catchplate. Seven spring coils. SF 625, Context 5039, Intervention 5040 (Gully), SG 300192 (Ditch). Early Roman. (Fig 94) 70. Incomplete. Spring present but very corroded, pin missing. Transverse groove decoration on wings. Pronounced crest at head with punched transverse line decoration. Pronounced ridge down centre of bow (almost triangular cross section), punched decoration on front, longitudinal grooves down either side. Bow Hod Hill Type Brooches 71. 72. 73. 74. tapers slightly to foot. Perforated sub-rectangular catchplate. Eleven spring coils. SF 1804, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. Incomplete. Spring intact, pin broken. Grooved decoration on tips of wings. Triangular-sectioned bow, grooved decoration down centre with beaded edges and zigzag central lines. Small sub-rectangular catchplate. Nine spring coils. SF 1846, Context 6445, Intervention 1000 (Spring). Incomplete. Spring intact, pin broken. Beaded and grooved decoration on tips of wings. Bow with one beaded rib either side of central groove, central crest at top of bow slightly damaged in the middle. Small subrectangular catchplate with large perforation. Nine spring coils. SF 1847, Context 6445, Intervention 1000 (Spring). Incomplete. Flat-sectioned bow with single longitudinal ridge, zigzag line at top of bow, beaded further down. Triangular perforated catchplate. Bent. 5 spring coils. SF 1862, Context 6444, Intervention 1002 (Spring). Incomplete. Part of spring remains, crest at head. Pronounced ridge along bow, tapers slightly to foot. Sub-rectangular perforated catch plate, the upper part of the perforation drilled, the lower part thinner and extending towards middle of catchplate. Casting flash/working marks visible on reverse of bow. Three 561900 67. 561800 210 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 93 Springhead: distribution of Hod Hill brooches Chapter 3 Springhead Metalwork 70 211 71 72 73 75 76 74 81 77 78 80 0 79 Figure 94 Springhead: copper alloy brooches 70–81 80 40mm 212 Settling the Ebbsfleet Valley spring coils. SF 9145, Context 6447, Intervention 1001 (Spring), SG 300012 (Watercourse). Early Roman. 75. Incomplete. Wings decorated with five transverse grooves each side. Bow has grooved ridge down length, tapers slightly to foot. Triangular catchplate, upper edge concave. Pin missing. Eight spring coils. Leaded brass. SF 15343, Context 10808, Intervention 10808 (Layer), Property 11. Early Roman. 76. Incomplete. Pin missing. Crest continues as ridge with zigzag decoration down length of bow, numerous transverse striations on reverse of bow from filing during manufacture/production. Triangular catchplate. Eight spring coils. Leaded bronze. SF 18334, Context 17439, Intervention 17439 (Layer), SG 300599 (Layers), Property 3. Mid-Roman. See also SFs 178, 436, 504, 551, 634, 661, 714, 728, 1233, 1251, 1727, 1861, 1876, 1877, 1882, 9257, 15218, 15227, 15399, 15680, 15894, and 15924. aiii) With footknob 77. Incomplete. Bow has simple curve, tapers slightly. Raised ridge decoration along upper third of bow, Dshaped section with slight ridge on top. Foot ends with knob. Triangular catchplate with triangular perforation. Pin missing. Eight spring coils. Leaded bronze/ gunmetal. SF 181, Context 3988, Intervention 3081 (Ditch), SG 300046 (Ditch). Mid-Roman. 78. Incomplete. D-shaped sectioned bow, tapers slightly to moulded footknob. Crest continues down length of bow as ridge. Triangular catchplate. Pin missing. Spring coil shows differential preservation to body of brooch. Ten spring coils. Leaded copper alloy. SF 20457, Context 16894, Intervention 16902 (Pit), SG 300579 (Pits), Property 4. Mid-Roman. See also SFs 664, 1401, 9343, -506 and most likely SFs 9430, 15140 and 15264 of which only the feet and catchplates remain. b) Crest or groove on the upper bow only bi) Large brooches (longer than 45 mm) 79. Incomplete. D-shaped sectioned bow, tapers to foot. Beaded crest with groove at top of bow, four transverse grooves towards foot of bow. Triangular perforated catchplate. Left side of spring missing, pin broken. Three spring coils. SF 1410. 80. Incomplete. Two fragments – uncertain if from same brooch. 1: body of Colchester two-piece type, spring and pin missing. D-shape sectioned bow, tapers towards foot, longitudinal groove on upper part with zigzag decoration. Triangular perforated catchplate. 2: spring fragment: 3 coils with axial rod in situ and part of pin. Brooch: leaded brass. SF 18729, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. See also SFs 180, 721, 15226 and 18186. bii) Small brooches (shorter than 45 mm) T93A 81. Incomplete. D-shaped sectioned bow, decorated with deep central groove to midpoint, beading either side. Lower part of bow decorated with small transverse grooves to foot. Subrectangular-shaped perforated catchplate. Pin missing. Seven spring coils. Leaded gunmetal. SF 177, Context 3988, Intervention 3081 (Ditch), SG 300046 (Ditch). Mid-Roman. (Fig 95) 82. Incomplete. Spring and pin missing. Transverse groove decoration of two parallel lines on more complete wing. D-shaped sectioned bow, single groove on upper part, beaded decoration either side, tapers slightly to foot with raised tip. Triangular catchplate. SF 1272, Context 3289, Intervention 3289 (Artefact). 83. Incomplete. D-shaped sectioned bow, tapers to foot. Short groove at head, slight beading visible either side, transverse groove at foot. Triangular perforated catchplate. Most of spring and all of pin missing, one spring coil remaining. SF 15417, Context 16022, Intervention 16022 (Layer). 84. Incomplete. Small. Spring intact, pin missing. Dshaped sectioned bow, tapers to foot. Groove on upper half of bow. Sub-rectangular perforated catchplate (large perforation of irregular shape). Seven spring coils. SF 15848, Context 12465, Intervention 12446 (Other), SG 300349 (Pit), Property 2. Early Roman. See also SFs 1315, 1856, 9146, 9150, 9178, 15234, 15691 and -581. c) Plain bow T93B 85. Incomplete, in two parts. D-shape sectioned bow, tapers slightly to small moulded footknob. Short plain crest at top of bow. Part of spring and pin missing. Six spring coils. SF 20441, Context 17183, Intervention 17183 (Layer). Early Roman. A very similar but shorter brooch was found at Richborough (Bayley and Butcher 2004, 88 fig 67, 192). 86. Almost complete. Transverse grooves decorate wing tips. Oval-sectioned bow, tapers slightly towards foot. Triangular catchplate. Tip of pin missing. Seven spring coils. SF 18765, Context 17855, Intervention 19597, SG 300613 (Layers), Property 3. See also SF 1728. Additional to the brooches mentioned above, the very corroded brooch heads SFs 1587, 1713, and 9205 could at least be identified as belonging to the two-piece Colchester rather than the one-piece type. The same is true of the 18 feet with catchplates of SFs 374, 607, 700, 915, 1292, 1462, 1552, 1835, 1858, 1881, 9206, 9321, 9460, 15143, 15342, 15889, and 18871. The two-piece Colchester brooches are here classed according to the system devised for the Richborough assemblage by Bayley and Butcher (2004, 82–9, figs 62–9). They are a development of the one-piece type T90–91 via the dolphin brooches T94A (ibid, 157; Mackreth 1981, 137–8). The two-piece Colchester is a British form with a distribution mainly south-east of the Fosse Way, especially common in East Anglia and northern Kent (Bayley and Butcher 2004, 194 fig 170). At Colchester the type has been dated to c AD 50–65 (Hawkes and Hull 1947, 311; type IV), revised to Chapter 3 Springhead Metalwork 213 82 83 84 85 86 88 0 50mm 87 Traces of enamelling 89 Figure 95 Springhead: copper alloy brooches 82–90 90 Settling the Ebbsfleet Valley 214 Table 48 Springhead metal finds: ‘eye’ motif and derivative used in catchplate perforations of two-piece Colchester brooches T92–93 group Ring in triangle (‘eye’ motif) Triangle and round hole ai) Cat. No. 65–6 & SFs 1518, 1845, 1864 & foot SF 15043 SF 629 aii) bi) bii) foot only Cat. No. 73 & SFs 504, 1233, 1727, 15894 SF 15226 Cat. No. 84, SFs 1856, 15961 SFs 607, 1462, 15342, 15889 AD 50–70 by Crummy (1983, 12) for the Colchester B, which is equivalent to groups ai and aii, and AD 65–80 for the Colchester BB which is similar to groups b and c; none of the latter two groups were found at the Colchester Sheepen Site, which supports their later date. Among the brooches of group aii from Springhead there is one (SF 1251) which, unlike the others, has lines of rocker mark decoration either side of the central ridge. This has been called a major variant of the Harlow type by Mackreth (1995, 959–60). One specimen from Verulamium was found in the upper filling of a context dated 150–70 (Goodburn 1984, 22–3, fig 6, 25). While this variant is rare both at Springhead (one of 105 T92–93) and Richborough (one ai and two di) of 66 T92–93 (see Bayley and Butcher 2004, 83 fig 163; 88 fig 67, 193–4), it is proportionally more common at Canterbury Marlowe Car Park where three out of nine T92–93 have this form of decoration (Mackreth 1995, 959–60, fig 403, 14–16). Another detail which was observed in six of the 18 brooches of group ai at Springhead is a catchplate whose perforation takes the form of a ring set into a triangular or sub-triangular opening, and it is here proposed that this may be intended to represent an eye with the ring representing a staring pupil. This detail seems to be confined to brooches of group ai, probably because these larger brooches provide enough space for the motif (Table 48). Individual examples are not unknown from other sites (eg, Colchester, Hawkes and Hull 1947, pl 91, 41; Richborough, Bayley and Butcher 2004, 83, fig 62, 160; Canterbury Marlowe Car Park, Mackreth 1995, 960 fig 403, 10; West Thurrock, inhumation burial 17059, Schuster 2009, fig 8, 19570), but no assemblage with more than one example is yet known to the writer. Combinations of a triangle near the tip and a round hole near the inside edge of the catchplate are derivations of this motif, mainly found in the smaller variants but one also on an ai brooch, SF 629, at Springhead. In terms of the chronological development of this detail it is interesting to note that the ‘eye’-motif only occurs on ai brooches, whereas the triangle-and-hole motif occurs on examples of groups ai, aii, bi, and bii. Good parallels, still quite close to the original motif, were, for instance, found on brooches of group aii at Harlow (Gobel 1985, 73, fig 40, 52) and Colchester (Hawkes and Hull 1947, pl 91, 38), and of group bii at Richborough (Bayley and Butcher 2004, 87, fig 66, 187–8.191). Group aiii may be slightly later, in line with the later range of AD 75–125, if not earlier, suggested by Mackreth (1995, 961) for brooches with foot knobs more like Richborough group dii. Hull suggested a date of Vespasian or earlier for a specimen from Lullingstone villa, which is especially similar to SF 664 (Hull in Meates 1987, 63–4, fig 24, 58). Although irrelevant for the date, it should be noted that the foot of a two-piece Colchester brooch aiii, SF 9430, was found in the fill of Saxon SFB 5809. With 105 examples recovered during the HS1 excavations, the two-piece Colchesters are the most numerous type of brooch at Springhead, with more than two thirds found on the Sanctuary site or in the Ebbsfleet itself (Fig 96). Among the T92–93 brooches from the Sanctuary just under half (34 of 70) had perforated catchplates, while within the Roadside settlement and the Ebbsfleet this feature only occurred in just over a third of the brooches (14 of 35). Equally, the occurrence of fragmented brooches, either only heads or feet with catchplates, was markedly different in these two areas of Springhead: 14 feet and eight heads were found on the Sanctuary site; west of the Ebbsfleet the numbers are seven feet and two heads. Some of them show clear signs of deliberate breaking, eg, the foot SF 1462 with a lip from breaking on the inside edge, or the heads SFs 961, 981, or 9255, where the straightening of the bow happened before it finally snapped; although very corroded, the extremely twisted foot SF 607 is also the result of a deliberate break. It should be mentioned that some of the 21 feet mentioned here could well belong to a Polden Hill type brooch like SF 18743, but considering the large number of T92–93 brooches compared to only three (see below) with a Polden Hill spring arrangement, the resulting error will not be significant. The results of XRF-analysis of 13 two-piece Colchester brooches have shown that none was made of brass, only two of leaded brass, two of (leaded) bronze, and the remainder of leaded bronze and/or leaded gunmetal. The change from brass as the preferred material for the Colchester brooches to leaded bronze for the two-piece type has been linked to a reduced availability of brass and an increased supply of lead in the second half of the 1st century AD (Bayley and Butcher 2004, 155–6 fig 123). The results of the Springhead assemblage fit well with this outlined development. Dolphin brooches T94A 87. Incomplete. Both wings have double transverse grooves at tips. Head quite sharply angled, crest at top of head has two deep transverse ridges and extends onto bow with double longitudinal groove decoration. Bow ovalshaped cross-section; longitudinal grooves define edges, slight ridge down centre. Tapers to small Chapter 3 Springhead Metalwork Polden Hill brooches T95 Two Piece Colchester Brooches Incomplete, spring and pin missing. Wings each decorated with central band of diagonal ribs. Spring held in the Polden Hill manner, one closed end remains to hold now missing axial rod. Long, tapering, now slightly flattened bow with two central ribs joining in the middle to continue to foot as one. Triangular catchplate with two irregularly shaped holes. Quite similar to hinged dolphin brooch from Richborough (Bayley and Butcher 2004, 90, fig 70, 206) apart from 561900 89. 561700 The sprung dolphin brooch T94A is regarded as an intermediary stage between the one- and the two-piece Colchester types. The three from Springhead all share the construction with a backward hook holding the spring in place; apart from that they are quite different from each other. The type is generally dated to the first few decades after the conquest (Hawkes and Hull 1947, 311; Mackreth 1981, 137–8; 1985, 15; Bayley and Butcher 2004, 157). Cat No 87 is unusual in that it has a foot knob similar to that found in the two-piece Colchester group aiii. A sprung dolphin brooch from Soham, Cambridgeshire, has a similar knob, but the inside edge of the catchplate has a less step angle (Mackreth 1985, 16, fig 7, 95). On account of the unusual foot Mackreth suggested that this brooch may be slightly later than the general run of the type. The same may be true of SF 15927 and especially Cat No 88, both with catchplates which continue with a small web along the insides of the bow, a feature which is more commonly seen among the Polden Hill and T-shaped series (eg, Cat No 90). Good parallels for Cat No 88 have been found at Woodcock Hall, Saham Toney, Norfolk (Brown 1986, 24–5, figs 14–5, 60.65.77), but the webbed catchplate also occurs on a plain T94A from Hod Hill (Brailsford 1962, fig 6, C13). Dolphin brooches with rearward facing hooks are especially common in East Anglia (Brown 1986, 21–8, nos 60–86; Bayley and Butcher 2004, 157). 561800 moulded foot knob with slight collar above. Large triangular catchplate, small circular perforation at centre. Part of chord and single spring coil missing from right side. Eleven spring coils. SF 15039, Context 10013, Intervention 10013 (Ditch), SG 300364 (Ditch), Property 12. 88. Complete. Spring is separate but complete, chord held by a downward hook at the top of the bow. Wings decorated with transverse grooves and beading. Ovalsectioned bow, tapers to foot, longitudinal beaded ridge decoration. Triangular catchplate with web continuing half way up the bow. Thirteen spring coils. SF 15329, Context 10405, Intervention 10405 (Layer), Property 11. Mid-Roman. See also SF 15927 which is very corroded but probably belongs here. 215 er Riv Eb t lee bsf 172800 Unstratified brooches from Hill slope 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 96 Springhead: distribution of two-piece Colchester brooches 216 Settling the Ebbsfleet Valley the different spring mechanism. SF 563, Context 2875, Intervention 2874 (Pit). Early Saxon. See also SFs 1257 and 18743. The Polden Hill-manner of spring attachment is not very common at Springhead, a fact which need not surprise as it is generally considered a western phenomenon (Cool 1998, 29): only four definite occurrences and one possible are noted among the brooches of types T95 and T110. Those belonging to the Polden Hill series proper are of the early type T95. At Colchester, similar brooches have been dated preFlavian (Hawkes and Hull 1947, 311, pl 91, 42–3; Crummy 1983, 13, fig 8, 62–3). Unfortunately, there is nothing to confirm this dating at Springhead where two were found in Saxon contexts and one in the fills of the waterfront on the western side of the Ebbsfleet. T-shaped brooches 90. T110. Almost complete. Large, heavy brooch with its spring fixed by an axial bar set in the flanges of the plain cross bar in the Polden Hill manner. Chord held by small crest, tip of pin missing. Quite sharp, pronounced angle at head. Disc-shaped flanges on side of very upper part of neck of bow. Upper part of bow rectangular-sectioned, decorated with two enamelled (now orange/yellow) panels of three rectangles; small circular raised boss at base of panels also enamelled. Mid-point of bow has D-shaped section, double dotted punched lines link to lower part of bow; tapers to moulded foot knob. Two small moulded protrusions lead to double longitudinal ribs to foot. Solid triangular catchplate. Eight spring coils. Leaded copper alloy. SF 18383, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. See also SF 611 which is probably the bow and foot of a similar or slightly shorter brooch (eg, Bayley and Butcher 2004, 166 fig 138, T111). Cat No 90 is a developed T-shaped brooch; it has the Polden Hill-treatment of spring attachment, but its bow relates it to the hinged T-shaped brooches. Similar brooches have a mainly south-westerly distribution in Britain, with one from Croft Ambrey, Herefordshire, dated to AD 65–160, and others suggesting a date either side of the middle of the 2nd century (Bayley and Butcher 2004, 167). A comparable brooch from Nor’nour has a more elaborate crest but its panels lack the sub-divisions seen in the Springhead example (Hull 1967, 31 fig 11, 11). A complete example of a T-shaped brooch T111 (SF 1859, not illustrated as it only came to the author’s attention when all illustrations were finished), distinguished from those like Cat No 90 by its hinge whose axial bar is held by a narrow tube, was found in the channel fills of the Ebbsfleet. It has an unperforated head stud and remains of blue and other unidentifiable enamel in the two rectangular panels on its bow. It is unusual among the T111 in that it has a triangular knob pointing towards the foot where others of the same type have two leaf-shaped mouldings arranged in a V-shape pointing towards the head. A T111 brooch was found at Caerleon with pottery dating 130–60; the date range of the type probably covers the late 1st and early 2nd centuries (Bayley and Butcher 2004, 166–7, fig 138). The type has a south-western distribution (eg, Nor’nour: Hull 1967, 31–3 figs 11, 12–3; 12, 14–21; Butcher et al 2004, 20–2, fig 8, 301; Wanborough: Butcher 2001, 59, fig 24, 106). Fantailed brooches (Fig 97) 91. Incomplete. Hinged bow brooch with crossbar. Bow has flat cross section. Upper part has triangular ridge with beaded decoration either side. Lower part flares out, with three ring-and-dot decorations. Subrectangular catchplate with concave inner edge. Pin missing. Leaded bronze. SF 1426, Context 0. 92. Incomplete. Spring attachment as that of the Colchester-derivative brooches T92–93. Right side of spring survives, pin missing. Crest continues to middle of bow, grooved decoration on edge of silvered/tinned fantailed foot. Sub-rectangular catchplate, continues into low ridge on back of bow. Four spring coils. Leaded bronze. SF 1865, Context 6444, Intervention 1002 (Spring). See also SF 15040. Cat No 91 is one of a rare type which draws on typological and technological details found among some Colchester derivative (eg, from Stonea; Mackreth 1996, 297, fig 93, 9) and Aesica brooches (Hattatt 2000, 310, fig 169, 793.796), especially the Hook Norton type which has a similar pattern on the foot (cf Bayley and Butcher 2004, 151, fig 115), and ultimately goes back to the rosette brooches. A small number of close parallels for Cat No 91 have been found in the East Midlands and East Anglia; an outlier is recorded from Wiltshire (Hattatt 2000, 302, fig 161, 920). In the Stonea report, Mackreth (1996, 301) mentions similar brooches from Verulamium and Leicester but discusses them in conjunction with others which do not necessarily have a fantail foot, eg, the brooch from Stonea (see above). The dates range from the later 1st to the later 2nd century. Cat No 92 and SF 15040 belong to a relatively tight group of brooches that can be regarded as one of the prototypes of the Celtic fantailed brooches which usually have an enamelled foot and date to the late 1st and 2nd centuries. This prototype, named the ‘Maxey type’ by Hattatt (2000, 314, fig 173), is mainly found in East Anglia but he also lists one from Kent. One was found at Gorhambury, Hertfordshire (Butcher 1990, 116, fig 121, 16), and further examples from Kent include one from a later 1st century context at Lullingstone villa (Meates 1987, 64, fig 24, 56) and a foot from Richborough (Bayley and Butcher 2004, 99, fig 79, 239). A later 1st century date was also suggested by Crummy (1983, 164–5, fig 111, 2) for the brooch from Maxey, Cambridgeshire, on account of the spring Chapter 3 Springhead Metalwork 217 93 92 91 95 Silver wire Silver wire 99 94 97 96 98 102 Enamelling Surviving Tinning Silver wire 0 100 Figure 97 Springhead: copper alloy brooches 91–102 101 40mm Settling the Ebbsfleet Valley 218 construction similar to that of T92–93, while the fully fledged bow-and-fantail brooches have a hinged pin and usually a fixed headloop. Reversed fantail brooch 93. Incomplete. Double perforated lug on reverse of head for sprung pin (missing). Bow bent, resulting in head now sharply angled. Upper part of bow of D-shaped section, decorated by three longitudinal grooves. Lower part of bow in the shape of an inverted triangle with grooved borders. Sub-rectangular catchplate. Leaded copper alloy. SF 15926, Context 16149, Intervention 16148 (Ditch), SG 300473 (Pits), Property 10. Early Roman. This brooch shares many of the characteristics of the Maxey type, but the foot is turned 180º so that the end is pointed. While a number of reverse fantail brooches from East Anglia are listed by Hattatt (2000, 313, fig 172), their foot is usually more strongly profiled, and none have bows with longitudinal grooves like Cat No 93. Apart from one example from Saham Toney, Norfolk (Brown 1986, 28, fig 16, 87) with a spring construction like Cat No 93, all others have a reverse hook construction which suggests that the type developed in the latter half of the 1st century. The Springhead brooch was found in a ditch in property 10 with a pottery date range covering the later 1st and 2nd centuries. Headstud brooches T148C var. Incomplete. Hinged, pin missing. Wings decorated with twisted silver alloy wire inlay at tips and transverse grooves. Cavity in front of neck would originally have held ?enamelled stud, now missing. Bow also decorated with band of white metal. Moulded foot knob would have held held decorative stud, now missing. Triangular catchplate with web extending along length of reverse of bow to back of neck cavity. (Leaded) brass with silver alloy. SF 846, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. 95. T149B. Complete. Hinged, plain fixed headloop. Short wings have transverse grooved decoration. Rectangularsectioned bow, decoration of headstud and lattice including ten enamelled lozenge shapes (the first green–yellow, the others and the sides all corroded green). Two transverse mouldings above moulded foot knob. Sub-triangular catchplate. Leaded copper alloy. SF 704, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. See also SFs 15925 and 1959 (not seen by JS). SF 1887 may be a stud belonging to such a brooch. 94. The headstud brooches from Springhead belong to three different variants of the Lamberton Moor series. With its twisted silver wire on the long wings and the white metal inlay on the bow Cat No 94 is a more elaborate variation of the usually enamelled T148C, while SF 15925 is a sprung specimen of T148B with a bow decorated with rectangular cells of blue and green enamel. A third T148 from Springhead is illustrated by Böhme (1972, 49, Abb 6, 12). Both spring and rectangular cells have been suggested as indicating earlier forms of this type, eg as found on a brooch from Stonea, Cambridgeshire, dated to c AD 65–85 (Mackreth 1996, 308, fig 96, 44; 315; spring as indication of early sub-type refuted by Bayley and Butcher 2004, 167), and one hinged specimen with fixed headloop from Cottenham, dated before AD 100 (Mackreth 1985, 19–21, fig 9, 121). The end of the use of rectangular cells before AD 100 is supported by the evidence from Castleford, West Yorkshire (Cool 1998, 30–1). Both Springhead T148 brooches are lacking a headloop which would have been separate, while in Cat No 95 this forms part of the brooch. The separate, now missing studs of Cat No 94 suggest that this brooch belongs to the earlier run of the series, while the headstuds cast as part of the bow indicate a more developed stage (Crummy 1983, 13). Cat No 95 belongs to group 5a at Castleford which at that site was found in a context dating from the Flavian–Antonine period (Cool 1998, 30). Another early brooch of this variant of T149B was found at Chelmsford in a ditch context dated to before AD 100 (Butcher 1992, 72, fig 38, 24). Unfortunately, the wide pottery date ranges of the contexts at Springhead add nothing to this discussion, but SF 15925, stratigraphically related to the second phase of the smithy in property 10, would suggest a similar date. T148 and 149 are found throughout Britain although there are some variations depending on the sub-type; thus T148B has a more southerly distribution and is scarce in the north (Crummy 1983, 13), while T148C is more evenly distributed (Bayley and Butcher 2004, 198, fig 174). Trumpet-headed brooches 96. Incomplete, upper part of bow and head only. Probably has spring but due to corrosion cannot count coils, held on axial bar between two lugs behind flat oval head. Loop at end of rectangular lug cast in one with head. D-shaped section of bow, with double transverse moulding above break, too corroded to identify whether originally with acanthus or plain moulding. SF 810, Context 5750, Intervention 5750 (Layer). See also SFs 617, 647 and probably 18187. Cat No 96, SFs 617 and 647 are all missing the bow or at least the foot but are best classed according to the system suggested by Bayley and Butcher (2004, 93, fig 73, 220; 160–4), in which they belong to group C with a head based on a flat plate, a fixed headloop, and a spring held on a bar between two lugs. The foot and bow of SF 18187 is very similar to that of the Richborough brooch and thus likely to belong here, too, although the moulding on the bow also links it to a trumpet-related brooch from Alcester (Mackreth 1994, 175, fig 79, 61). Being the devolved copies of the standard trumpet brooches of the northern military area (group A), group C brooches date to the 2nd century and are distributed in the south and west of Britain (Bayley and Butcher 2004, 162; 197, fig 173). Chapter 3 Springhead Metalwork 219 Trumpet-head related brooches 97. Incomplete. Head only. Reverse of head is hollow, perforated lugs on side of head indicate where axial rod held now missing spring in place. Trace of headloop projection at top of head. Top edge of trumpet decorated with yellow–greenish enamelled convexsided triangles, below this are transverse triangular panels with blue and yellow enamel, on reverse of this part of brooch is a loop. Below this a semi-circular crest protrudes outwards, decorated with darker dots, probably of white metal. Remains of white metal also on ridge and in front of crest. Remainder missing. SF 18811, Context 16640, Intervention 16640 (Layer), SG 300491 (Layers), Roadside shrine. Cat No 97 is a colourful, enamelled variant of the Alcester type (T162) which is more commonly decorated with applied strips and coils of silver wire also found on the disc-on-bow variant discussed below (Bayley and Butcher 2004, 169; Hattatt 2000, 331, fig 190, 979–80.1539–40). Such decoration may originally have been present on the strips and dots of white metal remains still visible on the brooch. Alcester brooches are widely distributed in Britain in the 2nd century, although more common in the south (Bayley and Butcher 2004, 169; Cool 1998, 32). 98. Incomplete. Lug at top of head appears to be unperforated, spring between two lugs held by iron rod, pin broken. Semi-circularly shaped enamelled plate at mid point of narrow hexagonal bow; enamelled field now appears green, with three reserved metal spots covered with remains of white metal. Bands of white metal coating also along outer edge of head, centre of bow and foot and base of centre of bow. Lunula-shaped foot terminal also with white metal cover. Catchplate fragmented. Very corroded. Four spring coils. Brass. SF 1730, Context 6379, Intervention 1001 (Spring), SG 300015 (Deposits). Early Roman. This type is essentially half way between the disc-onbow type T166, discussed below, and the type with a pelta-shaped ornament on the bow, T167, but seemingly less common. Where the foot terminal is preserved, it always ends in a penannular shape like that found on the disc-on-bow brooches from Springhead (cf Cat Nos 100–1). Olivier (1996, 256–7, fig 11.10, 110) mentions 12 which are mainly distributed in the eastern half of Britain between the Thames and Humber estuaries, with outliers at Ilchester and Blandford. No firm dating has been established yet, and Springhead adds nothing new, but on account of affinities to the two types mentioned above, a similar date range in the 2nd century is suggested. A brooch with one lug behind a flat head found at Augst may be an imitation of such a brooch (Riha 1979, Taf 65, 1689). 99. Complete. Fixed head loop, groove decorates junction to head of brooch. Sprung pin with rear chord held between two perforated lugs on reverse of head by iron Plate 4 Bow brooch with blue and green enamel and white metal bands and dots which have lost their original decoration, probably of beaded silver wire (Cat No 99), leaded brass L 70 mm. Photo: E Wakefield axial rod. Upper part of bow has concave sides which broaden to ‘wings’/inverted crescent shape. Elaborate champlevé enamel in fields of now semi-translucent blue and green: at the top four inverted green triangles, followed by a concave-sided panel divided into four triangular fields by saltire-shaped ribs; the larger upper and lower fields filled with blue enamel, the narrower, lateral fields with concave outer sides, filled green. Below this follows a pelta-shaped field with blue enamelled background and a central circular ring-and-dot ornament: the ring of green enamel, the dot of the base metal covered with white metal. Foot has D-shaped cross section, median strip and dots either side with white metal coating. Bow tapers to ringed foot – cast in one – but white metal coating gives impression of penannular ring with rounded terminals. Subrectangular catchplate with short web along entire length of foot. Five spring coils. Leaded brass. SF 20115, Context 19398, Intervention 19398 (Layer), SG 300635 (Layers), Property 3. Early Roman. (Pl 4). See also SF 589. This very colourful brooch (Cat No 99; Pl 4) unites elements known from other types in an, as yet, unique combination: its lower part, the foot and the peltashaped part of the bow, relate it to the trumpet-headed brooches with straight foot and disc or half disc-on-bow (cf Cat Nos 98, 100–1). However, the head has a very different shape found in some fantail and fantailderivative brooches which are usually hinged (eg, Cool 1998, 44, fig 10, 51–2.55: Bayley and Butcher 2004, 170, fig 143, T163). The spring attachment with its two side lugs between which the rear-corded spring is attached is again reminiscent of the trumpet-headed types. The brooch was found in a context sealing the roadside ditch to the west of the bakery complex in property 3, which has a pottery date range spanning the period from the mid-1st to mid-2nd centuries. While the typological consideration would already suggest a date in the 2nd century, the context date confines this to its earlier half. 220 Settling the Ebbsfleet Valley Brooch SF 589 is most likely another trumpet-head related brooch, missing most of its bow and head but with a small knob in the middle of the bow, which is comparable to a Wroxeter-type brooch (T151) from Eye, Suffolk (Hattatt 2000, 331 fig 190,1538); however, that brooch also has a panel of chequered enamel on the bow, lacking in the Springhead specimen. A general 2nd century date may be suggested on that basis (Bayley and Butcher 2004, 169). Trumpet head related disc-on-bow brooches T166C 100. Incomplete. Spring held on bar between two lugs. Fixed headloop missing. Circular flat-backed waist moulding, four small lugs attached to its edge, centre of circle enamelled. Bow tapers to foot with penannular terminal; foot originally coated in white metal (tin/lead alloy). Hook on catchplate missing. Four spring coils. Leaded bronze. SF 1549, Context 3392, Intervention 3392 (Ditch), SG 300046 (Ditch). Early Roman. 101. Complete. Unperforated lug at top of head (imitates small headloop on Trumpet and Headstud types). Sprung pin, iron corrosion from axis adheres to reverse of head. White metal-coated dots on either side of trumpet head; differential corrosion possibly remains of silver sheet soldered onto head. Circular disc at top of bow, four small lobes protrude from its edge, centre of disc enamelled (now reddish colour), trace of silver thread around disc. Remains of applied beaded silver wire on lower part of bow which tapers slightly towards foot with annular terminal. White metal coating on top of terminal. Subrectangular catchplate. (Leaded) brass. SF 18741, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. See also SF 9359. Richardson first classified this type in 1960, already listing more than 30 examples from all of Britain, including the outer Hebrides. Their date range is mainly Antonine (Bayley and Butcher 2004, 170; Butcher 2001, 59–60 fig 24,115–6). Hull distinguishes four variants based on the treatment of the foot. All three brooches from Springhead belong to his Type 166C with annular or penannular terminal. A better preserved example of this variant, with the applied silver wires still in place, was found at Walbrook, London (Richardson 1960, 203 fig 2, 31). A further brooch of this type was found on the bakery, site A, at Springhead, but its foot is missing and thus it remains uncertain whether it belonged to the same variant (Penn 1957, 81 fig 14, 5; 98). Knee brooches Small knee brooches with cylindrical head T173A 102. Incomplete. Part of fixed headlloop attached to top of cylindrical head of spring cover. D-shape sectioned bow, tapers to foot which flares out again towards the end. Profile similar to knee type with sharp curve back from base of bow to foot, but angle at top of bow not so pronounced. Eight subrectangular panels, four either side of top of bow, originally held enamel decoration (now traces of yellow/green colour). Corroded remains of white metal band across top of spring cover and longitudinally along centre of bow and foot. Subrectangular catchplate, incomplete. Pin broken. Six spring coils. Brass. SF 18185, Context 17425, Intervention 17425 (-). (Fig 98) 103. Almost complete. Semi-cylindrical spring cover, rectangular-sectioned headloop attached. Top of bow is a lozenge-shaped panel filled with (now) pale yellow/green colour enamel and a dark (?black) central bead set in the centre and polished. Two transverse knobs protrude either side of the panel. D-shaped bow tapers to foot. White metal band across top of spring cover, around lozenge and a band along centre of bow and foot. Small sub-rectangular catchplate, pin bent and tip missing. Four spring coils. Leaded brass. SF 1729, Context 6379, Intervention 1001 (Spring), SG 300015 (Deposits). Early Roman. While rare on the German limes, in Britain T173A is found widespread south of a line between the Humber and the lower Severn Valley (Mackreth 1999, 222), both in military and civilian sites, although their generally small size indicates a more civilian use (Bayley and Butcher 2004, 180, fig 154). The composition of the two Springhead brooches conforms to the majority of the type which is made of brass (ibid, 180–1, fig 155–6). These brooches are often enamelled and/or applied with silver bands like the two examples from Springhead but plain brooches are also known as, for example, at Colchester (Crummy 1983, 14, fig 10, 69) or Fordham in Cambridgeshire (Mackreth 1985, 26, fig 11, 154). An example with a rectangular panel with concave sides was found in drain group 4 of the baths at Caerleon, dated Antonine–3rd century (Brewer 1986a, 171, fig 55, 12). This fits well within the main date range of c 125–225 suggested for the type by Mackreth (1999, 222). With bow of rectangular section expanding to square foot with transverse catchplate T176B 104. Variant. Incomplete. Spring in semi-cylindrical cover, pin broken. Rectangular-sectioned bow, pronounced arch at head, tapers to middle, flares out to rectangular transverse catchplate. Five spring coils. (Leaded) brass? SF 9374, Context 6682, Intervention 6682 (Artefact). Early Roman. In Hull’s typology, this brooch is of his type 176B, but the distinct semi-circular arch of the bow links it to Böhme’s type 21d for which she lists a number of parallels, mainly from Britain and the Upper Germanic limes (Böhme 1972, 21; 59 Fundliste 11: Taf 9, 462). The British examples include brooches from ditch fills at Richborough dated to 250–80 (Bayley and Butcher 2004, 101 fig 80, 241) and Springhead itself. The broader type is very common in the camps and forts of Chapter 3 Springhead Metalwork 221 104 103 Enamelling Surviving Tinning 0 50mm 106 107 108 105 110 109 Purple/white marbled glass 111 Green glass 114 113 Red glass? 112 Figure 98 Springhead: copper alloy brooches 103–16 116 115 Settling the Ebbsfleet Valley 222 the limes from the Antoninian period onwards and seems to go out of use around AD 200 (Böhme 1972, 21; Bayley and Butcher 2004, 179). In Britain the loose distribution reaches as far north as Newstead (ibid, 200, fig 175). P-profiled sprung brooches 105. Incomplete. The headplate has two lateral holes and a central small knob set off from the plate by a transverse moulding. Behind the head there is a lug with two holes to hold a double spring (or spring and chord?); part of iron spring bar survives in the lower, one coil of spring visible and in the upper hole part of ?chord or bar for second spring. Degenerated openwork decoration, reminiscent of trumpet-scrolls, on either side of bow (now flattened), central bar contains longitudinal ribbed decoration. Bow-foot junction marked by triangular-sectioned segment. Catchplate bent, extends only along two-thirds of foot length; foot flares towards triangular end. One spring coil. Leaded gunmetal. SF 828, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. This brooch shares certain characteristics with P-profiled brooches like Böhme Type 27c (1972, Taf 16, 691–5), eg, the foot and catchplate, and the headplate of type 27 is generally triangular with a central knob but lacking the lateral holes found in Cat No 105. The best comparison, however, is a very similar but sheath-footed brooch with a double spring, found in Well F.19 in an earlier excavation at Springhead (Hull 1970, 143–4, fig 2b). The well was constructed in the early decades of the 2nd century and remained in use possibly into the early 3rd (Harker 1970, 142). The details of the bow of Cat No 105 can also be found on a brooch from Stonea (Mackreth 1996, 312, fig 100, 88), and from the same site there is a silver brooch with slightly plainer bow with tendrils either side of top and end of the bow and a head with a wavy top edge and lateral tendirils where Cat No 105 has the holes (ibid, 298, fig 94, 17). The design of the bow of these and the Springhead brooches may be a RomanoBritish detail, although a brooch allegedly found at Augst has a similar bow design (Riha 1979, Taf 49, 1425), but there the decoration (thought to be ‘Germanic’) with drop-shaped openwork sits on the crest of the bow, not on its sides. While brooches with double springs are found in various types in the European Barbaricum as far east as southern Russia (Mackreth 1996, 304), the closer similarities among the brooches mentioned above suggest a date for Cat No 105 at the end of the 2nd, or more likely the early 3rd century. Plate brooches Flat petal-shaped disc with circular central motif Riha type 7.7/Feugère type 24f 106. Incomplete. Pin missing. ‘Petal’-shaped with ring decoration at centre. Possible perforation in middle. Two lugs for hinged pin on reverse, small catchplate. (Leaded) brass. SF 873, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. Cat No 106 is a rare variation of an otherwise not uncommon type with a small bone disc applied; one such disc is likely to have been attached to the centre of this brooch. The closest parallel is found at Augst (Riha 1979, Taf 59, 1565), and this better preserved example has a pattern of dotted twirls around the central disc. At Augst, the type is dated Claudio–Neronian to early 2nd century (Riha 1994, 157), and Feugère (1985, 344) suggests an end of production around AD 60/70. The distribution centres on western Switzerland and central France (ibid, 343, fig 52), with a small number in Britain where the brooches are mainly found in the south and east, although one is reported from Castleford (Cool 1998, 50, fig 13, 98; further parallels from Norfolk in Hattatt 2000, 342, fig 201, 513–4; Brown 1986, 36, fig 23, 163). Flat cruciform brooches with circular central motif T225/Feugère 24b1/Riha 7.4.1 107. Almost complete. Lozenge shape, slightly concave sides, upper and lower terminals decorated with two lobes and central groove, one of central terminals survives, undecorated. Circular depression in middle of plate, beaded ridge within, rivet hole (now empty) at centre. Two perforated lugs on reverse of head hold pin, tip missing; rectangular catchplate. Brass. SF 9218, Context 6629, Intervention 6621 (Ditch). Late Iron Age. See also SF 247. In contrast to the preceding type, T225 is fairly well attested in Britain, again mainly south of a line between the Wash and the Severn Estuary (Feugère 1985, 339, fig 49; Bayley and Butcher 2004, 154). At Augst the type is attested in Tibero–Claudian and Claudian contexts (Riha 1994, 154), while British finds are mostly postConquest (eg, Hod Hill, Brailsford 1962, fig 11, F3; Colchester, Hawkes and Hull 1947, pl 98, 165), although Mackreth (1996, 319) suggests a small number may have arrived earlier. It is generally assumed to have gone out of use by the late 1st century. While some examples (like that from Colchester) may have had an enamelled centre, those from Springhead are likely to have held a small knobbed rivet. Brooch in the form of a wheel T266B/Feugère 24c/Riha 7.6 108. Almost complete (tip of pin and central riveted stud missing). Face tinned/silvered. Solid outer ring with central open section divided by four spokes, perforation at centre possibly for holding decorative stud of bone or coral, now missing. Two perforated lugs on reverse hold pin. Subrectangular-shaped catchplate. (Leaded) brass. SF 1767, Context 6356, Intervention 6356 (Layer), SG 300099 (Deposits). Early Roman. Chapter 3 Springhead Metalwork Dating and distribution are similar to those of the preceding types, although this type is generally rarer. To the distribution map in Feugère (1985, 341, fig 50) can be added further British finds from Hod Hill (Brailsford 1962, fig 11, F4), Colchester (Crummy 1983, 16, fig 14, 86), Wanborough (Butcher 2001, 66, fig 26, 136), Hadrian’s Wall, and Oxfordshire (Hattatt 2000, 343, fig 202, 547.1004–5). An interesting observation is that, among the British examples, most lack the small lugs sometimes found around the outer edges of the wheels. Flat brooch with four round openings T245 109. Incomplete. Openwork frame, flat, with four large holes, the edge follows this shape. Four pointed lugs protrude from the corners where the scallops meet. Small pellet at centre. Top right corner of brooch is missing. One perforated lug on reverse of head (only base of 2nd one survives) would have held hinged pin (now missing), opposite this at foot of brooch is rectangular-shaped, rectangular-sectioned transverse catchplate. Leaded copper alloy. SF 18382, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. Hull has listed these ‘flat brooches with four round openings’ as his type 245 (cf Bayley and Butcher 2004, 239). There are a number from central southern England, including Silchester, Thunderbarrow Hill, Langton Matravers, and Winchester, with one from Nor’nour and the Springhead example marking the western and eastern fringes of the distribution. The most northerly so far seems to be one from Alchester in Oxfordshire (Lloyd-Morgan 2001, 224, fig 6.3, 7). Finds from dated contexts, like that from Winchester, suggest a date in the 2nd century, tending towards the second half of the century (Hull 1964, 89, fig 24, 6; 1967, 58, fig 23, 224), while the brooch from Alchester, found in an early–mid-4th century context is almost certainly residual. Early plate brooches set with glass or stone T224/Riha 7.8 a) Shaped like an eight-pointed star 110. Almost complete. Eight-pointed star shape with concave/scalloped edges. Front applied with tinned repoussé metal foil. A band of small raised dots follows shape of edge. Circular setting in centre, now empty. Complete hinged pin on reverse, pin slightly bent, subrectangular catchplate. Casting flash and other working marks visible on reverse. (Leaded) brass. SF 919, Context 6084, Intervention 6177/653, SG 300083 (Deposits). 111. Incomplete. Same as above. Green glass central setting. Eight-pointed star shape with concave/scalloped edges. A band of small raised dots follows shape of edge, followed by a circular rib around central setting. Hinge and catchplate on reverse of opposing points, part of pin remains but broken. Leaded gunmetal. SF 1844, Context 6445, Intervention 1000 (Spring). 223 b) Shaped like a lozenge with lobed corners 112. Incomplete. Lozenge-shaped ground plate with lobed corners, applied with repoussé foil. Red glass setting survives above catchplate, void where that above the hinge would have been. Perforated lugs on reverse hold part of pin, sub-rectangular catchplate. Very corroded. Leaded bronze. SF 1883, Context 6445, Intervention 1000 (Spring). 113. Almost complete – only part of surface decoration missing. Lozenge-shaped ground plate with lobed corners, applied with repoussé foil. Two bosses of opaque white and maroon marbled glass (K Hayward and T Goskar, pers comm) set above pin hinge and catchplate; surrounding this is a tinned repoussé decorated foil that has been applied after the glass settings. On reverse two perforated lugs and iron axial rod hold hinged pin in place, at foot a subrectangular catchplate. Lug for spring and catchplate connected by raised rib cast in one with plate. Leaded brass. SF 15634, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. Compared with the more common variant a, variant b seems to be much rarer with only two other specimens known from Avenches and Augst, both Switzerland (Riha 1979, 201, fig 32a; 1994, Taf 41, 2808); the glass roundels of the brooch from Avenches are shaped as human faces. A possible example from Britain was found in period 2 construction levels at Fishbourne (AD 75 or earlier; Cunliffe 1971, 106, fig 40, 40); at least the ground plate may have had a similar shape but the brooch is too corroded to be certain. A fourth brooch from Colchester has a fantailed foot and catchplate attached to one of the long sides, making it look much like the plate type of the rosette brooches, T238 (cf Cat No 39). A parallel for the marbled glass of Cat No 113 may be the glass setting of a star-shaped brooch from Augst, said to be made of white–blue glass (Riha 1979, 185; Taf. 59,1572). British examples of the star-shaped variant are known from Colchester (Crummy 1983, 16, fig 14, 77), Baldock (Stead 1986, 121, fig 49, 146), and Richborough (Bayley and Butcher 2004, 121, fig 94, 340 and further listed p 154). On the Continent, the earliest specimens have been found in Tiberian contexts, but the type is generally dated Claudian/mid-1st century (ibid, 154; Riha 1994, 158). This fits well with the evidence from Springhead, where Cat Nos 111 and 112 were found in a spring context with a samian date of AD 40–60. Disc brooches with central ornament Riha 7.2.1/Feugère 24a 114. Incomplete. Tinned, slightly raised border; central recessed area with central rivet hole which probably held ornamental stud. Approximately one third of disc missing. Two perforated lugs on reverse hold complete pin; small rectangular catchplate. Leaded brass. SF 9144, Context 6447, Intervention 1001 (Spring), SG 300012 (Watercourse). Early Roman. Settling the Ebbsfleet Valley 224 115. Incomplete. Composite plate of two joined discs, both with recessed centre and rivet holes. Grooved decoration around circumference. On reverse two lugs for hinge pin and remains of catchplate. Pin broken. (Leaded) gunmetal. SF 1709, Context 6380, Intervention 1003 (Spring). See also SF 1532. Members of this type can easily be confused with those represented by SF 1532 (see below), especially when badly preserved. Good parallels for Cat No 114 are known from Augst (1994, Taf 39, 2751–7), and the specimen from that site may hint at sub-division within Riha 7.2.1 where examples with a slightly raised rim like Cat No 114 never seem to have small lugs on the rim, which are an expressly mentioned characteristic of Hull’s T261 and Feugère’s otherwise similar type 24a (Feugère 1985, 335; in his list, he also mentions examples without lugs, such as from Colchester, cf Hawkes and Hull 1947, pl 98, 174.179). Cat No 115 with its two adjoining discs is a variant of the type; exact parallels were found at Wanborough (Butcher 2001, 63, fig 25, 129), and one at Augst has traces of red enamel in the central area (Riha 1994, Taf 39, 2767). A possibly similar brooch, found in the disturbed levels above the Harlow temple, is described as having small central bosses rather than holes for studs (France and Gobel 1985, 74 fig 41, 75). The dating is again mainly mid-1st century, which is well supported by Cat No 115 with a context coin date of AD 69, while Cat No 114 has a pottery date range of mid-1st to mid-2nd century. Brooch SF 1532 is very corroded and likely to have had a repoussé sheet metal soldered to the disc, which would relate it to Böhme’s type 44a, dated to the mid2nd century and continuing into the 3rd (Böhme 1972, 41–2; Taf 28, 1070–116). The Springhead brooch has a rib joining the double lugged hinge to the catchplate. This detail is also found on a brooch from the Saalburg fort (Böhme 1972, Taf 28, 1102) and the Germanic settlement Feddersen Wierde on the German North Sea coast (Schuster 2006, Taf 8, 61), and similar to a brooch from Richborough (Bayley and Butcher 2004, 130, fig 98, 374) which is, however, sprung like the majority of these brooches and those mentioned here. The type is found predominantly in the forts and vici of the Germanic-Raetian limes (Böhme 1972, 41–2). detail slightly better preserved in the Springhead example. Other brooches of the type have terminals in the shape of dolphins, acorns, or human faces (cf Riha 1979, Taf 66, 1700–06; 1994, Taf 45, 2888–97; Bayley and Butcher 2004, 121, fig 94, 346–7; Hattatt 2000, 343, fig 202, 1024).The type had its floruit in the second half of the 1st century AD (Riha 1994, 169; Taf 45, 2887), a dating which would suit the pottery and other finds found in the spring context with Cat No 116, while the coins date as late as AD 388. Toilet set brooch T199 (Fig 99) 117. Incomplete. ‘Umbonate’ plate type with one straight side where remnants of toilet set are attached. Pin hinged between two lugs. Four projecting lugs around edge, one missing, all probably enamelled. Raised central boss with six-petalled motif around it, alternately enamelled yellow and blue. Six larger petals cover part of brooch which is curved; straight side of brooch has two parallel lines of petal motif which are also alternately coloured yellow and blue. One hinge attached to straight side survives; attached to this (but bent backwards) is leaf-shaped nail cleaner with ringand-dot decoration at top of blade; single groove down length of blade, two grooved lines decorate junction and terminal. Terminal contains decoration of petal shape above sub-rectangular shape, both enamelled (now greenish colour). Scoop SF 20017 probably from this brooch. Leaded brass. SF 20014, Context 17759, Intervention 17759 (Layer), Property 3. Mid-Roman. These British brooches share a similar design with umbonate brooches T268, based on the ‘sunburst’ pattern of Roman origin. They are found mainly in the south of England and are dated around AD 100 (Butcher 2001, 61; Bayley and Butcher 2004, 172–3). The mixed nature of the context of Cat No 117 at Springhead adds nothing to the dating. ‘Buckler’, Tutulus, or disc brooches with conical centre surmounted by a knob or a button T269 118. Disc with looped attachment T242 var 116. Incomplete. Disc with recessed and perforated centre for inlay surrounded by omega-shaped twirl with everted terminals representing snakes’ heads. Tinned surface. Two perforated lugs on reverse hold broken hinged pin. Rectangular catchplate. (Leaded) brass. SF 1878, Context 6445, Intervention 1000 (Spring). 119. Riha puts a similar brooch from Augst, found in a context dated AD 70–220, in her rather broad type 7.21 of ‘brooches with medallions’ whose round discs can be surrounded by plant or animal ornaments in half profile. She suggests that the terminals may be snakes’ heads, a Feugère Type 25b. Almost complete. Main disc hollow on reverse, raised central stud. Outer flange with raised ‘lip’ on edge and beaded ridge decoration within, possibly enamel decoration around edge but now gone (surface very uneven, compared to beaded decoration), six small lugs symmetrically placed around circumference. Hinge of two perforated lugs on reverse at top, part of pin remains – rectangular-sectioned at top, tapers to circular-sectioned point. Small subrectangular-shaped lug on opposing edge for catchplate. Leaded brass/gunmetal. SF 362, Context 2222, Intervention 2222 (Artefact). Feugère Type 25b. Incomplete. Raised conical tutulus with small central knob in the shape of small cup with central cone and knob. Outer flange with six projecting roundels, four still retain green enamel, one with tiny central black glass bead pressed into base enamel without polishing (trace of holes in other lugs suggests Chapter 3 Springhead Metalwork 225 119 117 121 120 124 118 Surviving Tinning 122 125 123 Surviving Tinning 128 126 127 Remains of Gilding Glass Enamelling 129 0 130 Figure 99 Springhead: copper alloy brooches 117–30 50mm Settling the Ebbsfleet Valley 226 Plate 5 Tutulus brooch with star-shaped base (Cat No 120), (leaded) brass and enamel Diam 36 mm. Photo: E Wakefield 120. they may also have had central bead decoration). Hinge and catchplate attached to back of opposing roundels, one containing catchplate is bent at 90 degrees. Trace of flat-sectioned pin remains. Part of central cone is missing. SF 778, Context 5690, Intervention 5781 (Pit), SG 300050 (Pit). Mid-Roman T269 var. Incomplete. Star-shaped base with roundels at tips and in corners, which are filled with turquoise enamel. The triangular fields of the base each have three small circular depressions which are remains of small glass beads pressed into the corroded olive-brown enamel. Central tutulus crowned by circular central boss which is filled with orange enamel with five small depressions from small glass beads. None of the enamelled fields were polished. Hinge at back of head, catchplate fixed on lug, pin missing. (Leaded) brass. SF 9149, Context 6447, Intervention 1001 (Spring), SG 300012 (Watercourse). Early Roman. (Pl 5). The first two brooches are best classed in Feugère’s typology as type 25b which has a small cup on top of the tutulus, while those with only a knob are type 25a. Type 25b has a wider distribution than 25a, which covers midand southern Britain as well as central and eastern France, western Switzerland, and outliers in northern Germany, the Czech Republic, Italy, Serbia, Poland, and Syria (maps in Feugère 1985, 352–3, figs 53–4, additions for type 25b in Schuster 2006, 42). Feugère did not sub-divide type 25b further, but Cat Nos 118–9 present two of the main varieties with six or eight plain small lugs (Cat No 118) or larger discoid lugs which can be filled with enamel like Cat No 119; an exact parallel for the latter is known from the Saalburg fort (Böhme 1972, Taf 25, 967), while a very similar brooch from Richborough has an enamelled central cup but the lugs are set on a scalloped edge and have a pattern of concentric rings without enamel (Bayley and Butcher 2004, 130, fig 98, 377). The start date of Feugère 25b has been linked repeatedly to the Flavian grave 1, Grange Road, Winchester (Biddle 1967, 229–30, fig 4.14), but this brooch is more likely to be a transitory stage between Feugère types 25a and b (Schuster 2006, 42); however, there are parallels for type 25a with Flavian dates, eg, from Augst (Riha 1979, 186, Taf 60, 1587; 1994, 159–60, Tab 204). At Sulz, Rieckhoff-Pauli (1977, 17–19, Abb 6, 114–16; Abb 7, 122) was able to distinguish between the earlier (Claudio–Flavian to Domitian–Hadrianic) non-enamelled variations with smooth or tinned surfaces and knurled bands and later versions where enamel is already used. British examples of Feugère 25b continue in use throughout the 2nd century and possibly into the early 3rd (Bayley and Butcher 2004, 178); a very plain example like Cat No 118 from a 4th century context at Caerleon is most likely residual (Brewer 1986a, 171, fig 55, 23). Cat No 120 has been listed here as a variation of the type on account of its conically raised centre surmounted by an enamelled cup (Pl 5). The enamelled fields of the brooch do not appear to be polished and thus the now missing glass beads would have extended above the level of the base enamel. This technique has been identified as preceding the use of sections of glass rods set into the base enamel to produce a ring-and-dot pattern which is polished over. Unpolished beads appear as early as the beginning of the 2nd century, while the polished ring-and-dot inlays point towards the middle of the century (Riha 1979, 32; Feugère 1985, 364; but compare discussion of lozengiform brooch Feugère 26d1 in Schuster 2006, 41). A brooch from Woodyates, Dorset, shares the outline of the base plate but lacks the tutulus in the centre which is instead marked by a stud with conical head (Hattatt 2000, 356, fig 215, 589). So far, the closest parallel for Cat No 120 is a brooch from Suffolk (ibid, 354, fig 213, 1610) whose triangular rays have a narrower base. A similar brooch from Augst has a pyramid-shaped centre with a square plate showing an enamelled cross pattern. It is classed in Riha’s type 7.20 which is dated to the later 2nd century on account of the more complicated enamelled patterns (Riha 1979, Taf 65, 1697). Lozengiform brooch with frilled edges T240 var 121. Incomplete. Lozenge-shaped body with three stepped stages, central one of recessed field probably for enamel (now empty), second step has grooved ledge. Two complete ring-and-dot decorated projections survive on one edge, part of larger one visible on corner. Perforated lug remains on reverse for hinge, pin missing. Leaded brass. SF 18190, Context 17425, Intervention 17425 (-). Considering the projections on its side, this brooch is really an intermediary stage between brooches like a T227 from Richborough (Bayley and Butcher 2004, 127, fig 97, 364) or one from Nor’nour (Hull 1968, 49, fig 19, 146) and a T240 from the same site (ibid, 55, fig 21, 181), all of which would fall into Feugère’s type 26d1. The type is found in (mainly southern) Britain, Gaul, along the Rhine, in Belgium, and at least 12 are known in Germania magna (Feugère 1985, 362; Schuster 2006, 41). For dating see the remarks on enamel for the preceding three brooches. Enamelled disc with open centre T258 122. Incomplete. Circular, with large perforation in centre, recessed area enamelled (now greenish yellow colour), six small lugs protrude from rim. Two perforated lugs on reverse of head for hinged pin (missing), Chapter 3 Springhead Metalwork subrectangular-shaped catchplate at foot, bent, both are at the rear of protruding lugs described above. Leaded brass. SF 15633, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. Plate 6 Equal-ended brooch with arched, enamelled central Plate (Cat No 123), leaded bronze/gunmetal and enamel L 40 mm. Photo: E Brook Although a seemingly simple type, not many parallels can be quoted for Cat No 122. The closest comparison is an undated brooch from Colchester (Hattatt 2000, 345, fig 204, 530), and others are included in Riha’s type 7.12 which can be round or oval. The corroded enamel of Cat No 122 may have had different colours placed next to each other without separation, linking it to Riha 7.13. On the basis of this, a date around the middle of the 2nd century is proposed for this brooch. Equal-ended brooch with arched, enamelled central plate T229 123. Incomplete. Sharply curved rectangular-shaped bow, central panels tinned, beaded decoration along edges. Raised rectangular panel at centre bordered by beaded lines, decorated with six star or oak leaf shapes, four with orange enamel, the central two still contain residues of enamel, now turquoise. Moulded head and foot, two perforated lugs for hinge on reverse of head, subrectangular catchplate on foot. Pin missing. Leaded bronze/gunmetal. SF 342, Context 2235, Intervention 1002 (Spring). (Pl 6). Similar brooches are listed by Feugère as type 26c1a and found in southern Britain, France, along the Rhine and occasionally further east (Feugère 1985, 364–5; Riha 1979, 192). The pattern of small oak leaves is also found in various arrangements, for instance at Augst (ibid, Taf 62, 1627–8.1634–5.1639) and on related types, such as a T230 at Richborough (Bayley and Butcher 2004, 127, fig 97, 358). The date range is summarised by Bayley and Butcher (2004, 171; cf Feugère 1985, 364) and covers the late 1st and the beginning of the 2nd centuries. Lozenge, square or rhomboid-shaped brooches with ornamental projections at the angles T227 124. Incomplete. Perforated lug on reverse of moulded head, 125. pin missing. Lozenge (on its side)-shaped central plate, bordered by grooves and beading, two adjacent round holes in centre with four areas for decoration around them, two still have traces of enamel (now pale yellow colour), upper and lower areas with greenish corrosion products of enamel. One lug protrudes on left side with cross decoration, right lug missing. Foot and catchplate missing. Leaded gunmetal. SF 1863, Context 6444, Intervention 1002 (Spring). Incomplete. Lozenge-shaped panel set on its side, beaded decoration around edge, symmetrical enamelled decoration within, outer subtriangular fields opaque olive green with some orange specks and circular holes which contained glass beads, now missing; enamel in central field corroded to turquoise colour. Two perforated lugs on reverse of moulded 227 head, top of pin visible. Moulded foot with transverse ribs, sub-rectangular catchplate on reverse. Leaded gunmetal? SF 319, Context 2221, Intervention 2221 (Artefact). Similar to above, but with zoomorphic lugs T228 126. Incomplete. Central lozenge enamelled now corroded light green, divided into four quadrants set with three white dots each. Projection at bottom of lozenge connects to oblong foot decorated with double ringand-dot zoomorphic motif, subrectangular-shaped catchplate on reverse. One perforated lug of hinge on reverse of projecting head survives, pin missing. Leaded copper alloy. SF 9357, Context 6682, Intervention 6682 (Artefact). Early Roman. Symmetrical plate brooches can have a variety of different plate shapes, such as triangles, squares, ovals or, like the brooches listed under T227 and T228, lozenges (eg, Böhme 1972, Taf 24 and 25). A production date in the 1st half of the 2nd century can be assumed for all three brooches (Bayley and Butcher 2004, 176–7; Feugère 1985, 364). Considering the tendency towards larger enamelled fields in the course of the development of enamelled plate brooches, the sequence of 124–6 may also be chronological, although no context dates are available from Springhead to verify this. A division of the enamelled fields similar to Cat No 124 was found on a brooch with shorter lugs on Nor’nour (Hull 1968, 49, fig 19, 143). One vaguely similar to Cat No 126 comes from London (Hattatt 2000, 352, fig 211, 1101), while a brooch of similar shape but more complex multicoloured enamel was found in a phase 2 context of the Harlow temple, dated around AD 200 (France and Gobel 1985, 74, fig 41, 80). Similar brooch shapes are found widely distributed in southern Britain, northern Gaul, western Switzerland, and along the Rhine (cf Böhme 1972, 38; Feugère 1985, 358; Hattatt 2000, 352–3, fig 211–2). Skeuomorph plate brooch in the shape of a shoe sole T275/Feugère 28b2 127. Incomplete. Hinged, two perforated lugs on reverse of heel, pin missing. Pale opaque green enamelled background filled with opaque yellow glass beads, polished to height of background surface. Subrectangular catchplate on reverse of toe end. Leaded brass. SF 20032, Context 17932, Intervention 17648 (Pit), SG 300660 (Pits), Property 3. (Pl 7). Settling the Ebbsfleet Valley 228 Plate 7 Skeuomorph Plate brooch in the shape of a shoe sole (Cat No 127), leaded brass with enamel L 40 mm. Photo: E Brook Shoe sole shaped plate brooches are widely distributed in central and northern France, western Switzerland and the Upper Rhine valley as well as central southern Britain and Nor’nour (Feugère 1985, 376, fig 57). Those with a loop or a small disc at the heel (Feugère 28b1) are more common than the variant without, represented by Cat No 127 (Feugère 1985, 374). At Augst, the unenamelled variant Feugère 28b3 could be as early as Neronian–early Flavian (ibid, 200), while the date range of the enamelled species covers the 2nd and early 3rd centuries (Riha 1979, 203; 1994, 172); British finds fall well within this period (Mackreth 1996, 322). The eight shoe sole brooches from Nor’nour – all with heel loops or discs (Hull 1968, 59, fig 216–23) – have been interpreted as being votive gifts of unknown attribution (Butcher in Mackreth 1989, 192; see also the reassessment of Nor’nour as a shrine by Butcher et al 2000–1), and in Britain most have been found in civilian settlements. In a recent article Crummy has suggested that this brooch type is one of a number of types, including cockerel, fly, and purse brooches, which may have been associated with the cult of Mercury (Crummy 2007, 226–7). Zoomorphic plate brooch in the shape of a hare T211/ Feugère 29a14a 128. Complete. Zoomorphic plate brooch in the form of hare moving to right. Body enamelled, now opaque green with three dots of the reserved base metal along mid-line, eye almost translucent blue around central dot. Traces of white metal in a band around main enamel field and possibly also on the rear leg. On reverse bent catchplate behind head, double-lug hinge holding short spring and pin at tail end. Two spring coils. Leaded brass. SF 502, Context 2732, Intervention 2732 (Surface), SG 300159 (Deposit). Mid-Roman. (Pl 8). Hare brooches have been sub-divided by Feugère (1985, 383) into three varieties depending on the treatment of the body, type 29a14a with large enamelled Plate 8 Plate brooch in the shape of a hare (Cat No 128), leaded brass with translucent blue and green enamel and remains of white metal L 28 mm. Photo: E Wakefield panels being the most numerous (ibid, 406–7). Based on the variety of technological details noted among the zoomorphic plate brooches, he suggested a distinction of types attributable to at least three different workshops, type 29a14a belonging to ‘Atelier B’ which produced stylised outlines filled with large enamelled panels (ibid, 388). No location for the workshop was proposed, and it is not decided whether all of these brooches come from one workshop in Gaul or whether some may in fact have been produced in Britain where the use of a short spring between two lugs, as found on Cat No 128, is more common (Bayley and Butcher 2004, 174). Parallels for Cat No 128 – with small dots in the enamelled field – have been found at Thistleton, Lincolnshire, Bury St Edmunds, Suffolk, in Norfolk (Hattatt 2000, 362, fig 221, 162.614–5), and Lincoln (Brailsford 1964, 23, fig 11, 43); further analysis is needed to show whether the dots are glass beads set in the base enamel or reserved metal as is the case with the Springhead hare. Considering the development of enamel decoration discussed above, brooches with large enamelled panels like Feugère 29a14a are assumed to start in the Antonine period and continue into the 3rd century (Feugère 1985, 394); one from Augst was found in a context dated mid-3rd century and later (Riha 1979, Taf 67, 1737). Cat No 128 was found outside beam-slot building 300157 in a cobbled surface with a pottery date ranging AD 50–200. The location would not stand against a votive deposition, perhaps in relation to a wish for love or fertility, although other attributes associated with the hare in antiquity included cleverness but also cowardice (Dummer 1987; Smith 2006, 50; Crummy 2007, 228). A hare is also depicted on the nicolo gem of silver finger-ring SF -542 from property 8. At least two other zoomorphic brooches were found during earlier excavations at Springhead: a brooch in the shape of a sitting duck was found in building B.10 (Penn 1968a, 184–5, pl 2B), and one in the shape of a stag similar to Feugère 29a12c was found in 1957 (Penn 1957, 81, fig 14, 4). Oval flat brooches with broad outer band of enamel T260 129. Incomplete. Front separated into two recessed areas by broad band of reserved metal, central enamel possibly of reddish colour, pale yellow–greenish corrosion of enamel remains in outer band, with patches of lighter red on both long sides. Perforated lug on reverse holds axial bar and part of spring, the chord appears to be extended to form a loop, pin broken. Incomplete transverse catchplate with punched zigzag decoration on its underside. Leaded copper alloy. SF 20016, Context 17759, Intervention 17759 (Layer), Property 3. Mid-Roman. The outer band of enamel on Cat No 129 may have contained two colours like a similar brooch from Nettleton, Wiltshire (Wedlake 1982, 129, fig 54, 64), while the centre may have been enamelled or inlaid with a repoussé sheet or contained a separately made copper alloy mount with an intaglio (Mackreth 1996, 321). Chapter 3 Springhead Metalwork T260 is a British type, and parallels are known mainly from southern and eastern Britain, only a few from forts on Hadrian’s Wall (cf list in Bayley and Butcher 2004, 261, additions in Mackreth 1996, 321). While the remainder of the series has the catchplate aligned along the central axis, Cat No 129 is unique in having a transverse one. This type of catchplate is rare in British brooches but commonly found in other types like knee brooches T 176 (Cat No 104). However, there are British types with such catchplates, for instance the 1st century plate brooches T245 (eg, Cat No 109), and two enamelled knee brooches were recently found with inscriptions showing they were made in the Castleford area (Bayley 2005). The chord of the Springhead brooch, extended to form a loop, is equally unusual but may perhaps have served to suspend a necklace or something else from it. The date range for the type is mid-2nd to 3rd century (Mackreth 1996, 321; Bayley and Butcher 2004, 178), one from a context dated AD 400 and later from Richborough is certainly residual (ibid, 134, fig 100, 386). The context pottery date for the Springhead brooch fits well within the suggested range, while the latest coin dates to AD 367. Plate 9 Plate brooch with a late 3rd century coin as plate (Cat No 132), copper alloy L 24 mm. Photo: E Wakefield the 4th century, like gilding or stamped SSS-decoration. The three brooches from Springhead do not add anything new to the debate; a fourth brooch was found south of Watling Street in a late 3rd/4th century layer of temple I during the earlier excavations (Hull 1959, 19 fig 9, 1). Unclassified plate brooches 132. Oval gilt brooches with stone, glass or intaglio in centre T271 Complete. Oval disc with conical central glass setting surrounded by two zones of gilded punched dot decoration, separated by raised rib. Reverse tinned. Spring attached to perforated lug on reverse of head. Sub-rectangular catchplate. Four spring coils. Brass. SF 876, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. (Fig 100) 131. Incomplete. Fire-gilded copper alloy? sheet applied to front of oval-shaped plate. Base tinned on reverse. Three raised bands of decoration around central oval setting: middle band of zigzag lines, outer two imitate twisted chord. Central ?intaglio/stone/glass damaged and partly missing. Spring partly intact, attached to perforated lug on reverse with axial bar, pin missing. Semi-circular catchplate. Four spring coils. Brass. SF 1466, Context 0. See also SF 15536. 130. Plate brooch made from a coin. Lugs for spring and catchplate appear in x-radiograph to be cast as part of a strip to which coin is attached (soldered?). Very corroded. Numismatic description by N Cooke: ‘show side is reverse of antoninianus of the ‘Mars Victor’ type, depicting Mars helmeted, walking, holding a spear in his right hand and a trophy slung over his shoulder in his left hand.’ SF 20394, Context 16889, Intervention 16902 (Pit), SG 300579 (Pits), Property 4. MidRoman. (Pl 9). If the interpretation of the not very conclusive xradiograph is correct, this object belongs to a group of coin brooches and pendants found from the time of Nero and continuing in various shapes and borders into the Carolingian period (cf Berghaus 1994, 106–13, Abb 68–71). The reverse of the coin forming the plate of Cat No 132 is common from Probus to Diocletian, and without the obverse the date cannot be narrowed down more closely than late 3rd century; as it does look slightly ‘irregular’ it may be one of the copies struck between c 275 and 296 (N Cooke, pers comm). The pottery date range of the context is AD 50–300, which in this instance would support the numismatic date of Cat No 132. The context belongs to the upper fills of pit 16902 which was initially dug in the early 2nd century. 133. T271 is the later development of the preceding type, and it is also known in a round variant (T270). Differences in the metal used for the base plate of the two types have been suggested to indicate different workshops (Bayley and Butcher 2004, 179). The oval variant is particularly common in southern and eastern Britain, with some from Hadrian’s Wall, Yorkshire, and the East Midlands (ibid, 178–9; Mackreth 1996, 321; Hattatt 2000, 364–5, fig 223–4). A small number have also been found along the limes and in Germania magna (Böhme 1972, 68–9, Fundliste 42). The earlier tendency to date T270 and T271 to the 4th century has been refuted by Mackreth (1995, 977–9; 1996, 321), who suggested a floruit in the mid-2nd and 3rd centuries. However, Bayley and Butcher (2004, 179) point out that there are some technological details more common in 229 Incomplete. Symmetrical plate type. Generally lozengeshaped, but central opposing sides formed as open circles with beaded decoration on inner edge; each has roundels with ring-and-dot decoration above and below; only half of circle on right survives. Roundel at head missing, that at the foot intact with traces of enamel remaining. Central panel contains blue enamel (fragmentary towards centre), originally polished. Hinge of two perforated lugs on reverse of head, large triangular catchplate, bent. Leaded brass. SF 9258, Context 400104, Channel fills. A brooch found in a mid-2nd century or later context in the vicus of Castleford is comparable but not similar to this brooch. It has two lozenge-shaped units joined by lateral semi-circular bands with three lugs each (Cool 1998, 52, fig 14, 111). In fact, Cool links the Castleford Settling the Ebbsfleet Valley 230 131 132 Surviving tinning Drawn from X-Ray Gilding Enamelling 134 133 135 137 140 141 0 50mm 143 136 139 138 142 146 145 144 Figure 100 Springhead: copper alloy brooches (131–5), pins (136–44), armillae (145–6) Chapter 3 Springhead Metalwork brooch to Riha’s wheel-shaped enamelled brooches (Riha 1979, Taf 65, 1695) which are rare both in Britain and the Rhineland (Cool and Philo 1998, 33). Even though the enamel in the Springhead brooch is not too well preserved its rather simple use of enamel with just one unicoloured – although polished – blue field would suggest a slightly earlier date, perhaps in the first half of the 2nd century, than the more complicated brooches referred to above. 134. Incomplete. Possibly half of a ?symmetrical plate brooch. Double lobe at one end, two recessed areas, one contains red enamel, the other a greenish residue, on reverse of lobe is triangular lug (possibly remnants of a catchplate). Lobe attached to triangular panel with traces of blue enamel remaining. Openwork section outlined by concave bars and narrow central bar connects above mentioned part to central narrow rectangular panel/strip decorated with 3 beaded ridges. It is possible that the above pattern was repeated in reverse for symmetrical side of object/brooch. SF 15727, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. No immediate parallel has been found for Cat No 134, although it is likely that it was part of one of the more complex enamelled plate brooches of the later 2nd/early 3rd century, like examples from Lanslevillard, France (Feugère 1985, pl 152, 1908), Zugmantel, Germany (Böhme 1972, Taf 24, 927.942), or Augst, Switzerland (Riha 1994, Taf 44, 2884). Penannular brooches 135. Complete. Round-sectioned ring (leaded bronze/ gunmetal), ends flattened and coiled back at rightangles to the ring. Surface corrosion obscures any decoration. Pin (bronze/gunmetal) oval-sectioned, tapers to point, flattened where rolled around ring. Bronze/gunmetal. SF 737, Context 5637, Intervention 5637 (Layer), SG 300104 (Deposits). Early Roman. See also SFs 552, 630, and 1704. SF 9340 may be an annular brooch or simply a very bent buckle. Only four penannular brooches have been recorded at Springhead; they are confined to the area to the east and south-east of the Ebbsfleet (Fig 101). A fifth brooch was found during the earlier excavations (Penn 1957, 81, fig 14, 6). All have a more or less circular-sectioned ring with the ends flattened and coiled back onto the ring at right-angles, thus belonging to Hull Type P3 or Fowler Type C. All were found in early Roman contexts dated to the decades around AD 100. The type is commonly dated to the 1st century AD in Britain where it is mainly found in southern and south-eastern England (Fowler 1960, 164–6; Crummy 1983, 18; Bayley and Butcher 2004, 136, fig 102, esp 400–5). On the Continent, however, such simple brooches are also found in later contexts, such as those from VireuxMolhain, France, grave 8 (Lemant 1985, fig 12) or Tongeren, Belgium, graves 111 (gilded bronze) and 283 231 (Vanvinckenroye 1984, 190), which date to the middle of the 4th century. Generally, the later Fowler Type C brooches more commonly have a flattened ring, often with decoration on the ring (cf Fowler in Crummy 1983, 19; Schuster 2006, 51). Fragments of brooches – springs, pins, and feet Twenty-four fragments of brooches were found, including five heads with parts of the spring surviving, nine springs with varying length of the pin, and two pins of which one belongs to a hinged brooch. All these are too corroded to allow identification to type. Of the eight feet included here, seven are likely to belong to either one- or two-piece Colchester brooches, while another foot (SF 1797) has a triangular catchplate with a round hole near the inside edge and a flat foot which flares out the end. Metal Pins by Elina Brook with Jörn Schuster The typology used for the description and discussion of the pins found at Springhead follows that devised by Cool (1990). Eighty objects have been recorded as metal pins certainly or probably belonging to the Roman period. All are copper alloy, apart from five iron pins from the Roadside settlement (ARC SHN02) of which only one is certainly a pin (SF 15126). Only 58 metal pins could be identified to type, the remainder are unidentifiable shaft fragments that could equally have been part of needles or brooches. SF 15912 could have been part of a belt. The worked bone pins are described below by Allen (Chapter 13). Group 1 136. Incomplete. Hemispherical upper part of head, conical lower part, possible groove decoration (square pattern?) on top. Circular-sectioned shaft with possible groove just below head, tapers slightly towards broken tip. Slightly bent. Crummy Type 3. SF 354, Context 2222, Intervention 2222 (Artefact). See also SFs 249, 844, 979, 1842, 15956, 18004. One pin, SF 1842, could be attributed to this group, possibly of sub-group C. A further six are included here as they display stylistic traits best paralleled in this group (Cat No 136 and SFs 249, 844, 979, 15956, and 18004). Three have double conical heads like Cool 1990, fig 1.2; SFs 249 and 18004 have possible decoration on top but are corroded. Cat No 136 has a spherical upper part and a conical lower part (?G1) and possible decoration which may be four grooved lines in a square pattern on top of the head. SFs 844 and 979 are possibly of sub-group D (if conical lower part is of G1). SF 15956 has a slightly spherical head but is severely corroded so a closer identification is not possible. This group is a broad category appearing across Britain throughout the Roman period but especially during the 2nd–4th centuries (Cool 1990, 151). Settling the Ebbsfleet Valley Group 3 137. Incomplete. Almost conical head (wider than shaft) with double cordon below and finial knob above, circular-sectioned shaft, tapers slightly, bent and broken. SF 20417, Context 19593, Intervention 19592 (Pit). Early Roman. See also SFs 785, 1509, 15029, 15123, 15159, 15928, 18020, 18254. 561700 Nine pins can broadly be attributed to this type. SFs 785, 1509, 15123, and 15159 are very similar to one another: they relate to Group 3 in that they have a curved unit below a cordon on their head (although not a bulging one), and to sub-group A as the heads are narrower than their shafts; however, the decoration has most likely been cast rather than cut into the top of the shaft as Cool (1990, 154) suggests for this sub-group. So far no direct comparisons have been found elsewhere. SFs 15029, 15928, and 18254 are similar but have variations on the number of cordons below the finial knob. Cat No 137 and SF 18020 possibly belong to Penannular Brooches sub-group B as the features on the heads are wider than their shafts. The group is thought to be early Roman (1st/2nd century) and is found across Britain. The dating is supported by the pottery context dates from Springhead, which cluster around the later 1st and early 2nd centuries. The original lengths of SF 785 of 106 mm, 112 mm for SF 15159, and approx 110 mm for SF 18020 also indicate that they are possibly of this date (Cool 1990, 173). Group 5 138. Complete. Three bands of cordon decoration separated by one cylindrical barrel and one shorter squaresectioned block, both have repeated incised cross decoration. Circular-sectioned shaft tapers towards tip. SF 1563, Context 3547, Intervention 3546 (Pit), SG 300245 (Pits). Early Roman. Cat No 138 is the only example of this group, found to the east of the Ebbsfleet. It is complete (length 104 mm) and possibly belongs to sub-group C as it has cross-hatched decoration around its head. It also has elements of sub-group D with multiple horizontal grooves/cordons dividing the two cross hatched panels. Group 5 is thought to have been most common during the 2nd century and is quite widespread in its 561900 Miscellaneous SF 15524 is possibly a pin or a nail. It has a globular head that may have been wound around the shaft – or this may be due to corrosion making it appear to be the case. If it is a pin it would be part of Group 1. 561800 232 er Riv Eb t lee bsf 172800 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 101 Springhead: distribution of Penannular brooches Chapter 3 Springhead Metalwork 233 distribution in Britain (Cool 1990, 157). The diagonal crosses on the blocks also relate it to Group 11 subgroup B which is found in the Midlands (ibid, 164). which is mainly distributed on the north Kent coast, where these pins were in use in the 1st half of the 2nd century (Cool 1990, 164). Group 6 Group 12 139. 142. Complete. Oval-sectioned flattened spherical head. Three grooves directly below head form double cordon. Circular-sectioned shaft slightly facetted towards top, tapers slightly towards tip. SF 903, Context 6163, Intervention 6163 (Layer), SG 300087 (Deposits). Early Roman. See also SF 499. The two pins attributed to this group, equivalent to Crummy type 5 (Crummy 1983, 30), are both from the Sanctuary site. SF 499 has a smaller head which Cool suggested to be a later variant of the type (Cool 1990, 157). Comparisons were found at Gadebridge (c Neal 1974, 144, fig 64, 224) and Verulamium (Stead and Rigby 1989, 21, fig 13, 73), dating mid–late 2nd century/before AD 218. Cat No 139 has a larger head and is complete, with an average length of 96 mm (cf Neal 1974, 144, fig 64, 224). This type is mainly found in eastern Britain. Group 9 140. Complete. Head with double notched, perforated disc finial above rectangular-shaped moulding, this sits above curved barrel with possibly decorated ?cylinder below. Shaft of circular section, very slightly faceted and bent. SF 888, Context 5921, Intervention 5921 (Layer), SG 300104 (Deposits). Mid-Roman. See also SF 18306. The two pins of this group are almost identical, although the disc finial at the top of the head of SF 18306 is broken. Both pins fit in with the southeasterly distribution within Britain. The scant dating evidence known so far suggests that the type was in use by AD 125 (Cool 1990, 160). Group 10 One incomplete example (SF 1252) was found at the Sanctuary site. It belongs to sub-group A with a groove around the circumference on the lower part of its conical head, in use by AD 125 and more commonly found in the east of Britain (Cool 1990, 160). Group 11 141. Incomplete. Multiple block head, lower block largest, of circular section, possibly with diagonal groove or crosshatch decoration; central block is the smallest, of circular section with double horizontal groove decoration; domed finial. Shaft tapers slightly, broken. Leaded copper alloy. SF 18032, Context 17210, Intervention 17210 (Layer), SG 300654 (Layers), Property 3. Mid-Roman. This pin with its multiple block head and crosshatched and horizontal grooves belongs to Group 11A Almost complete, only very tip is missing, original length slightly longer than 75 mm. Upper part of head spherical, lower part conical, circular cross section. Horizontal groove around widest part of head, on top four pairs of grooved lines radiate forming cross shape. Circular-sectioned shaft, bent at right-angle. SF 1560, Context 3570, Intervention 3570 (Layer). See also SFs 435, 497, 15120, 15201, 15317, 18312, 18364. Eight pins can be attributed or related to this group whose main characteristic is the grooved cross pattern formed by pairs of lines on top of the head. Four are directly comparable to those within Cool’s typology (Cat No 142 and SFs 497, 15201, and 15317), whilst a further four have similar features (SFs 435, 15120, 18312, and 18364). Of the four that certainly belong to this group Cat No 142 and SF 15317 both have a horizontal groove that runs around the widest part of the head. However, they are not identical as the head of SF 15317 is slightly wider, and it is longer than the almost complete Cat No 142. SFs 497 and 15201 have the cross pattern but no additional horizontal groove. All four have a more spherical-shaped upper part of the head. Of the four pins related to this group a further two also have a spherical upper part of the head: SF 435 is fairly corroded but the cross pattern on top is visible, the difference is that there is a possible groove on the underside of the head as opposed to around the widest part. SF 18312 again has a more spherical head, but the decoration on top consists of five grooved lines radiating from the centre (as opposed to the cross pattern) with multiple horizontal grooves on the lower part of the head – this is similar in form to a Group 8 example (Cool 1990, 159, fig 6, 4) but lacks the lower upturned hemisphere (for a good parallel from Colchester see Crummy 1983, 31, fig 31, 500). The decision to relate it to Group 12 was based on the grooved decoration on its head. The final two pins related to this group are SFs 15120 and 18364. Both have a double conicalshaped head, but the decoration varies. SF 15120 has a cross pattern on top, but this is made of single lines as opposed to pairs of lines; pairs of grooved lines do occur but are located around the edge of the upper part of the head creating an almost zigzag effect, there is also a single horizontal groove on the lower part of the head (as with SF 435 above). This pin is complete, measuring 104 mm, suggesting it dates to the 1st/2nd century (Cool 1990, 173–4). SF 18364 has the cross pattern attributable to Group 12 but with additional shorter grooves in between extending to the edges of the upper part of the head, creating an almost notched effect, as well as the horizontal groove below. Settling the Ebbsfleet Valley 234 Table 49 Springhead metal finds: frequency of pin groups (after Cool 1990) Pin group ARC SHN02 ARC SPH00 Total similar plain form (Crummy type 1) for which she suggests a manufacturing date of AD 50–200 (Crummy 1983, 28). 1 2 6.1% 5 20.0% 7 Group 25 3 7 21.2% 2 8.0% 9 144. 5 0.0% 1 4.0% 1 6 0.0% 2 8.0% 2 3.0% 1 4.0% 2 0.0% 1 4.0% 1 0.0% 1 9 1 10A 11 1 3.0% 12 5 15.2% 3 12.0% 8 24 14 42.4% 8 32.0% 22 25 3 9.1% 2 8.0% 5 Total 32 100% 25 100% 58 The only dating available for the group is from a Walbrook deposit in London, again suggesting that it was in use by AD 125. With four pins from early Roman layers and only one from a middle Roman context, this general date range is also supported by the evidence from Springhead. The distribution is within north Kent and London (Cool 1990, 164). Group 24 143. Almost complete. Simple pin with domed, slightly pointed finial. Cross-section oval at head, more circular along shaft, tapers towards tip. Slightly facetted. Broken at tip, bent. SF 633, Context 0. See also SFs 294, 421, 506, 641, 1586, 9140, 9174, 15065, 15101, 15194, 15126, 15752, 15840, 15914, 18026, 18703, 18916, 18945, 20001, 20071, 20519. This is by far the most common group at Springhead with a combined total of 22 pins. It is a general category for a very simple form of pin and the Springhead examples have either a conical or slightly domed head. The only identifiable iron pin, SF 15126, belongs here, too. Only five are complete (SFs 294, 15194, 15126, 18026, and 18916), and three of these are over 120 mm long which may suggest they are of 1st/2nd century date (Cool 1990, 173–4). Four objects from this group have noticeably flatter domed heads (SFs 18703, 18916, 20001, and 20071). SFs 1586 and 15840 have flat heads and are possibly not hairpins but perhaps simple probes. The remainder are all quite similar, with more conicalshaped heads, some of which are very slightly facetted. Cool (1990, 170) suggests that this form may have been used throughout the Roman period but when compared with groups 3A and 5 (of simple form but with decoration cut into the head) it is possible to say that they may have been more common during the 2nd century. This seems to be supported at Springhead where three pins were found in early Roman layers, ten in mid-Roman, and only one in a late Roman layer. Crummy notes the similarity to bone hairpins of a Incomplete. Biconical head, oval cross-section. On upper part of head a band of grooved lines radiate from below groove around circumference, further groove around circumference on lower part of head. Circularsectioned shaft, tapers slightly, broken. SF 1706, Context 6380, Intervention 1003 (Artefact). See also SFs 9347, 15116, 15424, 15987. This group is a general category for those pins that have grooved decoration on their heads but which cannot be directly placed into any of the groups within the Cool typology. There are five from Springhead. Cat No 144 and SF 15424 are very similar to each other as they both have a double conical head with a band of grooved lines on the upper part that radiate from the centre below two horizontal grooved lines which in turn create an almost finial knob on top of the head. On the underside of the head is a further horizontal grooved line. The conical shape of the head is reminiscent of groups 10 and 12 which were both in use by AD 125, but the nature of decoration is different, and comparisons for this decoration have not been found so far. SF 9347 also has a double conical head (as with groups 10 and 12) with multiple diagonal groove decoration on the underside of the head and a beaded edge; a further difference is that the shaft is also decorated with a double cordon at its top and bulges towards the centre. The combination of cordon and double conical head is reminiscent of a Saxon type from middle Saxon Hamwic (type Ca2ii; cf Hinton and Parsons 1996, 27, fig 10, 23/2) which also has a slightly swelling shaft, but that example only has a single cordon and is undecorated. As this pin is a metal-detector find from the spring area, it cannot be securely dated. SF 15116 and 15987 both have more sphericalshaped heads. The first has vertical groove decoration with a single horizontal groove towards the base of the head, similar to Cool Group 13 (Cool 1990, 162, fig 8, 6 and 7), but the form is different and less elaborate. Similar pins have been found at Colchester (Crummy 1983, 31, fig 31, 499) and possibly Wanborough (Hooley 2001, 101 fig 40, 148), although neither matches very closely. The pottery context dates for the pin is late 1st-/mid-2nd century. SF 15987 is almost the same but lacks the horizontal groove at the base of the head. The Colchester example previously referenced is a close parallel although it has additional grooves at the top of the shaft. Wrythen-decorated spherical heads have been found in Saxon deposits at Hamwic (Hinton and Parsons 1996, 16, fig 7, type Ab1ii and Ab2ii), but these more commonly have swelling shafts and cordons at the top of the shaft. Such features are not present on the Springhead pin which was found in a pit in property 4 Chapter 3 Springhead Metalwork Only 18 bracelets or fragments of such have been identified among the metal finds, but it may well be that some unidentified fragments of bracelets can be found among the fragments of wire (eg, SF 15300). Another possible bracelet fragment, a sinuous wire with eight loops (SF 15994a), was found together with a thin copper alloy strip in early Roman post-hole 17012 in property 5. Similar patterns are known from early to middle Iron Age pectoral jewellery in Lower Saxony, Germany, for example a hoard from Issendorf (Häßler 1995, 128, Abb 4) and a cremation grave at Erichshagen-Wölpe near Nienburg/Weser (Häßler 2002, Pins ( = Cu alloy, = Cu alloy, type not identified, =Fe) 561900 Bracelets 223, Abb 116). Although the fragmentary preservation precludes any certainty in identifying its use, a more likely explanation for the Springhead find may be provided by an extendable armlet dredged from the Waal between Rossem and Tiel in the Netherlands, for which a 1st century AD date has been suggested (Sas and Thoen 2002, 175 no 94; R Jackson, pers comm). If this dating is correct, it is not unreasonable to assume a line of tradition to similar bracelets of the late 4th–early 3rd century BC like some of those found in the large spring deposit at Duchcov, Chech Republic (Berger 1882, Tab 4, 55–6) or in graves of similar date at, for instance, Epernay or Bussy-le-Chateau, both Marne, France (Charpy 1991, 245; 247). With 12 examples (plus two possible fragments), SF 15746 and one from an earlier excavation (Boyle in Boyle and Early 1999, 27 fig 17, 4), the most common form of bracelet is represented by group a. A recent reconsideration of these flat, penannular metal bands by Crummy (2005a) concludes that they are most likely a form of military award, an armilla. They were usually worn as a pair suspended from a band around the neck and awarded only to soldiers who were Roman citizens (Crummy 2005a, 100). As the present author only became aware of this study (F Pemberton, pers comm) after the catalogue and figure order were finalised, the armillae are here grouped with the bracelets, but in the 561800 with a pottery date range of 50–300. Table 49 shows the frequency of those pins that could be assigned to groups within Cool’s typology. It does not take into account the fragments or the uncertain SF 15524. Slightly more pins are found on the western than on the eastern side of the spring (Fig 102). In the Sanctuary they cluster near the temple 400035 but, interestingly, only two have been found in the spring itself, a marked difference compared to the many brooches recovered from there. In the Roadside settlement pins are mainly found in properties 3, 4, and 11. 235 er Riv Eb t lee bsf 172800 172700 A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 102 Springhead: distribution of pins 236 Settling the Ebbsfleet Valley tables of finds categories they are grouped under military equipment. In her discussion Crummy (2005a, 95–6, fig 3) distinguished four groups (A–D) of armillae, depending on the number of wreaths or textured bands. Group A has two bands, usually set towards the middle. This is the most common group, also borne out at Springhead, where there are seven examples (Cat No 145 and SFs 9199, 15192, 18725, 18868, 20161, -540). Group B has one central band (Cat No 146 and SFs 15403, 15530) and group C three, set symmetrically (SF 18726), while group D with three bands set asymmetrically was not found at Springhead. Armillae of the form under discussion here are mainly confined to the east of England, with only a small number found as far west as Gloucestershire and Somerset; only one was found previously in Kent, at Richborough, and no closely similar armillae have so far been found on the Continent (ibid, 94, fig 2; 98). The design is related to that found on Aucissa and Hod Hill brooches as well as military fittings worn by the invading army (ibid, 96). Based on this evidence, Crummy concludes that the armillae are awards specific to the early years of the Roman Conquest of Britain (ibid, 98). Of the 12 examples from Springhead, eight were found to the west of the Ebbsfleet, two with pottery date ranges of AD 50–150 and two with coin termini post quos of AD 117 and 134 respectively, while five others have coin dates in the later 4th century suggesting a high degree of residuality. a) Flat, wide penannular strips with longitudinal grooves and moulding – armillae Crummy group A. Fragment. Slightly curved band with flat, rectangular cross-section. Wide central groove runs along length of hoop, four further incised lines with dotted decoration run parallel (two either side), originally ridge with transverse S-lines. One end broken, the other has rounded edges with transverse groove and dotted decoration. Very worn. SF 326, Context 2221, Intervention 2221 (Artefact). 146. Crummy group B. Incomplete. Rectangular-sectioned. Punched cross-hatched decoration down centre, two grooved lines either side, some transverse grooves close to one end. Broken both ends. SF 9260, Context 400104, Channel fills. See also SFs 327, 9199, 15192, 15403, 15530, 18725, 18726, 18868, 20161, -540 (possibly also 15746). 145. b) Bracelets with transverse groove decoration and D- or oval-shaped cross-section (Fig 103) 147. Fragment. Penannular; D-shaped cross section. One end is the decorated terminal with four sets of doublebeaded ridges and a blunt end. Internal edge has groove along it, possibly from the manufacturing process. SF 15681, Context 16641, Intervention 16655 (Ditch), SG 300545 (Ditch), Roadside ditch 3. Early Roman. See also SFs 793, 473, 15822, 15730, 15915. A bracelet from Shepton Mallet is similar to Cat No 147 in that it has two registers of cordoned decoration but there the single ridges are not beaded (Smith in Leach with Evans 2001, 203, fig 55, 29). The type is generally rare and most numerous in the 3rd and 4th centuries. c) With snakeshead terminals 148. Incomplete. Penannular; D-shaped cross section, flattens to rectangular-sectioned terminal. Five transverse grooved lines and triple ring-and-dot motif decorate terminal, broken at other end. SF 9261, Context 400104, Channel fills. This bracelet belongs to Swift’s late Roman ‘snakeshead bracelets with type 1 terminals’ which are found in southern Britain and a restricted area in the Netherlands and Belgium south of the Rhine (Swift 2000, 153, 169, fig 215, 177). The fragment of a silver wire bracelet with a snakeshead terminal (SF 15133) was found during metal-detecting on the Roadside Settlement site. The design of the head recalls late Roman copper alloy bracelets with similar terminals known from Cottenham and Stonea in Cambridgeshire (Ireland in Taylor 1985, 12, fig 5, 53–5). However, better parallels exist in a group which may be of Romano-British origin including the pair of silver bracelets from the Castlethorpe hoard dated to the middle of the 2nd century (Cool 1979, 166, fig 1B; 168), but both these bracelets have a wider, flatter band. The possibility that the Springhead piece may have been part of a 2nd century finger-ring should not be discounted (eg, Ditton: Cool 1979, 167, fig 2C; Verulamium: Stead and Rigby 1989, 21, fig 13, 44; Guiraud 1989, 195, type 7), although the curvature of the hoop would fit better with a bracelet. d) Narrow, with crenellated outer edge 149. Incomplete. Half of hoop remains. Narrow rectangular cross section, crenellated outer edge. Thinner one end, very worn. SF 9141, Context 6379, Intervention 1001 (Spring), SG 300015 (Deposits). Early Roman. ‘Cogwheel’ bracelets of the late 3rd–early 5th century are more or less confined to Britain, south of a line between the Wash and the Severn estuary, with only one outlier in grave 78 at Oudenburg in West Flanders (Swift 2000, 127, 136, fig 163, 160; Crummy 1983, 41, fig 43, 1659). Interestingly, that grave also contained a snakeshead bracelet of the same type as Cat No 148 and one with multiple motifs (Swift b12). Based on this rare combination which, apart from the snakeshead bracelet, is replicated at Portchester, Swift takes this as a ‘rare example of personal ornaments enabling a precise movement from one area to another to be traced’, in this case from Britain to the Continent (ibid, 178–9). Whether this journey led past the Ebbsfleet at Springhead remains open to speculation. Chapter 3 Springhead Metalwork 237 148 147 149 150a 151 152 153 150b 150c 154 (gold) 155 158 156 (silver) Enamelling Stone 0 50mm 157 162 159 Glass Gold 160 161 163 164 Cornelian Figure 103 Springhead: bracelets (147–50), ear-ring (151), finger-rings (152–63), gemstone set in gold collet (164). Copper alloy, unless indicated otherwise 238 Settling the Ebbsfleet Valley Ear-rings 151. Plate 10 Gold finger-ring (Cat No 154) Diam 18 mm. Photo: E Brook Possible ear-ring fragment. Square-sectioned, slightly twisted, bent in U-shape, tight bend/hook at one end, the other tapering to thin tip. SF 9365, Context 6682, Intervention 6682 (Artefact). Early Roman. Only four possible ear-rings were identified among the metal finds from Springhead. One is a fragment of a loop of two twisted wires (SF 20531), which, if it is an ear-ring, corresponds to Allason-Jones type 5, known throughout the Roman period and distributed evenly throughout Britain, although less common in the eastern counties (Allason-Jones 1989, 7, map 12). A similar ring (SF 1788) has a loop of three twisted copper alloy wires like Allason-Jones type 6. According to Allason-Jones (ibid, 7, map 13) these ear-rings are more commonly found on civilian than on military sites, mainly along a corridor between Essex and the Severn Valley. Their date range is biased towards 4th century contexts; however, this does not seem to apply in this case where the ear-ring was found in the build up next to an early Roman trackway. The other two ear-rings (Cat No 151 and SF 18023) are of the simple type 1, with square and D-shaped section respectively. The type has an even distribution throughout England with a slight clustering around the Bristol Channel. The simplicity of the form precludes any tighter dating within the Roman period as the type is also known before and after (ibid, 2; map 1). Finger-rings Plate 11 Top: Intaglio showing two eagles on globes holding a garland between them (Cat No 164), red carnelian with gold collet L (with collet) 14 mm. Bottom: Silver finger-ring with nicolo intaglio showing ?hare facing right (SF -542; nicolo: L 88 mm). Photo: E Wakefield e) Possible bracelets 150. Incomplete fragments, possibly from same object but do not clearly join: a) square-sectioned strip, gently curved, broken one end, cube-shaped terminal at other decorated with incised crosses (visible on two sides), double cordon separates terminal from shaft (L 55.3 mm, max W 5.5 mm); b) shaft fragment, squaresectioned at ends, flattens to rectangular section in middle, broken both ends (L 72.7 mm, max W 3.8 mm); c) square-sectioned shaft, broken both ends, bent and twisted (L 53.7 mm, max W 3.6 mm). SF 911, Context 6023, Intervention 6023 (Layer), SG 300114 (Deposits). Early Roman. See also SF 1610. Twenty-four rings have been identified as finger-rings, but there may be more among the plain copper alloy rings classed as fittings in the database. Of note is the golden ring Cat No 154 (Pl 10), as it might indicate an elevated social status of its owner; in theory at least gold was only allowed to be used for rings worn by persons of senatorial and equestrian (partially) status. During the reign of Tiberius the right was granted to persons of freeborn ancestry. It is doubtful, however, whether this law was strictly enforced after the 1st century AD, and in AD 197 soldiers were granted the right to wear gold rings by the emperor Septimius Severus (Guiraud 1989, 174, note 1; Henig, note in Goodburn 1984, 19; Henig 1995, 1000, no 186). The only other precious metal rings are Cat No 156 and SF -542 of silver, the latter still with a nicolo intaglio depicting a hare (Pl 11, bottom); the remainder of identifiable finger-rings are cheaper trinket rings of copper alloy, most having lost their settings. Their distribution is shown in Figure 104. Plain hoop 152. Complete. D-shaped cross section. Very abraded, ?transverse line decoration. One point of hoop worn away more. SF 1287, Context 3325, Intervention 3325 (Artefact), SG 300046 (Ditch). See also SFs 378, 503, 643, 9382 and 15331 Chapter 3 Springhead Metalwork Guiraud Type 7c. Complete. Circular-sectioned wire strip coiled three times. Max W 6.12 mm. (cf Guiraud 1989, 195). SF 9217, Context 6629, Intervention 6621 (Ditch). Late Iron Age. Plain, coiled finger-rings of multiple coils are easy to make and occur in many periods and regions (Guiraud 1989, 195; Schuster 2006, 72). The relatively high number of such rings from mid-1st century AD contexts at Sheepen, Colchester has been suggested by Crummy (1983, 47) as indicating an early date range for the type. This is supported by an example from Harlow found in the Belgic layers while a second was found in the demolition debris above the temple (Gobel in France and Gobel 1985, 84, fig 43, 46–7). Plain hoop with bezel 561700 Guiraud type 2g. Complete. Very worn. Hoop widens at top into flat bezel with lozenge/diamond-shaped empty field indicated by faint lines, sides decorated with further transverse lines – more on one side than the other. Bezel/top decoration confined by two transverse 154. Rings ( = Finger, =? Finger, = Other) In Gaul and the German provinces rings of this type and the similar type 2h are particularly frequent in the north-east and along the Rhine; of those analysed by Guiraud 16.7% were of gold (Guiraud 1989, 184–5, fig 18–9). British examples in copper alloy are known from Wanborough (Hooley 2001, 91, fig 35, 95–6). 155. Complete. Annular hoop. Oval cross section, thicker and wider towards top. Subrectangular raised bezel, no decoration visible. Very worn. SF 1256, Context 3188, Intervention 3187 (Ditch), SG 300030 (Ditched enclosure). Late Iron Age A possible comparison for this very crude ring might come from a late Roman context in Colchester where a white-metal ring with transverse grooves has a small rectangular bezel (Crummy 1983, 50, fig 52, 1790). Plain hoop with setting 156. Guiraud type 4e. Almost complete. Rectangularsectioned hoop widens to flattened bezel, separate oval 561900 Coiled 153. lines on only one side of the hoop, may have worn away on the other side. Gold. SF 20117, Context 16711, Intervention 16711 (Layer), SG 300594 (Layers), Property 4. (Pl 10) See also SF 806 561800 Cat No 152 may be a very abraded example of a ring with transverse grooves (cf Crummy 1983, 48 fig 50, 1770). 239 er Riv Eb t lee bsf 172800 172700 Au Ag A2 Key: Late Iron age Early Roman Mid-Roman Late Roman 0 100m Based upon the Ordnance Survey® Land-Line® digital data with the permission of Ordnance Survey on behalf of The Controller of Her Majesty's Stationary Office, © Crown Copyright 2007. All rights reserved. London & Continental Railways Limited/Union Railways (North) Limited both of, 3rd Floor 183 Eversholt Street, London NW1 1AY Licence No. 100047146. Figure 104 Springhead: distribution of rings 240 157. Settling the Ebbsfleet Valley collar soldered to bezel, setting is missing. Silver alloy. SF 15726, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. Guiraud type 4e var, Riha 2.1.11 (Riha 1990, reference after Bertrand 2003). Incomplete. Rectangular-shaped flattened bezel with part of angled shoulders, most of hoop missing, quite thin, rectangular in section. Small teardrop-shaped black stone in raised setting at centre of bezel. SF 20013, Context 17759, Intervention 17759 (Layer), Property 3. Mid-Roman. In Gaul and the German provinces rings of Guiraud type 4e (37.7% gold) are especially common in the Rhine–Main area and near the mouth of the Rhône; some with inscriptions of Christian content or favouring the emperor date to the 4th century (Guiraud 1989, 189, fig 28; 191). A parallel for Cat No 157 in copper alloy comes from Antigny, France (Bertrand 2003, 46, pl 9, 44). 158. Guiraud type 2a var. Complete. Oval-sectioned annular hoop, widens slightly towards oval-shaped bezel. Remains of enamel or glass paste decoration still visible, now pale green colour. Max W 12.9 mm, min. 3.3 mm. SF 900, Context 6168, Intervention 6168 (Layer), SG 300087 (Deposits). Early Roman. Cat No 158 is unusual in that it has an oval bezel with its axis orthogonal to the axis of the hoop. 159. Incomplete. Rectangular section, widens to bezel which has slight ridge around it, possibly originally decorated or holding a setting, now very worn. Possibly penannular ring (or breaks worn to sharp point). Max W 8.03 mm, thickness 1.02 mm. SF 1657, Context 3845, Intervention 3844 (Ditch), SG 300252 (Ditch). Saxon (ring ditch around grave 3903) Although found in the ring ditch around Saxon grave 3903, this ring would not be out of place in a Roman context, eg, Guiraud types 2a or h (1989, 181), and might well be residual. Even though this part of the Sanctuary site was almost devoid of Romano-British features, a pit of that period was only c 10 m south of the ring ditch. necessarily of Roman date. SF 18713 is the bezel with large setting and missing gem of a ring Guiraud type 2a. Profiled hoop with setting 161. Guiraud type 3a. Complete. Oval-shaped translucent light green glass gem of very crudely cut figure with upper arms at right angle to body and lower arms hanging down perpendicularly, upper legs visible but below this slightly chipped to lower edge. Hoop has Dshaped cross section, widens towards oval bezel with angled shoulders, lower part of hoop slightly bent. SF 15635, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. The shape of the ring is typical of the late 2nd/3rd century (Guiraud 1989, 185), the motif of the cast glass intaglio joins this ring to a group of similar gemstones from Britain south of the Fosse Way, which look like a person holding a stave in each hand, possibly depicting the deity Virtus with spear in the right and pouring a libation with the left hand. The group has been linked with the beginning of signet use amongst the peasantry in Lowland Britain during the 3rd century, the motif possibly derived from the radiate coinage of the time (Henig 1974, 164 with note 18, fig 3, type 4, pl 17, 549–52; Taylor 1985, 11, fig 4,11). 162. Fragment. Oval bezel and adjoining parts of shoulders with transverse mouldings, hoop completely missing. Bezel filled with corroded, now brown, glass paste. SF 9350, Context 6682, Intervention 6682 (Artefact). Early Roman. The shape of Cat No 162 places it among the profiled rings of the 3rd or 4th century, see for example a ring with a cabled hoop from Colchester (Crummy 1983, 48, fig 50, 1785). Profiled hoop with bezel 163. Round bezel with stamped decoration of a head with a ?crested helmet within three concentric rings, the inner two beaded. The hoop to either side of the bezel is of triangular shape with longitudinal cable moulding. Lower part of the hoop missing. SF 9354, Context 400104, Channel fills. Guiraud type 2a. Incomplete. Two fragments join, Dshape sectioned hoop, widens evenly to oval-shaped recessed bezel, stone/intaglio now missing. Loop broken both ends. SF 726, Context 5256, Intervention 5256 (Layer), SG 300145 (Deposits). Mid-Roman. Cat No 163 can loosely be associated with rings Guiraud type 3f which are often of precious metal (eg, Richborough; Bushe-Fox 1949, pl 35, 93) and some set with coins rather than gems (Guiraud 1989, 185, 187, fig 25). The form of the ring suggests a 3rd century date. See also SF 214 with a translucent white glass intaglio, possibly showing a phallus or a very degraded standing figure (eg, like Henig 1974, p. 18, 563). SF -542 (Pl 11) is a silver ring of Guiraud type 2d set with a nicolo intaglio depicting a hare facing right. SF 1350 has a D-shaped sectioned hoop of equal width with a small circular recess for a gem setting (Henig type II ring). It is a metal detector find from the top soil and not Other finger-rings (not illustrated) SF 1669 has an interlocking bezel whose loose ends are wrapped around the ends. This type, Guiraud type 6d, is fairly common in Gaul and the Rhine–Main area throughout the Roman period (Guiraud 1989, 193–4, fig 39), British examples are known, for instance, from Wanborough, Wiltshire (Hooley 2001, 91–2, fig 35, 101) and a small bracelet from a context dated after 330 from 160. Chapter 3 Springhead Metalwork Richborough (Wilson in Cunliffe 1968, 98, pl 41, 156). The type continues into the Saxon period (Walton Rogers 2007, 133, fig 4.21d, left) as exemplified by an example from cremation grave 1465 at Spong Hill, Norfolk, dated to the earlier 7th century (Hills 1977, fig 123). One from Verulamium was found in a medieval layer over building XXVIII and is probably residual from the Roman layers (Goodburn 1984, 30, fig 10, 57). A slightly simpler variant of the type, without the central spiral (similar to Guiraud type 6a), comes from grave 1470 at Spong Hill and is paralleled by SF 528 from grave 2827 in the Saxon cemetery to the east of the Ebbsfleet at Springhead (see Schuster, Vol 4, Chap 3). It is possible that these Saxon rings were parts of necklaces rather than finger-rings. SF 20003 may be a finger-ring with a thick D-shape sectioned hoop and the remains of the casting sprue as a bezel, imitating a jet ring (eg, Caerleon: Brewer 1986b, 145, fig 47, 22; Brewer also refers to a similar ring in gold in the British Museum). It was found to the northeast of the bakery structure in property 3. 241 Drawn from X-Ray Niello Silver ? 166 171 165 Gemstone with a note by Martin Henig 164. Oval cornelian of flat section with oblique sides (Henig 1974, 55, fig 1, Flat 1), approximately one-quarter missing, in plain, oval gold collet with lower rim and only very slight intermittent notch at top to hold intaglio. The top of the stone is broken at an oblique angle from the upper left to the middle right side. The intaglio comprises two eagles, the head and upper body of that on the right missing. The left eagle has its breast facing forward and right, its head facing back to its upper left, standing on/holding a small globe in its talons. The right bird may be similar to the above but its back seems to be facing the viewer; the scene is obscured by various chips and damages on this side. Both birds were holding a garland between them, the stem of which extending to the left of the head of the left eagle. A base line runs from the right bird’s tail to at least below the globe of the left. Cornelian and gold. SF 9267, Context 6682, Intervention 6682 (Artefact). Early Roman. (Pl 11, top) This intaglio with its gold collet was most likely set into a finger-ring, as exemplified by a silver finger-ring from the Marlowe Car Park site, Canterbury, dated to the late 2nd century (Henig 1995, 1002, fig 419, 187). The motif of an eagle on a globe is on occasion found on the reverse of Roman coins, for example on coins for Augustus (RIC 82, C 247 (Tiberius)) or Septimius Severus (BMC 423, 21 (Caracalla); RIC 239, 191c (Caracalla) with the legend ‘CONSECRATIO’). For comparisons on intaglios, Martin Henig kindly provided the following note: ‘… the gem did indeed show two eagles holding a garland between them and both 168 167 170 169 0 50mm Figure 105 Springhead: copper alloy fittings 165–71 standing on globes, signifying world-rule. There is an eagle standing on a thunderbolt atop a globe on an agate intaglio from Aldborough, North Yorkshire (Henig 1974, pl 26, no App69). For an eagle, wings partly spread, simply standing on a globe, a better parallel is an agate/onyx from Pompeii set in an iron ring (Pannuti 1983, 149, no 271). For two eagles holding a garland between them but both standing on low altars or bases note a cornelian in Romania (Gramatopol 1974, 82, no 566, pl 27). Also note two eagles on bases but holding individual wreaths on a yellow jasper from Aquileia (Sena Chiesa 1966, 381, no 1266, Tav 64).’ Settling the Ebbsfleet Valley 242 Fittings (Fig 105) 165. Rectangular plate with pattern of niello-inlaid hearts extending on lines from central line, separated by dots; four studs at corners on back. Surface quite corroded but niello decoration visible beneath. SF 9201, Context 0. 166. Strap or belt fitting. Incomplete. Rectangular, silver sheet applied to front. Sheet with openwork decoration of four transverse lines (a fifth visible on broken edge) with heart-shaped ends and triangular notches in the middle. Openings filled with ?niello inlay. Two fragmented rivets near broken end on back. Fairly corroded. SF 158, Context 3996, Intervention 3223 (Ditch), SG 300030 (Ditched enclosure). Late Iron Age. See also SF 165 which lacks openwork decoration. 167. Extended crescent-shaped hoop with tips joining in three collars and knob. Ridged profile. Perforation in wider part. Slight knob at top end probably remains of suspension loop or hinge. SF 1721, Context 6379, Intervention 1001 (Spring), SG 300015 (Deposits). Early Roman. 168. Crescent shape central part, rectangular terminal one side with cordon, knobbed terminal on other side. SF 15632, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. The two very similar crescentic fittings/pendants Cat No 167 and 168 were found in contexts either side of the Ebbsfleet area. The small holes both items have in the upper part of the crescent find parallels in such pendants from Butzbach and Stockstadt on the Upper Germanic limes (Oldenstein 1976, 162–4, Taf 45, esp 442.445). All of them belong to the larger group of lunula pendants, widely used as amulets over a long period. The German parallels for the Springhead examples suggest a date in the 2nd or 3rd century. 169. Incomplete. Small openwork plate belt or strap fitting. Rectangular shape, circular-sectioned, riveted stud attached to reverse at complete end. SF 18915, Context 17710, Intervention 17710 (Layer), SG 300600 (Layers), Property 3. Mid-Roman. A similar rectangular fitting from Osterburken on the Upper Germanic limes includes the letters ‘IOVIS’. Oldenstein (1976, 199, Taf 65, 846) dates this and other parallels to the late 2nd and early 3rd centuries. 170. 171. Incomplete. Possible belt or balteus fitting. Flat rectangular cross-section. Ornate shape, symmetrical with curved, zigzag, and straight sides. Remains of rivet on underside. SF 875, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. Circular ‘flower’ shape with six petals and central hole. Flat cross-section. Six petals radiate from middle. SF 1249, Context 3005, Intervention 3005 (Artefact). Early Roman. Four very similar fittings but with more lancetshaped petals were found together with other belt fittings of a military cingulum in a late 4th century grave in KölnAachenerstraße (Böhme 1974, Taf 76, 7). The motif itself is long-lived and does already occur on earlier – 2nd/3rd century – fittings, eg, from the limes forts at Zugmantel and Saalburg (Oldenstein 1976, 188, Taf 57, 704–6), although with a higher number of petals and a different attachment consisting of mostly two riveted studs. A later 14th century example was mounted on a brass buckle frame found at Billingsgate, London (Egan and Pritchard 1991, 113, fig 73, 520). Three fragments of a fine double loop-in-loop chain (SF 20465) were found in roadside ditch 1 on the western side of the Ebbsfleet. Toilet or Medical Equipment While there are a number of objects like the spoon probes or ligulae that might have had a use in the handling or application of medical ointments, there are no obvious medical instruments. The amphoriskos Cat No 187 (Pl 12) as a possible container for a medical ointment might be the only exception. The assemblage on the whole appears more related to body care and cosmetics. (Fig 106) 172. Incomplete toilet set. Circular-sectioned suspension loop, two instruments still attached: one complete small spoon/ear probe (rectangular-sectioned shaft, bent over at top to attach to loop, circular spoon) and part of nail cleaner (angled head, rectangular section, broken blade). Three loose fragments all possibly from pair of rectangular-sectioned tweezers, flaring slightly towards inward-turned straight jaws, two pieces join. Faint ringand-dot decoration on exterior surface of tweezers. SF 350, Context 2183, Intervention 2183. Mid-Roman. 173. Incomplete toilet set. Circular-sectioned ring terminal with free moving tools attached: a) tweezers, complete, L 52.69 mm W 5.23 mm, two fragments join, rectangular-sectioned blades flare out slightly to straight-edged jaws, single grooves decorate edges; b) ear scoop, L 47.17 mm W 4.29 mm, rectangularsectioned shaft, circular bowl; c) nail cleaner, complete, L 35.34 mm W 12.9 mm, leaf-shaped rectangularsectioned blade, tip missing, line decoration along edges and very faint on blade, possibly imitating ribs of a leaf. Suspension loop in same plane as blade. SF 15694, Context 16641, Intervention 16655 (Ditch), SG 300545 (Ditch), Roadside ditch 3. Mid-Roman. 174. Toilet set. Nail cleaner and tweezers detached. Nail cleaner L 39.3 mm, tweezers not measured as incomplete, scoop (L 65.85 mm) with looped terminal attached to ring which also holds two broken shafts and ?chain (covered by corrosion). Iron. SF 509, Context 2766, Intervention 2766 (Layer), SG 300186 (Deposits). Mid-Roman. See also SFs 771, 1879, 15309, 15996, 15998, 18025. Chapter 3 Springhead Metalwork 243 172 173 175 177 176 174 178 182 186 181 180 179 50mm 0 No.187 only: 0 183 184 185 187 Figure 106 Springhead: toilet and/or medical equipment (172–87). Copper alloy, apart from 182 (iron), 183 (silver), and 187 (lead) 50mm Settling the Ebbsfleet Valley 244 Nail Cleaners With one exception (SF 809) belonging to Crummy’s Baldock type (Crummy 2001, 3, fig 2) the 11 individual nail cleaners from Springhead as well as those which are part of the nine toilet sets all have a suspension loop which lies in the same plane as the blade. SF 723 has a shape like a nail cleaner from a context dated to AD 60–65 at Wanborough (Hooley 2001, 110, fig 44, 191) but lacks the ornament of three compressed saltires. The slightly more swollen shoulders of SF 1870 relate it to one from the Sheepen site at Colchester (Hawkes and Hull 1947, pl 100, 34). The two nail cleaners from the copper alloy sets both have more pronounced shoulders; in the case of Cat No 172 they are distinctly angular. 175. Complete. Very small leaf-shaped blade, rectangular section, narrow groove extends slightly up one side from points, small transverse moulding between blade and circular suspension loop. SF 1806, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. 176. Incomplete. Broken flat-sectioned ring terminal. Flatsectioned leaf-shaped blade, broken. SF 1870, Context 6444, Intervention 1002 (Spring). 177. Complete. Rectangular section, shaft offset from the suspension loop by plain junction, shaft tapers to points. SF 723, Context 5125, Intervention 5124 (Posthole). Early Roman. 178. Complete. Lozenge-shaped suspension loop continues in plain junction to oval-shaped blade head, tapers slightly towards middle then flares out at points. SF 385, Context 2263, Intervention 2214 (Pit), SG 300073 (Pits). Mid-Roman. See also SFs 734, 809, 930, 1814, 15299, 18319, 18764. 182. Incomplete. Pair of plain tweezers with blades widening towards missing jaws. Iron. SF 20534, Context 10086, Intervention 10254 (Fill), SG 300731 (Quarry). Early Roman. See also SFs 826, 869, 9173, 15693, 15724, 15733, 15735, 18951. Toilet Spoons and Probes 183. A similar small scoop from a mid–late Roman context at Colchester was made of copper alloy sheet (cf Crummy 1983, 60, fig 64, 1898). 184. 185. 186. Tweezers 179. 180. 181. Almost complete. Parallel-sided blades flare out slightly at straight-edged jaws. Rectangular-sectioned. Tips of jaws slightly damaged. SF 1869, Context 6444, Intervention 1002 (Spring). Complete. Parallel-sided rectangular-sectioned bars, straight-edged jaws bent inwards slightly. Single groove along each edge, incised cross decoration bordered by single transverse groove below terminal loop on both sides. SF 15351, Context 11919, Intervention 11892 (Other), SG 300384 (Structure), Property 12. Mid-Roman. Complete. Rectangular-sectioned blades, both sides decorated with three punched lozenge shapes made up of nine smaller squares. Straight edged jaws bent inwards. SF 488, Context 2670, Intervention 2670 (-), SG 300161 (Deposit). Mid-Roman. A pair of tweezers from Baldock, found in a pit dated AD 90–120, has a comparable pattern of three groups of squares, although not as finely worked as Cat No 181 (Stead and Rigby 1986, 133 fig 57,312). Ear scoop. Incomplete. Strip of metal rolled, flattened and trimmed one end creating slightly cupped scoop. Top of shaft broken but beginning of loop remaining. Probably part of a toilet set. Silver. SF 15729, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. Toilet implement. Incomplete. Scoop/spoon with rectangular-shaped terminal with perforated lug on reverse, petal-shaped enamel on front (light green colour). Small moulding below to join to circularsectioned shaft, widens into elongated scoop, tip missing. Probably from toilet set brooch Cat No 117. SF 20017, Context 17759, Intervention 17759 (Layer). Mid-Roman. Tool. Two circular cross-sectioned rods, one is twisted around the other and continues to form looped terminal, now broken. Rod at other end is flattened and ?broken. Part of tool/implement from toilet set (cf Crummy 1983, 62, fig 67, 1943). SF 15062, Context 10016, Intervention 300370 (Quarry). Mid-Roman. Ligula. Small, flat, circular spoon; facetted shaft, swelling slightly towards middle of lower 3rd where it has octagonal section. Other end blunt, no point. Bowl of long cupped scoop probe, wider near base than tip, found with it, but clearly not part of ligula (broken from shaft at top), L 30.21 mm, W 7.25 mm. SF 20085, Context 17913, Intervention 17913 (Layer), Property 3. Mid-Roman. See also SF 383. There are also six olive probes (SFs 878, 15195, 15207, 18001, 20092, 20520) and one blunt-ended probe and narrow spatula (SF 514). Other Toilet/Medical Equipment An end-looped pestle from a cosmetic set (SF 9154; Jackson 1985, 180, fig 6, 47–9) comes from the channel fills of the Ebbsfleet. Possible mirror fragments have been found in five context (SFs 318, 944, 9143, 15067, 20523). 187. Medicine container (Amphoriskos). Small amphorashaped vessel, hollow but quite squashed. Lead. SF 1436. (Pl 12). Chapter 3 Springhead Metalwork Plate 12 Amphoriskos, possibly used as a container for medicine (Cat No 187), lead L 72 mm. Photo: E Wakefield Cat No 187 is a stray find from the slope to the east of the Ebbsfleet and the Sanctuary. No similarity exists with the well-known medieval lead ampullae with scallop or round bodies (Spencer 1990, 57–9, 85–92, figs 170–89); the round body shape could well be based on late antique ‘Menas ampullae’ (Effenberger and Severin 1992, 48, Abb 38). Its shape suggests that it imitates a type of amphora of possibly eastern Mediterranean, perhaps Graeco-Roman or Rhodian forms (eg, Peacock and Williams 1986, figs 37, 39). No exact parallels are known to the author; however, a small ‘lead amphora flask’ on offer on an antiquities website, and purported to be from Europe, (www.gilliscoins.com/antiquities/ celtic_and_roman/pic_wro5185.htm; accessed 18 May 2008) comes closest in terms of its shape and general appearance. This 50 mm high vessel also has a pointed body which in contrast to the Springhead amphoriskos is decorated with a pattern of circles, saltires, and triangles arranged in zones on the body of the object; its handles, which are triangular like those on Cat No 187, are attached at either side of the shoulder area but only continue to the base of the neck. Small lead ampullae looking like flat-bottomed amphorae are known from the eastern parts of the Roman Empire where they were used as medication containers; quite a few of these have been found with texts including the word ‘λυκιον’ written on them or attached with a label (Simpson 1854, pl 2; Sjöqvist 1960, pl 20). The Lykion in question is a plant of the Berberis family used for medicinal purposes. The 2nd century medical author Galen claims that the most powerful Lykion comes from India. The plant had many medical uses, but most importantly it served as a medication against a wide variety of eye infections and inflammations because of its astringent and soothing properties. British doctors reported its widespread application in India during the 19th century and it was used for those conditions on British soldiers returning from Egypt (Simpson 1872, 50–2; Boon 1983, 9–10). Apart from Lykion, other medications were also sometimes carried in lead ampullae (E Künzl, pers comm), as exemplified by the labels ‘xeron presidi’ on a 245 phial from Lauriacum, Austria, (Ruprechtsberger 1974) or ‘aitnaion’ on a small lead container from Italy (Ferrandini Troisi 1981). Apart from its use as a container, the amphoriskos itself could also have had an apotropaic property, which can also be assumed for amphora-shaped 4th century strap-ends (cf Simpson 1976, 198–200) like those from Lullingstone (Meates 1987, 73, fig 30, 163), Richborough (Bushe-Fox 1949, pl 36, 112–13) or Bremen-Mahndorf, grave 208 (Böhme 1974, Taf 9,16) which are stylistically close to lanceolate or discoid strap-ends with lateral animal tendrils (eg, ibid, 74, Abb 28; Taf 115, 10). Textile Manufacture or Working of Textiles (Fig 107) 188. ?Wool comb. Iron. SF 9398, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. 189. Spindle whorl. Disc-shaped, perforation at centre. 17 g. Lead. SF 559, Context 2831, Intervention 2831 (Layer), SG 300186 (Deposits). Early Roman. Although it is uncertain whether identification as spindle whorls is correct for some of the perforated lead discs (SFs 1446, 1511, 1722, 9136, 15078, and 18647) are similar to such discs from South Shields where they have been described as caulking (Allason-Jones and Miket 1984, 331, figs 8.96, 8.97, 8.99). At the medieval Tyske Bryggen at Bergen, Norway, the seven lead spindle whorls found there range in weight from 9.5 g to 71.3 g, their diameters vary from 19 mm to 28 mm and the heights from 8.5 mm to 22 mm (Øye 1988, 49–50, fig II.14); at Springhead these measurements are: weight: 12–46 g, diameter: 20–29 mm, and height: 4–11 mm. SF 15078, which comes from a mid-Roman quarry in property 12 and weighs 46 g, could well have been used as a weight of 10 sextulae (nominal weight 45.5 g). Other possible lead spindle whorls have biconical (SF 1512, 64 g and 15923, 32.2 g), or D-shaped profiles (SF 1711, 72 g), and thus the weight range of the Springhead whorls is comparable to that from Bergen. Needles 190. 191. Complete. Flat spatula head, rectangular eye, circularsectioned shaft bulges out slightly at centre. Crummy Type 2a (cf Crummy 1983, 66, fig 70; 1977). SF 1562, Context 3547, Intervention 3546 (Pit), SG 300245 (Pits). Early Roman. Simple needle with eye set in short groove near the top of the shaft. Iron. SF 15214, Context 10608, Intervention 10608 (Layer), SG 300407 (Layers), Property 11. Early Roman. This iron needle is similar to Crummy Type 3 of copper alloy with a groove above and below the eye. Settling the Ebbsfleet Valley 246 189 192 188 from X-ray 191 190 196 193 195 194 0 50mm Nos.188–191 & 195: 0 50mm Nos.192–194 & 196: Figure 107 Springhead: wool comb (188), spindle whorl (189), needles (190–1), domed stud (192), box handle (193), spout (194), pot-mend (195), rivet (196). Copper alloy, apart from 188 and 195 (lead) and 191 (iron) The iron needle is similar to a 1st/2nd century example from Walbrook in London (Manning 1985a, pl 15, D22). With a length of 106 mm it sits near the middle of the range of needles discussed by Manning (ibid, 35–6). See also SFs 384, 15104, 15627, 18305, 20018, all of Crummy type 3. SF 15390 may be the heavily bent shaft of a needle. Household Utensils and Furniture 192. Box fitting. Domed stud with central recess marked by circular outer ridge. Riveted to square base sheet. Corrosion products on upper side of fitting contain spelt glume fragments (C Stevens, pers comm). This item has possibly been burnt or cremated, which suggests it may belong on a burial casket. SF 976, Context 6378, Intervention 1001 (Spring), SG 300015 (Deposits). Mid-Roman. 193. Small box handle. Very thin. Rectangular with recessed central panel. One corner badly bent. Small circular perforations in two opposing corners. Within central panel a fine piece of copper alloy wire has been threaded through two hoops (these fixed on reverse by twisting ends and bending backwards against panel), itself forming small U-shaped loop (cf medical etui from Wehringen, Germany, grave 7; Nuber 2000, 170, Abb 141). SF 18721, Context 17709, Intervention 17709 (Layer), SG 300600 (Layers), Property 3. See also fittings from caskets in graves 6104 and 6345 and lock-bolt SF -515 Chapter 3 Springhead Metalwork 247 Two bell-shaped lock-pins (SF 15753 and 18270) come from property 2. Ring-key SF 9373 could have served to lock a small casket. It appears to belong to Guiraud’s type 5a of general Roman date, which is common in eastern and north-eastern Gaul and along the Rhine (Guiraud 1989, 191, fig 33). Parts of Metal Vessels 194. Zoomorphic spout. Lower jaw forms spout, snout/nose upturned and decorated with grooved lines forming triangle. Oval groove is possible eye on left side only. Triangular ‘ears’ protrude from top. Top also decorated with incised lozenge-shaped lines. SF -523, Context 66, Intervention 66 (Tree-throw hole), Property 7. Early Roman. Found under floor of early Roman building 300522, as was Cat No 265. Strainer bowls with zoomorphic spouts representing fish heads (eg, Felmersham, Bedfordshire; Watson 1949, pl 5a–b, 42, figs 3–4), a dog or lion (eg, Marlborough, Wiltshire; Robinson 1998, 147, fig 1) or an ox head (eg, Kirmington, Lincolnshire; May 1971, 254, fig 1) appear to be a native British product (Sealey 1999, 121), linked to the consumption of beer, and possibly mead, rather than wine (ibid, 123–4). The oldest bronze example comes from the late 1st-century BC burial at Welwyn Garden City (Stead 1967, 24, fig 12), and no bronze examples are found after Nero or the Boudiccan revolt; in pottery, the earliest strainers of the carinated CAM 323 form are dated to the 2nd quarter of the 1st century AD, and they go out of use by the early 2nd century (Sealey 1999, 121–2; Mills, this vol, Chap 1). Three strainer bowls with similar spouts were recently discovered in a hoard consisting of three trullei and two strainer bowls at Kingston Deverill, Wiltshire (Worrell 2006, 458–62, figs 31–2) and at Chettle, Dorset (A Fitzpatrick, pers comm). A fragment of a vessel with five perforations (SF 590) was found in the channel fills of the Ebbsfleet and may have been part of a strainer bowl similar to the type that Cat No 194 belonged to. Of interest in conjunction with the presence of at least one strainer bowl or its fragments at Springhead is fitting Cat No 297 (see below) which looks remarkably similar to a crescentic plate which projects from the back of the head of the Felmersham spout, presumably originally supporting a firmer attachment of the spout to the rim of a vessel (Watson 1949, 41–2, figs 3–4). However, there are no traces on the rim of Cat No 297 which would suggest an attachment of a similar kind. The bottom fragment of a strainer (SF 15970) with a perforation pattern similar to den Boesterd 52 (den Boesterd 1956, 19, pl 13, 52) was found in tank 16831 in property 4. Dipper and strainer sets of this and similar forms date to the later 1st and 2nd centuries. Three fragments of metal vessel rims are too small to identify a type (SFs 1603, 1783, 9440), as is a folded strip (SF 1900) perhaps from the body of another vessel. Plate 13 Type 1 lead plug on a vessel in a shelly fabric (Fig 46, 636) from Context 2319. Photo: E Wakefield An escutcheon (SF 9224) for a bowl or bucket was found in quarry 300204; a similar piece comes from South Shields where it was suggested to belong to an uncommon 2nd century type (cf Allason-Jones and Miket 1984, 216, no 3.732). SF 9001 is a fragment of a handle with plain leaf-shaped terminal from a metal vessel similar to one from Colchester (Crummy 1983, 72, fig 76, 2045). SF 9412 is an iron handle, possibly for a bowl, found in deposits pre-dating the sanctuary. An iron swivel and loop (SF 20552) with chain of figure-of-eight-shaped links was found in quarry pit 10016. It may have served for the suspension of a cauldron (cf Manning 1985a, 138, pl 64, S4) but, among other uses, could also have been part of a dog chain. A similar swivel and loop was found at Northfleet villa (see Schuster, Chap 4, Cat No 10). SF 18294 from the waterfront area in property 4 may be a leg of a small stand for a dish or bowl (cf Pompeii: Ward-Perkins and Claridge 1976, nos 320–1). Pot-mends (Figs 107 and 108) 195. Two almost D-shaped sectioned bars with flattened sides facing each other, joined by two circular-sectioned shafts at opposing ends. Lead. SF 15427, Context 16022, Intervention 16022 (Layer). See also pot-mends of type 1: SFs 263, 265, 346, 406, 520, 842, 863, 9172, 9451, 15404, 15405, 15407, 15408, 15410, 15617, 15740, 18207, 18208, 18292, 18378, 18844, 18873, 18882, 18917, 20101, 20398, and a vessel in a shelly fabric from Context 2319 (Fig 46, 636 and Pl 13) and an oxidised flagon from context 16776; type 2a: Cat No 195, SFs 894, 1708, 9370, and samian vessels from contexts 16687 and 16863; type 2b: SF 15617. The 34 pot-mends from Springhead are all of lead. They can be sub-divided into two types. Type 1 is the most frequent with 27 examples. It has irregularly shaped inner and outer discs (relative to the position on the vessel), normally joined by a narrower middle part which creates a notch around the circumference. The inner and outer discs can be of similar size, as SF 406 or the very carefully crafted SF 18378 with two almost circular discs and fragments of the ceramic vessel still held in the middle. Others have discs of considerably different sizes, like SF 15410. Fabric imprints on SF 9451 on the larger inside surface of the pot-mend suggest that the person carrying out the repair had at     HU 5LY  *OXHGSRWVKHUGV  Settling the Ebbsfleet Valley 248 (E EVI  OHH W  $ .H\ /DWH,URQDJH (DUO\5RPDQ 0LG5RPDQ /DWH5RPDQ 2WKHUSRWPHQGV  W\SH W\SH HU 5LY E (E H VIOH  W  $  P %DVHGXSRQWKH2UGQDQFH6XUYH\Š/DQG/LQHŠGLJLWDOGDWDZLWKWKHSHUPLVVLRQRI2UGQDQFH6XUYH\RQEHKDOIRI7KH&RQWUROOHURI+HU0DMHVW\ V6WDWLRQDU\2IILFH‹&URZQ&RS\ULJKW $OOULJKWVUHVHUYHG/RQGRQ &RQWLQHQWDO5DLOZD\V/LPLWHG8QLRQ5DLOZD\V 1RUWK /LPLWHGERWKRIUG)ORRU(YHUVKROW6WUHHW/RQGRQ1:$</LFHQFH1R Figure 108 Distribution of lead pot-mends and glued pot sherds Chapter 3 Springhead Metalwork 249 197 200 198a 201 202 199 198b 203 206 209 204 207 205 0 208 50 mm 0 50 mm Nos.197, 199, 207 & 208 only: Figure 109 Springhead: objects employed for weighing and measuring (197–203) and writing/written communication (204–209). Copper alloy: 197, 207–8; iron: 204–6 (206 with copper alloy); lead: 198–203 (203 with iron) least some basic understanding of techniques used by later, medieval, pewter workers which have recently been summarised by Egan (1998, 189, fig 153, 240, fig 188; see also Schuster 2006, 92–3). SF 9172 may be a possible variant of type 1; with a length of 70 mm it is quite large and one side roughly shaped like a bow tie with rounded, semi-circular ends connected by a thinner bar in the middle. The other side has one T-shaped and one irregular end, and the notch around the middle still contains remains of some unidentified oxidised ware. Type 2 is of a cramp-like shape with an upper bar, normally of D-shaped section, with two shafts at or near the ends. The four examples of variant 2a have shafts continuing into another D-shaped bar; while in the one of variant 2b the shafts end in two flatter points bent inwards at right-angles to the shafts. At Springhead, evidence for pot-mends of type 2 is confined to samian vessels and takes the form of drilled or filed holes, two vessels even had the pot-mends still in place. While potmends of type 1 are commonly used on coarsewares (see Fig 46, 636; Pl 13), type 2 pot-mends are also occasionally found on such fabrics, for example at Kingscote (Redknap 1998, 112, fig 64, 15.23–4). A third type of pot-mend, consisting of copper alloy strips with separate rivets, is known from grave 24 at Verulamium, KHL, dated c AD 30–55 (Stead and Rigby 1989, 279, fig 92, 24.2); if dismantled, such fittings would be extremely difficult to distinguish from simple strips with rivet holes either end. None was noted among the material from Springhead but small strip fragments like SF 1885 could have been part of such a mend, among many other uses, for instance, on leather. Evidence for organic ties and glues is discussed elsewhere in this volume (Seager Smith and Marter Brown, Chap 1). The distribution of lead pot-mends follows that of the glued pot sherds (Fig 108) but is more restricted, focusing on properties 3 and 4 to the west of the Ebbsfleet, and the southern area of the Sanctuary and pre-Sanctuary contexts. Based on a sample of 76 vessels from Silchester with holes pierced deliberately after firing, some of which were subsequently repaired, Fulford and Timby (2001) have argued against Settling the Ebbsfleet Valley 250 deliberate ‘piercing, as a kind of ritual killing, being part of the ritual of deposition.’ They suggest as more likely that the holes were pierced for uses such as timing devices or food preparation (ibid, 296). While these are certainly feasible explanations, the concentration in the Sanctuary area might have some connection to ritual practices, while those in properties 3 and 4 could perhaps rather be related to the lead working carried out in these properties. 196. Rivet. Folded towards centre and bent back again lengthwise. Slightly bent. SF 52007, Context 411, Intervention 411. Spoons Four spoons have been recorded, including an almost complete spoon with round bowl (SF 15951) and the bowl of another (SF 924), belonging to Crummy type 1 of the later 1st and 2nd centuries (Crummy 1983, 69). The two others are fragmented oval bowls only (SF 9265 and 9361). Weighing and Measuring The finds of scales and weights at Springhead do not show any particular concentration. While some may be indicative of unspecified commercial activities, those found in layers associated with the riverbed may also have served a votive function. Scales (Fig 109) 197. Balance. Incomplete, balance arm. Circular cross section, tapers slightly both ends, both tips are missing. Base of small lug mid-way along shaft. Badly corroded. SF 1809, Context 6444, Intervention 1002 (Spring). A badly corroded copper alloy rod (SF 18311) found unstratified in the ARC SHN02-area may be a further balance arm. Neither this nor Cat No 197 are sufficiently well preserved in the area of the lug for suspension to allow for a basic dating on the basis of Steuer’s (1990, 44, Abb 1) typology. Weights Twenty-three weights have been recorded at Springhead. Table 50 gives an overview of the various shapes, which appear not to be linked to specific weight classes. Unfortunately, only six weights were found in secure Roman contexts, most of the remainder coming from overburden deposits. Of the disc-shaped weights, only Cat No 199 is from a secure context, but it may be either a weight or a token; if the latter, it may be that the ornament was meant to represent a Chi-Rho, although the ornament’s crude manner of execution precludes any certainty. A similar object from the fortress baths at Caerleon has some faint scored lines on one side but certainly no Chi-Rho and was described as a small lead disc or counter (Zienkiewicz 1986, 190, fig 65, 13). If no similarity to a Chi-Rho was intended on the Springhead disc but the weight of the object was of interest, the symbol could have been intended to represent a double cross which was used as a symbol for the denarius, eg, on a lead tag also from Caerleon (Hassall and Tomlin 1989, 342–3, fig 8). This suggestion is supported by the object’s weight of 3.5 g which is reasonably close to a drachma of 3.41 g, the standard weight of a denarius after Nero’s debasement of that coin denomination (Chantraine 1961, sp 622). Cat No 199 was found in an early Roman chalk quarry which also contains some 3rd/4th century coins in the upper fills. Thus, in terms of the date range of the fill both interpretations are feasible. A plain copper alloy disc of that weight was found at Colchester (Crummy 1983, 101, fig 105, 2512); its dimensions, apart from its weight, are similar to the lead disc SF 1822. However, double crosses are also found on much later weights, as on a possibly Carolingian weight weighing 68.8 g from Mainz, Germany, which has a more accurately incised double cross on both sides (Wamers 1994, 175, fig 103, M09). The inclusion in the table of the unstratified cylindrical weight SF 1310, which was cut off from a lead baton, is based on its weight of almost exactly one sextula. Somewhat more uncertain is SF 18647 with a weight of almost 16 scruples, found in the channel fills of the waterfront in the Roadside Settlement area. It is of irregular cylindrical shape with a central hole and may have been used as a small suspended weight on a steelyard, but a use as net sinker is equally feasible. The same applies to SF 18438 with a weight of 7 g. The cone-sectioned weight SF 9395 was found in deposits overlying the riverbed; considering its weight of almost exactly half a Roman pound it may well be a Roman weight, while SF 15406, which was found in a spread above properties 3 and 4, may be a 3rd of a Celtic pound. With 309.1 g (Anderson 2001a, 117; Frere 1972, 150 gives 310.05 g) the Celtic pound was lighter than the Roman and a 3rd would weigh 103.03 g. The biconical weight SF 15113 was found in the fill of late Roman Grave 10150, while the other two biconical weights were found overlying Romano-British features at the base of the slope in the Sanctuary area. All have remains of suspension loops for use on steelyard scales. An ounce weight of similar truncated spherical shape to Cat No 201 was found at Wanborough, Wiltshire, in a later 4th century context (Hooley 2001, 115, fig 47, 246). Apart from a central dot it also had the Greek symbols to denote one uncia; from the time of Diocletian onwards it had become customary in the West to use Greek numerals on weights (Steuer 1990, 43). While truncated spherical copper alloy weights are a very Chapter 3 Springhead Metalwork 251 Table 50 Springhead metal finds: weights Square Uncertain Cylindrical Cone-section 260a) Cat No 198a); 19.3 g, may represent ¾ of an uncia = 20.47 g, or 17 scruples = 19.33 g 1310 4.4 g, represents 1 sextula = 4.548 g 9395 166 g, may represent 1 semis = 6 unciæ = 163.73 g 260b) Cat No 198b); 315 g, represents 11½ unciæ = 313.8 g 1226 6g 18664 Cat No 202 108, represents 1 triens = 4 unciæ = 109.15 g 83 2 g (CuA) 1582 114 g, may represent 1 triens = 4 unciæ = 109.15 g 18647 18 g, may represent 16 scruples = 18.192 g 15406 98 g 261 100 g 1333 5.6 g, may be 5 scruples = 5.685 g 18712 Cat No 203; 317 g, probably equivalent to 11½ unciæ = 313.8 g 239 4g 15113 32.8 g 1392 Cat. No. 201; 26.0 g (CuA), possibly 1 uncia = 27.288 g 20046 61 g (heptagonal) 9222 4.6 g, represents 1 sextula = 4.548 g 1822 Cat. No. 200; 4.2 g may represent 1 sextula = 4.548 g Biconical Truncatedspherical Disc 1874 31 g 50931 95 g 18680 10 g 9323 Cat No 199; 3.5 g, may represent 1 drachma = 1 denarius = 3.411 g SF numbers and weight, Cat No if included in catalogue and illustrated. All lead unless marked CuA = copper alloy. Roman weights according to Chantraine 1961, sp 620 common Roman type which continues into the 6th century, they again become very widely used throughout Europe from the 10th century onwards (Steuer 1997, 46). The truncated spherical or bun shaped weight SF 50931 was found in an evaluation trench which produced pottery of the 1st–3rd centuries. The two smaller SFs 1226 and 1333 are both hemispherical; both are unstratified and their use as weights is uncertain. The only two square weights were found during metal detecting of a mixed surface layer in property 3. Their square shape would allow placing them in the late Roman period when this form became quite common for weights (Steuer 1990, 43) which usually had the weight units displayed on them in either Roman or Greek numerals. Equally, a much later date is feasible also, as similar weights are known from late medieval/early post-medieval contexts like a weight from Schleswig, Germany, which has a ring-punch mark very similar to Cat No 203, although arranged in a way that makes it appear as a number ‘8’ (Steuer 1997, 67, Abb 34). With a weight of 46 g, SF 15078, recorded as a spindle whorl, weighs almost exactly the equivalent of ten sextulae (45.48 g) and may have been used as a weight. 198. 199. Two weights. a) small disc-shaped weight. 19.3 g. b) larger biconical-shaped steelyard weight, slightly flattened at bottom end, pointed at top where iron corrosion indicates the remains of a suspension loop; 315 g. Both lead. SF 260, Context 2183, Intervention 2183 (-). Mid-Roman. Irregular, roughly cut out subcircular disc; incised decoration of what looks like a six-armed cross on both sides, one possibly a Chi-Rho; 3.5 g. Lead. SF 9323, Context 6568, Intervention 6571 (Other), SG 300204 (Quarry). Early Roman. If a Chi-Rho was indeed intended to be shown, it may be compared with equally crude examples from grafitti on 4th century bowls found in London and Caerwent (Thomas 1981, 89, fig 5, 5.9; Green 1976, pl 24c). 200. 201. 202. 203. Thick disc, possibly with impression on one side. 4.2 g. Lead. SF 1822, Context 6445, Intervention 1000 (Spring). Complete. Spherical with flat top and bottom. The top is less affected by corrosion and has a small silvery rectangle in the centre. 26 g. SF 1392. Square shape, rectangular section. Two ring-punch marks on one side just off centre. 108 g. Lead. SF 18664, Context 19026, Intervention 19026 (Other). Square shape, rectangular section with two slightly convex sides; iron shaft through centre folded back onto base, protrudes at top: remnants of suspension loop. 317 g. Lead. SF 18712, Context 19026, Intervention 19026 (Other). Writing and Written Communication 204. Stylus Manning type 1. Shaft slightly bent. Iron. SF 18248, Context 11320, Intervention 11320 (Layer), SG 300456 (Layers), Property 10. Early Roman. See also SF 15898 which may be a copper alloy variant of the type. 252 Settling the Ebbsfleet Valley Stylus Manning type 2. Iron. SF 15698, Context 16642, Intervention 16642 (Layer). Mid-Roman See also SF 9465 and possibly 15322 (bent at right-angles). SF 20329 is a stylus Manning type 4, SF 244 may be a copper alloy stylus. 206. Fragment of knife handle, with diagonal appliqués with lower edges decorated with a volute, and small part of iron tang. SF 268, Context 2183, Intervention 2183 (-). Mid-Roman. is conical shape. Lead. SF 977, Context 6378, Intervention 1001 (Spring), SG 300015 (Deposits). Mid-Roman. See also possible lead seals SFs 1769, 9380, 18688. Knives with similar handles, frequently counted among toilet equipment (eg, Boon 1991, 22, fig 1a), have recently been identified as pen knives for sharpening the calamus during writing and are thus listed here with other writing accessories (cf Feugère 2003). Cat No 206 belongs to the type with sawn handle for which numerous parallels can be cited, especially from early Roman contexts, amongst others Alba, France (ibid 10, fig 2, 2) or Newstead (Curle 1911, pl 60, 12). Bridles 205. 207. Circular seal-box. Complete. Base has three perforations: one in centre, two towards edges. Two rectangular notches on opposing sides at right-angled plane to hinge axis. Hinge intact (not free moving) – double perforated lugs and ?iron pin hold perforated lug of lid. Lid bent backwards at right-angles to base. Incised/stamped ?decoration (of transverse lines) inset in circle around possible rivet. Rivet suggests possible further decorative plate (eg of bone) may have been at centre. SF 1572, Context 3547, Intervention 3546 (Pit), SG 300245 (Pits). Early Roman. See also SFs 367, 1836 (bases only), 9200, 18727. 208. Lanceolate seal-box. Complete – lid and base separate, contents from within box removed by conservator and retained in small tube. Tear-drop shape. Lid has small lug at narrow end (that opposite the hinge), upper surface has possible dotted decoration, some dots create swirling lines (others may be result of corrosion). Base has four perforations in bottom, three at wider end, one at narrow end. On the exterior shallow depressions can be seen around these holes. Double perforated lugs at one end hold iron rod for hinge, at opposing end is further small lug (corresponds to lug on lid). Along edges are two rectangular notches, these lie at right angles to the hinge axis. SF 15744, Context 12061, Intervention 12059 (Post-hole), SG 300329 (Post-holes), Property 2. See also SF 15618 (base fragment). A similar seal-box lid from Great Walsingham has four lanceolate petals in a cruciform arrangement (Bagnall Smith 1999, 41, fig 4, 48). A comparable object from Colchester was considered too thin for a seal-box lid (Crummy 1983, 167, fig 204, 4647). 209. Seal. Oval shape, ends more pointed, flat one side where small circular depression can be seen, one hole goes through to upper surface where it is split into two smaller holes, two grooves then run off at right-angles probably where ?string ran out of mould. Upper surface Objects Associated with Transport Harness (Fig 110) 210. One half of snaffle-bit with ring and link with tubular loop, loop at other end of link missing. Iron. (cf Manning 1985a, pl 28–9, H11 and H14). SF 20393, Context 16890, Intervention 16902 (Pit), SG 300579 (Pits), Property 4. Mid-Roman. Snaffle-bit SF 15084 from Grave 10079 is of the same type. 211. Snaffle-bit. One ring and link with thick tubular loop, other end, link and ring missing. Iron. SF 20438, Context 19580, Intervention 19581 (Ditch). Mid-Roman. Buckles 212. Strap buckle, complete. D-shaped frame, D-shaped cross section. Rectangular terminal with T-shape cut out, two grooves along opposing edges on one side only (cf Bishop and Coulston 2006, 191, fig 124, 7; Oldenstein 1976, Taf 76, 1008). SF 15631, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. Strap distributors 213. Phalera. Flat circular disc head, raised circular decoration at centre. On reverse is rectangular/Dsectioned bar forming a right angle (originally a Ushape, but part missing). ATC33MH. SF 50995. Such phalerae are most likely part of military horse harnesses, used in conjunction with pendants similar to Cat No 294–5 (see discussion below). Hipposandals 214. 215. 216. Right side wing of a hipposandal (Aubert 1929, 1F). Strip with lower edge bent at right angles where it would continue into the sole which is missing. Upper edge tapering triangularly towards front (compare Cat No 215). Iron. SF 20425, Context 17896, Intervention 17896 (Layer), SG 300642 (Layers), Property 3. Mid-Roman. Right side wing of a hipposandal (Aubert 1F). Strip with lower edge bent at right angles where it would continue into the sole which is missing. Upper edge tapering triangularly towards front (compare Cat No 214). Iron. SF 20458, Context 16894, Intervention 16902 (Pit), SG 300579 (Pits), Property 4. Mid-Roman. Left side wing of hipposandal (Aubert 1F). Wider part slightly curved, with two slots (x-ray). Iron. SF 20152, Chapter 3 Springhead Metalwork 253 210 212 211 214 216 213 215 217 218 219 0 Nos. 210, 211, 214–218: 50mm 0 50mm Nos. 212, 213 & 219: Figure 110 Springhead: objects associated with transport (210–19). Iron, apart from 212–3 and 219 (copper alloy) Settling the Ebbsfleet Valley 254 221 222 220 224 223 0 50mm Nos. 223–226: 0 50mm Nos. 220–222: 226 225 Figure 111 Springhead: objects associated with transport. Copper alloy (220–2), iron (223–6) Chapter 3 Springhead Metalwork 217. 218. Context 12131, Intervention 12131 (Layer), SG 300336 (Layers), Property 2. Early Roman. Two fragments join together. Right side wing of hipposandal with long end, ends in a sharp point protruding beyond the outline of the square part. SF 20319, Context 17177, Intervention 17179 (Pit), SG 300556 (Pits), Property 5. Early Roman. Might be developing towards Aubert’s second series. Flat base with rear downward-facing hook only. Iron. SF 1510a), Context 3418, Intervention 3228 (Pit), SG 300214 (Pits). Mid-Roman. Terrets 219. Incomplete. Wide D-shaped frame, approximately onethird missing, widens in middle, rectangular section. Straight rectangular-sectioned bar. Decorated on both sides: raised curvilinear tendrils with additional groove decoration and yellow enamel dots, recessed area around this is enamelled in champlevé technique (now green colour) – this decoration repeated on other side. SF 18271, Context 12132, Intervention 12132 (Layer), SG 300336 (Layers), Property 2. Early Roman. (Fig 111) 220. Incomplete. D-shape sectioned ringed terret. Skirted base has scalloped edges. Knobs at opposing ends. Grooved decoration visible around collar, stamped dot decoration radiates from top in lines (only visible on one side but possibly on other also) and around edge of base. SF 15396, Context 16157, Intervention 16157 (Layer), SG 300491 (Layers), Shrine. 221. Double ringed head sits at right-angles to moulded hollow base; two knobbed upturned protrusions at sides with flat circular head surviving on one side. Remains of square-shaped shaft protrude from underside. SF 15910, Context 16022, Intervention 16022 (Layer). 222. Incomplete. Fragment of harness fitting with remains of loop and skirt with knob at one end of crescentic terminal. SF 248, Context 2183, Intervention 2183 (-). Mid-Roman. The ornate terret Cat No 219 belongs to a group of similar objects found in the south and east of Britain (Jope 2000, 157; pl 292–4). Their method of fixture on the yoke is shown by MacGregor (1976, 40, fig 3). These terrets have a red champlevé enamel patterning on the flat surface of a wide frame. The technique itself was used in Britain from the 1st century BC onwards and by the early 1st century AD began to be employed on a variety of fittings like the terrets similar to Cat No 219 (Jope 2000, 159). The Springhead terret was found in a late 1st century AD context which is part of the makeup layers prior to the construction of the temple in property 2. That the ornament was still in use by the late 1st century can be demonstrated by two neck collars of Flavian date from Stichill and Plunton Castle in Scotland, which show an ornament called a ‘swash N’ by 255 Leeds (Ulbert 1977, 41, Abb 3.5–6, after Leeds 1933, 110). Related ornaments are also found on some trumpet brooches (eg, Cool and Philo 1998, 48, fig 12, 73), and the dragonesque brooches are another instance of a flourishing of ornaments based on indigenous Celtic design after the Conquest (MacGregor 1976, 127–9, fig 6). Simple terrets are widely known throughout the Roman Empire, with a focus on its north-western provinces. Their function and position on Roman yokes and horse collars has been reconstructed by Alföldi and Radnóti (1940, 310, Taf 23,6; cf also Radnóti 1961, 31, Abb 13 left), based on near complete yokes and documentary evidence from Pannonia. Terrets with iron spikes like Cat No 221 were certainly nailed into the wooden part of a yoke, and this arrangement would be needed when a third animal was harnessed as a lead in front of two others. Terrets with loops could be strapfitted but are also found on the sides of a yoke, as described above, where they were set into a hollow and secured with a small stick. The hoops through which the reins were running are most commonly round or oval, but there are also pelta-shaped hoops and, much rarer, double hoops like Cat No 221. A British example with double hoops was found at Chinnor (Worrell 2006, 446, fig 15), a few more are known from Pannonia and both Germanies (Alföldi and Radnóti 1940, 315 with note 34, pl 25, 2). Interestingly, the three terret fragments Cat No 220–2 all have more or less pronounced croissantlike terminals at the small sides of the base collar. Terrets with distinctly croissant-like terminals like Cat No 221 and 222 have thus far mainly been confined to Raetia and the two Germanic provinces, with a focus in the Rhine-Main area, and a possible manufacture of such terrets was suggested for the legionary fort at Mainz (Koch 1997, 167). A further four were found outside the limes in Germania magna as far north as the southern North Sea coast (Schuster 2006, 80). The distribution has, however, received a significant addition in Britain with 12 terrets with croissant-like terminals recorded by the Portable Antiquities Scheme. The objects are mainly found in the east of England, from Surrey to Lincolnshire with concentrations in Suffolk and Norfolk, but one very good example was found in Wales near Langstone, Newport (S Worrell, pers comm). The terminals of Cat No 220 are less pronounced and more like an end-knob. Parallels for this detail appear so far to be confined to Britain, with examples from London (Webster 1958, 87, fig 6, 166 with a square lower loop), Bancroft, Buckinghamshire (Hylton 1994, 319, fig 151, 150), and Ferrybridge Henge, West Yorkshire (Duncan et al 2005, 153, pl 27, 156, fig 116, 4) as well as the above mentioned double hooped terret from Chinnor where the knobs appear in the middle of the long sides of the skirt, while the London example has knobs there and at the raised longitudinal ends. The Springhead examples fit well within the known date range of the type, which covers the period between the late 1st and 3rd centuries (Koch 1997, 165). Settling the Ebbsfleet Valley 256 0 50mm Figure 112 Springhead: lead guttering (227) Cart fittings Animal driving implements 223. 226. 224. 225. Possible lynch pin (Manning type 2a). Spatulate head, bent at c 60º. Square-sectioned shaft. Iron. SF 20137, Context 12220, Intervention 12165 (Pit), SG 300346 (Pits), Property 2. Mid-Roman. Lynch pin. Spatulate head tapering towards stem, with turned-over loop at the top (Manning type 2b). Rectangular-sectioned stem with rebate starting c 42 mm from lower end. Iron. SF 9400, Context 2945, Intervention 2945 (Layer), SG 300163 (Deposit). Mid-Roman. Possibly pole or beam binding for cart? Arched plate with rectangular opening in middle of base with nail/rivet hole above it. Tapers in the middle and continues as a flat bar with another hole at its broken end. Iron. SF 20112, Context 19026, Intervention 19026 (Other). The identification of Cat No 225 is by no means certain, but it is not inconceivable that it could have served to protect the tip of the draught-pole of a cart from wear. Other possible bindings are known from Newstead (Curle 1911, 288, pl 15, 4; Manning 1985a, 75, fig 21, 2), which appears to be flat and looks like it has a square loop at the wider end, and a slightly curved binding from Warrington (Jackson 1992, 79, fig 43, 8, 82). Ox goad. Socket of two coils, increasing in width towards the long point. Iron. SF 20537, Context 10808, Intervention 10808 (Layer), Property 11. Early Roman Although it probably does not apply to this example because of its size, many ox goads (like that from Gadebridge Park: Neal 1974, 164, fig 71, 404, or Northfleet, see Schuster, Chap 4, Fig 124, Cat No 12 ) could well rather be ink pen nibs. At Vindolanda three such ‘ox-goad’ pen nibs were found with a wooden shank still attached to the iron nib, some still with ink traces. A narrow hole bored down the centre of the shank allowed a use similar to a fountain pen (Birley 2002, 35, fig 26). The use of the ox goad to drive animals is best illustrated by the probable 2nd/3rd century Piercebridge Plough Group, a tiny bronze model depicting a man holding an ox goad in his right hand, ploughing with a bull and a cow (Manning 1971, 126, fig 1, pl 44a); the combination of a male and a female animal in the plough team and the ploughman’s head covered with a hood suggest the statuette represents a ritual act (ibid, 134). In Gaul ‘ox goad’ like objects are frequently found in late Celtic and early Roman sanctuaries and temples and uses discussed include both the above as well as candle/lamp holder (cf Nickel et al 2008, 153 with note 341 and further refs), Chapter 3 Springhead Metalwork and the latter might also provide an interesting explanation for Cat No 226. A support for the interpretation as lamp holder is provided by very similar objects with a cup formed by two lateral flanges rather than a spiral, known as cupped candleholders, dating from the high medieval period onwards (eg, Egan 1998, 142, fig 108). Buildings and Services (Fig 112) 227. Fitting. Incomplete. ?Guttering. Rectangular shape, rectangular cross section, one side bent and folded. Orange patches from iron corrosion attached. Lead. SF 18014, Context 16837, Intervention 16831 (Pit), SG 300570 (Pit), Property 4. Mid-Roman. A lead spout of a shape that might have fitted a piece of guttering like Cat No 227 was found at Lullingstone villa in the final destruction levels of the bath (Meates 1987, 93, fig 40, 221). A strip of possible lead guttering (SF 9453) with six nails was found in a timber- and claylined tank (5917) to the rear of temple 400035 in the Sanctuary complex. A c 460 mm long, more or less rectangular lead sheet with nails through its edges found between two pipe collars at the Roman villa site at Dicket Mead near Welwyn, has been interpreted as a patch applied to a water pipe (Rook 1987, 151–2, fig 61, 4). From the villa at Lullingstone there is a 750 mm long lead sheet strip with nail holes along both long sides, interpreted as a sealing for a wooden box or a tank (Meates 1987, 92, fig 39). Only four T-clamps (SFs -559, 9137, 20461, and context 2762) and one L-clamp were identified among the iron objects from Springhead. Tools Knives Twenty-one knives have been recorded from Roman contexts, those identifiable are listed here. (Fig 113) 228. Manning 1b. Parallel-sided blade, tip formed by back turning down sharply at angle of c 30° in lower fifth of blade. Blade runs on into plate tang without shoulder. End of tang missing. Two rivet holes. Iron. SF 15685, Context 12077, Intervention 12079 (Post-hole), SG 300341 (Post-holes). Mid-Roman. 229. Manning 7b. Handle with two decorated bone plates attached by two copper alloy rivets and loop at end. Back of blade begins straight and then falls in a concave curve towards tip. The cutting edge has a gentle Scurve. Iron. SF 18759, Context 17043, Intervention 17043 (Layer), Property 3. Mid-Roman. 257 This type of knife may have been used as a razor and is common during the 1st and early 2nd centuries. Parallels are known from Britain and the Upper German province (Manning 1985a, 111–2). 230. Manning 11. Tip and end of tang missing. Iron. SF 20372, Context 16860, Intervention 16860 (Layer). Early Roman. 231. Manning 11b, blade more 12a. Spiral terminal in same plain as blade. Strongly curved cutting edge. Iron. SF 677, Context 5269, Intervention 5268 (Gully), SG 300177 (Gully). Early Roman. 232. Manning 13. Knife with slightly arched back continuing line of tang. Straight cutting edge. Iron. SF 20082, Context 17759, Intervention 17759 (Layer), Property 3. Mid-Roman. 233. Similar to Manning 13. Narrow blade (two pieces) with rectangular-sectioned tang extending at level with back, 3rd piece with flat disc-like terminal seems to be part of tang. The cutting edge rises gently towards missing tip. Iron. SF 20179, Context 12351, Intervention 12351 (Layer), SG 300323 (Layers), Property 2. Mid-Roman. See also SF 15163 which is very corroded and bent but may also belong to type 13. 234. Manning 18. Incomplete, tang missing. Convex back, falling in a straight line towards tip and turning slightly downwards towards missing tang. S-shaped cutting edge, narrowest part in end third (cf Manning 1985, pl 55, Q56, but not like outline of type 18a!). Iron. SF 9399, Context 2766, Intervention 2766 (Layer), SG 300186 (Deposits). Mid-Roman. 235. Manning 22. Socketed knife, blade with straight back and cutting edge widest in first third from tip, rising in gentle curve towards tip, the end missing. Socket hexagonal near end, opening on left side of handle. Iron. SF 15653, Context 11919, Intervention 11892 (Other), SG 300384 (Structure), Property 12. Mid-Roman. 236. Manning 23. Tanged knife with slightly concave back, curved cutting edge and upturned tip. Tang now bent almost at right-angles. Blade L 63 mm, W 23 mm, Tang L 36 mm. Iron. SF 15697, Context 16642, Intervention 16642 (Layer). Mid-Roman. 237. Manning 23. Knife with curved blade and tip turned up above level of short tang. Complete. Iron. SF 18249, Context 11320, Intervention 11320 (Layer), SG 300456 (Layers), Property 10. Early Roman 238. Parallel-sided blade with back curving down towards missing tip. The handle is of the same width, the end is broken off. Iron. SF 15664, Context 16022, Intervention 16022 (Layer). 239. Cutting implement? with spirally rolled-up tang. Width of blade 16.22 mm, diam of spiral terminal 23.34 mm. Iron. SF 20544, Context 19196, Intervention 19196 (Layer). Mid-Roman. 240. Socketed knife. Rectangular socket. Back of blade Sshaped, straight but corroded cutting edge. Rivet hole near broken end of open socket. Iron. SF 18658, Settling the Ebbsfleet Valley 258 228 229 230 231 236 237 238 232 233 240 241 234 239 235 242 0 50mm Nos. 228–241: 0 50mm No. 242 only: Figure 113 Springhead: knives and knife handles. Iron (228, 230–40), iron, bone and copper alloy (229 and 241), copper alloy (242) 241. 242. Context 17573, Intervention 17573 (Layer), SG 300675 (Layers), Property 3. Early Roman. Fragment of knife handle. Rest of tang between two bone plates decorated with longitudinal grooves. Copper alloy rivet survives near broken edge (cf Crummy 1983, 109, fig 111, 2933). Iron. SF 1800, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. Folding knife handle. Rectangular shape with openwork dog chasing hare on a base with notches in the centre and near the ends. Groove in bottom for blade to be folded into, hinge for blade was probably located behind the dog’s hind legs where the handle is damaged. SF 15728, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. A very similar rendition of the motif is known from Richborough (Bushe-Fox 1949, pl 36, 118); others such as that from the Thames at Hammersmith (Wheeler 1930, 78, fig 19, 4), and a handle from Canterbury, Marlowe Car Park, has oblique lines on the base where Cat No 242 has its notches (Lloyd-Morgan in Garrard 1995, 1034, fig 441, 452). The possible religious connotation of the motif as well as the ritual aspects of hare coursing are discussed by Smith (2006, 50; cf also Vaughan 2001, 326–30). Chapter 3 Springhead Metalwork 259 243 244 50mm 0 247 245 246 250 251 252 249 248 255 253 258 257 Figure 114 Springhead: cleavers, shears, and other iron tools (243–58) 256 254 Settling the Ebbsfleet Valley 260 Wood-working Tools 248. 249. 250. Drill bit. Spoon bit with pyramidal head with flat tip and square-sectioned stem with long round-ended spoon. Iron. SF 20197, Context 12411, Intervention 12411 (Layer), SG 300325 (Layers), Property 2. Late Roman. Drill bit. Pyramidal point and square-sectioned stem with what seems to be remains of spoon. SF 20440, Context 19197, Intervention 19197 (Layer). Mid-Roman. Drawknife blade. Straight back and edge, one end with triangular tang in same plane as blade, tapering from back but with step from cutting edge, the other tang bent at right angles to plane of blade. Iron. SF 15665, Context 16022, Intervention 16022 (Layer). A similar drawknife from Abbeville, France, which is however twice as long, has been suggested to be a cooper’s or cartwright’s tool (Champion 1916, 219, pl 3, 63645). Chisels Plate 14 X-radiograph of possible iron dividers (Cat No 259) L 92 mm. Photo: J Watson No distinction between metal, wood, or masonry working has been attempted in the identification of the chisels as their poor condition precludes certainty of identification (cf Manning 1985a, 8–9, 21). 251. Cleavers (Fig 114) 243. Cleaver with triangular blade and rounded pommel. Iron. SF 416, Context 2559, Intervention 2555 (Ditch), SG 300047 (Ditch). Mid-Roman. 244. Socketed cleaver (Manning type 3). Iron. SF 15666, Context 12374, Intervention 12374 (Layer), SG 300326 (Layers), Property 2. 245. Manning type 4. Open socket for thick handle (opening at end c 30 mm in diameter). Iron. SF 20270, Context 16134, Intervention 16464 (Pit), SG 300506 (Pits), Property 10. Early Roman. 252. 253. Mortise chisel. Square section at upper end, top missing, tapering to rectangular section at slightly splayed cutting edge which is bevelled on one side (cf Manning 1985a, pl 10, B40). Iron. SF 20392, Context 16917, Intervention 16917 (Layer). Mid-Roman. Chisel with triangular blade, rounded cutting edge and sub-rectangular shaft, its end missing. Iron. SF 20541, Context 16775, Intervention 16775 (Layer). Mid-Roman ?Chisel. Upper half rectangular-sectioned, swelling towards its middle. Axis of lower half at c 90°, splayed towards rounded cutting edge. Iron. SF 15323, Context 10647, Intervention 10646 (Pit), SG 300410 (Pits), Property 11. Mid-Roman. Leatherworking Tools Shears 254. 246. 247. Blade and handle of shears. Straight-sectioned blade with back curved towards tip. Rectangular-sectioned arm with omega-shaped spring. Iron. SF 9415, Context 2674, Intervention 2674 (Layer), SG 300164 (Deposits). Mid-Roman. Fragments of shears. Back of one blade with gentle curve, the other too short to determine. Both blades with curved sections and ovalsectioned arms. Iron. SF 9421, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. Awl. Square-sectioned tang and round-sectioned shaft (Manning type 4b). Iron. SF 20542a, Context 17157, Intervention 17157 (Layer). Early Roman. Tools of Uncertain Use 255. Possibly awl or small tracing punch, bent. Middle square-sectioned, one end subtriangular section, the other subrectangular. Iron. SF 18663, Context 11875, Intervention 11874 (Post-hole), SG 300367 (Postholes), Property 12. Early Roman. Chapter 3 Springhead Metalwork 256. 257. 258. 259. Rectangular-sectioned blade/bar, tapers to one end, which seems broken off. Possibly flat chisel for stone trimming? Iron. SF 20360, Context 16685, Intervention 16685 (Layer), SG 300578 (Layers), Property 4. Mid-Roman. Triangular bar, possibly unfinished piece. Pointed end not broken (x-ray). Iron. SF 20470, Context 16744, Intervention 16740 (Pit), SG 300546 (Pit), Property 6. Mid-Roman. Hook with suspension hole. Iron. SF 9413, Context 2674, Intervention 2674 (Layer), SG 300164 (Deposits). Mid-Roman. Possible dividers (Pl 14). Two arms joined by hinge at ?rounded ends, tapering gently towards tips which may be broken. Iron. Mineralised remains analysis (JW): Re-x-rayed to try and see what the organic layers were, but no sign of a wooden case, just layers of wood fragments and straw. Stereo-radiography revealed several parts to this object including at least two flat strips, folded together, and short lengths of chain. To identify what this object is will probably have to require the removal of the extraneous organic materials and reveal the metal pieces. SF 20295, Context 16463, Intervention 16464 (Pit), SG 300506 (Pits), Property 10. Early Roman. Fasteners and Fittings Nails No systematic analysis of the more than 3000 iron nails and nail fragments has been attempted in the course of this study. However, a rapid scan of the radiographs shows that the most common type of nail is the flatheaded variety Manning type 1b, followed by Manning type 1a with pyramidal head. There are also at least four examples of nails with a triangular head like Manning type 2. The head as the main feature of nail typology is a detail which is often affected by corrosion and thus prevents any closer identification. (Fig 115) 260. Manning type 1a. Large nail or bolt with domed head. Iron. SF 20223, Context 12630, Intervention 12630 (Layer), Property 2. Mid-Roman. 261. Manning type 1b. Iron. SF 20545, Context 19198, Intervention 19198 (Layer). Mid-Roman. 262. Manning type 2. Nail with flat triangular head. Iron. SF 20446, Context 17313, Intervention 17309 (Pit), SG 300556 (Pits), Property 5. Mid-Roman. Studs Plain studs 263. Sub-rounded flat head, eight small protrusions, alternating ribs and dots, around perimeter on underside of head; rectangular-sectioned shaft slightly bent. SF 9251, Context 400104, Channel fills. 261 Similar copper alloy studs are known from Chichester (Down 1978, 300, fig 10. 35,73) and two from Richborough, one of which was found in a group of what the excavator assumed to be the stock of a metal worker. The group was found in a deposit dated AD 55–75 and also contained a number of lorica segmentata fittings (Bushe-Fox 1932, 12, fig 2, 3980; 82, pl 11, 27). Bushe-Fox (ibid, 80) reports similar studs from the two 1st century AD ships found in Lake Nemi in Italy (McGrail 1993, 47, fig 36B) where they may have fixed the lead sheathing of the tenoned hull. A similar construction was used for the ship from Grand Conglué, France, a reconstruction section of which shows small nails or studs fixing the sheathing to the hull (Throckmorton 1972, 71, fig 10; the shape of the studs is not shown). Similar protrusions have also been found on Viking period and medieval clench nails, for example at Dublin (McGrail 1993, 24 fig 20D; 46–7 fig 36C), a preserved medieval wreck from Bole, Telemark, Norway, and an impression of such a nail noted on a late 12th century ship at Bergen (ibid, 46). It is worth noting that while the Roman nails all had an alternating pattern of four dots and four ribs, the later nails only had four dots. Similar rib-and-dot patterns are also known from Roman period iron hobnails, eg, from the Treveran temple on the Martberg near Pommern on the river Mosel in Germany (Nickel et al 2008, 313, Abb 200, 10; 321, Abb 207, 9–10). One possible explanation for the patterns on both the ship and the hobnails could be to prevent a twisting or turning of the nails during use. If Cat No 263 was indeed, as seems likely, used for fixing lead sheathing to the hull of a ship, it would be an indication that the ship so treated was built in the Roman tradition, fitted for use in more temperate climes than north-western Europe where ship-worm, against which the sheathing afforded some protection, did not occur. In the English Admiralty sheathing was not considered necessary for the rebuilt Resolution in 1698 ‘unless for a foreign voyage’ (Lavery 1987, 60), and the surgeon Robert Spotswood reports about the preparation for a voyage to the West Indies in 1744 that ‘a ship destined to remain long within the tropicks is sheathed & fill’d [ie, pliable pieces of wood fastened along the whole bottom with nails whose heads are larger than a crown piece] as thick as can but prevent the destructive corrosion of worms’ (Bingeman et al 2000, 219, after Spotswood 1793, 13). 264. Large stud with sub-rectangular-sectioned tapering shaft and large sub-circular head. Iron. SF 20084, Context 17710, Intervention 17710 (Layer), SG 300600 (Layers), Property 3. Mid-Roman. Two sets of small studs (SFs 945 and 20525) with flat circular heads and short, riveted shafts were found in a shallow drainage ditch in temple 400033 in the Sanctuary site and in property 12 of the Roadside settlement. Another variety is SF 15241 with slightly domed head and a narrow flange. Parallels from the Rhineland date to the end of the 2nd and the 3rd centuries (cf Oldenstein 1976, 171, pl 48, 516–27). Settling the Ebbsfleet Valley 262 261 263 266 267 268 262 269 264 270 260 265 271 275 272 273 277 0 274 276 0 50mm 50mm Nos. 263, 265–268 & 275: Figure 115 Springhead: nails, studs, fasteners, fittings, and lock bolt. Iron (260–62, 264, 269–274, 276–7), copper alloy (263, 265–8, 275) The most numerous variety of studs at Springhead are the 32 dome-headed studs, with head diameters ranging between 11 mm and 39 mm and clusters between 22–4 mm and 31–2 mm. Some of the larger studs and three with diameters of 15 mm, 19 mm, and 20 mm have the hint of a flat rim; one (SF 892) has a down-turned rim similar to Crummy 1983, no 3157. SF 15360 is a domed stud with two ribs and a wide flange whose iron shaft was soldered to the back. SF 20507 has a conical rather than a domed head. Thirteen studs have flat heads, with diameters from 15 mm to 37 mm. Such studs could have been used for a variety of purposes, among them decoration of boxes or upholstery (cf Crummy 1983, 117, fig 120, 3151, 3173, 3187; Oldenstein 1976, 166–7, Taf 46, 451–69). SF 18018, from the large SFB 300555 in property 5, is a copper alloy bolt with a chunky head. Chapter 3 Springhead Metalwork A possibly riveted stud with a flat, circular head and scalloped rim (SF 692) was found in the abandonment deposits above beam-slot building 300157 in the Sanctuary. The head has a decoration of three crossing lines creating six heart-shaped panels. The head was probably coated in white metal and the lines inlaid with niello. Such studs were found in mid-1st century contexts at Colchester (Crummy 1983, 129, fig 144, 4175; 133, fig 151, 4208 and 4212). A second, riveted stud (SF 1779) has a niello-inlaid floral ornament. It was found in the probable colluvial/build-up deposits by the side of the trackway 300045 which may be contemporary with the clayfloored circular structure 400028. The ornament is similar to motifs found on pre-Flavian/Flavian belt fittings (Grew and Griffiths 1991, 57 fig 4, 1–3); four exactly similar studs are reported from Wallbrook, London (Webster 1958, 87, fig 6, 151c). These and the following two studs could be components of military belt or strap fittings. Two studs with unusual heads are unfortunately unstratified metal-detector finds. SF 243 has an almost square head with two convex ridges separated by an angular groove. Possible comparisons come from Niederbieber and Weißenburg in Germany (cf Oldenstein 1976, Taf 51, 588–9). A square stud from Atworth villa, Wiltshire, has a ridged profile with a raised centre (cf Bircher in Erskine and Ellis 2008, 83, fig 23, 79). SF 258 has a hexagonal head with six incised petal leaves radiating from the centre and separated by dotted lines. SF 15345 from property 11 is a late 2nd/early 3rd-century riveted stud with domed head and wide rim (cf Oldenstein 1976, Taf 49, 561–2). 263 Various Fittings 266. Incomplete. Flat cross-section. Shape is irregular but would have been symmetrical, curved sides, lozengeshaped point at one end. Four decorative grooves, imitating leaf veins, visible on front. Hole at centre. SF 829, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. This object could be a heart- or leaf-shaped pendant similar to such items described from 2nd century contexts from Britain and the Rhineland by Oldenstein (1976, 127, Taf 29–30, esp 207–8); some from Rheingönheim, which seem typologically earlier, even have fully moulded ribs and circular holes instead of the kidney-shaped openings observed in the later types. While this interpretation would place Cat No 266 among the military equipment, another comparison allows an intriguingly different interpretation. One of the chain head-dresses from the Roman temple at Wanborough, Surrey, has attached to it a small pendant in the shape of an ivy leaf, with veins, some of which end in a punched dot, incised on both faces (O’Connell and Bird 1994, fig 23; 101, pl 15). Unfortunately, Cat No 266 is broken and its mode of attachment is no longer visible. As it is a metal detecting find from the overburden above the Sanctuary area, there is no indication of any other component indicative of a headdress. However, next to the casket in grave 6345 lay a ring with four chain fragments attached to it. While this could be the suspension arrangement for a small bowl or lamp, such chain distributors are also part of priestly chain head-dresses found, for example, at Wanborough (ibid, fig 24, pls 11 and 17) and Cavenham (Green 1976, 213). Decorated studs 265. Lion-headed, slightly damaged on edge at top. Reverse filled with lead holding central iron shaft. SF -513, Context 143, Intervention 143 (Spread), Property 7. Early Roman. This lion-headed stud was found under the floor of early Roman building 300522, as was zoomorphic spout Cat No 194. Six similar but slightly smaller bosses originally held the lock-plate of the box found in grave 6345. Two more come from the channel fills of the Ebbsfleet (SFs 1824 and 1832) and an individual, slightly more detailed example was found in the channel fills of the waterfront at the Roadside settlement. Parallels for this very common type of boss are found in a bedding trench, probably earlier than AD 75, at Fishbourne (Cunliffe 1971, 117, fig 48, 125) and at Richborough in a context pre-dating the late 3rd century (Bushe-Fox 1949, pl 44, 168). Borrill (1981, 320–1, table 46) lists another ten sites from south-east England and East Anglia where lion-headed studs were found, mainly from Flavian contexts but ranging from Vespasianic to Antonine in date. 267. Fitting, complete. Elongated diamond shape with convex scooped edges. Rectangular cross-section, two studded rivets on reverse. Very slightly bent. White-metal coating on surface. SF 845, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman See also SF 1752 with only one studded rivet. The shape of the two fittings Cat No 267 and SF 1752 appears similar enough to suggest that they belonged to the same piece of equipment, possibly a belt or bridle strap. Although the author is unaware of parallels with similar shaped plates, it might be that the Springhead fittings are evidence of a very localised production and thus indicating a situation comparable to that discussed by Oldenstein (1976, 188–9) for mid-2nd century elliptical fittings from Saalburg and Zugmantel. 268. Repoussé disc with central ornament of eight-leaved flower surrounded by a beaded field and a cabled rim. Edges slightly squashed and torn. Brass. SF 329, Context 2221, Intervention 2221 (Artefact). Settling the Ebbsfleet Valley 264 280 279 278 50mm 0 282 281 283 284 Nos. 281–282 only: 0 50mm Figure 116 Springhead: Latchlifters and keys. Iron (278–280, 283–4), copper alloy (281), copper alloy with iron (282) This may be the head of a stud or ornamental fitting of a helmet. A similar ornament can be found on the back of one of the Roman helmets found in the Thorsberg bog in Northern Germany, buried there in the first half of the 3rd century (Raddatz 1987, Nr 400, Taf 86–7). Two conical copper alloy studs or knobs with spherical finials and iron shafts (SFs 1312 and 20002) can be compared to such objects from South Shields and other places in Britain and Germany with date ranges from the 2nd to the 4th centuries (cf Allason-Jones and Miket 1984, no 3. 733–4.736). Hooks 269. 270. L-shaped wall hook. Iron. SF 20538, Context 10956, Intervention 10955 (Pit), SG 300410 (Pits), Property 11. Early Roman. Hook or bent rod with rectangular section. Iron. SF 20540, Context 16001, Intervention 16001 (-). Mid-Roman. Joiner’s dogs 271. Elongated oval plate with shafts either end, one broken off. Iron. SF 15978, Context 16861, Intervention 16861 (Layer). Mid-Roman. Chapter 3 Springhead Metalwork Brackets or tie-strips 272. Corner brace, possibly for box. Shorter arm with rounded, perforated end, slightly wider than width of strip. Iron. SF 20377, Context 16863, Intervention 16863 (Layer), Property 4. Late Roman. See also SF 15202. 280. Double-spiked loop Keys 273. 281. End of one arm bent to meet the end of the other with straight end. Iron. SF 20138, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. Rings 274. 275. Ring with stone attached by corrosion products. Iron. SF 9418, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. Fragment. Oval-shaped loop, rectangular cross-section. Two small lugs attached at sides and bent at rightangles to loop. SF 20237, Context 19565, Intervention 19565 (Layer), SG 300362 (Layers), Property 1. Handles 276. Ring with double-spiked loop. Iron. SF 15164, Context 10405, Intervention 10405 (Layer). Mid-Roman. Locks, Keys, and Latchlifters Lock bolt 277. Sprung lock bolt with semi-circular handle, the bent ends of which fit into the tubular ends of a thick strap with probably three springs extending from the other end (although this is not clear in the x-ray). Iron. SF 15249, Context 10326, Intervention 10324 (Pit), SG 300396 (Pits), Property 11. Mid-Roman. An exact parallel for this sprung bolt was found in burial 6 at Baldock (Stead and Rigby 1986, 70, fig 31, 27), where it was part of a box which is presumed to have had a sliding lid as the hinges found in the grave were found away from the remains of the box (ibid, 65, fig 28; 68, fig 30). A padlock bolt (SF 9317) was found in the area of the Ebbsfleet. Other parts of boxes are discussed under the casket from Grave 6345 (see below, Roman graves). Three bell-shaped lock-pins come from properties 2 and 3. SF 18270 has a copper alloy shaft cast in one with the head (cf Crummy 1983, 125, fig 137, 4143), while the other two (SFs 15753 and 18948) are of the same type with an iron shaft which is now lost. Context 16422, Intervention 16419 (Pit), SG 300506 (Pits), Property 10. Mid-Roman. Flat rectangular sectioned handle, the end is missing. Blade of rectangular section at 90º to plane of handle. Iron. SF 20375, Context 16861, Intervention 16861 (Layer). Mid-Roman. Zoomorphic key handle in the shape of a dog rising out of a calyx. The dog’s ears lie flat against its back, its extended front paws clutch the shoulders of a hare whose head projects to the front. Square iron terminal at other end with remains of shaft still in socket; iron corrosion filling part of flat-bottomed groove in base of object. SF 18694, Context 19026, Intervention 19026 (Other). This handle belongs to a group of keys with anthropomorphic or zoomorphic handles for which a number of examples can be quoted from Britain, in the shape of a ‘docile lion’ from Fishbourne (Cunliffe 1971, 121, fig 50, 144) and Verulamium (Goodburn 1984, 46, fig 18, 165), a boar, also from Verulamium (Adamson and Niblett 2006, 157, fig 46, 60) or a dog with a long snout from Richborough, area V, dated to AD 50–80 (Bushe-Fox 1949, 129; pl 36, 117). While the motif of a dog chasing a hare is very common in Britian, eg, the folding-knife handle Cat No 238 (for a discussion of the motif on folding knife handles see Vaughan 2001, 326–30), the style of the handle itself – a horse, lion, or dog rising out of a calyx – finds better parallels in the Rhineland, where such key handles are dated as early as the 1st century AD (eg, Simpson 2000, 145, pl 19, 1; Menzel 1966, Taf 66; Kaufmann-Heinimann 1998, 35–6, Abb 12), with a key handle from Heddernheim, Germany, particularly similar in the treatment of the dog’s face (ibid, Abb 12, 3). However, a key handle with a lion rising out of a calyx was also found at Baldock in a 3rd or 4th century context (Stead and Rigby 1986, 137, fig 59, 370), and a recent find of such a handle showing a lion with a ram’s head between its front paws was made at Winthorpe, Nottinghamshire (Worrell 2006, 439–40 fig 8). Although a metal-detector find, it should be mentioned that Cat No 281 was found in the overburden in the area of property 3, immediately adjacent to the temple in property 2 in the Roadside settlement. 282. Latchlifters (Fig 116) 278. Latchlifter with looped handle and sub-circular sectioned blade with rectangular tip bent upwards. Iron. SF 15209, Context 10687, Intervention 10660 (Ditch), SG 300386 (Ditch), Roadside ditch 1. Early Roman. 279. Latchlifter with loop at end of handle holding a ring and sub-circular sectioned blade. Iron. SF 15878, 265 283. Slide-key with rectangular sectioned handle. Edges with transverse mouldings before shank and before loop of bow. Iron nail with rectangular sectioned shaft attached to bit (separated during conservation). SF 9135, Context 6541, Intervention 6540 (Other), SG 300014 (Pits). Manning type 2. Slide-key with rectangular handle and sub-rectangular hole at its end. Bit with probably two broad teeth (thin dividing line in x-ray) (cf Manning 1985a, 93, pl 41, O52). Iron. SF 20435, Context 17043, Intervention 17043 (Layer), Property 3. Mid-Roman. Settling the Ebbsfleet Valley 266 285 286 287 288 0 50 100mm Figure 117 Springhead: iron shackle (285) and spade shoes (286–8) See also SF 1833. 284. L-shaped lift-key with rectangular handle and triangular top with convex sides and round hole. Almost square sectioned stem. Triangular bit with three teeth. Iron. SF 15387, Context 16141, Intervention 16141 (Posthole), SG 300505 (Post-holes), Property 10. A small lever lock key (SF 910), perhaps for a small box, was found in floor deposits associated with rectangular building 400029 which is probably a smithy. A tapering iron strip, probably a barb-spring padlock key, SF 18659, was found in post-hole 17921 in property 3. 286. Spade shoe Rees type 1a. Slightly less than half of a spade shoe with U-shaped blade. Inner edge of mouth and arm with V-shaped groove to hold wooden blade of spade. Iron. SF 9416, Context 2674, Intervention 2674 (Layer), SG 300164 (Deposits). Mid-Roman. A very similar example comes from Chesters Roman villa (Rees 1979, 324; 373, fig 114a). Rees mentions a further type 1a shoe of 2nd century date from Springhead (ibid, 398, table 10). A spade shoe of this type was found at Stonea together with the completely preserved wooden parts of the spade made of ash (Malim 2005, 119, fig 59). Objects Associated with Agriculture, Horticulture, and Animal Husbandry 287. (Fig 117) 285. Part of shackle with rectangular loop attached to looped end or circular eye. Iron. SF 9417, Context 2675, Intervention 2675 (Layer), SG 300163 (Deposit). Mid-Roman. 288. As only one half of Cat No 285 survives it cannot be assigned to one of Manning’s (1985a, 81–3, figs 22–3) types. The shackle was found in the layers above temple 400035 in the Sanctuary site and may have served as a hobble for animals rather than for humans. Such hobbles have been found in Viking graves in Scandinavia and were still used in Iceland until the 20th century (Roesdahl 1992, 232, No 16). An iron hoe (SF 1432, Pl 15) was found during metal detecting in the upper fill of quarry 300207. With its two-pronged fork and triangular blade it belongs to a type of hoe which is likely to be the ascia-rastrum Spade shoe Manning type 1d. Fragment of blade with deeply grooved side arm; short tang continuing from the cutting edge at the top of the arm. Back of arm straight and very wide. Iron. SF 20154, Context 12411, Intervention 12411 (Layer), SG 300325 (Layers), Property 2. Late Roman. Spade shoe Manning type 1d. Grooved side arm of a spade shoe with fragment of flat arm flange at top of arm. Iron. SF 20370, Context 16855, Intervention 16855 (Layer), SG 300553 (Layers), Property 5. Early Roman. Chapter 3 Springhead Metalwork 267 described by classical authors; a tool used for weeding, aerating the soil and tending to young plants (Rees 1979, 309–10; 341–4, figs 85–7). This type of hoe appears to have been introduced to Britain early in the Roman period and is predominantly found in forts (eg, Richborough; Bushe-Fox 1949, pl 61, 338), towns and small settlements rather than villas (Rees 1979, 348, map 6). A probable reaping hook (SF 9401) was found in the abandonment deposits overlying beam-slot building 300157. A reaping hook (SF 20072) Manning type 1 was found in property 5. Military Equipment Military awards in the form of flat, bracelet-like bands, so-called armillae, were only identified as such after the sequence of the catalogue and drawings had been fixed and are therefore discussed under bracelets (see above). In the tables (Tables 51–2) considering function groups for the various site entities at Springhead they have, however, been counted in the category ‘military equipment’. Plate 15 Iron hoe (SF 1432) L 228 mm. Photo: E Wakefield 292. Arms (Fig 118) 289. Pilum. Pyramidal head and short length of ?roundsectioned stem. Iron. SF 20422, Context 19593, Intervention 19592 (Pit). Early Roman. The identification is not certain. A similarly corroded, unprovenanced head in the British Museum has been grouped among the drill bits (cf Manning 1985a, pl 12, B59). 290. Spearhead Manning Group IIa. Socketed, tip of blade missing. Iron. SF 20202, Context 12310, Intervention 12310 (Layer), SG 300326 (Layers), Property 2. Late Roman. A catapult bolt head (SF 9464, not illus) of Manning type 1 was found in an upper layer of Viewing platform 300173, thus possibly not related to the use of that structure. These bolt heads, which were probably in use throughout the Roman period, have been found on many sites in Britain and on the Continent, especially in Germany (Manning 1985a, 175, pl 83–5). 291. Socketed ferrule with square sectioned stem, a rivet hole in the lower part of the split socket. Iron. SF 20472, Context 16744, Intervention 16740 (Pit), SG 300546 (Pit), Property 6. Mid-Roman. This ferrule has a close parallel at the Roman fort of Carrawburgh (Brocolitia) on Hadrian’s Wall with a rivet hole in a similarly low position of the socket (Manning 1976, 46, fig 13, 26). Tanged dagger. Remaining length of tang c 50 mm, tapering towards the tip; very slightly waisted blade has a flat mid-rib. Remains of right valve of oyster corroded on to top of blade/beginning of tang. Iron. SF 18367, Context 12656, Intervention 12666 (Pit), SG 300348 (Pits), Property 2. Early Roman. The shape of the blade of Cat No 292 suggests a relation to Scott’s type C daggers, although with a width at the shoulder of only 27 mm this piece is slightly narrower than these (35–45 mm; Scott 1985, 154). The blade profile of Cat No 292 is very similar to one from Kingsholm (Manning 1985a, pl 75, V19), but as that blade is not one of the standard 1st century military types, Scott is somewhat doubtful of its association with the Roman army, ‘although in the context of Kingsholm this is probable’ (Scott 1985, 159). Type C daggers are tentatively dated to the 1st century AD and probably continued into the Trajanic period (ibid, 156). 293. Scabbard mount. Straight sided frame with lateral facets, flaring outwards slightly then stepped back to upper part which is ribbed and tapers to a point. SF 866, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. Cat No 293 belongs to a well known type of scabbard mount (Carnap-Bornheim IIIC1) of the late 2nd/early 3rd century, which is found predominantly along the German limes, but also in Britain, Romania, and Gaul; considerable numbers have also been found beyond the limes in northern Germany and especially some of the Danish bogs like Vimose (cf von Carnap-Bornheim 1991, 35–6; 100, Fundliste 12, Karte 2; Oldenstein 1976, Taf 12–3, 45–50; Schuster 2006, 86–7). SF 626 may be a possible belt or sheath fitting, and the iron staple SF 1645 could be a cramp for a scabbard. Settling the Ebbsfleet Valley 268 289 294 293 290 Nos. 289–292: 50mm 0 295 291 292 296 297 298 299 300 0 50mm Nos. 293–301: Figure 118 Springhead: weapons and military equipment. Iron (289–92), copper alloy (293–301) 301 Chapter 3 Springhead Metalwork Cavalry Harness Pendants Belt- and Apron Fittings 294. 297. Horse harness fitting. Oval plate, flat section. Perforated lug attached one side, broken bar on other side. Incised decoration on one side consisting of an S-shaped twirl. SF 1799, Context 6436, Intervention 6436 (Surface), SG 300083 (Deposits). Early Roman. Better preserved examples of this pre-Flavian type with a design of leaves, tendrils, and berries on the tinned surface, which also show the likely outline of the tendril for Cat No 294, were found at Camerton (Jackson 1990, pl 7, 77) and in the Fremington Hagg hoard (Webster 1971, 113, fig 12, 19–20). 295. Trifoliate horse harness fitting. Rectangular crosssection, probably leaf-shaped form with outer leaves ending in small acorns. Remains of rectangular sectioned suspension loop at top edge. Possibly niello and white metal covering, despite cleaning any surface decoration is obscured by corrosion products. Photographed before and after conservation. SF 1766, Context 6356, Intervention 6356 (Layer), SG 300099 (Deposits). Early Roman. Pendants like Cat Nos 294–5 are normally joined to strap distributors like Cat No 213 which is here grouped under objects associated with transport. They belong to the 1st century AD (eg, from Magdalensberg: cf Bishop and Coulston 2006, 120, fig 70, 3; Newstead: Lawson 1995, 994, fig 418). In his discussion of the Canterbury harness hoard Lawson (1995, 995) suggests that such pendants had been reserved for cavalry horses, and were not generally issued but may have been awarded, perhaps for ‘bravery or outstanding equestrian skill or … used as a sign of rank.’ Armour Fittings 296. Possibly from a buckle fitting/plate or lorica fitting. Rectangular shape, one end has rounded edges, rectangular section. Two perforated lugs at one end for hinge. Two grooved ring motifs and linear border decoration. Two rivets at centre of rings (cf Bishop and Coulston 2006, 99, fig 56, 19 and 20). SF 1854, Context 6444, Intervention 1002 (Spring). SF 18760, and possibly also 18379, may be a fitting of early Roman Corbridge type lorica segmentata (cf Bishop and Coulston 2006, 99, fig 56, 7 and 10–11; Crummy 1983, 117, fig 120, 3148), while SF 940 is a rectangular hinge plate (Bishop and Coulston 2006, 99, fig 56, 15 and 22). SF 15330 resembles a strap union link from Camerton (Jackson 1990, pl 6, 63, but lacking thin backing plate). 269 Semi-circular ?fitting with decoration of two spirals, almost rectangular cross-section. On the back near this break is a small strip with subtriangular section set in a layer of irregular corrosion which might be the remnants of a solder. Corroded enamel in sunken areas of front. SF 287, Context 2183, Intervention 2183 (-). Mid-Roman. The fitting, Cat No 297, does not show any obvious method of attachment but it may have been a broken part of a larger object, unsuccessfully repaired with the ?soldered metal strip on the back. The motif of the enamelled panel is common in late Iron Age and early Romano-British art, eg, on an unprovenanced strap union (Jope 2000, pl 294k) or other bridle or belt fittings (cf Ulbert 1977, 40, Abb 3). The best comparison, however, is found on the early 1st century AD fish-head spout from a strainer bowl found at Felmersham, Bedfordshire, where a crescentic plate projects at the back of the head, presumably originally supporting a firmer attachment of the spout to the rim of a vessel (Watson 1949, 41–2, figs 3–4). Another fitting, possibly attached to two leather straps, is SF 20418 for which there are parallels from Mariaweiler-Hoven and Zugmantel, Germany (Oldenstein 1976, 201–2, Abb 6, Taf 67, 879). 298. Elliptical plate with two large riveted studs on reverse. Possible pattern on upper surface (cf Oldenstein 1976, nos 715–19). SF 9213, Context 6619, Intervention 2856 (Pit). Mid-Roman. Another type of strap fittings is represented by Cat No 267 and SF 1752. With their concave notches on all four sides they are slightly unusual but it is not impossible that they were part of military gear. They may be a variation of fittings like Oldenstein (1976) nos 733–6. 299. 300. 301. Phallus shaped fitting. Trace of circular shaft on reverse. SF 15270, Context 11347, Intervention 11347 (Layer), SG 300439 (Layers), Branch Road. Incomplete. Strap-end for military apron. Rectangular cross-section. Strip folded over at one end forming loop for ?bar. Rectangular section/plate beneath loop has three transverse lines incised. Lower part of object with crescentic openwork decoration, broken at end. On reverse is a riveted projection (cf Crummy 1983, 136–7, fig 157, 4236; Allason-Jones and Miket 1984, 227–8, no 3.790). SF 830, Context 5707, Intervention 5707 (Sanctuary overburden). Mid-Roman. Tabula ansata with loop on one long side. Fragment. Rectangular sheet with grooved edges either end, ‘ansata’ at one end, the other broken. Attached to one long side of rectangle is fragment of probably circular disc. Two prongs/rivets on reverse. (cf Crummy 1983, fig 157, 4239). SF 15414, Context 16022, Intervention 16022 (Layer). Settling the Ebbsfleet Valley 270 303 302 306 305 308 307 304 0 50mm 0 50mm Nos. 302, 303 & 306 only: Figure 119 Springhead: figurines and other objects associated with religious practices and beliefs. Lead (302–3, 306), silver (305), copper alloy (304, 308), copper alloy with iron (307) Objects Associated with Religious Practices and Beliefs (Fig 119) 302. Figurine (Pl 16, right). Female in long, flowing dress with multiple longitudinal lines indicating folds of fabric. The hair above the forehead is either laid in plats forming a wreath or she wears some sort of diadem. The left arm appears to hold a cornucopia, the right arm extends downwards to support a rudder or steering oar. A four-spoked wheel rests against her lower right leg. The distorted base is formed out of the casting sprue. Casting flashes can be seen along the sides of the outline of the figurine. On the back side, two irregular round bulges in the area of the lower back may be remains of two casting channels. Lead. SF 50957, Context 200wb (chainage: 2ATC-29A+2ATC-23B), Intervention 200wb. Mid-Roman. See also SF 20114 (Pl 16, left). The attributes cornucopia, rudder or steering oar, and wheel identify both figurines as the goddess Fortuna or Fortuna-Rosmerta. The type represented in the two pieces from Springhead is well-known in stone reliefs, for instance, from Frankfurt-Heddernheim (Nida; kept in Wiesbaden Museum; Espérandieu 1931, 86, no 129; Webster 1986, 63, fig 2, 6) or Jagsthausen (Espérandieu 1931, 425–6, no 673). A more accomplished figurine of Fortuna or Abundantia in a similar pose is represented by a small copper alloy figurine from Colchester which, however, lacks the wheel found at the left foot of the Springhead figurines (Crummy 1983, 142, fig 168, 4264). Lead figurines are generally very rare; the author is only aware of one possible lead Fortuna, reported through the Portable Antiquities Scheme and found in Kent (PAS find KENT1713, and E Durham, pers comm). The flashes from the casting have not been removed on either figurine. Slight variations in the details of the folds visible on the backs of the figurines suggest that both were created using the same model but were not cast in the same two-piece mould. The crude craftsmanship, coupled with the fact that the figurines are made of lead, suggests that they were cheap trinkets sold and, most likely, also produced, locally. That one is bent suggests it was perhaps used as a votive, an interpretation which is further supported by its deposition in the channel fills of the waterfront in property 4. In contrast, Cat No 302 appears to have been deposited closer to the metalworking area discovered to the south of the route of the A2. Chapter 3 Springhead Metalwork 303. 304. 271 Figurine of a horse. Head fragment separate, appears to be broken deliberately. Faint ridges running from the poll to the mouth, across the nose, and along the neck probably indicate the bridle. Body has casting sprue behind forelegs. Body L 31.45 mm, W 9.41 mm, H 16.09 mm. Head L 13.52 mm, W 3.69 mm. Lead. SF 18476, Context 17710, Intervention 17710 (Layer), SG 300600 (Layers), Property 3. Mid-Roman. Bell. Incomplete. Pyramidal shape with lozengeshaped suspension loop at top and four corner lugs at bottom corners, three remaining. Iron clapper missing but corrosion remaining. Corrosion on outer surface on half of object. SF 1813, Context 6445, Intervention 1000 (Spring). The use of bells is discussed below (see Roman grave goods: grave 12222). 305. Shield (Pl 17). Oval, two long edges have scalloplike decoration mid-way, flat bulges on either end. Rectangular cross-section. Small circular raised central boss. The object has not been polished to hide the rough casting surface. Silver alloy. SF 18709, Context 19026, Intervention 19026 (Other), metal-detector find, Property 3 near the boundary to Property 2. If Cat No 305 is indeed a miniature shield, it adds an as yet unknown variant to the corpus of such objects. Miniature Iron Age shields like those from the Salisbury Hoard are usually hide-shaped, oval, or hexagonal and of copper alloy (Stead 1991, figs 12–21). A chalk model of a shield was found at Garton Slack (Stead 1971, 32, pl 4d). While all but one of the Salisbury shields have separate handles riveted across the back of the shield behind the umbo, handles were not always fitted (ibid, 31 appx 2G; Knowles and May in May 1996, 270–1). The shape of the Springhead shield is closer to the outline of an enamelled plate brooch from the General Post Office site in London whose enamelled field exactly reproduces the hide-shape familiar from the Salisbury shields (Stead 1991, 25, fig 12). The 2nd century date of this brooch suggests that the Springhead shield was made and deposited during the Roman occupation of the site. Apart from the three miniature shields from the temple site at Worth, Kent, such objects are not usually found in southeastern England but are more common in ‘regions attributed to the Durotriges, Dobunni, Corieltauvi and Parisii’ (Knowles and May in May 1996, 271, fig 11.18). However, miniature shields are not confined to Britain but are also known from late Iron Age contexts on the Continent, eg, on the Titelberg in Luxembourg (Metzler et al Plate 16 Front and back view of Fortuna figurines (SF 20114) left: L 53 mm and (Cat No 302) right: L 76 mm, both lead. Photo: E Wakefield Plate 17 Miniature silver shield (Cat No 305) L 27mm. Photo: E Wakefield Settling the Ebbsfleet Valley 272 Rolled Unrolled x2 Plate 18 Lead sheet (SF 18667) L 55mm b) Unfolded, lines and marks highlighted in c) L 140 mm. Photo: T Goskar 2000, 436, Abb. 5, 1–2) or Mouzon, Dépt ChampagneArdenne, France (Müller 2002, 125, Abb 85). It is possible that the disc-shaped fitting, Cat No 316, which has been folded twice before its deposition, is another model shield of an as yet unparalleled shape. 306. Incomplete. Model helmet cheek piece? Fairly flat sheet, one edge curved with three ribs along side. Possible dolphin shaped cast decoration in centre with lines radiating from its convex edge. Lead. SF 1898, Context 6445, Intervention 1000 (Spring). Although no comparisons are known for this model cheek piece, the votive use of helmet cheek pieces is known from the 4th century BC Samnite sanctuary at Pietrabbondante, Italy, where real pieces were nailed to the walls (Müller 2002, 98–9, Abb 65). A cheek piece of an auxiliary cavalry helmet dredged out of the Tyne near South Shields shows a dolphin incised below the image of a Dioscurus (Allason-Jones and Miket 1984, 213, pl 7, 3.723). 307. 308. Thin circular token, wrapped around a nail. SF 571, Context 2855, Intervention 2856 (Pit). Mid-Roman. Sheet fragment, folded/rolled to form small container. Irregular shape. SF 15622, Context 12000, Intervention 12000 (Layer), SG 300326 (Layers), Property 2. Late Roman. The placing of a single nail or a small number of nails has been considered in the context of ritual, magic, or the warding-off of evil (eg, Black 1986, 223; Dungworth 1998, 153). The circumstances of discovery of Cat No 307, found in ritual shaft 2856 together with a chained dog, emphasise the ritual and/or apotropaic connotations of the object. Cat No 308 may be a container for a similar purpose, and the possible circular fitting, Cat No 316, may have been bent into its current shape for the same reason. A small lead rod with wider, flattened ends (SF 9276) was found in the channel fills of the Ebbsfleet. Comparable objects which are quite common in the Netherlands and north-west Germany have been interpreted as lots, ie, implements used to ‘draw straws’, or – less likely – weights or rules (cf Schuster 2006, 97–8). An object from the Roman villa at Dicket Mead near Welwyn (Rook 1987, 151, fig 61, 2) may be related, but with a length of 84 mm it would be the longest example. Metal sheet fragments – defixiones? Nine folded metal sheet fragments were identified as possible curse tablets or defixiones, eight were of lead and one of copper alloy. It was possible to unfold five of these without risk of damaging the objects and check for writing, but none was found (numbers of those unfolded are in bold): SFs 691 (copper alloy), 967, 9379, 18293, 18667, 18674, 18736, 18819, and -555. However, SF 18667 (Pl 18) was found to have numerous faint lines, some possibly letters but none forming any words. It has a lobate border very similar to those on sheets from Chelmsford, interpreted as an applied ornament (Drury 1988, 99 fig 66, 87), and Caerleon, possibly re-used as a weight (Zienkiewicz 1993, 115, fig 41, 7). An alternative may be that it derives from a bowl with a Chapter 3 Springhead Metalwork beaded edge like on some pewter bowls, eg, from Wanborough (Anderson 2001b, 119, fig 49, 1–3). Twenty-seven lead sheet fragments with one or more straight cut edges, often folded, have been included in the category metalworking in the quantification tables. Objects and Waste Material associated with Metalworking Of the 273 objects in this category, two are gold (Pl 19): SF 1459 is an end of a small ingot which has been flattened but not processed further; SF 770 is a small gold link; both are metal-detecting finds from the east side of the river. There are 224 lead and 40 copper alloy objects, mostly sheet, strip, or unidentifiable waste fragments which cannot be related to a specific production process; almost half were found in dump layers in the waterfront area of property 4 on the western side of the Ebbsfleet. Seven casting sprues may be evidence for the process of casting on site but may equally well have been part of scrap metal collected for reuse; two found in the watching brief south of the A2 may relate to either of the workshops located in this area in the 2nd and 3rd centuries (cf Boyle and Booth in Boyle and Early 1999, 40–1; Penn 1968b, 257). Of the six iron bars, three come from property 2, c 100 m north-east of these workshops. While none of the others shows any typologically distinctive detail, Cat No 309 is a sword-shaped bar (Allen 1967, 308–10) belonging to Crew’s Bearwood type (Crew 1995). Such bars are assumed to have been produced and deposited in the Middle Iron Age (between 250/200 BC and 50 BC and the birth of Christ; Hingley 1990, 92, 111; 1997, 13) and are found mainly in southern Britain and the West Midlands, with a small number found further east along the Thames as far as Hammersmith and Bigbury in Kent (Allen 1967, 313, fig 2; Hingley 1990, 93, fig 1; 96, fig 2). Considering the possible symbolic/ritual character of such bars, it is worth noting that Cat No 309 was found in a late Roman layer in the overburden above the temple in property 2, where it may have been curated for several hundred years before its final deposition. (Fig 120) 309. End of sword-shaped currency bar with tubular socket. Iron. SF 15864, Context 12312, Intervention 12312 (Layer), SG 300326 (Layers), Property 2. Late Roman Plate 19 Gold link (SF 770) left: L 11mm and gold ingot end (SF 1459) right: L 13 mm. Photo: E Wakefield 273 Objects the Function or Identification of which is Unknown Possible Toilet or Medical Equipment (Fig 120) 310. Octagonal shaft, widens in middle, broken one end, at other is moulded collar and square ‘plate’ with groove along side, knob at top (or broken part of continuing shaft?). SF 18019, Context 17194, Intervention 17185 (Other), SG 300555 (Structure), Property 5. Mid-Roman. Possible Household Utensils 311. Possibly part of fire shovel. Potsherd attached by corrosion. Iron. SF 20083, Context 17710, Intervention 17710 (Layer), SG 300600 (Layers), Property 3. Mid-Roman. A fire shovel with a handle similar to Cat No 311 was found in the Mithraeum at Carrawburgh (Manning 1976, 39; 56, fig 23, 149), another one from Verulamium has a twisted stem (Frere 1972, 165, fig 60, 6). Possible Fittings 312. Incomplete. Fitting for unidentified use. Decorated central circular disc, flat. Outer band with radiating grooves, central motif within single line border is a cross with three billets arranged in clover leaf-like fashion at either end of four arms. Central disc void on reverse but outer edges with curved flanges. On one side a leaf/fan-like shaped arm extends, rivets at tip and at two points along bar close to decorated disc (visible in x-ray). This was probably mirrored on the opposing side of the disc but is now broken; only one rivet hole still visible in x-ray. SF 819, Context 5901, Intervention 5917 (Other), SG 300128 (Pit). Mid-Roman. This object from clay-/timber lined tank 5917 to the rear of temple 400035 in the Sanctuary site comes from the lower fill and should thus have a firm mid-Roman date. It may have a remote resemblance to harness fittings like one from the Canterbury hoard (cf Lawson 1995, 987, fig 414, 145), but the present author has so far not come across a fitting with comparable projections at the back. Elements of its decoration suggest that it is considerably more recent than the context date would suggest. The central ornament resembles what would heraldically be described as a cross bottony. A fingerring from a 7th century grave at Tawern-Röler, Germany, has a similar ornament, although featuring a cross potent (Böhner 1958, Taf 22, 17). However, a probably Roman bronze clasp from a necklace found in the area of the late Roman/Merovingian cemetery ‘bei der Niederburg’ in Gondorf, Germany, has stylised Settling the Ebbsfleet Valley 274 313 312 310 309 314 316 315 311 319 318 320 317 321 0 50mm 50mm 0 Nos. 310, 312–3, 316 & 321 : Figure 120 Springhead: iron currency bar (309) and objects of unknown or uncertain function. Iron (311, 314–5, 317–20), copper alloy (310, 312–3, 316), and silver (321) leaf-shaped arms closely comparable to Cat No 312 (Ludwigshafen Wilhelm-Hack-Museum, Coll Wilhelm Hack No a1405; M Schulze-Dörrlamm, pers comm). 316. 313. 317. 314. 315. Incomplete. Double loop-in-loop chain, broken both ends. SF 889, Context 5938, Intervention 5938 (Surface), SG 300108 (Floor). Mid-Roman. Possible handle. Iron. SF 15686, Context 12077, Intervention 12079 (Post-hole), SG 300341 (Postholes). Mid-Roman. Possible handle, one end socketed? With remains still in socket. Iron. SF 18016, Context 17194, Intervention 17185 (Other), SG 300555 (Structure), Property 5. Mid-Roman. 318. 319. Disc shape fitting, two sides folded in. Raised decoration of two concentric circles, six circular perforations around edges. SF 15155, Context 10399, Intervention 10399 (Other). Perhaps a strip or binding. Iron. SF 15875, Context 16022, Intervention 16022 (Layer). Sub-rectangular sectioned hollow bar, perhaps part of a lock. Two halves of rectangular void through centre stand at right-angles to each other. Iron. SF 20056, Context 11974, Intervention 11892 (Other), SG 300384 (Structure), Property 12. Mid-Roman. Bar of sub-rectangular section with ends bent at right angles to form points, a third point wrought out of the bar in its middle. Perhaps a three-pronged joiner’s dog. Chapter 3 Springhead Metalwork 275 324 322 323 325 329 326 50mm 0 327 328 Figure 121 Springhead: objects of unknown or uncertain function. Iron (322, 324–7) and lead (323 and 329) 320. 321. Iron. SF 20371, Context 16855, Intervention 16855 (Layer), SG 300553 (Layers), Property 5. Early Roman Stem/tang of sub-rectangular section, tapering to end. Other end widens sideways and to front to form flat ?plate at oblique angle to stem, broken off on two sides. Iron. SF 20258, Context 16482, Intervention 16481 (Pit), SG 300677 (Pits). Early Roman. Fitting. Incomplete. Small rectangular plate. Dotted border decoration, two internal panels created by central line of ring-and-dot; both panels with curved line with transverse grooves and larger ring-and-dot on the inner side of the end points. One corner missing. One side bends up slightly. Silver. SF 9186, Context 400104, Channel fills. The small silver fitting Cat No 321 was found in the channel fills of the Ebbsfleet. Its decoration, similar to but not the same as on military belt plates (eg, Böhme 1974, Taf 129, 5; 139, 16; 142, 7), would suggest a late Roman date. Any form of attachment must have been fitted to the now missing parts. Of similar outline but lacking the oblique opening is a lead shovel of unknown purpose from Lullingstone villa, found in a 3rd century level (Meates 1987, 93, fig 40, 226). 324. This blade may have belonged to a knife or perhaps more likely a pair of shears, but the preservation prevents certainty of identification. 325. 326. 327. Possible Tools (Fig 121) 322. Incomplete. Drawknife? Blade set at right-angles to tang at line of back, its lower sides curving out towards blade. Iron. SF 822, Context 5451, Intervention 5450 (Ditch). Early Roman. 323. Fragment, perhaps fitting for a plane’s bench mouth. Subrectangular shape with rounded corners, wedgeshaped in section with oblique opening in the middle. Lead. SF 18922, Context 17710, Intervention 17710 (Layer), SG 300600 (Layers), Property 3. Mid-Roman. Lanceolate blade with short tang. Iron. SF 15197, Context 11442, Intervention 11441 (Ditch), SG 300366 (Ditch), Property 12. Mid-Roman. 328. Possibly a knife blade with a wooden handle (analysed with SEM, but too degraded to identify), attached to the curved blade by a copper alloy rivet. Iron. SF 20535, Context 10242, Intervention 10233 (Ditch), SG 300387 (Ditch), Roadside ditch 1. Early Roman. Incomplete. Perhaps nail extractor or stem with looped end. Iron. SF 20539, Context 16001, Intervention 16001 (-). Mid-Roman. Open socket at one end, the other bent at right angles and slightly tapering. Iron. SF 18009, Context 11281, Intervention 11211 (Pit), SG 300465 (Pits), Property 10. Mid-Roman. Perhaps punch or nail. Iron. SF 15969, Context 16836, Intervention 16835 (Beam-slot), SG 300570 (Pit), Property 4. Possibly Associated with Metalworking 329. Oval shape, flat. Stamp decoration on one side, very worn. Lead. SF 9198. 276 Settling the Ebbsfleet Valley Roman Grave Goods by Jörn Schuster with Jacqui Watson (mineralised remains analysis) 3. Grave goods are not counted in the catalogue sequence; they are listed by individual grave, and on the grave plans in Vol 1 the objects are referred to by their small finds numbers (individual nails not numberd on plans, illustrated SFs in bold). Grave 3142 (Vol 1, Fig 2.54) SF 1268. Leg ring with expanding fastening. Iron. Context 3182, Intervention 3142, SG 300048 (Burial). Mid-Roman. Grave 3142 also contained one hobnail (SF 1269). 4. 5. Grave 3428 (Vol 1, Fig 2.54) SF 1584. Fragment of knife blade; tip, shoulder, and tang missing. Iron. Context 3429, Intervention 3428 (Grave), SG 300048 (Burial). Mid-Roman. Grave 6104 (Vol 1, Fig 2.20) 1. SF 913. Mineralised remains analysis (JW): nine large iron nails with sections of mineral preserved wood: oak (Quercus sp.). At least one nail represents the join between two radial surface planks, c 22.4 mm thick. Context 6012, Intervention 6104 (Grave), SG 300078 (Burial). Early Roman. 2. SF 914. Seven iron nails, one with stone attached by corrosion products, one with dog bone. Mineralised remains analysis (JW): six large nails with sections of mineral preserved wood: oak. At least one nail represents the join between two radial surface planks, c 24.2 mm thick. Context 6012. 3. SF 9433. Irregularly shaped lump, possibly waste. Copper alloy, but probably with high lead content. Context 6012. ‘Casket burial’ Grave 6345 (Vol 1, Fig 2.20; Fig 122 and Pl 20) 1. SF 970. Lock-plate. Rectangular lock-plate: one corner badly damaged. Rectangular cut-out (L 18.9 mm, W 4.8 mm), for clasp, and L-shaped keyhole (L 15 mm, W 16.2 mm) both slightly off centre. Small club/L-shaped plate protrudes from below keyhole at right-angles to plane of plate. Plate originally attached with six lionheaded studs (three either side), one in situ, three separate but retained, max diam 19.1 mm. Nail shafts remain in two central opposing holes but heads no longer attached – separate studs 2 and 5 probably came from these as corrosion patterns line up. Separate iron nail (L 23.72 mm) with wood attached. On reverse of plate fragments of mineralised wood adhere in three places. See associated SFs 971–4. 2. SF 971. Lock bolt. From lock-plate 970. Now in two pieces which join. Six circular holes (two lines of three) in central rectangular block, one end continues tapering 6. from rectangular cross-section to square, the opposing end is rectangular sectioned but with slight cut-out on one side. Slight white discoloration at break. SF 972. Studs. From lock-plate 970. Lion-headed studs 1 and 4. Square sectioned shafts of nails protrude from reverse. SF 973. Nails. Three nails, square sectioned, flat heads with mineral preserved wood: beech (Fagus sp.). These represent the joint between two sides, with the thickness of one being 11.5 mm. L 27.61 mm, L 34.12 mm, and L 29.20 mm. Iron. SF 974. Plate. From lock-plate 970. Large fragment: rectangular but broken along one edge, L 55 mm, W 27 mm, H 3 mm; small perforation visible in one corner, matching hole on opposite corner with radial surface of wood preserved. Smaller fragment: L 19 mm, W 18 mm, T 2 mm. Two edges form corner, mineralised wood attached on one side, possibly small perforation. Iron. This is the cover for the lock mechanism on the inside of the box as no leather is preserved between the metal and the wood. Possible fragment of human mandible attached. Context 6355, Intervention 6345 (Other), SG 300079 (Graves). Early Roman SF 975. Suspension chain. Rectangular sectioned ring with four chain fragments attached, links clearly visible in x-ray. One further chain fragment separate. Tips of some have iron corrosion/staining attached. Context 6355. Suspension chains like this could have multiple purposes, eg, for the suspension of bronze lamps (cf Ward-Perkins and Claridge 1976, no 135) or scale/balance pans (ibid, no 248). Distribution rings for a chain with much larger rings are also part of the priestly chain head-dresses found at Wanborough (O’Connell and Bird 1994, fig 24) and Cavenham (Green 1976, 213). Notes on the casket from Grave 6345 by Jacqui Watson The cremation casket itself is c 250 x 175 mm from plan, but the height is uncertain. It was a wooden box with the sides nailed together, then covered with leather and decorated with copper alloy fittings, including a lock with six lion-headed studs. This is a common form of casket construction found with early Roman burials (Table 51). The box was made from beech, like most other cremation caskets. Although the burial appears to be complete and undisturbed the casket itself lacks a number of expected fittings. There are no hinges or hasp to connect the lid to the sides of the casket. Also only three nails remain, which is barely enough to hold together a single corner of a casket. Possibly the casket is incomplete, like the examples from Brougham, Cumbria (Mould in Cool 2004, 393–6), or was originally made with dowelled or articulated joints that did not require nails. A box with an identical lock-plate held by six lion-headed studs was found in the late 1st century grave 171 in the St Pancras cemetery at Chichester (Down 1971, 85, fig 5.16, 171k). Chapter 3 Springhead Metalwork 1 4 2 5 3 6 277 970 973 1 4 2 5 974 972 6 = Mineralised wood 971 0 50mm 975 Figure 122 Lock-Plate (SFs 970–974) and suspension chain (SF 975) from casket burial in grave 6345 at Springhead Plate 20 Copper alloy lock-Plate (SF 970) L 109 mm, with lock bolt (SF 971), lion-headed studs (SF 972), and iron nails (SF 973) from casket in grave 6345. Photo: E Wakefield Settling the Ebbsfleet Valley 278 Grave 10079 (Vol 1, Fig 2.97) SF 15084. Snaffle bit. Incomplete, one ring and one half of bit with tubular loop remaining (cf Cat No 210 and Manning 1985a, pl 28–9, H11 and H14). Iron. Context 10080, Intervention 10079 (Grave), SG 300363 (Burials). Late Roman. Grave 10079 also contained at least 25 coffin nails and two hobnailed boots (x-rays of hobnailed boots (left boot SF 15031, 9158–61, right SF 15032)). The snaffle bit is not marked on the grave plan as it was recovered from the backfill. Grave 10150 (Vol 1, Fig 2.97) Weight (not illus). Biconical shape, perforation at one end for suspension loop; the hole at the other end is blocked, probably by remains of the shaft of the suspension loop. 32.8 g. SF 15113, Context 10151, Intervention 10150 (Grave), SG 300363 (Burials). Late Roman. Grave 12222/Pot Burial 12224 SF 15748. Small bell (not illus), double grooved line decoration around circumference towards base, triangular suspension loop. Iron clanger in place but not free moving. Context 12223, Intervention 12222 (Artefact), SG 300344 (Burial), Property 2. Mid-Roman. A similar bell was found in the richly furnished late Flavian grave II at Grange Road, Winchester (Biddle 1967, 243, fig 9, 23). The grave also contained one nail (SF 20121). Both the bell and the single nail may have had an apotropaic use in this context. With regard to the nail, a more convincing example can be quoted from West Thurrock, Essex, grave 17062, where a single copper alloy nail was found in a clearly non-functional position in a wooden box containing the inhumation burial of a small child (Schuster 2009, fig 10, 19597; for the apotropaic use of bells worn by children as well as soldiers cf Forrer 1919, 1030; Philpott 1991, 163; Schuster 2006, 94). Table 51 Springhead metal finds: Romano-British caskets Site Skeleton Green, Hertfordshire # (Borrill 1981) Godmanchester, Cambridgeshire # (Watson forthcoming) Springhead, Kent # Mansell St, London (Watson 1997) Dimensions (mm) Materials c 300 x 250 x 150 410 x 290 x 150 250 x 175 x >80 240 x 170 x 90 # = caskets used to hold cremation burials Beech, small iron nails, leather cover, decorative brass fittings Beech, small iron nails, leather cover, decorative CuA fittings Willow or poplar, iron nail, leather cover, decorative CuA fittings Discussion of the Iron Age and Roman Metal Small Finds Assemblage The Metal Small Finds Assemblages compared across the Various Spatial Entities The following section will examine the differences and/or similarities discernible in the composition of the functional categories of the metal small finds assemblages from various spatial entities excavated at Springhead. Admittedly, these categories only provide a rather crude tool for comparison, as some like ‘fittings’ include a wide range of objects, while others like ‘writing’ or ‘weighing and measuring’ are very narrowly defined and thus generally less numerous. However, Table 52 clearly demonstrates that the quantities of the different categories vary considerably across the site, and this allows some consideration of the nature of the activities carried out within the different site entities. Late Iron Age enclosure 400012 More than half the metal finds from enclosure ditch 300030 (mainly the upper fill) forming part of late Iron Age enclosure 400012 are personal objects, including seven brooches which are exclusively mid-1st century AD types, with some, like rosette brooch Cat No 30, very likely of pre-Conquest date. Other personal items comprise finger-ring Cat No 155, two hobnails, and two 1st century strap fittings covered in silver foil. A knife is the only tool from the ditch. A perforated token (SF 1263), possibly a very worn or deliberately smoothed coin, can be compared to similar objects interpreted as amulets from temple 10 at Balkerne Hill, Colchester, which has recently been attributed to Mercury (Crummy 2006, 64, fig 32, 6–7). Viewing platform 400045–8 Of the 40 objects from this feature, 24 belong to the early and 16 to the mid-Roman phase. Overall, a third of these are personal items. Interestingly, all personal items of the early period are brooches, mainly post-Conquest types but including two of the four penannular brooches from Springhead. Other finds include a possible lead spindle whorl, two knives, and a probable iron flesh hook. No brooches were found in mid-Roman layers; the items from this phase include two pins, an iron toilet set, and a knife. ‘Bakeries’ 400037–41 These five features only contained 24 objects. Of the ten personal items, six are brooches dating to the 2nd half of the 1st century AD. Feature 400039, where six out of the nine objects were personal, yielded two finger-rings; 400041 contained a two-piece Colchester brooch, a cramp-like pot-mend and a double-spherical object (SF 623) of unknown function, perhaps a pin head. Early road 400009 and associated features Of the 35 objects in this group, 21 were found in or on the road and only four in the associated ditches. The p g g p Table 52 Springhead metal finds: total no metal small finds of Late Iron Age and Roman periods Site entity LIA enclosure 400012 Viewing platform 400045–8 ‘Bakeries’ 400037–41 Early road & assoc. features Post-road, pre-sanctuary Sanctuary complex Pool & area to east Toilet/ medical Textile Household Metrology Script Transport Construction Tools Fittings Agricultural Militaria Votive Metalworking Unknown Total 12 1 4 6 23 52.2% 4.3% 17.4% 26.1% 100% 13 1 1 1 3 5 3 13 40 32.5% 2.5% 2.5% 2.5% 7.5% 12.5% 7.5% 32.5% 100% 10 1 4 9 24 41.7% 4.2% 16.7% 37.5% 100% 8 1 1 4 1 7 1 2 10 35 22.9% 2.9% 2.9% 11.4% 2.9% 20.0% 2.9% 5.7% 28.6% 100% 11 1 1 1 9 1 1 7 32 34.4% 3.1% 3.1% 3.1% 28.1% 3.1% 3.1% 21.9% 100% 48 8 2 3 1 1 1 10 33 5 3 1 4 66 186 25.8% 4.3% 1.1% 1.6% 0.5% 0.5% 0.5% 5.4% 17.7% 2.7% 1.6% 0.5% 2.2% 35.5% 100% 83 7 1 6 7 2 1 50 3 4 5 50 219 37.9% 3.2% 0.5% 2.7% 3.2% 0.9% 0.5% 22.8% 1.4% 1.8% 2.3% 22.8% 100% 317 28 12 34 22 10 4 1 21 245 6 18 6 45 300 1069 (ARC SPH00) 29.7% 2.6% 1.1% 3.2% 2.1% 0.9% 0.4% 0.1% 2.0% 22.9% 0.6% 1.7% 0.6% 4.2% 28.1% 100% Springhead total 490 20.3% 56 2.3% 18 0.7% 64 2.7% 41 1.7% 18 0.7% 25 1.0% 3 0.1% 70 2.9% 582 24.1% 9 0.4% 37 1.5% 11 0.5% 268 11.1% 720 29.9% 2412 100% Area west of pool (ARC SHN02+51724) Property 2 Property 3 Property 4 Property 5 Property 10 Property 11 Property 12 Roadside ditches 1–3 172 28 6 29 17 8 19 2 49 326 3 19 4 215 401 1298 13.3% 2.2% 0.5% 2.2% 1.3% 0.6% 1.5% 0.2% 3.8% 25.1% 0.2% 1.5% 0.3% 16.6% 30.9% 100% 15 2 3 1 3 1 9 35 1 2 1 12 57 142 10.6% 1.4% 2.1% 0.7% 2.1% 0.7% 6.3% 24.6% 0.7% 1.4% 0.7% 8.5% 40.1% 100% 36 9 1 11 6 1 4 6 61 4 2 123 98 362 9.9% 2.5% 0.3% 3.0% 1.7% 0.3% 1.1% 1.7% 16.9% 1.1% 0.6% 34.0% 27.1% 100% 8 1 2 1 2 17 2 23 56 14.3% 1.8% 3.6% 1.8% 3.6% 30.4% 3.6% 41.1% 100% 4 3 1 3 2 9 22 18.2% 13.6% 4.5% 13.6% 9.1% 40.9% 100% 12 1 1 1 5 10 28 58 20.7% 1.7% 1.7% 1.7% 8.6% 17.2% 48.3% 100% 20 6 1 1 1 2 20 2 4 28 85 23.5% 7.1% 1.2% 1.2% 1.2% 2.4% 23.5% 2.4% 4.7% 32.9% 100% 13 4 2 1 1 1 7 59 1 10 39 138 9.4% 2.9% 1.4% 0.7% 0.7% 0.7% 5.1% 42.8% 0.7% 7.2% 28.3% 100% 11 2 1 2 14 2 1 14 47 23.4% 4.3% 2.1% 4.3% 29.8% 4.3% 2.1% 29.8% 100% Chapter 3 Springhead Metalwork Spring and channel fills Personal Combined totals for the whole of Springhead, the Ebbsfleet, and the areas to the east, the areas to the west, as well as selected individual site entities; therefore, individual numbers do not add up to make combined total 279 Settling the Ebbsfleet Valley 6 5 45 23 5 31 24 41 6 186 100.0% 2 1 14 11 3 14 8 10 3 66 35.5% 3 3 10 5.4% 2 1.1% 8 4.3% 1 1 2 4 3 1.6% 1 0.5% 1 1 0.5% 1 0.5% 6 7 3 7 2 33 17.7% 1 1 1 1 1 1 Pits 2925, 2954 & 5353 Trackway 300045 Enclosing ditch Ritual shaft Portico building Pit align N of P Bld Temple Late deposits Assoc features Total 3 2 15 5 2 4 4 12 1 48 25.8% 2 5 2.7% 1 3 1.6% 1 0.5% 1 1 1 1 2 1 2 8 3 3 4 2.2% Total Votive Tools Construction Transport Metrology Household Textile Toilet/ medical Personal Function Group in Sanctuary complex Table 53 Springhead metal finds: small finds per function group from features in Sanctuary complex Fittings Agricultural Militaria Metalworking Unknown 280 seven brooches found in road surface 300082 are probably all pre-Flavian, with Cat No 70 as the latest brooch. Other objects from the road include the small nail cleaner Cat No 175, the knife handle fragment Cat No 241, a lead spindle whorl (SF 9136), a chain fragment and link, as well as 1st century horse harness pendant Cat No 294. While it is possible that all these objects were lost during the use of the road, this situation presents a stark contrast to the western side of the Ebbsfleet where roadside ditches 1–3 contained 47 objects and the road surface of Watling Street only six. Nine objects are associated with graves 6104 and 6345, the latter including a casket with a lock-plate fixed with lion-headed studs and a chatelaine or suspension chain with ring. Post-road – pre-sanctuary features 400027 The 32 objects from features associated with this phase include five mid–later 1st century AD brooches, three pins of which probably two date to the 2nd century, a probable lorica fitting, a small copper alloy ring or bead, and a slightly larger ring, as well as finger-ring Cat No 158 and a nail cleaner. A very corroded lead spindle whorl, two latchlifters, and a handle of a probable third as well as two studs and a circular fitting with suspension loop also belong to this group which consists of levelling deposits pre-dating the construction of the temple (400035). Sanctuary complex A breakdown of the metal small finds assemblages of the various structures associated with the Sanctuary complex is shown in Table 53. Of the three pits, which are earlier than the remainder of the Sanctuary complex, only pit 2925 contained datable objects; a very fragmented Nauheim-derivative and a two-piece Colchester brooch, suggesting a date in the 2nd half of the 1st century AD. This pit also yielded a ferrule (SF 665) with thread-like grooved decoration. The function of such objects remains as yet unexplained but, apart from Silchester, most have been found on sites with military associations (Williams 2005, 11). Trackway 300045 yielded only five metal objects, but two of these are brooches found in different layers: mid1st century Colchester brooch Cat No 17 belongs to the early Roman phase; a later two-piece Colchester brooch was found in a layer dated to the early 2nd century. With a total of 45 objects, the ditch (400017) that encloses the Sanctuary to the east produced the largest number of metal finds from this complex. This figure includes a leg ring and a pair of shoes represented by 12 hobnails (counted as one in the table) from Grave 3142 and a knife blade from Grave 3428. The personal items are dominated by eight brooches, mainly developed forms of two-piece Colchester brooches of the later 1st century but including two 2nd century brooches (Cat Nos 100 and 119). Other personal items include a bracelet, a plain finger-ring, and at least two hobnailed shoe-soles. Two fragmented toilet instruments, one of only four metal vessel fragments from Springhead, as Chapter 3 Springhead Metalwork well as two knives, a cleaver, and an iron staple that may belong to a sheath or scabbard are also among the identifiable finds from the ditch. Ritual shaft 2856 (see Vol 1 Chap 2, Fig 2.55) contained 23 metal small finds which were not distributed evenly in the 15 fills. The five personal objects include four fragmented hobnailed shoes, one of which was found in a ceramic vessel. It was not possible to establish the nail pattern of the soles, but as two groups of hobnails were found in layers 6619 and 5285 – the 2nd and 3rd lowest from the bottom of the shaft – a pair of shoes may have been deposited in each instance. Layer 6619 also contained military fitting Cat No 298 and a possible hinge for a door or window pivot (SF 9411). A Hod Hill brooch of the second half of the 1st century AD was found in layer 2986. Among the 20 dog skeletons from the shaft, two were buried with their iron chains or leads still attached; interestingly, one was found in the lowest layer of the shaft while the other came from the uppermost layer containing articulated or semi-articulated animal bones (2855). The latter layer also contained the most intriguing votive object from Springhead: a circular copper alloy token folded around a nail (Cat No 307). As mentioned above, the placing of a single nail or a small number of nails has been considered in the context of ritual, magic, or the warding-off of evil (eg, Black 1986, 223; Dungworth 1998, 153), and it is difficult to imagine that such properties were not also invoked in this instance. Only five objects were found in the portico building (400020–1), including a two-piece Colchester brooch and a hobnail, while the pit alignment to its north (400023) contained 31 metal finds. Pit 2214 is one of the few features in the Sanctuary complex which contained toiletry equipment: a ligula and a nail cleaner. Pit 2227 is the only of this group to contain brooches, and both Cat Nos 14 and 22 date to the middle decades of the 1st century AD; a slightly later pin (Cool G12) was found in the layer above, and the uppermost layer contained one of the only eight metal needles from Springhead. No datable metal finds were recovered from pit 2236 which contained a blade fragment, a possible small weight, a hobnail and a strip binding fragment. A possible sheath fitting strip (SF 626) is the only metal object from pit 2931. The 24 metal objects from features associated with the temple (400033) include a fragmented Nauheimderivative brooch, the hare brooch Cat No 128 (Pl 8), a pair of tweezers with an unusual pattern on its arms (Cat No 181) as well as a nail cleaner, a toilet set, and a fragment of a rectangular mirror. An S-bent hook (SF 9405) may have been a cauldron hanger, while a bent lead strip (SF 9453) with six iron nails may have had a constructional function, perhaps as a guttering. The three objects possibly associated with metalworking include two wire fragments and a folded lead sheet, and thus may simply be waste of some unknown process. With 41 records the amount of metal objects from later deposits overlying the temple is almost twice as large as that of the temple contexts. The 12 personal objects 281 include seven brooches, predominantly two-piece Colchester and Hod Hill types dating to the later 1st century AD but also one headstud brooch (Cat No 95) of the late 1st/early 2nd century. Two pins are of 2nd century types. Of note are three agricultural objects: a spade shoe (Cat No 286), the fragment of a shackle (Cat No 285), and a reaping hook, as well as an iron wool comb (Cat No 188), two shears (Cat Nos 245–6), two knives, a hook (Cat No 258), and a lynch pin (Cat No 224). While the personal objects may have been deposited in the vicinity of the temple as votive offerings, the agricultural and other tools may be indicative of activities related to the keeping of animals and the processing of wool. The dating of the brooches and the pins suggests that these objects were residual in these layers, but whether or not this is also the case with the agricultural and other objects is difficult to ascertain as it is not inconceivable that they may relate to activities associated with the use of the temple; equally, they could well belong to a phase after the ritual use of the temple ceased and the area was used for more domestic/agricultural purposes. The Ebbsfleet (spring and channel deposits) 300009, 400007–8, 400068 The second largest assemblage of metal small finds was recovered from the spring and channel deposits, yielding 217 objects. In absolute numbers, it contained the largest amount of personal objects, 83 accounting for 37.9% of the Ebbsfleet assemblage, which is the second highest percentage after the late Iron Age enclosure 400012. Of these 83 personal objects 68 are brooches, mainly dating to the latter half of the 1st century AD but including types ranging in date from the early/mid-1st to the later 2nd century. Among the early types are two Langton Down, two rosette, four early plate, six Aucissa, five Hod Hill, five Nauheim-derivative brooches, and one Maxey-type brooch. Very remarkable is the stark difference in the numbers of Colchester and two-piece Colchester brooches: there is only one of the former but 28 of the latter. Based on this evidence, it appears that the deposition of brooches in the Ebbsfleet was introduced only after the Roman Conquest as even the earliest of the above mentioned types would straddle the Conquest. By contrast, 2nd century brooches are already much less frequent with only nine examples, but these include a number of the more colourful enamelled types like Cat Nos 98, 103, and 124 as well as the unusual rhomboid plate brooch Cat No 133, a tutulus, a trumpet-headed, and one complete and two fragmented T-shaped brooches. No brooches of the full 3rd century or later were found in the Ebbsfleet but other personal items like the two bracelets Cat Nos 148 and 149 and finger-ring Cat No 163 belong to this later period, while the five pins or fragments of such may belong to the earlier phase and lunulate fitting Cat No 167 to the 2nd/3rd century. The seven toiletry implements from the Ebbsfleet include three joining fragments of a round mirror (SF 9143), two nail cleaners, two tweezers, a fragmented 282 Settling the Ebbsfleet Valley toilet set, and a pestle (SF 9154) of a cosmetic grinder set. Two perforated lead discs (SFs 1722 and 1874) are counted among textile equipment as they may have been spindle whorls but this identification is rather tentative. Among the six household items is the aforementioned boss as well as a vessel handle (SF 9001), a fragmented oval spoon bowl (SF 9265), and the bow tie-shaped potmend SF 9172. The six items of metrological use include a balance bar (Cat No 197), a lead weight (Cat No 200), and four lead discs which may initially either have been used as weights or tokens (SFs 9249 and 9275) or possibly seals (SFs 1893 and 9068) before being deposited in the Ebbsfleet. Writing is attested by a seal-box base and a lead seal (Cat No 209). The three military items include armilla Cat No 146 (a military award of the immediate post-Conquest period), buckle or lorica fitting Cat No 296, and a buckle (SF 1853) similar to examples from South Shields (Allason-Jones and Miket 1984, 193–4 no 3.628) and Corbridge. The category fasteners and fittings is the second largest finds category from the Ebbsfleet, accounting for 50 objects. These include, amongst others, two slide keys, four sheet fragments which may have been part of boxes, as is likely for two lion-headed studs (SFs 1824 and 1832) similar to Cat No 265. A boss with a square back plate (Cat No 192) was certainly part of a box and is therefore listed under household items. Other fittings from the Ebbsfleet deposits include a T-clamp, six copper alloy, and not more than seven iron nails, a plain handle (SF 1899), and a decorated strap (SF 613) which may have been part of a box, perhaps as a handle, six copper alloy rings, and a ferrule (SF 1873) similar to SF 665 found in the Sanctuary complex but with only three ridges. Of particular interest is the stud, Cat No 263. Such studs have been used to fix the lead-sheathing of ships built in the Roman ship-building tradition as opposed to the indigenous Romano-Celtic tradition in which sheathing was not used. The stud was found in the western area of the Ebbsfleet near properties 2 and 3, which is where the landing places for boats and small ships sailing up the Ebbsfleet is supposed to be located. The four items associated with religious practice include a bell (Cat No 304), a lead sheet in the shape of a miniature cheek piece of a helmet with the image of a dolphin on it (Cat No 306), a small rod (SF 9276) which may be a ‘lot’, ie, an implement used to ‘draw straws’, or – less likely – a weight or ruler, similar to such objects from the Netherlands and Northern Germany (cf Schuster 2006, 97–8), and a folded copper alloy sheet (SF 691) which may have been a curse tablet, but it was not possible to unfold it and check for writing. Another 51 metal-detector finds (identified as from context 6682), recovered from the spoil heap of a pipe trench dug through the Ebbsfleet, have not been counted among the Ebbsfleet assemblage as their position within the depositional sequence could not be ascertained. However, it is likely that a large part of these 51 objects were deposited in the Ebbsfleet similar to the assemblage discussed above. The 51 objects include 12 brooches, ear-ring Cat No 151, two finger-rings (including Cat No 162), a ring-key (SF 9373), a pin (SF 9347), and a folded lead plaque (SF 9379) without writing but showing a curvilinear line, possibly an animal’s tail, on one surface. Roadside settlement property 2 Of the 142 metal small finds from this property, which includes a temple, only 15 are personal items, including eight brooches. The latter mainly belong to the 2nd half of the 1st century AD (Nauheim-derivative and twopiece Colchester brooches), while brooch Cat No 134 and finger-ring Cat No 156 found in the temple overburden which also contained all the Nauheimderivatives, date to the 2nd century. Other finds from the temple overburden, which contained the majority of the finds from property 2 (87 items), include a silver toilet spoon, a pair of tweezers, two knives, a cleaver, an iron currency bar (Cat No 309), and an early Roman spearhead (Cat No 290). The only other weapon from this property was a 1st/early 2nd century dagger (Cat No 292) from pit 12666. One of the post-holes defining the temenos contained one of only six seal-boxes (Cat No 208) from Springhead; a second one was found in the sub-soil within the area of this property. Other finds of note from this property include a late Roman bracelet (SF 15822), also from a temenos post-hole, of a type most common in Britain, a possible gouge (SF 20146) from a pit southwest of the temple, and three objects associated with transport comprising hipposandal (Cat No 216), terret (Cat No 219), and lynch pin (Cat No 223), all from features probably pre-dating the temple. Two studs or box nails (SFs 15753 and 18270) may have belonged to the same box but one comes from the temple overburden while the other is from a pre-temple layer. While most of the objects included in the metalworking category comprise lead waste and sheet cut-offs from the temple overburden, there are also two iron bars, one found in oven 12002 south-west of the temple and another from a temenos post-hole. The small bell SF 15748 from grave 12222, a pot burial containing one or possibly two neonates, is the only votive find in property 2, although it is likely that a single nail found in this grave also had some apotropaic purpose in this context. Cat No 308, a sheet fragment, folded/rolled to form a small container, was found in the sub-soil above the temple. The object may have had a similar use to that of the small token wrapped around a nail (Cat No 307) found in ritual shaft 2856 in the Sanctuary complex. Roadside settlement property 3 The largest individual assemblage of recorded metal small finds from any identifiable unit at Springhead was recovered from property 3, amounting to 362 objects, and most of these, 243, were found in channel fills (400175) in the waterfront area to the north-east of the bakery complex. As was the case in property 2, personal objects do not dominate the assemblage in a way seen in the area to the east of the Ebbsfleet (ARC SPH00) Chapter 3 Springhead Metalwork where they rarely account for less than a quarter of individual assemblages. On the western side of the Ebbsfleet (the Roadside settlement) this quantity is never reached in any of the individual property assemblages. In property 3, the 36 personal objects only account for 9.9%, but in absolute numbers this still makes it the third largest assemblage from Springhead. Seventeen personal objects were retrieved from the channel fills (400175) in the waterfront area, including 12 brooches of types dating to the mid-1st and 2nd centuries. A probable belt fitting (Cat No 169) is slightly later, dating to the late 2nd/early 3rd century. Another eight personal objects were found in the layers to the north-east of the sunken-featured structure, possibly a bakery (see Vol 1, Chap 2). These layers contained 49 of the 362 metal small finds from this property, including only three brooches: a later 1st century two-piece Colchester brooch, toilet set brooch Cat No 117, and mid-2nd/3rd century brooch Cat No 129. Other personal items from this layer comprise finger-ring, Cat No 157, and an unfinished ring as well as two pins of probable 2nd century date. Apart from one of only eight metal needles from Springhead and a toilet spoon (Cat No 184), which is likely to belong to the toilet set brooch, most of the other finds from this layer are unidentifiable fragments of lead sheet or waste which may be related to an unidentified metalworking activity. Other features within property 3 produced some of the more eye-catching brooches from Springhead, including the shoe-sole brooch, Cat No 127 (Pl 7), and the most unusual Cat No 99 (Pl 4), both dated to the earlier half of the 2nd century, as well as a small late rosette brooch and two two-piece Colchesters. Of the eight pins four belong to types in use throughout the Roman period and two are more specific to the early 2nd century. The quantity of toilet implements from this property is among the lower percentages from Springhead, but the absolute number of nine items is similar to that from the Sanctuary complex on the other side of the Ebbsfleet and the Ebbsfleet itself. Apart from the toilet spoon, a pair of tweezers (SF 18951), the fragments of at least two toilet sets with nail cleaners and tweezers were found, as well as a ligula (Cat No 186), a small single olive probe (SF 18001), two possible mirror fragments, and an unidentified tool (SF 20414). Knife, Cat No 229, may have been used for shaving, while the socketed knife, Cat No 240, will have served some heavier duty. With 11 objects, property 3 produced the largest number of items in the category ‘household’, including eight lead pot-mends of which all but one come from the channel fills in the waterfront area. Pot-mends have the densest distribution in the area of property 3 (Fig 108), but the reason for this is not clear as is the exact purpose of these objects; rather than simply mending pots they may be related to a certain type of process or even have some ritual significance (see Pl 13 and earlier discussion under ‘pot-mends’). Among the household items is a small animal leg-shaped fitting (SF 18294) which may 283 have been part of a vessel or a vessel stand. With two keys and a latchlifter, property 3 yielded the largest number of such items from any property in Springhead. A small fitting with handle (Cat No 193) certainly belonged to a small box or etui. Six items are or may have been used for some metrological purpose, including two or possibly three weights and a folded disc (SF 18438). The latter is reminiscent of the token folded around a nail (Cat No 307) in ritual shaft 2856 but does not contain a nail. Four military items were discovered in the channel fills (400175) in the east of property 3. Three of these are armillae (SFs 18725–6, 18868), making this the largest concentration of these post-Conquest military awards in any property at Springhead. The fourth item is a conically shaped stud, possibly from a helmet (SF 20002). Property 3 is unusual in the context of Springhead in that it produced two votive figurines, a deliberately bent lead figurine of Fortuna (SF 20114, Pl 16, left) and a small figurine of a horse whose head had been broken off but both pieces were found together (Cat No 303). In both cases the mutilations appear deliberate and are likely to be associated with the act of deposition. Although not included in Table 52 because it was found by metal detector in the overburden below the access road to the modern nursery, the small votive silver shield Cat No 305 should be mentioned here as the area it was found in falls into property 3, close to the boundary of property 2 adjoining to the south. Remarkable in the assemblage are the 123 items classed under ‘metalworking’; 119 of these are lead and take the form of irregular off-cuts or folded bits of sheet, some have clearly been partly melted or are drops of molten metal. Of these waste items, 109 were found in the channel fills (400175), which makes it difficult to link these items to any specific feature identified in the property. Considering the prevalence of lead, the activity resulting in this waste probably did not require high temperatures, and any installation involved in this activity may not have left any discernible trace apart from its waste. It is possible that the large amount of lead waste is related to the dismantling of the lead guttering from pit 300570 and similar pits in properties 3 and 4. Roadside settlement property 4 The eight personal objects from this property include the only golden finger-ring from Springhead (Cat No 154; Pl 10). Of the four brooches, three date to the 1st century AD and an unusual plate brooch made of a radiate coin of the later 3rd century. One of the two pins belongs to Cool’s group 25 which may span the Roman period. Only one item in the ‘household’ category was found in a clay and timber lined tank (16831); it is a piece of sheet metal (SF 15970) with lines of small holes radiating from the centre, probably part of a late 1st/2nd-century strainer. At the bottom and towards one corner of the tank a large piece of lead guttering with a drain hole (Cat No 227) was found in situ. 284 Settling the Ebbsfleet Valley Roadside settlement property 5 In contexts which can be assigned to property 5, only 22 metal small finds were found. Among the four personal items are three pins: two unidentifiable shanks and one of Cool’s early Roman group 3 from a sunken-featured building. A length of wire bent in a continuous S-shape (SF 15994) was found in post-hole 17012; the wire may be the central part of an extendable armlet, possibly similar to one dredged from the Rhine near Nijmegen (Sas and Thoen 2002, 175, no 94). Of the three transport objects, one is a Roman hipposandal (Cat No 217) and two are horseshoes of which one comes from the same pit (17179) as the hipposandal; it is likely, however, that both horseshoes are related to the use of a later, medieval track which runs along the western side of the Ebbsfleet. The two agricultural objects from the property are a reaping hook (SF 20072) and a spade shoe (Cat No 288), both of forms which were in use by the Roman period. Roadside settlement properties 6 to 9 Only 28 metal small finds were recorded from features associated with these four properties. Of note is a ferrule (Cat No 291) which is probably part of a spear, from property 6, as well as a zoomorphic spout (Cat No 194) and a lion-headed stud (Cat No 265) found under the floor of early Roman building 300522 in property 7. Of the ten objects from property 8, a silver finger-ring with a nicolo intaglio probably depicting a hare (SF -542; Pl 11, bottom) and an armilla fragment (SF -540) – part of a Conquest period military award – should be mentioned here, as well as the only two items from property 9, found in the same post-hole: a very corroded dolphin brooch and a Cool group 3 pin. These assemblages are too small to suggest any particular use for the properties in question. Roadside settlement property 10 This property, where a smithy was located during the later 1st and early 2nd centuries, produced one of the smaller assemblages of metal small finds from the Roadside Settlement area, amounting to only 58 items. Interestingly, however, this is the second largest assemblage west of the Ebbsfleet in terms of the proportion – but not the absolute number – of personal objects with 20.7%. Of the nine brooches from this property, a mid-1st century Colchester brooch and a Hod Hill-derivative brooch (Cat No 63) were found in the build-up adjacent to Watling Street, and the unusual reversed fantail brooch Cat No 93 comes from a ditch pre-dating the smithy. One Nauheim-derivative and three two-piece Colchester brooches were found in layers associated with the first phase of the smithy and only one two-piece Colchester, a headstud brooch, and a simple twisted-wire ear-ring with the second. A pit (16464) pre-dating the first smithy phase contained cleaver Cat No 245 and an unidentified iron tool (Cat No 259) which may have been a pair of dividers, but its extremely corroded condition and a chain attached to it make this identification uncertain. Other metal small finds associated with the smithy include an iron stylus (Cat No 204), two knives (Cat No 237 and SF 18247), and an L-shaped lift key (Cat No 284) from the first phase, and a spoon probe (SF 15195), a lead spindle whorl (SF 15923), and two unidentifiable tools (SFs 15395 and 15860), one perhaps a punch or a chisel, from the second. While the 28 items classed as unknown include some sheet and other fragments, these appear not to be obvious off-cuts which would have required them to be classed as metalworking debris as was the case in other assemblages. Thus, together with the lack of tools to be expected in a smithy, the assemblage of metal small finds does not add anything towards the understanding of the processes carried out in the smithy. However, it should be mentioned that both crucible fragments and a considerable amount of metalworking slag have been retrieved from property 10 (see Andrews, Chap 5). None of the datable metal small finds was later than the early 2nd century, which is consistent with the evidence of the pottery. Roadside settlement property 11 The assemblage of metal small finds from this property located at the junction of Watling Street and the branch road has the highest proportions of both personal and toilet/medical items recorded in any property west of the Ebbsfleet; with 7.1% the latter category is the highest for all of the Springhead entities. More than a third (31) of the 85 metal small finds was recovered from two extensive layers, 10405 and 10808, of which the former contained a dolphin brooch (Cat No 88) and two Hod Hill brooches (Cat Nos 54 and 62) as well as a deliberately bent and broken spoon probe (SF 15207), a fragmented, bent olivary probe (SF 20520), and a tinned mirror fragment (SF 20523); while 10808 yielded three two-piece Colchester brooches suggesting a possibly slightly later date of deposition in the 1st century AD. Most of the other finds come from or near to the area of the sequence of circular buildings in the south-eastern corner of property 11, near the junction of Watling Street and the branch road. Two probable dene holes contained one or more iron rings of varying sizes and one also contained a two-piece Colchester brooch (SF 15234) and an iron strap. The very ornate Hod Hill brooch Cat No 52 was recovered from the early gully parallel to the branch road. Of the pits in the vicinity of the circular buildings, ten contained one or more metal small finds. Of the seven personal items recovered from these pits six are pins and one a bent wire which may have been a bracelet (SF 15300); the latter was found together with a nail cleaner (SF 15299). Pit 10170 yielded a Cool group 3 pin and a discoid fitting with two perforations, of unknown use, but very similar to one found at Weißenburg and other forts on the Upper Germanic-Raetian limes (cf Oldenstein 1976, 176; Taf 51, 598). Pit 10338 contained three pins. The only metal find from pit 10324 was a sprung lock bolt (Cat No 277), very similar to the bolt belonging to the box from burial 6 at Baldock (Stead and Rigby 1986, 70, fig 31, 27), but apart from two worked bone fragments, one of Chapter 3 Springhead Metalwork them probably a pin, and almost 1 kg of later 1st/2nd century pottery, no metal or bone fittings were recovered which would indicate the deposition of an entire box. While there is only one unidentifiable fragment associated with the earlier phase of the circular building, there is an L-shaped wall hook (Cat No 269), a chisel (Cat No 253), a hobnail, and the sheet fragments of a possible mirror from its second phase. A Colchester brooch comes from a layer pre-dating the earliest phase of the circular building. It is interesting to note that of the ten brooches found in this property, only two were found in a pit or gully and the other eight come from levelling layers while all the pins were found in pits associated with the circular building. Roadside settlement property 12 With 138 objects this property produced one of the bigger assemblages of metal small finds, but a number of the features within it only produced one or two finds. Just over a third (49) was found in two big brickearth quarries predating the establishment of the boundary between properties 11 and 12. The fill of the larger quarry 300370 contained 32 objects, including four brooches, a pin, a fragmented implement from a toilet set (Cat No 185), and a mirror fragment (SF 15067). The date ranges of these objects fall into the later 1st and early 2nd centuries, covering the earlier period of the pottery date range which extends to the later 3rd century. Other objects worth noting include an iron swivel hook with chain still attached (SF 20552), perhaps used to suspend a cauldron, and a spindle whorl or perhaps weight (SF 15078), weighing 46 g, which is almost exactly 10 sextulae. The smaller quarry 300371 contained an assemblage of largely unidentifiable fragments of metal but including a pair of iron tweezers (Cat No 182) and a possible bow fragment of a Hod Hill-type brooch. Only two objects were found related to the aisled barn 400119, including a 190 mm long bladed tool (SF 15298), perhaps a drawknife which may be compared to one from Abbeville, France (Champion 1916, pl 3, 63645). The fill of SFB 400120 yielded 26 metal small finds, comprising one two-piece Colchester brooch (SF 15399), a pair of tweezers with cross decoration (Cat No 180), and a socketed knife (Cat No 235). Other than that the fill contained six nails, two studs, and 12 unidentified objects, mainly sheet or strip fragments. As is the case with the quarry fills, it cannot be determined whether the material deposited in the SFB relates to the use of the building or whether the feature simply provided a convenient location to dump material derived from levelling or construction work in this or another property. The small later Roman cemetery 300363 (probably 3rd/4th century) on the edge of and beyond the northwestern boundary of property 12 comprised five graves of which three contained metal small finds. Two nails only were found in grave 10046. Grave 10079 had 28 metal finds; there were at least 23 nails which, apart from one on the right, were found on the left side of the north-west-facing inhumation burial. The nails could 285 have been part of a coffin, but their concentration along only one side of the burial may perhaps be due to the fact that the body was placed on a reused board which still contained nails along one side from its previous use. The body was buried with hobnailed boots, and the grave also contained half a snaffle bit (SF 15084), but it is not clear whether this was part of the fill of the grave or whether it was a placed grave good. Most of the finds from Grave 10150 are likely to be part of the backfill rather than being grave goods. They include a biconical lead weight (SF 15113), a nail, a piece of slag as well as the fragment of a bone pin with a globular head of Crummy’s type 3 which is dated mainly to the 3rd/4th century (Crummy 1983, 22), the latter possibly a grave good though its position is unrecorded. Roadside ditches 1–3 In contrast to the ditches alongside the road in the Sanctuary area on the eastern side of the Ebbsfleet which were devoid of metal small finds, the ditches along Watling Street and the north-western branch road contained 47 finds. Of these, 30 come from ditch 3 whose course could be followed in the areas of properties 3–5 for approximately 70 m. The excavated length of roadside ditch 1 is similar to ditch 3, but only 8 metal finds were retrieved from it. The frequency of finds in ditch 2, to the rear of the smithy property 10, is markedly higher: nine objects were found in the c 25 m exposed during excavation. From the datable metal small finds it appears that the ditches had essentially been filled in by or during the first half of the 2nd century. Of the seven brooches from the ditches only one fragmented two-piece Colchester comes from ditch 2, the other six were all from ditch 3, mainly belonging to its later re-cuts and including at least three two-piece Colchesters and one probable Hod Hill brooch; the fourth, final phase of roadside ditch 3 contained a Langton Down brooch (Cat No 26). Ditch 1 only contained three fragments of a fine copper alloy chain (SF 20465) and a spherical-headed pin (SF 15116); a simple Cool group 1 pin was found in ditch 3. The only personal object which may be later is bracelet Cat No 147. While toiletry implements are among the less frequent finds at Springhead, it is interesting to note that one arm of a pair of tweezers and a toilet set were found in ditch 3, but considering the fact that the adjacent property 3 yielded seven toiletry implements this may provide an explanation of the origin of the ditch fill, and this may also apply to the early Roman spoon SF 15951 and one of the two armillae fragments (SF 20161). Another armilla fragment (SF 15192) comes from ditch 1, which also contained one of the eight latchlifters (Cat No 278) found at Springhead. The six objects found in the road surfaces of Watling Street include one Hod Hill brooch (Cat No 55), a simple ear-ring (SF 18023) and the fragments of a toilet set. The scarcity of finds from the road surface suggests that it was regularly swept, but as it was metalled items that were lost on it could also be retrieved more easily. 286 Settling the Ebbsfleet Valley Communication with the Gods of Springhead: Votive Objects – Objects used as Votives In this section, an attempt shall be made to discuss the evidence for religious practices discernible through the study of the metal small finds assemblage from the HS1-excavations and to answer questions concerning the nature of the cult or cults practised at Springhead, the spheres of religious practices represented in the assemblage and the objects used in these practices. Exchanges between human and deity Rituals concerning exchanges of vows and votives to secure the help of the gods are a common feature of prehistoric and antique religion. At the sphere of the Roman state, the exact observance of rituals to ensure the favour of the gods for the state and its inhabitants is described with the term religio; it was the ritual adherence to the exact wording and actions of the cult that were of importance while a spiritual involvement was not required (Müller 2002, 16). The more private devotion of the population seeking religious fulfilment and personal devotion to a deity, aspects which according to Müller (ibid, 17) are rarely mentioned in antique literature, fall into the definition of superstitio which covers aspects such as popular belief, superstition, and magic. Henig (1984, 32) points out that ‘magic is not religion but rather a debased offshoot from it which assumes that the gods can be controlled by man’, contrasting it to religious prayer ‘addressed to gods who are free agents, not obliged to answer it.’ To promote the success of a request the devotee would vow to present a gift (the act is described by the Latin term nuncupatio) if the deity granted his or her wish. The fulfilment of the vow by the devotee is called solutio, and this stage of the process is often documented with the inscription VSLM (votum solvit libens merito) ‘vow paid freely and deservedly’ frequently found on altars or votive plaques (Müller 2002, 19, Abb 7; Bagnall Smith 2008, 153). Nuncupatio and solutio are constituent and consecutive parts of the ritual practice of a formal votum which would have reached Britain with the Roman army at the time of the Conquest (Bagnall Smith 1999, 48, 51), if not before. In her discussion of the votive objects from Great Walsingham, Bagnall Smith (ibid, 49–50) explains the presence of the 22 seal-boxes there as evidence of the first stage of the votum. Based on a theory proposed by Derks (1995) who examined evidence for the ritual of the vow in Gallo-Roman religion, focusing on the Rhineland, the seal-boxes are interpreted as all that remains of a nuncupatio in the private sphere. This would not require permanent material but could be written on a wooden writing tablet, while those of a more private nature might need to be written on the inner surfaces of the tablet which would then be tied with a cord, sealed and the seal put in a seal-box for protection. Bagnall Smith (ibid, 50) claims that the large number of sealboxes at Great Walsingham provide strong evidence for the formal practice of the nuncupatio but that the actual number of seal-boxes found at a shrine is in effect irrelevant, because these documents ‘would not have remained in a sanctuary beyond the time stipulated in the legal wording of the vow.’ The theory laid out above could provide an explanation for all or at least some of the seven sealboxes or seal-box fragments from Springhead, especially the two leaf-shaped boxes found in property 2 near the temenos boundary and in the temple overburden (Cat No 208 and SF 15618). The base of one seal-box (SF 1836) was found in the channel fills of the Ebbsfleet which could imply that it had become a votive gift as part of the solutio after the wish made in the first part of the vow had been granted. However, more mail-related uses of sealboxes have been discussed for some of the 13 seal-boxes recovered at Castleford where six were found in a building which could have been the starting point or the destination of the packages protected by seals (Cool in Cool and Philo 1998, 101), and this use should not be discounted for some of the boxes from Springhead. The Springhead Metalwork Assemblage in comparison to other ‘Votive’ Assemblages This section compares selected metalwork categories with those from other Romano-British sanctuaries which have produced larger numbers of metal objects, in order to understand the nature of the cult or cults at Springhead as far as this can be discerned from quantitative variations in the assemblages. As can be seen in Table 54, most of the sites are in southern and south-western England; Harlow, c 40 km north-west of Springhead, is the only south-eastern site. Most of the quantifications in the table were taken from Woodward and Leach (1993, 332 tab. 20); those for Lydney, Nettleton, Henley Wood, and Woodeaton were amended using the quantifications in Smith (2001); quantities for brooches from Harlow were slightly amended on the basis of a footnote in Haselgrove (2005, 411, note 118) where a brief summary has been given of more recent excavations carried out in the 1980s. This circumstance also serves to illustrate the scarcity of fully published assemblages from religious sites in the southeast of England with sufficient quantities to allow meaningful comparison. The sites listed in Table 54 show clear variations in the amounts of objects across the various categories with distinctive peaks occurring in one to six specific categories. The peaks occur in the following categories, (peaks in non-metalwork categories are in italics. They are not included in Table 54 but were originally considered by Woodward and Leach (1993, 332)): • Springhead: brooches, rings, pins and toilet articles; • Uley: copper alloy rings, miniature clay pots, finger rings; • Lydney: bracelets, pins, spoons; • Nettleton: brooches, pins, bracelets, styli, spoons, finger-rings; • Henley Wood: counters, brooches; Chapter 3 Springhead Metalwork 287 Table 54 Springhead metal finds: comparison of selected metal object types in votive assemblages Springhead Uley Lydney Nettleton Henley Wood Woodeaton Harlow (Woodward & Leach 1993) (Wheeler & Wheeler 1932) (Wedlake 1982) (Watts & Leach 1996) (Kirk 1949) (France & Gobel 1985) 10 1 7 Votive Figurines 3 18 Rings 49 52 Leaves Plaques 11 1 Parts of body 1 18 9 3 2 Letters Miniature tools/weapons Total 3 1 1 2 19 45 53 7 2 3 14 6 1 1 19 8 116 73 2 6 42 17 Fittings Iron rings 22 Metal vessels Candlesticks Total 6 3 15 2 3 2 22 23 3 2 3 0 1 1 Jewellery Brooches 309 40 32 >112 33 132 96+56 Bracelets (+armillae) Pins 10+8 33 c 300 57 5 28 4 80 8 320 63 14 54 15 Finger-rings 22 38 6 24 12 32 19 Total 429 119 c 658 >256 64 >246 190 Spoons 4 14 >40 29 Toilet articles 54 6 3 21 5 29 19 Total 58 20 43 50 5 36 19 Personal items 7 Writing/literacy Seal-boxes 6 1 ? 1 Styli 7 8 1 33 10 27 ? 5 Total 13 9 1 34 10 27 5 Harlow with additions based on Haselgrove (2005, 411 note 118) • Woodeaton: brooches, plaques, pins, toilet articles, miniature tools/weapons; • Harlow: brooches. In the cases of the smaller assemblages like Henley Wood and Harlow, Woodward and Leach (ibid) cautioned that the collection of objects may not be representative of the original assemblages in use on the sites, and it also true that the prevalence of brooches at sites like Harlow or Springhead might in part be due to chronological factors. This caveat is certainly worth bearing in mind but the distribution patterns of brooches and pins at Springhead caution against explaining the use of brooches in ritual circumstances purely on grounds of chronology. With only six out of 80 pins their distributions clearly avoids the Ebbsfleet pool (Fig 102) where brooches are the most numerous type of object (see below). On the other hand, rings are found in the pool, just like the brooches, and the distribution of all three types of objects is less discrete on land. The main deity venerated at Uley was Mercury, and direct evidence of his cult there is provided by statues, altars, figurines, caducei, plaques, and inscriptions on lead tablets. Other categories of finds from the site suggested to be linked with the god include the rings, finger-rings, and coins, and possibly also votive legs and a plaque fragment with a leg (Woodward and Leach 1993, 333; Henig in ibid, 112 and 174). While the former allude to the god’s role as a guardian of trade and commerce, the latter two may be ascribed to his role as a god of travellers effective in the cure of diseases impeding movement. The miniature weapons found at Uley were linked to Mars and considered to be substitutes for full size weapons found in deposits dated to the 1st century AD. Other deities represented at Uley include Sol, Jupiter, Cupid or Victory, and a naked child, possibly Bacchus (Woodward and Leach 1993, 333). The miniature clay pots found in considerable numbers at Uley where interpreted as containers for small votive 288 Settling the Ebbsfleet Valley offerings of a few grains of cereal or small amounts of wine or beer (Henig in ibid, 112). The temple at Lydney is known to have been dedicated to Nodens or Nodons, possibly a god of hunting and fishing, who was equated to Mars on two votive plaques from the site (Henig 1984, 51). Figurines and depictions of hounds have been interpreted as indicating a cult of healing at the site (ibid, 55; Wheeler and Wheeler 1932, 39–43, cited after Woodward and Leach 1993, 333), and similarly the large numbers of pins and bracelets have been taken as indicators to a healing cult comparable to those given to Greek temples by female worshippers in preparation to give birth (ibid, 41–2). It is possible that the more than 40 spoons from the site also played a role in the veneration of Nodens at Lydney. Nettleton’s principal deity was Apollo, once referred to as Cunomaglos (the hound-prince) on an altar (Toynbee in Wedlake 1982, 136, pl 34). The hunter connection for Apollo at Nettleton has been discussed by Woodward and Leach (1993, 334), who suggested that the large number of pins and spoons may point to his guise as archer and healer similar to Nodens at Lydney. Other pagan gods venerated at Nettleton include Diana and Silvanus, both also hunters, as well as Mercury and Rosmerta, and an unidentified, probably local goddess (Wedlake 1982, 104). At Harlow, Henley Wood, and Woodeaton no inscriptions indicate a principle dedication of the cults. However, several depictions of deities at Woodeaton provide a glimpse at the pantheon worshipped at the site (Green 1976, 177–8). These include Venus, a kilted Celtic goddess, Mars, Minerva, Cupid, and possibly a Celtic Hercules. Six bronze eagles suggest an association with Jupiter. Votive chain mail and miniature weapons would appear to belong to the worship of Mars, whereas the two bronze snakes, a statuette of a nude female, the female toilet articles and miniature bracelets suggest a cult of healing at the site (ibid; Woodward and Leach 1993, 334). At Harlow, a limestone helmeted head of Minerva was found in excavations in the 1980s (Haselgrove 2005, 411, note 118), and the only figurine found at Henley Wood depicts a Celtic goddess wearing a peculiar head-dress and a torque, interpreted by Henig (in Watts and Leach 1996, 133) as a native counterpart to Greek Aphrodite or Roman Venus. Other aspects of the assemblage like the brooches, and perhaps the rings and an infant burial, could all be associated with a cult of fertility and fecundity of both humans and animals (Watts and Leach 1996, 145). As summarised by Woodward and Leach (1993, 334) the three groups of ex votos, firstly miniature weapons and tools, secondly personal objects, and thirdly the combination of rings, discs, or tokens and money ‘can be seen to represent three major cult forms present in the Romano-Celtic religion of Britain.’ The first group belongs to a martial cult, the second denotes a cult devoted to fecundity and healing, while the third group is indicative of Mercury. Which Deities were Venerated at Springhead? Similar to Harlow, Henley Wood, or Woodeaton, this question is difficult to answer at Springhead where inscriptions, dedications or even explicit, inscribed curse tablets have so far not been forthcoming (cf list of Springhead temples in Green 1976, 228), and equally none of the graffiti found on sherds refers to deities (Seager Smith and Marter Brown, Chap 1). The only metal small finds which could be readily identified are the two lead figurines depicting Fortuna (Cat No 302 and SF 20114), both probably made at Springhead itself using the same model but not the same mould. Considering the rather crude workmanship, a use in any form of official cult related to the fortune of the Emperor and the state seems less likely than a more personal sphere, for example relating to the wish for luck in gambling, protection from mischance in the bathhouse, or a safe return home (Henig 1984, 77–9). The contexts in which the two figurines where found – the waterfront in property 4 and the watching brief trench south of the route of the A2 – unfortunately do not add any further clues. Especially in Gaul and the Rhineland Fortuna-Rosmerta is often coupled with Mercury, ‘the most popular classical deity in Britain, as in Gaul, … the god of the shops and market-places, as his name indicates’ (Webster 1986, 60). Prior to the HS1-excavations the only figurines of deities recovered from Springhead were of pipeclay, depicting Venus (see Mepham, Chap 10; Jenkins 1958, pl 1), and Green (1976, 228) also mentions a bone Genius Cucullatus. Evidence for a possible statue of Jupiter at Springhead is provided by a gilded copper alloy thunderbolt from Site B (Penn 1958, 92, fig 9.4; Bagnall Smith 2008, 160, fig 8.5). A brooch in the shape of a three-dimensional duck (Hull T213) was found in a later 2nd century level in Building B.10 at Springhead (Penn 1968a, 184, no 2, pl 2b). Crummy (2007, 228) suggested that water birds ‘may relate to local deities of rivers and springs’, but their significance is not yet certain. Apart from the two Fortuna figurines the only other figurative metal object from the HS1 excavations is the small horse figurine Cat No 303 whose head was broken off deliberately and both pieces deposited together in property 4. This horse is not the first from Springhead as fragments of a clay figurine, possibly of a horse, were found in the ditch surrounding Temple IV excavated in the 1960s (Penn 1964, 172). In Romano-Celtic iconography, horses are connected with Epona, the Celtic horse goddess, usually depicted seated on a horse or between two ponies (Green 1976, 14; Euskirchen 1993). However, it is possible that the Springhead horses are related to the cult of Mars who, in Britain and Gaul, is frequently depicted on horseback, eg, at Stragglethorpe, Lincolnshire, or Willingham Fen, Cambridgeshire, while this aspect is less pronounced in his classical Graeco-Roman guise, although not unknown (Henig 1984, 51; 53, figs 13–4; Green 1976, Chapter 3 Springhead Metalwork 11; 30). Perhaps this connection to a warrior god might also provide an explanation for the presence of the small silver shield Cat No 305 and the model cheek piece Cat No 306. A shield-shaped brooch was found at Lydney (Stead 1991, 25–6), a major temple dedicated to the worship of Mars Nodens or Nodons (Henig 1984, 51). Green (1975, 59) considers the possibility of model spears, swords, and shields to be ‘offerings by a soldier to a soldier’s deity – Mars or a local equivalent – or merely to a local multi-functional god.’ The probable model cheek piece of a helmet (Cat No 306) is decorated with a dolphin, an animal suggesting a connection with Neptune or Ocean (Henig 1974, 154; see also altar dedicated to Neptune dredged from the Tyne at Newcastle (Liversidge 1973, 389, fig 148)). Rather than a direct veneration of these gods, however, the depiction on the model of part of a defensive weapon may perhaps rather be a votive by a soldier or a soldier’s relative asking for a safe journey into the afterlife, bearing in mind that other sea-creatures like hippocamps have been connected with the myth of the soul’s wanderings to the Islands of the Blessed (Henig 1974, 154), and as such dolphins might be suitable animals to be shown on a tombstone; an example is the stone to T Valerius Pudens from Lincoln (Brailsford 1964, 59, fig 28, vi, b.3). The safe-journey aspect might also be what is implied by the depiction of a dolphin below a dioscurus on a real cheek piece of an auxiliary cavalry helmet dredged out of the Tyne near South Shields (Allason-Jones and Miket 1984, 213, pl 7, 3.723) or the dolphin on a leather panel, perhaps a wallet or bag, from the Blackfriar’s Roman barge (Merrifield 1965, pl 32; Waterer 1967, pl 12). Considering the use of a model cheek piece as a votive object, the use of such pieces is known from the 4th century BC Samnite sanctuary at Pietrabbondante, Italy, where real pieces were nailed to the walls as votives (Müller 2002, 98–9 Abb 65). ‘Shields and spears are symbols of protection from threats of a personal or general nature’ (Webster 1986, 61), and the same can be expected of other pieces of armour. That a model was used in this instance, as are those mentioned above, concurs with the substitution of real weapons by models observed at other sanctuaries; a practice which according to Green (1975, 56–7) can be traced back to at least as early as the Bronze Age. Returning to the symbolism on the cheek piece, it should be remembered that the dolphin was also incorporated into late antique Christian art (Thomas 1981, 92–3 fig 8, 8–10). Apart from a possible Chi-Rho grafitto on a weight (Cat No 199), no other indications of Christianity were found at Springhead. Furthermore, while the cheek piece was found in the Ebbsfleet, and thus without context date, the weight was found in chalk quarry 300204 belonging to the early Roman phase, consequently rendering a Christian interpretation of the graffito even less likely. The wheel-shaped plate brooch Cat No 108, found in layers pre-dating the sanctuary to the east of the Ebbsfleet, may be a votive offering of similar significance 289 to model wheels found in Britain and Gaul. The wheel was a symbol of the Celtic sky-god, Taranis, identified with Jupiter (Green 1975, 58–9 and list p 62; Henig 1984, 59). However, the wheel has been a sun symbol since prehistory and was therefore probably already used as a talisman before becoming associated more specifically with a particular deity. Considering the two Fortuna figurines whose attributes include a wheel, it may equally have been a general ‘good-luck’ charm (cf Webster 1986, 61), and as such may add further support for a veneration of Fortuna at Springhead. While the above examples might provide some insight as to identifiable deities venerated at Springhead, the distribution patterns of certain object types appear to be sufficiently discrete to suggest deliberate selection in the types of votive offerings dedicated to different deities. The most obvious case in point concerns the distribution of brooches and metal pins. The distribution map of metal pins (Fig 102) shows only six of the 80 pins (7.5%, or 7 = 8.6% if those from pipe trench 6682 were included) were found in the Ebbsfleet, which contrasts strongly with the 68 (21.9%, or 80 = 25.8% including context 6682) brooches found in the spring and channel fills, out of a total of 310 late Iron Age and Roman brooches. This strong discrepancy is further emphasised when bone pins are included in the consideration, as these were equally lacking from the Ebbsfleet but very common in the area of the temple in property 2 (Allen, Chap 14). That this selection is highly likely to represent a real discrimination is further emphasised by the fact that the distribution of nonferrous metal finger-rings and other rings (Fig 104) is essentially similar to that of the pins, but includes many more rings in the Ebbsfleet and considerably fewer in the area to the west of the round structure in property 11. Which Objects were Employed in a Religious Context? What becomes immediately apparent from Table 52 is the stark contrast in the quantity of personal items between the eastern (Sanctuary) and western (Roadside Settlement) halves of Springhead. While one could argue that most personal items found in dry areas of the site were simply lost during everyday life activities, this explanation does not hold true for the large quantity of objects found in the spring and channel fills of the Ebbsfleet. There, a deliberate deposition rather than an accidental loss is much more likely. In many cultures water has been considered sacred, and the veneration of sacred springs in antiquity is a well documented phenomenon (Müller 2006, 111). A religious or ritual reason for deposition in watery contexts can therefore be assumed (Henig 1984, 17; Müller 2002, 56–92). The deposition of an object in water, and the same is true, for instance, of caves or rock crevasses, has the purpose of bringing it closer to the deity and at the same time taking it beyond human reach. The same reasons apply to the deliberate destruction of objects prior to deposition 290 Settling the Ebbsfleet Valley (ibid, 31). In the case of the brooches from Springhead there are 68 broken and 17 bent examples of which at least 47 appear to be deliberately broken and 12 deliberately bent. As with deposition in the sacred areas of temples or shrines on dry land, where objects would be placed in the soil, proximity to a sacred place was what mattered because it provided the liminal context, the interface between the human and the divine (Crummy 2006, 56). For the attribution of an object to a functional category in Table 52 this means that the use of an object originally created for one purpose – in the case of a brooch its use as a closing mechanism for garments – can differ from the use that can be deduced from the circumstances of its final deposition. Considering the above remarks about deposition in springs, the final use of an object would thus have to be described as votive. It is easy to accept personal or toiletry objects like brooches, pins, finger-rings, nail cleaners, or tweezers as votive gifts because such objects, representing the individual, might seem appropriate to establish a personal relation to the divine force (Cunliffe 1988, 360). If this interpretation can be accepted for these categories of objects, what about objects listed in other categories? An example comes from the spring of Sulis Minerva at Bath, where Cunliffe (ibid) distinguished five categories: personal items, professional items, paterae and other vessels, coins, and curses. The scope and the reason for throwing personal items into the spring has been explained above, and due to its composition at Bath, Cunliffe suggested that these largely represented the female sphere, either because the devotees offering such items were predominantly female or because it was considered appropriate to present ‘female’ objects to a goddess. A small collection of items, including a model ballista washer, an ink well, lead and pewter ingots, a knife, a whetstone, and spindle whorls, on the other hand, was tentatively related to a professional, predominantly male sphere reflecting a range of trade and craft activities carried out at or near Bath (ibid, 360–1). The vessels of various date groups could have been gifts to the goddess, some perhaps deposited after a long use as implements for rituals such as libations. While no doubt offerings, Cunliffe considered various reasons for the deposition of coins: some probably being votives in the formal sense discussed above, ie, fulfilment of a vow, while others ‘may have been motivated by little more than casual superstition—the desire to remain on the right side of the goddess … just in case’ (ibid, 361). The deposition of the metal curse tablets in the spring of Sulis Minerva brought their contents to the attention of the goddess, the act of deposition may have been cathartic for the aggrieved and, if known to the person cursed in the formula, may have caused the perpetrator concern or despair. Considering the symbolism behind the votive offering of brooches, bracelets, finger-rings, pins, etc, Webster (1986, 60) offers some explanations of what it was the persons offering such objects wanted to achieve. Thus, brooches can be seen as symbols of a bond, mirroring their functional purpose of holding together folds of a garment, and at the same time their pin pierces the cloth, which ‘can be seen as an image of a union made effective by sexual penetration.’ The image of rings and bracelets as symbols of union between the sexes still persist today in the form of the wedding ring, and ‘pins have an obvious sexual shape and symbolize the need for satisfactory penetration’ (ibid, 61). Webster goes on to suggest that apart from the sacrificial aspect of removing from human use objects by breaking them, the breaking of rings or bracelets, or making small rings from broken bracelets, ‘could have symbolized wishes for existing unions to be ended, so that new ones could be established’ (ibid). With some restraints concerning the lack of curse tablets and complete vessels at Springhead, similar explanations can reasonably be assumed for the objects found in the Ebbsfleet and the Sanctuary area, and probably also for a great many more features at Springhead like the temple in property 2, the viewing platforms or the round structure in property 11 at the junction between Watling Street and the north-western branch road. Another sphere of objects with possible religious connotations is that represented by priestly regalia, of which there may be at least two items from Springhead: the fragmented ivy leave-shaped fitting Cat No 266, found in the Sanctuary overburden, and the ring with four lengths of fragmented chain attached to it (SF 975), the latter found in grave 6345 which contained the cremated remains of a c 25–35 year old female individual. Whether the object lay just inside the south-western side of the casket or immediately outside it could not be decided during the excavation; it was certainly not exposed to the heat of the pyre and, therefore, it is unlikely that it found its way into the grave as part of the pyre debris. The fragmented ring-and-chains could have been deposited in the grave as a pars pro toto, a deliberately fragmented item put in the grave as a representation of the whole object. The interpretation of these two objects in a religious context has to remain uncertain due to their fragmentary preservation, which would also allow for other explanations, including box or strap fitting and suspension chain for a scale/balance pan or a lamp. To conclude this section, a tentative explanation for the presence in the Ebbsfleet’s spring and channel fills of some of the 50 objects in the category ‘unknown’ may be offered. This is a collection of – often very corroded – fragments of strips, sheet metal, lumps, bars, wires, etc. While some of these objects may have been complete at the time of their deposition and now are merely too corroded for further identification, others may simply have been pieces of scrap metal thrown into the pond by the poorer visitors of Springhead who could not afford to obtain nice, complete objects for sacrifice to a deity Chapter 3 Springhead Metalwork but wanted to imitate their wealthier contemporaries who they observed throwing metal objects into the Ebbsfleet. It is known that certain materials were attributed specific amuletic properties, and writing about amulets and ex-votos Henig (1984, 187) notes that ‘noble metals, especially incorruptible gold, suit the 291 high purposes of personal protection, while base lead is the metal for retribution and cursing.’ If this is the case, could it not be possible that for certain wishes or sacrifices the material of the votive object was of greater importance than its shape and even waste material would suffice? 308 Settling the Ebbsfleet Valley Blank Page Chapter 1 Introduction 405 Bibliography Note: the use of ‘a’, ‘b’ etc suffixes to denote publications by the same person in the same year has been standardised across the four volumes so that some entry suffixes here may not run in true consecutive order. Abegg, A, 1989 Eine wohlhabende Frau aus Belginum Grab 2370, in Gräber – Spiegel des Lebens. Zum Totenbrauchtum der Kelten und Römer am Beispiel des Treverer-Gräberfeldes Wederath-Belginum (A Haffner), 299–316. Verlag Philipp von Zabern, Mainz Adamson, S and Niblett, R, 2006 The small finds, in Niblett, R, with Manning, W and Saunders, C, Verulamium: excavations within the Roman town 1986–88. Britannia 37, 53–188 ADS, 2006 CTRL Digital Archive, Archaeology Data Service, http://ads.ahds.ac.uk/catalogue/ projArch/ctrl Alföldi, A and Radnóti, A, 1940 Zügelringe und Zierbeschläge von römischen Jochen und Kummeten aus Pannonien, Serta Hoffileriana. Festschrift für Victor Hoffiler, 309–319. Zagreb Allason-Jones, L, 1989 Ear-rings in Roman Britain. Brit Archaeol Rep 201, Oxford Allason-Jones, L, 1996 Roman Jet in the Yorkshire Museum. Yorkshire Museum, York Allason-Jones, L and Miket, R, 1984 The Catalogue of Small Finds from South Shields Roman Fort. Soc Antiq Newcastle upon Tyne Monogr 2, Newcastle Allason-Jones, L and McKay, B, 1985 Coventina’s Well. A Shrine on Hadrian’s Wall. Chesters Museum/Alan Sutton, Stroud Allen, D, 1967 Iron currency bars in Britain. Proc Prehist Soc 33, 307–35 Andersen, S T, 1970 The relative pollen productivity and pollen representation of north European trees and correction factors for tree pollen spectra. Danmarks Geologiske Undersøgelse R.II 96, 1–99 Andersen, S T 1973 The differential pollen productivity of trees and its significance for the interpretation of a pollen diagram from a forested region, in Quaternary Plant Ecology (eds H J B Birks and R G West), 109–15. Blackwell, Oxford Anderson, A C, 1980 A Guide to Roman fineware. VORDA Res Ser 1, Highworth, Wiltshire Anderson, A, 2002 Facemask, in Cataractonium: Roman Catterick and its Hinterland. Excavations and Research, 1958–1997. Part II (P R Wilson), 200–1 Counc Brit Archaeol Res Rep 129, York Anderson, A S, 2001a Lead objects, in Anderson et al 2001, 117–8 Anderson, A S, 2001b Pewter objects, in Anderson et al 2001, 119–20 Anderson-Stojanovic, V R, 1987 The chronology and function of ceramic unguentaria. Amer J Archaeol 91(1), 105–22 Anderson, A S, Wacher, J S and Fitzpatrick, A P, 2001 The Romano-British ‘small town’ at Wanborough, Wiltshire, Britannia Monogr 19, London Andrews, P 2009 West Thurrock: late prehistoric settlement, Roman burials and the medieval manor house. Channel Tunnel Rail Link excavations 2002. Trans Essex Soc Archaeol Hist 40, 1–77 Applegate, D, 2007 London Essex Stamped ware from Shorne Marshes, Kent Archaeol Rev 170, 212–13 Armbruster, B, 2004 Goldsmith’s tools at Hedeby, in Land, sea and home (eds J Hines, A Lane, and M Redknap), 109–23. Maney, Leeds Armbruster, B R, with Eilbracht, H, 2006 Technological aspects of the Viking age gold treasure from Hiddensee, Germany, Hist Metall 40(1), 27–41 Arthur, P, 1978 The lead glazed wares of Roman Britain, in Arthur and Marsh (eds) 1978, 293–355 Arthur, P and Marsh, G (eds) 1978 Early Fine Wares in Roman Britain. Brit Archaeol Rep 57, Oxford Arthur, P, 1986 Roman amphorae from Canterbury. Britannia 17, 239–58 Aubert, X, 1929 Évolution des hipposandales. Revue des musées vols 19–21, pp 5–9, 53–6, 75–8 Aveling, E, 1997 Chew, chew, that ancient chewing gum. Brit Archaeol 21 http://www.btriarch.ac.uk/ba/ba21/ ba21feat.html Aveling, E M and Heron, C, 1999 Chewing tar in the early Holocene: an archaeological and ethnographic evaluation. Antiquity 73, 579–84 Bagnall Smith, J, 1999 Votive objects and objects of votive significance from Great Walsingham. Britannia 30, 21–56 Bagnall Smith, J, 2008 Aspects of votive offering in south-east Britain, in Ritual Landscapes of Roman South-East Britain (ed D Rudling), 153–70. Oxbow, Oxford Bakker, J Th, 2001 Les boulangeries a moulin et les distributions de ble gratuites, in Ostie: Port et porte de la Rome antique (ed J-P Descoeudres), 179–85. Musée Rath, Genève Barber, B, and Bowsher, D, 2000 The Eastern Cemetery of Roman London: excavations 1983–1990, Mus London Archaeol Service Monogr 4, London 406 Settling the Ebbsfleet Valley Barford, P M, Elder, J, and Stow, S, 1995 Other objects of fired clay in The Archaeology of Canterbury vol 5: excavations in the Marlowe Car Park and surrounding areas, Part 2: the finds (K Blockley, M Blockley, P Blockley, S Frere and S Stow), 1183–4. Canterbury Archaeol Trust, Canterbury Barker, G 1985 Prehistoric farming in Europe. Cambridge Uni Press, Cambridge Bateman, N, 2000 Gladiators at the Guildhall, Mus London Archaeol Service, London Bayley, J, 2005 Roman enamel and enamelling: new finds from Castleford, Yorkshire, in Annales du 16e Congrès de l’Association Internationale pour l’Histoire du Verre, 72–4. AIHV, Nottingham Bayley, J, and Butcher, S, 2004 Roman Brooches in Britain: a technological and typological study based on the Richborough Collection. Rep Res Comm Soc Antiq London 68, London Beard, M, North, J, and Price, S, 1998 Religions of Rome. Volume 2: a sourcebook. Cambridge Univ Press, Cambridge Bechert, T, 1973 Römische Fibeln des 1. und 2. Jahrhunderts n. Chr., Funde aus Asciburgium 1, Rheinhausen, Duisburg Bedwin, O, and Bedwin, M, 1999 A Roman Malt House: excavations at Stebbing Green, Essex 1988. E Anglian Archaeol Occas Pap 6, Chelmsford Berger, Š., 1882 Bronzy Duchcovské. Článek první. Památky Archologické 12/1, 71–80 and 109–115 Berghaus, P, 1994 Münzfibeln, in Wamers 1994, 106–15 Bertrand, I, 2003 Objets de parure et de soins du corps d’époque romaine dans l’Est picton (Deux-Sèvres, Vienne), Mémoire 23. Association des publications chauvinoises, Chauvigny Betts, I, Black, E W, and Gower, J, 1997 A corpus of Roman relief-patterned tiles in Roman Britain. J Roman Pott Stud 7 Betts, I M, and Foot, R, 1994 A newly identified late Roman tile group from southern England. Britannia 25, 21–34 Biddulph, E, 2005 Roman pottery in Bradley, P, Charles, B, Hardy, A and Poore, D, Prehistoric and Roman activity and a Civil War ditch: excavations at the Chemistry Research Laboratory, 2–4 South Parks Road, Oxford, Oxoniensia 70, 141–202 Biddulph, E, 2008 Form and function: the experimental use of samian ware cups. Oxford J Archaeol 27(1), 91–100 Biddulph, E, Compton, J, and Martin, T S, forthcoming The late Iron Age and Roman pottery, in Heybridge: a late Iron Age and Roman settlement, excavations at Elms Farm (M Atkinson and S J Preston). E Anglian Archaeol, Chelmsford Bidwell, P T, 1979 The Legionary Bath House and Basilica and Forum at Exeter. Exeter Archaeol Rep 1, Exeter City Council and University of Exeter Bidwell, P, and Croom, A, 1999 The Camulodunum/ Colchester type series, in Symonds and Wade 1999, 468–87 Bingeman, J M, Bethell, J P Goodwin, P and Mack, A T, 2000 Copper and other sheathing in the Royal Navy. Int J Naut Archaeol 29(2), 218–29 Birbeck, V, Jones, G P and Seager Smith, R H, 2008, A Roman pottery kiln, kiln furniture and new vessel forms from Alice Holt Forest. Hampshire Stud 63, 110–28 Bird, J, 1984 The samian ware, in Excavations in the Darent Valley, Kent (B Philp), 95–7, Kent Monogr 4, Dover Bird, J, 1986 Samian wares, in The Roman Quay at St Magnus House, London. Excavations at New Fresh Wharf, Lower Thames Street, London 1974–78 (L Miller, J Schofield and M Rhodes), 139–85. London Middlesex Archaeol Soc Spec Pap 8, London Bird, J, 1989 The samian ware, in The Roman House with Bacchic Murals at Dover (B Philp), 67–74. Kent Monogr 5, Dover Bird, J, 1995 The samian ware from Springhead 1991–94, in A Roman Site at Springhead (Vagniacae) near Gravesend (B Philp and M Chenery), 27–31. Kent Spec Subject Ser 9, Dover Bird, J, 1996 The samian ware, in The Roman Villa Site at Orpington, Kent (B Philp), 62–8. Kent Archaeol Rescue Unit, Dover Bird, J, 1998 The samian, in The Tremple of Mithras, London. Excavations by W F Grimes and A Williams at the Walbrook (J Shepherd), 106–7. English Herit Archaeol Rep. 12, London Bird, J, 1999 Decorated central and east Gaulish samian, in Symonds and Wade 1999, 75–119 Bird, J, 2005 The lion and the mule: from lamps to samian, in Image, Craft and the Classical World Essays in Honour of Donald Bailey and Catherine Johns (ed N Crummy), 69–79. Éditions Monique Mergoil, Montagnac Bird, J, 2006 Samian ware from Peppers Hill, in ADS 2006 Bird, J, Graham, A H, Sheldon, H and Townend, P, with Bailey, J, Locker, A, Chapmen, H and Harries, J, 1978 Southwark Excavations 1972–74: vols 1 and II. London Middlesex Archaeol Soc/Surrey Archaeol Soc Jnt Publ 1, London Bird, J, and Marsh, G, 1981 The samian ware, in The Excavation of the Roman Forts of the Classis Britannica at Dover 1970–1977 (B Philp), 178–202. Kent Monogr 3, Dover Birley, A, 2002 Garrison Life at Vindolanda: a band of brothers. Tempus, Stroud Birley, R, 1977 Vindolanda: a Roman frontier post on Hadrian’s Wall. Thames and Hudson, London Bishop M C, 1996 Finds from Aldborough: a catalogue of small finds from the Romano-British town of Isurium Brigantum. Oxbow Monogr 65, Oxford Bishop, M C, and Coulston, J N C, 2006 Roman Military Equipment: from the Punic War to the fall of Rome, 2 edn Oxbow, Oxford Bibliography Black, E W, 1986 Romano-British burial customs and religious beliefs in south-east England. Archaeol J 143, 201–39 Black, E W, 1987 The Roman Villas of South-East England. Brit Archaeol Rep 171, Oxford Black, E W, 1995 The building materials: voussoirs and wall-tiles, in Blockley et al 1995, 1267–93 Blagg, T F C, 1984 Roman architectural ornament in Kent. Archaeol Cantiana 100, 65–80 Blagg, T F C, 1990 Building stone in Roman Britain, in Stone: quarrying and building in England, AD 43–1525 (ed D Parsons), 33–49. Phillimore, Chichester Blagg, T F C, 2002 Roman Architectural Ornament in Britain. Brit Archaeol Rep 329, Oxford Blanning, E, 2006 The Archaeology of Beehive Querns in Kent. Unpubl undergrad dissert, Univ Canterbury Blockley, K, Blockley, M, Blockley, P, Frere, S, and Stow, S, 1995 Excavations in the Marlowe Car Park and Surrounding Areas. Archaeol Canterbury 5, Canterbury Blumstein, M, 1956 Roman pottery from Hoo. Archaeol Cantiana 70, 273–7 Böhme, A, 1972 Die Fibeln der Kastelle Saalburg und Zugmantel. Saalburg-Jahrbuch 29, 5–112 Böhme, H W, 1974 Germanische Grabfunde des 4. bis 5. Jahrhunderts zwischen unterer Elbe und Loire. Münchner Beiträge zur Vor- und Frühgeschichte 19, Munich Böhme-Schönberger, A, 1998 Die provinzialrömischen Fibeln bei Almgren, in 100 Jahre Fibelformen nach Oscar Almgren. Internationale Arbeitstagung 25.–28. Mai 1997, Kleinmachnow, Brandenburg (ed J Kunow), 351–66. Forschungen zur Archäologie im Land Brandenburg 5, Wünsdorf Böhner, K., 1958 Die fränkischen Altertümer des Trierer Landes. Verlag Gebr. Mann, Berlin Boon, G C, 1991 Tonsor Humanus: razor and toiletknife in antiquity. Britannia 22, 21–32 Boon, G C, 2000 The other objects of copper-alloy, in M G Fulford and J Timby 2000, 338–57 Booth, P, 2006a The pottery and other ceramic finds, A Romano-Celtic temple and settlement at Grimstock Hill, Coleshill, Warwickshire (J Magilton), 103–65. Trans Birmingham and Warwickshire Archaeol Soc 110 Booth, P (ed), 2006c Roman pottery from Section 1 of the Channel Tunnel Rail Link, Kent, in ADS 2006 Booth, P, 2007 Roman pottery, in Excavations at Barrow Hills, Radley, Oxfordshire, 1983–5, Volume 2: the Romano-British cemetery and Anglo-Saxon settlement (R Chambers and E McAdam), 33–8. Thames Valley Landscapes Monogr 25, Oxford Booth, P, Bingham, A-M and Lawrence, S, 2008 The Roman Roadside Settlement at Westhawk Farm, Ashford, Kent, Excavations 1998–9. Oxford Archaeol Monogr 2, Oxford Borrill, H, 1981 Casket burials, in Skeleton Green, A Late Iron Age and Romano-British site (C Partridge), 304–21. Britannia Monogr 2, London 407 Boyle, A and Early, R, 1998 Excavations at Springhead Roman Town, Southfleet, Kent. Oxford Archaeol Unit Occas Pap 1, Oxford Brailsford, J W, 1962 Hod Hill, Volume 1: antiquities from Hod Hill in the Durden Collection. Brit Mus Press, London Brailsford, J W, 1964 Guide to the Antiquities of Roman Britain, 3rd edn. Brit Mus Press, London Braithwaite, G, 2007 Faces from the Past: a study of Roman face pots from Italy and the western Roman provinces of the Roman Empire. Brit Archaeol Rep S1651, Oxford Brewer, R J, 1986a The bronze brooches, in Zienkiewicz 1986, 168–72 Brewer, R J, 1986b Other finger-rings and settings, in Zienkiewicz 1986, 144–5 Briggs, D, Haselgrove, C and King, C, 1992 Iron Age and Roman coins from Hayling Island temple. Brit Numis J 62, 1–62 Brigham, T, 1990 The late Roman waterfront of London. Britannia 21, 99–183 Brigham, T, 2001 The Thames and the Southwark waterfront in the Roman period, in London Bridge: 2000 years of a river crossing (B Watson, T Brigham and T Dyson), 12–27. Mus London Archaeol Service Monogr 8, London Brigham, T, Goodburn, D, Tyers, I, and Dillon, J, 1995 A Roman timber building on the Southwark Waterfront, London. Archaeol J 152, 1–72 Brodribb, G, 1987 Roman Brick and Tile. Alan Sutton, Stroud Brown, A E, 1994 A Romano-British shell-gritted pottery and tile manufacturing site at Harrold, Beds. Bedfordshire Archaeol J 21, 19–107 Brown, R A, 1986 The Iron Age and Romano-British settlement at Woodcock Hall, Saham Toney, Norfolk. Britannia 17, 1–58 Brown, A E and Sheldon, H L , 1969 Post-excavation work on the pottery from Highgate, London Arch 1(3), 60–5 Burnham, B C and Wacher, J S, 1990 The ‘Small Towns’ of Roman Britain. Batsford, London Bushe-Fox, J P, 1932 Third Report on the Excavations of the Roman Fort at Richborough, Kent. Rep Res Comm Soc Antiq London 10, Oxford Bushe-Fox, J P, 1949 Fourth Report on the Excavations of the Roman Fort at Richborough, Kent. Rep Res Comm Soc Antiq London 16, Oxford Butcher, S, 1990 The brooches, in Excavation of the Iron Age, Roman and Medieval Settlement at Gorhambury, St Albans (D S Neal, A Wardle and J Hunn), 115–20. Engl Herit Archaeol Rep 14, London Butcher, S, 1993 Brooches of copper alloy, in Woodward and Leach 1993, 149–57 Butcher, S, 2001 The brooches, in Anderson et al 2001, 41–74 Butcher, S with Penhallurick, R and Williams, D F, 2000–1 Roman Nor’nour, Isles of Scilly: a reconsideration. Cornish Archaeol 39–40, 5–44 408 Settling the Ebbsfleet Valley Callender, M H, 1965 Roman Amphorae. Oxford Univ Press, Oxford Carnap-Bornheim, C von, 1991, Die Schwertriemen bügel aus dem Vimose (Fünen). Kleine Schriften aus dem Vorgeschichtlichen Seminar der PhilippsUniversität Marburg 38, Marburg Catherall, P D, 1983 A Romano-British pottery manufacturing site at Oakleigh Farm, Higham, Kent. Britannia 14, 103–41 Champion, B, 1916 Outils en fer du musée de SaintGermain. Revue Archéologique 3 (5th ser), 211–46 Chantraine, H, 1961 s v uncia, in Paulys Realencyclopädie der classischen Altertumswissenschaft (ed A F Pauly), vol 9 A/1, Col 604–65. J B Metzler, Stuttgart Charpy, J-J, 1991 The Champagne region under Celtic rule during the fourth and third centuries B C, in The Celts (S Moscati, O H Frey, V Kruta, B Raferty and M Szabó), 243–50. Bompiani, Milan Charters, S, Evershed, R P, Blinkhorn, P W and Denham, V, 1995 Evidence for the mixing of fats and waxes in archaeological ceramics. Archaeometery 37(1), 113–27 Clarke, G, 1979 Pre-Roman and Roman Winchester. Part 2, The Roman cemetery at Lankhills. Winchester Studies 3, Oxford Clarke, G, 1982 The Roman villa at Woodchester. Britannia 13, 197–228 Cleal, R M J, 1991 Fired clay, in Redeemed from the Heath: the archaeology of the Wytch Farm Oilfield (1987–90) (P W Cox and C M Hearne), 149–56. Dorset Natur Hist Archaeol Soc Monogr 9, Dorchester Cool, H E M, 1979 A newly found inscription on a pair of silver bracelets from Castlethorpe, Buckinghamshire. Britannia 10, 165–8 Cool, H E M, 1990 Roman metal hair pins from Southern Britain. Archaeol J 147, 148–82 Cool, H E M, 1998 The brooches, in Cool and Philo 1998, 29–57 Cool, H E M, 2004 The Roman Cemetery at Brougham, Cumbria. Excavations 1966–67. Britannia Monogr 21, London Cool, H E M, 2006 Eating and drinking in Roman Britain. Cambridge Univ Press, Cambridge Cool, H E M and Philo, C, 1998 Roman Castleford, Vol. 1: the small finds. Yorkshire Archaeol 4, Wakefield Corney, M, 2000 The brooches, in M G Fulford and J Timby 2000, 322–38 Cox, P and Woodward, P, 1987 The Kimmeridge shale, in P J Woodward, The excavation of an Iron Age and Romano-British settlement at Rope Lake Hole, Corfe Castle, Dorset, in Romano-British Industries in Purbeck (N Sunter and P J Woodward), 165–72. Dorset Natur Hist Archaeol Soc Monogr 6, Dorchester Creighton, J, 2006 Britannia: the creation of a Roman province. Routledge, London Crew, P, 1995 Currency Bars and other forms of trade iron. Hist Metall Soc Archaeol Data Sheet 8 Crummy, N, 1983 The Roman Small Finds from Excavations in Colchester 1971–9. Colchester Archaeol Rep 2, Colchester Crummy, N, 2000 Bone working in Roman Britain: a model for itinerant craftsmen, in L’artisanat romain: evolutions, continuites et ruptures (Italie et provinces occidentales) (ed M Polfer), 97–109. Éditions Monique Mergoil, Montagnac Crummy, N, 2001 Nail-cleaners: regionality at the clean edge of Empire. Lucerna 22, 2–6 Crummy, N, 2005a From bracelets to battle-honours: military armillae from the Roman conquest of Britain, in Image, Craft and the Classical World: essays in honour of Donald Bailey and Catherine Johns (ed N Crummy), 93–106. Instrumentum Monogr 29, Montagnac Crummy, N, 2005b The mixed grill over-egged. J Roman Pott Stud, 12, 59–63 Crummy, N, 2006 Worshipping Mercury on Balkerne Hill, in A Victory Celebration: papers on the archaeology of Colchester and Late Iron Age–Roman Britain presented to Philip Crummy (ed P Ottaway), 55–68. Colchester Archaeol Trust, Colchester Crummy, N, 2007 Brooches and the cult of Mercury. Britannia 38, 225–30 Crummy, P, 1984 Excavations at Lion Walk, Balkerne Lane, and Middlesborough, Colchester, Essex. Colchester Archaeol Rep 3, Colchester Cunliffe, B W, 1968 Report on the Excavations of the Roman fort at Richborough, Kent. Rep Res Comm Soc Antiq London 23, Oxford Cunliffe, B W, 1971 Excavations at Fishbourne, 1961–1969, Vol 2: the finds, Rep Res Comm Soc Antiq London 26, 27, London Cunliffe, B (ed), 1988 The Temple of Sulis Minerva at Bath. Volume 2 The finds from the Sacred Spring. Oxford Univ Comm Archaeol Monogr 16, Oxford Cunliffe, B and Poole, C, 1991 Danebury: an Iron Age hillfort in Hampshire. Vol 4: the excavations 1979–1988: the site. Counc Brit Archaeol Res Rep 73, London Cunliffe, B W and Poole, C 2008 The Danebury environs Roman programme: a Wessex landscape during the Roman era. Volume 2 – Part 7 Dunkirt Barn, Abbotts Ann, Hants 2005–2006. Oxford Univ Comm Archaeol Monogr 71, Oxford Curle, J 1911 Roman frontier post and its people: fort of Newstead in the parish of Melrose. Maclehose, Glasgow Dannell, G B 1986 Decorated samian forms, in Baldock: The excavation of a Roman and Pre-Roman settlement 1968–72 (eds Stead, I M and Rigby, V), 213–22. Britannia Monog Ser 7, London Dannell, G B, 1999 Decorated South Gaulish samian, in Symonds and Wade 1999, 13–74 Dannell, G B, Dickinson, B M, Hartley, B R, Mees, A W, Polak, M, Vernhet, A, and Webster, P, 2003 Gestempelte Südgallische Reliefsigillata (Drag. 29) aus den Werkstätten von La Graufesenque. Verlag des Römisch-Germanischen Zentralmuseums, Mainz Bibliography Darling, M J, 1994 Guidelines for the Archiving of Roman Pottery. Study Group for Roman Pottery Guidelines Advisory Document 1 Darrah, R, 1982 Working unseasoned oak, in Woodworking Techniques before AD 1500 (ed S McGrail), 219–29. Brit Archaeol Rep S66, Oxford Davies, B, Richardson, B, and Tomber, R, 1994 The Archaeology of Roman London Volume 5: A dated corpus of early Roman Pottery from the City of London. Counc Brit Archaeol Rep 98, York Davies, M, 2004 Cabriabanus – a Romano-British tile craftsman in Kent. Archaeol Cantiana 124, 163–82 Derks, T, 1995 The ritual of the vow in Gallo-Roman religion, in Integration in the Early Roman West: the role of culture and ideology (eds J Metzler, M Millett, N Roymans and J Slofstra), 111–27. Dossiers d’Archéologie du Musée National d’Histoire et d’Art 4, Luxembourg Detsicas, A P, 1967 Excavations at Eccles, 1966. Archaeol Cantiana 82, 162–78 Detsicas, A P, 1968 Excavations at Eccles, 1967. Archaeol Cantiana 83, 39–48 Detsicas, A P, 1977 First-century pottery manufacture at Eccles, Kent, in Roman Pottery Studies in Britain and Beyond: papers presented to John Gillam, July 1977 (eds J Dore and K Greene), 19–36. Brit Archaeol Rep S30, Oxford Detsicas, A 1983 The Cantiaci. Alan Sutton, Gloucester Dewey, H, Bromehead, C E N, Chatwin, C P and Dines, H G, 1924 The Geology of the Country around Dartford. Explanation of Sheet 271. HMSO, London Dickinson, B, 1990 The samian ware in A Roman, Anglian and Medieval Site at Blackfriars Street, Carlisle (M R McCarthy), 213–36. Cumberland and Westmorland Antiquarian and Archaeol Soc Res Ser 4, Stroud Dickinson, B M, 2000 The samian, in The Roman Baths and Macellum at Wroxeter Excavations by Graham Webster 1955–85 (ed P Ellis), 282–302. English Heritage, London Dickinson, B, Hartley, B R and Pengelly, H W, 2004 Fabric supply: Class S, samian wares, in Cool 2004, 345–50 Dines, H G, Holmes, S C A and Robbie, J A, 1954 Geology of the Country around Chatham (One inch Geological Sheet 272, New Series). HMSO, London Down, A, 1971 The Roman cemetery at St. Pancras, in Chichester Excavations 1 (A Down and M Rule), 53–126. Phillimore, Chichester Down, A, 1978 Chichester Excavations 3. Phillimore, Chichester Driel-Murray, C van, 2001 Footwear in the NorthWestern provinces of the Roman Empire, in Stepping through Time (O Goubitz, C van DrielMurray, and W Groenman-van Waateringe), 337–76. Stichting Promotie Archeologie, Zwolle Drury, P J, 1988 The Mansio and Other Sites in the Southeastern Sector of Caesaromagus. Counc Brit Archaeol Res Rep 66, London 409 Drury, P J and Rodwell, W R, 1973 Excavations at Gun Hill, West Tilbury. Essex Archaeol Hist 5, 48–112 Dudd, S N, and Evershed, R P, 1999 Unusual triterpenoid fatty acyl ester components of archaeological birch bark tars. Tetrahedron Letters 40(2), 359–62 Dummer, J., 1987 S.v. Hase, in Lexikon der Antike (eds J Irmscher, with R Johne, 219. Gondrom, Bindlach Durham, E, and Poole, C, 2008 The building materials, in The Danebury environs Roman programme: a Wessex landscape during the Roman era. Volume 2 – Part 3 Fullerton, Hants, 2000 and 2001 (B W Cunliffe, and C Poole), Oxford Univ Comm Archaeol Monogr, Oxford Duncan, H B with Cool, H E M. and Stead, I M, 2005 The small finds, in Ferrybridge Henge: the ritual landscape (ed I Roberts), 152–7. Yorkshire Archaeol 10, Leeds Dungworth, D, 1998 Mystifying Roman nails: clavus annalis, defixiones and minkisi, in TRAC 97: proceedings of the Seventh Annual Theoretical Roman Archaeology Conference (eds C Forcey, J Hawthorne and R. Witcher), 148–59. Oxbow, Oxford Eckardt, H, 2002 Illuminating Roman Britain. Monographies Instrumentum 23, Éditions Monique Mergoil, Montagnac Edlin, H L, 1949 Woodland Crafts in Britain: an account of the traditional uses of trees and timbers in the British countryside. Batsford, London Effenberger, A and Severin, H G, 1992 Das Museum für Spätantike und Byzantinische Kunst/Staatliche Museen zu Berlin. Philipp von Zabern, Mainz am Rhein Egan, G, 1998 The Medieval Household. Daily living c. 1150 – c. 1450. Medieval finds from excavations in London 6, Museum of London, London Egan, G and Pritchard, F, 1991, Dress Accessories c. 1150 – c. 1450. Medieval finds from excavations in London 3, Museum of London, London Egging Dinwiddy, K and Schuster, J, 2009 Thanet’s longest excavation. Archaeological investigations along the route of the Weatherlees – Margate – Broadstairs wastewater pipeline, in Kentish sites and sites of Kent. A miscellany of four archaeological excavations (Andrews, P, Egging Dinwiddy, K, Ellis, C, Hutcheson, A, Phillpotts, C, Powell, A B and Schuster, J), 57–174. Wessex Archaeol Rep 24, Salisbury Erskine, J G P and Ellis, P, 2008 Excavations at Atworth Roman Villa, Wiltshire 1970–1975, Wiltshire Archaeological and Natural History Magazine 101, 51–129 Espérandieu, É, 1931 Receuil générale des bas-reliefs, statues et bustes de la Germanie Romaine. Éditions G van Oest, Paris/Brussels Evans, C and Hodder, I (eds), 2006 Marshland Communities and Cultural Landscapes: the Haddenham Project: volume 2. MacDonald Inst Archaeol, Cambridge 410 Settling the Ebbsfleet Valley Evans, J, 1987 Graffiti and the evidence of literacy and pottery use in Roman Britain. Archaeol J 144, 191–204 Evans, J, 2001 Material approaches to the identification of different Romano-British sites types, in Britons and Romans: advancing an archaeological agenda (eds S James and M Millett), 26–35. Counc Brit Archaeol Res Rep 125, York Evans, J 2004 The pottery vessels, in Cool 2004, 333–64 Evans, J 2007 Styles of pottery deposition at a Roman rural site in Hampshire, in Roman Finds: context and theory (eds R Hingley and S Willis), 176–85. Oxbow, Oxford Evans, K J, 1974 Excavations on a Romano-British site, Wiggonholt, 1964, Sussex Archaeol Collect 112, 97–115 Farrar, R A H, 1973 The techniques and sources of Romano-British Black Burnished Ware, in Current Research in Romano-British Coarse Pottery (ed A P Detsicas), 67–103. Counc Brit Archaeol Rep 10, London Ferrandini Troisi, F, 1981 Un vasetto per medicinale con iscrizione greca. Archeologia classica 33, 329–31 Feugère, M, 1985 Les fibules en Gaule Méridionale: de la conquête à la fin du Ve siècle après J-C. Revue Archéologique de Narbonnaise, Supplément 12, Paris Fitzpatrick, A P, 1985 The Celtic coins, in France and Gobel 1985, 49–66 Fölzer, E, 1913 Die Bilderschüsseeln der ostgallischen Sigillata-Manufakturen. Römische Keramik in Trier I. Marcus and Weber, Bonn Forrer, R, 1919 Römische Mühlen-, Töpferei- und Handelsbetriebe, Metallwerkstätten und Waffenfunde in Straßburg. Cahiers Alsaciens d’Archéologie, d’Art et d’Histoire 10, 37–40, 988–1078 Fowler, E, 1960 The origins and development of the penannular brooch in Europe. Proc Prehist Soc 26, 149–77 France, N E and Gobel, B M, 1985 The Romano-British Temple at Harlow, Essex, West Essex Archaeological Group, Gloucester Frayn, J M, 1978 Home-baking in Roman Italy. Antiquity 52, 28–33 Frere, S S, 1972 Verulamium Excavations I. Rep Res Comm Soc Antiq London 28, Oxford Frere, S S, 1984 Verulamium Excavations III. Oxford Univ Comm Archaeol Monog 1, Oxford Frere, S, and Stow, S, 1983 The Archaeology of Canterbury, vol 7: excavations in the St George’s Street and Burgate Street areas. Canterbury Archaeol Trust, Canterbury Fulford, M G, 1975 The pottery, in Excavations at Porchester Castle, Vol 1: Roman (B W Cunliffe), 270–367. Rep Res Comm Soc Antiq London 32, Oxford Fulford, M G and Bird, J, 1975 Imported pottery from Germany in late Roman Britain. Britannia 6, 171–81 Fulford, M G and Timby, J, 2000 Late Iron Age and Roman Silchester: excavations on the site of the ForumBasilica 1977, 1980–86. Britannia Monogr Ser 15, London Fulford, M G and Timby, J, 2001 Timing devices, fermentation vessels, ‘ritual’ piercings? A consideration of deliberately ‘holed’ pots from Silchester and elsewhere. Britannia 32, 293–7 Garrard, I P, 1995 Other objects of copper alloy and silver, in Blockley et al 1995, 1005–62 Gillam, J P, 1970 Types of Roman Coarse Pottery Vessels in northern Britain, 3 edn. Oriel Press, Newcastle upon Tyne Gillam, J P, 1976 Coarse fumed ware in north Britain and beyond. Glasgow Archaeol J 4, 57–80 Gobel, B.M, 1985 Brooches, in France and Gobel 1985, 70–81 Going, C J, 1987 The Mansio and Other Sites in the Southeastern Sector of Caesaromagus: the Roman pottery. Counc Brit Archaeol Res Rep 62, London Going, C J, 1992a Economic ‘long waves’ in the Roman period? A reconnaissance of the Romano-British ceramic evidence. Oxford J Archaeol 11(1), 93–115 Going, C J, 1992b The other pottery, in The Temple and other Sites in the North-eastern sector of Caesaromagus (N P Wickenden), 93–115. Counc Brit Archaeol Res Rep 75, London Going, C J, 1993 Roman pottery from the Grubenhäuser, in Excavations at Mucking 2: the Anglo-Saxon settlement (H Hamerow), 71–2. Engl Herit Archaeol Rep 21, London Going, C J, Hassall, M W C and Wright, R P, 1987 The graffiti, in Going 1987, 102–4 Goodburn, D, 1991a A Roman timber-framed building tradition. Archaeol J 148, 182–204 Goodburn, D, 1991b Wet sites as archives of ancient landscapes, in Wet Site Excavation and Survey (eds J Coles and D Goodburn), 51–3. WARP, Exeter Goodburn, D, 1991c New light on early ship and boatbuilding in the London area, in Waterfront Archaeology: proceedings of the third international conference, Bristol 1988 (ed G Good, R Jons and M Ponsford), 105–11. Counc Brit Archaeol Res Rep 74, York Goodburn, D, 1994 Spontning I London under Viking tid och medeltid, 17–23. Utblik, Gotland Goodburn, D, 1995 From tree to town, in Brigham et al 1995, 33–59 Goodburn, D, 1998 The death of the wildwood and birth of woodmanship in SE England, in Hidden Dimensions (ed C Bernick), 130–8. UBC Press, Vancouver Goodburn, D, 2001 Wooden remains as an archaeological resource: some insights from the London Wetlands, in Estuarine archaeology: the Severn and beyond (ed S Rippon), 187–96. Archaeology in the Severn Estuary 11, Bristol Goodburn, D, 2004 Summary Analysis of the Woodwork found During PCA Excavations at Token House Yard, Bibliography City of London. Unpublished report for PreConstruct Archaeol Goodburn, D, 2005 The Wood, in Seeley and Drummond-Murray 2005, 191–97 Goodburn, D, 2008 Worked wood, in Smith, A, Brown, L and Brady, K, A Romano-British landscape at Brockley Hill, Stanmore, Middlesex: Excavations at Brockley Hill House and the former MoD site. Trans London Middlesex Archaeol Soc 59, 136–138 Goodburn, D, 2009 Woodwork evidence in Great houses, moats and mills on the south bank of the Thames: medieval and Tudor Southwark and Rotherhithe (eds Blatherwick S and Bluer, R). Mus London Archaeol Monog 47, London Goodburn, D, forthcoming The Roman woodworking evidence, in Excavations in Borough High Street, Southwark (Brown, G and Pickard, C). PreConstruct Archaeol Monog, London Goodburn, D and Redknap, M, 1988 Replicas and wrecks from the Thames area. London Archaeol 6(1), 7–10, 19–22 Goodburn, R, 1984 The non-ferrous metal objects, in Frere 1984, 19–67 Goodburn, R, Hassall, M C W and Tomlin, R S O, 1978 Roman Britain in 1977. Britannia 9, 404–85 Goodman, W L, 1964 The History of Woodworking Tools. G Bell, London Gosden, C and Lock, G, 2003 Frilford: a RomanoBritish ritual pool in Oxfordshire. Curr Archaeol 184, 156–9 Gose, E 1972 Der gallo-römische Tempelbezirk im Altbachtal zu Trier. Trierer Grabungen und Forschungen 7, Mainz Gramatopol, M, 1974 Les pierres gravées du Cabinet numismatique de l’Academie Roumaine. Coll Latomus 138, Brussels Green, C, 1978 Flavian ‘ring and dot’ beakers from Londinium: Verulamium form 130 and allied types, in Arthur and Marsh 1978, 109–18 Green, M J, 1975, Romano-British non-ceramic model objects in south-east Britain. Archaeol J 132, 54–70 Green, M J, 1976 The Religions of Civilian Roman Britain. Brit Archaeol Rep 24, Oxford Green, M J, 1981 Model objects from military areas of Roman Britain. Britannia 12, 253–69 Greene, K, 1978 Mould-decorated Central Gaulish glazed ware in Britain, in Arthur and Marsh (eds) 1978, 31–60 Greene, K, 1979 Report on the Excavations at Usk 1965–76: the Pre-Flavian finewares. Univ Wales Press, Cardiff Grew, F and Griffiths, N, 1991 The pre-Flavian military belt: the evidence from Britain. Antiquity 109, 47–84 Grocock, C and Grainger, S, 2006 Apicius: a critical edition with an introduction and an English translation of the Latin recipe text Apicius. Prospect Books, Totnes Grunberg, J M, 2002 Middle Palaeolithic birch-bark pitch. Antiquity 76, 15–16 411 Guiraud, H, 1989 Bagues et anneaux à l’époque romaine en Gaule. Gallia 46, 173–211 Haalebos, J K, 1979 PRIMUS, CELADVS, und SENICIO. Rei Cret Rom Fautores Acta 19/20, 121–35 Haalebos, J K, 1986 Fibulae uit Maurik. Oudheidkundige Mededelingen, Supple 65, Leiden Hally, D J, 1983 Use alteration of pottery vessel surfaces: an important source of evidence for the identification of vessel function. N Amer Archaeol 4, 3–26 Hammerson, M 1988 Roman coins from Southwark, in Excavations in Southwark 1973–76 and Lambeth 1973–79 (ed P Hinton), 417–27. London Middlesex Archaeol Soc/Surrey Archaeol Soc Jnt Publ 3, London Harding, D W, 1974 The Iron Age in Lowland Britain. Routledge and Kegan Paul, London Harker, S R, 1969 Springhead – the well, F.19. Archaeol Cantiana 85, 139–48 Harrison, L, 2008 Roman brick and tile, in Booth et al 2008, 259–66 Hartley, K F, 1972 The mortarium stamps, in Frere 1972, 371–81 Hartley, K F, 1984 The mortarium stamps, in Frere 1984, 280–91 Hartley, K F, 1998 The incidence of stamped mortaria in the Roman Empire, with special reference to imports to Britain, in Form and Fabric; studies in Rome’s material past in honour of B R Hartley (ed J Bird), 199–218. Oxbow Monogr 80, Oxford Hartley, K F, 1999 The stamped mortaria, in Symonds and Wade 1999, 195–211 Hartley, K F and Tomber, R, 2006 A mortarium bibliography for Roman Britain. J Roman Pott Stud 13 Haselgrove, C, 2005 A trio of temples: a reassessment of Iron Age coin deposition at Hayling Island, Harlow and Wanborough, in Haselgrove and Wigg-Wolf (eds) 2005, 381–418 Haselgrove, C and Millett, M, 1997 Verlamion reconsidered, in Reconstruction Iron Age Societies. New Approaches to the British Iron Age (eds A Gwilt and C Haselgrove). 282–96. Oxbow Monogr 71, Oxford Haselgrove, C and Wigg–Wolf, D (eds), 2005 Iron Age Coinage and Ritual Practices. Studien zu Fundmunzen der Antike 20, Mainz Hassall, M W C and Tomlin, R S O, 1989 Roman Britain in 1988: II Inscriptions. Britannia 20, 327–45 Häßler, H-J, 1995, Die vorrömische Eisenzeit, in Geschichte des Landes zwischen Elbe und Weser 1. Vor- und Frühgeschichte (eds H-E Dannenberg and H-J Schulze), 121–59. Landschaftsverband Stade, Stade Hattatt, R, 2000 A Visual Catalogue of Richard Hattat’s Ancient Brooches. Oxbow, Oxford Hawkes, C F C and Hull, M R, 1947 Camulodunum: first report on the excavations at Colchester 1930–1939. Rep Res Comm Soc of Antiq London 14, London 412 Settling the Ebbsfleet Valley Heard, K, and Goodburn, D, 2003 Investigating the Maritime History of Rotherhithe. Mus London Archaeol Service Archaeol Stud 11, London Henig, M, 1974 A Corpus of Roman Engraved Gemstones from British Sites. Brit Archaeol Rep 8, Oxford Henig, M, 1984 Religion in Roman Britain. Batsford, London Henig, M, 1993 Votive objects: images and inscriptions, in Woodward and Leach 1993, 89–111 Henig, M, 1995 The Roman finger rings, in Blockley et al 1995, 1000–4 Hermet, F, 1934 La Graufesenque (Condatomago). Paris Heron, C, Evershed, R P, Chapman, B, and Pollard, A M, 1991 Glue, disinfectant and ‘chewing gum’ in prehistory, in Proceedings of a Conference on the Application of Scientific Techniques to Archaeology (eds P Budd, B C G Chapman, C Jackson, R C Janaway and B C Ottaway), 325–31. Oxbow, Oxford Hills, C, 1977 The Anglo-Saxon Cemetery at Spong Hill, North Elmham, Part 1. E Anglian Archaeol 6, Dereham Hingley, R, 1990 Iron Age ‘currency bars’: the archaeological and social context. Archaeol J 147, 91–117 Hinton, D A, 1996 Southampton finds Vol 2: The gold, silver and other non-ferrous alloy objects from Hamwic, and the non-ferrous metalworking evidence. Southampton Archaeol Monogr 6, Southampton Hinton, D A and Parsons, A L, 1996 Pins, in Hinton 1996, 14–37 Holman, D J, 2000 Iron Age coinage in Kent: a review of current knowledge. Archaeol Cantiana 120, 205–33 Holman, D J, 2005a Iron Age coinage and settlement in east Kent. Britannia 36, 1–54 Holman, D J, 2005b Iron Age coinage from Worth, Kent and other possible evidence of ritual deposition in Kent, in Haselgrove and Wigg-Wolf (eds) 2005, 265–85 Holmes, A G, 1989 A Romano-British site at Shedfield, Hants. Proc Hampshire Fld Club Archaeol Soc 45, 25–41 Holmes, S, 1995 Seal boxes from Roman London. London Archaeol 7(15), 391–5 Hondius-Crone, A 1955 The Temple of Nehalennia at Domburg. J M Meulenhoff, Amsterdam Hooley, D, 2001, Copper alloy and silver objects, in Anderson et al, 75–116 Howe, T, Jackson, G and Maloney, C, 2001 Archaeology in Surrey 2000. Surrey Archaeol Collect 88, 343–63 Huld-Zetsche, I, 1993 Trierer Reliefsigillata Werkstatt II (Materialen zur römisch-germanischen Keramik 12). R Habelt, Bonn Hull, M R, 1959 The fibulae, in Penn 1959, 48–9 Hull, M R, 1963 The Roman Potters’ Kilns of Colchester. Rep Res Comm Soc Antiq London 21, London Hull, M R, 1964 Brooches, in Winchester Excavations 1949–1960, vol 1 (B Cunliffe), 88–90. City of Winchester Mus Library Comm, Winchester Hull, M R, 1967 The Nor’nour brooches, 28–64 in Excavations on Nor’nour in the Isles of Scilly, 1962–6 (D Dudley). Archaeol J 124, 1–64 Hull, M R, 1970 Brooches, in Harker 1970, 143 Hume, I N, 1954 Romano-British potteries on the Upchurch Marshes. Archaeol Cantiana 68, 72–90 Hylton, T, 1994 Other non-ferrous objects, in Williams and Zeepvat 1994, 303–19 Jackson, R, 1985 Cosmetic sets from Late Iron Age and Roman Britain. Britannia 16, 165–92 Jackson, R, 1990 Camerton, the Late Iron Age and Early Roman Metalwork. Brit Mus Publ, London Jackson, R, 1992 The iron objects in Roman Warrington: excavations at Wilderspool 1966–9 (J Hinchliffe and J H Williams with F Williams), 78–86. Brigantia Monog Ser 2, Univ Dept of Archaeol, Manchester Jenkins, F, 1958 The cult of the ‘pseudo Venus’ in Kent, Archaeol Cantiana 72, 60–76 Jenkins, F, 1981 Pipe-clay statuettes, ‘theatre mask’ and sculptured chalk objects and their possible significance, in The excavation of the Roman forts of the Classis Britannica at Dover, 1970–1977 (B Philp), KARU, Dover Jenkins, F, 1995 The pipeclay figurines, in Blockley et al 1995, 1180–3 Jersey, P de, 2005 Deliberate defacement of British Iron Age coinage, in Haselgrove and Wigg–Wolf (eds) 2005, 85–113 Jessup, R F, 1928 A Romano-British settlement at Springhead. Antiq J 8, 337–43 Jones, G P 2009 in Egging Dinwiddy and Schuster 2009 Jones, M U, 1972 Potters’ graffiti from Mucking, Essex. Antiq J 52(2), 335–8 Jones, M U and Rodwell, W J, 1973 The Romano-British Pottery Kilns at Mucking. Essex Archaeol Hist 5, 13–47 Jope, E M, 2000 Early Celtic Art in the British Isles. Clarendon, Oxford Keller, P T, 1988 Quern production at Folkestone, South East Kent: An interim report. Britannia 20, 193–201 Kenyon, K, 1948 Excavations at the Jewry Wall site, Leicester. Rep Res Comm Soc Antiq London 15, Oxford King, D, 1980 Petrology, dating and distribution of querns and millstones: the results of research in Bedfordshire, Buckinghamshire, Hertfordshire and Middlesex. Bull Inst Archaeol London 23, 65–126 Koch, J K, 1997 Ein Jochbeschlag aus Ergersheim, Lkr Neustadt a d Aisch-Bad Windsheim. Beiträge zur Archäologie in Mittelfranken 3, 162–71 Koch, U, Welck, K v, and Wieczorek, A, 1996 Die Bevölkerung Nordgalliens: Einheimische und Fremde, in Die Franken – Wegbereiter Europas. Vor 1500 Jahren: König Chlodwig und seine Erben (eds A Wieczorek, P Périn, K v Welck, and W Menghin), 840–52. Exhibition Catalogue Reiss-Museum Mannheim, Mannheim/Mainz Bibliography Laubenheimer, F, 1985 La Production des Amphores en Gaule Narbonnaise. Centre de Recherches d’Histoire Ancienne 66 / Annales Littéraires de l’Université de Besançon 327, Paris Laubenheimer, F, 2003 Amphorae and vineyards from Burgundy to the Seine. J Roman Pott Stud 10, 32–44 Lawrence, S and Smith, A, 2009 Between Villa and Town. Excavations of a Roman Roadside Settlement and Shrine at Higham Ferrers, Northants. Oxford Archaeol Monogr 7, Oxford Lawson, A K, 1995 A hoard of Roman harness fittings, in Blockley et al 1995, 984–1000 Leach, P, 1993 The pottery, in Woodward and Leach, 219–49 Leach, P with Evans, C J, 2001 Excavation of a RomanoBritish Roadside Settlement in Somerset: Fosse Lane, Shepton Mallet 1990. Britannia Monogr 18, London Lemant, J.-P, 1985 Le cimetière et la fortification du basempire de Vireux-Molhain, Dép. Ardennes. RömischGermanisches Zentralmuseum Monographien 7, Mainz Liversidge, J, 1973 Britain in the Roman Empire. Cardinal, London Lloyd-Morgan, G, 2001 Objects of copper-alloy, bone, antler, jet and shale, in Excavations in the Extramural Settlement of Roman Alchester, Oxfordshire, 1991 (P M Booth, J Evans and J Hiller), 221–34. Oxford Archaeol Monogr 1, Oxford Lowther, A W G, 1935 An early Iron Age oven at St Martha’s Hill, near Guildford. Surrey Archaeol Collect 43, 113–5 Lowther, A W G, 1972 Appendix G: the ventilator, in A Roman building at Chalk, near Gravesend (D E Johnston). Britannia 3, 146–7 Lowther, A W G, 1976 Romano-British chimney-pots and finials. Antiq J 56, 35–48 Lowther, A W G, 1987 Roller impressions on Plaster in Meates 1987, 46 Lucquin, A, Javier March, R, Cassen, S, 2007 Analysis of adhering organic residues of two coupes-a-socles from the Neolithic funerary site ‘La Houge Bie’ in Jersey: evidences of birch bark tar utilisation. J Archaeol Sci 34, 704–10 Ludowici, W, and Ricken, H, (eds), 1948 Katalog VI meiner Ausgrabungen in Rheinzabern 1901–1914. Die Bilderschüsseln der römischen Töpfer von Rheinzabern, Tafelband. Speyer Lyne, M, 2006 The late Iron Age and Roman pottery from Thurnham Roman Villa, Thurnham, Kent, in ADS 2006 Lyne, M A B, forthcoming The organisation of the Bestwall Black Burnished Ware (BB1) industry, in Excavations at Bestwall Quarry,Wareham 1992–2005: Volume 2: the Iron Age and later landscape (L Ladle). Dorset Natur Hist Archaeol Soc Monogr, Dorchester Lyne, M A B and Jefferies, R S, 1979 The Alice Holt/Farnham Roman Pottery Industry. Counc Brit Arachaeol Res Rep 30, London 413 Macdonald, J, 1977 Pagan religions and burial practices in Roman Britain, in Burial in the Roman World (ed R Reece), 35–8. Counc Brit Archaeol Res Rep 22, London MacGregor, A, 1985 Bone, Antler, Ivory and Horn: the technology of skeletal materials since the Roman period. Croom Helm, London MacGregor, M, 1976 Early Celtic Art in North Britain: a study of decorative metalwork from the third century BC to the third century AD. Leicester Univ Press, Leicester Mackreth, D F, 1978 The Roman brooches, in Down 1978, 277–88 Mackreth, D F, 1981 The brooches, in Skeleton Green: a late Iron Age and Romano-British site (C Partridge), 130–51. Britannia Monogr 2, London Mackreth, D F, 1985 The Roman brooches, in Prehistoric, Roman, Saxon and medieval artefacts from the southern fen edge, Cambridgeshire (A Taylor), 13–29. Proc Cambridge Antiq Soc 74, 1–52 Mackreth, D F, 1989 The Roman brooches from Chichester, in Chichester Excavations 6 (A Down), 182–194. Phillimore, Chichester Mackreth, D F, 1994 Copper alloy and iron brooches, in Roman Alcester: southern extramural area, 1964–1966 excavation, Part 2 Finds and discussion (S Cracknell and C Mahany), 162–77. Counc Brit Archaeol Res Rep 97, York Mackreth, D F, 1995 Pre-Roman and Roman brooches, in Blockley et al 1995, 955–82 Mackreth, D F, 1996 Brooches, in Excavations at Stonea, Cambridgeshire, 1980–85 (R P J Jackson and T W Potter), 296–327. Brit Mus Press, London Mackreth, D F, 1999 The brooches, in The excavation of a ceremonial site at Folly Lane, Verulamium (R Niblett), 216–22. Britannia Monogr 14, London Malim, T, 2005 Stonea and the Roman fens. Tempus, Stroud Manning, W H, 1971 The Piercebridge plough group, in Prehistoric and Roman Studies, Commemorating the Opening of the Department of Prehistoric and RomanoBritish Antiquities (ed G de G Sieveking), 125–36. Brit Mus, London Manning, W H, 1974 Objects of iron, in Neal 1974, 157–87 Manning, W H, 1976 Catalogue of the Romano-British Ironwork in the Museum of Antiquities Newcastle upon Tyne, Dept Archaeology, Univ Newcastle upon Tyne Manning, W H, 1985a Catalogue of the Romano-British Iron tools, Fittings and Weapons in the British Museum. Brit Mus Press, London Manning, W H, 1985b The iron objects, in Inchtuthil, the Roman Legionary Fortress Excavations 1952–65 (L F Pitts and J K St Joseph), 289–99, Britannia Monogr 6, London May, J, 1971 An Iron Age spout from Kirmington, Lincolnshire. Antiq J 51, 253–9 May, J, 1996 Dragonby: report on excavations at an Iron Age and Romano-British settlement in North Lincolnshire 1. Oxbow Monogr 61, Oxford 414 Settling the Ebbsfleet Valley Marsden, P, 1994 Ships of the Port of London: first to eleventh centuries AD. Engl Herit, London Marsh, G, 1978 Early second century fine wares in the London area, in Arthur and Marsh 1978, 119–223 Marsh, G, 1981 London’s samian supply and its relationship to the development of the Gallic Samian Industry, in Roman Pottery Research in Britain and North-West Europe: papers presented to Graham Webster (eds A C Anderson and A S Anderson), 173–238. Brit Archaeol Rep S123, Oxford Marsh, G, and Tyers, P, 1978 The Roman pottery from Southwark, in Southwark Excavations 1972–1974, 533–82. London Middlesex Archaeol Soc/Surrey Archaeol Soc Jnt Publ 1, London Martin-Kilcher, S, 1983 Les amphores romaines a huile de Betique (Dressel 20 et 23) d’Augst (Colonia Augusta Rauricorum) et Kaiseraugst (Castrum Raurucense). Un Rapport preliminiaire, in Prod. Y Com. Del Aceite en la Antiguedad (eds J M Blazquez and J Remesal), 337–47. II Congresso, Madrid Mattingly, D, 2007 An Imperial Possession: Britain in the Roman Empire. Penguin, London Mazza, P P A , Magi, M , Colombini, M P , Giachi, G , Landucci, F, Lemorini, C, Modugno, F, Ribechini, E, Martini, F and Sala, B, 2006 A new Palaeolithic discovery: tar-hafted stone tools in a European Mid-Pleistocene bone-bearing bed. J Archaeol Science 33(9), 1310–18 McCarthy, M 1991 The Roman waterlogged remains and later features at Castle Street, Carlisle: excavations 1981–2. Cumberland and Westmorland Antiq and Archaeol Society Res Ser 5, Kendal. McDonnell, G, 1983 Tap slags and hearth bottoms. Curr Archaeol 20, 81–3 McDonnell, G, 1995 Ore, slag, iron and steel, in Iron for Archaeologists: a review of recent work on the archaeology of early ironworking sites in Europe (eds P Crew and S Crew), 3–7. Plas Tan y Bwlch Occasional Paper 2, Tan y Bwlch, Gwynedd McGrail, S, 1993 Medieval Boat and Ship timbers from Dublin. Medieval Dublin excavations 1962–81, Ser B, Vol 3, Dublin McKinley, J I 2004b Human remains, pyre technology and cremation rituals, in Cool 2004, 283–310 Meates, G W, 1979 The Roman Villa at Lullingstone, Kent. Volume 1 – the site. Kent Archaeol Soc, Maidstone Meates, G W, 1987 The Roman Villa at Lullingstone, Kent. Volume II – the wall paintings and finds. Kent Archaeol Soc, Maidstone Medlycott, M, 1996 A medieval farm and its landscape: excavations at Stebbingford, Felsted, 1993. Essex Archaeol Hist 27, 102–81 Mees, A, 1995 Modelsignierte Dekorationen auf südgallischer Terra Sigillata, Konrad Theiss, Stuttgart Mepham, L N and Morris, E L, 1992 The pottery, 18–21, in Excavations at Norton Bavant Borrow Pit, Wiltshire (C A Butterworth), Wiltshire Archaeol Nat Hist Mag 85, 1–26 Merrifield, R, 1965 The Roman City of London. Ernest Benn, London. Metzler, J, 1995 Das treverische Oppidum auf dem Titelberg. Zur Kontinuität zwischen der spätkeltischen und der frührömischen Zeit in Nord-Gallien 1. Dossiers d’Archéologie du Musée National d’Histoire et d’Art 3, Luxemburg Metzler, J, Bis, R, Gaeng, C and Méniel, P, 2000 Vorbericht zu den Ausgrabungen im keltischrömischen Heiligtum auf dem Titelberg, in Kelten, Germanen und Römer im Mittelgebirgsraum zwischen Luxemburg und Thüringen (eds A Haffner and S von Schnurbein), 431–45. Habelt, Bonn Millett, M, 1979 An approach to the functional interpretation of pottery, in Pottery and the Archaeologist (ed M. Millett), 35–48. Institute of Archaeol Occ Pub No 4, London, Millett, M, 1986 The pottery, in Excavations on the Romano-British Small Town at Neatham, Hampshire 1969–1979 (M Millett and D Graham), 63–94. Hampshire Fld Club Archaeol Soc Monogr 3, Winchester Millett, M and McGrail, S, 1987 The archaeology of the Hasholme logboat. Archaeol. J 144, 69–155 Mills, J M, 1997 Samian ware, in Asthall, Oxfordshire: Excavations in a Roman ‘Small Town’ (P Booth), 106–13. Thames Valley Landscapes Monogr 9, Oxford Mills, J M, 1998 Samian, in Excavations alongside Roman Ermine Street, Cambridgeshire, 1996: the archaeology of the A1(M) Alconbury to Peterborough road scheme, (P Ellis, P, Hughes, G, Leach, P, Mould, C and Sterenberg J) 68–70. Brit Archaeol Rep 276, Oxford Mills, J M, 2004 The samian pottery, in The RomanoBritish Roadside Settlement at Wilcote, Oxfordshire III. Excavations 1997–2000 (A R Hands), 66–102. Brit Archaeol Rep 370, Oxford Mills, J M and Corney, M, 1993 Samian, in Excavations at County Hall, Dorchester, Dorset, 1988 in the Northwest Quarter of Durnovaria (R C J Smith), 42–3. Wessex Archaeol Rep 4, Salisbury Mills, J M and Woodward, P J, 1993 Shale and jet, in Woodward et al 1993, 139–45 Milne, G, 1985 The Port of Roman London. Batsford, London Milne, G and Milne, C, 1982 Medieval Waterfront Development at Trig Lane, London. London Middlesex Archaeol Soc Spec Pap 5, London Mitchell, N, 1998 Tile, in Boyle and Early 1998, 32–3 Monaghan, J, 1983 The Woodruff Collection. Archaeol Cantiana 99, 199–217 Monaghan, J, 1987 Upchurch and Thameside Roman Pottery: a ceramic typology for northern Kent, first to third centuries AD. Brit Archaeol Rep 173, Oxford Morris, C, 2000 Craft, Industry and Everyday life: wood and woodworking in Anglo-Scandinavian and medieval York. Archaeol York 17/13, York Mould, Q, 1998 The lead-alloy artefacts, in Cool and Philo 1998, 121–8 Bibliography Mould, Q, Carlisle, I and Cameron, E, 2003 Craft, Industry and Everyday life: leather and leatherworking in Anglo-Scandinavian and medieval York. Archaeol York 17/16, York Müller, F, 2002 Götter, Gaben, Rituale, Religion in der Frühgeschichte Europas. Kulturgeschichte der antiken Welt 92, Mainz am Rhein Müller, F., 2006 Sakrale Untiefen. Die Spuren vorgeschichtlichen Kultes in Gewässern, in Unterwasserarchäologie und Geschichtsbild: Akten des 2. Internationalen Kongresses für Unterwasserarchäologie (eds A Hafner, U Niffeler and U Ruoff), 110–21. Antiqua 40, Basel Mutz, A, 1976 Römisches Schmiedehandwerk. Augster Museumshefte 1, Augst Nayling, N and McGrail, S, 2004 Barlands Farm Romano-Celtic Boat. Counc Brit Archaeol Res Rep 138, York Neal, D S, 1974 The Excavation of the Roman Villa in Gadebridge Park, Hemel Hempstead, 1963–1968. Rep Res Comm Soc Antiq London 31, Oxford Neito, X and Puig, A, 2001 Excavacions arqueològiques subaquàtiques a Cala Culip 3 Culip IV: la Terra Sigil·lata decorada de La Graufesenque. Monografies del CASC 3, Girona, Catalonia Nickel, C, Thoma, M and Wigg-Wolf, D, 2008 Martberg. Heiligtum und Oppidum der Treverer Vol I. Der Kultbezirk: die Grabungen 1994–2004, Teil 1, Berichte zur Archäologie an Mittelrhein und Mosel 14, Koblenz Nuber, H U, 2000 Eine Grablege reicher Landbesitzer in Wehringen, in Wamser et al (eds), 166–70 O’Connell, M G and Bird, J, 1994 The Roman temple at Wanborough, excavation 1985–1986, Surrey Archaeol Collect 82, 1–168 Oldenstein, J, 1976 Zur Ausrüstung römischer Auxiliareinheiten. Bericht der Römisch-Germanischen Kommission 57, 49–284 Olivier, A C H, 1996 Brooches of silver, copper alloy and iron from Dragonby, in Dragonby, Report on Excavations at an Iron Age and Romano-British settlement in North Lincolnshire 1 (J May), 231–64. Oxbow Monogr 61, Oxford Orton, C, Tyers, P and Vince, A, 1993 Pottery in Archaeology. Cambridge Univ Press, Cambridge Oswald, F, 1945 Decorated ware from Lavoye. J Roman Stud 35, 49–57 Oswald, F and Pryce, T D, 1920 An Introduction to the Study of Terra Sigillata. Longmans, Green and Co, London Øye, I, 1988 Textile Equipment and its Working Environment, Bryggen in Bergen c 1150–1500. Bryggen Papers 2, Oslo Palmer, J 2001 Roman Purbeck stone: a new database. Proc Dorset Natur Hist Archaeol Soc 123, 104–9 Pannuti, U, 1983 Museo Archeologico Nazionale di Napoli. Catalogo della Collezione Glittica. Cataloghi dei Musei e Gallerie d’Italia 1, Rome Paynter, S, 2008 Metalworking remains, in Booth et al 2008, 267–301 415 Peacock, D P S, 1977 Pompeian Red Ware, in Pottery and Early Commerce; characterization and trade in Roman and later ceramics (ed D P S Peacock), 147–162. Academic Press, London Peacock, D P S, 1987 Iron Age and Roman quern production at Lodsworth, West Sussex. Antiq J 67, 61–85 Peacock, D P S and Williams, D F, 1986 Amphorae and the Roman economy: an introductory guide. Longman, London Pengelly, H, 1993 Samian, in Woodward et al 1993, 202–4 Penn, W S, 1957 The Romano-British settlement at Springhead. Excavation of the bakery, Site A. Archaeol Cantiana 71, 53–105 Penn, W S, 1959 The Romano-British settlement at Springhead. Excavation of Temple I, site C1. Archaeol Cantiana 73, 1–61 Penn, W S, 1960 Springhead: Temples III and IV. Archaeol Cantiana 74, 113–40 Penn, W S, 1964 Springhead: the temple ditch site. Archaeol Cantiana 79, 170–89 Penn, W S, 1968a Springhead: miscellaneous excavations. Archaeol Cantiana 83, 163–92 Penn, W S, 1968b Springhead. Archaeol Cantiana 83, 257 Perrin, J R, 1999 Roman pottery from excavations at and near to the Roman small town of Durobrivae, Water Newton, Cambridgeshire, 1956–58. J Roman Pott Stud 8, 1–141 Philp, B, 1973 Excavations in West Kent, 1960–70, Kent Archaeol Rescue Unit, Dover Philp, B, 1980 London Essex stamped ware from Kent. Kent Archaeol Rev 61, 21–3 Philp, B, 1991 The Roman Villa Site at Keston, Kent: first report (excavations 1968–78). Kent Monogr 6, Dover Philp, B, 1994 Warning: stolen finds from Roman Springhead. Kent Archaeol Rev 116, 131–2 Philp, B, Parfitt, K, Willson, J, Dutto, M, and Williams, W, 1991 The Roman Villa Site at Keston, Kent. Kent Archaeol Rescue Unit, Dover Philpott, R, 1991 Burial Practices in Roman Britain. Brit Archaeol Rep 219, Oxford Pitts, M, 2005 Pots and pits: drinking and deposition in late Iron Age south-east Britain. Oxford J Archaeol 24(2), 143–61 Plouviez, J, 1985 The late Romano-British pottery, in West Stow: the Anglo-Saxon Village, vol. 1: Text (S West), 82–5, E Anglian Archaeol 24, Ipswich Pollard, R J, 1982 The Roman Pottery of Kent. Unpubl PhD thesis, Univ Reading Pollard, R J, 1983 The coarse pottery, including the kiln products, in A Romano-British pottery manufacturing site at Oakleigh Farm, Higham, Kent (P D Catherall) Britannia 14, 121–38 Pollard, R J, 1987 The other pottery, in Meates 1987, 164–283 Pollard, R J, 1988 The Roman pottery of Kent. Kent Archaeol Soc, Maidstone 416 Settling the Ebbsfleet Valley Poole, C, 1995 Loomweights versus oven bricks, in Danebury: an Iron Age hillfort in Hampshire.Volume 6 (ed B Cunliffe), 285–6. Counc Brit Archaeol Res Rep 102, York Poole, C, 1998 Ceramic Building Material from the Shorne-Farningham Pipeline unpubl rep for Engineering Archaeological Services, Blaenau Ffestioniog, Gwynnedd Poole, C, 2000 Structural daub and baked clay, in The Danebury environs programme. The prehistory of a Wessex landscape volume 2 – Part 6 Houghton Down, Stockbridge, Hants, 1994 (eds B W Cunliffe, and C Poole), 115–19. Oxford Univ Comm Archaeol Monogr 49, Oxford Poole, C, 2006 Fired Clay from Beechbrook Wood in ADS 2006 Poole, C, 2008 Wall daub, painted plaster and mortar, in Cunliffe and Poole 2008, 166–73. Pretty, K, 1997 Appendix 1: small finds, in The Baths Basilica Wroxeter, Excavations 1966–90 (P Barker, R White, K Pretty, H Bird and M Corbishley), 249–58. Engl Herit Archaeol Rep 8, London. Pringle, S, 2004 A Spatial Analysis of Specialised Ceramic Building Materials in Early Roman London from the Conquest to the Flavian period (circa AD 50–80). Unpubl MSc dissert, Univ College London Pringle, S, 2005 Assessment of Ceramic Building Material from Ebbsfleet Roman Villa, Kent (ARCEBB01). unpubl assess rep Oxford Archaeology Pritchard F A 1986 Ornamental stonework from Roman London. Britannia 17, 169–90 Pritchard, F and Chapman, H, 1986 Basketwork, in The Roman Quay at St Magnus House, London: excavation at New Fresh Wharf, Lower Thames Street, London 1974–78 (L Miller, J Schofield and M Rhodes). London Middlesex Archaeol Soc Spec Pap 8, London Pryce, T D, 1949 Decorated samian, in Fourth Report on the Excavations of the Roman Fort at Richborough, Kent (ed J P Bushe-Fox), 160–83. Rep Res Comm Soc Antiq London 16, Oxford Pyke, J A, 1983 London Essex stamped ware from Otford. Kent Archaeol Rev, 72, 26–7 Rackham, O, 1976 Trees and woodland in the British landscape. Dent, London Raddatz, K, 1987 Der Thorsberger Moorfund. Katalog. Teile von Waffen und Pferdegeschirr, sonstige Fundstücke aus Metall und Glas, Ton- und Holzgefässe, Steingeräte, Offa-Bücher 65, Neumünster Radnóti, A, 1961 Ein Jochbeschlag der römischen Kaiserzeit, Saalburg-Jahrbuch 19, 18–36 Rajewski, Z, 1970 Pech und Teer bei den Slawen, Zeitschrift für Archäologie 4, 46–53 Redknap, M, 1998 Objects and waste material associated with metal-working, in Excavations at Kingscote and Wycomb, Gloucestershire (J R Timby), 111–3. Cotswold Archaeol Trust, Cirencester Reece, R, 1991 Roman Coins from 140 sites in Britain. Cotswold Stud 4, Cirencester Reece, R, 2002 The Coinage of Roman Britain. Tempus, Stroud Rees, S E, 1979 Agricultural Implements in Prehistoric and Roman Britain. Brit Archaeol Rep 69, Oxford Regert, M, Delacotte, M, Menu, M, Pètrequin, P and Rolando, C 1998 Identification of Neolithic hafting adhesives from two lake dwellings at Chalain Jura, France. Ancient Biomolecules 2, 81–96 Regert, M, Vacher, S, Moulherat, C and Decavallas, O, 2003 Adhesive production and pottery function during the Iron Age at the site of Grand Aunay (Sarthe, France). Archaeometry 45(1), 101–20 Remesal Rodriguez, J, 1986 La Annona Militaris y la Exportacion de aceite Betico a Germania. Universidad Complutense, Madrid RIB I, 1965 The Roman Inscriptions of Britain. Vol I: Inscriptions on stone (eds R G Collingwood and R P Wright). Clarendon Press, Oxford RIB II, 1991 The Roman Inscriptions of Britain. Vol II: Instrumentum domesticum (personal belongings and the like). Fascicule 3: Brooches, rings, gems, bracelets, helmets, shields, weapons, etc (eds R G Collingwood and R P Wright). Alan Sutton, Stroud RIB II, 1994 The Roman Inscriptions of Britain. Vol II: Instrumentum domesticum (personal belongings and the like). Fascicule 4.Wooden barrels, stilus-tablets, objects of wood, leather, oculists' stamps, etc (eds R G Collingwood and R P Wright). Alan Sutton, Stroud RIB II, 1995b The Roman Inscriptions of Britain. Vol II: Instrumentum domesticum (personal belongings and the like). Fascicule 8: Graffiti cut on coarse pottery before and after firing, etc (eds S S Frere and R S O Tomlin). Sutton, Stroud RIC, The Roman Imperial Coinage, vols I–VII. Spink, London Rice, P M, 1987 Pottery Analysis. A Sourcebook. Univ Chicago Press, Chicago Richardson, B, 1986 Pottery, in The Roman Quay at St Magnus House, London: excavations at New Fresh Wharf, Lower Thames Street, London 1974–78 (L Miller, J Schofield and M Rhodes), 96–138. London Middlesex Archaeol Soc Spec Pap 8, London Richardson, B and Tyers, P A, 1984 North Gaulish pottery in Britain. Britannia 15, 133–42 Richardson, K M, 1960 A Roman brooch from the Outer Hebrides with notes on others of its type. Antiq J 40, 200–13 Ricken, H and Fischer, C, 1963 Die Bilderschüsseln der römischen Töpfer von Rheinzabern. Textband mit Typenbildern zu Katalog VI der Ausgrabungen von Wilhelm Ludowici in Rheinzabern 1901–1914. Materialen zur römisch-germanischen Keramik 7, Bonn Rieckhoff-Pauli, S, 1977 Die Fibeln aus dem römischen Vicus von Sulz am Neckar. Saalburg-Jahrbuch 34, 5–28 Rigby, V, 1973 Potters’ stamps on Terra Nigra and Terra Rubra found in Britain, in Detsicas (ed) 1973, 2–24 Bibliography Rigby, V, 1989 Pottery from the Iron Age cemetery in Stead and Rigby 1989, 112–210 Riha, E, 1979 Die römischen Fibeln aus Augst und Kaiseraugst. Forschungen in Augst 3, Augst Riha, E, 1994 Die römischen Fibeln aus Augst und Kaiseraugst. Die Neufunde seit 1975. Forschungen in Augst 18, Augst Ritterling, E, 1905 Das frührömische Lager bei Hofheim i. T. Ausgrabungs- und Fundbericht. Annalen des Vereins für Nassauische Altertumskunde und Geschichtsforschung 34, 1–110 (Addenda 397–423) Rival, M, 1991 La charpenterie navale Romaine: matériaux, méthodes, moyens. CNRS, Paris Roberts, W I, 1982 Romano-Saxon Pottery. Brit Archaeol Rep 106, Oxford Robinson, P, 1998 A Late Iron Age or early Roman spout in the Stourhead Collection. Wiltshire Archaeol Natur Hist Mag 91, 145–7 Rodwell, W, 1978 Stamp-decorated pottery of the early Roman period in eastern England, in Arthur and Marsh 1978, 225–93 Roe, F 1998a The worked stone, in Boyle and Early 1998, 29–31 Roe, F, 2008b Worked stone, in Booth et al 2008, 188–195 Roesdahl, E (ed), 1992 Wikinger, Waräger, Normannen: die Skandinavier und Europa 800–1200. Philipp von Zabern, Mainz Roest, J van der, 1988 Die römischen Fibeln von ‘De Horden’. Berichten van de Rijksdienst voor het Oudheidkundig Bodemonderzoek 38, 141–202 Rogers, G B, 1974 Poteries Sigillées de la Gaule Centrale I. – Les motifs non figurés. Gallia Supplement 28, Paris Rogerson, A and Dallas, C, 1984 Excavations in Thetford 1948–1959 and 1973–1980. E Anglian Archaeol 22, Dereham Rollo, L, 2001 The Iron Age and Roman pottery, in Monument 97, Orton Longueville, Cambridgeshire: a late pre-Roman Iron Age and Early Roman farmstead (D F Mackreth), 46–77. E Anglian Archaeol 97, Cambridge Rook, T, 1987 The Roman villa site at Dicket Mead, Lockleys, Welwyn. Hertfordshire Archaeol 9 (1983–86), 79–175 Roth, H, and Wamers, E (eds), 1984 Hessen im frühen Mittelalter. Archäologie und Kunst. Thorbecke, Sigmaringen Rowsome, P, 2000 Heart of the City: archaeology at 1 Poultry. Engl Herit/Mus London, London Ruprechtsberger, E M, 1974 Eine Bleiphiole mit griechischer Inschrift aus Lauriacum. Römisches Österreich 2, 55–60 Russell, M, 1997 Relief-patterned daub in Betts et al (eds) 1997, 47–50 Salter, C J, 1998 The iron objects, in Boyle and Early 1998, 20–2 Sas, K and Thoen, H (eds), 2002 Schone Schijn, Romeinse juweelkunst in West-Europa. Peeters, Leuven Scaife, R G, 2000 Holocene vegetation development in London, in Sidell et al 2000, 111–17 417 Schlott, C, 1999 Zum Ende des spätlatënezeitlichen Oppidum auf dem Dünsberg (Gem. BiebertalFellingshausen, Kreis Gießen, Hessen). Forschungen zum Dünsberg 2. Montagnac Schuster, J, 2006 Die Buntmetallfunde der Grabung Feddersen Wierde. Chronologie – Chorologie – Technologie. Probleme der Küstenforschung im südlichen Nordseegebiet 30, Feddersen Wierde 6. Isensee, Oldenburg Schuster, J, 2009 Metalwork, in Andrews, P, West Thurrock: Late prehistoric settlement, Roman burials and the medieval manor house. Channel Tunnel Rail Link Excavations 2002. Trans Essex Soc Archaeol and Hist 40, 1–77 Scott, I R, 1985 Daggers, in Manning 1985, 152–9 Scott, I, 2009 Small finds, in Lawrence and Smith 2009, 198–249 Seager Smith, R H, 1993 Roman pottery, in Excavations at County Hall, Dorchester, Dorset, 1988 (R J C Smith), 41–61. Wessex Archaeol Rep 4, Salisbury Seager Smith, R H, 2009 The Romano-British pottery, in Cambourne New Settlement: Iron Age and Romano-British Settlement on the Clay Uplands of West Cambridgeshire (J Wright, M Leivers, R Seager Smith and C J Stevens). Wessex Archaeol, Salisbury Seager Smith, R H and Davies, S M, 1993 Roman Pottery, in Woodward et al 1993, 202–89 Sealey, P R, 1999 Finds from the Cauldron Pit, in The Archaeology of Ardleigh, Essex: excavations 1955–1980 (N R Brown), 117–24. E Anglian Archaeol 90, Chelmsford Seeley, F and Drummond-Murray, J, 2005 Roman Pottery Production in the Walbrook Valley: excavations at 20–28 Moorgate 1998–2000, City of London. Mus London Archaeol Service Monogr 25, London Sena Chiesa, G, 1966 Gemme del Museo Nazionale di Aquileia. Associazione nazionale per Aquileia, Aquileia Shaffrey, R, 2007 The puddingstone rotary querns from Springhead Roman town, Kent. Lucerna 33, 6–10 Sidell, J, Wilkinson, K, Scaife, R and Cameron, N, 2000 The Holocene Evolution of the London Thames: archaeological excavations, 1991–1998. Mus London Archaeol Service Monogr 5, London Simpson, G, 1987 The decorated and plain samian pottery, in Meates 1987, 153–64 Simpson, J Y, 1854 On some ancient Greek Medical Vases for containing Lykion; &c., and on the modern use of the same drug in India. Proc Soc Antiq Scotland 1 (1851–54), 47–52 Simpson, C J, 1976 Belt-buckles and strap-ends of the later Roman Empire: a preliminary survey of several new groups. Britannia 7, 192–223 Simpson, G, 2000 Roman Weapons, Tools, Bronze Equipment and Brooches from Neuss – Novaesium excavations 1955–1972. Brit Archaeol Rep S862, Oxford Simpson, G and Rogers, G 1969 Cinnamus de Lezoux et quelques potiers contemporains. Gallia 27, 3–14 418 Settling the Ebbsfleet Valley Sjöqvist, E, 1960 Morgantina: Hellenistic medicine bottles. Amer J Archaeol 64, 78–83 Smith, A, 2001 The Differential Use of Constructed Sacred Space in Southern Britain from the Late Iron Age to the 4th century AD. Brit Archaeol Rep 318, Oxford Smith, K, 2006 Guides, Guards and Gifts to the Gods: domesticated dogs in the art and archaeology of Iron Age and Roman Britain. Brit Archaeol Rep 422, Oxford Smith, T P and Betts, I M, 2006 Building material, fired clay and wall plaster from Thurnham Roman Villa, Thurnham, Kent, in ADS 2006 Smith, V T C, 1997 The Roman road at Springhead Nurseries. Archaeol Cantiana 117, 51–67 Spencer, B, 1990 Pilgrim Souvenirs and Secular Badges. Salisbury Mus Medieval Cat 2, Salisbury Spotswood, R, 1793 Manuscript of a Naval Surgeon (1720–1800). Unpublished document Stacey, R, 2004 Evidence for the use of birch bark tar from Iron Age Britain. PAST 47, 1 Stanfield, J A, 1930 Further examples of Claudian ‘terra sigillata’ from London. Antiq J 10(2), 114–25 Stanfield, J A and Simpson, G, 1990 Les Potiers de la Gaule Centrale. Revue Archéologiques Sites, Horssérie 37, Gonfaron Stead, I M, 1967 A La Tène burial at Welwyn. Archaeologia 101, 1–62 Stead, I M, 1971 Yorkshire before the Romans: some recent discoveries, in Soldier and Civilian in Roman Yorkshire (ed R M Butler), 21–43. Leicester Univ Press, Leicester Stead, I M, 1986 The brooches, in Stead and Rigby 1986, 109–25 Stead, I M, 1991 Many more Iron Age shields from Britain. Antiq J 71, 1–35 Stead, I M and Rigby, V, 1986 Baldock: the excavation of a Roman and pre-Roman settlement, 1968–72. Britannia Monogr 7, London Stead, I M, and Rigby, V, 1989 Verulamium: the King Harry Lane site. English Heritage Archaeol Rep 12, London Steuer, H, 1990 Spätrömische und byzantinische Gewichte in Südwestdeutschland. Archäologische Nachrichten aus Baden 43, 43–59 Steuer, H, 1997 Waagen und Gewichte aus dem mittelalterlichen Schleswig. Funde des 11. bis 13. Jahrhunderts aus Europa als Quellen zur Handels- und Währungsgeschichte. Zeitschrift für Archäologie des Mittelalters Beiheft 10, Köln Swan, V, 1984 The Pottery Kilns of Roman Britain. RCHME Supple Ser 5, HMSO, London Swift, E, 2000 Regionality in Dress Accessories in the Late Roman West. Monographies Instrumentum 11, Montagnac Symonds, R P, 1992 Rhenish Wares: fine dark coloured pottery from Gaul and Germany. Oxford Univ Comm Archaeol Monogr 23, Oxford Symonds, R P and Wade, S M, 1986 A large Central Gaulish glazed vessel with applied-moulded decoration from Colchester. J Roman Pottery Stud 1, 55–7 Symonds, R P and Wade, S M, 1999 Roman Pottery from Excavations in Colchester, 1971–86, Colchester Archaeol Rep 10, Colchester Taylor, A 1985 Prehistoric, Roman, Saxon and medieval artefacts from the Southern Fen Edge, Cambridgeshire. Proc Cambridge Antiq Soc 74, 1–52 Taylor, M, 2007 The samian, 242–4 in Highstead near Chislet, Kent, Excavations 1975–1977 (P Bennett, P Couldrey and N Macpherson-Grant). Archaeol Canterbury NS 4, 242–4 Taylor-Wilson, R, 2002 Excavations at Hunt’s House, Guy’s Hospital, London Borough of Southwark. Pre-Construct Archaeol Monogr 1, London Terrise, J-P, 1968 Les Céramiques Sigillées Gallo-Romaines des Martres-de-Veyre (Puy-de-Dôme). Gallia Supple 19, Paris Tester, P J, 1961 The Roman villa in Cobham Park, near Rochester. Archaol Cantiana 76, 88–109 Tester, P J and Caiger, J E L 1954 Excavations on the site of a Romano-British settlement in Joyden’s Wood, near Bexley. Archaol Cantiana 68, 167–83 Thomas, C, 1981 Christianity in Roman Britain to AD 500. Batsford, London Thompson, I, 1982 Grog-tempered ‘Belgic’ Pottery of South-eastern England. Brit Archaeol Rep 108, Oxford Throckmorton, P, 1972 Romans on the sea, in A History of Seafaring Based on Underwater Archaeology (ed G F Bass), 65–86. Thames and Hudson, London Timby, J R, 1998 Excavations at Kingscote and Wycomb, Gloucestershire. Cotswold Archaeol Trust, Cirencester Timby, J, 2007 What happens in the temple precinct? Recent work at Higham Ferrers, Northants. Study Grp Roman Pott Newslet 44, October 2007 Tomber, R and Dore, J, 1998 The National Roman Fabric Reference Collection: a handbook. Mus London Archaeol Service Monogr 2, London Tomlin, R, 1979 Graffiti on Roman bricks and tiles found in Britain, in Roman Brick and Tile (ed A McWhirr), 231–51. Brit Archaeol Rep S68, Oxford Tomlin, R S O, 1997 II. Inscriptions, 455–72 in Roman Britain in 1996 (B C Burnham, L J F Keppie, A S Esmonde Cleary and R S O Tomlin). Britannia 28, 355–472 Toynbee, J M C, 1962 Art in Roman Britain. Phaidon, London Tyers, P A, 1996 Roman pottery in Britain. Batsford, London Tyers, P and Marsh, G, 1979 The Roman pottery from Southwark, in J Bird et al 1979, 533–82 Bibliography Tylecote, R F, 1986 The Prehistory of Metallurgy in the British Isles. Inst Metals, London Ulbert, G, 1977 Die römischen Funde von Bentumersiel. Probleme der Küstenforschung im südlichen Nordseegebiet 12, 33–66 Ulrich, R B, 2007 Roman Woodworking. Yale Univ Press, New Haven Urem-Kotsou, D, Stern, B, Heron, C and Kotsakis, K, 2002 Birch-bark tar at Neolithic Makriyalos, Greece. Antiquity 76, 962–7 Van der Werff, J H, 2003 The third and second lives of amphoras in Alphen aan den Rijn, The Netherlands, J Roman Pott Stud 10, 109–116 Vanvinckenroye, W, 1984 De romeinse zuidwestbegraafplaats van Tongeren (Opgravingen 1972– 1981). Publikaties van het Provinciaal GalloRomeins Museum te Tongeren 29, Tongeren Vaughan, S, 2001 Objects of bone and related materials, in Anderson et al, 322–32 Vernhet, A, 1976 Creation Flavienne de six Services de Vaisselle a la Grafesenque, Figlina 1, Documents du Laboratoire de Cérmologie de Lyon; Publications de la SFECAG, 13–27 Völling, T, 1994 Fibelformen der vorrömischen Eisen- und der römischen Kaiserzeit. Bericht der Römisch-Germanischen Kommission 75, 147–282 Völling, T., 1998 Die Fibeln Almgren Fig 2, 18, 19 und 22, in 100 Jahre Fibelformen nach Oscar Almgren. Internationale Arbeitstagung 25.–28. Mai 1997, Kleinmachnow, Brandenburg (ed J Kurnow), 39–51. Forschungen zur Archäologie im Land Brandenburg 5, Wünsdorf Wacher, J 1974 The towns of Roman Britain. B T Batsford, London Walker, D R, 1988 The Roman coins, in Cunliffe (ed) 1988, 281–358 Walton Rogers, P, 2007 Cloth and Clothing in Early Anglo-Saxon England, AD 450–700, Brit Archaeol Rep 145, York Wamers, E, 1994 Die frühmittelalterlichen Lesefunde aus der Löhrstraße (Baustelle Hilton 2) in Mainz. Mainzer Archäologische Schriften 1, Mainz Wamser, L, with Flügel, C and Ziegaus, B (eds), 2000 Die Römer zwischen Alpen und Nordmeer. Philipp von Zabern, Mainz Ward-Perkins, J B, 1939 Excavations on Oldbury Hill, Ightham. Archaeol Cantiana 51, 137–81 Ward-Perkins, J and Claridge, A, 1976 Pompeii AD 79. Exhibition Cat, Roy Academy Arts, London Waterer, J W, 1967, A leather panel with pierced decoration, in A ship of the Roman period: from Blackfriars, in the City of London (P Marsden), 56–7. Guildhall Museum, London Watson, J, 1997 The Reconstruction of a Roman Jewellery box from Mansell Street, London. Anc Monum Lab Rep 88/1997, Portsmouth Watson, W, 1949 Belgic bronzes and pottery found at Felmsersham-on-Ouse, Bedfordshire. Antiq J 29, 37–61 419 Watts, L and Leach, P, 1996. Henley Wood, Temples and Cemetery: excavations 1962–69 by the late Ernest Greenfield and others. Counc Brit Archaeol Res Rep 99, York Webster, G, 1958, The Roman military advance under Ostorius Scapula. Archaeol J 115, 49–98 Webster, G, 1986, What the Britons required from the gods as seen through the pairing of Roman and Celtic deities and the character of votive offerings, in Pagan Gods and Shrines of the Roman Empire (eds M Henig and A King), 57–64. Oxford Univ Comm Archaeol Monogr 8, Oxford Webster, G A, 1971 A hoard of Roman military equipment from Fremington Hagg, in Soldier and Civilian in Roman Yorkshire (ed R M Butler), 107–25. Leicester Univ Press, Leicester Webster, P, 1974 The samian ware, in Excavations on a Romano-British Site, Wiggonholt, 1964 (K J Evans). Sussex Archaeol Collect 112, 97–151 Webster, P (ed), nd Catalogue of the Decorated Samian from the early excavations at Caerwent. www.cardiff. ac.uk/learn/archaeol/caerwent/MasterIndex Wedlake, W J, 1982 The Excavation of the Shrine of Apollo at Nettleton, Wiltshire, 1956–1971. Rep Res Comm Soc Antiq London 40, London Weerd, M de, 1988 Schepen voor Zwammerdam: bouwwijze en herkomst van enkele vaartuigtypen in West- en Middeneuropa uit de Romeinse tijd en de Middeleeuwen in archeologish perspectief. Univ Amsterdam, Amsterdam Wendrich, W, 1999 The World According to Basketry: an ethno-archaeological interpretation of basketry production in Egypt. CNWS Publications 83, Leiden Wessex Archaeology, 2006a Kingsmead Quarry, Horton, Berkshire, Parts of Extraction Phases 4–7, Postexcavation Assessment Report with Updated Project Design for Analysis and Publication. Unpubl client rep 54635.03 Wheeler, R E M, 1930 London in Roman Times, London Mus Cat 3, London Wheeler, R E M and Wheeler, T V, 1936 Verulamium: a Belgic and two Roman cities. Rep Res Comm Soc Antiq London 11, Oxford Whiting, W, Hawley, W and May, T, 1931 Report on the Excavation of the Roman Cemetery at Ospringe, Kent. Rep Res Comm Soc Antiq London 8, Oxford Wickenden, N P, 1992 The Temple and Other Sites in the North-eastern Sector of Caesaromagus. Counc Brit Archaeol Res Rep 75, York Wild, J P, 1970 Button-and-loop fasteners in the Roman provinces. Britannia 1, 137–55 Williams, D F, 1993 Amphorae, in Woodward et al 1993, 214–9 Williams, D F, 2004 Purbeck marble in Roman and medieval Britain, in Purbeck Papers (ed D A Hinton), 126–31. Univ Southampton Dept Archaeol Monogr 4, Southampton Williams, D and Carreras, C, 1995 North African amphorae in Roman Britain: a re-appraisal. Britannia 26, 231–52 420 Settling the Ebbsfleet Valley Williams, R J and Zeepvat, R J, 1994 Bancroft: a late Bronze Age/Iron Age settlement, Roman villa and temple-mausoleum. Vol 2: finds and environmental evidence. Buckinghamshire Archaeol Soc Monogr 7.2, Aylesbury Williams, S, 2005 Tubular ferrules, Lucerna 29, 9–11 Willis, S H, 1998 Samian pottery in Britain: Exploring its distribution and archaeological potential. Archaeol J 155, 82–133 Wilmott, T, 1982 Excavations at Queen Street, City of London, 1953 and 1960, and Roman timber lined wells in London. Trans London Middlesex Archaeol Soc 33, 1–78 Winbolt, S E 1925 Roman Folkestone. Methuen, London Woodfield, C, 2005 Rare tazze, paterae and a broad hint at a lararium from Lactodorum. J Roman Pott Stud 12, 209–12 Woodward, P J, 1987 The excavation of a Late Iron Are settlement and Romano-British industrial site at Ower, Dorset, in Romano-British Industries in Purbeck (N Sunter and P J Woodward), 44–124. Dorset Natur Hist Archaeol Soc Monogr 6, Dorchester Woodward, A and Leach, P, 1993 The Uley Shrines: excavation of a ritual complex on West Hill, Uley, Gloucestershire: 1977–9. Engl Herit Archaeol Rep 17, London Woodward, P J, Davies S M and Graham A H, Excavations at the Old Methodist Chapel and Greyhound Yard, Dorchester, 1982–1984. Dorset Natur Hist Archaeol Soc Monogr 12, Dorchester Worrell, S, 2006 Finds reported under the Portable Antiquities Scheme. Britannia 37, 429–66 Young, C J, 1977 The Roman Pottery Industry of the Oxford region. Brit Archaeol Rep 43, Oxford Zienkiewicz, J D, 1986 The Legionary Fortress Baths at Caerleon: II, the finds. Cadw, Cardiff