COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS FROM THE TWELFTH AND THIRTEENTH CENTURIES Robert Webley A particular group of medieval copper-alloy mounts have long been identiied and discussed in small inds studies, particularly in England. Known here as ‘binding strips’, such mounts have been seen as archetypal of the twelfth and thirteenth centuries and synonymous with high-status sites. Recent publications discussing such mounts, both in England and on the Continent, have prompted an up-to-date appraisal of these objects. These international publications present different interpretations of the functions of such mounts which seem to be reconcilable only by the proposition that they had multiple uses. However, the distinctive form of the ‘octopus’ mounts noted herein, and of other related mounts, appears to represent speciic shield decoration related to the early heraldic charge of the carboncle or escarbuncle, the name of which is derived from the Old French for boss,1 and which appears, notably, on the shield on the mid twelfth-century enamelled funeral plaque of Geoffrey Plantagenet, Count of Anjou. Introduction and deinition The mounts discussed here (see Plate 1a and b) were picked out over thirty-ive years ago as having basic shared characteristics in a chapter by Alison Goodall on medieval bronzesmithing in a collaborative volume about medieval industry.2 They are generally shallow D-shaped in cross-section, having a lat reverse, and a curved front. They are strip-like, that is long and thin; and often they bifurcate, sometimes rejoining to form openwork, geometric shapes, sometimes branching again. At intervals there are perforations for attachment by rivets. The strip itself is often decorated with transverse or angled grooves (gadrooning), or strips can feature regular perforations within successive oval lobes (see Plate 1c). Because they are thin, and often gilded, they are assumed to have had a decorative purpose. Embellishments can vary, and include hollow bosses of various forms set along the length of a strip, and elaborate G. J. Brault, Early Blazon. Heraldic terminology in the twelfth and thirteenth centuries with special reference to Arthurian literature (2nd edn., Woodbridge 1997), pp. 139-40, 188 with igs. 248-50. More elaborate decoration, perhaps with the precious stone or stones which comprise the other meaning of the word, is portrayed in other depictions of the escarbuncle in the Rolls of Arms; see the brief commentary by H. Stanford London in Rolls of Arms: Henry III (Aspilogia 2: London 1967), pp.167-8, no. 6. 2 A. R. Goodall, ‘The medieval bronzesmith and his products’, in D. W. Crossley (ed.), Medieval Industry (CBA Research Report 40: London 1981), pp. 63-71 at 70. 1 The Coat of Arms 3rd ser. 13 (2017), no. 233/234, pp. 1-12. 1 THE COAT OF ARMS terminals, also of various forms. Due to the variety of decoration within the range of such strips there can be some subjectivity in deciding whether or not a given mount belongs to this group. On the periphery of this group of ‘binding strips’ are so-called ‘octopus’ mounts (see Plate 1d). Fragments of such mounts were often initially discussed undistinguished from binding strips; their arms share the key characteristic D-shaped crosssection, and are sometimes gilded and decorated with transverse grooves. However, while binding strips have been assumed to have been mounted onto lat surfaces,3 and when found bent are assumed to have been damaged in the process of removal, fragments of ‘octopus’ mounts are invariably curved in proile. They also often feature an expanded terminal, perforated for attachment, set at an angle to each arm. Their function will be elaborated below, but their form alone suggests that they were attached to a lat surface and sat over a dome. Historiography Mounts that fall into this group have a long historiography, particularly in England. A very early example is Hume’s 1863 discussion of such mounts from Meols on the Wirral;4 by 1913 the pieces from Rayleigh Castle, Essex, were being cross-referenced to similar inds from the excavation at Castle Hill, Folkestone, Kent, by General Pitt-Rivers.5 The irst synthesis of binding strips came within the site report of Ascot Doilly Castle, Oxfordshire, when in 1959 Jope and Threlfall listed ifteen sites that had yielded binding strips, including the Husterknupp, near Cologne.6 It was Jope and Threlfall who identiied the recurrence of such strips on high status sites of the twelfth and thirteenth centuries, castles and manors—an association reiterated in Goodall’s synthetic chapter of 1981.7 Though many sites have been published since these key works of 1959 and 1981, adding more examples to the corpus, their authors tended to refer back to these writings rather than offer new discussion. In the last few years, however, there has been a re-engagement with binding strips, starting with the wide ranging overview provided by John Clark in his discussion of the binding strips from South Mimms Castle, Hertfordshire.8 Furthermore, they featured in recent surveys of copper-alloy objects ascribed to the Anglo-Norman period in England (c. 1066-1200), both by Steven Ashley and by Oliver Creighton and Duncan Wright, as being typical speciically of 3 A. R. Goodall, ‘Medieval copper alloy’, in G. Beresford, Goltho: the development of an early medieval manor c.850-1150 (London 1987), pp. 173-6 at 176. 4 A. Hume, Ancient Meols: some account of the antiquities found near Dove Point on the sea coast of Cheshire (London 1863), pp. 192-7 with plate 20. 5 E. B. Francis, ‘Rayleigh Castle: new facts in its history and recent explorations on its site’, Transactions of the Essex Archaeological Society 12 (1913), pp. 147-85 at 165. 6 E. M. Jope and R. I. Threlfall, ‘The twelfth-century castle at Ascot Doilly, Oxfordshire: its history and excavation’, Ant. J., 39 (1959), p. 219-73 at 267-8. 7 See note 2 above. 8 J. Clark, ‘Decorative metalwork’, in J. Kent, D. Renn and A. Streeten, Excavations at South Mimms Castle, Hertfordshire 1960-91 (London & Middlesex Archaeological Society Special Paper 16: London 2013), pp. 64-7 at 64-6. 2 COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS Figure 1: Distribution map of ‘binding strips’ and ‘octopus’ mounts. Selected key sites are listed in the order in which they appear in thetext: 1 Meols; 2 Ascot Doilly Castle; 3 Der Husterknupp; 4 South Mimms Castle; 5 Castle Acre Castle; 6 Goltho Manor; 7 Loughor Castle; 8 Waterford; 9 Vintry, London; 10 Château de Nogent-sur-Seine. Map © R. Webley. the twelfth century.9 They were also alluded to by Lewis in a recent discussion of medieval non-ferrous metalwork discovered in rural England, though the binding strips under discussion here are neither fully isolated nor analysed.10 Finally, there is the notable contribution of Vincent Legros, who gathered together numerous examples found in France and compared them with selected mounts from English sites.11 This last, however, raises questions, the French historiography seeing such mounts as harness ittings,12 at odds with a dominant association with boxes or caskets suggested in England and Germany.13 S. Ashley, ‘Anglo-Norman elite objects from castle and countryside’, in J. Davies, A. Riley, J.-M. Levesque, C. Lapiche (edd.), Castles in the Anglo-Norman World (Oxford 2016), pp. 281-98 at 282-3; O. H. Creighton and D. W. Wright, The Anarchy: war and status in 12thcentury landscapes of conlict (Liverpool 2016), p. 91. 10 M. J. Lewis, ‘Mounts for furnishings, padlocks and candleholders: understanding the urbanisation of medieval England through metal small inds recorded by the Portable Antiquities Scheme’, in B. Jervis, L. G. Broderick and I. Grau-Sologestoa (edd.), Objects, Environment, and Everyday Life in Medieval Europe (Turnhout 2016), pp. 157-85 at 160-5. 11 V. Legros, Archéologie de l’objet métallique aux époques médiévale et moderne en Picardie. Approaches typologique et fonctionnelle = Revue Archéologique de Picardie 2015 no. 1/2, pp. 97-9. 12 For example E. Louis, Recherches sur le château à motte de Hordain (Nord) = Archaeologia Duacensis 2 (1989), p. 85; B. Schnitzler, Vivre au Moyen Âge: 30 ans d’archéologie médiévale en Alsace (Strasbourg 1990), p. 423. 13 For example Ashley (note 9 above), p. 282; L. Clemens, ‘Die hochmittelalterliche Niederungsburg von Dockendorf (Kr. Bitburg-Prüm). Ein Vorbericht’, Funde und Ausgrabungen im Bezirk Trier 32 (2000), pp. 71-94 at 86. 9 3 THE COAT OF ARMS The reinvigoration of the discussion of binding strips, and the related ‘octopus’ mounts, prompts us to revisit these objects and consider again the orthodoxy regarding their dating and contextual associations. Furthermore, around four times the number of indspots of such strips have been traced by the author than were known in the early 1990s: for this study over 450 mounts have been documented from over 110 sites (see Figure 1).14 The remainder of this article will discuss the form, distribution, dating and function of binding strips, thus offering an up-to-date appraisal of these objects. Form As noted, although the ‘binding strips’ here deined have certain basic characteristics, the strips are found in a multitude of forms (see Figure 2). Even with the largest assemblages,15 where binding strips number in the tens and form a signiicant proportion of the copper-alloy inds, as at Castle Acre Castle, Norfolk, for example, it is hard to join fragments and thus reconstruct original forms, with very few exceptions—a notable one being pieces found at the Oude Huys, Helmond (Noord-Brabant).16 However, in reporting on large assemblages certain authors have been better able to comment on general groups within their given corpus. The set of around forty examples from Goltho, Lincolnshire (see Figure 2), was divided by Goodall into strips the main part of which had regular perforations within successive oval lobes (see Plate 1c), and strips whose main part had straight edges and were decorated with bosses along their lengths (see Plate 1a and b).17 Within the collection of around thirty inds from Loughor Castle, West Glamorgan, four groups were deined, the irst two approximating to Goodall’s groups from Goltho.18 The other two groups were plain straight strips with regular circular widenings (presumably rivet holes), and closed, geometric forms. The groupings established at Loughor may seem like a good basis for a typology of binding strip forms, but such an endeavour is argued here to be lawed. The divi14 J. M. Lewis, ‘Excavations at Loughor Castle, West Glamorgan, 1969-73’, Archaeologia Cambrensis 142 (1993), pp. 99-181, noted (at 144) two dozen sites that had yielded binding strips. 15 For example those from Castle Acre Castle, Norfolk (see A. R. Goodall, ‘Objects in copper alloy’, in J. G. Coad and A. D. F. Streeten, ‘Excavations at Castle Acre, Norfolk, 1972-77: Country house and castle of the Norman earls of Surrey’, Archaeological Journal 139 (1982), pp. 235-40 at 235-8); Goltho Manor, Lincolnshire (Goodall, ‘Medieval copper alloy’ [note 3 above], pp. 173-6); South Mimms Castle, Hertfordshire (Clark, ‘Decorative metalwork’ [note 8 above], pp. 64-6); Loughor Castle, West Glamorgan (Lewis, ‘Excavations at Loughor Castle’ [previous note], pp. 142-146); Boves Castle, Picardy (V. Legros, ‘Le mobilier métallique de Boves (1996-2000)’, in P. Racinet (ed.), ‘Recherches puridisciplinaires sur un territoire picard de l’époque antique à la révolution française. Boves. Étude du mobilier, 1996-2006’, Revue Archéologique de Picardie 1-2 (2012), pp. 91-105 at 94, 96, 99). 16 N. Arts, ‘Oorlog in Helmond: Archaeologische wapenvondsten uit het Oude Huys (circa 1175-1400)’, Brabants Heem 47.3 (1995), pp. 85-91 at 87 with illn. 4. 17 Goodall, ‘Medieval copper alloy’ (note 3 above), pp. 173-6. 18 J. M. Lewis, loc. cit. (note 14 above). 4 COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS Figure 2: ‘Binding strips’ and ‘octopus’ mount fragments from Goltho, Lincolnshire. After A. R. Goodall, ‘Medieval copper alloy’, in G. Beresford, Goltho: the development of an early medieval manor c.850-1150 (London 1987), pp. 174-5, ig. 154-5. Shown here half-size. sions are not clear cut, and are exposed by larger fragments which show the occurrence of elements of different ‘groups’ on the same mount: two examples will sufice. Firstly, there are binding strips from a number of sites with successive perforated oval lobes which continue as plain straight-sided strip.19 Secondly, there are strips which have the same terminal form but different main elements, such as a foliate terminal on a strip of the third Loughor type found at South Mimms,20 and the same terminal on 19 For example from Castle Acre Castle (Goodall, ‘Objects in copper alloy’ [note 15 above], p. 237, ig. 43.1); from Goltho Manor (Goodall, ‘Medieval copper alloy’ [note 3 above], p. 174, ig. 154.21); from Hordain Castle (Legros, Archéologie de l’objet métallique [note 11 above], p. 99, ig. 71A). 20 Clark, ‘Decorative metalwork’ (note 8 above), p. 65, ig. 49. 5 THE COAT OF ARMS a strip of the irst Loughor type found in the parish of Ryther cum Ossendyke, North Yorkshire (see Plate 1a).21 Furthermore, the different ‘types’ have been found on all site types across a wide geographical distribution. Where they occur at the same site, for example, Castle Acre, South Mimms or Goltho, strips of different ‘types’ occur in the same chronological horizons. This being the case, a meaningful typology of binding strips cannot be established on current evidence, though it might remain an aspiration for the future. For now, we can content ourselves with the fact that within the variety of different arrangements of binding strips various features recur in disparate locations. The foliate terminal described above (see Plate 1a), for example, as well as occurring at South Mimms and as a stray ind in North Yorkshire, is paralleled at Waterford, Ireland, Lesnes Abbey, Greater London, and Carisbrooke Castle on the Isle of Wight.22 Equally widespread are strips with bosses that sit in openwork, wheel-like roundels, being known from inter alia ‘Le Village’, Bezannes, Marne (see Figure 4, p. 9 below) and Hordain Castle, Nord, in France, and Goltho and the Vintry, London, in England (see Plate 1c).23 Certain binding strips, however, are rather out of the ordinary: two examples can be noted here. Unusual for England is a mount found in Winchester, Hampshire, which depicts a moulded lion passant on an expanded oval knop.24 Particular to south west France are strips with circular expansions which depict designs imitating coins struck by a variety of issuers in the mid to late twelfth century.25 Distribution The preceding discussion has already made reference to binding strips and ‘octopus’ mounts having been found at various locations across Europe: modern-day England, Ireland, France and Germany. This distribution across a large area of northwest Europe is a key aspect that has become apparent in compiling an up-to-date dataset PAS NLM-E21147. For Waterford see O. M. B. Scully, ‘Metal artefacts’, in M. Hurley, O. M. B. Scully, R. M. Cleary and S. W. J. McCutcheon, Late Viking Age and medieval Waterford: excavations 19861992 (Waterford 1997), pp. 428-89 at 480, ig. 15:17.4; for Lesnes Abbey see A. Clapham, ‘Explorations at Lesnes Abbey: report presented 1910’, Woolwich District Antiquarian Society Annual Report 15 (1910), pp. 83-173 at 160, ig. 39.1; for Carisbrooke Castle see A. P. Fitzpatrick, ‘Objects of copper alloy’, in C. J. Young, Excavations at Carisbrooke Castle, Isle of Wight, 1921-1996 (Wessex Archaeology Report 18: Salisbury 2000, pp. 134-9 at 136, ig. 48.10. 23 For Bezannes see D. Gucker, Bezannes, Le Village. Fouilles archéologiques préventives de l’Institut National de Recherches Archéologiques Préventives (forthcoming), inv. 0I 1606; for Hordain Castle see Legros, Archéologie de l’objet métallique (note 11 above), p. 99, ig. 71A; for Goltho Manor see Goodall, ‘Medieval copper alloy’ (note 3 above), p. 174, ig. 154.18. The inds from the Vintry are unpublished; see Museum of London, inv. VRY89[V912](3039). 24 D. Hinton, ‘Fittings from reliquaries and other ine caskets, i: gold, silver, lead or pewter, and copper-alloy ittings’, in M. Biddle (ed.), Object and Economy in Medieval Winchester (Winchester Studies 7 ii: Oxford 1990), pp. 762-80 at 773, ig. 220.2353; see also Ashley (note 9 above), p. 282. 25 R. Chareyron, ‘Rivets monétiformes’, Revue Numismatique 165 (2009), pp. 203-20 at 20910; see also Artefacts IND-6006, IND-6007, IND-6008. 21 22 6 COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS Figure 3: Distribution map of ‘binding strips’ and ‘octopus’ mounts. Stars represent castle sites; black dots represent urban sites; grey rings represent issuing mints on which coin designs are based; and grey dots represent all other sites. Map © R. Webley. (see Figure 3). This has previously been under-recognised, particularly in the English scholarly literature, despite early references to examples from the Husterknupp, Germany.26 The Continental literature has been far better at referring to English examples,27 though the current dataset demonstrates the full extent of the European distribution of these mounts, which includes Belgium, the Netherlands and Switzerland, in addition to the countries already mentioned. This distribution crosses political boundaries and is perhaps best interpreted in socio-economic terms, as will be discussed below. Jope and Threlfall (note 6 above), p. 267; S. Moorhouse, ‘Excavations at Burton-In-Lonsdale: a reconsideration’, Yorkshire Archaeological Journal 43 (1971), pp. 85-98 at 96; Lewis, ‘Excavations at Loughor Castle’ (note 14 above), p. 144. 27 For example J.-M. Lassure, La civilisation matérielle de la Gascogne aux XIIe et XIIIe siècles : Le mobilier du site archéologique de Corné à l’Isle-Bouzon (Gers) (Toulouse 1998), p. 532; Clemens (note 13 above), p. 86. 26 7 THE COAT OF ARMS The association of gilt ‘binding strips’, and indeed ‘octopus’ mounts, with castles and manors has a long and persistent pedigree in the English historiography, present since Jope and Threlfall’s 1959 note. If this was a product of a historic emphasis on castle excavations, then one would expect an adjustment in recent years. Although urban inds of binding strips have featured increasingly,28 minimal numbers have been found elsewhere, for example in ecclesiastical contexts or on deserted medieval settlements.29 Noteworthy is the fact that inds of such mounts reported to the Portable Antiquities Scheme (PAS) in England and Wales are also minimal compared with the overall large quantity of medieval material. Indeed, only thirty-one records have been veriied for inclusion in the present study, of which four are deemed to be ‘octopus’ mounts. This may be a result in part of decay in the plough zone, or may be due to a lack of recognition on the part of inders, but they, along with urban inds, only provide a modest adjustment to the general association with high status sites, predominantly castles. Given that the primary consumers of these mounts appear to have been of elite status, it may have also been the case that the production of many such mounts took place within the domains of castle owners.30 Dating If the association of ‘binding strips’ with castles and manor sites was argued strongly by the work of Jope and Threlfall and others, then so too was the dating of such mounts to the twelfth and thirteenth centuries (and particularly the twelfth). This dating is signiicant in the context of discussions of a dearth of non-ferrous metalwork across Europe in general until the later twelfth century.31 If we leave aside the exceptional examples bearing coin designs, there is little about the mounts that is diagnostic of date on stylistic grounds, other perhaps than the perforations on the roundel from Bezannes echoing objects from the Romanesque period that use motifs inspired by contemporary architectural arcading (see Figure 4).32 Consequently, the dating of For example F. A. Pritchard, ‘The small inds’, in A. G. Vince (ed.), Aspects of Saxo-Norman London 2: Finds and Environmental Evidence (London & Middlesex Archaeological Society Special Paper 12: London 1991), pp. 120-278 at 151; J. Brenan, ‘Furnishings’, in G. Egan (ed.), The Medieval Household, Daily Living c.1150-c.1450 (London 1998), pp. 65-87 at 69; Clark (note 8 above), p. 66. 29 An example of the former is Lesnes Abbey: Clapham (note 22 above), p. 160, ig. 39.1). An example of the latter is Westbury, Buckinghamshire: R. Ivens, P. Busby and N. Shepherd, Tattenhoe and Westbury: two deserted medieval settlements in Milton Keynes (Buckinghamshire Archaeological Society Monographs 8: 1995), p. 348, ig. 151.46. 30 For a discussion of non-ferrous metalworking on elite sites see L. Bourgeois, ‘Production et distinction: l’artisan au château (Nord-Ouest de l’Europe, Xe-XIIe siècles)’, in D. Barthélemy and J.-M. Martin (edd.), Richesse et croissance au Moyen Âge. Orient et Occident (Paris 2014), pp. 151-82. 31 Bourgeois (see previous note), p. 154; Ashley (note 9 above), p. 281. 32 Ashley (note 9 above) , pp. 288, 290, 293-4, igs 18.7, 18.11, publishes a number of buckle plates and harness pendants with this decorative conceit and cites various comparanda. Although the globular headed rivets often found on ‘octopus’ mounts are found on a number of twelfth-century objects, such as reliquaries, they cannot be considered as irm dating markers. 28 8 COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS Figure 4: ‘Binding strip’ with openwork roundel from Bezannes, ‘Le Village’, Marne (inv. OI 1606). Scale 1:1. Photo © Laboratoire LandArc. the contexts of deposition for over 290 mounts from almost 80 sites has been charted using the ‘balanced average’ technique (see Figure 5a); admittedly, it is dificult to assess the discrepancy between use date and deposit date. This notwithstanding, it seems that the focus on the twelfth and thirteenth centuries has been justiied in the case of binding strips, though with a high point around the second quarter of the twelfth century. We can be fairly conident in the dating of a number of examples to this high point, with some English inds associated with destruction during the ‘Anarchy’ of the reign of Stephen (1135-1154),33 and a cache of mounts found at Nogentsur-Seine Castle, Aube, in a context of the same period.34 At the earliest end of the spectrum (see Figure 5a), it has been assumed that examples found in late AngloSaxon deposits at Portchester, Hampshire, and Wareham, Dorset, were intrusive.35 At the other end of the date range, Hinton suggested continuity for binding strips into the mid-ifteenth century,36 while Clark noted an item from a context of such a date in Northampton.37 A charting of dates for binding strips from urban sites (see Figure 5b) shows a stronger ‘tail’ in the early fourteenth century suggesting perhaps greater residuality in such contexts, or possibly longer curation of the object to which the mounts were attached. Certainly, binding strips found at various castles in a badly twisted condition suggest that their removal was part of a destructive event, and both that the objects to which they were attached did not survive in the long term, and that the strips were not themselves curated for reuse.38 Clark (note 8 above), p. 65, cites Wareham Castle, Dorset in this respect. Legros (note 11 above), p. 97. 35 Hinton (note 24 above), p. 766. 36 See previous note. 37 Clark (note 8 above), p. 65. 38 Lewis (note 14 above), p. 142; Clark (note 8 above), p. 66. 33 34 9 THE COAT OF ARMS Figure 5: top (a), balanced average chart for all mounts discussed; below (b), balanced average chart for mounts from urban sites. Both charts © R. Webley. Function As noted, there are discrepant views on the function of these ‘binding strips’ in the international literature. Many options have been suggested: book mounts, altar mounts, reliquary mounts, box/chest/casket ittings, cabinet ittings, harness ittings and shield mounts. The predominant English association is with boxes such as caskets due to the broad association of binding strips with fragments of wood and possible bone plates at Goltho;39 this attribution is followed in Germany too, and in relation to the southwestern French mounts with coin designs.40 However, as noted, more generally in France such mounts are seen as being from harnesses, with most recent support for this suggested by a group of such mounts found with equestrian equipment at Nogent-sur-Seine.41 In the absence of either unequivocal evidence or easy answers the strips can further be considered in and of themselves. The lat reverses of binding strips suggest that they would have decorated lat surfaces, and may have helped leather retain leather, or some other perishable mateGoodall (note 3 above), p. 176. For Germany see Clemens (note 13 above), p. 86; for the southwestern French mounts perhaps from boxes used to store money or trade equipment see Chareyron (note 24 above), p. 218. 41 Legros (note 11 above), p. 97. 39 40 10 COPPER-ALLOY ‘BINDING STRIPS’ AND SHIELD MOUNTS rial, on wood. One of the closest parallels for the strips under discussion can be seen on a reliquary shrine from Cologne where gilt copper-alloy mounts overlay a linen and silk clad box.42 Though very few examples of binding strips survive complete, many are not just long, over 360 mm in some cases, but also expansive where they bifurcate or branch (as much as 100 mm across). It is this latter quality, possessed by many mounts, that makes them seem unlikely as harness ittings. However, in the absence of irm evidence a pragmatic interpretation sees ‘binding strips’ as all being inter-related, but applied to a variety of objects that circulated at an elite level.43 Finally, and more consistently, ‘octopus’ mounts have been identiied as decorative mounts for shield bosses, across German, Dutch and French writings.44 Though identifying the function of such mounts has not been essayed in the English scholarly literature,45 the general reference by Lewis to the shield on the mid-twelfth-century enamelled portrait of Geoffrey Plantagenet at Le Mans is persuasive in terms of a probable depiction of such a mount on the boss (see Plate 2).46 This identiication is therefore reiterated here. It is likely that other binding strips, some of which have formal similarities to some octopus mounts, were also mounted on to such shields, especially the straight examples for which Lewis saw approximations on certain late twelfth-century capitals at Monreale Cathedral, near Palermo in Sicily,47 and which can be glimpsed on Geoffrey Plantagenet’s shield. Together, they could have produced radial designs such as that on a kite-shaped shield depicted on a ceramic slab found at Montech, Tarn-et-Garonne.48 It is noted that although the modal number of arms on octopus mounts is indeed eight, as became convention on depictions of the escarbuncle,49 in the loruit of such mounts the number varied—and when not eight there were generally fewer. Conclusion Both copper-alloy ‘binding strips’ and ‘octopus’ mounts have a role to play is discussions of material culture of the Central Middle Ages. Even if we cannot be certain about the function of each given example, we can here suggest the presence of many binding strips on shields, in association with ‘octopus’ mounts. We can, furthermore, be more conident about their dating, focused on the twelfth century, and their social milieu, namely high status sites such as castles. The wide distribution of such objects across parts of northwest Europe can now be appreciated far more fully. As 42 A. Legner, Ornamenta Ecclesiae. Kunst und Künstler der Romanik (3 vols., Cologne 1985), p. 343, cat. E107. 43 V. Legros, personal communication. 44 H. W. Böhme, Das Reich der Salier 1024-1125: Katalog zur Ausstellung des Landes Rheinland-Pfalz (Ausstellung im Historischen Museum der Pfalz, Speyer, vom 23. März bis 21. Juni 1992) (Sigmaringen 1992), p. 102; Lassure (note 27 above), p. 532; Clemens (note 13 above), p. 88. 45 See Goodall (note 3 above), p. 176; Ashley (note 9 above), p. 282. 46 Lewis (note 14 above), p. 144. 47 See previous note. 48 F. Sarret, ‘Une brique médiévale ornée’, Archéologie du Midi médiéval 1 (1983), pp. 145-6. 49 Brault (note 1 above), p. 140. 11 THE COAT OF ARMS such, they are a pan-European object type of the twelfth century, prior to the widespread and mass produced copper-alloy accessories of the thirteenth and fourteenth centuries. Like the sets of tabulae discussed by Creighton, they helped engender a process he termed the ‘Europeanisation of elite culture’, through their consumption in largely seigneurial settings.50 In terms of the development of early armory and blazon, ‘octopus’ mounts and some of the other fragmentary strips, seemingly forming the extended arms of the elaborate boss or escarbuncle, provide tantalising material evidence for an early armorial charge. The evidence suggests that it may have derived from non-ferrous decoration applied to the shield rather than previous suggestions attributing it to the shield’s structural ironwork.51 Indeed it may have been the development of heraldry that led to the demise of such applied decoration, which became increasingly redundant as heraldry became formalized in the course of the twelfth and thirteenth centuries.52 O. H. Creighton, Early European Castles, Aristocracy and Authority, AD 800-1200 (London 2012), p. 112. 51 A. C. Fox-Davies, A Complete Guide to Heraldry (London and Edinburgh 1909), pp. 90-1. 52 This article is based on an earlier version published in February 2017 as Cahiers LandArc 19: Analyse et interprétation des appliques en alliage cuivreux dites «binding strips» des XIIe-XIIIe siècles en europe de l’ouest. I would like to thank the following: John Clark and Dan Nesbitt (Museum of London) for providing archival access; Jean Soulat (Laboratoire LandArc), Luc Bourgeois (Université de Caen), Amélie Berthon (Evéha), Vincent Legros (Université de Picardie) and Ben Rijns (The Coinhunter Magazine) for providing references and discussing these objects; Steven Ashley for advice on the heraldic connotations of such mounts; and Laura Burnett (PAS) for reading drafts of this note and her support. 50 12 PLATE 1 ‘Binding strips’ and an ‘octopus’ mount. All shown 1:1. From top: (a) ‘binding strip’ from Ryther cum Ossendyke parish, North Yorkshire (PAS NLM-E21147). Photo © North Lincolnshire Museum, by courtesy of the PAS. (b) ‘binding strip’ from Bungay parish, Suffolk (PAS NMS-202011). Photo © Norfolk County Council, by courtesy of the PAS. (c) ‘binding strip’ with successive perforated oval lobes from Vintry, London (inv. 3039). Photo © Museum of London. (d) ‘octopus’ mount from Walcot near Folkingham parish, Lincolnshire (PAS LIN-B35B23). Photo by courtesy of the PAS. See pages 1-2. PLATE 2 Enamelled grave cover of Geoffrey Plantagenet, Count of Anjou, from Le Mans Cathedral, c. 1151. With (right) detail of boss mount on shield. See page 11. As illustrated in E. Taburet-Delahaye and B. Drake Boehm (edd.), L’œuvre de Limoges: émaux limousins du Moyen Âge (Paris 1995), p. 99.