The Use and Reuse of Stone Circles

by Richard Bradley

  The finds from grave 2 are puzzling. They comprise a pair of spiral rings, very neatly made, but markedly larger in diameter and rod size than the typical spiral finger- and toe-rings of the Scottish Iron Age (Clarke 1971). They represent some form of fitting or fastening. Their regular, horizontal placement side by side indicates they have not fallen from a lost organic object. Organic traces on one ring show it was in contact with pelt and textile, though unfortunately the detailed arrangement is lost. They lay beside a cattle skull, and the beast’s hide may have been the source of the pelt traces. Likewise, the textile traces as they survive seem to be in casual contact rather than clearly functionally related. However, pelt (and perhaps textile) could be explained as remains of sandals or other footwear, with these rings functioning to attach them to a toe. Spiral rings are known from a few burials as toe-rings, both singly and in pairs. It is normally assumed these were ornamental, and this may be the case in some instances, but others can be linked to footwear. A female burial from Rath, Co. Meath had three toe-rings (a singleton and a pair). The pair were quite large (dimensions are not given in the report), with one on each foot; the better preserved was clearly fitted over two adjacent toes, and the excavator suggested they were attached to a sandal (Schweitzer 2005, 96–98). O’Brien (1999, 21, 57) has noted that the wearing of a single sandal has cross-cultural associations with kingship, though the presence of two rings here (and in a comparable burial from Mine Howe; Hunter in prep) indicates pairs of shoes in these Scottish cases.

  ORGANIC REMAINS ON A COPPER ALLOY RING

  Penelope Walton Rogers

  Graves in which organic materials are well-preserved, such as the Bronze Age burial at Whitehorse Hill, Devon, are extremely rare. In most instances, clothing and other organic materials decay rapidly after burial and their appearance has to be reconstructed from sparse mineral-preserved remains adhering to metalwork. In one of the Waulkmill inhumations, semi-mineralised remains of textile, animal pelt and a strand of woody material were preserved in association with the copper alloy spiral ring, sf.22. They had become detached from the object, but corrosion marks and imprints showed which side had faced the metal. The pelt and textile had both been in direct contact with the metal, although they did not appear together in any of the fragments: it is possible, therefore, that the ring lay between the two. The pelt lay with its hair side towards the metal, and the woody strand lay outside it, on its skin face.

  The textile was medium-coarse, with yarn, 1.0 mm in diameter, spun in opposite directions in warp and weft (indicated Z × S). The weave was unclear, but in general character the textile resembles the twills and tabbies sometimes found in Iron Age sites such as the cemeteries in Yorkshire (Crowfoot 1991) and in Roman Britain (Wild 1970). A small number of records from burials in the region between Dundee and Dunbar demonstrate that a similar range of textile-types had already been established in Scotland in the Pre-Roman Iron Age (summarised in Walton Rogers forthcoming). The Waulkmill textile had been made from fine wool, 10–32 microns diameter in the sample examined, and was non-pigmented (originally white). Sheep with fine white fleeces first appear in any numbers in Britain in the Roman period (Ryder 1983, 177–80), although in this instance the extraction of fine underwool from double-coated prehistoric stock cannot be discounted.

  The animal pelt was made up of fibres 11–48 microns diameter. Most were densely pigmented (originally brown) but included a few non-pigmented (originally white) fibres. There were no visible medullas (central air channels) and the scale pattern, visible only in patches, was irregular mosaic with rippled margins. The species could not be identified from this, although small fur-bearing mammals and sheep could be discounted (Appleyard 1978).

  The single woody strand could perhaps represent basketry (assuming that it is not just detritus from the grave fill). Basketry techniques were used in prehistoric and Roman times, not just for the kinds of rigid containers that we would recognise today, but also in flexible materials for mats, bags and capes. They are well-represented in prehistoric burials (Henshall 1950, 151–55; Ballin Smith 2014, 67–74), though less in evidence in the graves of Roman Britain.

  To summarise, the remains preserved with the ring indicate that the body is likely to have been buried in clothing and accompanied by other organic accessories.

  The dagger found in 1898

  Fraser Hunter

  Description: NMS X.EQ 290. Three non-joining iron fragments from a fine dagger, two from the tang and one from the blade (Figs 3.5, 3.23).

  Description: NMS X.EQ 291. Fragment of horn, presumably from the handle; 17 × 10 × 2.5 mm.

  The object must have been complete when buried (as there are traces of leather sheath on the blade) but was broken on recovery. The fine lentoid blade section, although badly disturbed by corrosion, indicates it was a thin parallel-sided dagger rather than a knife. The rectangular-sectioned tang (9 × 5 mm) had a complex horn handle, with at least two separate cylindrical segments separated by a thin iron washer. A transverse horn pommel 12 mm thick at the end was retained by an oval iron washer (11 × 8 mm) with the tang burred to hold it. Tang: min. length 61 mm; blade: surviving length 33 mm.

  The gaming counters

  Hilary Cool and Mark Hall

  This section considers the significance of the glass counters from the 1898 Waulkmill burial (Hilary Cool) before moving on to a wider consideration of both groups of material and their context in the history of gaming (Mark Hall) (Figs 3.5, 3.24).

  Figure 3.23. The dagger found in the late nineteenth century (Alan Braby/NMS).

  THE GLASS GAMING COUNTERS FROM THE 1898

  WAULKMILL BURIAL

  Hilary Cool

  Both types of glass gaming piece (mid-blue and polychrome) from the Waulkmill burial(s) are unusual (see Table 3.3). Glass gaming counters are common in the first to mid-second century (Cool et al. 1995, 1555–56), but these are monochrome, opaque and overwhelmingly made of white and black, or sometimes very dark blue glass. Polychrome examples are known with inlaid dots (notably at Lullingstone and Lankhills; Cool and Price 1987, 123–25, 139–42; Clarke 1979, 251–54), but this is a separate development belonging to the second half of the fourth century. Other colours are much rarer; when they occur they again tend to be opaque and often fall outside the normal date range of glass counters (e.g. from a burial at Ospringe, Kent, of late second century date; Whiting 1925, 93–95 and plate facing 96). Translucent-coloured monochrome counters occasionally occur but are always rare. At Castleford, for example, only two of the 43 plano-convex counters were translucent (both blue; Cool and Price 1998, 190). Translucent blue counters, when found in dated contexts, tend to be contemporary with the black and white ones, i.e. first to second century (Cool and Price 1998, table 35). The translucent mid-blue of the Waulkmill examples is rather unusual, and they seem to fall outside the usual Roman provincial pattern.

  Figure 3.24. The gaming counters from 1898 (NMS).

  Polychrome counters other than the late fourth century ones are exceedingly rare. There is one from a pre-Flavian pit at Usk which might be from a recycled piece of vessel glass (Price 1995, 129, 134, no. 84). It had dark green ground with opaque red and yellow elements, and a pillar-moulded bowl was also found on the site with the same colours. The same colour combination is known from Cirencester (McWhirr 1986, 240, no. 15), where it is just described as millefiori glass; whether it was recycled vessel is not stated. Given this colour combination is a regular one in the early cast tradition, recycling seems likely. Strangely the option of using broken bits of recycled polychrome glass to make small objects does not seem to have been adopted that often in Britain, as beads made in this way are equally uncommon (see Wilmott et al. 2009, 351–52, 354, fig. 403, no. 5 for discussion).

  This makes the Waulkmill polychrome counters even more unusual than the blue ones from the same site. The elements preserved would suggest they could have come from a polychrome vessel which included canes with opaque yellow centres (or spirals) in a gr
ound that included both deep translucent blue and translucent peacock (a green/blue turquoise shade). Most unusually there are also elements of red glass ranging from mid-red opaque shade to a darker translucent one that fringes the yellow in places. It is very difficult to make a red glass that remains translucent, and in the ancient world where red glass occurs it is normally in the opaque red form. Interestingly the fact that this glass is translucent in places is best seen on the broken edge, which probably suggests that when complete the red would have appeared opaque. This colour combination is unusual in the early Roman cast glass tradition, so it is unlikely the raw material to make the counters came from a broken vessel of that date.

  There is an outside possibility that the glass came from one of the later cast millefiori vessels (Cool and Price 2008a; 238, fig. 10.2, no. 755; 2008b, D10-8), though these are fairly unusual in Roman Britain (and indeed elsewhere in the western empire). This would be broadly contemporary with the stem fragment in the cemetery (most likely of late second to early third century date of manufacture). Given the general rarity of the late cast polychrome vessels there is a slight problem over how one would have travelled to Scotland, possibly suggesting that the recycling could have occurred south of the frontier or indeed elsewhere within the empire. Certainly a similar combination of colours (though normally opaque) is seen in the cast chequerboard inlays and vessels that appear to belong to this late tradition (e.g. Gudenrath and Tatton-Brown 2003, 27, fig. 4). Price (2000, 104, no. 1), discussing a fragment of chequerboard glass from Frocester, drew attention to the Waulkmill counter, possibly being slightly misled by the drawing in Curle’s publication (1932, 390, fig. 68). From the photographs of the pieces I have seen, the different elements do not form regular lozenges as suggested in that paper.

  The Waulkmill polychrome glass counters are highly unusual and thus difficult to date. A possible set of comparanda comes from the Rhineland where in a later third century, rich, female grave in Bonn, Josefstraße there were 15 bichrome counters amongst the total of 27 glass counters found (Haberey 1961, 328–30, abb. 7.15, taf. 63.2). Nine were described as an opaque azure blue, three opaque black, three deep purple with opaque white veining (marbling), and twelve in a simple millefiori technique in green with opaque yellow tubes. Haberey related them to the early millefiori vessel industries though noting a lack of marbled or millefiori early Roman counters on the Rhineland. At the time he was writing, it was not known that this millefiori tradition continued in the East into the late Roman period. The examples of polychrome counters known to him were all from late graves, though he gave no details, and it should be noted that elsewhere in the discussion he is clearly referring to the late fourth century Lullingstone type (Haberey 1961, 329). He dated the grave to the period c.AD 250–275 and it is clearly after AD 251–253 as it contained a coin of Volusianus. The jewellery certainly fits a third century milieu. Some of the glass vessels are types that are coming into use during the second half of the century, so a deposition date in the final quarter of the century would also be possible.

  Table 3.3. Catalogue of the Waulkmill gaming counters from 1898.

  THE WAULKMILL GAMING COUNTERS:

  DISCOVERY AND GAMING CONTEXT

  Mark Hall

  The discovery of a further group of gaming counters from Waulkmill is an exciting affirmation of the site’s importance in demonstrating the entanglement of indigenous and Roman practices–indigenous in the burial practice, location and ethnicity of the incumbents (which Rome might have labelled ‘Pictish’), and Roman in the inclusion of gaming pieces in the burial rite.

  As reported above, the salvage recovery of the first group of counters in 1898 gave us a set of six glass counters and six similarly-sized quartzite discs (natural pebbles) (Coles 1905, 215–17). Twelve pebbles of brown, grey and whitish quartzite, only one of which resembled the earlier quartzite discs, may well represent a separate find (Callander 1915). As presently constituted in NMS and Aberdeen University collections (see Table 3.3), the material comprises two blue glass counters, two fragments of blue glass counters, two fragmentary polychrome glass counters, six brownish quartzite-pebble counters (two notably darker than the others), and a thick, nearly colourless cylindrical piece of glass which is a beaded stem fragment of a beaker or flask (Ingemark 2014, 51–54). Ingemark (2014, 52–53) raises the possibility of its reuse as a gaming-piece but favours the interpretation as a pars-pro-toto gift, arguing that it ‘would differ quite significantly from all the other pieces in its crudeness and awkward shape’. However it does not handle awkwardly at all, and its difference of form does not prevent it being used in a game that relied on different pieces such as ludus latrunculorum or successor games (see below). The extensive wear on one end of the stem fragment also argues for its use as a gaming piece. It is likely that this item was committed to the grave as a gaming piece.

  The 2012 group of Waulkmill stones cannot be so straightforwardly defined as a set of gaming counters as the 1898 group (Table 3.3; Fig. 3.25). Number 2.1 is a manufactured disc of mica schist comparable in size to the quartzite counters originally found at Waulkmill. The pebble, number 2.3, is a very close match to the original finds, and though not quite as symmetrical and slightly smaller, number 2.4 is also a reasonable match. However, the remaining quartzite pebbles (plus one sphere of conglomerate) are much more irregular and fit Callander’s description (1915, 206) of 11 of the additional 12 pebbles as ordinary ‘chuckie’ stones (the twelfth was an obvious brown quartzite counter) (Fig. 3.25). Callander suggested they were brought there by a natural process (i.e. water deposition) rather than human agency, but the new finds suggest deposition by human agency for both groups. The nature of the current group and its possible deposition in a bag suggests we may be dealing with a miscellaneous assortment of amulets and trinkets, a point we return to below. Almost all have differential use-wear on one surface, supporting use in gaming at some stage in their lives; variations in the degree of polish suggest they are a composite group rather than a single set. All the stone counters (1898 and 2012) are at home in the local geology (Simon Howard pers. comm.).

  Table 3.4 Catalogue of finds from the Waulkmill excavation 2012 (Group 2) (Fraser Hunter, with geological identifications by Simon Howard). All are unmodified natural pebbles apart from 2.1.

  No. Description Size (mm)

  2.1 Mid-grey disc, split on bedding planes to create a flat form, the convex edges abraded to shape. Mica schist with quartzite vein 28 .5 × 28 × 4.5

  2.2 ‘Egg’ amulet. Pyriform pebble with flattened oval section, giving it the form of a flattened elongated egg. Off-white with patterning from natural inclusions: dark linear streaks, irregular dark blobs and regular russet red blobs. Handling polish all over; dark staining on part of one face. Flatter than the usual ‘egg amulets’ (cf Hunter 1993, 331; Stevenson 1967), but belongs to the same category. Spotted quartzite; the red inclusions are garnet, the black an iron-magnesium mineral. 48 × 28 × 15.5

  2.3 Pale brown pebble, naturally circular with oval section. Use-polish on both faces and some staining on one. Fine-grained quartzite with slight iron staining and some mica. 27 × 25.5 × 11.5

  2.4 Pale yellow-brown pebble, sub-circular in plan, irregular pyriform in section with one flatter surface which shows preferential use-polish and was probably the base. Fine-grained quartzite with slight iron staining. 20.5 × 20 × 11.5

  2.5 Irregular squat pyriform pebble, slight handling polish all over, most pronounced on the flattest side (on which it naturally suits). Fine-grained quartzite with slight iron staining and some clay minerals. 23.5 × 21 × 20

  2.6 Irregular beige sub-oval pebble with attractive variegated blue-grey veining. Extensive use-polish, most pronounced on flattest face. Fine-grained quartzite with natural ‘pollutants’ causing some discolouration. 24.5 × 18.5 × 17.5

  2.7 Irregular mid-grey sub-spherical pebble with worn facets from cleavage planes and pitting (especially on domed surface) from decay of mica inclusions. Use-pol
ish all over, especially on flattest surface. Quartzite with sheared decayed mica. 19.5 × 16.5 × 15

  2.8 Pale yellow sub-spherical pebble, slightly elongated in one plane, slightly flattened on one face which bears more use-polish than the rest. Fine-grained quartzite with slight iron staining. 19.5 × 17 × 14

  2.9 Sub-spherical pale grey granular pebble, with use-polish most pronounced on the slightly flatter side. The naturally smooth erosion skin has been abraded at both ends, probably from post-depositional erosion (as suffered also by the other softer, granular stone 2.11). Quartzite, some mica. 20 × 19 × 16.5

  2.10 Off-white, very irregular pebble; slight use-wear especially on the flattest side. Fine-grained quartzite with slight iron staining. 18 × 16.5 × 12

  2.11 Near-spherical mid-brown pebble, surfaces worn (probably from erosion during burial). Hints of use-polish. Fine-grained angular conglomerate predominantly made from granite minerals. 22 × 20.5 × 19

  In considering the original discovery, local laird Duguid Milne recognised the blue glass counters were very similar to examples he had seen from Roman Sicily, in the museum at Palermo, and interpreted them as glass buttons (Coles 1905, 215). Coles’s (1905) analysis quotes Dr Joseph Anderson as suggesting the pieces were gaming-pieces (one he made three years earlier in connection with the finds from Camelon noted below). Since the turn of the twentieth century such pieces have become widely interpreted as counters for gaming or reckoning. They are found in varied colours, materials and sizes from across the Roman Empire and beyond its frontiers. Recent analysis suggests that such games were spread through the Germanic and Celtic lands of Europe via contact with Rome (Hall and Forsyth 2011). The two key Roman games appear to have been ludus latrunculorum and XII scripta/alea (later tabula), but these are unlikely to have remained unaltered throughout their histories. It is likely that they were improvised upon to create new socially/culturally embedded or relevant games, including what became the tafl group, especially hnefatafl and its Irish and Brittonic equivalents fidcheall and gwyddbwyll (Hall and Forsyth 2011, 1326–27). The game(s) may have arrived in Tarland as Roman gifts to a powerful household or they may have arrived second hand as part of an already-customised indigenous/Pictish game.