A small pottery assemblage was recovered from only one of a series of evaluation trenches excavated to the south of the Devil’s Arrows, Boroughbridge. The sherds represent cooking pot and storage jar types of a domestic character dating to the Iron Age: late 1st Millenium BC. The fabrics, which are hard fired with much poorly sorted quartz sand temper, and the weak–profiled vessel profiles, compare well with the Iron Age assemblage from Roecliffe. Temper characteristics are indicative of local manufacture.

Between four and five vessels were represented by rims and bases, but no profiles were reconstructable. There was also a possible loom weight fragment. The fabrics could be divided into three types: Finer fabric: [114] Two jars/cooking pots. Coarse fabric: [128] [143] [145] large storage jar, hard brittle body, profuse coarse quartz sand tempering. Intermediate fabric: [135] smaller grain sized temper.

Condition: All unweathered showing sharp fractured edges except for on small rim [114] with slight rounded edges. T1. Surface. Small sherd (1.5cm) Weight 5gm. A possible rim ? pointed profile to lip. Hard slightly laminated; Dark grey. Temper fine quartz sand. Wall thickness 6mm. U/S T1. Two pieces, one a flat shale fragment Flake (6×3.5×0.9cm) 15gm. Irregular moulded external surface probably from a loom weight. Hard compact matrix; dark grey to brown–grey. Temper abundant fine quartz sand. [114] Three sherds (two rims) Weight 40 gm a. Rim and upper body profile (2 joining pieces, >7cm.), out–turned thickened rim, about 15mm. diameter, barrel profiled jar. Hard laminated fabric; dark grey exterior, well smoothed surface; orange–brown core and interior, irregular surface with elongated cresentic scars. Temper 10% range, mostly small sand – chips and sub–angular quartz grains, scarce rock fragments >7mm. (sandstone). Wall thickness 7mm. Patches of carbonised incrustation on body and under rim. b. Small rim (>2.5cm.) Slightly weathered. Thin out–curving profile, rounded lip. Hard compact laminated fabric; dark grey with some buff–brown toning. Horizontal smoothing inside the lip. Temper 10%, poorly sorted sand, sub–angular quartz grains, rare rock fragment >2mm. Traces of carbonised layer on exterior below rim and inside lip. [128] Three wall sherds (>6cm.) Weight 809m. a. Two sherds, fractured along ring build junction. Coarse fabric; hard, heavy, laminated; orange–brown exterior, dark grey core and interior. Temper 15% sand, poorly sorted sub–angular quartz grains. Wall thickness 9–10mm. Shallow vertical finger furrowing on exterior. Patches of carbon on the interior. b. One sherd. Coarse fabric; hard laminated; orange–brown, toned dark brown to dark grey. Temper 15% profuse fine sand and scattered sub–angular quartz grains >2mm. Wall thickness. Wear patch on exterior. [135] Surface Wall Sherd (>4cm.) Weight 15gm. Hard laminated fabric; dark grey, buff exterior. Temper 20% fine quartz sand. Wall thickness 12mm. [143] 23 sherds and flakes including a base (>8cm). Weight 420gm. Base angle, 17cm. dia. thickness 4cm>. and lower body sherds showing fracture at ring build junctions. Coarse fabric; hard, heavy, laminated, brittle fracture; Oxidised orange–brown exterior, grey–brown interior, greyish core. Temper 20%, much poorly sorted quartz sand, sub–rounded grains >1mm., scattered sub–angular grains. Sparse rock fragments >3mm. quartz and ferriginous sandstone. Wall thickness 12mm thickening towards base to 20mm. Probably all part of the same large storage jar. [145] Rim Sherd Weight 259m. Slightly inward sloping rim, rounded lip. Diameter uncertain Coarse fabric; hard laminated, brittle fracture; oxidised brown–orange exterior; irregular interior surface, grey–toned brownish. Temper 20% much sub–angular to sub rounded quartz >5mm., scarce sandstone fragments >4mm. Wall thickness 14mm.

The flint assemblage totals 9 struck pieces, with a total collective weight of 27.4g. This lithic assemblage is unfortunately to small and of to poor a contextual integrity to be statistically assessed, however a few salient points can be made:

The material is in a moderately fresh state with a third of the material having been subjected to breakage (contexts = 151, 301, 422). In two of these instances (contexts = 151 &422) the damage appears to be related to modern ploughing etc. Slight plough damage is also present on a retouched flake from a topsoil context (400). Slight traces of patination are present on only 2 of the flints, both of which are flakes (record numbers 3 &5).

With the possible exception of an irregular flake (record number 1) from context 128 non of the pieces shows any trace of attrition characteristic of utilisation.

A surprisingly high proportion of the assemblage comes from the later stages of lithic reduction; there being at least 5 pieces (55.5%) from tertiary stages of lithic reduction. Context 308 (record number 4) produced a rather small (8.9g in weight), corticated, 3 platformed (2 opposed) irregular micro–flake core of olive grey till flint. Such pieces are characteristic of some of the regions Later Neolithic and Early Bronze Age assemblages.

All but 1 of the pieces are from till / gravel derived flint sources. The significant exception to this being a slightly broken blade (record number 2) of greenstone / epidiorite (this specimen is technically an epidiorite) from context 151. The material is similar to that from stone axe petrology groups I–III and it is possible (although in this instance probably not!) that the flake has come from an axe. The origin of this particular raw material is thought to be Cornwall (Clough 1988, 7) though current work (T.G Manby, personal communication) may yet locate a nearer raw material source; possibly in the Langdale area of the central Lake District.

The incidence and composition of the assemblage is given in the table below:

The topsoil (context 400, Record 7) produced a small heavily plough damaged tertiary flake with a small area of nondescript, convex, semi–invasive retouch to its transverse distal end.

A small fragmentary retouched flake (record 9) was recovered from context 422. The piece has a slight resemblance to a chisel arrowhead but is probably just a nondescript retouched flake.

TABLE: The Assemblage Composition. Debitage. Total No Broken Weight in g Content Flakes &Chippings = 5 (1) 6.9 128, 301, 308 (×2), 416 Blades = 1 (1) 5.9 151 Cores = 1 NA 8.9 308 Retouched Implements. Edge Retouched Flakes = 1 NA 4.0 400 Miscellaneous Retouched Flakes = 1 (1) 1.7 422 Totals 9 (3) 27.4

( ) = Number of broken examples.

There are no clearly chron–diagnostic pieces in the small assemblage although the core and retouched flakes are most suggestive of a Later Neolithic / Early Bronze Age date. The greenstone / epidiorite blade is slightly more suggestive of a Later Early to Early Middle Neolithic date. It must be stressed that due to the nature of the assemblage the pieces may not all be of the same date.

The following 3 points are the major aspects of flint assemblage: The assemblage is in a moderately fresh state. Most of the struck assemblage comes from the later stages of lithic reduction. The presence of a greenstone / epidiorite flake may prove to be of future significance.

The current assemblage is of little significance other than caveat 7.3.

There are no recommendations.

The enclosed constitutes all the reportage and archive that is required for this assemblage.

The aim of the environmental sampling was to determine the potential for preservation of, in particular, charred plant remains (the site is not waterlogged) and to determine if deposits appropriate for later radiocarbon dating were present – should development proceed.

Bulk samples were taken from ditch and pit fills. these were manually floated in the laboratory with both flot and residue retained upon 500m mesh. After drying the residues were scanned for artefactual and ecofactual material. The flots were sorted under a stereomicroscope, at magnifications of up to x40, and notes made of the matrix components. Charred plant remains were sorted out, identified by comparison with modern reference material and numbers of each type counted. All data, both matrix and botanical, were numerically coded and input to the suite of programs PHYTOPAK (Huntley, Huntley and Birks, 1981) for manipulation in tabular form.

Six samples were processed and assessed, the details are presented in table 1. All produced at least some charred plant remains (table 2) which were generally in a good state of preservation. There is no clear evidence that they were moved around once deposited. Most of the material was cereal grain but chaff and a few weed seeds were recorded. Given the good preservation it seems reasonable to suggest that this lack of processing debris is real and that maybe cereals were being used nearby but not processed. Most of the identifiable grain was hulled barley. Oats were the next most abundant and may have been either the wild or cultivated species since the diagnostic floret bases were not recovered. Rye and wheat, including spelt and probable bread wheat, were present too. Spelt glumes and a spikelet fork reinforced the presence of spelt wheat.

Table 1: sample details Context Notes 111 AB Fill from within rectangular structure (148) which appears to have been burnt in situ. Substantial amounts of charcoal were associated with both the structural remains and the central fill. The excavator intends two radiocarbon assays from this feature. 19 litres of sandy soil produced a moderate flot of flaky charcoal and a very small amount of industrial waste. Reasonable numbers of charred cereal grains were recorded – principally hulled barley but rye and indeterminate wheat as well. About one third of the grains were not identifiable to species. Hazelnuts and a sloe stone, also charred, attest further dietary material probably. One spelt glume base was present. 114 AA Fill of a substantial pit (115); high charcoal content noted although no trace of in situ burning was recorded. Prehistoric pottery was recovered. 22 litres dark brown sandy silt processed. Considerable numbers of well preserved charred plant remains were present; these were fragmentary in many cases although otherwise well preserved with no sign of abrasion. The majority of the grains consisted of hulled barley but hexaploid wheat and oats were also common. One probable rye grain was recorded. A selection of weed seeds was present mostly indicating nutrient enriched and damp ground. Hawthorn and hazel nut shell suggest local fruits. 139 AA Primary fill of a substantial ditch identified through geophysical survey. Prehistoric pottery was recovered from the fill. Although 23 litres of brown/orange sandy loam produced only 50ml of flot. This was very rich in tiny fragments of calcined mammal bone and a few fragments of burnt bone. At least some of the large pieces were apparently human although identification needs to be checked. About 25% of the charcoal was twiggy or roundwood with rather less than usual of the flaky oak charcoal of other contexts. Waterlogged elderberry seeds (Sambucus nigra) scored 2 indicating frequent occurrence. The charred material was not abundant but spelt glume bases were the most common. Spelt spikelet fragment, oat awn fragment and a rye rachis node were all recovered with one indet. cereal grain and one barley grain. A fragment of a large legume provides limited evidence for other food plants. 329 AA Fill of post hole 330. 8 litres brown sandy loam produced a moderate sized flot of charcoal. The charcoal included both the flaky oak characteristic of many samples from this site and a small amount of twiggy material – some oak but other species present too, including a conifer. A tiny amount of industrial spatter was present as was the occasional fragment of mammal bone. A few cereal grains were recorded – mostly hulled barley but also oats and a possible wheat although it may have been rye. Oat awn fragments were present. One burnt blackberry pip was recorded as were several waterlogged blackberry seeds. 416 AA Fill of ditch 417. A number of fractured pebbles recovered from the fill may represent pot boilers and worked flint from the fill may provide a date once examined. 17 litres of brown sandy clay produced a very small flot of charcoal with industrial waste. This latter was metallic–coloured and honeycombed. Hulled barley and one each of oats and probable bread wheat grains were recovered. 430 AA Sample from pit with high proportion of burnt bone. 16 litres of dark brown loam produced a moderate sized flot of charcoal and the very occasional fragment of calcined bone. Only one spelt glume base was present.

These features clearly are in receipt of plant material with cereal grains being common and well preserved. The presence of rye and, perhaps cultivated, oats could suggest a date later than the prehistoric period – possibly post– Roman which would make the site of more importance regionally. The low levels of weed seeds are interesting since a more rural site, as this apparently is, would be expected to be processing its own grain. However, the area excavated is small and need not be typical of the archaeological site as a whole. Material appropriate for radiocarbon dating is present and it is recommended that AMS dates are sought on cereal grains wherever possible rather than upon the oak charcoal. The trial trenches have demonstrated the potential for further environmental work to be considered as essential should further intervention be undertaken. If the material does indeed date to the post Roman period it adds to the continuity of occupation of this part of North Yorkshire and could add considerably to our knowledge of crop husbandry at that time – otherwise of very limited extent only.

Table 2: Matrix and botanical data Bar Lane, Boroughbridge: BL98 Bio.code 2208 2207 2206 2205 2204 2209 Context number 139 430 114 416 111 329 Site code no. 47 47 47 47 47 47 Excavation 75 75 75 75 75 75 Volume floated (litres) 23 22 17 14 09 Volume flot remaining (ml) 50 200 300 10 100 75 Industrial waste 01 01 01 01 Calcined mammal bone 04 02 01 Charcoal fragments 02 05 05 04 04 05 Clinker/cinder 01 Coarse sand/gravel 03 01 Mammal bone 01 Coal 01 Charred material (counts) ca Chenopodium album 03 ca Fallopia convolvulus 01 cc Avena grain 09 01 02 cc Cerealia undiff. 01 03 01 10 04 cc cf. Secale cereale 01 cc Hordeum hulled 56 04 17 06 cc Hordeum indet. 01 cc Secale cereale grain 01 cc Secale rachis node (undiif=2115) 01 cc Triticum (hexaploid) 05 cc Triticum cf. aestivum 01 cc Triticum sp(p). grain 03 01 cc Triticum spelta 01 cg Gramineae >4mm 04 ch Sieglingia decumbens 01 cr Galium aparine 01 cr Rumex obtusifolius–type 01 cr Sonchus asper 01 cs Triticum spelta glume 06 01 01 cs Triticum spelta spikelet 01 ct Corylus avellana nut frag. 01 02 ct Crataegus monogyna 01 ct Prunus spinosa 01 ct Rubus fruticosus 01 cw Carex (lenticular) 01 01 cw Carex (trigonous) 01 cx Bromus sp(p). grain 01 cx Legume >4mm 01 01 wt Sambucus niger 02 wt Rubus fruticosus 01

Excavation adjacent to the Devil’s Arrows produced fragments of animal bone from shallow pits of prehistoric date.

For this assessment, the hand recovered bones in each bag were scanned and the following information recorded by context for cattle, sheep/goat and pig, if present: Fused Bone; Unfused Bone; Zones; Lower Jaw; Loose Teeth, Measurable Bone. These categories give an indication of the presence of ageable bones and confirm the impression gained of the general level of preservation. Fragments were usually only counted as identifiable if they encompassed a recognisable 'zone', as defined by Rackham (unpubl.). The presence of species other than cattle, sheep/goat and pig was noted. Cattle and sheep size ribs and vertebrae are referred to by the categories Large Ungulate and Small Ungulate respectively.

Finds of animal bones were sparse, being present in only five features. The state of preservation was poor. Only context 139, the largest group, produced two bones with reasonably intact surfaces. The enamel in the teeth is decaying, causing the enamel to flake off. The articular ends of long bones are abraded and therefore unmeasurable. Long bone shafts are disintegrating. Only two species are present, cattle and horse. The bones of these animals are large and robust and therefore take longer to decay than the bones of smaller animals, such as sheep. Horse is represented only by teeth, a single cheek tooth from context 139 and the remains of two adjacent incisors from context 502. The latter are still held together by mud, rather than bone, and indicate that part of a jaw was originally deposited. The cattle skeletal elements present are: scapula in context 102; two astragali, one humerus and an acetabulum from context 139; a calcaneum and metacarpal in context 352; a metapodial in context 502; a metapodial and first phalanx in context 510. All the epiphysial ends present are fused. Most of these elements are extremely robust and have a higher survival rate than other parts of the body. Cattle teeth were seen in all contexts except for context 510. These teeth were from the permanent adult dentition and all appeared to be in wear, indicating adult animals. Again teeth are robust and have enhanced survival. Two fragments of burnt bone were seen in context 352, one of which is part of a distal cattle metacarpal.

Both the cattle bones and horse teeth are from small sized animals, thus are domestic, farmed animals not wild, hunted animals. The cattle bones suggest that the occupants consumed beef. It is unclear whether the horse teeth represent the use of horse meat for food or ritual use for the head or skull.

Detailed further work is not warranted on this assemblage. However, sites of this period producing animal bone are not common in the region. Therefore some mention should be made in the site report about the features producing animal bone, its poor condition and the species present.

Table X: Phase Context Cattle & LAR Sheep/Goat & SAR Pig Other Species Pres Comments 102 ZT P 139 ZFT Horse P 352 ZFT P Burnt bone ×2 502 TZF Horse P 510 ZF P

Key: F = Fused bone present U = Unfused bone present Z = Bones with zones present J = Lower jaw present T = Teeth present M = Measurable bone present LAR = Large ungulate SAR = Small ungulate Preservation: G = Good A = Average P = Poor

The cremation remains were washed through a 2mm sieve and stones/pebbles removed by hand prior to drying. There was apparently little loss of very small fragments of bone/ dust (few were visible in the wash–through from the sieve). Following drying the bone fragments were passed through 10mm and 5mm sieves and the residues weighed. Recognisable human bone fragments include pieces of skull (×16), a possible rib fragment, femoral shaft with linea aspera, long bone shaft (large and small diameter) were recovered from sample BL 98 430. There are no indications as to sex or age at death.

Retained by 10mm sieve = 55.2g Retained by 5mm sieve = 31.9g Retained by 2mm sieve = 3.4g

The volumes/weights of bone in each context is small and none represents the remains of a complete individual.

The fragments are small but there is an absence of very tiny fragments and bone dust (as evidenced by the wash–through material). This probably indicates selective picking of cremated remains from the pyre for burial, with no deliberate breaking of bone.

The colour of bone fragments indicates incomplete oxidation consistent with a lower temperature burning; fragments are grey/dark blue with occasional black charring. It is noticeable that the inside of the long bone fragments is a darker colour than the outside (i.e. lower temperature and less efficient burning of the larger bones of the body)

Whilst skull fragments and long bone shaft fragments are present, there is a lack of certain other bones such as teeth roots (which tend to survive, especially at lower temperatures), and vertebral fragments (either bodies or processes). Only three phalangeal fragments were noted in total, and only two pieces of bone with articular surface. This perhaps suggests very selective retrieval of material for burial with much of the body being left at the pyre site or buried elsewhere.