130 excavations at kingscote and wycomb,...

47
130 least two from Britain, one from Moorlinch in the Polden Hills, undated (Bulleid and Gray 1911, 195, fig 42.4) and the other from Poundbury, Dorchester, where it came from a late Iron Age/early Roman grave (Mackreth 1993a, 84-87, fig 63.1). On balance, the date-range is likely to run from the latter part of the first century BC at the latest to the third or fourth decade of the first century AD: had it been much later, the association with Roman sites should have been stronger. 86 Fig 69 Half the spring with the pin is missing. Only the head of the bow survives. It is thin and relatively broad with a line of stamps down the middle making two rows of square indentations. FW143b. The ornament on this group was carried out by using a stamp with cross-cuts to produce short lines of square punch-marks. There can be no doubt that the inspiration for the decoration lay in the ornament found on the Nauheim, even to the use of wavy lines between borders. The distribution of the type shows conclusively that it was centred on Hants and Berks with a sizeable overspill into Glos and Wilts. In tribal terms, these are the lands of the Atrebates and, as there was a very low incidence of Colchesters in these lands, it should be seen as a mark of that tribe. The two examples from Wroxeter ought to have arrived there with the army in 55 I 60 at the earliest, otherwise the dating is, second century and later contexts being omitted: Fishbourne, two examples, 43-c 75 (Cunliffe 1971, 100, fig 36.6, 13); Hod Hill, before 50 (Brailsford 1962, 7, fig 7.C25); Wilcote, Oxon, Claudian (Hands 1993, 31, fig 24.14); Silchester, before 60 (Boon 1969, 47, fig 6.3). Considering the numbers known, this is a pitiful list. As the decoration is so close to the Nauheim, and that passed out of manufacture in the first century BC, the type ought to be early, and certainly earlier than the last quarter of the first century for survivors in use, and it can only be a matter of time before at least one specimen is published from a pre- conquest context. 87 Fig 70 The first coil of the spring survives. The bow is fairly broad at the top and tapers to a pointed foot. The upper bow has two cross-grooves top and bottom, a bordering groove between and a series of cuts along each margin. FW39. 88 Fig 70 There seems to be half an axis bar through the surviving half spring. The bow is like that of the last brooch, only having a central furrow with two cross-grooves top and bottom. FW44. 89 Fig 70 The spring is missing. The bow has a thin rectangular section with, half-way down to the top of the catch-plate, a cross-groove with two Vs cut in below. FW107. 90 Fig 70 The pin is missing. The bow has a narrower rectangular section than that in the last brooch and two cross-grooves marking the central section. FW17. 91 The spring is missing. The bow is even narrower and, here, has a median groove down its upper part. FW133. Excavations at Kingscote and Wycomb, Gloucestershire 92 The spring is missing. The brooch is a plain version of the last. FW70. These are Nauheim Derivatives (see after Wycomb 17 for comments on the Drahtfibel Derivative) as the rectangular section of their bows shows. Five are decorated and this should be a sign that they are earlier than the bulk of plain examples which tend to have indefinite bow sections. Brooches with bordering grooves and rocker-arm ornament down the triangular upper bow, but without the framed catch-plate which is the mark of the parent (eg Jackson and Dix 1987, M74, fig 22.6), should be counted as being pre- conquest in manufacture and that this is to be expected is shown by a brooch from Fox Holes Farm dated 80-20 BC (Mackreth 1989c, 132, fig 76.5); one from Baldock should also have been pre-conquest (Stead and Rigby 1986, 109, fig 40.15), and one from the King Harry Lane cemetery fortunately occurs in Phase 1, safely pre-conquest whether the phase starts at 15 BC or AD 1 (see after Brooch 4, Stead and Rigby 1989, 354, fig 154.317.4). There is, as yet, no means of assessing how long the true descendants of the Nauheim, those with decoration recalling the parent, remained in manufacture. Even allowing a long life, none is likely to have survived in use as late as AD 75. As for Brooch 92, although it may have a regular rectangular section, it could still have lasted in manufacture longer than the decorated ones, which means that it could have lasted to near the end of the first century. 93 The spring is missing, as is most of the catch-plate. The bow has a lozenge section. FW142. 94 The start of the spring is present. The bow has a circular section, a marked recurve in profile and ends in a bulbous foot-knob. FW4. Brooches with lozenge sections like 93 are uncommon and none offers evidence for a proper date-range: probably first century AD. The recurve in the profile of Brooch 94 and the foot-knob show that the maker was influenced, either directly or at second-hand, by the Aucissa series. As most are post-conquest in arrival, a post-conquest date is likely, but as members of the earlier part of the sequence had arrived in Britain before the conquest, it is just possible that the maker of this brooch received his influences from those as opposed to the named type. Aesica 95 Fig 70 The spring was held in a version of the Polden Hill manner (see above 15), but how the chord was held is uncertain. The brooch was a two-part casting of which only the bow survives. The head of the bow had been either cast or bent back so that the main face ran down to form the back of the spring case. There is a small, rectangular depression which may have housed a hook or loop to hold the chord. The bow has a broad top and splays out to a knob on each side before sweeping back to the centre where there is a small projection above a moulding, and then a peg to which the fantail foot was once fitted. There is a hole in the middle of the bow and this, as well as corrosion products

Upload: doankiet

Post on 18-Mar-2018

216 views

Category:

Documents


1 download

TRANSCRIPT

130

least two from Britain, one from Moorlinch in the Polden Hills, undated (Bulleid and Gray 1911, 195, fig 42.4) and the other from Poundbury, Dorchester, where it came from a late Iron Age/early Roman grave (Mackreth 1993a, 84-87, fig 63.1). On balance, the date-range is likely to run from the latter part of the first century BC at the latest to the third or fourth decade of the first century AD: had it been much later, the association with Roman sites should have been stronger.

86 Fig 69 Half the spring with the pin is missing. Only the head of the bow survives. It is thin and relatively broad with a line of stamps down the middle making two rows of square indentations. FW143b.

The ornament on this group was carried out by using a stamp with cross-cuts to produce short lines of square punch-marks. There can be no doubt that the inspiration for the decoration lay in the ornament found on the Nauheim, even to the use of wavy lines between borders. The distribution of the type shows conclusively that it was centred on Hants and Berks with a sizeable overspill into Glos and Wilts. In tribal terms, these are the lands of the Atrebates and, as there was a very low incidence of Colchesters in these lands, it should be seen as a mark of that tribe. The two examples from Wroxeter ought to have arrived there with the army in 55 I 60 at the earliest, otherwise the dating is, second century and later contexts being omitted: Fishbourne, two examples, 43-c 75 (Cunliffe 1971, 100, fig 36.6, 13); Hod Hill, before 50 (Brailsford 1962, 7, fig 7.C25); Wilcote, Oxon, Claudian (Hands 1993, 31, fig 24.14); Silchester, before 60 (Boon 1969, 47, fig 6.3). Considering the numbers known, this is a pitiful list. As the decoration is so close to the Nauheim, and that passed out of manufacture in the first century BC, the type ought to be early, and certainly earlier than the last quarter of the first century for survivors in use, and it can only be a matter of time before at least one specimen is published from a pre­conquest context.

87 Fig 70 The first coil of the spring survives. The bow is fairly broad at the top and tapers to a pointed foot. The upper bow has two cross-grooves top and bottom, a bordering groove between and a series of cuts along each margin. FW39.

88 Fig 70 There seems to be half an axis bar through the surviving half spring. The bow is like that of the last brooch, only having a central furrow with two cross-grooves top and bottom. FW44.

89 Fig 70 The spring is missing. The bow has a thin rectangular section with, half-way down to the top of the catch-plate, a cross-groove with two Vs cut in below. FW107.

90 Fig 70 The pin is missing. The bow has a narrower rectangular section than that in the last brooch and two cross-grooves marking the central section. FW17.

91 The spring is missing. The bow is even narrower and, here, has a median groove down its upper part. FW133.

Excavations at Kingscote and Wycomb, Gloucestershire

92 The spring is missing. The brooch is a plain version of the last. FW70.

These are Nauheim Derivatives (see after Wycomb 17 for comments on the Drahtfibel Derivative) as the rectangular section of their bows shows. Five are decorated and this should be a sign that they are earlier than the bulk of plain examples which tend to have indefinite bow sections. Brooches with bordering grooves and rocker-arm ornament down the triangular upper bow, but without the framed catch-plate which is the mark of the parent (eg Jackson and Dix 1987, M74, fig 22.6), should be counted as being pre­conquest in manufacture and that this is to be expected is shown by a brooch from Fox Holes Farm dated 80-20 BC

(Mackreth 1989c, 132, fig 76.5); one from Baldock should also have been pre-conquest (Stead and Rigby 1986, 109, fig 40.15), and one from the King Harry Lane cemetery fortunately occurs in Phase 1, safely pre-conquest whether the phase starts at 15 BC or AD 1 (see after Brooch 4, Stead and Rigby 1989, 354, fig 154.317.4). There is, as yet, no means of assessing how long the true descendants of the Nauheim, those with decoration recalling the parent, remained in manufacture. Even allowing a long life, none is likely to have survived in use as late as AD 75. As for Brooch 92, although it may have a regular rectangular section, it could still have lasted in manufacture longer than the decorated ones, which means that it could have lasted to near the end of the first century.

93 The spring is missing, as is most of the catch-plate. The bow has a lozenge section. FW142.

94 The start of the spring is present. The bow has a circular section, a marked recurve in profile and ends in a bulbous foot-knob. FW4.

Brooches with lozenge sections like 93 are uncommon and none offers evidence for a proper date-range: probably first century AD. The recurve in the profile of Brooch 94 and the foot-knob show that the maker was influenced, either directly or at second-hand, by the Aucissa series. As most are post-conquest in arrival, a post-conquest date is likely, but as members of the earlier part of the sequence had arrived in Britain before the conquest, it is just possible that the maker of this brooch received his influences from those as opposed to the named type.

Aesica

95 Fig 70 The spring was held in a version of the Polden Hill manner (see above 15), but how the chord was held is uncertain. The brooch was a two-part casting of which only the bow survives. The head of the bow had been either cast or bent back so that the main face ran down to form the back of the spring case. There is a small, rectangular depression which may have housed a hook or loop to hold the chord. The bow has a broad top and splays out to a knob on each side before sweeping back to the centre where there is a small projection above a moulding, and then a peg to which the fantail foot was once fitted. There is a hole in the middle of the bow and this, as well as corrosion products

The Finds

on the lower part, may be related to a repousse plate applied to the face of the bow. FW93.

The origin of Aesicas in the developed Rosette Brooch, and their basic distributions have been discussed (Mackreth 1982a). Little has occurred since then to alter the general conclusions. The complex casting of Brooch 95 suggests an early place in the overall sequence which can run into the earlier second century, as an example from Halam, Notts, shows (Bishop and Freeman 1993, 183-6, fig 10.3). There is a hint that the type had a pre-conquest origin: a badly damaged brooch from the King Harry Lane cemetery ought to have been an Aesica (Stead and Rigby 1989, 350, fig 151.306.4). The grave was unphased, contained a developed Rosette, a developed Leontomorphe, a plain Langton Down and a late Colchester; the single pot belonged to the late pre-Roman Iron Age tradition. This grave cut another of Phase 2 which contained a range of imported wares, none of which needs to have been Claudian. If the grave with the brooches belonged to Phase 3, then it need not have been later than 45 (see comments on Brooches 1-4).

Looking at the Aesicas which come from the area which, as far as the Colchester Derivatives are concerned, has produced most of the Kingscote brooches, Wilts and Glos have more than any of the other counties to the north, west or south-west. Wiltshire seems to have more examples with complex castings than Glos and all single-piece castings have Polden Hill spring systems (see above, Brooch 15). Kingscote has already produced two Aesicas (Glos Museum, 104/1969, 1.54/1971), both with the Harlow spring system (see above, Brooch 8), and one from Cirencester has a rearhook. These serve to show that influences came from the east, even if they were ironed out by the end of the first century, as an example of the main single-piece casting group shows (Mackreth 1989a 188, fig 26.1, 67). The present example is not, as may appear at first sight, easy to classify: the pierced lugs at the end of the wings should point to a Polden Hill spring system, but there is no obvious hook, forward- or rearward-facing, or pierced crest to hold the chord. There is, however, a hole in the bow which may once have been used to mount a device for securing it. As none of the brooches which had applied repousse decoration is similarly provided with holes, there is a suspicion that it is, possibly, an experimental piece and therefore may date before c 60/5.

Birdlip

96 Fig 70 The pin was hinged, the axis bar being housed in stubby plain wings. The rest of the brooch consists of a short bow above a long foot. The bow has convex sides defined by a groove. Down the centre runs a groove, with punch-marks along it, which divides, the two branches running to the corners. The top of the foot has a beaded cross-moulding above a forward-facing curved plate which ends in a blunted point. The rest is plain, save for a groove along each margin and a divided projecting moulding at the bottom. FW60.

97 Fig 70 The pin and the wings are the same as in the last

131

brooch. The bow here is narrow and deep, the back face being broader than the front. The sides are concave and, at the bottom, merge into the top of the foot which projects forward and is curved in plan. The foot has concave sides and a slightly curved lower edge. On the front is engraved a median line between lines following the sides and meeting in the middle. Within the basically triangular forms at each end is a curve rising from the top and bottom margins; the semi-circle so formed is divided to form two plain lenticular areas, the other spaces are filled with a herringbone pattern. Site 2, (16), Area 11, Period 5.

The origin of the Birdlip has not been firmly established. If the beak-like feature, such as that on Brooch 96 here, is important, then the earliest form can be identified and it will only be a matter of time before the intermediate stages are discovered. The suggested original type is a member of a class of brooch discussed by Stead (1976a) and examples with framed catch-plates (eg Brown 1986, 15, fig 8.5) should take it back to the middle of the first century BC, those with fretted catch-plate (eg Hawkes 1940, 492 and fig) should carry it forward into the earliest part of the first century AD,

and a small group of almost standard items all with a series of small circular holes in their catch-plates (eg Stead 1976a, 404, fig 2.4) should point to dating in the region of AD 25/30, one being as late as c 50/5 (Frere et a/1987, 185, fig 165.3). Brooch 97 is associated with 96 because of the general form, the 'beak' not being an absolute requirement, sometimes being so turned up that it forms a loop with the bow above (eg Reading Mus, 03206). Both also have hinged pins, and this occurs frequently in the main group, although the Birdlip Brooch itself and most of the highly elaborate ones have integral four-coil-internal-chord springs. There is no really clear pattern in the distribution. However, hinged­pin brooches seem to be more frequent in Lines and in a band running down to Hants. There is, however, a remark­able lack from the lands of the Catuvellauni. Dating is difficult. The only one recorded by the writer as coming from a definite context is the Birdlip Brooch itself, and much depends on art-historical argument based on the knife handle and mirror. The safest conclusion is that Brooch 97, with its fine engraving, may be associated with another from Wattlesborough Heath (private coll), on which engraving occurred on the back and under the foot of the bow, and that these two represent a milieu which could have produced the Birdlip Brooch itself, as well as another from Ham Hill, Soms (Fox 1958, fig 59), these having finely pierced catch-plates which reflect pre-conquest practice, but which are, perhaps, more safely assigned to the middle years of the first century AD. This is indicated by a fine silver, parcel-gilt example with a rearhook from Weeting, Norf (private coll), which should not be later than 60/5.

Late La Tene, Continental

98 Fig 70 The spring-case has a panel on the front formed by a line across the top and evidence for at least one line down each side. A cross-moulding separates the case from the bow. This is reeded with three major flutes divided by

132

1.87

1.95

)1

r .I

1.98

1.99

0

I! 11, I!

' I

Excavations at Kingscote and Wycomb, Gloucestershire

1.88 1.90

1.89

1.96 1.97

1.100

50 m

Fig 70 Sites 1, 2 and field-walking: 1 Objects of personal ornament/dress: copper-alloy brooches. Scale 1:1

The Finds

minor ones into which mouldings are introduced at the top to accommodate the splay here. The catch-plate has a piercing whose three sides are curved. FW14.

This and Wycomb 18 are standard Langton Downs, lacking beading and without any waisting in the front view, or recurve in profile. The King Harry Lane cemetery (Stead and Rigby 1989) is the only site published in Britain so far which can be used for analytical purposes. It reveals that beaded Langton Downs are generally earlier than unbeaded ones: beaded, Phase 1 -five examples, Phase 2-two examples, Phase 3 - one example; unbeaded, Phase 1 -two examples, Phase 2 - none, Phase 3 - eight examples. Phase 4 does not figure as few graves were assigned to it. Using the revisions suggested for the dating of the phases (see after Brooch 4), Phase 3 would end about 40/5. However, other site dating shows that the latter end of the Langton Down production continued in use to c 55/60, even if most survivors had entered the ground or melting pot by 50/5. As far as Brooches 98 and Wycomb 18 are concerned, little emerges from looking at the straightness or otherwise of the reeded Langton Down.

99 Fig 70 The Colchester-style spring (see above Brooch 1), is housed in a case made by folding a thin sheet forged from the casting over the top, bottom and sides of the spring. The bow is thin and well curved. It has a groove down each margin and two sunken wavy lines formed by using a punch on alternate sides of a straight ridge. The lower end of the bow rises from a disc with a plain centre surrounded by a zone of sunken large bosses, then a narrow plain band followed by a sunken zone with a line of ring­and-dot stamps in the bottom. The foot issued from the back of the disc and is broken off before its front face could be exposed. The brooch was cast as a straight item with a disc in the middle and was hand-worked to its final form. Site 1, Quarry (2), 40ins.

Only one site in Britain provides useful information on the typological development of the Rosette, and its close relative the Leontomorphe, and that is the King Harry Lane cemetery. Before dealing with the statistics, it is necessary to look at the typology as it is presented in the report (Stead and Rigby 1989, fig 49, F, G and Ll). The different manu­facturing styles are not arranged in a chronological frame­work (Stead and Rigby 1989, 93-4), but the graves listed under each 'style' reveal the fairly obvious: central discs separate from the fantail foot are earlier than those which have the disc cut from the same piece of plate. Those lacking a proper bow between the disc and the spring-case are later still. This can be used as a rough guide for those in unphased graves. The dating of the cemetery (see after Brooch 4) is not really at issue here: all that need be noted is that the present brooch belongs to a typological stage earlier than any at St Albans, which means that it is first century BC, and probably before c 20 BC.

Trumpets

Brooches 100-10 have or had their springs mounted on a sheet metal tube running through the coils and a large loop behind the head of the bow. The tube itself is kept in

133

position by the ends of a wire loop over the head, the waist of the loop being caught by a collar. No loop and collar survives and only Brooches 102 and 104 retain their tubes.

lOO Fig 70 The head is covered with relief ornament of Celtic type. An element rises from a buried boss at the bottom on each side to interlock with another running down the middle. The knop, made up of a marked central disc with multiple petals on each side, has double cross­mouldings above and below. The lower bow has a panel on each side, filled with more relief ornament consisting of a Z motif linking two buried bosses, and a cross-moulding at the bottom. The foot-knob has four cross-mouldings, with a boss underneath, under a flute. The lower bow had one side partially removed in antiquity. FW57.

101 Fig 71 The upper bow has relief decoration. In this instance, it rises from a comma motif on each side and has an elaborate trumpet element tucked underneath. The knop is a repeat of that on the last brooch, except for a more bulbous centre. The lower bow has two incised S motifs on each side, thin enough to have once been inlaid with a different metal, and ends in three cross-mouldings. The bulbous foot-knob has a thin cross-moulding under a flute. FW58.

This discussion begins by presenting all the dating evidence for Trumpets known to the writer, excluding brooches like Brooch 117. The advantage will be that the whole picture can be seen, and on what slender grounds it rests, bearing in mind the hundreds that are known. Once the presentation has been made, it should be easier to see how the various sub-varieties present in this collection relate to the overall picture. Those brooches with their springs housed between pierced plates, like Brooches 111-16 and Wycomb 21-2, are indicated by DL; those with hinged pins by H; those with relief or inlaid ornament by R; and those with enamelled decoration by E. For those marked K, see after Brooch 106. All examples from contexts later than the early third century have been left out and the dating is arranged by initial date.

The dating is: Tewkesbury, K, 50-140 (Hannan 1993, 66-7, fig 19.9); Baginton, Warks, two examples, one E, before 70/5 (Hobley 1967, 110, fig 19.9; 1973, 66, fig 19.8); Nettleton, 69-117, coin-dating (Wedlake 1982, 127, fig 53.54); Whitton, Glam, K, 70-95 (Jarrett and Wrathmell 1981, 175, fig 70.24); Holcombe, Devon, R, 70-180 (Pollard 1974, 138-40, fig 22.2); Prestatyn, two examples, one R, 70s-160 (Mackreth 1989b, 95, fig 38.17-18); Corbridge, c 75-90/5 (Daniels 1959, 156: Snape 1993, 32, fig 4.6); Leicester, K, late first century (Clay and Pollard 1994, 145, fig 74.24); Caerleon, 80-100 (Brewer 1986b, 170, fig 54.3); Harlow, DL, 80-100? (France and Gobel 1985, 79, fig 41.74); Derby, DL, 80-140 (Mackreth 1985, 291-3, fig 128.29); Watercrook, c 90-95 (Potter 1979, 209, fig 84.8); Prestatyn, c 90-100/20 (Mackreth 1989b, 96, fig 39.20); Caerleon, E, c 100 (Wheeler and Wheeler 1928, 162, fig 13.14); Higham Ferrers, Northants, E, second century (Meadows 1992,88-9, fig 7.2); Whitton, Glam, R, second century (Jarrett and Wrathmell 1981, 174, fig 70.21); Wroxeter, R, 110-30 (Bushe-Fox 1913,

134

26, fig 9.7); Watercrook, c 110-40? (Potter 1979, 208-9, fig 84.6); Halstock, Devon, after c 110/20 (Mackreth 1993b, 79, fig 14.19); Wroxeter, DL, c 110-30 (Bushe-Fox 1913, 26, fig 10.8); Silchester, early second century (Cotton 1947, 144, fig 8.1), c 120-160/70 (Cotton 1947, 45, fig 8.2), Hadrianic­Antonine (Boon 1969, 47, fig 6.7); Caerleon, K2, DL, Hadrianic-Antonine (Brewer 1986b, 170, fig 54.4); Brough­on-Humber, c 125 (Wacher 1969, 93, fig 39.33); Chelmsford, DL, 120/5-160/70 (Drury 1988,94, fig 62.10); Verulamium, H, 125-50 (Wheeler and Wheeler 1936, 207, fig 44.29); Carlisle, late 120s-150/80 (Mackreth 1990, 109, fig 100.9); Wroxeter, R, before 130 (Kenyan 1940, 224, fig 15.4); Wroxeter, E, before c 130 (Atkinson 1942, 206, fig 36.H65); Wroxeter, R, c 130-155/70 (Atkinson 1942,205, fig 37, H61); Caerleon, K2, DL, 130-60 (Wheeler and Wheeler 1928, 164, fig 13.16); Dover, E, DL, 130-160/70 (Philp 1981, 151, fig 33.77); Whitton, Clam, K, before 135 (Jarrett and Wrathmell 1981, 174, fig 70.22); Tewkesbury, 140-60 (Hannan 1993, 67-8, fig 19.10); Verulamium, 145-50 (Frere 1984, 25, fig 7.30); Newstead, E, four examples, and other plain ones not noted here, mid-second century (Curie 1911, 322, pl 85.11-12, pl86.13-14: Frere 1987, 136); Rudston, H, 140-60 (Stead 1980, 95, fig 61.9); Camelon, c 140-60/5 (Christison 1901, 401, fig 36); Mumrills, c 140-60/5 (Macdonald and Curie 1929, 553-5, fig 115.1); Winterton, Antonine (Stead 1976b, 201, fig 101.28); Wroxeter, mid-second century (Bushe-Fox 1913, 26, fig 9.6); Watercrook, 150-70 (Potter 1979, 210, fig 84.12); Derby, E, two examples, another DL, 150-75 (Mackreth 1985, 289-93, fig 127.25, 28, 31); Watercrook, two examples, c 150-220 (Potter 1979, 209, fig 84.5,9); Wilcote, Shakenoak Farm, DL, second half second century (Brodribb et a/1971, 110, fig 47.66); South Shields, c 160 (Miket 1983, 113, fig 72.85); Manchester, 160-earliest third century (Bryant et a/1986, 67, fig 5.5, 3194); Caerleon, 160-230 (Brewer 1986b, 170, fig 55.11); Camerton, K, before 180 (Wedlake 1958, 221, fig 51.15); Hockwold, Norf, E, late second century? (Mackreth 1986, 63-4, fig 40.3); Derby, R, late second century (Mackreth 1985, 191-3, fig 127.27); Rampton, Notts, late second century or earlier (Ponsford 1992, 117, fig 23.3); Brancaster, late second century (Hinchliffe and Green 1985,42-4, fig 28.4); Rudston, Yorks, post-Antonine (Stead 1980, 95, fig 60.8); South Shields, H, late second century (Miket 1983, 117, fig 75.157); Derby, R, late second-early third century (Mackreth 1985, 191-3, fig 127.26); Exeter, late second-early third century, two examples, one H, (Holbrook and Bidwell 1991, 237, fig 102.27-8); Carlisle, late second-early third century (Mackreth 1990, 109, fig 100.10); Dover, c 190-210 (Philp 1981, 151, fig 33.78); South Shields, E, early third century (Miket 1983, 116, fig 75.126).

The demonstrably earliest brooches remain those from Baginton, Warks, despite arguments for a pre-Roman conquest origin based on art-historical criteria relying on items themselves largely undated (Boon and Savory 1975): the data in the last paragraph reveal the fact that not one brooch can be shown conclusively to be mid-first century. One may have started as early, and two might go back that far, but there is no trace of there having been continuous

Excavations at Kingscote and Wycomb, Gloucestershire

manufacture from an earlier time. The Trumpet thus came into being between 50 and 75. Thereafter, the type in its various manifestations continued in production until the third quarter of the second century, the bulk of examples having been consigned to the ground before 175. There is a dribble dating into the early third century and it is a moot point as to how many of these are truly residual and any later than 200 should be seen as being just that. This analysis excludes any Trumpet which has a disc or other shape in place of a knop, such as Brooches 117 and Wycomb23.

Turning to Brooches 100-1, these, along with inlaid ones, tend to have a western distribution. The dating, as set out above, is a little ambiguous: eight are dated, of which four, possibly five, are earlier than c 125, one covers the middle of the second century and the remaining two can be counted as being residual. In which case, it may be that all had ceased to be made by 125 and were passing out of use in the next quarter-century. This contrasts with enamelled ones, these showing a steady distribution through the whole of the suggested true floruit, with the same chance of being residual later as all the other kinds.

102 Fig 71 The trumpet head is less pronounced than usual, has a median arris, runs up to an oval head-plate and ends below in a double moulding. The knop is a triple central moulding with a flute above and below. The lower bow has a single moulding at the top, a repeat of the arris and a simple two-part foot-knob. Site 1, Quarry (2), level 60ins.

103 Fig 71 Almost exactly the same as 102; the mouldings next to the knop are now two lenticular elements and the foot-knob is a simple projection. FW3.

104 Another repeat, this time with a buried ridge down the trumpet and a groove down the lower bow. The foot­knob is like that of Brooch 102. FW16.

105 As 103, with a ridge down the trumpet and with the foot-knob of Brooch 102. FW11.

106 As 105, but the foot-knob lacks the cross-moulding. FW10.

107 The head is lost, but the rest is essentially as 105. FW118.

The distribution of the variety to which these, and Wycomb 19, belong is mainly from the southern Severn valley into the Midlands. There are two further examples from Kingscote (Glos Mus, 6/1968, 22/1968) which has now produced far more than any other site, but in such a large collection one would hesitate to suggest that this was necessarily near the centre of production. The dating is indicated in the analysis above by K, and reveals that, far from being typologically devolved, it is an individual variety dating from the last quarter of the first century, two examples, into the first half of the second. The only one later than 150 could well have been residual in its context.

The Finds 135

1.101 1.102

1.103

1.110 1.111

,,,

0 50 mm

1.119 1.123 .

I~~) I I I

Fig 71 Sites 1, 2 and field-walking: 1 Objects of personal ornament/ dress: copper-alloy brooches. Scale 1:1

136

108 The head is lost; the brooch is placed here because the style of knop points to the same spring-fixing arrangement as the others in this group. The trumpet and the lower bow have median arises. The knop consists of three cross­mouldings divided by flutes, the middle one cut to suggest petals. The simple foot-knob is cut is resemble petalling. FW119.

109 Only the trumpet survives. This has a marked concave face on each side. FW117.

Neither is complete and, although the head of Brooch 108 does not survive, it is likely to have had a single loop, these outnumbering the other systems by approximately two to one. Only the general date-range discussed after Brooch 101 can be applied here.

110 Fig 71 The trumpet is very short with a nib on the head intended to prevent movement of the loop and collar over the trumpet. Above and below the knop is a pair of cross mouldings, the knop itself having grooves giving the impression of two lenticular bosses dipping in the centre. The lower bow has a median arris and the foot-knob is in two parts. The return of the catch-plate has a groove across top and bottom with a V between. Site 2, (44), Room 9, Phase 2.4.

A diminutive brooch with two vague parallels known to the writer. One from Woodeaton, Oxon, has slightly different details, and no dating; the other, from Tewkesbury, is dated 140-60 (Hannan 1993, 67-8, fig 19.10), but the upper bow is too different to be regarded as a close example. The ornament on the catch-plate return, however, should not be later than the early second century.

Brooches 111-17 have or had their springs mounted between a pair of pierced lugs behind the head.

111 Fig 71 On the head is a cast-on loop and pedestal. The trumpet has a groove around its top edge. The petalled knop has a double moulding above and a single one below. The lower bow has a groove down each side, an arris down the centre, and a foot-knob made up of two cross­mouldings. Site 2, (32), Room 9, Period 5

The loop is merely a replacement of the movable loop­and-collar found on Trumpets whose springs are mounted on single loops. Brooch 111 does not seem to belong to a particular group and can only be dated on general grounds. In the comments following Brooch 101, those with a pair of pierced lugs are indicated by DL and this system begins in the late first century and occurs fairly evenly throughout the main floruit ending 150/75. The distribution of the system, despite the number of varieties, is mainly in the Midlands, the Marches and the southern Severn Valley area extending into Wilts, but examples occur relatively frequently in the east and south-east of England. It is un­common in the northern Pennines and the north generally. One important group stands out: it has a knop made up of four almost equal cross-mouldings (eg Mackreth 1985, 291-3, fig 128.31) and occurs in the southern Pennines and the immediately adjacent areas.

Excavations at Kingscote and Wycomb, Gloucestershire

112 On the head is a loop and pedestal. The trumpet is less rounded and has a groove round its upper edge and a trace of a median arris between pairs of ridges running away to the edge. The knop is reduced to two pairs of lenticular bosses separated by a flute. The lower bow has a groove down the middle and another pair of ridges. The foot is a simple divided projection. FW19.

113 Fig 71 A repeat of the last brooch. FW113.

Brooches like these, and Wycomb 21-2, are marked by K2 in the dating list given after Brooch 101. The distribution favours the southern Severn Valley and spreads into South Wales and to Wilts, with occasional ones occurring else­where in southern England. The dating is very weak and only the general date-range for the Trumpet overall can be suggested: later first century to 150/75, with the possibility that one or two may have survived in use as late as 200.

114 Only the trumpet survives: oval at the top, with a groove round its upper edge and the trace of a pedestal; the underside is flat. FW122.

115 Essentially the same as the last brooch, the pedestal with its two cross-grooves survives. The section at the fracture is lenticular. FW124.

116 Like 113, there is no groove round the upper edge of the trumpet which retains a substantial part of the clay core inside. FW123

There is not enough present for any to be safely assigned to a particular group. The same dating as that suggested for the previous two examples is offered here.

117 On the head is a cast-on loop and pedestal. The trumpet has traces of an 'eye' on each side and a groove round the top. The knop is replaced by a pelta, also bearing traces of applied ornament. The lower bow has a narrow central face and a chamfer on each side. The foot is made up of a biconical boss under a cross-moulding. Despite being polished in modern times, it is clear that the brooch had once been tricked out with the usual compliment of applied decoration (Richardson 1960, 212, fig 2.31). FW27.

This brooch, along with Wycomb 23, is a member of a large family made in a variety of sizes, of which Wycomb 23 belongs to the larger group and 117 to the smaller. The family is distinguished by, first, applied white metal trim, second, the use of discs, D-shapes and peltas in place of the knop on the ordinary Trumpet, and third, the frequent substitution of a pelta for the foot-knob. The distribution suggests that it was made somewhere in the south-east or eastern parts of England, examples thinning out to the west and north. The dating for this variety, and incidentally for the use of white metal trim, is: Newstead, SO-early third century (Curie 1911, 332-4, pi 87.27: Hartley 1972a, 54); Camerton, 90-100 (Wedlake 1958, 224, fig 51.19); London, two examples, early second century (Wilmott 1991, 118, fig 82.303-4); Camelon, 140-65 (Christison 1901, 401, fig 39); Wroxeter, mid-second century (Kenyon 1940, 224, fig 15.3); South Shields, c 150 (Miket 1983, 113, fig 71.84); Canterbury, c 150-400 (Blockley 1988, 110-11, fig 14.18); Strageath, c 158

The Finds

(Frere and Wilkes 1989, 150-1, fig 76.56); Scole, Norf, mid Antonine-early third century (Rogerson 1977, 132, fig 54.8); Springhead, Kent, before c 175 (Penn 1957, 98, fig 14.5); Leicester, before c 180 (Kenyan 1948, 251, fig 81.1); Verulamium, late second/ early third century (Richardson 1944, 93, fig 4.5). There are, in fact, not many examples and the message is, accordingly, more diffuse than for the main series of Trumpets (see after 101). However, with the exception of the one from Camerton, all could fit comfortably between the early second century and c 175: only one needs be later and this would fit the picture provided by the other Trumpets, providing it is allowed that the one from Camerton is dated too early (see after Brooch 173).

Unclassified

118 The spring, complete with wire loop, is mounted on a lug behind the head (see above, Brooch 100). The brooch is, in effect, a single, narrow trumpet with a restricted head ending in a circular head-plate. On the front is an excres­cence which, in profile, approximates to a semi-circle with a bite out of it. The foot is a version of the usual two-part design. Site 2, (350), Quarry 2, Period 3.

An individual design, this is the common form and belongs essentially south-west of a line joining Hants and Glos. There is an elaborate form in which the top of the trumpet head is extruded, bent down to where the knop on a Trumpet should be, and then bent up and out (eg Pitt Rivers, 1888, 117, fig b). One of these has decoration on the catch-plate return. There is also one from Nor'Nour (Dudley 1967, 36, fig 154.49) which has a decorated upper bow and a loop beneath, which recalls some Birdlips on which the beak is bent up to touch the upper bow (see after Brooch 97). Two others, which have a boldly projecting plate with a shaped profile, and a cast-on loop and pedestal above a spring mounted between two pierced plates (Hattatt 1985, 112, fig 46.444; excavations, Dr G Webster, to be published), both lie north of the main distribution, but more examples are needed before this can be seen as being significant. One elaborate form from Camerton is dated to 150-200 (Wedlake 1958, 224, fig 51.17); no other known to the writer is dated. As the elaborate form can have a decorated catch-plate, it should belong to the later first and early second centuries.

Headstud

119 Fig 71 There had been a Polden Hill spring system (see above, Brooch 15). The poor condition of the brooch means that some of the details are obscure. There are the remains of a loop and pedestal. Each wing splays back from the bow, has a vertical groove at its end and a ring-and-dot motif on the main face. The stud has a crest above and a peak at the bottom. The setting in the stud had been separately made and then held by a rivet through the bow. FW20.

The plain bow and swept back wings, along with the loop and pedestal, should place this in a small group the

137

members of which recorded by the writer come mainly from the north with a thin scatter over the rest of Roman Britain. The group is practically undated: Carlisle, late second/ early third century (Mackreth 1990, 108, fig 100.6). However, the writer has no parallel for the Polden Hill spring system and loop and pedestal. Only a general date­range can be offered: late first into the earlier second century, perhaps running on to 150?

Unclassified

The spring is housed in the head, an axis bar through the coils being seated in pierced lugs, the chord fitting snugly into the semi-circular bottom of the head-plate.

120 Fig 71 The head-plate has a flat top, a semi-circular bottom and a moulding right round the forward edge. The bow is like that of Brooch 40, only with a rounded back. The chief feature is the enamelling. On each side is a line of lozenge cells with infilling triangles. There is red in the former and traces of mid to dark blue in the latter. Down the centre of the upper part of the bow run two ridges which end in a circular flat area which may have traces of a white metal solder. The foot is essentially a fat truncated cone with two cross-mouldings above and a single one below set back from the edge of the cone. There may have been a circular hole in the catch-plate. FW68.

121 The head-plate has, above a moulding, a crest with concave cut-outs on each side each bordered by a trumpet form. The bow has a flat face at the top with two longi­tudinal cells for enamel, now missing. The cells are stopped by a moulded section like that on Brooches 111-12. Only an indistinct part of the lower bow remains. FW61.

Both these and Wycomb 24 belong to a family which has a variety of ornament and two sprung-pin systems, that shown on Brooches 120-21 and that on Wycomb 24. The straight-sided upper bow, frequently enamelled, with a quadrant profile, is common, ending below in a boss or some ornament which can recall the knop on some varieties of Trumpet. The group seems to have evolved from a Colchester Derivative using the Polden Hill system (Webster and Daniels 1970, 18-19, fig 4.3) and quickly transmuted into recognisable versions of the family (Shrewsbury, Rowley's House Mus, X24, X26-27). This is not the place to explore all the varieties and their basic distributions: none of the three here is at variance with such considerations. Two brooches, one an unfettled casting, the other a lead pattern, point to at least one area where part of the family was produced: the southern Pennines (Mackreth 1983, 56, fig 3.5,6). Dating is weak: Winchester, 43-140 (Cunliffe 1964, 88, fig 24.3); Colchester*, 60/1-c 150 (Crummy 1983, 13, fig 9.66); Prestatyn*, 70s-160 (Mackreth 1989b, 88, fig 36.3); Silchester, 100-20 (Cotton 1947, 145, fig 8.3); Derby, 115-40 (Mackreth 1985, 293-4, fig 128.33); Biglands, 125-180/97 (Potter 1977, 171, fig 11.16). Those marked* are prototypes, and probably indicate an origin in the later first century; the rest of the series probably came to

138

an end c 150/75. However, more are needed to clarify the position.

122 The pin was hinged, the axis bar being housed in a semi-circular projection behind the base of the head-plate. This has a rounded top rising in an arc over the bow to be topped by loop. The thin bow is broad at the top wit~ only a slight taper down to the fracture below which all IS lost. Site 2, (268), Building VII (Room 17), Phase 6.3.

Although only the head of the bow survives, the rest of the bow probably tapered to a fairly broad projecting foot having one to three cross-mouldings; the profile is frequently recurved and ornament often occurs round the top edge of the head-plate. The feature leading to this conclusion is the use of a hinged pin housed in a projection along the bottom of the head-plate. The distribution reveals that the group is local, extending southwards into Wilts and Soms. Dating is poor: Nettleton, 69-117, coin dating (Wedlake 1982, 127, fig 53.49); Marshfield, probably second century (Blockley 1985, 145, fig 45.25). The Nettleton piece has a decorated catch-plate return and the tentative conclusion is that the date-range probably runs from the late first to c 150/75.

'Aucissa'-Hod Hill

Brooches 123-42, with the possible exception of 125, have or had the axis bars of their hinged pins housed in the rolled­over head of the bow. Brooches of this family tend to have two-part foot-knobs: a bulbous boss with a cross-moulding at the top. Only variations will be noted.

123 Fig 71 The head-plate has a bead-row next to the pin, then a flute followed by a name in relief. The name reads SVGG, with an incomplete letter at the end. The upper bow has the usual bordering ridge and sunken bead-row down the centre of the swollen front to be found on named Aucissas. The lower bow has a chamfer down each side and ends in a separately made foot-knob which is more of a plain ring than the two-part design found on the named type. FW13.

124 A standard uninscribed Aucissa, ie, like the last, but with no name and the usual two-part foot-knob. Site 1, Quarry (1), level35ins.

125 Although very heavily worn and with the head missing, this item has all the signs of having been a standard Aucissa. FW120.

126 The head is missing. The upper bow has a central flute with a groove on each side. At the junction with the plain foot is a section marked by a cross-groove above and below with a cut-out on each side between. There is no trace of a peg for the separately made foot-knob. FW134.

The first three are related to the named Aucissa, and Brooch 123 introduces a name new to Britain-SVGG (Frere and Tomlin 1991, 1-14), one not recorded by Behrens (1950). The other two, although without the name, are Aucissas, and Brooch 123 can be counted as belonging to the same strand

Excavations at Kingscote and Wycomb, Gloucestershire

in the overall development. The named type stands at the end of a development beginning in the Alesia in the first century BC (Duval 1974). Several examples of t~e inter­mediate stages, as well as the Alesia itself, were Imported into Britain before the conquest. However, no Aucissa as such is known as yet to have been introduced so early and the type seems to have ceased being made by c 40, if not shortly before, as its progeny, the Hod Hill, was fully developed at the time of the Roman conquest. The distribution of the Aucissa proper in Britain gives a strong sign of when it was finally passing out of us~. Mos~ are south-east of the Fosse Way military zone, with military sites in central southern England showing well. The deep south-west and East Anglia, defined as Norfolk and Suffolk are very poorly represented. While the Dumnonian peninsula seems to have resisted Romanisation for some time, the general lack in the lands of the Icem should have a bearing on the general dating of the Aucissa in Britain. Beyond the Fosse Way, the type is so excessively rare as hardly to be present at all, except for a large group at Wroxeter. The writer has discussed the Wroxeter military assemblage (Webster forthcoming): the Aucissa had passed out of use before any serious attempt was made to bring the lands beyond the Fosse Way permanently into the Province. Therefore, 55 I 60 is a reasonable estimate for the final period of disuse. The lands of the Iceni were not formally taken into the Province until the death of Prasutagus and not really until the suppression of Boudica's rebellion, Aucissas seem not to have been around when this happened. As for Brooch 126, the treatment of the bow at the junction of the upper with the lower part relates the item to the Aucissa, but the bow section and the nature of the lower bow both show that this could not be an Aucissa as such.

127 Fig 72 There is no head-plate. The upper bow has a sunken bead-row down the middle and a wide flute on each side. There is a small step at the top of the plain lower bow and the separately made foot-knob is like that on Brooch 123. FW18.

128 The upper bow has a marked flute down the centre, with a sunken bead-row on each side, stopped at the bottom by two cross-mouldings divided by a flute. The lower bow is plain and the foot-knob fitted on a peg. FW9.

129 Fig 72 The upper bow has four vertical ridges, with cross-cuts, divided by flutes. The lower bow is flat, wide at the top with rounded shoulders and has a punched-dot central line with arcs peeling off that. The foot-knob was made separately. The brooch was tinned or silvered. FW7.

These three, and Wycomb 25, to some measure fill the gap between the Aucissa and the Hod Hill. The feature which sets them apart from these is the separately made foot­knob. Brooch 127 has a recognisable version of the Aucissa upper bow, and Brooch 128 has an upper bow cop~ing an alternative upper bow design running parallel with the style used by the Aucissa before the named type came into being. Brooch 129 is entirely a Hod Hill, except for the foot. Unlike the Aucissa, or the Hod Hill (see below), one

The Finds

\.

lil i

I

1.127

1.138

1.145

I,

1.152

1.129

1.147

1.153

1.133

1.142 1.143

1.149 1.151

1.155

0 50 mm

Fig 72 Sites 1, 2 and field-walking: 1 Objects of personal ornament/ dress: copper-alloy brooches. Scale 1:1

139

140

example of this stage in the development has come from a pre-conquest deposit: Skeleton Green, AD 15-40 (Mackreth 1981a, 134-5, fig 72.53). The rest are poorly dated and, as in the case of the Aucissa itself, the British distribution perhaps reveals more about the date than anything else. The general type is well represented in the lands of the Catuvellauni and fairly well in the lands beyond the north­west limits of the tribe, amongst the Atrebates and amongst the Dobunni. To some extent this is a false picture as much depends upon the excavation of Roman military sites in these areas. However, the lands of the Iceni are poorly represented and there is only one from Wroxeter. All this should show that brooches of this category had passed out of use by 55 I 60.

130 The head is missing. The upper bow is a repeat of that of Brooch 128, but with a projecting boss on each bottom corner. The lower bow is flat and in the shape of a lozenge. The brooch was tinned or silvered. FW43, (Field IV, Site 5).

131 The upper bow has three beaded ridges down the middle and a flute and a bordering ridge on either side. The lower bow is flat with a sunken bead-row down the middle; the foot-knob is separated from an extra moulding by a flute. FW12.

132 The upper bow has a median beaded ridge separated from bordering ridges by flutes. The plain and flat lower bow has two cross-mouldings and flutes at the top. FW5.

133 Fig 72 The upper bow has a pair of prominent cross­mouldings and flutes top and bottom. Between, the bow splays out towards the bottom with a broad centre having a sunken bead-row down the middle, a flute on each side and a bordering ridge. The lower bow is straight-sided with a groove down the median arris and another down each side and across the foot, which projects slightly, replacing the foot-knob. FW41.

134 Fig 72 The upper bow is similar to that in the last, but has a flute on each side of the central bead-row and the cross-mouldings top and bottom are replaced in favour of a single one below. The lower bow is flat and appears to end in a fish tail. FW62.

135 A variation on the upper bow of the previous two; it does not splay out, the broad faces on the last have been reduced to narrow ridges, and the bordering ridges seem not to be beaded. The lower bow tapers to the foot-knob. The brooch was tinned or silvered, possibly differentially. FW38.

136 The head is lost. The upper bow splays out towards the bottom and has five ridges, divided by flutes, the middle three beaded. The lower bow is set off from the upper by a flute and a step and is broad at the top, with a line of cross-cuts down the middle, tapering markedly to the missing foot-knob. FW75.

137 The upper bow has a prominent cross-moulding and flute above and below and three vertical ridges between, divided by wide flutes between. On each side is a wing

Excavations at Kingscote and Wycomb, Gloucestershire

ending in a boss. The lower bow has a central cross-ridge with a flute on each side. FW52.

138 Fig 72 The design has been reduced to a symmetrical arrangement about a horizontal central bar, with a groove along each side and two small mouldings at the end. Above and below this is a set of three narrow cross-mouldings divided by flutes. FW74.

139 The upper bow has a wide central flute with a narrower one on each side next to the bordering ridge. The flat and plain lower bow has a flute and cross-moulding at the top and tapers to the foot-knob. The brooch was tinned or silvered. Site 1, Quarry (1), level40ins.

140 The upper bow consists of a flute with cross-cut borders. The lower bow, set off from the upper by two flutes and cross-mouldings, is flat and tapers outwards to a straight line at its base. The brooch was tinned or silvered. FW54.

141 Only part of the top survives. This has a median arris, and a narrow cross-moulding at its top and tapers towards to the fracture. FW115.

Wycomb brooches 26-31 are included in this discussion. Had it not been for the Swain collection, Hod Hills would have been poorly represented in this assemblage. In general, all Hod Hills were provided with a silvery finish, probably by using tin, and in some cases, this was done with enough precision for parts of the design to gleam gold to the silver of the rest. No Hod Hill has been convincingly published from a pre-conquest context, although it is almost a matter of chance that this has not happened as the brooch arrived fully formed, and in great variety, with the army of conquest. One example from Skeleton Green may have been pre-conquest, but it was not unequivocally sealed beneath the flood (Mackreth 1981a, 141-2, fig 72.55) and another from Baldock was assigned to the first quarter of the first century AD, which must be incorrect as it would then pre-date its parent (Stead and Rigby 1986, 120, fig 47.112). The Hod Hill is not present in the King Harry Lane cemetery; in fact, its precursors are hardly there either: Phase 3, a Rosette design with a rolled-over head (Stead and Rigby 1989, 314, fig 114.153.2); Phase 3, a rolled-under head and an alternative bow design to the Aucissa proper (Stead and Rigby 1989, 326, fig 127.205.3); Phase 3, a 'Bagendon' type known to be pre-conquest (Stead and Rigby 1989, 330, fig 132.233.3); unphased, an alternative bow design to the Aucissa proper (Stead and Rigby 1989, 314, fig 116.158.2); unphased, an Aucissa type (Stead and Rigby 1989,366, fig 163.362.2). These few examples serve to show that there are no grounds for thinking that any of the main phases should run later than 40/5 at the latest (see after Brooch 4).

Virtually all Hod Hills display the two-part design of the Aucissa and the examples here offer a fair sample of the range of designs, ranging from plain derivatives such as Brooches 131-2 to the more extravagant Brooch 138. The eponymous site is instructive both for the variety of designs and their dating: before 50 (Brailsford 1962, 9, figs 8-9,

The Finds

C54-C79: Richmond 1968, 117-19) which, again, points to the early close of the King Harry Lane cemetery. None of the present collection, save for Brooches 141 and Wycomb 31, would be out of place there. Wycomb 31, because of its separately made foot-knob, is early (see above) and may possibly owe something to the Langton Down. Brooch 141 could have been a simplified version of another straight­through design (eg Bushe-Fox 1932, 78, pl 9.11), or a two­part scheme in which there is little division between the upper and lower bow (eg Olivier 1988, 48, fig 20.58). There is no reason to think that this, or any of the others, should be seen as being later than the general range which, as the distribution of the type in Britain shows, was passing out of use between 60-70, very few surviving to be taken north by the Roman army. There is one strand, however, which persists on the continent to give rise to a further series of designs of which Brooch 142 (see below) is an example.

Unclassified

142 Fig 72 The axis bar of the hinged pin, now missing, was housed in the rolled-over head of the bow. The upper bow is recessed for enamel, now missing, around two reserved wavy lines formed by using a punch in alternate sides of cast ridges. The top and sides of the panel are cross­cut. Under the panel is a flute and a cross-moulding. The lower bow is narrow and ends in a two-part foot-knob with a small nib underneath. FW49.

143 Fig 72 The wings, with a flat front and a cross-cut top border, housed the axis bar of the hinged pin. The upper part of the brooch is a lozenge divided vertically by a buried wavy line formed by using a punch. On each side is a triangular cell for enamel now discoloured. The outer edge of the lozenge is cross-cut. Beneath the lozenge, and under two cross-mouldings, is a zoomorphic terminal made up of a rounded end with two dimples for eyes and a projection for beak or snout, the neck has crescent-shaped stamps representing fur, scales or feathers. FW40 (Field Ill, Site 7).

144 The head and the foot are lost. The upper part of the brooch has a triangular cell for enamel, now missing, edged first by cross-cuts, then by a flute and a bordering ridge. A boss on the lower corner remains and the narrower chamfered lower bow has a cross-moulding at the top. FW72.

These three brooches were made on the continent. Although Brooch 144 is incomplete, it shows all the signs of belonging to the same school as the other two, with a semblance of the rolled-over head of the Hod Hill and a foot ending, most probably, in a foot of the same basic form as the previous two. Brooch 142 shows the clearest traces of its origin in the Hod Hill: the foot-knob, the two-part design, the cross-mouldings and the treatment of the upper part which, like the rest of the group, is enamelled, are still recognisable. Dating in Britain is poorly focused; those marked by a Z have terminals like that on Brooch 143: Colchester, Z, Trajanic (Hull 1958, 117, fig 47.3);

141

Chelmsford, Z, 140-70 (Drury 1988,94, fig 62.9); Tripontium, Antonine (Cameron and Lucas 1973, 136, fig 23b.11); Camerton, Z, before 180 (Wedlake 1958, 230, fig 53.48); Richborough, c 200 (Bushe-Fox 1932, 79, pl 10.16). The best single source on the continent is the large collection from Augst (Riha 1979) which is, however, beset with problems of residuality. The date-range there runs from the second half of the first century (Riha 1979, 157, Taf 46.1377; 160, Taf 48.1405) or specifically the last third (Riha 1979, 159, Taf 47.1397). One is dated Nero-Vespasian (Riha 1979, 160, Taf 48.1406) which is dangerously early, but should indicate a date of at least 75. Otherwise, the dating is very like the British, mainly second century and running into the third, although those that do were probably residual in their contexts.

'Strip'

The axis bar of the hinged pin in the following two brooches is housed in the rolled-under head of the bow.

145 Fig 72 The rolled head forms vestigial plain wings. The bow has a marked recurve with two cross-cuts at the point of inflection. Above these is a central flute with a sunken bead-row on each side. The foot is pointed. FW56.

146 Field-walking find from North Wraxall, Glos. The roll was extruded as a wing; the surviving one has three vertical grooves next to the bow. This is broad at the top with a swollen face between bordering grooves each of which has a line of rocker-arm ornament along it. There is a small dimple in the centre at the top.

Both brooches belong to the family known as the 'Strip', a convenient term. An iron example from Wycomb, Brooch 32, has a rolled-over he(ld and is discussed there. The type is not as simple as it appears. The several strands in its development have much to say about its initial date. First, there is the distribution: overwhelmingly in Dorset, Wilts and Soms, the lands of the Durotriges, and it uses the hinged pin so much the mark of earlier brooches of the region. The influences run from the Nauheim or its decorated derivatives, the Alesia-Aucissa sequence and the Langton Down. Of themselves, none of these parent types need be pre-conquest, except for the evidence from the Alesia-Aucissa derivatives. The writer has deliberately invoked the pre-Aucissa part of the sequence for two reasons. Firstly, the rolled-under head of the type copies the Alesia and its immediate successors; second, the decoration found on these derives from stages of development before the Aucissa had come into being. Two chief features may be noted: the reflection of feather stamps (Farwell and Molleson 1993, 88, fig 63.3; Stead and Rigby 1986, 113, fig 46.106; Gregory 1992, 126, fig 114.33) and the use of 'eyes' on the head-plate (Sharples 1991, 155, fig 130.1; Webster 1981, 61, fig 26.19; Green 1987, 97, fig 67.17). There is no need to look for an early date for the use of 'eyes' on the continent. A brooch from Baldock shows that examples had arrived by the first quarter of the first century AD (Stead and Rigby 1986, 113, fig 46.105), and the existence of the whole

142

of the continental-influenced Strip type should show that examples had arrived in the lands of the Durotriges at least as early as that. The dating available does not, naturally, emphasise enough the early beginnings, as so many more Roman sites have been excavated than purely Iron Age ones: Ower, Dors, before 25/30 (Woodward 1987a, 96, fig 52.212); Maiden Castle, three examples, 25-70 (Wheeler 1943,261, fig 84.23-5); Dorchester, Poundbury, 40-70 (Green 1987, 97, fig 66.11); Maiden Castle, 43-70 (Wheeler 1943, 261, fig 84.17); Braughing, 45-55 (Olivier 1988,46, fig 20.52); Halstock, before 50 (Mackreth 1993, 78, fig 14.11); Hod Hill, 27 examples, before 50 (Brailsford 1962, 7-10, fig 6, C15, fig 7.C30-42, fig 10, C81-7, C89, C90-4: Richmond 1968, 117-19); Gussage All Saints, four examples, mid-first century (Wainwright 1979, 113, fig 86.3016, 3026, 3061, 3066); Camerton, mid-first century (Wedlake 1958, 229, fig 53.42); Dorchester, Poundbury, mid-first century (Richardson 1940,441, fig 4.1); Stoke Abbott, 12 examples 50-60 (Webster 1960, 96-7, fig 6.9; fig 7.24, 27-9; 1965, 144, fig 6.3,4; 1981a, 61, fig 25.5,7,14; fig 26.18,19); Halstock, before 60 (Mackreth 1993, 79, fig 14.9); Camerton, two examples 55-90 (Wedlake 1958, 229, fig 55.42A, 43); Nanstallon, 60-80 (Fox and Ravenhill 1972, 92, fig 18.2); Exeter, c 75-80, and late first century (Holbrook and Bidwell1991, 237, fig 102.23-4). The very few later than 100 have been omitted: the type began very probably at the latest in the first quarter of the first century AD and had passed from use by 75/80, if not slightly earlier. South Cadbury has produced at least 45 examples, and these should surely provide yet more confirmation of this date-range.

Knee

147 Fig 72 The hinged pin is housed between a pair of pierced lugs behind the rectangular plate forming the head of the brooch. The bow, with a S profile, has two bars at the top each with a triangular section. The lower part has a thin rectangular section. The catch-plate, a simple hook, is at right-angles to the axis. FW30.

An example of the Knee brooch imported from the continent, the dating of the whole group of such imports into Britain is: Mumrills, three examples, c 140-160/5 (MacDonald and Curle 1929, 555, fig 115.2-4); Camelon, c 140-160/5 (Christison 1901, 401, fig 40); Strageath, 142-58 (Frere and Wilkes 1989, 150, fig 76.54); Caerleon, four examples, 160-230 (Brewer 1986b, 172, fig 55.16-19); Carlisle, late second/early third century (Mackreth 1990, 109, fig 101.11). Those from Newstead have been omitted: the date-range runs from 80 to c 200 and obscures the detail that none seems to date from the early second century. On the other hand, those from Caerleon giving the impression that the type could run into the early third century may also confuse because of the time-span given to their context.

Crossbow

148 The brooch had been made from sheet metal. The six­sided onion-profiled knobs have a wire band at their base. The hollow hexagonal-sectioned wings have brackets on

Excavations at Kingscote and Wycomb, Gloucestershire

their front, each three curved lobes in front. There is a dimple where the hollow bow had been which may be related to how it had been fastened. FW130.

Although only the wings survive, the brooch can be assigned to Keller's Typ 5 or 6 (Keller 1971, Abb 11) on the basis that only those two included brooches made of sheet metal. The ordering of Typen depends on the occurrence of examples in coin-dated graves (Keller 1971, Abb 12) and this works admirably for all except 5 and 6, where the evidence suggests that 5 runs concurrently with 4: 350-80. Brooch 148, on the basis of the section of the wings, in proportion to their length, may belong to Typ 6, but rather more is required before this can be asserted. The break­down in coin-dating for Typ 6 does point to a later date than for 5, but such brooches had a high status and it is more possible that, in the first instance, such brooches were not available to the classes of people buried in ordinary cemeteries in South Bavaria.

Zoomorphic

149 Fig 72 How the original pin was fitted is not entirely clear. Either it had been repaired using a bilateral spring mounted on a surviving lug of a hinged pin, or the original spring was replaced. The brooch is in the form of a duck floating on water. The head and beak are simply conveyed, each wing has a single cell, rounded at the front and pointed at the back, for mid to dark blue enamel, the back of the body is less well defined and contains a yellow enamel. FW50.

Three-dimensional representations of birds appear to be British in origin, at least when it comes to the enamelled series. The dating is: London, early second century (Wilmott 1991, 119, fig 82.307); Wroxeter, before mid-second century (Bushe-Fox 1916, 25, pl 16.13); Verulamium, two examples, 150-5/60 (Frere 1972, 118, fig 31.21-2); Colchester, late second century (Hull 1963, 145, fig 81.1). These few examples point to the second century as having been the floruit.

Unclassified

150 The hinged pin was mounted between a pair of lugs. The brooch has a curved profile and convex sides viewed from the front. There is a serrated step on each side. In the middle is a rectangular raised platform with beaded borders above and below and two small horizontal cells, probably for enamel. Above and below the platform is a series of cross-mouldings. Site 2, (146), Area 16, unphased.

Of continental origin, this particular design is probably concurrent with one which has a broader central section and narrow terminals ending in knobs which hide the catch-plate and the lugs for the hinged pin. Dating in Britain is weak: Caerleon, 110-40 (Wheeler and Wheeler 1928, 164, fig 14.19), Hadrian-Antonine (Brewer 1986b, 170, fig 54.7); Verulamium, 270-310 (Frere 1984, 29, fig 8.46). The last date is clearly inapplicable. The few dated examples from Augst suggest an earlier origin for the group, if group

The Finds

it be: Flavian-150 (Riha 1979, 193, Taf 62.1635); late first century (Riha 1979, 193, no 1632).

Plate

151 Fig 72 The sprung pin is mounted between a pair of pierced lugs. In the form of a rectangular shield, the umbo, now missing had been mounted on the central reserved circle from which issue into each half a pair of wings. Between the wings is a mid-blue enamel, while down the outside is a red. FW95.

Brooches in the shape of objects seem to be a reasonably uniform group and encompass axes, daggers, buckles, purses, flagons, lamps, shields as here and, above all, the soles of footwear. None seems to be specifically British except for a group of axes and another of shields like the one here. The dating is: Verulamium, shield, 80-150 (Wheeler and Wheeler 1936,209, fig 45.34); Chichester, sole, late first-mid to late second century (Mackreth 1989, 192, fig 26.2, 84); Quinton, Northants, plain sole, 100-70 (Friendship­Taylor 1979, 28, fig 63.478); Wilcote, Shakenoak Farm, sole, second century (Brodribb et al 1973, 108, fig 53.179); Verulamium, sole, 135-45 (Frere 1984,29, fig 9.50); Cramond, purse, c 140-200 (Rae and Rae 1974, 193, fig 14.2); Caerleon, sole, 160-230 (Brewer 1986a, 172, fig 55.14); Camerton, axe, 180-350 (Wedlake 1958, 232, fig 54.55); Dover, flagon, 190-210 (Philp 1981, 150, fig 32.71). The message is clear, such brooches are second century, but what is at issue is how many actually survive in use into the third.

Brooches 152-70 have or had their hinged pins mounted between two pierced lugs.

152 Fig 72 The plate consists of a lozenge with a pendant fantail. The lozenge has a rough surface once sealed by a decorated plate trapped under mouldings applied to the corners of the lozenge. FW51.

Two chief designs are related by the item missing from the present brooch: a blue glass cabochon 'gem'. One form is the present one, with or without the fantail, and the other is a plate with an eight-cusped margin. Both are fitted with repousse sheet to keep the 'gem' in place. On brooches like 152, the sheet is clipped under the knobs; on the cusped design, solder alone is used. The dating is: Chichester, 43-late Flavian (Mackreth 1989, 192, fig 26.2, 81); Colchester, 44-8 (Niblett 1985, 116, fig 76.41); Baldock, 50-70 (Stead and Rigby 1986, 120, fig 49.144); Richborough, 50-80 (Bushe-Fox 1949, 110, pl 25.10); Colchester, 61-5 (Niblett 1985, 116, fig 76.42); Quinton, Northants, before 70-80 (Friendship-Taylor 1974, 49, fig 18.5); Baldock, 70-90 (Stead and Rigby 1986, 120, fig 49.142); Chichester, second century (Mackreth 1989, 192, not illus); Silchester, after c 175 (Boon 1969, 47, fig 6.10). Those which belong to the second century were obviously residual in their contexts and the bulk of the rest date to before 90 and it may be that most had passed out of use by 80, having appeared immediately after the conquest.

153 Fig 72 The thin plate is shaped as a circle with an extension to hide the catch-plate. The circle has two

143

concentric grooves containing four poorly spaced ring-and­dot motifs around a circular hole. There is a projection on either side, with a double one at the top. The extension has rounded shoulders and tapers to a bifurcation divided by a groove. The brooch was tinned or silvered. FW104.

154 Fig 72 The centre is a recessed circle with a bordering groove. Around this is a series of 10 petals with rounded tops. The brooch was tinned or silvered. FW85.

Brooches 153 and Wycomb 33 belong to the same family and, although moulded relief is not really a feature of the family, Brooch 154 can safely be placed with these other two. The chief features of the family which unite a whole range of designs are the use of either bifurcated terminals, or a circular recess with a raised annulus in it, both often appearing together, as in Wycomb 33. Scribed circles, niello inlay, punched dots and the use of turned bone bosses also appear. The dating is, omitting the five later than 100: Bagendon, 20/5-43/5 (Clifford 1961, 184, fig 36.6); Bancroft, mausoleum, late pre-Roman Iron Age-late first century (Mackreth 1994, 295, fig 134.29); Colchester, 43-8 (two examples) and 43-61 (Hawkes and Hull1947, 325, pl98.170, 177, 179); Lockleys, Welwyn, Claudius (Ward-Perkins 1938, 352, fig 2.2); Longthorpe, Claudius-Nero (Dannell and Wild 1987, 87, fig 21.11); Braughing, 45-150 (Olivier 1988, 49, fig 21.70); Colchester, four examples, 49-61 (Hawkes and Hull 1947, 325, pl 98.165, 172-4); Hod Hill, three examples, before 50 (Brailsford 1962, 13, fig 1l.F2, F3, FS: Richmond 1968, 117-19); Stoke Abbott, 50..,-60 (Webster 1981a, 61, fig 25.6); Colchester, 61+ (Hawkes and Hull 1947, 325, pl 98.171); Richborough, pre-Flavian (Bushe-Fox 1949, 109, pl 25.6); Wroxeter, Flavian (Atkinson 1942, 208, fig 36.H86); Colchester, before 100 (Crummy 1983, 17, fig 14.85). The picture is clear. Made before the conquest, members of the family arrive with the army of conquest and had virtually all ceased to be used by 65/70.

155 Fig 72 The plate is a simple lozenge whose whole field is recessed for enamel or millefiore. FW72.

156 A repeat of the last, but with the lower part missing. FW106.

Because of its basically similar character, Wycomb 34 is covered here as well as these two. As vehicles only for the display of either enamel or millefiore, or both, that is the chief feature. The dating available to the writer is: Dorchester, Poundbury, first to fourth centuries (Green 1987, 47, fig 67.19); Nettleton, with first and second-century pottery (Wedlake 1982, 132, fig 54.74); Colchester, probably 150-300 (Crummy 1983, 17, fig 14.81). All in all, a pitiful showing.

157 Fig 73 The circular brooch had a tab at the top and at the bottom a circular addition with a recessed ring-and-dot having orange enamel in the dot. The main plate has two zones for enamel or millefiore, with a reserved circle in the centre, and a line of radial cross-cuts on the band between. The inner zone is empty, the outer has a surviving patch of alternating white and blue enamel or frit squares. FW91.

144

I I 11

1.170

.! :.•·

I' I:

1.171

(

I

~

Excavations at Kingscote and Wycomb, Gloucestershire

1.158

1.166

' I

1.169

~-- ~~·:;, w--lillLl)J

I;

- -r

If

( I -.

,, 11

I I I I

// ~---\./

0 50 mm --==~-==~-

Fig 73 Sites 1, 2 and field-walking: 1 Objects of personal ornament/dress: copper-alloy brooches. Scale 1:1

The Finds

158 Fig 73 The circular plate has three recessed zones around a central circular hole. The filling of the zones is indistinct, but the outer and middle ones probably had blocks of millefiore alternating with enamel, the inner may have had millefiore only. FW81.

159 The open-work design is made up of a pelta with a beaded arc below and a circular tab, filled with orange enamel, at the bottom. The pelta is recessed for two alternating colours of enamel, one turquoise, the other discoloured, and has five equally spaced bosses round the outer edge. The junction with the arc is masked with a small projection. Site 2, (296), Area 16, Period 5.

All three are imports from the continent. The third is an eclectic design, but there is no particular reason to think that it is markedly different in date from the first two. The dating in Britain for all examples belonging to the same general school recorded by the writer, and omitting those later than the mid-third century, is: Verulamium, after the late first century (Wheeler and Wheeler 1936,209, fig 45.36); Baldock, 120-50 (Stead and Rigby 1986, 122, fig 49.148); Verulamium, 125-50 (Wheeler and Wheeler 1936, 209, fig 45.37); Wroxeter, before the mid-second century (Bushe-Fox 1916, 25, pl 16.11); Camelon, two examples, c 140-160/5 (Christison 1901, 405, pl A.1,3); Cramond, c 140-200+ (Rae and Rae 1974, 193, fig 14.1); Braughing, 150-200 (Olivier 1988, 50, fig 21.72); Amersham, Bucks, 225-50 (Mackreth 1992b, 157, fig 28.10). Despite the numbers known, the dating is poor and only indicates the second century, possibly not from the beginning, with the possibility that some examples may have survived in use into the third.

160 Fig 73 A circular brooch with a concave truncated cone rising from the dished interior. The cone is topped by a recessed platform with a small boss in the middle. Around the inner edge of the dish are two engraved lines and around the periphery are six small bosses. FW97.

161-3 Repeats of Brooch 160. FW87, FW92 and FW96.

164 As Brooch 160 except for a beaded zone where the two engraved lines are. Site 1, Fll, Building 1.

165 Part of the rim of a brooch like 160. Site 2, (53), Room 9, Period 3.

166 Fig 73 A repeat of Brooch 165. FW143a. Seven, including Wycomb 35, of this British brooch in

these assemblages is a large number. The distribution fails to reveal any bias, once these are removed from the picture: had there been a concentration on this area, it should show itself by increasing numbers as it is approached, and this does not appear. The dating is: Verulamium, pre-Flavian (Stead and Rigby 1989,20, fig 12.35) and second century (fig 12.34); Wroxeter, before 120 (Bushe-Fox 1916, 25, pl 16.12); Gadebridge Park, second half of second century (Neal1974, 128, fig 55, 29); Duro-Europos, 164-256 (Frisch and Toll 1949, 40, pl 9.21). Of these, the first should perhaps be discounted until there is more evidence for a confirmed first-century date, the rest point to the second century at least.

145

167 Fig 73 A circular brooch with a cast-on loop above the pin and a circular extension below with a recessed ring­and-dot motif on it. The main plate has a wide brim, then a circular groove from which rises a dome on which there is a set of crescent-shaped cells forming a swirl. The outer part is recessed with a reserved design forming a sinuous line from which issue, on each side at the bottom, a leaf with three dimples on it. The rest of the spaces are filled with stylised grape clusters. FW100.

There is an exactly similar brooch from Manchester (unpub), and that is undated: probably second century.

168 The brooch is hexagonal with the remains of a loop above the pin and enough traces to suggest a projection on each corner. Inside the border, the centre rises to a platform recessed for enamel, now decayed, containing a reserved ring-and-dot. FW84

The writer knows of a few hexagonal brooches with raised centres as here, but not enough to form a useful distribution, and none is dated: probably second century.

169 Fig 73 The circular plate has a tall pierced tab above the pin and a projection on each side; the extension with the catch-plate is missing. The plate has, rising from a groove, a shallow dome on which is reserved a four-leaf design in a recessed cell for enamel. In the middle is a recessed annulus filled with red enamel. FW79.

The typical form of enamelled decoration on the dome is for two, or three if the diameter is large, concentric rows of narrow triangles. Alternatives consist of single rows of very long triangles, or less common versions like the present specimen, or oddities (eg Hattatt 1987, 188, fig 61.1076). Other variations consist of a wide border, with scalloped edges sometimes, bearing on occasion a row of triangles, and there is even a version which has the disc mounted as a greatly enlarged knop element under a narrow bow and a short foot. The writer has recorded two of these from Walsingham/Wighton, Norf (unpub). The distribution, even allowing for the biases in the writer's recording of brooches, seems to show the eastern side of England has more than the western side from Cornwall into Cumbria. The dating at present is: Wroxeter, c 100 (Bushe-Fox 1913, 26, fig 10.9); Carlisle, 105/15-late 120/30 (Mackreth 1990, 112, fig 101.22); Harlow, c 120-200 (France and Gobel1989, 80, fig 41.82); Ravenglass, c 130-200 (Potter 1979, 67, fig 26.2); Camelon, c 140-160/5 (Christison 1901, 405, pl A.2); Strageath, 142-58 (Frere and Wilkes 1989, 151, fig 76.57); Canterbury, mid-second century (Williams 1947,84, fig 9.1). This is a poor tally, considering the numbers known, but the picture is fairly consistent: from near or at the beginning of the second century up to the last quarter.

170 Fig 73 The circular plate has seven projections on three of which deep blue glass bosses survive. Covering most of the plate is a raised platform in which is an enamelled design. From the periphery rise eight semi­circular cells filled with a deep blue glassy enamel or frit. In the centre is a lozenge filled with emerald green enamel, the space between these having a pale blue frit. The cusped or

146

scalloped-edged central figure is found on several British brooches, although the writer has not recorded one quite like this example. The dating is almost certainly the same as for the next three items. FW82.

Brooches 171-3 have their sprung pins mounted between pierced lugs.

171 Fig 73 The circular plate is recessed for enamel, now all missing, with a reserved band made up of six circles joined by narrow arcs. In the middle is a reserved circle. FW90.

172 Fig 73 A repeat of the last, but damaged; there is midnight blue enamel in the outer zone and red in the inner. FW89.

173 Fig 73 The circular plate has a central circular cell with two concentric ones around that. The inner one has a line of reserved dots in it. The only enamel to survive is a pale blue one in the outer zone. FW80.

These three brooches are British and can have the spring mounted between two pierced lugs or one. Other patterns have triskeles, pentaskeles, stars, voided circles or single dots in an annulus, scalloped borders, and so on. What they also all once had in common was white metal trim soldered to the surface so that the bright enamel would show through a silvery beaded rim or have silvery rosettes rising from coloured fields. The technique appears to be entirely British and its use on particular varieties of Trumpet has been noted (see after Brooch 117).

The dating of the plate brooches is, including the enamelled series (indicated by E) leading to the family to which Brooches 174-7 belong: Nettleton, two examples, 69-117, coin dated (Wedlake 1982, 130, fig 54.68-9); Ewhurst, Surrey, 120-200 (Hanworth 1968, 33, fig 14.3); Watercrook, c 120-200+ (Potter 1979, 211, fig 84.15); Nettleton, E, 141-50, coin dated (Wedlake 1982, 128, fig 54.64); Overstone, Northants, before mid-second century (Williams 1976, 126, fig 13.107); Boxmoor, Herts, before 160? (Neal 1976, 78, fig 45.3); Hockwold, Norf, three examples, late second century? (Gurney 1986, 64, fig 40.5-7); Ilchester, late second century (Leach 1982, 247, fig 116.25); Dover E, c 190-210 (Philp 1981, 150, fig 32.73); Chichester, c 200 (Mackreth 1978, 288, fig 10.48,1), third century (Mackreth 1978, 287, fig 10.28, 55); Colchester, two examples, before 250 (Crummy 1983, 17, fig 14.80,82); Vindolanda, with mid­third century material (Bidwell 1985, 117-9, fig 39.3); Colchester, c 250-300 (Crummy 1983, 17, fig 14.83); Canterbury, 270-90 (Frere et al 1982, 121, fig 59.10); Verulamium, with late third century coins (Wheeler and Wheeler 1936, 209, fig 45.35); Gloucester, late third century? (Hurst 1986, 39, fig 22.4); Lullingstone, with coins of Allectus (Meates 1987, 65m fig 24.61).

The beginning seems to be in the early second century; the early part of the date-range for the one from Nettleton can be ignored. For the first time, dating covering the whole of the third century has been included: generally speaking, when there are enough dated specimens of British brooch types, a discontinuity occurs at the end of the second

Excavations at Kingscote and Wycomb, Gloucestershire

century. However, here, this does not happen. The data shows a continuous sequence through to the end of the third century before the break-down occurs. The third century can be difficult to deal with because of the poor quality of the dating evidence. Therefore, it is all the more remarkable that only for this and the next group does a consistent third-century picture emerge. The two early brooches from Nettleton should be discounted as they are so much at odds with the rest of the information. There is a cluster which suggests that the whole group or family, with its white metal encrustation, begins in the second quarter of the second century and continues into the third century, at least to its mid-point. It is worth bearing in mind that the cluster at the end of the third century may only really be there because of the tendency to place archaeological deposits perceived to be third century either early or late. Only two of the precursors (E above) of the gilded brooches in the next group have dates showing that they belong at least to the second half of the second century, but the sample is too small to draw any firm conclusions (see after Brooch 177).

Brooches 174-7 have or had their bilateral springs mounted on a single pierced lug behind the plate.

174 The plate is circular and recessed for enamel, now discoloured, with six reserved dots in it. In the centre is a riveted stud in the shape of a chess pawn. Site 2, (349), Quarry 2, Phase 4.7.

175 The oval plate has a blue glass gem set en cabochon in the middle. Round that are two zones defined by ridges and with cabling between. In the outer zone is a line of stamps. These form a relief saltire and have been half lapped. Site 2, (220), Room 5, Phase 4.1.

176 Fig 73 Small. The circular plate has an empty central zone with one zone around that containing a line of stamped saltires in relief. The front of the brooch is gilded and the back is tinned or silvered. FW98.

177 Fig 73 As 176, but oval; the cell in the middle is empty and the zone around that has two lines of stamps. The outer one is a boldly curved S and the inner two triangles touching at their apices. FW99.

Brooch 174 is placed here because of its spring system; its dating has been covered in the previous discussion. The remaining three are all developments from similar brooches and can be oval or round and with enamel in one or two zones round the central cell which was never enamelled, but fitted with one of at least three applied decorative items: a repousse copper-alloy sheet (eg Philp 1981, 150, fig 32.73), a dome or stud (eg Hunter and Kirk 1953, 59, fig 26.10), or a separate copper-alloy casting with a cell containing a paste intaglio in its top surface (eg Hattatt 1987, 252, fig 79.1207). The last two gave rise to two of the central motifs found on the gilded series (eg ibid fig 80.1220; Brooch 175) and a version of the stud on Brooch 174 is the possible origin of the most extreme decoration found in this family (eg Hattatt 1985, 151, fig 63.548). The British dating

The Finds

is: Manchester, c 160-earliest third century (Bryant et al 1986, 65, fig 5.5, 3102); Hockwold, Norf, shrine, late second to late fourth century? (Gurney 1986, 64, fig 40.8); Dorchester-on-Thames, post-Antonine (Frere 1962, 137, fig 27.8); Brancaster, with third century pottery (Hinchliffe and Green, 1985, 44, fig 28.5); Inworth, Essex, 250/60-370 (Going 1987, 81, fig 40.2); Fishbourne, 280/90 (Cunliffe 1971, 106, fig 40,43: Cunliffe 1991a, 162); Maxey, Cambs,late third/fourth century (Crummy 1985, 164, fig 111.6); Nettleton, shrine, 360-70 (Wedlake1982, 148, fig 63.5). For such a common type, the number of dated items is meagre, but does show that the family cannot be purely fourth century, a commonly held view. However, the range is so wide that all that can really be said is that the gilded group began at least towards the end of the second century and ran through the third. The two brooches from shrines should perhaps be disregarded as there is now good evidence that such items were retained as priestly equipment well into the fourth century. Examples found on the continent in datable contexts shed a little light on this fairly obscure picture: Wierden, Netherlands, a pair with a cremation in a second half of the second century pot (Van Es and Verlinde 1977, 80, fig 28.28); Esch, two examples, Netherlands, 200-50, possibly 225-50 (Van den Hurk 1977, 108, figs 25, .26); Saalburg and Zugmantel, three examples before 260 (Bohme 1972, Taf 29, 1132, 1134 and 1133 not illus); Augst, with third-century pottery (Riha 1979, 88, Taf 13, 309). These, in contrast to the British evidence, show that the type, on the continent at least, was in favour from the end of the second century until at least the middle of the third. Looking at the enamelled series again (see after Brooch 173), as there is no good reason to think that the two were made concurrently, the gilded brooches should have continued in use until at least the end of the third century, but better evidence is needed.

Penannulars

All have rings with circular sections, save for the last two, and the terminals are turned back along the top of the rings, unless otherwise described. The comments on manufacture are based on direct observation of the tell-tale seams, usually with a slightly helical twist, except in the case of Brooch 186 where the marks are basically parallel with the axis of the ring.

178 The surviving terminal is coiled at right angles to the plan of the ring. FW112.

Only terminals coiled as this, or more so, are considered here. The dating is: Gussage All Saints, first century AD at latest (Wainwright 1979, 112, fig 86.3044); Maiden Castle, 25-70 (Wheeler 1943, 264, fig 86.6); Wakerley, Northants, probably pre-conquest (Jackson and Ambrose 1978, 220, fig 57.9); Colchester, 49-61 (Hawkes and Hull 1947, 326, fig 59.2); Hod Hill, before 50 (Brailsford 1962, 12, fig 11, E8: Richmond 1968, 117-19); Chichester, Flavian-early second century (Mackreth 1978, 287, fig 10.28, 57); Colchester, second century to 250 (Crummy 1983, 18, fig 16,99); Scole, Norf, 110-60 (Rogerson 1977, 134, fig 55.12); Rudston, up to

147

late second century? (Stead 1980, 95, fig 62.20); Wakerley, Northants, probably second century (Jackson and Ambrose 1978, 220, fig 57.7). The dating shows that the type persisted through the first century into the second, but possibly not into the third. There are later examples; many belong to a distinct sub-group in which the terminals hardly form a coil, hence the comment above.

179 Each terminal has a simple flute in the middle. Site 1, Quarry (1), level30ins.

180 The surviving terminal is like the last, but has two flutes. The brooch was forged from rolled or folded sheet. FW114.

181 As Brooch 178, including the manufacturing tech­nique. FW113.

182 As Brooches 180-1, but with three flutes. Site 2, (201), Room 4, Phase 4.1.

Wycomb Brooch 36 is also considered here. There is no good evidence that the number of flutes in such terminals has any chronological significance: the brooches are so easily made that it is almost a matter of chance how many flutes are created. In the section on dating, the number of flutes is indicated in [] brackets, plain ones are marked [0]: Maiden Castle, [0], 25-70 (Wheeler 1943, 264, fig 86.7); North Cerney, [1], Glos, Claudius-Nero (Trow 1988, 51, fig 24.26); Nettleton, [0], 41-5, coin dated (Wedlake 1982, 133, fig 55.82); Bagendon, [2], 43/5-47/52 (Clifford 1961, 184, fig 36.9); Hod Hill, [1] two examples, before 50 (Brailsford 1962, 12, fig ll.Ell, E16: Richmond 1968, 117-19); Stoke Abbott, Dors [2] three examples, 50---60 (Webster 1960, 97, fig 7.21-2; 1981, 62, fig 59.13); Longthorpe, Cambs, [2][3], before 60/5 (Frere and St Joseph 1974, 46, fig 24.14-5), [3], Claudius-Nero (Dannell and Wild 1987, 87, fig 21.12); Prestatyn, [2], 70s-160 (Mackreth 1989b, 98, fig 40.28); Verulamium,[2], 80-150 (Wheeler and Wheeler 1936, 210, fig 45.39); Bancroft, [1], Bucks,late first to late third century (Mackreth 1994, 302, fig 137.53); Carlisle, [3], late first-105/15 (Mackreth 1990, 113, no 28); Leicester, [2], early /mid second century (Clay and Pollard 1994, 143, fig 74.18); Caerleon, [2], Hadrian-Antonine (Wheeler and Wheeler 1928, 166, fig 14.23); Tewkesbury, [5], 140-60 (Hannan 1993, 68, fig 19.12); Lullingstone, [0], late second century (Meates 1987, 65, fig 24.63); Old Winteringham, [0], Severan (Stead 1976,202, fig 102.36); Leicester, [2], early third century (Clay and Pollard 1994, 145, fig 75.28); Wilcote, Shakenoak Farm, [2], later third century (Brodribb et al 1971, 110, fig 47.71). There is a good cluster in the middle of the first century, which ought to mean that some are earlier than the conquest, then there is a break before a trail running to the middle of the second century, a hint of some lasting into the third, but after that, no real presence: fourth-century contexts have nothing to do with the main sequence and so have been omitted.

183 Each terminal has cut grooves, three on one and one and a half on the other. Forged from rolled or folded sheet. Site 1, between Buildings 1 and 2

148

Here is an example of what could well have been carelessness on the part of the maker: an unequal number of flutes and one at least incomplete. The dating is almost certainly the same as for the previous four brooches.

184 The terminals have a cut at each end with a concave chamfer between on each side. FW83.

185 Each terminal has two cuts at each end with a narrow cross-flute in between. Forged from rolled or folded sheet. FW103.

186 Like the last, but with a single cut at each end. Made from drawn wire, or forged from rolled or folded sheet. FW86.

187 Fig 73 Although like the last, each terminal is more carefully shaped so that there is a waist between the two cuts. Forged from folded or rolled sheet. FW101.

The terminals on these four brooches are more carefully formed. Even so, there are many examples on which it is hard to tell whether a curved chamfer on the sides of the terminal between the grooves at each end is meant (Brooch 184), rather than a proper saddle (Brooches 186-7), and Brooch 185 compounds the matter. In the following listing, A marks those which are definitely like Brooch 184, B those like Brooches 186-7, and C those which are indefinite and cover items like Brooch 185. The dating, except for fourth­century examples, for all similar styles is: Bagendon, A, B, 20/5-43/5 (Clifford 1961, 184, fig 36.8, 10); Hod Hill, A, B, before 50 (Brailsford 1962, 13, fig ll.E17, E18: Richmond 1968, 117-19); Colchester, B, 49-61 (Hawkes and Hull 1947, 327, fig 59.7); Longthorpe, Cambs, A, before 60/5 (Frere and St Joseph 1974, 46, fig 24.13); Stoke Abbott, Dors, A, 50-60 (Webster 1981a, 62, fig 25.11); Nettleton, A, probably first century (Wedlake 1982, 133, fig 55.84A); Prestatyn, A, two examples, 70s-160 (Mackreth 1989b, 98, fig 49.26-7); North Cerney, Glos, B, before mid-second century (Trow 1988,50, fig 24.25); Whitton, Glam, A, before 160 (Jarrett and Wrathmell 1981, 177, fig 71.29). Again there is a cluster in the earlier and mid-first century followed by a trail into the middle of the second century. The two Prestatyn brooches could have come from the underlying mid-first-century site, and the same might apply to the one from North Cerney, but this point is not pressed.

188 Here, each terminal is a boss with two basal mould­ings. Site 2, (339), Quarry 2, Phase 4.7.

The dating is: Maiden Castle, 150-100 sc (Wheeler 1943, 262, fig 86.2); Hod Hill, before 50 (Brailsford 1962, 12, fig 11.E2: Richmond 1968, 117-19); Prestatyn, 70s-160 (Mackreth 1989b, 98, fig 40.25); Colchester, 75/80-80/5 (Crummy 1983, 18, fig 15.95); Ilchester, early second century (Leach 1982, 247, fig 117.28); Rudston, c 140-60 (Stead 1980,95, fig 62.21); Strageath, 142-58 (Frere and Wilkes 1989, 151, fig 76.58); Vindolanda, 223-5 (Bidwell1985, 119, fig 39.12); Axminster, Devon, 225-mid-fourth century (Silvester and Bidwell 1984,49, fig 89.4); Rudston, third century (Stead 1980, 95, fig 62.23). There is a surprisingly wide date-range, ranging from the second century BC to the earlier third century at

Excavations at Kingscote and Wycomb, Gloucestershire

least. Although the Maiden Castle specimen could be dismissed as irrelevant, so little is known about penan­nulars in the two hundred years before the Roman conquest that it would be unwise to do so.

189 Fig 73 The surviving terminal has a knurled disc at the end and a fat knurled moulding next to it. FW111.

A distinctive type which is kept separate from the last one for the moment. The terminals come in two forms: plain knurled knobs and knurled knobs (A in the list) with a basal moulding (B in the list). The dating is: Skeleton Green, B, before 43 (Mackreth 1981a, 143, fig 72.58); Exeter, A, late first century (Holbrook and Bidwell 1991, 240, fig 103.35); Carlisle, B, late first century-105/15 (Mackreth 1990, 113, fig 103.27); Chichester, A, late first/early second century? (Down and Rule 1971, 81, fig 5.14.302e); Newstead, A, two examples, B, 80-c 200 (Curle 1911, pl 88.9,12,16); Prestatyn, B, c 90-135/40 (Blockley 1989,98, fig 40.24); Leicester, B, c 130-225 (Clay and Mellor 1985, 69, fig 38.7); Mumrills, A, 140-60/5 (Macdonald and Curle 1929, 555, fig 115.7); Derby, A, mid-second century? (Brassington 1980, 18, fig Se); Cappuck, A, 140-c 200 (Stevenson and Miller 1912, fig 11.2); Carlisle, B, 150-80s (Mackreth 1990, 113, fig 103.26), A, two examples, late second/ early third century (Mackreth 1990, fig 103.2.1,24); Leicester, B, early third century (Kenyon 1948, 252, fig 82.8); Old Winteringham, B, third or fourth century (Stead 1976, 198, fig 100.120); Carlisle, A, late third to mid/late fourth century (Mackreth 1990, 113, fig 103.25); Old Winteringham, B, after 320 (Stead 1976b, 201, fig 102.33). Apart from the one at Skeleton Green, which cannot be ignored, the bulk of the dating runs from the latter part of the first century to the latter part of the second, with a strong hint that some should have survived in use into the early third century. Those after 250 should have been residual in their contexts.

190 Fig 73 Although heavily worn, the chief features are certain. The ring bears distinct traces of groups of ribbing with, in a space, a trace of a pair of helical grooves. The terminals are simply shaped with a waist in the middle of each side, but the double nick in the end shows that there had been modelled ears and that the other end had been a snout. FW102.

This, and Wycomb 37-8, are patently all of the same general type. The zoomorphic terminals on Wycomb 38 are less clear on 190 and it is only the shaping of the chamfers on the side of Wycomb 37 which indicates the same kind of design. In short, there is a variety of terminal ranging from the obvious to the distinctly unobvious. In the present instances, there is the ribbing on the ring, continuous in the cases of Wycomb 37-8 and grouped on 190. Wycomb 37-8 have terminals formed by turning the ring back on itself; 190, on the other hand, appears to have the terminal formed in the end of the ring. This detail is important: it shows that typology goes out of the window as the turning back is supposedly earlier than the blunt end. On the other hand, the better the zoomorphic head, the later it should be. Again, we are in the hands of the manufacturers who may not have had such a sharp eye for such differences as

The Finds

archaeologists. The dating offered here distinguishes between continuous ribbing (C) and broken ribbing (B): Brancaster, C, fourth century (Hinchliffe and Green 1985, 42, fig 28.2); Milton Keynes, B, early /mid-fourth century (Marney and Mackreth 1987, 133, fig 41.18); Bancroft, Bucks, B, mid-fourth/early 5th century (Mackreth 1994, 302, fig 137.55); Barnsley, Glos, B, after 350 (Webster and Smith 1982, 142, fig 30.86); Birdoswald, C, c 370 and later? (Kilbride-Jones 1980, 147, fig 52.2); Wilcote, Shakenoak Farm, C, late fourth century (Brodribb et al1973, 108, fig 109.180); Feltwell, C, c 400+ (Gurney 1986, 30, fig 22.2). Not an impressive list, but one which should be considerably augmented when excavations at Wroxeter have been published. If there is a sound conclusion to be drawn, it should be that all these are mid-fourth century or later, and it is possible that continuous ribbing is entirely late fourth­century and later.

191 The ring has a thin rectangular section. Each terminal is more coiled than bent back over the ring, and each has an incised saltire between grooves cut across the ends and nicks in the sides. The surface of the ring has a series of ring-and-dot motifs with arcs between, whose intention was to create a running scroll, but the execution leaves something to be desired. Site 2, (327), Area 19, Phase 6.4.

An unusual brooch because of the combination of decoration, ring section and the slight suggestion of coiling in the terminals. The latter feature is the chronological indicator, and why the discussion of Brooch 178 dealt with more fully coiled terminals. Dating is difficult, but should become clearer when Wroxeter (Barker forthcoming) and Canterbury (Marlowe sites) have been published: Ilchester, decorated ring and saltire on terminals, before earliest fourth century (Leach 1982, 247, fig 117.32); Kelvedon, Essex, decorated ring, 300-50 (Rodwell1988, 57, fig 45.32); Baldock, groups of ribbing, fourth century (Stead and Rigby 1986, 122, fig 49.155); Barnsley, Glos, iron, plain, after 350 (Webster and Smith 1982, 144, fig 37.102); Chichester, plain, late fourth century (Mackreth 1978, 305, fig 10.39,127); Towcester, plain, 350-75/400 (Brown and Woodfield 1983, 104, fig 35.4); Verulamium, plain, fifth century (Frere 1984, 31, fig 9.55); Exeter, decorated ring, saltire on terminals, fifth century plus (Holbrook and Bidwell1991, 240, fig 103.38). All ought to be fourth century at the least, and two are dated to the fifth.

192 The section of the surviving half ring is like that of the last, but there is no clear terminal, unless there had been a cast raised area at the end. There is not enough evidence that this had been a brooch for any further discussion. Site 2 (141), Building VII (Room 17), Phase 6.3.

Fragments

There are many pieces of pin and spring, none of which adds anything useful to the collection in terms of types or dating.

149

Iron brooches By Ian Scott

Iron brooches commonly fall into three basic categories: penannular brooches, simple bow brooches formed from wire with sprung pins, and simple bow brooches with flat section bows and hinged pins. For penannular brooches see Fowler (1960).

193 Fig 74 Penannular brooch, formed from thin wire of indeterminate cross-section. The terminals are simple rolled­over loops. Only a fragment of the pin survives. Fowler Type C, dating to the first century. Diameter 43mm. Find 87. Unstratified.

194 Fig 74 Bow brooch, fragment. This is formed from wire and has a sprung pin, which is bent out of position, and a plain bow. The catch-plate is missing. Nauheim Derivative brooch. Length 67mm. Find 91. 'Outside south wall Room 1 in disturbed clay.' Unstratified.

195 Bow brooch, fragment. The brooch is formed from wire and had a sprung pin, which is missing, and a plain bow. The catch-plate is lost. Length 39mm. Similar to No 2. Find 94. Outside south wall Room 1 in disturbed clay. Unstratified.

196 Fig 74 Bow brooch with a plain tapering flat cross­section bow, a hinged pin, and triangular catch-plate. The pin is missing. Length 56mm. Find 95. Unstratified.

Beads By the late Margaret Guido

There is nothing recognisably unusual among the beads from Kingscote Site 2. The occupation does not suggest any overlap with early Romano-British lingering traditions, and there is no post- or late Roman period Germanic intrusion (as for instance at Cirencester). The main dates reflected in the assemblage indicate a period almost entirely within the third and fourth centuries AD.

GLASS

197 Translucent green cylinder. Diameter 3mm, length 9mm. Third/fourth century. Find 77/123. (349), Quarry (2), Phase 4.7.

198 Mid blue translucent cylinder. Diameter 5mm, length 12mm. Probably third-fourth century. Find 77/124. (349), Quarry (2), Phase 4.7

199 Bright turquoise translucent annular bead. Diameter 4mm, length 1.5mm. Probably late Roman. Find 77/125/1. (349), Quarry (2), Phase 4.7.

200 Green translucent annular bead. Diameter 4mm, length 1.5mm, with short fragment of fine copper-alloy wire inside. Find 77/125/2. (349), Quarry (2), Phase 4.7.

201 Annular blue bead, with white trail. Diameter 20mm, length 8-12mm. Find 77/150. (349), Quarry (2), Phase 4.7.

202 Glass bead-bright blue translucent cylinder. Diameter 5mm, length 10mm. Probably third/fourth century but

150

1.193

·""-

I

c:::==-='@ ~

1 259

~] --

~\ I I \

1.266

~ I

'0~ .. ,'. 0 ' ··.. .~

1.278 1.279

0

Excavations at Kingscote and Wycomb, Gloucestershire

'I il

!/ j

1.260 c-~--~ -~---

1

/~ /."' ~ -u

50 mm

1.280

I ~I ~~ ~'

/ I \

I

:\

1.276

-=-~= ~

// ~

1.281

1.262

c----•:f~=-;j==

Fig 74 Site 2: 1 Objects of personal ornament/dress: iron brooches, finger rings. Scale 1:1

The Finds

common type and could be a little earlier or later. Find 77/100. (352), Quarry (2), Phase 2.1.

203 Annular, bright blue bead. Diameter 8mm, length 5mm. Find 76/27. [207], Room 4, Period 4.

204 Blue glass bead, length 8mm, 7mm square. This type of faceted bead generally late Roman. Find 76/39. (234), Room 6, Phase 4.1.

205 Globular translucent bottle glass bead. Diameter 8mm, length 7mm. Not closely datable but Roman. Find 262. (51), Area 11, Phase 4.6.

206 Translucent globular sky blue glass bead. Diameter 8mm, length 6mm. Roman. Find 76/74. (219), Room 5, Period 5.

207 Four small biconical blue glass beads characteristic at Lankhills, Winchester, of fourth/ early 5th century. Diameter 6mm, length 3mm. Find 76/36. [219], Room 5, Period 5.

208 Bright translucent blue globular glass bead. Diameter 7mm, length 5mm. Roman. Find 76/71. (63), Room 8, Period 5.

209 Translucent bottle glass drop bead. Diameter 9mm, length 11mm. Not closely datable within Roman period. Find 77/116. (32), Room 9, Period 5.

210 Translucent bottle glass drop bead. Diameter 7mm, length lOmm. Roman. Find 77/122. (22), Room 9, Period 5.

211 Green glass bead. Pentagonal with pinched ends. Length 8mm, width 5mm. Common cAD 340-400. Find 77/171. (22), Room 9, Period 5.

212 Flattened cylinder glass bead, bright translucent turquoise. Length 11mm, section 6 by 3mm. Probably c AD

250-400. Find 84. (29), Room 14, Period 5.

213 Small globular bright blue glass bead. Diameter 4mm, length 4mm. Find 284a. (61), Room 12, Period 6.2.

214 Translucent green cylinder glass bead. Diameter 4mm, length 3mm. Find 284b. (61), Room 12, Period 6.2.

215 Emerald green opaque segment of cylinder glass bead. Length 18mm, section 3 by 2mm. Find 284c. (61), Room 12, Period 6.2.

216 Emerald green long hexagonal glass bead. Length 11mm, width 5mm. Third/fourth century. Find 284d. (61), Room 12, Period 6.2.

217 Bright translucent turquoise blue, green blob with traces of ?yellow ?wave. Length 8mm, section 8 by 7mm. Third/fourth century. Find 284e. (61), Room 12, Period 6.2.

218 Bottle glass bead. Badly made, probably intended to be long and square-sectioned. Length 15mm, section 8 by 5mm. Roman. Find 181. (16), Area 11, Period 5.

219 Oblong translucent ?bottle glass bead. Diameter 5mm, length 9mm. Probably second/fourth century. Find 234. (16), Area 11, Period 5.

151

220 Blue bead, roughly made, square-sectioned at one end and round the other. Length 9mm, 3mm square section. Late Roman. Find 309. (16), Area 11, Period 5.

221 Globular translucent blue bottle glass bead. Broken one end and could be part of a segmented bead. Diameter 6mm, length 6mm. Roman. Find 257. (56), Area 13, Period 5.

222 Opaque emerald green bead, square-sectioned. Length 5mm, 3mm square section. Late Roman. Find 311. (93), Area 16, Period 5.

223 Green cylinder bead. Diameter 2.5mm, length 6mm. ?Third/fourth century. Find 633. (296), Area 16, Period 5.

224 Annular translucent blue bead with white wave. Diameter 13mm, length 4mm. Very common from Romano­British to Anglo-Saxon period. Find 378a. (135), Room 15, Period 5.

225 Small turquoise green, blue semi-translucent globular bead. Diameter 4mm, length 4mm. Late Roman, third/ fourth century. Find 378b. (135), Room 15, Period 5.

226 Semi-translucent blue chip bead. Diameter 4mm, length 3mm. Probably late Roman. Find 438a. (141), Room 17, Phase 6.3.

227 Small globular translucent blue bead. Diameter 4mm, length 6mm. Probably late Roman. Find 438b. (141), Room 17, Phase 6.3.

228 Small globular translucent blue bead. Diameter 4mm, length 3mm. Probably late Roman. Find 438c. (141), Room 17, Phase 6.3.

229 Blue translucent cylinder. Length 14mm, diameter 3mm. Probably late Roman. Find 438d. (141), Room 17, Phase 6.3.

230 Small turquoise green, blue square-sectioned, semi­translucent. Length 3mm, 3mm square. Probably late Roman. Find 438e. (141), Room 17, Phase 6.3.

231 Broken translucent turquoise green-blue cylinder. Length 7mm, diameter 5mm. Probably late Roman. Find 438f. (141), Room 17, Phase 6.3.

232 Globular dark glass. Diameter 6mm, length 3mm. Probably late Roman. Find 438g. (141), Room 17, Phase 6.3.

233 Two dark green or blue glass bead fragments, flat­sectioned, oval. Length .12mm, 3mm thick. Fragments do not join. Probably late. Find 438h. (141), Room 17, Phase 6.3.

234 One tiny blue globular bead. Diameter 2.5mm, length 3mm. One square-sectioned long semi-translucent. Length 6mm, width 3mm. Probably late Roman. Find 508a. (268), Room 17, Period 3 I 6.3.

235 Globular glass bead with gold foil enclosed; perhaps broken off the common segmented type. Imported. Diameter 5mm, length 4mm. Probably late Roman. Find 508b. (268), Room 17, Period 3/6.3.

236 Translucent long blue cylinder bead. Diameter 3mm,

152

length 13mm. Probably mid or late Roman. Find 469. (136), Room 17, Period 3/6.3.

237 Small square-sectioned blue bead. Length 4mm, 3mm square. Late Roman. Find 516a. (270), Room 17, Period 3/6.3.

238 Small slightly conical dark glass. Diameter 6mm, length 4mm. Late Roman. Find 516b. (270), Room 17, Period 3/6.3.

239 Semi-translucent turquoise blue bead, long square section. Length 19mm, 4mm square section. Probably third/fourth century but unusual so not closely datable. Find 661. [262], Room 17, Period 3/6.3.

240 Flat-sectioned round bottle glass bead. Diameter 10mm, 4mm thick. Generally late Roman. Find 690a. [328], Area 18, Phase 6.4.

241 Long blue rectangular. Length 8mm, 3mm square section. Late Roman. Find 690b. [328], Area 18, Phase 6.4.

242 Translucent green bead on wire. Long diamond facets on all four sides. Length 13mm, width 2-4mm. Unusual and not closely datable within Roman period. Find 706. (332), Area 18, Unphased.

243 Small green glass bead, globular bottle glass, translucent. Diameter 6mm, length 7mm. Not datable for certain within Roman period. Find 711. (334), Area 11, Unphased.

244 Very small blue glass bead, square-sectioned, broken. Length 3mm, 3mm square section. Late Roman. Find 702. (320), Road, Unphased.

245 Semi-translucent green cylinder bead. Diameter 3mm. Length 14mm. Find 676a. (325), Unphased.

246 Semi-translucent green cylinder bead. Diameter 2.5mm. Length 13mm. Find 676b. (325), Unphased.

247 Segmented bead, translucent blue glass, with fine copper-alloy wire. Diameter 3mm, length 11mm. Find 676c. (325), Unphased.

248 Segmented bead, dark blue, black glass. Diameter 4mm, length 14mm. Third/fourth century. This type is common on late Roman sites. Find 676d. (325), Unphased.

249 Oblong blue bottle glass. Diameter 5mm, length 10mm. Probably third/fourth century. Find 686a. (326), Unphased.

250 Globular bottle glass. Diameter 6mm, length 8mm. Probably third/fourth century. Find 686b. (326), Unphased.

251 Small globular translucent bright blue. Diameter 5mm, length 5mm. Probably third/fourth century. Find 686c. (326), Unphased.

Melon beads

252 Pale blue, decayed surface. Diameter 17mm, length 12mm. Find 607. (302), Quarry (1), Phase 2.1.

Excavations at Kingscote and Wycomb, Gloucestershire

253 Pale blue-white, with very large ribs. Diameter 12mm, length 9-6mm. Find 699. (140), Area 16, Period 5.

254 Poor surface survival, very decayed turquoise blue. Diameter 22mm, length 20mm. Find 414. (141), Room 17, Phase 6.3.

JET

255 Annular jet bead. Diameter 4mm, length 2mm. Find 310. [48], (Room 15), Building Ill, Period 3.

256 Jet bead. Rectangular, with slightly concave upper face, drilled twice longitudinally. Length 12mm, width 9mm, 3-4mm thick. Cf Woodward and Leach 1993, fig 127.5, from Phase 7-8. Find 76/28. (201), Room 4, Phase 4.2.

257 Jet bead fragment. Circular with concave upper surface and flat back, drilled twice. Diameter 14mm, 5mm thick. Find 76/73. (22), Room 9, Period 5.

258 Jet fragment, very fragmentary but possibly from bead. Find 764. [328], Area 18, Phase 6.4.

BONE

259 Fig 74 Globular bone bead, lathe-turned with well­polished surfaces. Diameter about 25mm, length 20mm. Find 747. (342), Quarry (4), Phase 2.1.

Finger rings By Linda Viner

All are of copper alloy, except two of glass (nos 280-1). Most of the finger rings derive from third- to fourth-century contexts. The collection, viewed as a group, is not overtly rich. Several of the rings are of flimsy construction and form and may have been ear-rings but their condition and frag­mentary nature precludes clear identification to function, with reworking to suit secondary use a strong possibility in a number of cases.

In form, the rings are characteristic of the third century, and nos 267 and 277 reflect the typical fourth-century decorative motifs common on bracelets of the time. Parallels from Colchester for no 260 might suggest a mid­first-century date. The finger ring 280 is closely paralleled in form by a jet example from Uley.

COPPER ALLOY

260 Fig 74 Fragmentary, very fine wire band drawn out to form spiral twist. Internal diameter about 18mm, 1mm thick. Cf Crummy 1983, fig 50, 1756, dated cAD 49-50, 55, 'the design clearly has native British links'. Find 734. (345), Quarry (4), Phase 2.1.

261 Fig 74 Internal diameter about 20mm, with flat triang­ular shoulders, narrowing to band beyond a single trans­verse groove. Oval bezel with raised rim measuring 10 by 8mm, setting missing. Find 718. (341), Quarry (4), Phase 2.1.

262 Fig 74 Oval hoop, in poor preservation, 18 by 17mm.

The Finds

Blue glass setting in flat bezel. Find 77/226. (352), Quarry (2), Phase 2.1.

263 ?Finger ring. Strip of sheet metal expanding to circular bezel comprising plain flat disc of 6mm diameter. Find 289/4. (86), Area 11, Building II, Period 3.

264 Penannular finger ring, incomplete. Band with rectangular section, 3mm by lmm, flattened at shoulder and decorated with two crossbars separated by notches. (Cf Henig 1993a, fig 132.4, Structure 1, Phase 6.) Find 76/26. (201), Room 4, Phase 4.2.

265 Distorted. Internal diameter not determinable. Hoop expanding from 2mm to 7mm at bezel. Circular, oval bezel incomplete. Find 570. (51), Area 11, Phase 4.6.

266 Fig 74 Oval band of circular section with elaborate bezel comprising circular dished setting, flanked by small depressions filled with ?enamel. Further decoration on shoulders consists of shallow transverse grooves. Find 77/239. (348), Quarry (2), Phase 4.7.

267 Fig 74 Penannular finger ring of rectangular section, 2mm by lmm, tapering to plain terminals. Outer face decorated with irregularly spaced transverse grooves, between alternating notches-motifs common on bracelets. Find 77/234. (349), Quarry (2), Phase 4.7.

268 Fig 74 Plain hoop of D-section expanding to 8mm width at flat bezel. Ring setting of dished oval, 6mm by 8mm, now empty. Internal diameter about 16mm. Find 77/165. (349), Quarry (2), Phase 4.7.

269 Shoulder and part of bezel, flattened D-section Smm by 3mm, expanding to 8mm at beginnings of bezel. Find 76/79. (63), Room 8, Period 5.

270 Plain band of finger ring, of oval section at widest point of fracture, Smm by 2mm, tapering to lmm circular section. Internal diameter about 20mm. Find 77/185. (32), Room 9, Period 5.

271 ?Flat circular bezel from finger ring, lOmm diameter, 2mm thick, with vestiges of ring hoop of rectangular section. Overall ring diameter not determinable. Find 76/70/1/3. (22), Room 9, Period 5.

272 ?Finger ring. Cast strip, possible representing remains of bezel and ring setting? Length of strip 12mm. Find 77/158. (16), Area 11, Period 5.

273 Five fragments only joining to make a plain band of D-section, 1-3mm, expanding to wider section at bezel. Ring diameter unknown. Find 645. (16), Area 11, Period 5.

274 Penannular ring, internal diameter 20mm, with blunt square-ended terminals. Outer face decorated with alternating V-shaped notches. Cf Colchester (Crummy 1983, fig 50.1766, from grave deposit, dated cAD 320-450). Find 399. (141), Room 17, Phase 6.3.

275 Internal diameter c 2mm, lOmm width at circular, oval bezel. Setting missing. Find 537. (272), Area 11, unphased.

153

276 Fig 74 Bezel only of finger ring with oval glass setting flanked by transverse ribs. ?Intaglio. Cf Henig 1993a, fig 132.2, Phase 4, dated c 100-310, 17. Find 77/164. (90), road, unphased.

277 Fig 74 Penannular finger ring, employing band cut down from bracelet of rectangular section, internal diameter 19mm. Outer face decorated with six randomly spaced punched dots with raised circle. Find 77/185. (356), area north of road, unphased.

278 Fig 74 Complete ring in two fragments. Oval loop, 20mm by 22mm, with triangular shoulders projecting at an angle to the bezel. The oval contains a pale blue setting. Cf Henig 1978b, 38, fig 1, Type VIII, described as the characteristic ring of the third century. Find 363. Topsoil. Unstratified.

279 Fig 74 ?Finger ring or ear-ring. Band of metal, l.Smm square with fractured terminals. Internal diameter about 16mm. Outer edge decorated with continuous transverse grooves. Cf Crummy 1983, fig 50.1770, site clearance. Find 76/10. Topsoil. Unstratified.

GLASS

280 Fig 74 Fragment only, black glass, of flat D-section expanding towards bezel, with the outer face decorated with oval bosses. Cf example in jet from Uley: Henig 1993a, fig 133.7, Context 600, above structure XIX, Phase 8 (medieval and modern). Find 77/142. (349), Quarry (2), Phase 4.7.

281 Fig 74 Complete, black glass. Internal diameter 20mm. Flat oval bezel11mm by lOmm. Find 77/87. Topsoil. Unstratified.

Glass ring settings now divorced from ring

282 Blue glass, oval setting, with convex faces. Length 21mm, width 15mm, Smm thick. Find 77/133. (63), Room 8, Period 5.

283 Blue glass with white striations; oval, with flat base and slightly convex upper surface. Shallow cross-shaped depression in upper surface. Length lOmm, width 8mm, 2mm thick. Find 660. (320), road, unphased.

Bracelets By Linda Viner

Of 37 copper-alloy bracelets, 11 are of the common twisted cable form, of either two or three strands, with hook-and­eye terminals, in use from late first to fourth centuries. Of particular note is the visual effect achieved in 295 where varying metal composition has produced a three-coloured tone.

The majority of the bracelets are of the light bangle form classified by Cool (1993, 89) as the dominant bracelet of the fourth century, employing a variety of decorative motifs, notably alternating notches to give a zig-zag effect, and crenellations. Examples of the former group from

154

Poundbury, Dorset, suggest that this group at least may have developed in the third century.

Multiple unit bracelets, such as nos 317, 321-2 and 324, were contemporary with the light bangle type and date mainly to the fourth century, with examples from Poundbury (Cool 1993, fig 66, 18 and 19) and Lankhills, Hants (Clarke 1979, 301ff, fig 37).

The simple wire hook-and-eye bracelets, nos 285-8, are another fourth-century type. Their survival here with other similarly dated bracelet forms should be expected, despite their fragmentary nature. Other examples may be classified in the unidentified category, but survival is too incomplete to distinguish function.

COPPER ALLOY

Wire

284 Two lengths of wire, oval section 2mm by 1mm, lengths of 90 and 70mm. Find 51. (352), Quarry (2), Phase 2.1.

285 Fig 75 ?Bracelet or finger ring made of thin wire, of 1mm circular section. Distorted, but spiral twist at one end may indicate overlapping or expanding terminals. Find 77/218. (348), Quarry (2), Phase 4.7.

286 Wire, 67mm length in total, 2mm diameter. One end with slight hook may suggest original terminal. Find 76/76/3-7. (63), Room 8, Period 5.

287 Four lengths of wire, no joins, 1mm diameter. Find 400/1-4. (141), Room 17, Phase 6.3.

Two-strand cable

288 Two strands with overall diameter of 1.5mm, one strand extended to form hooked terminal. Surviving length 36mm. Find 77/223. [13], Room 9, Period 4.

289 Strands of 1.5mm circular diameter twisted together, with one strand extended as terminal hook. Surviving length 38mm, overall diameter 4mm. Find 77/223. [19], Room 10, Phase 6.1.

290 Two strands twisted, overall diameter 5mm, two lengths of 50mm and 22mm. Find 77/145. [19], Room 10, Phase 6.1.

291 Two fine wires twisted, surviving length 80mm, diameter overall 2mm. Find 532. (268), Room 17, Period 3/6.3.

Three-strand cable

292 Fig 75 Terminal only, surviving length 46mm. Two strands of circular section twisted around central core to form a cable. Remains of looped terminal, with strip of sheet copper alloy wrapped around terminal junction. Overall diameter 5mm. Find 259. (51), Area 11, Phase 4.6.

293 Overall diameter 2mm, and surviving length 20mm. Find 77/198. (22), Room 9, Period 5.

Excavations at Kingscote and Wycomb, Gloucestershire

294 Surviving lengths 13mm and 12mm. One end exhibits evidence of terminal with strand extended and twisted transversely around end. Find 279. (61), Room 12, Phase 6.2.

295 Overall diameter 4mm tapering to 3mm, surviving length 63mm. Composition of each strand appears to vary and has resulted in three-colour effect within cable. Find 382/1. (135), Room 17, Phase 6.3.

296 Three-strand originally, with one strand unwound and missing. Surviving length 30mm, overall diameter 2mm. Find 429. (139), Room 17, Phase 6.3.

297 Fig 75 Complete bracelet, internal diameter 45mm. One strand of wire with hooked terminals and two additional strands wrapped around core to give cable design. Find 500. (141), Room 17, Phase 6.3.

298 Fig 75 Surviving length 70mm. One strand extended into hook, the second wound around end, while the third is introduced after 20mm to create a diameter expanding from 2-4mm overall. Find 736. (325), unphased.

Plait

299 Bracelet or necklace. Three strands of very fine wire plaited together; one terminal survives with spiral binding. Surviving length 70mm. Find 77/221 (348), Quarry (2), Phase 4.7.

Bead-imitative (See Crummy 1983, 44, fig 46 for parallels).

300 Circular section, decorated with transverse grooves to give beaded effect. Surviving length 40mm, 2mm diameter. Find 76/8. (153), Room 1, Period 5.

301 Wire of 2mm diameter expanding to oval section of 3mm by 2mm, with one end lopped. The decoration comprises a rod of copper alloy wrapped around the wire. Surviving length 60mm. Find 77/153. (22), Room 9, Period 5.

Transverse grooves

302 Surviving length 25mm, rectangular 3mm by O.Smm, decorated with transverse grooves. Find 670/1. (325), unphased.

Transverse grooves in groups

303 Fig 75 Circular section of 2mm diameter, with transverse grooves on outer face. Bands of metal binding at IOmm intervals along surviving length. One terminal only survives, pinched flat, with rounded end. Find 109. (28), Room 15, Period 4/5.

304 Fig 75 Flat band of rectangular section, 4mm by 1mm, surviving length 26mm, decorated with groups of two transverse grooves. Find 430. (139), Room 17, Phase 6.3.

305 Fig 75 Complete bracelet, internal diameter 57mm, with overlapping terminals. Circular section decorated with groups of transverse grooves. Find 769. (141), Room 17, Phase 6.3.

The Finds

1.285

1.298 1.303 I I

- ,..

1.307

1.292 1 l

1.304 W,_L..,.II)-.. :=d!crr!§=-1]

{0, \1

/ /

t;\ - ,.,

1.305

1.316 I I

Fig 75 Site 2: 1 Objects of personal ornament/ dress: bracelets. Scale 1:1

155

1.297

-11 1.317 /I

0 50 mm

156

Crenellated

306 Rectangular section 3mm by 1mm, surviving length 42mm. Find 77/94. (70), Room 8, Phase 4.2

307 Fig 75 Internal diameter 65mm, of rectangular section 2 by 0.5mm, narrowing towards small hooked terminal. Find 77/184. (346), Quarry (2), Phase 4.7.

Alternating notches

All are distorted or too fragmentary to determine diameter.

308 Surviving length 72mm, rectangular section 3mm by 1mm. Find 76/31. (201), Room 4, Phase 4.2

309-10 Surviving lengths 80mm and 20mm, rectangular section 3 by 1mm and 2 by 1mm. Find 77 /154; 77/81/1. (348), Quarry (2), Phase 4.7.

311 Surviving length 30mm, rectangular section 3mm by 1mm. Find 382. (135), Room 17, Phase 6.3.

312 Surviving length 35mm, rectangular section 3mm by 1mm. Finds 728. [328], Area 18, Phase 6.4.

313-5 Three pieces, surviving lengths 18mm, 80mm and 30mm, rectangular section. Finds 77/85,77/196,670/3.347, road; (90) road and (325), unphased.

Miscellaneous

316 Fig 75 Flat strip bracelet, rectangular section of 4mm by 1mm, with decoration as illustrated. Surviving length 52mm. Find 77/201. (71), Room 8, Phase 4.1.

317 Fig 75 Rectangular section of 5mm by 2mm narrow­ing to 2mm by 2mm at hooked terminal, the face decorated with diagonal grooves. Surviving length 42mm. Find 76/18/1. (220), Room 5, Phase 4.2.

318 Fig 76 Rectangular section 6mm by 1mm, surviving length 20mm, with randomly placed cross-hatched lines on outer face. Find 77/168. (153), Room 1, Period 5.

319 Fig 76 Internal diameter about 50mm, rectangular section 5mm by 3mm, with outer face decorated with closely packed diagonal grooves. Find 77/90. (65), Room 7, Period 5.

320 Fig 76 Complete, with overlapping ends, and internal diameter 48mm. Rectangular section 3mm by 1mm. Decoration as illustrated. Find 148. (28), Room 15, Period 4/5.

321 Fig 76 Rectangular section 2.5mm by 1mm, surviving length 18mm, with decoration as illustrated. Find 77/95. (19), Room 10, Phase 6.1.

322 Short length of bracelet, 10mm, of D-section with cross-hatched lines on outer face behind two incised transverse grooves at the rounded terminal. Section 6mm by 4mm. Find 19/2. (61), Room 12, Phase 6.2.

323 Fig 76 'S' decoration, on short strip of 15mm, of rectangular section 4mm by 0.5mm. Cf Crummy 1983, fig 44.1700. Find 268. (82), Room 12, Phase 6.2.

Excavations at Kingscote and Wycomb, Gloucestershire

324 Fig 76 Diameter 60mm, with decoration as illustrated. Find 677. (325), unphased.

BONE

Bone bracelets are represented by seven examples, of which five have been chosen to illustrate the treatment of the terminals and/ or varying cross-sections. Three exhibit green staining suggesting copper-alloy bands were used to secure the strips. Bracelet no 330, in contrast, has a small drilled hole for either a stud or an organic tie. Clarke (1979, 313-4) in discussing 42 examples of bone, or antler, brace­lets recovered from graves at Lankhills argued for a date from the mid to late fourth century, a date confirmed by three examples from Uley (Woodward and Leach 1993, 168).

Diameters, unless indicated, have not been measured owing to the fragmentary nature of individual pieces, and the natural tendency to straighten out.

325 Bracelet fragment, plain band of rectangular section, 2mm by 2mm, surviving length 43mm. Find 172. (37), (Room 15), Building III, Period 3.

326 Fig 76 Plain band, well polished, of oval section 7mm by 3mm. Surviving length 17mm. Find 77/246. (348), Quarry (2), Phase 4.7.

327 Fig 76 Bracelet terminal, heavily stained suggesting originally bound with copper alloy. Surviving length 32mm, D-section of 3mm by 1.5mm. Find 99. (32), Room 9, Period 5.

328 Fig 76 Terminal heavily stained by copper alloy. Surviving length 16mm, of D-section 4mm by 2mm. Find 211. (38), Room 15, Period 4/5.

329 Fig 76 D-shaped section, with one square-ended terminal surviving, bound with thin copper-alloy sheet, the outer face of which is decorated with incised zig-zag motif. Length 85mm, 6mm by 3mm in section. Find 506/2/1. (268), Room 17, Period 3/6.3.

330 Fig 76 Bracelet with rectangular section of 3mm by 2mm, well smoothed and finished. One terminal missing, the second exhibits a small drilled hole. Surviving length 92mm. Find 738/1. (325), unphased.

331 Fig 76 Short section of bracelet, surviving length 20mm, rectangular section 4mm by 2.5mm. Find 738/2. (325), unphased

JET

332 Fig 76 Jet bracelet with rectangular section, and outer face decorated with three horizontal grooves. Sufficient survives to suggest slightly oval inner face and octagonal outer face. Parallels from Colchester and Silchester respec­tively would suggest late second to early fourth-century (Crummy 1983, fig 38.1568; Lawson 1975, fig 5.44). Find 76/47. (63), Room 8, Period 5.

The Finds

,,

1.321

I I 11

1.323 I I 1.324

I I

111

.

. .

I I

/ /

I ~

1330 f I

- t3 -

0 50mm ..... -=====---====--

/ /

@

'\ -

1.328

I ,,

-~-~ 0

1.331

~ .

I - -,,

I

rZ2l

Fig 76 Site 2: 1 Objects of personal ornament/ dress: bracelets. Scale 1:1

157

1.327 I I I I

-

- I

1.332

I I I I /

~ /

i I . '

I. ~ I

I I I I I I I

158

333 Fragment of jet, suggestive of bracelet but insufficient survives to indicate diameter or form. Find 263. (61), Room 12, Phase 6.2.

SHALE

The majority of the shale bracelets are plain, and mirror many of the forms outlined by Lawson (1975, 248) in his analysis of the assemblage from Silchester. Many have angular facets on the interior resulting from the method of cutting in manufacture. Three are decorated, with a fourth exhibiting grooves which may or may not be deliberate in decorative detail.

Plain shale bracelets, of circular, oval or D-shaped cross­section

334 Internal diameter about 60mm. Width 8mm, 6mm thick. D-shaped cross-section with internal ridge. Find 76/13. (234), Room 6, Phase 4.1.

335 Internal diameter 50mm. Width 6mm, 5mm thick. D­shaped cross-section. Find 76/7. (154), Room 1, Phase 4.4.

336-9 Four fragments, diameter unknown. Oval cross­section. Finds 77/244/1, 76/63, 77/244/1, 77/82. (348), Quarry (2), Phase 4.7.

340 Internal diameter about 40mm. Width 4mm, 4mm thick. D-shaped cross-section with pronounced ridge. Find 77/181. (231), Room 6, Period 5.

341 Internal diameter about 44mm. Width 4mm, 3mm thick. Oval cross-section with internal ridge. Find 419. (139), Room 17, Phase 6.3.

342-3 Two pieces, internal diameter about 46mm and 50mm. Width 4mm, 3mm thick. One D-shaped, the other with an oval cross-section, both with internal ridges. Find 576/1, 576/2. (268), Room 17, Period 3/6.3.

344 Internal diameter about 46mm. Width 6mm, 5mm thick. D-shaped cross-section with internal ridge. Find 731. [328], Area 18, Phase 6.4.

345 Internal diameter 64mm. Width 5mm, 4mm thick. D­shaped cross-section with internal ridge. Find 574. (272), Area 11, unphased.

346 Internal diameter about 40mm. Width 7mm, 5mm thick. Oval cross-section with internal ridge. Find 558. (100), unphased.

Decorated shale bracelets

347 Fig 77 Plain, D-section of 15mm by 10mm, internal diameter 80mm, with six transverse cuts on one face. Find 685. (352), Quarry (2), Phase 2.1.

348 Fig 77 Plain outer face decorated with ring-and-dot motifs. Surviving length 28mm, diameter about 50mm, 6mm by 5mm thick. Find 76/64. (22), Room 9, Period 5.

349 Fig 77 Internal diameter 50mm, of rectangular section,

Excavations at Kingscote and Wycomb, Gloucestershire

with two ribs of decoration on outer face. Original width probably 7-8mm and 3mm thick, if the crenellated rib represents the centre point of the band. Find 88. (29), Room 14, Period 5.

350 Fig 77 Rectangular section with slightly convex outer face which is decorated with four lathe-turned grooves, with two further grooves on each of the short ends. Diameter about 60mm, section 8mm by 4mm. Find 420. (136), Room 17, Period 3/6.3.

Ear-rings By Linda Viner

351 Fig 77 Both ends fractured, but thin metal and form suggest remains of ear-ring. Fine metal band of rectangular section, the outer edge only decorated with shallow transverse groove::;. Diameter about 18mm, in section 0.5mm by 1mm. Cf Allason-Jones 1989, Type 2e. Find 77/162. (350), Quarry (2), Period 3.

352 Fig 77 Penannular ring, formed from three strands of wire twisted together. Diameter 20mm. Described as Type 6 by Allason-Jones; parallel examples occur on civilian sites with a bias towards fourth-century contexts (1989, 7). Find 608. (302), Quarry (1), Phase ?2.1 or 4.5.

353 Fig 77 Fine wire bent to form loop, with second thin strand wound around first, the two interlinked to form decorative loops. Length 15mm, O.Smm diameter wire. Find 77/172. (349), Quarry (2), Phase 4.7.

354 The appearance of Type 4 (Allason-Jones 1989), but appears to be made from triangular sheet of thin metal, rolled to form tube and then curved. Length 13mm. Find 667/9. (325), unphased.

?Ear-rings, wire

355-6 A number of fine-wire fragments suggest possible use as ear-rings. In particular two pieces with both ends tapered to worked points: Find 76/25. (234), Room 4, Phase 4.1 and Find 76/12, (16) Area 11, Period 5. Cf Woodward and Leach 1993, fig 127, 14: find 4361.

Pins By Linda Viner

Pins with simple or elaborate heads are represented by 22 copper-alloy, 1 jet/shale, and 40 bone examples. In a number of instances shafts only survive, with insufficient evidence to ascribe function to either pin or needle: 17 copper alloy, 64 bone, 4 jet/shale. (For needles see p 167.)

COPPER ALLOY

Pins with heads of bead, reel and spool motifs-a type which seems to have been introduced early in the second century, and perhaps to have gone out of production in the third century (Crummy 1983, 28).

357 Fig 77 Bead, two reels, bead, reel. Tip missing, straight

The Finds

,_ ~

..__

1.360 1.361 @

-· -·

\\

159

/ 1.349 I / I I I

I , . . ·, ' . : I

-~ :'I I ® 'I'

\

1.351 1.350 '

\

1-(~ '

I -I -· 1.352

1.353 ~-~ 10 . :' ., --~ -· - ·. . . I . -· . •:

1.357 1.358 11

1.362 1.363 \\

1.365 ~ 1.366~ 1.370 ~

-· I I -·-·

1.371

-. 11

-· -. -.

11

0 50mm --====---====-...... Fig 77 Site 2: 1 Objects of personal ornament/ dress: bracelets, ear-rings, pins. Scale 1:1

160

shaft of 2mm diameter, tapering; surviving length 61mm. Find 77/215. (349), Quarry (2), Phase 4.7.

358 Fig 77 Bead, reel, bead, grooves. Surviving length 42mm, diameter 2mm. Find 303/1. (96), Area 13, Period 3.

359 Fig 77 Reels, bead, reels. Length 95mm, 3mm diameter, tapering to point. Poor surface finish, bent. Find 356. (51), Area 11, Phase 4.6.

360 Fig 77 Bead, reels, bead, groove. Complete, length 120mm, 3mm diameter. Find 562/1. (51), Area 11, Phase 4.6.

361 Fig 77 Reels, bead, reels, with conical head. Complete, length 95mm, diameter 2mm. Find 700/1. (170), Room 3, unphased.

362 Fig 77 Reels, bead, reels, with flat head. Tip missing. Surviving length 49mm, 2mm diameter. Find 572/1. (292), Area 11, unphased.

363 Fig 77 Reels, baluster, reels. Complete, length 109mm, 3mm diameter tapering to fine point. Find 705. (332), Area 18, unphased.

Pins with a faceted cuboid head; examples at Colchester have a similar date range to examples in jet and bone, all from contexts dated post AD 250 (Crummy 1983, 29).

364 Complete, length 56mm. Fine shaft of 1mm tapering to point, with separate cuboid head of 4mm. Find 76/66. (348), Quarry (2), Phase 4.7.

365 Fig 77 Complete, length 68mm. Head separate and encasing shaft of 1.5mm diameter tapering to point. Find 225. (39), Room 10, Phase 6.1.

366 Fig 77 Surviving length 61mm, shaft of 2mm diameter, with cuboid head of 5mm. Find 274. (61), Room 12, Phase 6.2.

367 Complete, length 70mm. Cuboid head of 5mm, 2mm diameter shaft tapering to fine point. Find 416. (139), Room 17, Phase 6.3.

Miscellaneous (not classifiable to types above)

368 Shaft only, surviving length 96mm, 2mm diameter, with tip and head missing. Three transverse grooves around one end might suggest pin head. Find 77/228/1-2. (352), Quarry (2), Phase 2.1.

369 Length 98mm, 3mm diameter, tapering to blunt point. Possibly not a hairpin. Find 76/57. (348), Quarry 2, Phase 4.7.

370 Fig 77 Head and shaft combined to produce complete pin, length 66mm, with cuboid head. Top surface of head decorated with small scoops to create cross design. Find 77/130. (22), Room 9, Period 5.

371 Fig 77 Pin with composite head of ?lead solder. Fine wire shaft of O.Smm diameter, tip broken. Surviving length 25mm. Vestigial remains of copper alloy on upper surface of composition with faint radiating lines from top. Find 108. (28), Room 15, Period 4/5.

Excavations at Kingscote and Wycomb, Gloucestershire

372 Head only, surviving length 7mm. Solid hexagonal tube encasing shaft. Find 212/6. (38), Room 15, Period 4/5.

373 Fig 78 Complete, length 118mm. Elaborate head comprising circular cup above five-sided reel with circular reel below. Setting probably held a glass bead or stone. Find 77/91. (64), yard, Period 5.

374 Complete, length 50mm, globular head of 3mm, on shaft of 1mm diameter. Find 428. (139), Room 17, Phase 6.3.

375 Shaft of pin, 2mm diameter, with composite (?lead solder) head, flattened oval (6mm by 4mm). Find 77/146/4. (356), north of Room 1, unphased.

376 Fig 78 Pin, complete, length 56mm, bent towards tip. Cuboid head with cuts in upper face to give cross design. Find 659. (320), road, unphased.

377 Surviving length 65mm. Shaft of 2mm tapering towards point. Head decorated with eight transverse grooves below small conical head. Shaft bent at right angles. Find 77/2. Room 17. Unstratified.

378 Fig 78 Head and shaft only. Head decorated with small conical tip, with spiral groove below. Square section of 1mm, tapering to circular shaft. Surviving length 38mm. Find 54. Unstratified.

BONE

Bone pins with a plain conical head (Crummy 1983, 20, Type 1). Examples at Colchester occur in deposits ranging in date from ?Flavian to fourth century. Uncatalogued shaft fragments were recovered from (51), Area 11, Phase 4.6 and (345), Quarry (3), Phase 2.1.

379 Fig 78 Surviving length 42mm, 5mm diameter, with straight tapering shaft. Find 9/3. (352), Quarry (2), Phase 2.1.

380 Fig 78 Tip missing, length 53mm, 6mm diameter tapering to 4mm at point of fracture. Slightly conical head, with irregular outline, but well smoothed and polished finish. Find 77/211/2. (349), Quarry (2), Phase 4.7.

381 Tip missing, surviving length 70mm, 4mm tapering straight shaft. Find 691 I 5. [328], Area 18, Phase 6.4.

Pins with one to four transverse grooves beneath a conical head (Crummy 1983, 21, Type 2). Similar to Type 1 but decorated with groove or grooves cut below the head. Type 2 pins were found at Colchester in deposits dating from the ?pre-Flavian period onwards, and a terminal date would appear to be c AD 200, with collaborative dating from deposits in York and Portchester.

382 Three transverse grooves. Complete, length 92mm, 3mm diameter, tapering to fine point. Find 77/80. (352), Quarry (2), Phase 2.1.

383 Complete, one incised groove. Length 75mm, 2mm tapering to fine point. Find 231. (38), Room 15, Period 4/5.

The Finds

.0 1.373 1

\I

1.411

~ V

1.3764}

- .

!\ J. ·I

il ~

1.378

11

1.396~ \;.:::'·'~~ ,..OZ>'

I

I I

~

@ 1.405 '' .~. ·.·

i! •I

!I

I

1.379

1:

IJ. I '

i:i 11:

'I' li· r,l

I I

~

1.404~

~

1.406 ® ... !.:

.. , :.\.~.

!~

Fig 78 Site 2: 1 Objects of personal ornament/ dress: pins. Scale 1:1

1.380~ 1.387 . ·~

·i i:

dill :I '•l I 1:1 I;

I'! :,

:! I I' ,,

·:-~ IJ

ij I I

G

I I

I I

0 50mm

1.408 ~

11

0

--=====---===:::::::.-

I I

11

il I

!I I' I

©·

1.410@ \9

I

I•

.J

I I

e)

16'1

162

384 Irregular grooves or reels below conical head. Surviving length 76mm, 3mm diameter. Blunt tip but no suggestion that this is reworked. Find 95/3. (30), Room 10, Period 5.

385 Four transverse grooves. Surviving length 36mm, diameter 4mm, with straight tapering shaft. Find 95/1. (30), Room 10, Period 5.

386 Two transverse grooves. Tip missing, surviving length 42mm, 3mm diameter. Find 529/6. (141), Room 17, Phase 6.3.

387 Fig 78 Two transverse grooves. Tip missing, surviving length 39mm, 3mm diameter. Find 691/10. [328], Area 18, Phase 6.4.

388 Three transverse grooves. Surviving length 65mm, 5mm diameter. ?Rodent gnawing around head. Find 712. (334), Area 11, unphased.

389 Two transverse grooves. Tip missing, surviving length 28mm, 2mm diameter, straight tapering shaft. Find 455. (137), Room 12, unphased.

390 Two transverse grooves. Tip missing, surviving length 72mm, 4mm diameter tapering with angular but smoothed shaft. The head appears to have undergone some reworking. Find 715. (337), road, unphased.

Pins with more or less spherical head (Crummy 1983, 21-2, Type 3, where they are further subdivided as to shape: globular or elliptical, semi-circular or elliptical, lenticular, or hemispherical). Although 137 examples are recorded from Colchester, the date of appearance and termination are vague, and a conservative starting date is offered as c AD

200, surviving to the end of the Roman period.

391 Complete, length 80mm, shaft diameter below head of 2mm with swelling to 3mm, tapering to fine point. Find 77/104/ 2. (350), Quarry (2), Period 3.

392 Small spherical head of 4mm diameter, with very marked swelling in shaft to 5mm. Extreme tip miss­ing, length 78mm. Find 77/148/2. (350), Quarry (2), Period 3.

393 Fig 78 Extreme tip only missing. Inverted cone, slightly oval in outline, 8mm by 6mm, tapering to thin waist above shaft. Swelling in shaft, surviving length 70mm. Find 76/32. (201), Room 4, Phase 4.2.

394 Small spherical head of 5mm diameter. Complete, length llOmm, with straight shaft. Find 77/170. (349), Quarry (2), Phase 4.7.

395 Fig 78 Complete, with bulbous head. Length 103mm, 8mm diameter head; 4mm at point of maximum swelling in shaft. Find 691 I 6. [328], Area 18, Phase 6.4.

396 Fig 78 Head only, angular facets with rough smoothing to form spherical head. Surviving length 25mm, 7mm diameter head on shaft of 3mm diameter. Find 77/159/2. (90), road, unphased.

Excavations at Kingscote and Wycomb, Gloucestershire

397 Complete, spherical head of 5mm diameter. Length 75mm, with swelling in shaft. Find 658/1. (320), road, unphased.

398 Spherical head of 5mm diameter, on shaft of 2.5mm, with only slight swelling. Angular facets, but smoothed. Complete, length 68mm. Find 720/1. (325), unphased.

399 Elliptical head, diameter 8mm, on shaft of 3mm, tapering to point with slight swelling in shaft. Complete, length 89mm. Find 554/1-2. (100), unphased.

Pins with a faceted cuboid head (Crummy 1983, 22-3, Type 4), with a date of appearance of cAD 250 or later. Cf examples in copper alloy and jet/shale.

400 Complete, length 76mm. Well-cut faceted head, 5mm square, length 7mm, on straight shaft. Find 77/191/3-6. (350), Quarry (2), Period 3.

401 Surviving length 40mm, head 4mm by 4mm by 5mm, on shaft of 2mm diameter with slight swelling. Find 77/129/2. (350), Quarry (2), Period 3.

402 Cuboid head, with very irregular facets. Angular shaft but well polished. Surviving length 68mm, 3mm diameter. Find 192. (42), Room 9, Phase 4.4.

403 Complete, length 65mm. Head 4mm by 4mm by 6mm, on shaft of 2mm diameter, with slight swelling in length tapering to point. Find 76/42. (218), Room 5, Period 5.

404 Fig 78 Cuboid head, 5mm by 5mm by 7mm. Surviving length 43mm, straight shaft, smoothed but knife cuts remain evident. Find 529/4. (141), Room 17, Phase 6.3.

405 Fig 78 Cuboid head, with irregular facets. Complete, length 87mm, with slight swelling in shaft which displays pale green staining. Find 506/1/2. (268), Room 17, Period 3/6.3.

Pins with one to five reels beneath a conical or ovoid head (Crummy 1983, 23, Type 5). Contexts at Colchester would suggest a date range within the fourth century.

406 Fig 78 Complete, crudely cut conical head above single reel. Oval section 5mm by 4mm tapering to point. Length 70mm. Find 77/140/1. (350), Quarry (2), Period 3.

407 Conical head, with angular cuts to create flattened sides on two opposing faces above single reel. Angular shaft with slight swelling. Surviving length 62mm. Find 77/241. (42), Room 9, Phase 4.4.

408 Fig 78 Small conical head above two reels. Angular crude cuts to form head and shaft which has a slight swelling. Surviving length 67mm. Find 77/134. (348), Quarry (2), Phase 4.7.

409 Fig 78 ?Type 5. Faceted conical head above single reel. Three half-moon cuts below reel around top of straight shaft. Complete. Length 75mm, 5mm diameter. Find 517/2. (270), Room 17, Phase 6.3.

The Finds

1.'412 I

I I

0 1.416 •

I I

- -@

1.414 ~

I

1.426

1.428a I

0

1.415 ~

I '!. !

1.430 I

I ~ 0

Fig 79 Site 2: 1 Objects of personal ornament/ dress: pins, buckles, pendants, studs. Scale 1:1

163

,,

50 mm

164

410 Fig 78 Conical head above single reel. Surviving length 30mm, 5mm diameter head, on shaft of 2mm diameter. Find 691/8. [328], Area 18, Phase 6.4.

Pins with a reel- or bead-and-reel-shaped head (Crummy 1983, 24-5, Type 6). Present in contexts at Colchester of the late third to fourth centuries.

411 Fig 78 Baluster-like head, 8mm diameter on shaft of 4mm diameter, with two incised lines around upper and lower reels. Surviving length 24mm. Find 76/24. (220), Room 5, Phase 4.2.

412 Fig 79 Head comprises flat-topped reel above bulbous bead and one further reel. Straight shaft of 2mm diameter, head of 5mm diameter. Surviving length 44mm. Find 529/3. (141), Room 17, Phase 6.3.

Miscellaneous (not classifiable to groups listed above)

413 Fig 79 Conical head decorated with incised diagonal lines, above single reel. Complete, length 100mm, 4mm diameter tapering to point with straight shaft. Find 77/129/3. (350), Quarry (2), Period 3.

414 Fig 79 Elaborate head, with incised lines around hemispherical base surmounted by two reels. Complete, length 95mm, maximum diameter 6mm. Find 182. (16), Area 11, Period 5.

415 Fig 79 Complete, length 90mm. Head of rectangular section 5mm by 3mm, 6mm long, with longitudinal grooves on each face. Well polished, 2mm diameter, tapering to point. Find 506/1/1. (68), Room 17, Period 3/6.3.

jET/SHALE

416 Fig 79 Jet pin with faceted cuboid head, 6mm square. Well-cut, circular shaft of 5mm diameter, smoothed and polished. Tip missing. Surviving length 28mm. Crummy 1983, 27, fig 24.449-53: Type 2, assigned to period from the late third to the fourth century. Find 77/169. (22), Room 9, Period 5.

Pins/needles

In addition 185 shafts were present, with or without points, with the head missing and with identification to pin or needle not possible. Of these 117 were of copper alloy; 64 were of bone and four were of jet.

Buckles By Linda Viner

In comparison to contexts at Cirencester, and given the range of personal items which are represented, there is a paucity of buckles in the assemblage from Kingscote. The collection totals eight copper-alloy, and two iron examples.

COPPER ALLOY

417 Fragment only, suggestive of buckle loop. Surviving

Excavations at Kingscote and Wycomb, Gloucestershire

length 30mm, width 20mm, of rectangular curved section. Small indent on surface suggests position of rest for buckle tongue. Find 647. (230), Room 6, Period 5.

418 Fig 79 D-shaped buckle with separate tongue. Incised decoration on upper face. Cross-bar missing. Length 18mm, width 16mm, rectangular section of 2mm by 3mm. Find 166. (38), Room 15 .. Period 4/5.

419 Fragment of D-shaped buckle in copper alloy, with iron corrosion filling the one surviving end. Width 25mm, length about 25mm, with sub-rectangular section. Find 212/5. (38), Room 15, Period 4/5.

420 Fig 79 Openwork design, ?from buckle. Surviving length 60mm, width 30mm. Square and rectangular sections within the design. Find 269. (83), yard west of room 12, Period 5.

421 Fig 79 D-shaped buckle, with loop and bar cast in one piece. Oval section to loop, with circular section for bar. Length 33mm, width 28mm. Find 396. (141), Room 17, Phase 6.3.

422 Fig 79 Fragment of circular buckle? with rectangular section. Length 75mm, width 40mm. Small nick on inner face suggests position occupied by tongue. Find 664. (322), Area 18, unphased.

423 Fig 79 D-shaped buckle. Length 20mm, width 15mm. Find 668. (325), unphased.

IRON BY /AN SCOTT

424 Fig 79 Rectangular buckle, with a loop of sub­rectangular cross-section. The simple pin is attached by a rolled-over loop. Length 53mm. Possibly not Romano­British. Find 84. Unstratified.

425 Fig 79 D-shaped, or flattened oval buckle. The loop is of rectangular cross-section. A fragment of the rolled-over loop for the pin or tongue survives. Length 39mm. Find 86. Room 4. Unstratified.

Studs or decorative mounts By Linda Viner

COPPER ALLOY

426 Fig 79 Stud with hemispherical section, decorated with alternating transverse grooves and beads. Cf Allason­Jones and Miket 1984, 3.878. Find 13/2/2. (352), Quarry (2), Phase 2.1.

427 Fig 79 Stud, triangular outline formed from grouped leaf pattern radiating from central boss. Hollow back, would suggest fixed by solder rather than nail or pin. Overall length 30mm, width 27mm, 1mm thick. Find 77/107. (220), Room 5, Phase 4.2.

428 Fig 79a,b Two studs of similar design, triangular outline, with stamped and embossed decoration. Remains of solder on reverse of each. Slight variation in size, lengths

The Finds

of 25mm and 22mm, widths 20mm. Thin sheet metal 0.5mm thick. Find 77/132, 1-2. (220), Room 5, Phase 4.2.

429 Three fragments of thin sheet metal, with embossed decoration of curvilinear lines. Find 708/1-3. (332), Area 18, unphased.

Fastening

430 Fig 79 Remains of cast oval plate with hooked end. Concave back. Incomplete, surviving length 25mm, 25mm wide. Find 625. (299), Quarry (3), Phase 2.1.

Tags

431 Tag? Hollow tube of rolled sheet metal, upper surface decorated with incised lines in imitation of plait. Length 16mm, oval 5mm by 3mm. Find 196/1. (39), Room 10, Phase 6.1.

432 Tag? Similar to 431, but solid and narrowing to blunt end. No obvious join with 431, but highly likely. Length 6mm, 4mm tapering to 2mm oval section. Find 196/2. (39), Room 10, Phase 6.1.

Iron hobnails and other shoe fittings By Ian Scott

433 Hobnails. Heavily mineralised, 12 to 15 examples. Manning Type 10. Length not measured. Find 120. (188), from alongside wall of Room 2 in fire-box, Phase 4.2.

434 Fig 85 Shoe, or boot, cleat. Lenticular cleat with incomplete arms. Length 33mm. Find 123. (15), between Rooms 8-9, Phase 4.2.

435 Fig 85 Shoe, or boot, cleat. Lenticular cleat with bent but extant arms. Length 33mm. Find 124. (261), Room 17, Period 3/6.3.

436 Hobnail with cuts forming a asterisk or star pattern on the head. Length not measured. Find 116. (400), on dry­stone wall around stoke hole, west of Room 2, unphased.

437 Fig 85 Shoe, or boot, cleat. Elongated narrow cleat with two extant arms. Length 39mm. Find 126. Unstratified.

438 Fig 85 Shoe, or boot, cleat with a circular body. The cleat has bent arms. There is a cone-shaped stud in the centre of the expansion. Length 23mm. Find 128. Un­stratified.

2 TOILET, SURGICAL OR PHARMACEUTICAL INSTRUMENTS

Nail cleaners By Linda Viner

Nail cleaners have been classified according to the typology established by Crummy (1983), with examples of Types la and 3 represented in the site assemblage.

TYPE la (Crummy 1983, 57, fig 62.1869-70). A simple type, cast, with a more-or-less straight, flat, fairly thick shaft. The

165

suspension loop rises directly from the shaft, and is on the same plane. Crummy suggests a mid to late first-century date for this type, which may survive into the second century.

1 Fig 80 Complete. Length 40mm, width 5mm, lmm thick, of flat rectangular section, pierced for suspension. Find 77/81/2. (348), Quarry (2), Phase 4.7.

TYPE 3 (Crummy 1983, 58, fig 62.1875). Plainer than Types 1 and 2, with an almost straight shaft, rectangular to square in section at the top, with crude moulding above the shaft. Probably mid to late third century.

2 Fig 80 Length 52mm, rectangular section 4mm by lmm, with transverse lines and feathering on one face. The shaft extended and looped to provide suspension loop. Find 76/48. (70), Room 8, Phase 4.2.

Fragment

3 ?Nail cleaner, fragment only of rectangular section, 7mm by lmm, tapering towards suspension end, with angular cleaning tips. Find 77/247/3. (348), Quarry (2), Phase 4.7.

Other types

4 Fig 80 Elaborate head, with moulding comprising zones of hatched cylinders below suspension loop. Oval shaft tapering to tip, one broken. Length 64mm, width 6mm. Find 280. (61), Room 12, Phase 6.2.

Chatelaine set By Linda Viner

5 Fig 80 Suspension loop and one toilet instrument surviving. Suspension loop consists of a length of wire, looped and held by rectangular clip. The nail cleaner consists of a rectangular blade extended to circular section to form loop, with end twisted spirally around shaft. Length 50mm. Find 161/1-2. (36), Room 9, Period 3/4.

Tweezers By Linda Viner

6 Fig 80 Complete, length 55mm. Parallel-sided, 3mm wide, with open spring. Find 77/112. (349), Quarry (2), Phase 4.7.

7 Fig 80 Parallel-sided blades, 8mm wide. Complete, length 55mm. An incised marginal groove runs parallel to each edge and over the loop. Cf Crummy 1983, 59, fig 63.1883. Find 39. (16), Area 11, Period 5.

8 Fig 80 Parallel-sided blades, 7mm wide. Complete, length 56mm, with four randomly placed incised circles decorating the face of each blade. Iron corrosion products survive in the loop suggesting suspension attachment. Find 675. (326), Area 18, Period 5.

9 Flared arms, 5-3mm wide. Plain, length 45mm. One arm missing. Find 77/157. (348), Quarry (2), Phase 4.7.

166 Excavations at Kingscote and Wycomb, Gloucestershire

2.2 2.1

2.5 l)

-( '' • :I /

11 I I

• I

f\ --I - I I I V V

2.6 2.7 2.8

,jl li:!IJ/1.

:::(' , .. '

~I l'' '11 ·J

I,: I I~/ .,

.w~ I I

I !lj . '11 0 ,I

I I I

1,,1'

I:,···. 0

1\ .:· 'lii\i

0 50 mm

2.13

- ~ 2.20

,, \

11

2.16 11

-· I I

11

2.19

Fig 80 Site 2: 2 Toilet, surgical or pharmaceutical instruments. Scale 1:1

The Finds

10 Parallel-sided, 4mm wide. Plain, length 48mm, one arm only survives. Find 77/179/2. (348), Quarry (2), Phase 4.7.

11 Parallel-sided, 4mm wide. Plain, complete, length 46mm. Find 76/69. (64), yard north of Building IX, Period 5.

12 Flared arms, 6-2mm wide. Plain, length 50mm. Tip of one arm missing. Find 328. (124), Room 12, unphased.

Toilet spoons By Linda Viner

13 Fig 80 ?Head of pin or toilet scoop. Flat oval plate, 3mm wide, above bead-and-reel decoration, on 2mm diameter shaft. Surviving length 41mm. Find 77/128/1. (352), Quarry (2), Phase 2.1.

14 Shaft of 90mm, 2mm diameter tapering to point, with remains of head, suggesting oval flat scoop. Find 77/114. (352), Quarry (2), Phase 2.1.

15 Head only, very corroded. Flat oval scoop, 6 by 3mm, on shaft with 2mm diameter. Surviving length 68mm. Find 344. (112), Quarry 3, Phase 2.1.

16 Fig 80 Complete, length 132mm, with small circular flat spoon of 6mm diameter. Find 77/113. (349), Quarry (2), Phase 4.7.

17 Head only, flat oval scoop 7mm wide, on 2mm diameter handle. Surviving length 58 mm. Head at an angle of 45 degrees to the shaft. Find 589. (292), Area 11, Period 3.

18 Head only of spatula, oval flat scoop 6mm wide, on oval shaft. Incomplete, surviving length 8mm. Find 688. [328], Area 18, Phase 6.4.

19 Fig 80 Toilet spoon made by rolling sheet of metal into type of 2.5mm diameter, tapering at one end to point, with flattened scoop at the other. Shaft bent. Length lOOmm. Find 623. (310), Room 12, unphased.

Razor or eraser handle By Martin Henig

20 Fig 80 Eraser handle in the form of a bust of Minerva. Her features are highly schematised, and although the crest of her Apulo-Corinthian helmet is shown, her aegis is not clearly indicated. The physiognomy is simplified and mask­like with prominent somewhat rounded eyes, long nose and slit mouth.

The bust stands on a flattened trapezoidal pedestal ornamented with horizontal lines and with a notch at the base for affixing to an iron spatula blade. Dimensions: height 52mm, width 13mm. (Previous publications: Henig 1979, 19, fig 16; Henig and Paddock 1993, 89, 91, fig 3.9).

Somewhat similar handles come from Ospringe, Kent (Whiting 1920, pl opp 65) and Woodeaton, Oxon (Kirk 1949, 40, pl iv). They have also recently been discussed in a paper by Boon (1991, especially 30-2, fig 4g-j) as razor handles and by Feugere (1995) as erasers for waxed tablets. The latter is more probable. Find 216. (28), Room 15, Period 5.

167

Mixing palettes By Alejandra Gutierrez and Fiona Roe

Two palettes were identified, neither made of very durable material, and so the working of a relatively soft substance is indicated, whether for cosmetic or medicinal use. Both fragments have the typical bevelled edges and rectangular shape of other known examples (Roe forthcoming (a)). One palette (21) represents the outcome of prestige trade, being made of a white marble of Mediterraneal'l, origin. Small fragments of white marble are difficult to identifY, ·and there are a number of possible sources. Gratitude~is-extended to Mr Michael Ballance for the suggestion that this piece might be Dokimion marble from near Afyon in Turkey. The second palette (22) is of immediately local origin, being made from Lias, available from within a few miles.

It was not unusual for Roman palettes to be made of exotic materials. A find from Charlton King's, Glos, for example, was made from Campan Vert (Rawes 1991,70, fig 16). Other fragments of marble from Dokimion, though not of the white variety, are known to have been found at Fishbourne (Cunliffe 1971, 16, no 9), and also in London (Pritchard 1986, 186).

21 Fig 81 Mixing palette, fragment. White marble. Thick­ness 8mm. Bevelled edges, weathered but no obvious signs of wear. (141), Room 17, Phase 6.3.

22 Fig 81 Mixing palette, two joining fragments. Lias. Length 96mm, thickness 12mm. Bevelled edges, very smooth and fine surfaces, slightly worn on the upper surface. (141), Room 17, Phase 6.3 and (325), unphased.

3 OBJECTS USED IN THE MANUFACTURE OR WORKING OF TEXTILES

Sewing needles By Linda Viner

The typology adopted follows that established by Crummy (1983, 65).

TYPE 1: needles with a pointed head. All are of bone.

TYPE la, with rectangular eye

1 Surviving length 60mm,- circular section of 4mrh diameter at widest point. Find 77/120. (352), Quarry (2), Phase 2.1.

2 Fig 81 Complete, length 105mm, with reworked point. Oval section 4mm by 3mm, tapering to tip. Find 77/208. (349), Quarry (2), Phase 4.7.

TYPE lb, with round eye

3 Fig 81 Single drilled eye. Complete, length 106mm, diameter 3mm tapering to point. Green staining along length. Find 77/197. (349), Quarry (2), Phase 4.7.

4 Surviving lengths .of 38mm and 24mm, probably from same needle. Tapering shaft of 4mm diameter, stained green along length. Find 77/180. (349), Quarry (2), Phase 4.7.

168 Excavations at Kingscote and Wycomb, Gloucestershire

2.22

2.21 I

-~ ~o·· .... · -1 ~

I I

~

0 250mm ~~==~---===~--

32 ~ -~ 3.3 3.5 3.7

-~ 3.13

.~ 3.15 '

, I

~ 1- d

., I

~ I. ,I

il ~ - ·=· 1:

i: !!/ :.I .. ,

~ .I

:il :I 'I' !! :!1

,. I I

I ,, I I I I I I I

I -<?2 L. I

I ~ -~ -f) --!! -~ I'

r 11 ,.

I /,

0 50 mm

Fig 81 Site 2: 2 Toilet, surgical or pharmaceutical instruments; 3 Items for textile working/manufacture. Scale 1:1

The Finds

TYPE le, figure-of-eight eye

5 Fig 81 Complete, length 108mm. Diameter 4mm tapering to very fine point. Tip stained green. Find 77/182. (349), Quarry (2), Phase 4.7.

6 Head only, fractured below eye. Surviving length lOmm, oval section 4mm by 2mm, with green staining around head. Find 694/1. (332), Area 18, unphased.

TYPE 2: needles with a flat spatulate head. An are of bone. Type 2b, figure-of-eight eye

7 Fig 81 Tip missing, surviving length 80mm, oval shaft 4mm by 3mm tapering towards broken tip. Find 735/1. (345), Quarry (3), Phase 2.1.

8-12 Five fragments where insufficient survives to assign to Types 1 or 2. All are of bone: three of which have green staining, one a well-polished finish. Three show part of a rectangular eye, one a figure-of-eight eye and one a possible figure-of-eight eye. Finds 77/104 (352), Quarry (2), Phase 2.1; Finds 77/151/1; 77/186 (x 2) and 77/233, all (349), Quarry (2), Phase 4.7.

TYPE 3: needles with a groove above and below the eye. All are of copper alloy.

13 Fig 81 Complete, length 107mm. Find 77/236. (351), Quarry (2), Phase 2.1.

14 Shaft only, surviving length 108mm, with lower section of eye only. Diameter 3mm. Find 77/118. (352), Quarry (2), Phase 2.1.

15 Fig 81 Tip missing, surviving length 140mm. Oval section 3mm by 2mm, tapering towards tip. Find 723. [328], Area 18, Phase 6.4.

Spindle whorls By Linda Viner and Alejandra Gutierrez

16 Perforated pottery disc. Base sherd of BBl dish (burnished line squiggles on underside). Diameter about 36mm, 9mm thick. Find 761. (345), Quarry (3), Phase 2.1.

17 Spindle whorl, fragment. Medium-grained arkosic sandstone, calcareous, slightly banded, with a little mica. Central hole 5mm diameter, 15mm thickness, 33mm total diameter. (121), Quarry (3), Phase 2.1.

18 Fig 81 Shale spindle whorl, fractured through centre. Diameter 32mm, 8mm thick (suggesting overall thickness of c 16mm). Groove on dished face 6mm from perforation. Cf Crummy 1983, fig 71.2002, usually dated to late third century; Lawson 1975, 272. Find 392. (141), Room 17, Phase 6.3.

19 Shale spindle whorl, complete. Diameter 35mm, 12mm thick. Biconical with finely incised ribbed faces. Unstratified.

20 Shale spindle whorl, fractured. Diameter 30mm, surviving to lOmm thickness. Lightly incised groove around perforation and two grooves surviving at maximum girth. Unstratified.

169

4 HOUSEHOLD UTENSILS AND FURNITURE

Spoons By Linda Viner

The typology presented here follows that of Crummy (1983, 69). All are of copper alloy.

TYPE 1: spoons with a round bowl. This type dates from the second half of the first and the second century.

1 Fig 82 The character of the bowl/handle junction, and sufficient of the bowl survives to suggest a circular bowl with diameter of c 30mm. Surviving length 132mm. Find 77/110. (352), Quarry (2), Phase 2.1.

TYPE 2: spoons with pear-shaped bowls. This type appears to have been in production by the first half of the second century.

2 Bowl only, 50mm long, 30mm wide, of thin sheet metal with white metal coating. Short stump only to suggest handle of rectangular section at bowl/handle junction. Find 298. (97), road, Period 4.

3 Fig 82 Complete, length 150mm, handle bent. Find 102. (29), Room 14, Period 5.

4 Bowl only, 52mm long, 32mm wide, of thin metal, with white metal coating. Very shallow profile. Fractured at handle junction. Find 520. (270), Room 17, Phase 3/ 6.3.

TYPE 3: spoons with a mandolin-shaped bowl. The type was probably produced throughout most of the Roman period.

5 Fig 82 Bowl only, with short stump of handle. Surviving length 59mm. Find 407. (261), Room 17, Phase 3/6.3.

6 Fig 82 Fragments of bowl, with flattened lip, mandolin­shaped, handle junction missing. Surviving length 30mm, maximum width 22mm. Find 346. (100), unphased.

Unclassified type

7 Spoon handle. Surviving length 118mm, circular section of 4mm near bowl/handle junction, merging to square section of 2mm near pointed tip. Find 77/111. (352), Quarry (2), Phase 2.1.

8 Bowl of spoon, fragment of thin metal, probably from Type 2. Surviving length 20mm, width 26mm. Find 131. (28), Room 15, Period 4/5.

Hook

9 Possibly a fragment from a lamp hook, with oval section and blunt tip. Cf Crummy 1983, 80, fig, 84.2114; and complete examples quoted in Loeschcke 1919, Taf 2 and 21; Wilson 1968, pl53, 203; and Hawkes and Hull1947, pllOO, 33. Find 76/49. (201), Room 4, Phase 4.1.

170 Excavations at Kingscote and Wycomb, Gloucestershire

4.1 4.3 4.5 I I

1

\ I ' -· ),

,I

·-

0 50 mm

4.10 ...

l I\

fr c

Fig 82 Site 2: 4 Household utensils and furniture. Scale 1:1

The Finds

Furniture attachment

10 Fig 82 Moulded plaque or attachment, with flat back and fractured end. Surviving length 36mm, overall width 36mm, 4mm thick. Find 77/141. (352), Quarry (2), Phase 2.1.

Bindings and fittings

This section encapsulates fragments of sheet metal which have been deliberately pierced, suggesting use as bindings or fittings to boxes or pieces of furniture. The fragility and form of the fragments would preclude use as structural fittings within the buildings. Cf Crummy 1983, 85-9 for discussion of box fittings.

See p 190 for fastenings and fixings which may be associated with bindings and fittings.

11 Three fragments of thin sheet metal. Find 289/5-7. (86), Area 11, Period 3.

12 Plate of sheet metal, 30 by 27mm, pierced in four corners with rivets to secure second sheet of metal having irregular fractured borders. Find 76/9. (171), Room 3, Phase 4.4.

13 Strip of copper-alloy sheet, of U-section, surviving length 22mm, width 6mm, with sides of 4mm. Find 709/2. (339), Quarry (2), Phase 4.7.

14 Distorted sheet metal, with holes punched through on four outer margins, parallel to edge. Length 63mm, width 55mm, 0.5mm thick. Find 77/237. (349), Quarry (2), Phase 4.7.

15 Fig 82 Thin sheet metal plate, with three raised edges, and one hole for nail or rivet. Maximum width 38mm, surviving length 46mm, 0.5mm thick. Find 77/210. (199), Room 4, Period 5.

16 Fig 82 Rectangular sheet, 0.5mm thick, length 90mm, width 34mm, pierced by six holes. Find 314. (16), Area 11, Period 5.

17 Fig 83 Square plate 1mm thick, 22mm square, pierced in each corner, with remains of iron stud in the centre protruding from one face. The upper surface of the iron corrosion is partially coated with thin copper alloy, and on the reverse two short flanges of copper alloy are folded flat against the plate. Find 264/1. (61), Room 12, Phase 6.2.

18 Fig 83 Square plate 1mm thick, 60mm square, with small tag on one side. Hole pierced in centre with scar of stud/boss suggesting diameter of stud about 12mm. One face highly polished. Find 264/2. (61), Room 12, Phase 6.2.

19 Fig 83 Thin sheet metal strip as illustrated, pierced with iron stud. All edges appear finished despite corrosion in places. Cf Crummy 1983, 85, fig 90.2175-76: winged lobate corner fittings from First Butt Road Box from grave 69 c 320-450. Find 266/8. (61), Room 12, Phase 6.2.

20 Fig 83 Thin sheet fragment with hole for nail or rivet. Surviving length 24mm, width 8-9mm, 0.5mm thick. Find 266/8. (61), Room 12, Phase 6.2.

171

21 Rectangular sheet folded in half and pierced with remains of copper-alloy stud. Length 44mm, 30mm wide. Find 380. (135), Room 17, Phase 6.3.

22 Thin sheet metal fragment, with raised rim and ridge along one edge. Lead solder survives on the reverse within the grooves. Surviving length 72mm, 26mm wide. Find 721/1/2. [328], Area 18, Phase 6.4.

23 Fig 83 Flat sheet plate, length 63mm, 35mm wide, 1mm thick, with sub-rounded corners. Two holes pierce plate, with copper-alloy rivet in situ in one. A second rivet survives loose, and is formed from thin sheet metal, rolled into tube and the head flattened. Length of rivet 12mm. Find 411. (146), Area 16, unphased.

24 Angled sheet, with hole pierced in one arm. Length 15mm, 8mm wide, 0.5mm thick. Find 358/2. 100, unphased.

25 Sheet fragment with irregular edges, and ridge across centre. Surviving length 52mm, width 32mm, 0.5mm thick. Find 358/3. (100), unphased.

Ring-keys

CfWoodward and Leach 1993,173, fig 132.10-12 and 14 for six examples from Uley; Crummy 1983, 84, fig 89.2163-70, and First Butt Road Box fig 90.2195.

26 ?Ring-key fragment. Find 605/8. (298), Area 16, Period 5.

27 Fig 83 ?Ring-key fragment. Find 282. (61), Room 12, Phase 6.3.

Bone inlay

Thin strips of bone, often decorated with incised geometric patterns, were used principally as casings for wooden boxes or caskets. Groups of inlay strips from boxes are illustrated in Henderson 1949, pl 58 and Wilson 1968, pls 61-2. A complete box of decorated bone is known from Heilbronn (Goessler 1932, 294). Larger items of furniture could also be decorated with bone inlay (Nicholls 1979; Crummy 1981).

Assorted shapes

28 Fig 83 Fragment of inlay, with flat back and convex upper surface decorated with ring-and-dot motifs, which survive in poor condition due to burning leaving the surface powdery and grey in colour. Surviving length 22mm, 10mm wide, 4mm thick. Find 77/161. [13], Room 9, Phase 4.4.

29 Fig 83 Two fragments, probably from same object. Flat back with convex upper surface, decorated with incised curvilinear design. Surviving length 18mm, width 15mm, 5mm thick. Find 390/3-4. (82), Room 12, Phase 6.2.

30 Fig 83 Complete, as illustrated. Length 22mm, width 16mm, 4mm thick, with cancellous tissue on base. Find 377. (137), Room 12, unphased.

172

4.18

-·-

0 50 mm

4.28

4.30 I 4.32

4.42

Fig 83 Site 2: 4 Household utensils and furniture. Scale 1:1

Excavations at Kingscote and Wycomb, Gloucestershire

4.341 4.351

-~-'. I

E?ZI

4.27~ /( ~

4.39

I

e2:2

4.40

-I-I

£2,g

- - - - __________________ ___J

The Finds

Plain rectangles and squares

31 Six square inlay pieces: four are 7mm square, 3mm thick; one is 9mm square, 5mm thick; and the sixth is fragmentary measuring 5mm by 3mm on two sides, 2mm thick. Find 76/72/1-6. (220), Room 5, Phase 4.2.

32 Fig 83 Square inlay, flat face, bevelled sides. 12mm square, 4mm thick. Find 180/2. (16), Area ll, Period 5.

33 Fig 83 Rectangular flat face, 17mm by 8mm, 7mm thick, wedge-shaped on reverse. The surface is charred white­grey with crazed appearance due to burning. Find 427. (139), Room 12, Phase 6.2.

Strips with ring-and-dot motifs

34-8 Fig 83 Five strips (two illus), with ring-and-dot motifs on upper face, from (63), Room 8. All are fractured and there are no obvious joins to allow matches to establish the overall length of individual strips. All have a flat upper face, with square-cut sides, and have a powdery white-grey colour due to burning. Finds 77/189/7, A9/12, 189/8, 189/10, 189/ll. (63), Room 8, Period 5.

Strips

39 Fig 83 Surviving length 22mm, 5mm wide, 2-3mm thick. Parallel sides and one end, square-cut edges to all faces, flat top decorated with diagonal incised grooves. Grey in colour. Find 77/189/7. (63), Room 8, Period 5.

40 Fig 83 Incomplete, surviving length 20mm, surviving width 7mm, 2-3mm thick. Powdery white. Surface decorated with longitudinal groove, and five grooves at right angles to the edge. Find 77/189/14. (63), Room 8, Period 5.

Circles

41-8 Fig 83 A group of seven discs (three illus) from (63), Room 8, in association with other pieces of inlay from the same context, would suggest their function as inlay rather than use as counters. All have irregular or semi-regular knife-trimmed edges, flat upper surface with deeply incised ring-and-dot motif. The backs in most cases show cancel­lous tissue, with variable thickness within each individual piece. An eighth piece from (22), Room 9 has similar signs of manufacture, but is decorated with a double ring-and­dot motif. Some are grey or black in colour with a powdery texture due to burning. Finds 77/189/1, 189/2, 189/3, 189 I 4, 189/5, 189/8, A9, (63), Room 8, Period 5; Find 69. (22), Room 9, Period 5.

Triangles

Plain triangles

49-54 Fig 83 Six examples of triangles (one illus), with straight knife-cut sides. White or grey in colour. Finds 77 /189/3A9, 189/4A9, (63), 189/6A9. (63), Room 8, Period 5; 307/1/4,307/2/1,307/2/4. (61), Room 12, Phase 6.2.

173

Ring-and-dot motif on triangle

55 Fragment of triangle, two cut edges, surviving length 18mm, surviving width 4mm, 2-3mm thick. Surface decorated with ring-and-dot motif. Find 77/189/9. (63), Room 8, Period 5.

Grooves on triangle

56 Triangle, length 20mm, 8mm wide, 3mm thick, decorated with two parallel grooves. Find 77/189/16. (63), Room 8, Period 5.

57-9 Three further examples. Finds 77/189/20, 189/17, 189/5 A9. (63), Room 8, Period 5.

Incised line parallel to borders of triangle

60-7 Fig 83 Seven examples (two illus) from (63), Room 8, and one from (61), Room 12. Finds 77/189/4, 19, 2., 3, 9, ll-12 All. (63), Room 8, Period 5; 307/1/5. (61), Room 12, Phase 6.2.

Wedge-shaped inlay

68 Knife-cut bevelled edges, with groove on face. Extreme tip missing. Surviving length 26mm, width 13mm. Find 307/2/6. (61), Room 12, Phase 6.2.

Heart-shaped inlay

Plain, with straight point

69-75 Fig 84 Seven examples (two illus), one from Room 8, the remainder from Room 12. Finds 77/189/22. (63), Room 8, Period 5; 307/2/2, 3, 5, 7, 745/1, 2. (61), Room 12, Phase 6.2.

Plain, with curved point

76-8 Fig 84 Three examples (one illus), with the point curving to the right. Grey or black in colour due to burning. Straight knife-cut edges, no bevel. Finds 77/189/21, 189/23, 189/24. (63), Room 8, Period 5.

79-80 Fig 84 Two examples (one illus), with the point curving to the left. Grey-black due to burning. Find 77/189/25, 189/17 A9. (63), Room 8, Period 5.

With incised crescent decoration

81-5 Fig 84 Five examples, (two illus), powdery white or burnt black. Find 307/1/1, 3. (61), Room 12, Phase 6.2; 77/189/1 and 6 All, 189/8, /10. (63), Room 8, Period 5.

With assorted decorative motifs

86 Fig 84 Heart-shaped, extreme tip missing. Charred and blackened surface. Compass double ring-and-dot decor­ation on face. Length about 23mm, width 17mm, 2-4mm thick. Find 77/189/14 A9. (63), Room 8, Period 5.

87 Fig 84 Heart-shaped, straight cut edges, with incised line as decoration outlining shape. Length 26mm, width

174

14mm, 3mm thick. Pale green staining in parts, on a white burnt bone. Find 307/1/2. (61), Room 12, Phase 6.2.

Unclassifiable, but probably heart-shaped

88 Inlay fragment, probably from the tip of a heart-shaped fragment, white powdery surfaces and very decayed. Surviving length 20mm, width 17mm, 2-3mm thick. Find 77/189/15 A9. (63), Room 8, Period 5.

89 One curving knife-cut edge to fragment which is probably heart-shaped. Burnt black. Surviving length 20mm, width 13mm, 2-3mm thick. Find 77/189/16 A9. (63), Room 8, Period 5.

Ring

90-3 Four fragments representing segments of rings, two with incised decoration. Grey or powdery white upper surfaces, burnt black on undersides. Finds 77/189/18-19, 15, 2 A9. (63), Room 8, Period 5.

Wedge-shaped strips

94 Five fragments, 2mm wide, of 20mm, 18mm, 15mm (2), and 12mm lengths; 3-4mm deep. Find 77/189/9-13 A9. (63), Room 8, Period 5.

95 Thin slithers, 21 fragments in total, lengths vary from 5-10mm, 1-2mm wide. All ends are fractured. Find 77/189/5 All. (63), Room 8, Period 5.

0

4.81 4.82

Fig 84 Site 2: 4 Household utensils and furniture. Scale 1:1

Excavations at Kingscote and Wycomb, Gloucestershire

96 Nine fragments, lengths of 10-13mm, 1.5-2mm wide, with depths of 4-5mm. Colour variation from powdery white, grey to dark grey. Find 77/189/13-21. (63), Room 8, Period 5.

97 Fig 84 Thin strip of bone, with wedge-shaped section. Length 70mm, width 2mm, depth 3-5mm. Find 307/2/8. (61), Room 12, Phase 6.2.

98 Length 62mm, 1-2mm wide, 6mm depth, with small patch of green staining. Find 307/1/6. (61), Room 12, Phase 6.2.

99 Two fragments, lengths 8 and 30mm, 1mm wide. Find 390/1-2. (82), Room 12, Phase 6.2.

IRON OBJECTS BY IAN SCOTT

Bucket fittings

100 Fig 85 Possible bucket handle mount, consisting of strip of thin rectangular cross-section, pierced by one countersunk nail hole. The end narrows and is curled into a loop or eye. Length 84mm. Find 64. Clay south of Room 3. Unstratified.

101 Fig 85 Possible bucket handle mount, consisting of strip of thin rectangular cross-section. Little survives of the plate. The end narrows and is curled into a loop or eye. Length 53mm. Find 65. Unstratified.

I I

4.79

50 mm

:1! 4.87

' n

4.97

I; I I -

The Finds

I

~ ~

1.434

I

~ 1.437

I I

@ I

f1 1.438

,,J. 11

7.16 7.19 7.20

\' - ~

- ! -.

. / •

11

•·:Ji:' ~ G~· . . -'

;

11

11 I

- ' -· \ - \ -·

11

I

~

I \ I rJ

IT '0- ~ n -0

r 7.21 !

1\i

-11-· I

li

I 8.1 •

11 1

7.13 :j 11

-1-·

j 1/

f: {'j- i

~ '

7.14 ' ' 11

- I -e

11

I '

7.15 ' Jl

11

0 100mm --==--===---.

-I

I 7' \~~;/ ,,-8.2

'

/

I -~·-. --~-~.,& , ...

8.4 =

Fig 85 Site 2: Ironwork: 1 Shoe-fittings; 4 Iron household items; 7 Styli; 8 Horse equipment. Scale 1:2

175

176

102 Fig 85 Bucket, or tub, handle mount, consisting of a crescentic plate with a centrally placed hook of square cross-section. The plate is pierced by two square nail holes. Length 80mm. Compare the larger example from Borough Hill, near Daventry (Manning 1985, 103-4 no P28). Find 66. Unstratified.

Candlestick

103 Fig 85 Socketed candlestick. The candlestick has a split tapering socket and four feet. The legs radiate out from the base of the socket and are of a sub-rectangular cross­section. The feet are flat and diamond-shaped. Length 100mm. See the example with tripod feet from Verulamium (Manning 1972b, 178 and fig 65.51). Find 67. Unstratified.

Lamp

104 Fig 85 Possible lamp hanger. Consists of a square section bar with twisted sections, and a hook or spike set at right-angles. Length 138mm. The twisted decoration indi­cates that this fragment is part of an implement used in conjunction with fire. It may be a lamp hanger, or part of a flesh hook. Find 69. Unstratified.

STONE OBJECTS BY ALEJANDRA GUTIERREZ AND FJONA ROE

Mortars

105 Fig 86 Mortar, fragment. Flat-bottomed bowl with a smoothly worn interior. Purbeck marble. 55mm height, 220mm diameter. (316), unphased.

Purbeck marble is well known as a material used for Roman mortars, and also for other items, including palettes (Dunning 1968). Other Roman Purbeck marble mortars have been found, for instance, down the Fosse Way at Shepton Mallet (Roe forthcoming a) and at Caerleon (Zienkiewicz 1993, 122).

Discs

Seven sandstone discs were found in Rooms 3, 9, 12 and 17. They are sub-circular in shape and 40-lOOmm in diameter. Six of the discs are made from Pennant sandstone, a fairly coarse-grained, micaceous red-grey sandstone which could have been obtained locally, probably from Rangeworthy, 12km or so away. The smallest disc (112) is made from a finer-grained stone which is possibly Old Red Sandstone. All of these discs could have been made from pieces of reused roofing tile.

There are a number of very similar Pennant sandstone discs from Wortley, Wotton-under-Edge, (Taylor and Bagnall 1992, 21, fig 10.212-14; 1994, 19), and also from Frocester (Price forthcoming). Suggested purposes include use as pot lids or weights, but the most feasible explanation may be that they were throwing discs, used in a game of skill.

106 Disc, 12-14mm thickness, about 85mm diameter. (171), Room 3, Phase 4.4.

Excavations at Kingscote and Wycomb, Gloucestershire

107 Disc, 17mm thickness, about 90mm diameter. Rough surfaces and edges. (36), Room 9, Period 5.

108 Fig 86 Disc, llmm thickness, about 100mm diameter. (36), Room 9, Period 5.

109 Disc, 15mm thickness, about 54mm diameter. Rough surfaces and edges. (61), Room 12, Phase 6.2.

110 Disc, 15mm thickness, about 80mm diameter. Rough surfaces and edges. (268), Room 17, Phase 6.3.

111 Disc, 11mm thickness, about 50mm diameter. (323), Area 18, unphased.

112 Fig 86 Disc, 8-10mm thickness, about 40mm diameter. (325), unphased.

Querns

Four fragmented upper stones from rotary querns were retained. Photographic evidence shows that other more complete stones were found, notably two large millstones and one smaller quem from Room 17 (Fig 27). The only type of dressing visible is the radial grooving on the grinding surface. The majority of the retained fragments are too small to reconstruct their original dimensions.

The stone used is the same for all the pieces, Upper Old Red Sandstone from the Forest of Dean, or the west bank of the River Wye. A number of querns made from the same sandstone and conglomerate occurred as loose finds on the site (Eagles and Swan 1972, 85).

Upper Old Red Sandstone quartz conglomerate and the associated sandstone were widely used for querns in Gloucestershire and surrounding areas during the Roman period. Local sites where Upper Old Red Sandstone has also been recorded include the nearby Wortley Roman villa (Taylor and Bagnall1990, 26, fig 13.158; 1992, 21, fig 10.215) and West Hill, Uley (Roe 1993, 199, fig 148).

113 Rotary quem, fragment. Upper Old Red Sandstone. 45-37mm thickness. Upper stone with traces of radial grooving on the edge. [19L Room 10, Phase 6.1.

114 Rotary quem, fragment. Upper Old Red Sandstone. Upper stone with radial grooving on edge and horizontal raised surface around exterior. (61). Room 12, Phase 6.2.

115 Rotary quem, fragment. Upper Old Red Sandstone. Upper stone with radial grooving; well worn. (61), Room 12, Phase 6.2.

116 Rotary quem, fragment. Upper Old Red Sandstone. 32-20mm thickness. Upper stone with traces of radial grooving near edge. (139), Room 17, Phase 6.3.

Whetstones

Thirty fragmented or complete whetstones were recovered, but only sixteen were securely associated with the main buildings. The rest were either from undated quarry fills, topsoil contexts or later phases. Their distribution over the