Newsletter 23 - January 2002 - Roman Finds Group

IucernaRoman Fi nds Group Newsletter 23ContentsRomandecoratedironstyli ...... 2Wax spatula handle from Yorkshire . ! . . 5NextRFGmeeting , . . . . . ! . . . ,.8Gold amulet-pendant from Shropshire . . 9Romanwatgrworks . . . , r . r . , . 10Bronzevessel fromMiddlewich . .. 11Ivory knife-handle from Silchester,, . . 12Nomoremystery?.. . r r . r . . . 13Boxappeal . . . . . r r . . . ! . r L4Bookreviewr . ! . . . r . . . . . . . 14FRGTOO-1700meeting. . r. . . . . 16Spindlerwhorlrpot. , . , , . . ,., L7The autumn meeting at Exeter . . . . 28Crosswordrby'Diggler'..,..... 32Confgrences.,rr!rt...r,33Booksetc. . r r ! . . . . . . . . t .34Infofmation for contributorsPost your contributions to Nina Cruffiffry, 2Hall Road, Copford, Colchester CO6 1BN, oremail them to nina.crummy@ntlworld.com,Deadlines May 30th for inclusion in the lulyissue, and November 30th for the Januaryissue.Emailed text should be either a .txt, .rtf, or.doc file. Illustrations should preferably besimple line drawings or uncluttered photos.Emailed pictures should be .tif or .jpg files.Remernber not to breach copyright law whensending illustrations.EditorialTttis issue is very full, and I am actremely gratefulto all the contributors fur their hard work.Writing eguipment has been the subject of muchstudy on the continent in recent years, and now arema*able group of s$li ftom Elms Farm,Heybridge (a site which has produced manyexceptional small frnds), has provided thestimulus for Hilary Major's typology of decontedstyli in Britain,There is a good crop of small articles, and a longcontribution from Bavaria. Though this is a studyof some unusual deposiB in two late ?ffi-centurygrav6, the piece has major implications reachingback into the Roman period. The paper is anobject /esson in avoiding preconceptiofrs, andexamining pot contents arefully.For your light relief and enteftainment during apeaceful tea-break, this issue includes the first ofwhat I hope will be a senes of ctosswords.'Digger's' archaeologial-historical crosswordshave been a feature of 'Essex Past and Prsent'for some time, so apologi* to Essex r*identswho may have long memori* for clues and theiranswers!Our new Meetings Co-ordinator, Ellen Swifr, hasbeen working overtime, ananging venu6 for thisautumn and spring 2003, as well as completingarrangemenB for Segedunum this spring. If youwould like to host a meeting, or have ideas for atopic, let her know. A booking form forSegedunum is included in this issue, and if youwant to know the full programme, do contactEllen by email, snail mail, or phone. Contaddetails on p 34.Ellen would be grateful if you could find the timeto frll in her questionnaire also enclosed. Sheaims to present the findings at a conference atDurfiam, Promoting Roman finds in context andtheory, ifi the autumn.Richard Hobbs, our Secretdry, will shortly bemoving departments within the British Museum.He is leaving Portable Antiquities, where he madea tremendous cantribution to the prajed, to go toPrehistoric and Early Europe, replacing C.atherineJohns. She will be sadly missed, but may thenhave more time to write articles for Lucerna.Nina Crummy

lu*rna -Iariuary !#$fRoman decorated iron styliManning (1985, 85) produced a baslctypology of Roman iron styli which has beenwidely used, and is based princlpally on thedegree of separation of the point and stem.All decorated styli were classed as type 4, avery broad and heterogeneous group. It isclear that the classification of decorated styliwlll bear refinemert, and the following is apreliminary note on an attempt to group anumber of published examples. Styli are theonly class of Roman iron object which areregularly embellished with moulded bandsand inlay in other metals, and consequentlyare potentially more closely datable thanrnost other ironwork, particularly if differentgroups can be identified and dated. It is alsodifficult to imagine a village blacksmithmaking the more elaborately decorated styli,and they surely must have been aspecialised product, produced at a limitednumber of workshops, and possibly animport from the continent.The impetus for this study was the finding ofa relatively large number of styli at ElmsFarm, Heybrldge, Essex. Thirfy sryli werefoutrd, all made of iron, of which thirteenwere decorated in various ways. Theyinclude one of the rnost elaboratelydecorated examples from this country (fig l,i). The site, which was excavated in 1gg3-95, and funded by English Heritage andBovis Hornes, was a Late lron Age andRoman rural settlement, with a centraltemple (see Atkinson and Preston lgg8 foran interim report). The Late Iron Age siteseerns to have been of some importance, buta slow decline appears to have set in by thesecond century. This is not the sort of sitewhere one would expect to find largenumbers of styli, especially the rnoreelaborately decorated ones, which must havebeen valued (and valuable?) possessions.The reason for the presence of so many styliat a relatively low status slte is not,however, considered here. I d0, however,think it is unlikely that they were being madeon the site.COI{STRAITTTS Oil THE STUDYIron styli of any kind may be difficult torecognise, as they tend to resemble nailswhen covered in concretion. Any decorationis usually hidden by corrosion, and itsrecognition is dependent on identificationfrom X-rays. None of the decorated stylifrom Elms Farrn were recognised as suchprior to X-ray. Few reports specificallymention X-rays, particularly older findspublicatioils, and in some reports the objectshave obviously been drawn without the useof X-rays, implying that decorated styli mayhave been rnissed during cataloguing.Published drawings of styli are oftenunsatisfactory, since they are usually at 1:Z(the normal reduction for ironwork), makingit difficult to see the details of thedecoration. The text doesnt always clarifythe drawing a lt ffi;ty, for exampl€, mentioninlald bands, but not how many. Withoutseeing the original objects, some styli canonly be provisionally grouped. Howev€r,given the tendency of ironwork todisintegrate during storage, I suspect thatnot all decorated styli survive, particularlythose from older excavations.THE RANGE OF DECORATIVETECHNTQUES EI|PLOYEDThe styli from Elms Farm employed a varietyof decorative techniques which incorporatemost of the repertoire found on styll. Theyfall lnto three main groups: moulded bands,incised decoration, and apptied non-ferrousmetals. There is also rare use of stampeddecoration.The moulded bands (the term 'moulded'does not here imply casting) can takevarious forms, from simple beading topolygonal sectioned bands (stylus handlesare normally circular in section). Three ofthe Elms Farm styli had bands with polygonalsectiofis, some with non-ferrous appliedbands, a feature which has been rarelyrecognised elsewhere. It is not somethingwhlch can be easily seen on X-rays, and canin most cases only be identified duringcleaning. It was therefore probably morewidely used than is evident from thepublisfied examples.Incised decoration most commonly takes theform of one or more circumferential grooves.other incised decoration includes spiralgroov€s, fine rilling, cross-hatchlfig, andobligue lines. Grooves may originally havecontalned non-ferrous inlay, so lt is, perhaps,rnisleading to differentiate between exampleswith inlaid grooves and those with groovesbut no inlay.

-r-a\! llghFlrAabcFig. 1 Group 1 a) Gestingthorpe b) Caister-on-Sea c) Elms Farm; Group 2Group 5 g) Gestlngthorpe; Group 6 h) Elms Farm; Group 7 i) Elms FarmeSOmmd) Harlow; Group 3 e) London; Group 4 Q Verulamium;il Gorhambury Group I k) London, St. Magnus House.kE \Ja,h.,iIsr)trrH

i#rry##' )dnuarTo !##lApplied non-ferrous metals were mostcommonly ln the form of wire, or narrowstrlps, set into a groove. The applied metalsfrom the Elms Farm styli were analysed, andproved to be exclusively brass. Virtually allknown examples are described as 'copperalloy', the exception being one fromBrancaster which has both brass strips andsllver lattice (Group 5; Sparey Green andHinchliffe 1985, 53, no77).T}IE GROUPSThe grioups are based on a sample ofseventy-six styli, from thirty-four sites. Thesample includes four styli which I consider tobe dubious, and sixteen styll which were notgrouped, elther due to incompleteness orlack of similar styli. Eight groups weretentatively identified (and may yet besubject to change!), based on the form ofthe stylus, and the style of decoration. Thecriteria used for the form were the method ofseparation of the point and handle (as withManning's typology), the general shape ofthe handle (i.e. straight or bulbous), and, tosome extent, the shape of the eraser.Unfortunately, the erasers are oftendamaged or missing, so although I considerthat the eraser shape may be relevant (andis one of the particular features of Group 2),it has generally been afforded lessimportance in this scheme than otherfeatures,It should be emphasised that, on the whole,the groupings are rather loose, and do notnecessarlly imply that the styli aFecontemporaty, ot were made in the sameplace. In a few cases, however, styli fromdlfferent sites are similar enough to suggestthat they did come from the same workshop.GROUP I (Fig 1, a-c)This is the most coherent gtuup, linked bythe unusual form of the point, which isstrongly separated from the handle, andbulbous. The erasens are semi-clrcular orstraight-sided where present. Five exampleswere identified, including hvo from ElmsFarm; the others are from Caister-on-Sea(Darling 1993, 101, no 422), Gestingthorpe(Manning 1985a, 51, no 200) andMagiovinium (Neal 1987, 52, no 106). Threeof the examples are similar enough tostrongly suggest that they are the productsof the same workshop (Ftg 1a-c; from ElmsFarm, Caister-on-Sea, and Gestingthorpe).The only securely dated example ls earlyRoman (Elms Farm). The bulbous point isalso present on a stylus from London wlth acopper alloy handle and iron polnt (Manning1985, N25). The handle is decorated wlthincised cross-hatching and polygonalsectioned panels. It ls prcbably from theWalbrook, and thereforc likely to be lst to2nd-century in date, suppofting an earlyRoman date for the type.The decoratlve motifs present include rilllng,spiral and circumferential gnooves,sometimes inlaid with wire, moulded beads,polygonal-sectioned bands and mouldeddecoration.GROUP 2 (Fig 1, d)Styli of group 2 are characterised by theshape of the eraser, and the style andpositioning of the decoration. The eraser iswaisted, and the decoration consists almostexclusively of low beaded elements withflanking grooves or slight ridges. Sixexamples were identlfied, from Bancroft vllla,Brough-on-Humber, Harlow, Lullingstone andRichborough. Where datable, they are frcmLate Roman contexts. Waisted erasers areprobably a late Roman form.GROUP 3 (FIg 1, e)The three deflnite and one probable memberof this group are typlfied by their long,slightly splayed erasers, and single reels atthe base of the handle. The three definlteexamples are all in the British Museum, andwere seen by Manning as a dlstlnct group,his type 4a (Manning 1985, N28-N30). Theprobable example is from Baldock, and isslenderer, with a shorter splayed eraser.The reel is present, but not flanked by agroove. One of the lncomplete Elms FarmsWli probably had a splayed eraser, andcould belong to this group. None of theexamples is dated.GROUP 4 (FIg 1, f)This group features seml-circular eraserc,with a handle that tends to swell sllghtlytowards the point, but not invariably. Therels sllght separation of the point and handle.The decoratlon consists principally of groovesat the Junction of the point and handle,sometimes with similar decoration below theeraser. The grooves are sometimes inlaidwith non-ferrous bands. Eight exampleswerc identlfled, five of which came fromdated conte)cs (although only one was4

luceme Janua*,2#$Jclosely dated), and a 3rd-century date seemslikely for most of the group. Although theyare not simllar enough to suggest a commonsource for all of them, most are from EastAnglia/Herts, including three from varioussites in Verulamium, which suggests thatthey are an eastem type, even lf not all fromthe same workshop.GROUP 5 (Fig 1,9)Thls is a somewhat disparate group in termsof decoration and eraser shape, but all havedistinctly differentiated points. The junctionbetween the handle and point is sharplydefined, with the point forming a splke set inthe mlddle of the end of the rylindricalhandle, which is sometimes slightly flared atthe point end. The decorative techniquesused are variable, including polygonalsectionedbands on one example fromLondon, and circumferential grooves,sometimes inlaid. The group includes theabove mentioned stylus from Brancaster,with a silver lattice inlay. It is unlikely thatthe group ls the product of a singleworkshop. Dating ranges from early 3rd to4th century, wlth one example from MaidenCastle (Wheeler L943,286, no 2) exhibitingthe waisted eraser typical of group 2, and ofpossible 4th century date.GROUP 6 (Fig 1, h)The members of thls group, while somewhatslmilar to Group 4, have parallel-sidederasers, and a copper alloy bead (or amoulded lron bead intaid with copper alloy)between the handle and point. Copper alloybands are present on the handles. Only fourexamples were identlfied, two fromAldborough (Bishop 1996, 34, nos 195 andL97) and two from Elms Farm, one of whichis very similar to the Aldborough styli. Thesecond is rather more elaborate,incorporatlng a polygonal-sectioned paneland crimping on the bead, and is possiblynot a member of this group. Both the ElmsFarm styli are from dated contexts: the one(fig l, h) more closely resembling theAldborough styli is late 4th century, whereasthe more elaborate ocample is from an earlyRoman context.GROUP 7 (Fig 1, I-r)This group, of which there were nineexamples, has a swelling at the junctionbetween the handle and point, morepronounced on some examples than otherc,with an undifferentiated point. It lncludes themost elaborately decorated stylus flom ElmsFarm, which has the handle entirely coveredin decoratlve elements. Most, however, havesimple decoration consisting ofcircumfercntial grooves, often rcund thethickest part of the handle (le at the Junctionof the handle and point). Five are fromdated contexts, from the second to 4thcentury, and there are possibly at least twosub-groups nepresented, mid-Roman andlate Roman.The Elms Farm stylus (Fig L, i) ls anexceptional piece of craftsmanship, thoughnow in poor condition, and cracked, wlth theeraser missing. The decoration consists ofbands of applied circumferential brass strips;three moulded polygonal-sectioned segmentswlth triangular facets pointing alternately upand down, wlth an applied strip of brassround each segment; and three panels ofapplied lattice made from very thin brassstrips. It ls unclear whether the lattice wasmade up before application to the stylus, orwhether it was laid strip-by-strlp dircctlyonto the iron. It came from a pit flll dated tothe mid 2d to mid 3d century.Group 8 (Fig I' k)The group has straight slded erasens, slenderhandles and undifferentiated points. Thedecoration ls slmple in all cases, consisting ofone or more grooves and/or inlaid bands.Nine examples were identified, plus twopossible examples. Of the five datedexamples, four are from 2nd-century, orposslbly 2d-century, contexts.To conclude, this preliminary look at thegrouping of decorated styli has .alreadyproduced some evidence for the o

ucerna Januery J##;'ReferencesAtkinson, M, & Preston, S, 1998 'The Late Iron Ageand Roman settlement at Elms Farm, Heybridge, Essex,excavations 1993-5: an interim report', Bribnnia 29,85-110Wheeler, R E M, L943 Maiden C-astle, Dorset Rep ResComm Soc Antiq London 12Bishop, M C, 1996 Finds from Roman AldboroughOxbow Monograph 55Darling, M J, 1993 Caister-on-Sea: exavations byCharles Green 7957-7955, E Anglian Archaeol 6OManning, W H, 1985 Abbgue of the Romano-EritishIron Tools, Fittings and Weapans in tfie British l{useumManning, W H, 1985a 'Ironwork' 46-58 in Draper, J.1985 Fscavations by-Mr. H.k C-oopu on tfie Roman siteat Hitl Farm, Cratingtlwrpe, Essex, E Anglian Arciaeol25Neal, D S, 1987 'Excavations at Magiovinium,Buckinghamshire, 1978-8A', Recards of Buclcs 29, I'I24Sparey Green, C & Hincfiliffie, J, 1985 'The small findsfrom the L974 and L977 $(ctvations' ln Hinchlift, J.Exavations at Brancaster 1974 and 1977 E AnglianArchaeol 25, 4I-62WAX SPATULA HANDLE FROM YORKSHIREOn the Portable Antiquities Database there isan entry for a metal-detector find fromBeadlam, near Helmsley, ifi North Yorkshire.Though described as a knife-handle in theform of a helmeted bust of Mars, the objecthas been so clearly described that it canreadily be identified as the ninth Minervabust wax spatula handle from Britain:'Inform of a helmeted bust with roundedshoulders. The head is rounded, while thebody and slot are flat, apart from a heartshapeddecoration on the breast. The face isworn, the nose completely eroded, with oneeye visible. At the bottom, there is a slot totake a blade. Length 45 mm.'Unfortunately, no image was made of theHelmsley handle when it was report€d,though its description matches the handlesillustrated here perfectly and it is also theright size. The heart-shaped decoration isthe aegis worn by Minerva and instantlyidentifies this image as that of the goddessrather than Mars, as originally supposed onthe database.The late George Boon suggested that theseMinerva bust handles were from razors(1991, 30-2), but recent work has shownthat they, and other related handlesdescribed by Boon, are from spatulae usedto smooth over sectiotrs, or whole pages, ofwriting on wax tablets. Wax spatulae havebeen the subject of considerable study onthe continent over the past few years, withthree main forms, A-C, identified by MichelFeugEre (1995). Type B splits into twosubtypeS, and type A into five, of whichthose with Minerva bust handles formsubtype A5 (Fig 3).0 1cm,,tFig 1: A5 handle from the Vicars Farm site, Cambridge.Drawing by Andrew Hall. Copyright CambridgeUniversity Archaeological Unit.Fig 2. Complete A5 wax spatula fuom Ospringe, afterWhiting 1923. Copyright M Feugdre. Scale in cms.

lucema January 200!FeugEre's initial research produced a list ofeight examples of variant A5 from Britain(1995 , 326). Unfortunately, one was aduplicate. (I am grateful to Jenny Hall forclearing this up.) To the remaining seven canbe added a recent find from Cambridge (Fig1), and now the Helmsley example. (AMinerva bust from Cheshire, cited in lackson& Potter 1996 as a parallel to the handlefrom Stonea, is of unusually thin curvedsection and more likely to be an appliqu6.)The revised British list is as follows:1. Kenchester, Herefordshire. Boon 1991, 3t, fig4, h; FeugEre 1995, 326, no 1.2. Kingscote, Gloucestershire. Henig & Paddock1993, 89, fig 3, 9; Feugbre 1995, 326, no 2, ftg 4,d.3. Woodeaton, Oxfordshire. Kirk L949, 40, no 2,gl 4, F; Pitts L979, 96-7, pl 2L3,218; Boon 1991,fig 4, j; FeugEre 1995, 326, no 3, fig 4, b. (Theremay be a second from Woodeaton: Kirk L949, pl4, E; Pitts L979, pl 23, 2L7; Boon 1991, 32,footnote 76.)4. Silchester, Hampshire. Toynbee 1954, 81; PittsL979, 97, no ZLg, pl 28; Feug6re 1995, 326, no4, fig 4, c.d,,5. Museum of London 12845, accessioned 1932,acquired earlier, Prcvenance unknown. Undgren1978, 96, pl 68; PitB L979,95, no 215; FeugEre1995, 326, no 6 (and duplicated as no 5, fig 4, d).6. Ospringe, Kent. Whiting L923, 66, pl oppmite65; Whiting et al 1931, pl 55, httom; Dunning1958, 16, pl 3; Boon 1991, 30, fig 4, gi Feuglre1995, 326, na 7, fig 2, a. Here Fig 2.7. Stonea Grange, C.ambridgshire. futter &Jackson 1982, fig 3, 4; Jackson & Potter 1996, figLLz,95; FeugEre 1995, 326, oo 8, fig 4, e.8. Cambridge, Cambridgeshire. Crummy 2001.Here Fig 1.9. Helmsley, Beadlam, North Yorkshire. PortableAnUquiUes Scherne: Finds ID yorymm4O4.In Britain the best-preserved A5 spatula, itsiron blade intact apart from some damagealong the edges, came from a grave atOspritrg€, Kent (Fig 2). Others from graveson the continent have been found withinkwells and other writing eguipment, suchas styli, and it is probable that at least someof the graves also originally contained writingtablets, long since decayed. The datingevidence suggests that A5 handles belong to,illi. lii.!:. 1:' ii. . r. t l'l'._ !.. ,i, l, l! l,,'A5 . .:,,,,'Hffijllirtf.r, r,i 1,'t iiijiti'ii ll'I :'Ir:iii i.ii'.,ll'''irll,Fig 3, Classification of wax spatulae.Copyright M Feugere.

iutema -]dnuary .?##lthe 2nd century, perhaps running into theearly 3d century.The Yorkshire A5 handle is the mostnortherly known in Britain, and it may beassociated with the occupation of the villa atBeadlam. The other examples are scatteredacross the south, from London (probably),Cambridge and Stonea Grange in the east toKingscote and Kenchester in the west. Thenumber from Britain compares to one eachfrom Holland, Austria, Hungary, Italy,Portugal, Romania and Morocco, four fromBelgium, eight from France, and ten fromGermany (FeugEre 1995, 326-331).Complete spatulas such as that fromOspringe are evidence for literary, and theuse of the image of Minerva on a writingutensil points to her function as goddess oflearning. There is, however, a sub-text tothese handles. Detached from its iron blade,the handle may take on an independent lifeas an image of the goddess in any of heraspects, of which perhaps the mostimportant may be as goddess of the healingarts, the Roman Minerva Medica, or in Britainperhaps conflated with a local goddess, egSulis Minerva at Bath. The two Minerva busthandles (probably only one from a waxspatula) found at the temple at Woodeatonwere probably votive offerings, as may havebeen the Cambridge find, which came from asmall settlement with monolithic postsettings dating to the 3d and 4h centuries.I would be grateful if anyone recognisingfurther e:

j' tt cg;ll E Jgn i,,ari. -ffrEA gold amulet-pendant from EatonConstanti ne, ShropshireIn July 2000, a small gold object was shownto Mike Stokes, of Rowley's House Museutn,Shrewsbury, by its finder, Mr Niall Mennice.Mr Stokes immediately recognised it as atubular Roman amulet-case pendant. Theobject was duly reported to the local Coroneras reguired by the 1995 Treasure Ad, aninquest was held in December 2000, and thependant was found to be treasure. It hassince been acquired by Rowley's HouseMuseum and the finder has received theappropriate reward.The object is a simple cylindrical tube of golddesigned to hang horizontally from its threesuspension loops on a chain or ribbon. Thecircular plates closing the ends of the tubeare very neatly soldered, BS is the lengthwisejoin of the tube. There is no decorationother than the lightly reeded surfaces of theloops. The object has been damaged bycrushing. Overall it is 2.5 cm in length, hasa diameter of 6 ffiffi, and weighs 4.4grammes. The metal is of extremely highpurity; semi-guantitative X-ray fluorescenceanalysis carried out in the Department ofScientific Research, British Museum, revealsa gold content of 99 per cent.Gold amulet-case pendant from Eaton Constantine,Shropshire. Length 25 mm. Photo: British MuseumTubular amulet-cases, often containing amagical inscription or a substance creditedwith supernatural power, have a very longhistory. In the Classical world, they occurredfrom Hellenistic to late-Antique times, butsimilar objects were also used in MiddleKingdom Egypt, in India, medieval Persia,and various Islamic and Jewish contexts upto post-medieval times and, fio doubt,many other times and places. The idea ofwearing an item of jewellery that containssomething of power or of special significanceto the wearer is still current in the wearing oflockets enclosing a picture or other memento(eg a lock of hair). The Graeco-Romanamulet-case combines that concept with theage-old protective power of a religioussymbol, the modern equivalents of whichinclude crucifixes, Stars of David, and thelike. A Romano-Egyptian portrait of the latesecond century which shows a young boywearing a gold amulet-case on a plaitedleather necklace is illustrated in thecatalogue Ancient Faces (Walker andBierbrier L997, no 94)There is a late-Roman gold amulet-case inthe Thetford treasurc, deposited at the endof the fourth century AD or the beginning ofthe fifth (lohns & Potter, 1983, no 30), andthe catalogue discussion includes a fullsurvey of the type, including a wide selectionof parallels. An earlier, but still valuable,discussion is to be found in the catalogue ofthe T6nEs, Algeria, treasure (early sthcentury AD) (Heurgon 1958, 57-9). As faras I am aware, until the discovery of theEaton Constantine pendatrt, the Thetfordexample and one from York (MacGregorL976, fig 8, 72) were the only specimensrecorded from Britain.In the absence of any association orcontextual information, close dating of theEaton Constantine amulet is difficult. Thereis no decoration to provide stylistic evidencepointing to earlier or later Roman taste, butthe relatively small size of the object may initself hint at an early Roman or MiddleEmpire date rather than one in the fourthcentury. Unfortunately, none of the small,decorated examples appears to have anyarchaeological context, but several of thelarger specimens do have associations datingthem to the third century and later, possiblysuggesting some rudimentary typologicalprogression based on size. Barbara Deppert-Lippitz assigns a late Hellenistic date to avery small (13 mm) amulet-pendant withgranulated ornament (Deppert-Lippitz 1985,no 35), though on stylistic considerationsalone, The Thetford amulet is nearly 40 mmlong, and some other examples from hoardsof irie 3rcr-5th centuries AD elsewhere in theRoman Empire are considerably larger. Oneof the two in the T6n&s treasure (Heurgon

i*cerna -jdnua*, J8#:1958, Pl V, no 5) has a length of 72 mm andweighs 2O g.The finder of the Eaton Constantine pendantand his associate also showed Mr Stokesother Roman objects found in the samegeneral area (within a 50 m radius of thefindspot), and these, including a sesfe rtiusand a dolphin-type fibula, werepredominantly 1$-to-2nd-century obiects. Onthis evidence, a date in the znd century mightbe tentatively suggested for the pendant.The high-purity gold is Wpical of goodRoman jewellery at all times and has nochronolog ical implication,The Eaton Constantine gold amulet-case isan interesting item of Roman jewellery and itis a most unusual find from Britannia. Thenew Treasure Act and the PortableAntiquities Scheme have improvedcommunication between metal-detectoristsand local archaeologists and museumcurators, and the fact that an importantsingle item such as this one is now safely ina museum collection, where it will remainavailable for study, is a credit to allconcerned.Catherine JohnsDept of Prehistaric and Early EuropeBritish MuseumLondanReferencesB Deppert-Lippitz 1985 Goldschmuck der R0merzeiti m RO m isch -Germa n ischen Zentra I m useu m (Bonn)J Heurgon 1958 Le Tr&or de Tdnis (Paris)C lohns & T W Potter 1983 The Thetford Treasure;Raman jewellery and silver (London)A MacGregor L976 Finds from a Roman sewer systemand an adjacent building in Church Street, York (York)S Walker & M Bierbrier 1997 Ancient Faces; mummyportraiB from Roman Egypt (London)Roman waterworksRecent excavations at 30 Gresham Street inthe City of London have proved particularlyrewarding. Earlier this year, a gilded bronzearm, possibly from a life-size statue of theemperor Nero, was discovered thrown into astagnant pond that had been filled by AD70.Now two examples of Roman water-liftingsystems have been excavated that allowedwater to be scooped up from the bottom ofdeep wells in a continuous loop, known inthe ancient world as a bucket chain.In summer 2001, two large deep oak-linedwells were excavated, still waterlogged afternearly 2000 years. The first smaller well,constructed in AD63, had been in use for lessthan ten years when the superstructure ofthe well collapsed. A number of woodencontainers were found at the bottom restingin a half-barrel, which may have acted as asump. These would have been linkedtogether to form a continuous bucket chainoperated either by a human treadmill oranimal powered. There are three relativelycomplete containers, hollowed from solidoak, with thin nailed lids and fragmentaryremains of about nine others. A narrow slothad been cut into one side to allow eachcontainer to fill with water (each box onlyhaving a capacity of about two litres). Theslots appear on opposing sides of thesurviving examples, suggesting that onewheel had perhaps turned two bucket chainssimultaneously. An oak roller, which mayhave been part of the gearing system, wasalso excavated from the well. This deviceclearly represents part of a new integratedsystem following the destruction of the townby Boudica in AD 60l6t.A remarkable collection of ironwork andcharred wood was discovered in the secondlarger well, constructed in AD 108/9 anddestroyed by fire some years later. Itconsisted of a series of sturdy angled ironbars and flat straps, linked by substantialpins, which once formed a continuous chainto hold about twenty buckets. The conditionof the ironwork is unbelievable and some ofthe uncorroded links are still articulated.Fragmentary evidence from the bucketsindicates that each was a rectangular boxnarrowing slightly at the opening. The ironchain would have revolved slowly around adrivewheel, powered by a human treadmill.The water supplied may have been for use atthe Roman amphitheatre, just north of thesite or at the small Cheapside baths, to thesouth, where a wooden container was alsofound during the 1955 excavations but notidentified. Finds from this well included alarge bronze cauldron, which had survivedintact but dented, a complete Highgate pot,stamped roof tiles and complete box-fluetiles.These are the most complete examples to bediscovered outside the Mediterranean andseem to be the finest-preserved throughoutthe Roman world. Time Team are producing10

,' { if *I- I}d J*,'l uarir "iiS$r-ra documentary about the site at GreshamStreet which will be screened some time inthe future. As a joint project between theMuseum of London and Time Team, it ishoped that a full-scale working version willbe built somewhere in the Museum next year- so watch this space!Jenny Hall,Earty London Dept,Museum of London,150 London Wall,Landon ECZY sHNi h a I I @ m useu m oflon don . oro . u kBronze vessel fromMiddlewichRecent excavations in Middlewich, off a sitebetween King Street and the River Croco,have uncovered the backs of properties thatwould have fronted on to the line of KingStreet, the Roman road from Wilderspool toSandbach. Apart from the Roman road itself,other remains were largely confined to pitsand ditches but the high water table led tothe survival of wattle-lined pits and a planklinedwell. Along with finds of an inscribedbarrel lid and writing tablet, these are someof the best remains of Roman timber fromCheshire, The planking from the well gives adendro date of AD 96-7.Amongst the copper-alloy artefacts is thedecorated base or umbo of a patera ortrulleus which came from a fairly shallow,shapeless feature cut into the natural sandsubsoil close to the Roman road.The vessel has been cast and lathe finished.The underside of the base has a footriilg,about 75 mm in diameter and 10 rnm high,and pronounced moulded concentric rings(Fig 1). In between the footring and themouldings are two concentric circles,possibly compass marks.The umbo is decorated with a roulettedborder and the raised area is decorated withS-shaped 'tendrils' coming from a centralpoint (Fig 2). There are four circles (?grapes)at the terminal of each tendril. There is ahalf-moon shape between each tendril whichthough not joined together, are on the line ofa faintly-drawn circle, possibly anothercompass mark.Fig 2. The decoration on the umbo. Scale t:2 appltx.Although I have found many examples ofsimilar vessels (Bishop and Coulston, 1993,fig 64; Boon 1961; Boon 1984; den Boesterd1956), I have been unable to find a parallelfor this sWle of decoration.Gillian DunnChester Archaeolagy27 Grosvenor StreetChe-s,terCHl zDDg . d u n n @9h estercc. g ou.. u kReferencesM C Bishop & I C N Coulston 1993 Roman militaryeguipmentM H den Boesterd 1956 Descriptian of the collectionsin the Rijlcsmuseum G M Kam at Niimegen 5; tfiebronze vesselsG C Boon 1951 'Roman antiquities at Welshpool',Antiq J 4L, nos L-2, 13-31Fig 1. The underside of the Middlewich vessel.G C Boon 1984 'A trulleus from Caerleon with a stampof the first cavalry regiment Thracians', Antig J 64, 4O3-711

it-' r+,rr*.a'I*n, Jq ts{i ?rr'}r+-;tIvory folding-knife handle from SilchesterExcavations by Reading University atSilchester in 2000 produced a remarkablefolding-knife with an ivory handle showing apair of dogs mating, Traces of leather notedat conservation suggest it was stored in aleather pouch, or perhaps wrapped in a pieceof leather. A conventionally-shaped scabbardwould not have been used, os the knife wasfound with its blade closed. The object isdescribed as shown in Fig 1, with the animalsupright with heads to the right.The animals are caryed in a naturalistic style,set on a small platfolTn, or ground and wereintended to be viewed as shown in theillustration. Seen from the other side theirfaces are obscured. As often happens withanimals carved to be viewed in profile, whenseen face on they appear asymmetrical,Their fur is shown by incised straight andwavy lines that follow the contours of theirbodies. Comparison with mating felinesclearly shows that these animals aredefinitely dogs, as female cats lie on theground during mating. Though it is difficultto be certain of breed and size, the dogs areneither tall long-limbed huntlng hounds nornecessarily small lap-dogs. Though thepossibility that this was a hunt scene wasconsidered, the stances of the dogs areprecisely those that would be assumed in lifeby a mating pair.,,,,,,, ,t,.,::; ,,:,; CmsFig 1. Ivory-handled knife from Silchester.Behind the dogs the ivory is solid andelliptical in section, with an incised cross setat the base on the front side and a singlediagonal line at the rear. There is adistinctive extension in height closest to thedogs, an exact parallel for which can be seenon an ivory handle from Huttwil, Switzerland.The part of the handle closest to the blade isrebated and has two grooves around thecircumference. This area was originallycovered by a band of copper-alloy sheet,now almost completely disappeared. Theblade is pivoted on a copper-alloy pin setthrough the lower side of this rebated area.The iron blade is short and wide, with astraight edge. The back angles upwardsslightly, then cutves down to the tip. Theedge of the blade slots into a central gap inthe platforffi, When open it would lie in astraight line with the underside of theplatfoffn, the angled back facing upwards.The blade form is similar to that of manyclasp-knives, but is particularly closelymatched on one from Vicgu€s, Switzerland.Boon identified small folding-knives like thisone as toilet knives, but the possibilityremains that they may be razors. The shortblade and the angled back would allow thethumb to control the blade quite precisely.The image of a solitary dog is often shownon the handles of folding-kniv€s, both metaland bone or ivory, and another well-knownscene is that of a hound chasing a hare.Another hunt scene shows a pantherspringing upon a deer. Other handles showerotic scenes between hurnatrs, both couplesand threesomes. The Silchester knifecombines these two themes in an image thatmust have been intended to be erotic,though, being two domestic animals of thesame species, the scene must have been onecommonly viewed in everyday life, and sothe eroticism is perhaps not as powerful as ifcoupling between a humdn and an animalwere depid€d, or two animals of differentspecies.Silchester, of course, has many dog burials,either whole or partial, and the pit thehandle came from also contained twocomplete dog skeletons. The handle, in thiscontext, takes on a religious aspect whichsubdues or subverts its erotic message toone of procreation and therefore the triumphof life over death.The pit is cut by the foundations of a late 3rdcentury Roman building, and contained lateT2

tri:f&,Ila -i#r ltuFri. J#{uznd to earlylmid 3rd century pottery (cL75|2OA-225/50). Further evidence of aritual aspect to the deposit came from twosherds, probably both from a single jar,which had both been perforated twice.Pierced sherds and vessels, as well as dogs,are a feature of Silchester's 'retigious' life.Hella EckardtSchool of Archaeology &Ancient HistoryUniversity of LecisterLeicuter LEI 7RHhe$@le.acJtkNina Crummy2 Hall RoadCopfordColchster CO6 18NEssexnina.crummy@ntrlworld.comAcknowledgementsWe would like to thank Gtherine Johns, DonaldBailey, Mike Fulford and Amanda Clarke for theirinput into this article.BibliographyG C Boon 1991 'Tonsor humanus: razor and toilet knifein antiquity', Britannia 22, 2L-32R Degen 1984 'Eine r6mische Kleinplastik: derSchaftirt von Cham-HagendoFn', Helvetia Archaeologica15, 169-184S Deschler-Erb 1998 Rilmische Beinartefakte ausAugusta Raurica 129-131, Taf. 6: 75-8A; 7: 81-84.J Garbsch 1975 'Zn neuen r6mischen Fiinden ausBayern, 1: R6mischen Rasiermesse/, fuyerischeVorgeschichBbldtter tn, 68-85M Fulfiord 2001 'Links with the past: pervasive 'ritual'behaviour in Roman Britain', Britannia 32, 199-218R Hattatt 1989 Ancient brooches and other artefacB,fig 36C Johns 1982 Sex or symbol, 113, 135-61984 Note on a fulding knife, in S S Frere,Verulamium excavations III, 58-9I Liversidge 1968 Britain in the Roman Empire, fig 66E von Mercklin 1940 'R6mische Klappmesserg riffe' , inSerta Hoffilleriana (Festschrift ftir Victor Hoffiller), 339-352E Riha 1986 Toilettgeriit und medizinischesInstrumente aus Augst und Kaiser-Augst, Forschungenin Augst 6, Abb 11C Smith 1908 A guide to the exhibition illustratingGreek and Roman life, fig 180R E M Wheeler 1930 London in Roman Times, fig 19No more mystery?Among the numerous series of amulets withphallic representations, a rare Wpe shows amore or less schematic phallus at one end,and at the other a clenched fist with thewell-known gesture of the 'fica' (the thumbprotruding between the firct and secondfingers), more prophylactic than obscene.The particularity of this family is theexistence of a circular central perforation, ofa diameter which seems too large for the useof the object as a pendant. Amulets of thistype exist in bronze, the most commoncategory, but also in bone. In both cases, asimplified typology would arrange them intotwo classes: more or less horizontal, andcurued or very cuved.Bone horizontal amulets, of which a goodexample was found at Colchester (Greep1983, L39, fig Lil, no 4258), seem to beslightly more frequent than the curved bonetype, to which the two objects from Londonbelong (Wardle 2001 , 7, with figure). Therarity of the curued form is perhaps moreapparent than real; the stylisation of theterminals is sometimes extreffi€, and atendency to break does not facilitate ttsidentification.Here is a short list of the 'bidules' of this kindknown to me. They keep their mystery as faras their precise use is concerned; thediameter of their central perforationsuggests, as indicated by Mrs Wardle, thatthey were set on a pole or stave.ffiffiI01 5ffi4ffiffiffiA'Fig 1 : Bone curved amulets with phallicrepresentations: L, Londoni 4, Augsburg; 5, Badeni 7,Rosinos de Vidriales.-cm1 & 2. London (Wardle 2001) (here Fig 1, t);3. Oberstimm (Sch6nberger 1978; ASS MUnchen1975.89), incomplete (phallic end broken);4. Augsburg, Kornhausgasse 4, Claudius-Nerolevel (Bakker 1985, 103, fig 50, no 2L),complete object (here Fig I, 4);5. Baden, Armee-Spital (Unz L974, 88, fig L, noL7), complete object (here Fig L, 5);13

l/,,,r'*x,-,yTfi ',i,*.rs t .iB {-l,t i,rlij"l,yv'Fr r.*:i'*vt .u$l _f *ht*b5. lfi/indisch (Unz L973, ftg 13, 151);7. Rosinos de Vidriales (Zamora), camp of the AlaU Flavia, Claudian-early Flavian (CareteroVaquero 1998-99, 56, fig 2, no ?.6) ; looking atthe illustration, the object, seen from above andby profile, seems to lack its two extremities(?)(here Fig. L,7).Pierre Abauzit,73 Rue Simone Weil,34500 B€ziers,FranceAcknowledgmentsWith warm thanks to M Feugbre for translatingand arranging the illustratians of this note.BibliographyL Bakker 1985 'Ausgrabungen an der Korhnausgassein der Provinzhauptstadt Augusta Vindelicum-Augsburg',Das Arch. lahr in Bayern, 101-104S Carretero Vacquero 1998-99 'Objetos de huesotrabajados del campamento del ala II Flavia enPetavonium (Rosinos de Vidriales, Zamora)', Lancia 3,Univ De L6on, 49-7LS Greep 1983 Note on the fist and phallus amulets inN Cruffiffiy, The Roman small finds from excavations inColchester 7971-9 {Colchester Archaeologiml Repoft 2),Colchester 1983H Sch6nberger ef al 1978 Kastell Oberstimm. DieGrabungen 7968-1971 (Limesfrorschungen, 18), BerlinL97BCh Unz L973 'R6mische Funde aus Windish imehemaligen Kantonalen Antiquarium Aarau',Jahresbericht Pro Vindanissa, IL-4LCh Unz L974 'R0mische MilitarftJnde aus Baden, einNachtra g', Jahresbericht Pro Vindonissa, 85-91A Wardle 2001, Mystery'widgets'. Lucerna XXII, 7Box appealIn 1998 and 1999 excavations wereconducted at Grateley South, Hampshire aspart of the Danebury Environs RomanProject. The excavation revealed an ovalenclosure of 1tt century Bc-ltt century ADdate and several Roman buildings. Both IronAge and Roman finds were recovered andinclude items of jewellery such as bracelets,finger rings and pins, and iron tools.Of some interest are the remains of twoboxes, one with iron fittings and the otherwith copper alloy. From a grave came agroup of joiners dogs, double spiked loops, aloop-headed pin with ring and a lock plate,while from a pit were recovered a group ofrectangular and curving plates and studs.The studs are of tripartite construction withan iron shaft and domed copper alloy headfilled with a white metal. This filling iscurrently being analysed by Dr PeterNorthover of the Department of Materials,Oxford University. Of interest is the dentatedecoration at the edge of the rectangularplates (Fig 1, opposite). A guick review ofthe literature has revealed little in the way ofcomparatiye decoration, and I would be verygrateful to learn of any similar examples.Emma HarrisonInstitute of Archaeologyi6 Beaumont StreetOxfordem m a . h a rrison @ a rch . ox. a c. u kREVIEWThe Romano-British 'small town' at Wanborough,Wil$hire, by A S Andersoo, J S Wacher & A PFitzpatrickBritannia Monograph 19, 2001 380 pF, 11 plates,L26 figsThis monograph presents the results of sixseasons of excavation by three site directorsat Wanborough, a small town on ErminStreet between Silchester and Cirencester.The excavations took place between 1966and L976, the earliest, by Ernest Greenfieldand lohn Wache r, in response to threatsfrom road building, the latest, by ScottAnderson, iD advance of a flood water lagoonscheme.Three main periods were detected, alldivisible into phases. Period 1 dates to Nero-Vespasian, when Ermin Street was built, anda single wooden building. There arenumerous brooches dating to this period,and indeed some pre-dating it, as well as theodd piece of military metalwork.In the }nd and early 3'd centuries (Period Z)the settlement took off, with numeroustimber buildings constructed, and one withstone foundations. A mansio, picked up byaerial photographs, lies to the east of theexcavated area, and may have been thenucleus of the settlement.Period 3 dates from AD 230 to 400+, a timewhen several buildings with stonefoundations (but probably timbersuperstructures) were erected both frontingonto Ermin street and at some distance14

luteme )anuaryJ$S.inFace A

j't t/'*rriJ!8,-tr a-:r.f"I ,. 7.4,n|'"..1J 6J L, id.l J il-: d; L; | | .i.a id J ,/i* L.'q: rtiit.A few unusual timber-framed buildingshad sleeper-beams laid directly on theground, and some had joists rest!-g +-stones to lift the wooden supersfudureabove the ground. These buildings lay closeto a stream and their designs may have beenattempts to raise the floors above dampground or flood-water level.The report on the structures by period isbrief, and the bulk of the votume consists offinds reporis. The various sections of thepoftery report take up about 130 pages, withuseful illustrations of the forms/fabrics foundin the individual period phases. There aret77 brooches, but this number includesmany pins and pin fragments. Almost all areillustrated, including the latter, though thepoint of illustrating a fragment of a pinlacking each end eludes me. Most of thebrooches date to the 1$ and 2nd centuries, asis expected for southern British assemblages,but most were residual in late contexts orunstratified. This must sound a warning forthe rest of the assemblage, and a similarpattern of dislocated deposition is apparentin the bone pins.The coin graph has a strong finish in the late4th century, and there are several pieces oflate metalwork to go wlth this peak, such asbuckles and strap-ends.As is often the case with excavations spreadover many years and producing large findsassemblages, some of the finds reports weresubmitted t7 years before the date ofpublication. In a few the subject matter isdiscussed in some detail and related to thecontext dates, others consist almost whollyof catalogue entries. It seems likely (andindeed typical) that the context details maynot have been available to those whofinished their reports early. This may accountfor the inclusion of an unstratified medievalthimble.As well as brooches, the copper-alloyassemblage includes a fine array of toiletinstruments, there is a neat little pewterhoard, and the iron objects include a widerange of metal- and woodworking tools, and,most unusually, 35 styli, of which only fourare illustrated. An eraser end with bronzedecoration is catalogued but not illustrated.This is proof positive of Hilary Majo/scomments on p 2 of this issue of thetreatment of styli in finds reports, and is inodd contrast to the illustration of the ?broochpin fragments mentioned above.There is therefore some imbalance inpresentation in this volume, but all in all thiswitl be a useful source to be mined forparallels, some of which may be stratified.The real potential of the Wanborough reportmay not be realised until research is doneinto finds assembiages from small towns ingeneral. As with Wilcote and Elms Farm, theWanborough excavations were rich in metalfinds, and it would be interesting to see ifthis is a real richness per cubic metre of soilshifted Cornpared to an urban site. Given thenumber of l$-century brooches and thecomparative lack of contemporary contexts,it should be real. If so, does the reason lie ina lack of disturbance of the archaeologicallevels over the centuries, or was there acontemporary reason? As so many of thebrooches were unstratified or residual, lackof disturbance is unlikely to be the answer(though it may be the nature of thedisturbance). HOwev€I, recent exCavations atSilchester (urban but not reoccupied in themedieval period, and so perhaps a goodtouchstone for Cornparison with unreoccupiedsmall towns) have turned uP manyunstratified and residual ltt-centurybrooches, but in nowhere near the numbersfrom Wanborough. A contemporary answer isperhaps then the rnost likely.Nina CrummyQuestionnaireEllen Swift would be very interested to hearyour views on careers in finds studies andperceptions of finds research in thearchaeological community and would be verygrateful if you would spare the time to fill inand return the questionnaire enclosed withthis issue. Please pass it on to other findsspecialists who might be interested.16

i{lfg,,i}f ie.ri,:'6r}, .JS# }Spindle, Whorl, Pot - a remarkable group ofgrave goods from Pfakofen, Lkr. Regensburg,BavariaIn 1993 in Pfakofen, near dispersedprehistoric settlement traces , a row-gravecemetery of early medieval date wasdiscovered. It had not previously beenrecognised through either earlier finds oraerial photograph surveys.l The find spot liesin terti ary hilly countrysid€, about 20 kmsouth of Regensburg (Fig t), on the northernedge of the village and about 250 m awayfrom the local church, which documentaryevidence of 1185 referred to as ecclesiamphaphinchaven.2runs via Plattling to Straubing-Alburg, whereit joins the alignment of the Roman road.3The cemetery at Pfakofen was excavatedbetween 1993-1996, and 515 graves wered iscovered. Two graves are particularlyimportant and are briefly described here:Grave J4 consisted of an adult femaleinhumation.o The dead woman lay in awooden coffin (Fig 2A). The skeleton hadbeen disturbed by grave robbers from theupper arms to the upper thighs, so that fewgrave goods were in their original position.Of 55 beads , only eight, found on the innerside of the right arm, were still strungFig 1. The arrow just below Regensburg points to'Pfakofen.The valley of the Grossen Laaber river is partof the early Bavarian settlement area, wellirrigatedand with fertile loess soil. Pfakofenis also connected to the so-calledOchsenstrasse (Oxen-road), which lies at thefoot of the slope along the right bank of theGrossen Laaber river. This ancient route hasbeen equated with the road of theNibelung€ff, which crossed the Donau atPfdrring and, cutting off a loop in the river,I First interim report: S. Codreanu-Windauer, Ein neuentdeclctes ReihengrAberfeld in Pfakofen, fuch. JahrBayern 1993, 121ff.J. Schmid, Die Urkunden-Regesten des Kollegiatstiftes'U.L. Frau n$ Alten Kapelle in Regensburg I(Regensburg l9l l) 2Fig 2. Plans of Grave 74 (A) and Grave 87 (B).together (Fig 2A, 7), the rest were scatteredin the robber pit. together with the remainsof thin silver sheeting, fragments of bronze,3 M. Roedig, Copien" Notizen und Arbeiten tiberZaitkofen und Pfakofen. Das Ftirstl. Thurn undTaxissche Herrschaftsgericht in Niederbayenq Hist. Ver.Oberpfalz, MS. O 151; W. Torbriigge, Die Landschaftenum Regensburg in vor- und ftihgeschichtlicher Zeit,Regensburg-Kelheim-Straubing Ftihrer an arch.Denkmiilern in Deutschl 5 (Stuttgart 1984) 36, 42 ff.o The anthropological report on the skeletal remains fromthe Pfakofen cemetery has yet to be published.17

an iron buckle and a mussel shell. An ironknife on the left thigh was undisturbed (Fig2A, 6), but a comb and comb-case may havebeen displaced by the robbers (Fig 2A, 5).Between the knees of the body laY,embedded in wooden fragments, smallunrestorable bronze fragments with a tinysilver rivet. The most remarkable feature ofthe burial was the recovery of four objectswhich had probably originally been placed ontop of the coffin at the foot end and so hadbeen missed by the grave robbers. Lyingclose together were a ceramic biconicalspindlewhorl , 2O mm long (Fig 3, 11), and abone hemispherical spindlewhorl, 31 mrn indiameter, decorated with fine rilling (Fig 3,12). The latter was found on the blunt end ofa bone spindle, 252 mm long, slightly thickeron the lower third, and terminating, in afinely-worked hook (Fig 3, 13). Above thespindle was a small pottery vessel withsloping sides and six spouts set around acentral opening, 65 mm in diameter.Grave 87 consisted of an undisturbedinhumation of a young girl also lying in awooden coffin. The burial contained a largeamount of bead jewellery (Fig 28, 3), whichhad been scattered over the whole area ofthe body. By the feet were a comb (Fig 28,1), a small cramp made from a strip of iron,11 mm !ong, and a hand-made bag-shapedbeaker with stamped decoratiotr, 92 mmhigh (rig 28, 2).Both these graves contained a large numberof beads, which enable the two burials to bedated. Sorne of the 55 beads which remainedin the adult female Grave 74 can be dated tothe last third of the 6th century AD.s Theycan be linked chronologically to Grave 27 atBasel-Bernerriffg, dated by coins to560/5 706; the dark blue beads withmulticoloured trails suggest a later date of cAD 600.Despite the damage done by the graverobbers, it is clear that the other furnishingsin the adult female burial Grave 74 were ofvery high quality. This is shown by therecovery of the small fragments of silversheeting (Fig 3, 2) and the silver rivet (fig 3,5U. Koclr, Probleme merowingerzeitlicher Glasperlenaus Stiddeutschland, Ann. 6 Congres Assoc. Intern, Hist.du Verre (Liege 1975), 138 ff.6 M. Martin, Das friinkische Graberfeld von Basel-Bernerring @asel 1976), 140, 261 tr.3).7 This is not the case for the young girl inGrave 87. Her undisturbed furnishings arecertainly striking because of the largequantity of glass beads recovered, but all areof common mass-produced types.The grave goods from the foot of Grave 74are the most important, particularly thespindle (fig 3, 13). Examples of wood, thernost commonly-used material, do notusually sulvive, and bone examples are alsorare. Therefore, in contrast to their relativelycommon whorls, spindles are not often foundin row-grave cemeteries.s Known examplesrange in length from 120 to 150 ffiffie, andtaper smoothly to a poifft, with the thickestpart (6-8 mm in diameter) lying near thecentre. At 252 ffiffi, the spindle fromPfakofen is much longer than other earlymedieval examples, and it also differs inhaving a hook at one end. That the find is aspindle and not a hooked needle isdemonstrated by its shape (the thickest partis in the lower third of the shaft) and by itsassociation with a whorl. The bone whorl (Fig3, 12) found the spindle's end has aperforation 75 mm in diameter, and so couldonly be placed 7A mm up from the spindle'stip. Bone spindlewhorls, as well as biconicalclay whorls such as Fig 3, 11, are oftenamong the grave goods placed in theinhumations of adult females and younggirts.lo Bone whorls of more or less7ibid. The other early graves, with the exception ofGrave 314, contained brooches; the early Crrave 314 andthe later Grave 40 produced dress fittings.t Classified by Martin (Footnote 6), 90, From this camethe new find of a spindle from the cernetery atStrasskirchen bei Straubing. We are grateful to H.Geisler for this information.e Sirnau Grave 63, length about 123 mm: R. KoctlKatalog Esslingen, Verdffentl. Staatl. Amt. flDenkmalpflege Stuttgart A l4{II (Stuttgart 1969), 93,Taf 23,81. Duggingen, 128 mm: Jahrb. SGU 49 (1962),97, Abb 58. Zuetern, 149 mm: A. Dauber, 'Neue Fundeder V6lkerwanderungszeit aus Baden', Bad. Fundber. 2l(1958), 157 fi, taf 56, 5. Croldbach 139 ilrm: B. Schmidt,Die spite Volkerwanderungszeit in Mitteldeutschlan{Katalog (Sudteil), Veroffentl. Landesmus. f Vorgesch"Halle 25 {Berlin 1970), 63, Taf 59, k. 5). Mtinnerstadt,length 122 mrn, or about 61 illm: R, Koch Bodenfundeder V6lkerwanderungszeit aus dem Main-Tauber4ebiet.Germ. Denkmiiler Vtilkerwanderungszeit A 8 (Berlin1967), 144, Taf 24, 4.to Overview in. U. Koch Das friinkische Greberfeld vonKlepsau im Hohenlohekreis. Forsch. u. Ber. z. Vor- u.Fnihgesch. Baden-Wiirttemburg 38 (Stuttgart 1990), 169tr Most recent: R. Reiss, Der merowingerzeitlicheReihengriiber&iedhof von Westheim. Anz. Germ. Nat.Mus., Wiss. Beibtinde 10 $,ltirnberg 1994), 140.18

lurena January';#836.r.ri €g! 6a cp oeo 6@ 0 O 0 0 B I 0 0 00 o O 0 tsp "''$ NE a @@ o ao o (D @@CI @D @6 @0@0 @@ @S @ Affiffiffi wwffiagMTmffi ffiilCeTreu 3ffi@@kFLtetffi0WMVMY,#.!lw["4,7w19

lucerna -!anua4, JC02e GDCICI[f OO DDos@o B0oo o000 oE80 O O O O O t s A'O @'e @c dG @-A e 0 o B ts I SS s s s s Ie€ gBC €E €€ B€€ €€Eg @@a@'o@'8@W..oe@8BgB8€e€@@@S@E@@BE1 BT z+3Fig 4. Grave goods from Grave 87. l...glass; 2...iron; 3...bone; 4...ceramic. Scale 1:2.hemispherical section appear to be typical ofthe second half of the 6'h century." Amongthe many forms of decoration used on bonewhorls, the fine concentric rilling on thePfakofen examples is relatively simple and isalso found on the whorl from Grave 7 atGnotzheim.12 The small biconical ceramicwhorl, io a brownish reduced fabric (Fig 3,11), is a common tyPe which cannot beclosely dated.13 It should be noted, however,that its perforation is so narrow that it couldonly be fitted onto the very tip of thespindle.Grave 74 and Grave 87 are also linked bythe deposition of pottery vessels, which isotherwise rare in Pfakofen and elsewhere inthe regionla, and in particular by therr U. Koch, Klepsau (Footnote l0), 169 q R. KochMain-Taubder Gebiet. (Footnote 9), 7 6.tz H. Dannheimer, Die germanischen Funde der spltenKaiserzeit und des fnihen Mittelalters in Mittelfranken.Germ. Denkmiiler V0lkerwanderungszeit A 7 @erlin1962), Taf 24, A 14.t3 R. Koch, Main-Taubder Crebiet. (Footnote 9),76.'o U. Koch Der Grabfunde der Merowingerzeit aus demDonautal um Regensburg. Crerm. Denkm6lerV6lkerwanderungszeit A l0 (Berlin 1963) passim. Th.Fischer, Das bajuwarische Reihengrflberfeld vonStaubing. Kat. Prihist. Staatssammlung 26 (Kallmtinz1993), passim.deposition of a pot at the foot of thedeceased. While the small bag-shapedbeaker in Grave 87 (Fig 4, 4) is of a Wp_ecommonly found in the- Regensburg areals,the vessel with the adult female in Grave 74is unique (Fig 3, 10). It belongs to theunusual group of ceramic _spouted vesselsfirst classified by E Grohne15. An example ofthe form, with three spouts, from the Romanperiod site at Btirs, AltmarklT, is fairly closeto the Pfakofen piece, with sloping sides anda central opening around which the spoutswere similarly arranged; the two vesselsmay have shared the same function, WhileGrohne suggested that modern spoutedforms might be used as flower vases, hefound no satisfactory explanation for those ofprehistoric or early historic date. Howev€r,he stressed the important point that theform was unsuitable for use as an oil-lamp,and therefore by association so are allspouted ceramic vessels.ls Despite Grohne'swork, the interpretation of similar forms aslamps has obstinately persisted in the15 Perhaps also U. Koclu Donautal (Footnote l4), 107 fi,Taf 97, 13.t6 E. Grohne, Die Koppel-, Ring- und Ttillengefiisse.Bremer Wissensch. Cresellschaft (Hrsg.), Abhandlungenund Vortrflg€ 6, Heft % @rernen 1932),94 ffGrohne (Footnote 16), t 18, Taf 43, a.""Crrohne (Footnote 16), l l l fr20

specialist literature. Thus, J Werner identifieda 70 mm high hand-made ceramic vesselwith four spouts and a central opening fromGrave 1 at Bollwerk as a lamp,te This richlyfurnishedfemale grave, dating to the thirdquarter of the 6th centu ryaa, aljo contained acomplete set of brooches, a silver perforatedspoon, a second ceramic vessel, and aweaving batten. Another vessel similar tothat from ffakofen came from Grave 9 atZilluzi.z1 This pot, again identified as a lamp,had a central opening and four spouts andwas placed at the foot of the grave, whichwas also dated to the third quarter of the 6thcentu ry." Somewhat different, but probablyattributable to the same function because ofits central opening and three spouts, is abaggy vessel, L7O rnm high, from the richlyfurnishedfemale Grave !4a at Schlotheim,dated to the second half of the 6th century.Behm-Blancke suggested that this vesselmight be interpreted as related to drinkofferingsor drinking ceremonies of theMigration Period.23 Pollen analysisza of thecontents of this pot and of another smallvessel from the same grave producedevidence for a honey-sweetened alcoholicdrink, such as mead, with the result that theSchlotheim vessel was presumed to havebeen used for a cult drinking ceremony.However, the fact that in each case these socalleddouble-, triple-, and ring-vessels werefound in well-furnished female graves25re J. Werner, Mtinzdatierte austrasische Crrabfunde.Germ. Denkmiler V6lkerwanderungszeit A 3(Berlidleipzig 1935), 85, Taf 6 A. Hriberner alsoconsidered this function for the spouted vessel fromWorms: W. Htiberner, Absatzgebiete fri.ihgeschichtlicherTopfereien in der Zone nordliche der Alpen. Antiquitas 6(Bonn 1969), 137 .'o The grave is dated by a silver coin of Totila (5al-552).I. Werner attributed it with a date c 600: Werner(Footnote 19), 39 fi, 85.2t B. Svoboda, Bohmen in der Volkerwanderungszeit.Monumenta Archaeologica 13 (Prag 1965), 284, Taf 87,10. (I urm grateful to G. Morgenstern, BLD Landshut,for the translation of this text).22ibid The buckle mentioned above with club-shapedprojections with transverse grooves might be similar tothat from Basel-Bernerring Grave 16, which is dated tothe third quarter of the 6tr century: Martin (Footnote 6),8'1, Taf, l, 7 .23G. Behm-Blanckg Trinkgaben und Trinkzeremonienim Totenkult der Volkerwanderungszeit. Alt-Thiiringenl6 (1979),171 ff24 H. Jakob, Pollenanalytische Untersuchung vonmerowingerzeitlichen Honigresten. Alt- Thtiringen 16(1979). 112 tr"2s Bhem-Blancke (Footnote 23), 2Ol tr The functiondefined by Behm-Blancke was also accepted in the mostmilitates against their use in drink ritualsassociated wittr male blood-brotherhoods, asargued by Behm-Blancke.26 The Pfakofenfind is therefore of particular importance inresolving the question of the function ofthese vessels. To this end, the pot's contentswere carefully saved for later research, andthe outside surface was only lightly cleaned.To establish the precise function of thespindle, the Archaeological Textiles Unit atthe Bayerisches Landesamt f0rDenkmalpflege in Bamberg was approachedand rapidly came up with the suggestion,based on comparison with other spindleswith hooked ends, that it could have beenused as a yarn-twisting spindle. Thisinterpretation was confirrned by anexarnination of the relevant literature.The spindle is one of the oldest and mostimportant tools known, used to produce notonly continuous thread for the manufactureof fine textiles, but also a range of threads ofvarious types (Figs 5-6). The fibres can bepulled out of a bundle of raw wool to thereguired thickness, stretched, straighten€d,and given variable elasticity and torsion (Z/Sspinning), before being cut off and woundup. Whorls of various shapes weigh down theFig 5. Men spinning. Beni Hassan, Egypt (after Forbes).spindle and also keep it steady. The whorlsalso add to the rotation movement needed toproduce an even thread. A small hook orincision at the top of the spindle keeps thethread in place during spinning. Lightrecent study of these unusual vessels: H-O. Muthmann,Mehrfachgefiisse (Rheinfelden 1987), 332 t, 348 f26 G. Frngerlin, Grab einer adligen Frau aus Gtittingen.Bad. Fundber. Sonderheft 4 (1964), 34. Pirling 4greeswith this sentiment regarding burial zfiA in CnefeldGellep: R Pirling Das r6misch-frenkische Crreberfeldvon Kreffeld-Gellep 196063. Germ. DenkmiilerV6lkerwanderungszeit B 8 @erlin 1974) 99 I Taf 82,12.2I

i#i#ri'i# .-!# fi *,A:11' i##i'spindlewhorls were ProbSblY used for wooland heavier ones for flat'? but the size andthe right would produceturning it to the leftthreads would then bethreads with Z-twist,s-twist (Fig 7). Therolled into balls andYY'Fig 6. Women spinning, on a Greek epinetron, Scentury BC (after Forbes).weight of both spindle and whorl are alsorelated to the desired thickness of thethread.z8 Whorls can be placed at either endof a spindle. To this day Egyptians place thewhorl at the base of the spindle whenspinning flax and at the top when spinningwoolze. To increase the weight of a spindle,two or even three whorls can be used, withtwo placed below the cone of spun threadand one below it.30 with this method theweight and rotation of the spindle issufficient for the production of yarn.In our grave a bone spindle of 252 mmlength was placed together with two whorlsof differing weight next to a vessel, Thespindle has a small hook and diagonal rillingat its upper end31. It appears that the bonewhorls would have been placed on the lowerend of the spindle while the ceramic examplecould have been placed onto either end32(Fig 2A, 1-4). It seerns likely therefore thatthe spindle was used with both whorls.Ancient literary and iconographic sourcesdocument several methods for producingthreads and yarns,33 Turning the spindle to"Elasticrty is a characteristic of woollen threads whichwas countered by the use of heavier whorls: R. J. Forbes,Studies in Ancient Technology, ry (New Yorlq 1987),154.2*ibid rs4, 167." ibid 15630ibid 159 f. Also see G. Cnenander-Nyberg, Spinningimplements of the Viking Age from Elisenhof in thelight of ethnological studies, Nesat Itr TextileSymposium in York (London 1987), 73 ff31 Such rilling is coilrmon: Forbes (Footnote 27), 154,160 fig 13, 168 fig U.32 M. von Kimakowicz-Winnicki, Spinn- undWebwerkzeuge, Mannus 2 (Leipzig 1930), 59 Abb 93.The whorl shown there is identical to our example and isidentified as a yarn spindle whorl." Forbes (Footnote ?7), 157 fig 12, 165 fig 15, 166 fig16. Also H. Bliimner, Technologie und Terminologie derYzIFig 7. Scheme of S- andyarn.Z-spinning used to prcduceplaced in a basket or small vessel.34 Theends of the threads would be pulled fromthese balls, knotted and then, using aspindle, twisted together into yarn. Yarnpossesses increased strength, weight andvolume and its surface characteristics aremore attractive.3sCatullus mentions the use of a spindle with ahook specifically for holding threads.36 Oncethe spindle was full, the threads were takenoff and wound into balls. He even mentionsdouble threads which were kept in smallbaskets or vessels ready for further work.Kimakowicz-Winnicki has documentedspinning technigues in the Siebenbtirgenarea of Romania where ancient traditionssurvived into the 20th century. Among themis a yarn-making process where several ballsare placed into bowls, with the threadsleading to a hanging hooked spindle (FigB)." The local women use this techniguewhen they do not have access to a spinningwheel.38Crewerbe und Ktinste bei Crreichen und Roern I (Leipzig1875), 107.34 Forbes (Footnote 2t), 168 fig 17; Kimakowicz-Winnicki (Footnote 32),62 Abb 99.3s See C. Crockett, Das komplette Spinnbuch (Bonn1980), 36 f. An additional whorl adds more momentumto the thread. G. Grenander-Nyberg, IJber die Z- und S-Drehung von Garnen in: I. H[gg, Ausgrabungen inI{aithabu 20 (Neumtinster 1984), 287 f.'u Forbes (Footnote 27), 167.37Kimakowicz-Winnicki (Footnote 32),62 Abb 99.38tbid 6r.22

through the spouts was uncertain. The vesselis quite small and could only have held tinyballs of very fine threads. As the threadswere pulled out the balls would easilybecome tangled. A better solution is to feedthe threads through the spouts from theoutside, then pull them together through thecentral opening to be spun into yarn (Fig10). To maintain the tension, each ball wasplaced in a small bowl or basket.Fig 8. SiebenbUrg method of spinning yarn (afterKi makowicz-Wi nnicki) .The fact that specific vessels occur both inthe literary and iconographic sources shedslight on the find in Grave 74, where thespindle and whorls were found immediatelynext to the spouted vessel. The latter wastherefore also probably used in textileproduction and so its contents wereexamined very carefully. Surprisingly,several fibres could be seen (Fig 9). Thesewere identified as wool-fibres of differentcolours (blue, reddish-violet, green andyellow). Some of the red fibres may be silk.It seems certain therefore that the vesseland spindle are functionally related.Fig 9. Wool fibres from the pot, Grave 74, Pfakofen.(Enlarged by about x 30).Next, using an exact replica of both spindleand vessel, an attempt was made toestablish their precise function.3e It wasassumed that the spouts would separate anddistribute the threads during yarn-making,but the direction in which the threads passed3' We would like to thank H. Voss, Bayer. Landesamt fDenkmalpflege Bamberg and L. Breinl, Bayer.Landesamt f. Denkmalpflege Regensberg for theproduction of these replicas.Fig 10. Reconstruction of spinning yarn by the methodsuggested by Grave 74. (Scale approx 1:6.)In the case of Grave 74 at Pfakofen thethreads must have been very fine indeed asis also indicated by the very slim spindle andthe light whorls. It is also likely that thethreads were treated in some way toincrease their strength (this would beespecially important if they were to be usedfor weaving); a smooth and lubricatedsurface would make their use in furthertextile working processes much easier.{Starch glue, tallow, wax and oils may havebeen used to treat the threads.al A spoutedvessel such as the one from Pfakofen mayhave contained fluids used to treat threadsbut, due to the difficulties of identifying suchsubstances, flo analysis was attempted.a2N This is especially important for fine threads andembroidery threads, etc.4t W. Sen[ Werkstoffkunde der Textilien II (Berlin1954), 135. Forbes (Footnote 27), 160, mentions thesuspension of threads 'in grease', and Kimakowicz-Winnicki (Footnote 32), 62, Abb 103, mentions threadsdragged through water.*2 We thank Dr M. Koller, Ddrner Institut Mtinchen forinformation and help on this subject. See also R. D.Bleck & M. Bdhmel, Chromatographische23

,r5{1$:-.i'il}' *,i.d,t? &,#r' j,' jJiili'In this context we should again return to thespouted vessel from the female burial 14a atSchlotheirn, Analysls there indicated thepresence of hon€Y, from which it argued thatthe vessel contained an alcoholic liquid. Thissuggestion was based on pollen analysisundertaken by H. lakob, but has since beenquestioned by both chemical analysis andexperimental archaeologYos. AS a result, thefunction of the Schlotheim vessel is stillunclear but the results from Pfakofensuggest that it was also used in textileprodugtion. It is Possible that the waxresidue identified by lakob was used to treatthreads. This link with textile production issupported by the high proportion of woadpollen.4During the analysis of the contents of thePfakofen vessel, numerous obiests of c 3 mmwere extracted and proved to be smallinsects, identified as a *pe of scate insectas,of which there are several species.6 In themedieval period two were used, togetherwith plant colouring agents, to produce reddyes.aT These are KermeS vermilio, orMediterranean kermes, and Porphyrophorapolonica, which produced the so-called Polishcochineal. Kermes vermilio has been usedsince prehistoric times. The insect lives onthe Quercus cocciferd, a bush-like oak andAnalysemethoden. Restaurierung und Museumstechnik 4(Weimar 1981), 15 ff43 Jakob (Footnote 24), ll2. Behm-Blancke (Footnote23), 171 tr After only 5 months buried in soil, honeycannot be identified with chromatography. Blecktherefore states @leck & B6hmel (Footnote 42), l8: 'Itis certainly impossible to detect sugar remains onarchaeological material such as sherds or the soilcontents of vessels Any attempts would be futile.'* Jakob (Footnote 24), ll2, 114.4s The insect could only be identified by morphologlcalcriteria because deposition in the soil had washed out allthe colouring, which therefore could not be analysedchemically. In contrast, the colouring agent is bound tofabric and threads through the mordant in the dye. Thisallows the colour to be detected on archaeologicaltextiles.46 We would like to thank Prof H. Schrnutterer,University of Gessen, Dr Ch. H. Fischer, Hahn-MeitnerInstitut Berliq Dr M. Koller, D6rner Institut Mtinchen,and J. Banclq Freiburg i. Br. for information andcomparanda.47R. Schoh Aus der Creschichte des Farbstoffirandelsim Mittelalter. Diss. Mtinchen (1929), 17 tr W. Born,Der ScharlactU in L. Niencki, Die Kunst des Fiirbens mitnatrirlichen Stoffen (Bern, Stuttgart, 1984), 85 tr.Kermes is an oriental word meaning 'worm', 'The termusually used in the medieval period is vermilium, ibid85.produces a highly-valued scarlet colour.€The Colouring substance they produce iscalled kermesic acid. Pliny states that thejuice of the young Kermes is matt and lacksstrength and that the juice of the olderinsects fades quickly, but nevertheless thisdye was very valuable.4e Recent research hasshown that Kermes vermilio vuas used tocolour the textiles in the grave of the Celticchief from Hochdof, but whether thedyestuff or already-dyed fabrics wereimported from the Mediterranean isunclear.5oThe huge importance of scarlet dye in theearly medieval period is indicated by anaccount of Charlernagne written by a monkfrom St Gall, Notker the Stutterer (c AD 840-912). He describes how Charlemagne worescarlet red wrappings on his shins and linentrousers Of the same colOur.Sl BeCauSe Oftheir valu€, the insects themselves wereknown as irnportant trade goods.s2In most cases, the species concerned wasprobably the Polish cochineal insect,Porphyrophora polonic€t, which in localfolklore is also called St lohn's blood.53 Thisis a parasite, which, ifi contrast to the veryspecialised Mediterranean species, lives onthe roots of a variety of middle and southernEuropean plants, in particular Scleranthusperennis, or Perennial Knawel (Fig 11). InPoland, Russia, the Ukraine, Prussia andSaxony these insects were used up untiln* E. Ploss, Ein Buch von alten Farben (Mtinchen, 1989),49.4e Bhimner (Footnote 33), 242 (and for generalcomments 215 tr).50J. Banch Die Textilfunde aus dem hallstattzeitlichenFtirstengrab von Hochdorf, Gemeinde Eberdingen (Krs.Ludwigsburg). Nesat V, Textilsymposium Neumtinster(Neum0nster, 1994), 49 f.51'Erat antiquorum ornatus vel paratura Francorumfasciloae crurales vermiculatae; et subtus eas tibialia velcoxalia lineq quamvis ex eodem colore ...', H- F-Haefele, Notker der Stammler, Taten des Kaiser Karlsdes Grossen, Monumenta Germ. Hist. Scriptores rerumgermanicarurn, Nov. ser 12 (Miinchen, 1980), 46- AlsoPloss (Footnote 48), 48 f and O. v. Gerkg Die Trachtder Crermanen in der vor- und frdhgeschichtlichm Zeit,Mannus Bibliothek 24 [,,etpzrg, 1922),49.52 Ploss (Footnote 48), 49. Scholz (Footnote 47), 17 Ssnggests there were two type$ of kermes in the 12*century.s3A. Verhecken & J. Wouters, The Coccid Insect Dyes,Ifistorical, Creographicat and Technical Data, Bull. Inst.Royal Patrimoine Artistique 22 (1988/89), 226.24

,,'Uf$,'fld' Jkfi f,,S,"i, "J#df.Iquite recently,il though it is usually arguedthat they were not used as early as theKermes vermilio.Casfrt'l-eat .Y(3*C}quite likely that the preferred host plantswere cultivated deliberately to increase theease of harvesting.The insects were killed in vinegar and thendried; in this state, they could be stored andtraded. Before use as a dye they had to besoaked in water for several days, thencrushed and placed into dye-vat. The pretreatedwool or silk was first placed into thecolouring bath and then carefully washed.t;is ,} ,**.;{rll.*t.Fig 11. Life-rycle of Porphyrophora polonica, with itshost-plant, Scleranthus perennis (after Born).The insects producing Polish cochineal aredescribed as very small, a description whichagrees well with those found in Pfakofen,which had an average size of 3 mm. fn theirdry state they appear shrunk and wrinkly.Wouters and Verhecken describe adultfemales as 4-7 mm !ong, a size reached by afew of the specimens in the Pfakofen vessel(Fig I2). Potish cochineal insects can also bedetected chemically by the presence ofcarminic acidss (Kermes vermilio produceskermesic acid).Traditionally, the dark red tiny insects werecollected from the roots of their host plants,a process which began with a religiousceremony on St lohn's day between 11 amand LZ noon. Hence the dye's name of Stlohn's blood.s6 Haruesting the insect was atime-consuming task and only about 60-100g of insects could be gathered each day. It is54 Scholz (Footnote 47), 19. Verhecken & Wouters(Footnote 53) , 224. Born (Footnote 47'1,88. Brockhaus,Konversationlexikon 9 (Leipag 1894), 159, 941.Verhecken & Wouters (Footnote 53), 226.s6"Born (Footnot e 47), 88. W. Wilmsen, Handbuch derNaturgeschichte ftlr die Jugend 3 (BerlfuU 1831), 450,talks about the hanresttng of St John's blood, which wasstill used as a red dye instead of American cochineal tillas late as 183 l. Brockhaus (Footnote 54), 159, 941 .tFig L2. Top row: modern Polish cochlneal insects (afterVerkecken & Wouters). Bottom row left: Pfakofienkermes insects in 3 stags of development. Left handinsect 4.8 mm long Bottom !'ow right: enlargement ofthe top right-hand insect from the three shown to theleft.This use of kermes for the colouring oftextiles poses important question withregards to the Pfakofen find. As the insectswere placed into the vessel whole theycannot have been used as a dye beforedeposition. The discovery of threads ofdifferent colours in the vessel alsodemonstrates that the pot did not contain adye-mix when last used and deposited. Theinsects were placed into the burial as rawmaterials and therefore as a grave good intheir own right. They were included as a parcpro toto for the expensive process of textilecolouring and as such fit well with the themeof textile-working represented in this femaleburial.This surprising result could be tested againstanother burial. The bag-shaped vessel fromthe young female burial Grave 87 was alsoexamined and yielded the same result itcontained whole kermes insects as gravegoods in their own right. No fibres were57 Verhesken & Wouters (Footnote 53) ,226 f.25

J#{tt';'f# -?a * uex,' i'SiiJdetected. Indirectly, this also confiffns thespecific use of the spouted vessel in theyarn-making process.Our two burials are so far the only exampleswhere it has been possible to Provescientifically the presence of a dye as agrave good, but it is possible that kermesinsects or other dyestuffs were morecommonly included in female burials. Thespouted vessel and another pot from thefemale burial at Schlotheim provide a clue tothis, as the report states that 'microscopicanalysis revealed a large number of blackshapeless objects which were certainly notcharred plant remainsds. There were also 'afew red objects which appeared to havetraces of sugar attached'.Se It seems likelythat kerrnes insects were present in thisvessel.In this context reference can also be made tothe very wealthy Nordic ship burial fromOseberg which contained the burial of ahigh-ranking woman, furnished with, amongother things, a decorated carpet and a set of50 weaving tablets and partly-workedbraid.60 Spindles were also foufld, as well asdye-plants such as woad, Isatis tinctoria. Thecontext of these finds suggests that this elitewoman was herself involved in spinniffg,dyeing and weaving.utApart from demonstrating for the first timethe deposition of dyestuffs as a grave good,the importance of the Pfakofen burials alsolies in the fact that they represent the oldestdocumented use of the Polish cochinealinsect in Central Europe. The oldest writtenaccount is found in the Capitulare de villis (cAD 800).62 The source mentions female'* Jakob (Footnote 24), ll2.t'ibid r13.* Ploss (Footnote 48) ,25.6t A. Bartel, when studying Merovingian fabrics oftenfound very small fragments of braid, probably oncecoloured. A Bartel & R Kn6chlern, Zu ernem Frauengrabdes sechsten Jahrhunderts aus Wagng am See, Lkr.Traunstein. Oberbayern. Crermania 7l[J (1993), 419 ffIn iconographic sources of the 8/9th century there arealso colourful bands in a variety of materials: H.Dannheimer &, M. Dopsctr, Die Bajuwaren.Ausstellungskatalog (Rosenheim, 1988), 223 Abb 147.E. Thiel, Creschichte des Koshims (Berliq 1980), 86,Abb 159, 87 Abb 158.62 Capitulare de villis 43: Ad genicia nostra, sicutinstitutum est, opera ad tempus dare faciant, id est linun;lananr, waisdo, vermiculo, warentiq pectinos laninas,cardones, saponern, unctunl vascula vel reliqua minutiaquae ibidem necessaria zunt. (You shall deliver, asordered and at the correct time, to the femaleworkshops (genetia) which are to receive theraur rnaterials necessary for te>ctileproduction from the imperial farms.63 Thesource mentions woad for the colour blueand madder and scarlet {vermiculo) for thecolour red, and certainly demonstrates thatthe use of these agents for dyeing was wellestablished by the I to gth century.6uDocuments from the time of Charlemagneillustrate that the necessary plants werealready cultivated by AD 800.5s The fact thatthe Kermes vermilio insects are mentionedtogether with plants produced on imperialfarms suggests that they were not importedfrom the Mediterranean. References to thedeliberate breeding of these insects occur inseveral medieval sources, most interestinglyin the Register of the monastery of StEmmeram in Regensburg from 1031.66Several villages are named as having tosupply vermiculi to the monastery.67Dyestuffs were of great importance tomonasteries and could be used for thecolouring of textiles, book illumination andfor medicinal purposes.6sOn most farms, pdyment in money replacedthose in kind by the l1th century, but twoworkhouses the required materials, namely flax, wool,woad, scarlet, madder, wool-combs, teasels, soap, frt,vessels and all the other small required objects). G.Franz, Quellen nfi Creschichte des deutschenBauernstandes im Mittelalter, Ausgewiihlte Quellen zurdeutschen Creschichte des Mittelalters 31 @armstadtt974),50 ff63 Scholz (Footnote 47), 19; for genetiae: E. Ennen,Frauen im Miuelalter (lvftinchen 1985), 88; A. Gelier,Birka III, Die Textilfunde aus der Grabern Oppsala1938), 40 ffuolbid.65 Ploss (Footnote 48), 39; for madder: G. Heuz6, DerKrapp in: L. Nencki, Die Kunst des Fiirtens mitnattiriStoffen (Bern, Stuttgart 1984), 125 tr.* Ph. Dollinger, Der bayerische Bauernstand vom 9 biszurn 13 Jahrhundert (MUnchen), 455 tr Dollingermentions the payment of vermiculi in St Remy in the 9ecentury (ibid, footnote 189); for more detail on the StEmmeram monastery sse: H. Heimpel, Das Crewerbe derStadt Regensburg im Mittelalter (Stuttgart 1926),35 fi,esp 44. Heimpel also mentions the payment of vermidiin the Register of Pnim (9* century; ibid, 4, footnote29). (The sources of the St Remy and Prtim dyestuffs isnot known.)u'Doilinger (Footnote 66), Harting (no 2), Scheuer (no3), Aufhausen (no 13), Hagelstadt (no l4), Dtinzling (nol5), Peising (no l7), Herrenwalthann (no l9), Ffrrring(no 31), Unterhartheim (no 32), F{ausen (no 35},Hohengebraching (no 49). AIso see P. Mai, Der StEmmeramer Rotulus des Gtiterverzeichnisscs von 1031.Verh. [fist. Ver. Oberpfalz 106 (1966), 87 ff,u* Ploss (Footnote 48), 36 ff26

farms still supplied colouring insects in the11 to 12th century: the monastery receivedvasculum vermiculi from Hausen andvermiculi coppos from Hohengebraching.6eThe source mentions containers but it isimpossible to estimate their size andtherefore the volume of insects; their valueranged from 6 to 54 Pfennig.'oAn important aspect of this source is thedirect link between the payment and use ofinsects and fabric-working. All the farmssupplying the insects also had to supply linenand woollen fabrics (pannum de lino/delana).7L These are therefore clearly farmsengaged in textile-working. The fact that theinsects had to be supplied at regularinteruals suggests that they weredeliberately cultivated and that the initialprocessing (killing in vinegar and drying)probably also took place on site. The sourcefrom 1031 probably records much oldertraditions and practices and it is thereforenot surprising that Pfakofen is located in anarea for which we have later accounts for thebreeding and processing of Kermes insects.It seems likely therefore that there was acontinuous tradition of dyeing using thisinsect from at least the 5th century onwards.The deliberate deposition of the insects inthe female Graves 74 and 87 at Pfakofensuggests that both women were somehowassociated with the cotlecti ng72 or processingof these insects, including the actual dyeing.It is also possible that the insectsrepresented a very valuable commoditywhich was produced in Pfakofen and formedthe basis of the community's wealth. Tradein thls valuable dyestuff is documented earlyon73 and Pfakofen's location near the socalledOxen-road would have been anadditional advantage.Grave 74 also provides concrete evidence forcomplex textile-working techniques such asthe production of yarn, which has alwaysbeen used for a multitude of purposes suchas weaving and stitching.'o Specific tools areu'Dollinger (Footnote 66), nos 35 and 49.70 The Register of St Emmeram for l03l gres thepayment for the value of a pig as 5{ Pfennig; this is alsothe average value for the vermicali: Dollinger (Footnote66), nos 37 and 38." S"e Footnote 68; also Heimpel (Footnote 66), 35 tr72 Perhaps children with their smaller hands were usedfor the collection ofthe tiny insects.73 Ploss (Footnote 48) , 49.'n Nahftiden in vielftiltigem Einsatz bei H.-J. Hundt, DieTextil- und Schnurreste, Die frtihgeschichtlichenot unusual in burials and our burialprobably reflects this woman's personal workin her community.Dr Silvia Codrean u-Winda uer,Bayerisch* l-a nd*amt ftir Denkmalpflege,Keplerstnsse 7,93U7 Regensburg,Germanycod rea n u . b lfd @t-o n I i n e. deAntja turtelffiyerisch* land*amt fiir Denkmalpflege,Au&enstelle Ba mberg Sdt/oss Seehof,96177 Memmelsdorf,GamanyText tnnslated by H Eckardt & /V Cntmmy.Picture creditsFigs L-4, 7 , 9: Bayerisches Landesamt fiirDenkmalpflege.Figs 5-6: R Forbes, Studi* in ancient technologyFig 8: after M von Kimakowicz-Winnicki, SpinnundWebwerleeug€, Mannus 2 (Leipzig, 1930).Fig 11: W Born, 'Der Scharlach', in L Niencki, DieKunst des Fdrbens mit natiirlichen Staffen (Bern,Stuttgtsrt, 1984).Fig I? top: A Verhecken & J Wouters, 'The coccidinsect dyes, historical, geographical, and technicaldata', Bull Inst Royal Patrimoine Artistique 22(1e88/8e)Fig 12 bottom: Bayerisches Landesamt fUrDenkmalpflege (A Bartel).Marschensiedlung beim Elisenhof in Eiderstedt 4(Franld,rrt am Main, l98l), 53 tr Zwirne alsNti,hmaterial mit grossem Anspruch auf Haltbarkeitwurde und auch zllm Zusammenfiigen von Schuhledern,zum Nflhen von Ledetfutteralen fiir Nadeln oder auchMesserscheiden verwendet: Heftfaden an Gtirtelschnalleaus Eggolflreirn, Lkr. Forcheinr, Grab l l l und Niihznirnfiir Lederfutteral aus Gelting, Lkr. Bad Tolz-Wolfratshausen, Grab 27: Dolarmentation BLfl), ABartel. Glatte geschlossene Zwirne gut vorbereiteteignen sich besonders als PerlenkettenfEden:Perlenkettenfiiden von Crelting Llff. Bad TOIr-Wolfratshausen, Grab 43: Dokumentation BLD, ABartel. Zwtne als Reparaturf[den in Verbindung mitMetallen und auch fiir Riemenverteiler,Riemenbeschldge, Nietlasche:/ H. Farke, Die Textilfundeaus dem friinkische Graberfeld von Alactu Alt Thtiringen25 (1990), 184, 187 fi, 192; H. Farke, Textile Reste anzurei volkerwanderungszeitlichen Vogelfibeltl AltThtiringen 26 (1991), 197 tr FrtihmittelalterlicheBrettchenbiinder und Biindchen in Materialkombination:Bartel & Kn0chlein (Footnote 6t). Daru auch Hundt(above), 69 tr Siehe auch die Brettchengarnitur vomOsebergschiff: Ploss (Footnote 48), 25. Eine Vielzatrlvon Brettchenbindern wedrden bei Creijer (Footnote 63),76 E vorgestellet.27

;'-, , -,.,.'4 -r* ,\ 'l '+ r* ,r" t ts{ i Tl rfi.+ 1lI,l".l'i"ff: ) l*1 J'#': l;'g) l' ;.*,*.RFG & HMS Autumn Meeting at ExeterStrong winds and driving rain did not preventabout 40 members of the RFG and theHistorical Metallu rgY Society gathering inExeter for a joint meeting. The day'sproceedings were started ofT by lohn Allen,Curator of Archaeology at Exeter's RoyalAlbert Memorial Museum, who described thecollectioffs, particularly those itemsconnected with metal-working. Amongstthese a few remarkable examples stand out:the Bronze Age rapier moulds from southwestDevon, the Coffinswell hoard ofcurrency bars, found in bundles tied up bybracken stems, and the tin ingots fromBigbury Bay, of uncertain date.In Exeter itself excavations have produced agreat deal of evidence for post-medievalmetal-working, ifl particular three importantbell foundries. The best preserved also madedomestic objects, such as skimmers andcauldrons. Some of the skimmer handlesconveniently bore the founder's name - JohnBirdall.Metal-working of the Roman period wasfound in the fabrica of the Roman fortress,where an aisled building appears to havebeen used as a finishing shop. Recentexcavations have mainly taken place in thesuburbs, often the industrial zone of Romantowns. A large array of small finds was outfor examination, including the military andother objects not already on display.Gill luleff then spoke on Ea rly ironworking onExmoor. A major multi-funded project toexamine ironworking on both Exmoor and itsmargins started four years ago and is nowwell underway and producing someinformative results. It is concerned with theevidence for raw material production , i€mining and smelting, and its impact on thelandscape.Exmoor's iron deposits comes from lodesextending east-west across the moor. Thereare also copper and lead deposits, andlead/silver deposits near Combe Martin(where there may be evidence for silverworking).The iron lodes outcropintermittently on the surface, where the oreweathers and oxidises, and so becomeseasily smeltable. Nineteenth-centuryextraction of iron was usually done by miningundergrouf,d, with the ore then sent toWales for smelting, but some exploitation ofthe lodes was also done from the surface,with the iron being smelted on the spot.This surface extraction shows up as eastwesttrenches across the moor, one of which,known as Roman Lode, is 650 m long and upto 40 m deep. Variations were noted in thesize of the spoil heaps alongside the lode,some small spoil-heaps being sited alongsidesmall interconnecting pits dug down into thelode and often associated with workplatforms. These spoil-heaps proved to bedump from smelting and contained charcoalas well as tap slag, thus enabling someradiocarbon dates to be done. Three sitesproved to be Late Iron Age to Early Roman indate, one was immediately post-Roman(450-615), and several were medieval. Theplace names of several of these sitessuggested that it would be worth while doinga place-name survey of the area (eg RomanLode, Sindercombe, Blacklake Wood).At Brayford slag could be seen everywhere inthe village, in gardens and gateways etc, butobviously very disturbed by modern activify.Excavation of a large section proved possiblewhere a tennis court had been terraced intothe hill, which turned out to be a enotrnousslag heap, full of tap slag, furnace liniog,charcoal efc. Several pot sherds, includingsamian and BB1 ware, and a coin were alsofound" The coin has not yet been dated. Thesize of the slag heap and the absence of anyevidence for the smithing of iron suggestedthat the smelted iron was exported, notreserued for local use. Beneath the slag heapwas a thick dump of grey clay, which couldnot be bottomed, but which was found tocontain a little slag, evidence of smeltingeven earlier than the lst- to 2nd-century datesuggested by the pottery.How metal and stone artefacts from RomanCornwall rnight demonstrate the acceptanceor rejection of Roman ideas was thesubstance of Henrietta Quinnell's talk. Thecomparative scarcity of metal objects suchas coins could be due to Cornwall's acidicsoil, but there also appears to be a longtradition against the digging of pits, a veryefficient middening system being usedinstead. What few artefacts there are frompits could often be seen as evidence forstructured deposition, one example being adeliberately folded woolcomb and a smalllead weight.28

,trf8,';r:i# .,jci;'j r,.?ri. ;-SiUtAt the enclosed settlement of Trethurgythere was a tradition of working the local softigneous rocks into weights and mortaria. Thelatter are usually about 30 cm in diameterand their rims mirror the chronologicaldevelopment of Roman pottery fomrs. Theirproduction may have been a householdindustry, but there is some slight evidencethat they were exported, as one has beenfound in Exeter. Larger mortars known asTrethurgy bowls (Quinnell 1993) are alsofound, about 5O cm in diameter, with openrectangular skeuomorphic handles on thesides (Fig 1), the form perhaps copyingthose on a tin bowl from the Treloy, nearNewquay (Penhallurick 1986). These mortarsare also found beyond the region, with onecoming from Lydney and another from asfaraway as Richborough.The stone weights vary from small to large,and most are slightly recessed on theunderside. The weisht data has beenexamined, but does not seem to conform toany particular standard. There seems tohave been some copying of Roman metalforms; the small round weight found with thewoolcomb mentioned above appears to copythe form of a bell, while some of the largerones resernble Roman pyramidal weights.The iron tools and other objects from sitesacross the county are Roman in style andcraftsmanship. Many have been chopped upfor recycling, which matches the lack ofevidence for iron extraction in the area. Thewoolcomb had been ingeniously repaired. Onthe original the teeth had been cut into apreformed block, just like the teeth onwooden and antler combs, but some hadbroken off and new ones had been weldedinto position. This repair was not visible withthe naked eye, but showed up clearly on anX-ray.Nails and other structural fittings, such asjoinert dogs, had been enthusiasticallyadopted by the inhabitants of RomanCornwall, and hobnails were also found insome numbers. The long-term curation ofRoman metalwork appears evident from thefact that the few brooches from the countyare usually of late 1s to znd century date butcome from late Roman features and are nottikely to be residual. They may have survivedso long if regarded as symbols of wealth orstatus.It appears that in Cornwall Roman thingshad been adopted where they served apractical purpose, but there was no evidencefor the 'conspicuous consumption' seen inmany areas of Roman Britain, insteadmetalwork was curated where possible, orrecycled if not.Neil Holbrook (Cotswold ArchaeologicalTrust) examined the evidence for coastaltrade around the south-west peninsula.Whereas much of Britain's imported Gallicpottery followed a route up the Rh6ne andthe Rhine and then across to the Thamesestuary, the North African amphorae andc6ramique i l€ponge from the south-westpointed to an Atlantic route across the Bay ofBiscay, while cross-channel trade hadalready been established in the Late Iron Agewith Armorican pottery coming in to theports at Hengistbury Head and Mount Batten,Plymouth.The presence of the Roman army during theearly years of the Roman period had been aconsiderable stimulus to trade, and theroutes established then appear to havecontinued after it moved on. Black burnishedware travelled from Poole Harbour to Exeterand Plymouth, and with it came products ofthe Kimmeridge shale industry, Purbeckmarble mouldings and veneers, as seen inExeter's bath-house, and later on Purbeckmarble mortars and buff-coloured tiles. It isnot possible to say, however, if these thingscame in on the back of less tangible imports,such as corn or salt.Among the products that flowed out of theregion was smelted iron, from the BlackdownHills in east Devon and from Exmoor Forest.Some may have gone to the SevernEstuary/Bristol Channel ports on the northcoast, some to the Channel ports on thesouth coast. From south Devon camepottery, found in Cornwall, Dorchester andLondon (the riverside site of New FreshWharf/St Magnus House), slate for the roofsof buildings in Exeter and villas in southSomerset, and tin from the south side ofDartmoor. Cornwall's distinctive Gabbroicpottery travelled very little, but a few piecesare found east of the Tamar, and, though theTrethurgy stone bowls described above donot appear to have been exported in largequantities, at least one reached Richborough.Cornish tin was also exported. Though tiningots have been found in the Thames atLondon, and much also went to the Mendipregion to be alloyed with lead to makepewter objects, pewter flagons from a wreckhave been washed up onto GoodringtonBeach in Torbay, suggesting some pewterworkingmay have gone on in the region.29

;i#i#;-.fif ,-bili'e t]' i?fre'Ot0cmTrethurgy bowls made of elvoff, aDevon andfine grained, oftenCornwall. Drawingslight coloured porphyritic rock, fouby Sandy Morris.nd in30

Itr{:g,,.$* _,b,'? t #{i, c-#iil}In summ dry , tin, pottery iron, buildingmaterials and some stone artefacts movedeast along the coast, where the ports ofHamworthy and Clausentum may have riseto be more important than Poole Harbour,while black burnished ware, shale objects,Purbeck stofi€, and tiles rnoved west.Neil ended by mentioning the site of Nornourin the Isles of Scilly, which produced largenumbers of brooches, Dea Nutrix figurines,miniature Fots, and coins. At first thought tobe a workshop, it is now more likely to be ashrifi€, perhaps used by sailors.lustine Bayley (English Heritage) then spokeabout Roman non-ferrous metalworking inthe south-west. Before examining theevidence for this area , Justine pointed outthat metalworking is a series of processesmining, smelting, refining, olloyitr9, castingand/or working. Each stage started with rawmaterials, and in Roman times the metal wasthe valuable factor in these stages, not thetime and effort, though these days thereverse is generally true.The distribution of sites with evidence fornon-ferrous metal-working shows some inCornwall, only one in Devon , a lot in theMendips, several in Dorset, and a scatter inGloucestershire.Tin ore, cassiterite, is specific to theCornwall, where it was mined and smelted toproduce ingots. These were exported to Bathand its immediate hinterland, where the tinwas alloyed with lead from the Mendips toproduce pewter, There are several finds ofpewter moulds from this region, thoughthere is also a scatter of moulds fromelsewhere.The evidence for lead smelting is usuallyconfined to the production (extraction)areas, i€ the Mendips, Wales, Derbyshire,and the northern part of the Pennines, butthere is also evidence for lead smelting fromLondon. Mendip lead contains silver, whichwas extracted as part of the smeltingprocess. There may be evidence for silverworking from Combe Martin in Devon.The debris produced by the various stages ofworking non-ferrous metals is generallycommon to all the materials and all theareas. The evidence for processing usuallyconsists of crucibles, droplets, and moulds,and , rdrely, ingots and ingot moulds mayalso be found. Manufacture can be shown bythe recovery of part obiects, such as rods,bars, and wires, or by offcuts, tools, andmoulds. Special vessels Can demonstratethat cupellation (extracting silver from silverleadalloys) or parting (separating silver fromgold) took place. Exeter has producedcrucibles containing traces of silver and gold,and there is evidence from Cirencester formelting gold scrap.The afternoon ended with talks on whatscience can offer the small finds researcher.Janet Lang, formerly of the BM and nowdoing post-graduate research at ReadingUniversity, explained the different techniquesof radiography and metallography, and wheneach is best applied. In terms of time,radiography is generally quick€r, especiallysirnple X-radiography, ds a surface has to beprepared for metallographic examination.The most sophisticated X-radiography is'real-time', ifl which the object can bescanned from various angles in a lead-linedroom, with the pictures observed in aadjacent room on a TV monitor and then themost useful views captured on filrn. Evenmore sophisticated (and thereforeexpensive) is the use of neutron radiographyto look at organic matter, particularly whereScreened by another material, such as metal.Metallography consists of examining themicro- and macro-stntcture of objects todiscover how they were made. The preparedsurface can be viewed using either anoptical, or a scanning electrofl, microscope'The optical microscope is used to look atlarge objects that cannot be put into an SEM,or where inlays are preseot, as these can bedamaged in the vacuum of a SEM.lanet then showed how these techniques hadbeen applied to specific objects, such as thepanther and spoons from the Hoxne hoard,and the grave goods from Mound L7 atSutton Hoo. Radiography is useful forshowing the details of construction (werevessels made by raising or by soldering partstogether), decoration (including decorationlost from the surface but perhaps surivingas traces in the core of the object, and initialworking (eg hammer marks). Metallographycan be used to distinguish between obiectsthat have been worked and annealed or cast,and can pick up where an object might havebeen partly cast or partly worked in someway {eg raising). Plating on a obiect showsup well on X-ray, but a micrograph willprovide more detail. For her final example,she shoyved the sword from Mound L7,31

ConferencesRescue open me{eting 23'd February, London.Theme: English Heritage & Archaeology. Guestspeakers: David Miles and Lord Redesdale.Venue: Museum of London. Further informationfrom Rescue The British Archaeological Trust,15a Bull Plain, Hertford, Hertfordshire SG14 lDX.Tel 01992 553377. Emait: rescue@rescuearqh a egl ggy.freeservq . co. u kShaping understanding: Form and order inthe Anglo-Saxon world, 4OO-11OO7th-9th March, LondonColloquium at The British Museum. Subiects coverthe visual arts, texts, scripts, languages, religiouspractice, and social structures. Further details:Leslie Webster, Department of Medieval andModern Europe, The British Museum, GreatRussell Street, London WC1B 3DG. Tel 020 73238209. Email l.wehstefQbritish-museum,ac.ukCantate domino: Vicars choral colleges inEngtish cathedrals 22"d-24th March, YorkVenue: St William's College, York. Theme:archaeological, architectural and documentaryevidence from all the English cathedrals withVicars Choral Colleges. Speakers include: WarwickRodwell, David Stocker, Tim Tatton Brown.Further details: Vicars Choral Conference, YorkArchaeological Trust, 13 Oglefofth, York YO1 7FG.Tel 01904 663000. Fax 01904 563024. Email:enqu i ri esQ.vorka rchAeolqgy. co . u kIFA annual conference for archaeologists25th -27th March, LeicesterIFa's 2AOZ conference will be held at theUniversity of Leicester. Title: Setting standards inarchaeology. Further details: ConferenceCommittee, Institute of Field Archaeologists,University of Reading, Z Earley Gate, PO Box 2,39,Reading RG6 6AU. Tel 0118 9316446. Email:admin . ifa,@virgin . netSouth East of England regional industrialarchaeology conference 23'd March, CranbrookThe Association for Industrial Archaeology's SouthEast of England Regional Industrial ArchaeologyConference (SERIAC), organised by the MedwayIndustrial Archaeology group. Venue: CranbrookSchool, Kent. Fufther details: Bob Barnes, 3Vespers Cottages, Cage Lane, South Marden,Ashford TN27 8QD, tel 0L233 770355.Computer applications and quantitatiYemethods in archaeology 2nd-5th April, CreteVenue: Heraklion, Crete. Main theme: The DigitalHeritage of Archeology. Plenary and parallels6sions, with papers, posters anclclemonstrations. Further details: CM20O2,Institute of Computer Science, Foundation forResearch and Technology - Hellas, PO Box 1385,7tt10 Heraklion, Greece. Emailcaa2002@ics.forth.grEnlightening the British: Knowledge,discovery and the museum ln the 18thcentury 4h-6u April, LondonVenue: British Museum. 3-day conference settingthe intellectual, social and political context for thefoundation of The British Museum (250 years oldin 2003). Further details: The Director's Office,The British Museum, Great Russell Street, LonclonwclB 3DG.Email : directorate@thebritishmuseum.ac.ukCarbon-I4 and archaeology th-13th April,OxfordThe 4th C14 and Archaeology Symposium will beheld in Oxford, organised by the OxfotdRadiocarbon Accelerator Unit, ResearchLaboratory for Archaeology and the History of Art,University of Oxford. Further details: Dr TomHigham, Oxforcl Radiocarbon Accelerator Unit,Research Laboratory for Archaeology ancl theHistory of Art, 6 Keble Road, Oxford OXl 3QJ.Email orau@archaeoloov-research.oxford.ac.ukSymposium for Richard Reee 20s-21* lpril,Cardiff2-day conference to celebrate the contribution ofRichard Reece to archaeology and numismatic.Venue: Cardiff University, School of History andArchaeology. Will look at interdisciplanary themesinvolving numismatic and archaeology. Furtherdetails: Peter Guest, School of History &Archaeology, Cardiff University, HumanitiesBuilding, PO Box 909, Cardiff CF10 3XU. Tel 02920876538. Email GuestP@cf.ac.ukInternational mitlstone colloquium 16h-19sMay, La-Ferte-sous-JouarreColloquium to discuss the quarrying, working,trade and use of millstones, anct their long-termhistory. Further details: Mouette Barboff &Francois Sigaut, Maison des Sciences de I'Homme(Salle 115), 54 Boulevard Raspail, 55006 Paris,France.tnternational Counci! for Archaeozoology2OOz conference 23d-28b August, DufiamICAZ 2OO2 will be held at the University ofDurham. Proposals for sessions, workshops,papers and posters are currently invited. Furtherdetails: IC Z 2OO2, Dept of Archaeology,University of Durham, South Road, Durham Dl.l13LE. Tel 0191 374 1139. Email: ieazZ00?@durhqm.ac.uk. Website www.nmnh.si.edu/icaz33

Books etcThe Romaro-British small town atWanborough, Wiltshireby A S Anderson, J S Wacher, & A P Fitzpatrick.See Review on p L4lI6 of this issue. BritanniaMonograph L9,2001, 380 pp, 11 plates, L26 figs.836 (till 30.4.A2), then 844, including p&p.Order with the form enclosed in this issue.Roman Alcester 3, northern extramural areabyPBooth&JEvansFinal volume in the series, focussing on the areaof this small town which may contain an earlyfort. Impoftant assemblages of pottery,metalwork, glass, faunal remains, etc.CBA Research Report I27. ISBN 9A277 L 22 2.L77 figs, 28 pls. €35.Order from York Publishing Services Ltd, 64Hallfield Road, Lay€fthorpe, York YO31 7ZQRoman Catterick and its hinterland, 1 & 2by P WilsonThe results of over 40 years of excavation of thisimportant small town. Covers relationshipbetween civilian and military populations, andtransition form Roman to medieval. Largeassemblages of finds.Vol L, CBA Research Report 128. ISBN 1 94277I23 0. 832. Vol 2., CBA Res Rep I29. ISBN 190277L 24 9. 832. Special pre-publication offer ofboth volumes for 845.Order from York Publishing Seruices Ltd, addressabove.L'aftisanat romain: 6volutions, continuit6s etruptures (Italie et provinces occidentales)edited by M PolferContains 20 papers presented at the }ndinternational conference on Roman crafts held atErpeldange in October 2001. Coverc glass, woodworking,tanning, crafts represented ontombstones, metalwork, bone-working, Fofte?f ,iron.Monographies Instrumentum 20. ISBN 2 9A730361 9. 260 pp. €36 euros + 4.80 euros p&p.Order from Aditions monigue mergoil, LZ rue desMoulins, F 34530 Montagnac, France.Bibliographie Instrumentum 1994-2001M FeugEre et al.5,000 entries presented alphabetically within 33thematic chapters covering functional categoriesof objects and different types of craft production.The data also comes as an electronic databasewhich allows users (Mac and/or PC) to savesearches and sorted data in a personal folder.Monographies Instrumentum t7. ISBN 2 9A7 30352 X. 276 pp + CD-rom.Order from dditions monigue mergoil, address above.RFG CommitfeeChair: Roy Friendship-Taylor, Toad Hall, 86Main Road, Hackleton, Northants NN7 zAD.Tel : 01604 8703 L2. Email: roy@friendshiptaylo [, freeselve. co . u kMinutes & General Secretary: RichardHobbs, Portable Antiquities Scheme, 4IRussell Square, London WC1. Tel : O2A7 3238611. Email : rhobbs@britishmuseufn.ac.ukTreasurer: Christine Jones , 2 Otto Close,Kirkdale, London SE26 4NA. Tel : (work)A2A7 239 2574. Email: christine.e.jones@consio,niA.comMembership Secretary: Angela Wardle, 1Stebbing Farm, Fishers Green, Stevenage,Herts SG1 zJB. Tel : (work) A2O7 566 932,2.Email : awardle@museumoflondon.org.ukMeetings Co-ordinator: Ellen Swift, Schoolof European Culture & Languages, CornwallisBuilding, University of Kent at Canterbury,Canterbury, Kent CT2 7NF. Tel : 0L227827898. Email : e.v.swift@ukc.ac. ukPublications Co-ordinator: Jenny Hall,Museum of London, 150 London Wall,London EC2Y 5HN. Tel: (work) 0207 8145739. Email : ihall@museumoflondon.org.UkNewsletter Editor: Nina Cruffiffiy, 2 HallRoad, Copford, Colchester CO6 1BN. Tel:01206 210255. Email: nina.crurnmv@ntlworld. comStepping through time, archaeologicalfootwear from prehistoric times until 18OOby W Groenrnan van Waateringe, C. can Driel-Murray, O. GoubitzInvaluable guide to excavated footwear of allages. This is not only a dated typology, but alsocontains discussions of the technical and historicalaspects, and fashion. There is an extensiveglossary and hundreds of drawings.396 pp. ISBN 9070LA 44 69. €35.Order from Oxbow Books, Park End Place, Oxford0x1 lHN.Crossurord answersdue-l '02 ualv '61 6eq-eal '9Ilalf,eJg 'tI Jellene-l 'ZT &seuAp ueH '6 Alqelndsto '9OuttftoN 'S f,lpnulel '? papeeN '€ uel3 'Z elonb 'Iurnoo1d aOeJolS 'EZ lseneJg 'ZZ laloA 'TZ 'eq-eH 'gr lale3elg 'LT AtOutS '5I salqel 'EI uelnrued.IIlaad'0It{3le]'Sa6VuoJI.Lauolsulan}'Issorrv34