The Ecclesiastical Significance of the St

Home , Flabellum, Pallium

ECCLESIASTICAE TH L SIGNIFICANCE TH F EO NINIAN'T S S ISLE TREASURE VERE bTH y Y REV. MONSIGNOR DAVID McROBERTS, S.T.L., F.S.A.SCOT. THE discovery t Ninian'S e n 1958i ,th f so , Isle Treasur raises ha e d many questions which will doubto n , discussee b , vera r ydfo long time befor definitivy ean e answers e foundar n particularI . fundamentae th , l questio e precisth f no e nature th f o e hoard has been discussed right from the start. Is the treasure a collection of ecclesiastical silver, perhap churce th s h plat somf eo e pre-Norse Celtic monasteryt i s i , Or ? the treasure of some secular household? Or again, is it a mixed lot of silver, partly ecclesiastical, partly secular, perhaps loot burie somy db e raider? fargenerao e S 1,th l verdict seem havo st e bee rejeco nt e exclusivelth t y ecclesiastical charactee th f o r treasure, both becaus itemf eo s included amon e silvegth r pommee (sucth s hd a an l the three cones, which are regarded as attachments of a sword, and the penannular brooches, which are regarded as quite definitely secular ornaments) and because of items whic e lackinhar g (suc s chalicha d patene hoaran eth s ecclesiasticaf i di ) l plate.2 circumstancee Th whicn si hoare founds hth dwa , withi arepre-Norse a nf th a o e church and concealed under a cross-marked stone, seem to indicate a priori that we are dealing with an ecclesiastical treasure. The approximate date of the hoard, about the year 800, would fit in with the assumption that the hoard is ecclesiastical plate, concealed during the Viking raids, which were being carried out, about that time, agains islane th t d monasterie Celtie th f sco church purposy M . thin ei s articlo t s ei suggest tha carefua t l investigatio possible th f no severae th use f o s l categorief o s items founhoare th n di (bowls, brooches, cones, etc. establisho )t woulr fa e o dg th case for accepting the ecclesiastical character of the entire hoard. Taking each group of items separatel shon ca we thayon t ther possibla s ei e interpretation which makes each group fit quite easily into an ecclesiastical context for the whole hoard. THE Six DRINKING BOWLS The most prominent part of the hoard is undoubtedly the collection of eight silver f thesdiscussee bowlso b similae x n ear Si groupa ca . forn s i d ra e man th : other two must be treated separately, because they are distinct in shape and pre- sumabl hemisphericaf o s i usen yi e on : l shape othee hangina th , s i r g bowl. silvex si e r bowlTh s whic similae har shapn ri e differ slightly fro anothee mon n i r sizdecorationd ean . The eace yar h abou inchex si t diametern si , wit hbasa e formed by the indentation of the bottom of the bowl, and they are all decorated with patterns of punched dots. Quite clearly these are six drinking bowls: but, are they ecclesias- tica r secularo l ? Theoreticall coursef o , is t y,i possible that some secular household 1 This article is an abstract of a paper read before a meeting of the Society on I ith December 1961. For a full descriptio t Ninian'S e th f nso Isle Treasure Antiquity,e ,se xxxm, 1959 . 241-68pp , als d magnificene oan ,th t photographs of the various objects in Aberdeen University Studies, No. 141, St Ninian's Isle Treasure. discussioe Se 2 Antiquity,n i loc. cit. 302 PROCEEDINGS OF THE SOCIETY, ig6o-6l sent these bowls to be buried in the church in the face of a threatened attack by the sea-raiders. In such an assumption, however, one would look for other unmistakably secular gear in the hoard, such as silver mountings for drinking horns, like the Burghea r Trewhiddldo e examples r harneso , s mounting r coinso s t these bu , ar e quite absent from the hoard and, when we remember that a cross forms part of the decoratio e bowl f som th no alsd f an seo that these bowl e onlgroue ar s yon f po object collectioa n si n where everythin shale w gs l elsa see , eis , capabl ecclesiasn a f eo - tical interpretation, the t becomeni s clear tha seculae th t r interpretatio thesf no x esi

FIG. I. Graffit waln o Materf o l Matuta temple, Rome, showing typef so chalice, after Cabrol-Leclercq n (ii), fig. 1894 (copyright Letouzey & An6) drinking vessels raises more problems than it solves. On the other hand, the diffi- culties are all largely solved if we accept that these bowls were buried here because they were part of the equipment of the monastic community of St Ninian's Isle. But if the bowls are ecclesiastical, what was their precise use? It is very unlikely that they were drinking bowls fro e monastimth c refectory: the e obviouslyar y much more important than that. Mad silvef eo skilfulld ran y decorated; their design including, in some instances, the cross; and buried here with other silverware that has a liturgical connection t seem;i s difficul escapo t t conclusioe eth n that thes silvex esi r bowls were six chalices used by the clerics of the St Ninian's Isle monastery for the euchar- NINIAN'T S E 3 TH 30 S ISLE TREASURE istic liturgy. It is true that these bowls do not have the normal shape that chalices assumed during the Middle Ages but that is not an insuperable difficulty since it is known earl e thatth yn i , centuries typeo tw chalice,f so s wer usen ecalixi e th , shape, with cup, ste basd man e (which becam universae eth l typ medievan ei l Europed an ) thepoculum shape desige th , whicf no base s ordinare hthawa n th d o f o t y drinkinp gcu onld sixthe typeha o seee yth b d chalictw f bow n -ni so o based an t e an e l Th e.ar century eucharistic graffito (fig. i) discovered by Father Hartmann Grisar, in 1894, on the wall of the Mater matuta temple in Rome.1 Thepoculum type of chalice became obsolete in the early Middle Ages, but it did exist and these six silver bowls from St Ninian's Isle, decorated with crosse buried san d among other liturgical equipment, are almost certainly eighth-century examples of this type of chalice. The fact that no patens were discovered with these six bowls in no way invalidates the conclusion that they are chalices because, up to about the ninth century, patens were not in general use with ordinary chalices. The small number of elaborate patens that have survived from the Dark Ages (or which are mentioned in documents) are all of them patenae ministeriales or patens used in conjunction with the large and elaborate ministerial chalices, which were used at Easter and other great festivals to com- municate a large number of the faithful. E HEMISPHERICATH L BOWL e nexTh t consideree bowb o t l alss d i t differsilve f i o o t bu rs fro othere mth y sb being completely hemispherical and by having no base. Since it has no base, it cannot have been used as a chalice and, having no escutcheons for cords or chains, t cannoi regardee b t hangina s da g bowl. o pointTw s emergy consideratioan n i e f thino s hemispherical bowl. Firstly, neithes sincha t ei r bas r arrangementeno hangingr sfo t musi , t have been placen di some kin stanf do hand e attendann helr th da o f n dsi o t whesecone th use n i n I d. place, its design of intersecting circles deliberately emphasises an equal-armed cross patee in the bottom of the bowl, which clearly indicates that the bowl had some liturgical or ecclesiastical use. A bowl, similar in shape to the hemispherical bowl from St Ninian's Isle, was found in a Viking grave at Nordre Kaupang in Norway, in 1951, and is described Aslay b k Liesto Actan i l Archaeological fac e spitn I th t f thaeo t Liestol speake th f so Nordre Kaupang bow hangina s a l g bowl simpls i t i , yhemisphericaa l bowl wito hn traces of any escutcheons. The remarkable feature of the Nordre Kaupang bowl is runie th c inscription which some owne scratches rha sides it n :d o ' I MUNTLAUKU - 'I ' n. . [the. ] hand-basi ' . Thi . n. s inscriptio presumabls ni y incomplett i t ebu doe thisf o indicat se typus bowlf e esimilao e eth Th . t Ninian'S r s Isle bowl would then be a hand-basin but would be intended for use in the liturgy and it would be used in conjunction with a ewer (perhaps of earthenware): the basin and the ewer would be held by some assistant, perhaps a deacon, when they were used for the

washin handliturgye e th th e miracle-storf n go i sTh . t ColumbaS f yo , actins ga Illustrate1 described dan Civiltaa L n di Cattolica, 16 ser. ix (1897) . 721-31pp , . 2 Acta Archaeologica, xxiv (1953) . 163-70pp , . 304 PROCEEDINGS OF THE SOCIETY, Ig6o-6l deacon to St Finnbarr in the celebration of the Holy Eucharist, bringing a jug (urceus] f wateo r fro e sprinm th d settin gan t gbesidi e altath e r (wit s contenthit s changed into wine), probably ewee refer th waterf o ro t s , which woul usee db n di conjunction wit hbowa l suc this ha s hemispherical bowl fro t Ninian'mS s Isle. E HANGINTH G BOWL seriee th e las f bowlsf so o Th t , discovere t Ninian'S e th n sdi Isle hoarde th s i , important silver hangin gfirse bowlth t s completi t :i e hanging discoverebowe b o t l d n Scotlane i verth yd firsan d t probabl s originae founit b n o yi t d l plac f useeo . Though most hanging bowls have been discovere n pagai d n Vikin d Saxoan g n burials, archaeologist agreee sar d that these bowls originat Christiae th n ei n Celtic area from about the fifth to the ninth century. All agree that these bowls had some ecclesiastical significanc t whaebu t their precise functio quits i s e nwa unknown . Aslak Liestol article th n i ,e referre aboveo dt ,suggestioe taketh p su n that hang- bowlg in s were water container washine th r sfo handf liturg e gholr o th fo yn r si y o Suttoe water th bowfise o n hi nTh Ho .l lends some colou thio t r s theory. Liestol's arguments are inconclusive and certainly the arguments based on the Nordre Kaupang bowl prove nothing, for that example is not a hanging bowl. Dr Francoise Henry, on the other hand, adopts the opinion that hanging bowls were used as lamps, or as reflectors of lamps, in Celtic churches.1 This opinion leans heavily on the identification as a lamp of the curiously-shaped tenth-century hanging bowl from Ballinderry crannog. The Ballinderry 'lamp' seems to be some kind of sieve for straining and pouring liquid and not everyone would agree that it is a lamp. But, apart from such evidenc Ballinderre th s ea y 'lamp' might provide, thero n s ei f permaneno e evidencus e th tr lampchurchee efo th Celtie n si th f cso provincef so Christendom. What evidence there is suggests that these churches were lit only during services by lamps carried by those taking part in the functions. There is the quain tcleri e storth f cyo Lugidius, falling asleep durin t ComgalPgS s mas Easten so r nigh lettind an tlamp e gth , whic carriede hh , fall int newle oth y consecrated Easter water, to St Comgall's great annoyance.2 That the important monastery church of lonwithous awa t Columba e timS lama tth f eo n pi , mus deducee b t d froe mth account of the saint's death as given by Adamnan. After Columba collapsed before altare th , Diarmai darknesgrope o t th d n ei ha tfino st d him, 'becaus brethree eth n had not yet come in with their lamps'. Unlike the descriptions of Continental churches by Venantius Fortunatus or Paulinus of Nola, Cogitosus, in his description churce ofth Kildaref ho nothins abouy ha , sa o gtt lamps illuminatin buildinge gth . Speakin generaln gi , therevidenco n s ei e whatsoeve r hanginfo r g bowls having been used in the churches of the Celtic lands as liturgical water-vessels or as lamps. A third possible explanation might be advanced: Celtic churches were influenced by ecclesiastical customs prevailing on the Continent and, during the seventh, eighth and ninth centuries, it was fashionable to present votive chalices to be hung up in churches: the Roman basilicas had great numbers of these votive chalices (the gifts of successive popes) and the palatine churches of Carlovingian princes also led J.R.S.A.L, LXVI (1936) . 209-46pp , . Sanctorum,a Ad . 345 p , . IV August.e di , I . t , 1 2 THE ST NINIAN'S ISLE TREASURE 305 thin i s fashion. Ther doubo n s ei t that report f thesso e rich votive gifts were duly brought back, by pilgrims and travellers, to the Celtic lands, and manuscripts, like those emanating fro coure mth Charle f o t Bale sth d (PI. XXIV, i)carver o , d ivories, like the panels of the Pola Casket (PI. XXIV, 2), showing the interiors of churches, with hanging chalices and vases, would help to spread the fashion. The hanging bowls of Celtic churches may simply have been votive gifts, presented to these churche accordancn i s e wit populae th h r Continental fashion n sucI . ha theory , hanging bowls would have, generally speaking utilitariao n , n purpose t woulbu , d simply express the piety and generosity of the donor. Such a supposition would, however, give some explanation of the enrichment of the insides of these bowls (as in the Sutton Hoo example or the Lincoln example from the River Witham), which makes e bowlth s more elaborat d costlan e s votivya e offerings t whicbu , h renderee th d bowls much less useful for any practical purpose. But, no matter what theory one accepts as to its function, there is general agree- ment that the purpose of the St Ninian's Isle hanging bowl was ecclesiastical and accordingly its presence here in no way invalidates the claim that the hoard is ecclesiastical in character THE BROOCHES The next important group of objects found in the St Ninian's Isle hoard are the penannular brooches, one large brooch and eleven smaller ones. These brooches are generally regarde s purela d y secular ornaments t thibu ,s assumptio t valino ds ni because, whil survivine eth g Celtic brooches giv theio clut o en s era real character, whether purely secular or ecclesiastical, there is sufficient literary and monumental evidence to prove that, at least in some instances, these brooches were ecclesiastical ornaments. legendA , preserve Booe th f Lecan ko n di , purporte th w givo t sstore ho e th f yo Delg Aidechta, testamentare th y brooch abbote th f lonf o , so a originated. According to the legend, Columba acquired the brooch from Pope Gregory the Great and it was worn ever after by the coarbs of Columcille at lona as part of their official insignia as heirs of the founder-saint. One of the bas-reliefs on the tenth-century Cros f Muiredaco s t Monasterboicha e show e 'Arresth s f Christo t ' (PI. XXV, i). Lik othee eth r sculpture than so t cross t depicti , s contemporary costume solo -tw e th : diers wear tenth-centur figure y th dres Chrisf d eo san dresses i t ecclesiastin a s da f co perioe th doub o abbotn da (n s mose ta ,th t important Celtifigure th n eci ecclesiastical world), he wears the tunic and the cloak, or amphibalus, and the cloak is fastened on the breas larga y b et penannular brooc antiquf ho e form. This conceivably represent contemporarsa y abbot, wearin testamentare gth y brooc monasterys hi f ho , just coare asth f Columcillbo t lonea a wor Delge eth Aidechta. e DelgTh Aidechta e abbotth f f lono o s s sufficienai shoo t t w that penannular brooches are not necessarily and exclusively secular. In fact, without straining the evidence unduly mighe on , t suggest tha timportane somth f eo t brooches, like eth Tara Brooch or the Hunterston Brooch, may have survived simply because they were ecclesiastical relic monastif so c founders. Giraldus Cambrensis, writin Itinerariums ghi 6 30 PROCEEDING SOCIETYE TH F O S , ig6o-6l Kambriae in the twelfth century, is almost certainly referring to such a brooch, preserve relia fifth-centura s f da co y saint: 'Further I mus, t omino t t mentioe th f no circlet (torguis), which they say belonged to St Cynawg, which is indeed like gold in weight, in quality and in colour. It extends in four sections, of pierced work (ex rimulis), all set out in the form of a circle and skilfully linked together. It is fastened across the middle by a kind of dog's head with snarling teeth protruding outward from its open jaws. The people regard this as one of the most potent of relics; so much so that no one would dare to commit perjury with his hand on this.' It is certainly not impossible that the large penannular brooch in the St Ninian'1s Isle treasur officiae th s ei l brooch Delge th , Aidechta, abbote t Ninian'th S f f o so s Isle. elevee case th f th eo n n I smaller penannular t Ninian'broocheS e th f so s Isle treasure, ther s alsi e o sufficient manuscrip d monumentaan t l evidenc o suggest e t that these brooches were accessorie liturgicae th f so l vestment sclerice worth y f nsb o the St Ninian's Isle monastery. A well known Dark Age monument from Inver- gowrie shows three clericliturgicae th n si l vestur periode th grou e f eo Th . p probably represents an important cleric, bishop or abbot, visiting the parochia of his diocese or monastery, bearing some relic of the patron saint (it might even be a brooch-relic), supported an lesseo tw ry dclericsb , presumably deacons accordancn I . e wite hth custom of the time, all three wear chasubles, but the deacons' chasubles are fastened at the shoulder by cross-marked disks. These disks appear on other monuments; we see them, for example, on the reliquary called the Corp Naomh, in the National Museum of Ireland, or on Stone No. 29 at Meigle. The fact that these disks are stylised brooches becomes clear whe examine nw e closel vestmente yth s wor thy nb e angels depicted on the Chi-Rho page of the Book of Kells (PI. XXV, 2): these angels dressee ar deacons da wead san r chasubles conventionalisee th , d fold whicf o s e har transformed int ocolourine wingth diskse d th san f g,o fastenin garmene gth e th t a t shoulders, shows quite unmistakably that these round objects are really circular brooches with a transverse pin. In view of this evidence from the monuments, one can readily accept that these smaller penannular brooches of the St Ninian's Isle treasure were accessories of the vestment deaconse useth monastie y th d b f so c community when they assistee th t da liturgy. However, since we do not have the complete treasure of the St Ninian's Isle community cannoe w , t argue fro numbee mth f theso r e brooche numbee th o st f o r deacons that might liturg e assismonasterye th th t a f ty o thit sBu .theor y does explain why there are so many of these smaller brooches and also why they are hidden away with the liturgical equipment of the monastery. THE CHAPES The two chapes discovered in the St Ninian's Isle hoard have been fashioned so mountee b ao st somn do e thin materia thid an sl fac possibilite t th rule t sou y that

1 I am indebted to Dr W. Douglas Simpson for this interesting reference. It may be useful to give the Latin descriptio e broochth f no , whic s somewhahi t involve d obscurean d : 'Praeterea torquee d quam Sancti Kanauci dicunt tacendum non censui. etenimEst aura tarn pondere naturaet quam colors simillima quattuorex frustris sicut ex rimulis patet orbiculariter in invicem insitis artificiose conserta et capite quasi canino rictuosis hinc inde dentibus extante per medium complexa.' (Giraldi Cambrensis Opera (Rolls Series), vi, pp. 25-26.) THE ST NINIAN S ISLE TREASURE 307 they may have been intended as mountings for a leather scabbard. The suggestion s thereforha e been made thachapese th t were stole-end thid san s interpretatios ni certainly reasonabl probable vien eth i f wo e liturgical characte othee th f ro r items in the hoard. It is to be noted, however, that the two chapes differ, to some extent, sizdesign d i ean n and, somewhas whili e eon t worn othe e minn i th , s ri t condition, being a replacement, presumably, of a worn-out older piece. That being so, we cannot envisage these two chapes as stole-ends for a stole of the usual type, such as the early example from St Cuthbert's tomb. It remains possible, of course, that such asymmetrical chape havy sma e decorate ende pallium-shapea dth f so d stole, where the two ends are not seen together. However, considering the strap-ends, or girdle- ends, which occu othen i r r hoards ,somewhae sucth s ha t later Trewhiddle hoard, t seemi s much more likelt Ninian'S chapeo e ytw th thaf e sso th tIsl e treasure ear ornamental ends for the girdles that would be used, to fasten at the waist, those albs or tunics, with richly embroidered hems, which are depicted on such Pictish monuments as Rosemarkie No. 4 or Meigle No. 24 and No. 29. E COMMUNIOTH E NKNIF TH SPOO ED NAN e spoo pronged Th nan d instrument discovere t Ninian'S e th sn di Isle treasure have been identified as a communion spoon for the administration of the Eucharist under both kinds and as a knife for the elaborate Fractio panis, which the Celtic liturgy shared wit e otheth h r Gallican liturgie f Westero s n Europe o otheN . r examples survive of such eucharistic knives or spoons from the Celtic liturgy but, even without such comparative material identificatioe th , n seems reasonable because a spoon for the administering of the intincted host would be a normal part of the church plate of any church which administered communion under both species in pre-medieval times. Also, some kin kniff do e woul necessare db complicatee th r yfo d Fractio panis envisaged in the Stowe Missal tract. In the general liturgical context of this collection of silverware, the identification of these two articles as communion Fractioe knife th th r s efo spoo a panis d highls nan i y probable. e particulaOn r e communioaspecth f o t n spoon claims attention e curiouth : s eucharistic symbol of the dog's head licking the spoon which is incorporated in the design. The origin and significance of the symbol are easily guessed: the symbol surely derives from the gospel-incident, where the Syrophoenician woman claims that 'the dogs under the table eat of the children's crumbs' and it suggests the humility with which the Christian should approach the Holy Eucharist. While this is evident, it has to be stated further that this particular eucharistic symbol is quite unknow rese Christendof th o t n ni m and, thagreaf o s ti t bein t i interes , gso fino t t d that this extremely unusual eucharistic symbol occurt anotheye n si r instance th n ei iconograph Darf yScotlande o kAg . The other instance of the eucharistic dog-symbol is to be found on the splendid cross-sla t Niggba Ross-shirn i , e (PI. XXVI), whic roughls hi y contemporary with the St Ninian's Isle treasure. The carved group in the triangular pediment of the Nigg monument looks t firsa , t sight, just morlike eon e occurrenc Paue d th f an l eo Anthony legend, so familiar in the Dark Age monuments of the Celtic provinces of 308 PROCEEDINGS OF THE SOCIETY, ig6o-6l Christendom closen o t r bu ,examination t prove i ,somethin e b o t s g quite different. elemente th Pau e l Anthond th Al an f l so y stor there t yar the transformee ebu yar d into something new ,symbolia c representatio concelebratioe th f no massf ne o Th . Holy Dove bear Eucharistie th s c Bread, bowineithen o priestso w rtw g lo sid e ,ear with their missals, concelebrating mass t mighI . t almos sculpturea . be t d representation of that incident in Adamnan's life of Columba, where Columba invited Cronan, the bishop, 'ut simul quasi duo presbyteri dominicum panem frangerent' - that together, as two priests, they should break the bread of the Lord. And the two animals, which appear in the composition, are no longer the two lions of the Paul and Anthony dogso legend r 'thtw fo ,t e dogbu , children'se undeth f table o t th r eea s crumbs't I . is quite clearly a repetition of the same dog-symbol that occurs on the St Ninian's Isle communion spoon. THE POMMEL AND THE CONE-SHAPED OBJECTS The exact nature of the remaining four objects in the St Ninian's Isle treasure bees ha n somethin mysterya f go suggestione Th . tha silvee th t r pomme thas a i l f o t sword and that the three silver cones are ornaments of a baldric, has not received unanimous approval mere th t e possibilitbu , y that these objects migh relatee b t o dt military accoutrements has been a stumbling-block to the acceptance of the whole

TTMflANTlAINPACEQVl ViX ITANNQX

FIG Representatio. 2 . n offlabellum Catacombe th n i Santf so ' Agnese, after Cabrol-Leclerc q(ii)v , fig. 4467 (copyright Letouzey & Ane) hoar collectiona s da ecclesiasticaf o l plate. However, ther anothes ei r interpretation of these objects which explains satisfactorily their peculiaritie alsd soan their presence her thin ei s hoar liturgicaf do l plate. Basin investigationr gou representationn so n si the contemporary Book of Kells, it is possible to demonstrate, with reasonable probability, that these silver objects are some fittings of a liturgical fan or ftabellum. greao tw t e ostricTh h feather fans which nowaday carriee ar s d whe e popnth e goes in solemn procession to the Vatican Basilica, are the sole modern survivors of the liturgical fans, which were in universal use throughout Western Christendom prior to the Middle Ages. To protect the elements of the Eucharist on the altar, it was the normal custom, in early centuries, for a cleric to stand by, during mass, drivo wavint n efa ofga f flied othean s r insects. Representation f theso s e fansr o , flabella, occur regularl Christiae th n yi n monument Europef so exampln A e . th n ei Catacomb Santf so ' Agnese (fig show) 2 . normae sth l ancienforn a f mo a ts fanha t i : long handle with a disk at the end, presumably a piece of parchment stretched on a circular fram d paintean e d with some design, whic liturgican hi l example s frei s - THE ST NINIAN'S ISLE TREASURE 309 quentl Chi-Rhe yth crossa r oninth-centuro A . y example, carvebaldacchine th n do o of an altar in Perugia (fig. 3), illustrates a smaller type with a short handle: and we can point to a particularly barbarous representation, carved in 725 on the altar of the abbey-churc f Ferentilloho Umbrin i , a (fig. 4); this shows three flabella with long handle whad san t appea basee b whicn o st o r h thestann yca d upright. e churcTh Darn hScotlani e kAg Ireland dan d use same dth e liturgical equip- menn identif ca s elsewher a te w y d soman e e representations of these eucharistic

FIG . Representatio3 . n offlabellum baldacchinn o altarf oo , Perugia, after Cabrol-Leclerc q(ii)v , fig. 4468 (copyright Letouzey & An6). monumentsn flabellaow r somn o ou well-know e f eo Th . t PetenS r Ston t Whithornea , dating fro sevente mth h century (PI. XXVII) displays aflabellum dise whicf th ko hs i decorate dhandla wits Chi-Rhbase d t unlikha h th et an ei no almos e d eth oan t contemporary examples from Ferentillo. Square flabella, also decorated with the Chi-Rho t witbu ha differen t typ f handleeo carvingo , e Isloccu tw f Raasath eo n n o ro s y 3IO PROCEEDINGS OF THE SOCIETY, ig6o-6l (PL XXVII); and Irish examples are to be seen at Inishkea North, Co. Mayo, and at Reask, Co. Kerry. Other monumental examples might be cited, but we turn to manuscripe th t evidence whic eves hi n clearer. It has been suggested that the Book of Kells was written, or at least begun, in lona t whethe writtes bu , wa t t Kellsloni n ra ni r ao t show i , ideae th s s currenn i t the Celtic churc perioa t ha d fairlt Ninian'S e y tim e th clos th f seo o et Isl e treasure

FIG . Thre.4 e flabella represente abbey-churc e altan dth o f ro Ferentillof ho , Umbria, after Cabrol-Leclercq v (ii) fig, 4470 (copyright Letouzey & Ane) and one thing that the Book of Kells does show very clearly indeed is the prominent place occupied in liturgical ceremonial by the flabellum. In the Madonna and Child page of the Book of Kells, for example (PL XXVIII), foue th r attendant angels hold flabella thein i r hands.1 These flabella e colourear d yellow and grey, with designs indicated on the disks in green and pink, which probably means that the artist had in mind flabella whose handles and frames would be of gilded wood or metal or of silver, with painted parchment stretched on the circular framewor diskse e bottoth f Th .ko m right hand ange than o l t same page carriesa floriated flabellum, coloured yellow to represent gilded wood or metal. This latter type of flabellum, madrea of conventiona or l eup l leave featherssor , occurs regularly Celtin i c art t Petereleventh-centure S : th n i , y Breac Maodhog, clerice e th th n d o , an book-shrine of St Molaise, or the two ecclesiastics, on Cross-slab No. 11 from St 1 These object usualle ar s y described commentatory b , iconographe th n o s Celtif yo c manuscriptd san Pictish stones 'sceptress a , t the quite bu ' yar e certainly liturgical flabella. T NINIAS E NS TH ISL E TREASURE Vigeans (fig. 5), all carry these floriated flabella. Other shapes of flabella occur: in the Temptation page of the Book of Kells, the ecclesiastic who stands in the door of the temple carries crosswise two flabella with heads formed like marigolds. The angel- symbol of St Matthew, on fol. 5r of the Book of Kells (PI. XXIX), carries zflabellum,

FIG . Cross-slab5 . t VigeansS , ; ecclesiastics holding floriated flabella whic greea s hnha disk (presumably representing enamelled meta colourer lo d parchment), with gold balls fixed to the frame of the disk. This last type offlabellum reappears on fol. tagv of the Book of Kells (PI. XXX), wher flabellao etw thif o s design, arranged crosswise, for backgroune mth eaco dt f ho symbole th foue th rf so evangelists . These pictures of flabellaBooe th Kell f kn o i e sar contemporary with the St Ninian's Isle treasure and they provide us with the clue we 312 PROCEEDING SOCIETYE TH F O S , ig6o-6l

FIG . Reconstructio6 . n offlabellum, t Ninian'S s Isle; surviving fragments (coned an s pommel) show blacn ni k THE ST NINIAN'S ISLE TREASURE 313 requir explaio et three nth e cones, treasuree founth n e Celtidi Th . c flabella depicted foln o . Booe 12gth Kellf f kvo o s hav serieea conef so scircula e fixeth o dt r frame th f eo disk and, attached by cords or chains to some of these cones, there hang bells or tassels. It will be seen at once just how easily the cones found in the St Ninian's Isle treasure fit into place in the make-up of one of these flabella and it seems evident that they simple ar y thre seriea f eo conef so s that once adorne circulae dth f o r e framon f o e these eighth-century liturgical e fansconeTh . s have eac a hbas e witsloto tw hs through which they coul fastenee db straa y r thondb po g ont diskframe e oth th f . eo Each con alss eha o holes perforate sides it whicy sb n di chaie hth r corno d carrying belle th r tasselso s , see foln no .Booe isg th f Kells f ko vo , coul attachee db e th o dt cones e pommeTh . fito sto l into suc htheorya legitimats i t i : assumo et s e thawa t i t used in conjunction with the cones since it was found in association with them, like silvef theo s i rmsimilas t i gilt ha t i , r decoratio showd nan same sth e amoun wearf o t . reasonabls i t I supposo et e tha attachmenn a t t suc this ha s silver pommel woule db added to the end of the long gilded wooden handle of \heflabellum to ornament it and protect it from wear. In the accompanying reconstruction of a Celtic flabellum (fig. 6), based on the examples showBooe th Kellsf ko n n i pommee th , conee thred th an lf seo have been shaded black to emphasise just how little of the fittings of the flabellum have actually survived in the St Ninian's Isle treasure. Several possible reasons might be adduced to explain why so little survived. It may be that, in the crisis of an imminent raid, some fittings were hurriedl t verno yd systematicallyan y torn off &. flabellum, whics hwa regulan i r use concealed ,an d wit othee hth r silver perhapr O . s these fittings wera n ei workshop flabellumpars a a ,f o t thabeins wa t g repaired beind an , g convenientlo yt hand wit x , othee wer bo ht th int e repu oth silver r againO .pommee e th th , d an l three cone havy sma e bee l thanal t remaine flabelluma f do preservethas wa t a s da relic of an early saint, perhaps the founder of the monastery, for flabella were preserved as relics in the Celtic church just in the same way as books and bells and bachul brooches.d san 1 In conclusion, I would sum up by saying that the only adequate explanation which provide coherensa t pictur contente th f eo f thiso s larchwoo unearthex dbo d on St Ninian's Isle is that it contains the church plate (or, more probably, part of the church plate) used by some eighth-century community of Celtic clerics for the eucharistic liturgy. In the alarm of an impending Viking raid, these clerics hurriedly gathered togethe concealed an r bowl-shapex dsi d silver chalices liturgicaa , l hand- basin, a hanging bowl or votive chalice, the abbot's brooch, eleven smaller brooches which were accessories of deacons' vestments, two ornamental girdle-ends, a communion spoon, a knife for the Fractio panis and a silver pommel and three other silver fittings fro meucharistia c flabellum. This theory give coherensa t explanatioe th f no whole hoard, with all its peculiarities, and there is sufficient evidence to make out a reasonably convincing case for such an interpretation. 1 The flabellum of Columcille was preserved among the saint's relics at lona down to the year 1034, when Annale loss th t seaa twa f Ulste t so .i Therw rothee ho tel ear f o l r reference flabellae th othef o t o s r Celtic saints, such as that of St Patrick at Armagh, which was still in existence, according to the Annals of the Four Masters yeae th rn i ,1128 . 314 PROCEEDING SOCIETYE TH F O S , ig6o-6l

SocietyThe indebtedis Universitythe to Aberdeenof providingfor blocksthe for figures i to 4 and 6 and for Plates XXIV, XXV and XXVIII to XXX, and theto following permissionfor reproduceto copyright material:The Board Trinityof College, Dublin (Plates XXVIIIXXV,and XXX); to 2 Messrs Letouzey et Ane (figures i to 4 and Plate XXIV, /); and The Mansell Collection (Plate XXIV, 2).