10 PROCEEDINGS OF THE SOCIETY, 1957-58.

II.

THE HUMAN HEAD IN INSULAR PAGAN CELTIC RELIGION.

BY ANNE BOSS, PH.D., F.S.A.ScoT.

INTRODUCTION.

A Scottish Tricephalos said to come from Sutherland.

Throug generosite hth Trusteee th Dice f yo th f kso Institute , Kilmarnock, importann a t carved hea kina t hithertf do dno o recorded from Scotlans dha been addeNationae th o dt lonlr Museum.fo g d beeha nt I exhibite1 e th n di Institute's museu earln a onle s mya th yfontd recor s provenancan , it f do e is '' Sutherlandshire.'' This head, 4-7 ins. high and 5-4 by 5 ins. broad, is shaped like a ball, truncated top and bottom. The top is hollowed into a conical cup about 3 ins. across and 2-6 ins. deep. The base is slightly concave. Three faces have been carved roun e sidesdth . Rounded incisions about J—. widin J e d abou an deptn i t. h^in outline three pair f eyeso s , chin droopingd san , triangular moustaches, whil lina e e continuous roun e stondth e indicates mouthse th . Broader lines -were use indicato dt innee eth r curvee th f o s cheeks and the sides of the drop-shaped noses. Between the faces there are similarly incised crosses, equal-armed, abou ins2 t . overall excep r onefo t , now damaged, abou. higin h1 t wit verha y uncertain horizontal strokt eno more than \ in. long. The ston granites ei , speckled black, whit pinkd ean . Mis . MacDonalsH d of the Geological Survey and Museum, South Kensington, has kindly examined it. She describes it as a medium-grained microline-granite and notes thavers i t yi t simila specimeno t r f biotito s e microline-granite from Norway. It does not resemble English or Irish granites in their collection, nor according to Miss MacDonald and Dr J. Phemister of the Survey's Scottish officet i Scottish s i lattee , Th r. doe t thinno s k tha a naturat l errati f Norwegiaco n granite woul foune db Sutherlandn i d . Some other continental origin is also possible as the Survey's collection of granites is not exhaustive. Unless the Sutherland provenance is incorrect, one may speculate whether the head was brought to Scotland by fugitives from the Romans, for a date 1 Acquire Johr M . Clarke nS y db , sometime F.S.A.Scot, fro ma Glasgo w antique shop before th e First World War. E HUMATH N HEA PAGAN DI N CELTIC RELIGION1 1 .

about the turn of the Christian era is possible for it, or whether possibly t i retaine s significancit d e sufficientl ye broughb lon o t g e t th ove y b r Norsemen.1

Scope of the Study. appearance Th thif eo s tricephalo pertinenr fo y s openwa t e questionsth s as to its place in archaeological chronology, its function and its artistic relationships. Alone, it can tell us little. Its three faces imply a religious significanc somf eo crossee eth sort d san , which tak place eth earf eo s would lead to the conclusion that the religion in question was Christianity. The drooping triangular moustaches are a typically Celtic feature, familiar to us from numerous example a TenL f eo s metalwor d certaikan n Bomano- British religious representations. This three-faced head divorced fros mit proper context can thus yield little specific information. It so happens, however, tha pagan i t n Celtic religious representative tradition attesten i , d Celtic social custom and in early Welsh and Irish written sources, the head figures naturalistically, or in bi- or tricephalic form, as an independent entity to a striking degree; to such an extent that it is possible from an examinatio more somf no th f eeo outstanding examples placo ,t e this Scottish tricephalos fairly confidentl originas it n yi l artisti religioud can s orbit. It is proposed to concentrate the following investigation of head symbolism on the evidence from Britain and Ireland. The Gaulish material is more plentiful, but the major examples have been discussed recently by Lambrechts, d aparan 2 t froe mth f o drawin o tw gr o attentio e on o t n continental monument r comparativfo s e t purposesproposeno s i o t t i d, study theo concludet mthin d i le s e context,ar mainle W . y froe mth Gaulish evidence, that the head was the bodily member which was especially venerate Celtse th y ,db tha heae regardes th t dwa seae they f dth b o tms a soule th centre vitae th ,th f le o essence , symboli regenerative th f co e forces of life. We have evidence from classical sources, fully substantiated by early insular vernacular tradition, that the head was prized by the as a war trophy. The fact that severed heads were impaled on stakes about their dwellings and temples implies that they had a dual significance for their owners, i.e. as irrefutable evidence of military prowess, and as amulets.

1 I am indebted to Mr E. B. K. Stevenson, Keeper of the National Museum of Antiquities for these3 kinr detailfo d d permissiosan publiso nt heade hth . 2 Lambrechts, Pierre, L' Exaltation Tetea l e d dans Penstea l danst e I'Art des Celtes, Bruges, 1954. " When they depart from battle they han heade gth f theio s r enemies fro e neckmth f thei o s r horse3 s . . . and nail the spectacle to the entrances of their homes." Strabo IV, 4, 5. (Loeb n, 247). "The heads of enemies of high repute, however, they used .to embalm in cedar oil and exhibit to strangers, and they would not deign to give them back even for a ransom of an equal weight of gold. ..." Strabo. 12 PROCEEDING E SOCIETYTH F O S , 1957-58.

The apotropaic significance of the human head is fully attested by the enigmatic antefixa from , Silchester, Dorcheste d Towcesteran r , where sculpted heads, or representations of heads in relief must have been quite common numerous a , s mould fragmentd san antefixf so a show.1 Their appearanc sucn eo h objects would serv streso et s their amuletic function. The frequency with which the head figures on Gallo-Roman religious monuments, both in the round and in relief, is sufficient to substantiate its religious connotatio givo t Gauln ni esatisfactord a an , y corpu evidencf o s e for insular comparison. The striking tricephalic heads from the territory of the Remi in particular illustrate the reverence accorded to the head. Here the is symbolised by an enormous bearded tricephalos, having a leaf-crown usualld an , y equated wit classicae hth l .2 These particu- lar representations would seem to testify to the concept of some autochthonous deity as a head alone, the head sufficing for the total being, the vital part, embued with the power of the whole. somn I e respects e studth , Celtif yo c religio evidences na d from i , facilitate wealte th y actuaf dhb o l religious representation classicay b d san l references which, although few in number, do contain some information which can help to guide the investigator through the pitfalls of interpretatio Romana. In Britain considerably less material evidence is available, and thi discouragingls si y scattere heterogeneoud dan mucd t seemingli s an f ho y unrelated large Th e. corpu f earlo s y Iris Welsd han h text hela n sca o pt certain extent, but only when the archaeological evidence has been examined and classified. A study of the confused corpus of religious representations d epigraphicaan l material does, however, indicate certain patternd an s tendencies which t uniquno f i ,Celtio et c religious idio temperamentd man , are distinguished by the emphasis laid upon them. One of these recurrent and predominant themes is the human head. It would, of course, have been possible to study the Scottish head from e pointh f vieo t f tricephaliwo c deitie generaln i s , s feli bii t i t that such representations and the concept that underlies them are but a single mani- festation of a more general cult of the head, and it -was thought more satis-

"When their enemies fall they cut off their headsand fasten them about the necks of their horse. . . s and carry them off as booty, singing a paean over them and striking up a song of victory, and these first fruits of battle they fasten by nails upon their houses . . . the heads of their most distinguished enemies they embalm in cedar-oi carefulld an l y preserv chesa thesd n ei an t e they exhibi strangerso t , gravely maintaining that in exchang thir efo s head someon theif eo rhimself n ancestorsma e ,th refuser o , e dth offer of a great sum of money. And some men among them we are told, boast that they have not accepted an equal weight of gold for the head they show. ..." Diodorus. (Loeb in, 173). " The consuls . . . got no report of the disaster till some Gallic horsemen came in sight, with heads hanging at their horses' breasts or fixed on their lances, and singing their customary song of triumph." . (Loe11 , 459), Liv26 b IV , .yX 1 P. 28 ff. . EsperandieuV 2 , Emile, Becueil General des Bos-Reliefs Statues t Busiese Gauledea l Romaine,, V pis. 3651, 3652, 3655, 3656, 3657, 3762, et seq. E HUMATH N HEA PAGAN DI N CELTIC RELIGION3 1 .

factory at this stage to regard it in the light of the wider cult of the head in Britain. British and Irish head representations can be divided up into several categories which will be examined according to their importance in relation- ship to the Scottish head rather than chronologically. The material falls into the following groups: 1. Sculpted heads set on bases, usually carved in one with the base or showing no trace of having at any time formed part of complete statues. These again can be sub-divided into: (a) Heads having attributes whereby themorn yca e readile yb connected with specific or religious cults, and (6) Heads similar to (a) but without attributes. 2. Head relien si altarn o f pillarsr so , havin t havinno r go g attribvitesr o , being brought into association with the attributes of some specific deity. It is not proposed to include here monuments which are clearly funerary portraits. Although heads, rather than busts tend figuree b o sepulchran t o d l monuments, spriny thisma g froe mth Celtic concep f deat o td decapitatio han s beinna g mor r leso e s synonymous actuae th d l an ,cul t significanc sucf eo h representations canno e demonstratedb t importane Th . t antefixa from Caerleon (Isca Silurum) will als discussee ob d here. 3. Heads figuring on metalwork objects having a ritual significance such s bucketsa , cauldrons, personal adornments d weaponsan , , which were originally venerated as being inhabited by independent spirits. These various representations "will be discussed and described in an attemp removo t Scottise eth h tricephalos from isolatio placd nan e it in the context to which it rightly belongs. As tricephalic representation comparativele sar y rar Britainn ei s a d an , they are the most striking distinguishing feature of the Scottish head, they will be discussed first.

. 1 Tricephaloi. , la (i) Sutherland (PI. III). Three-faced head, with hollotope th , wn probabli libationsr yfo , thus serving as a small, portable shrine. For full description of this head v. p. 10 &. above.

(ii) Corleck, Co. Cavan1 (PL IV). mose Th t significant hea relation di Scottise th o nt h head comes from Ireland. bees ha n t I worked fro mbloca sandstonf ko e int oheaa d having three facese th , features in fairly low relief, and bearing a striking resemblance to the Roquepertuse janiform heads.2 The eyes are wide and round, the mouth small, nose straight 1 Raftery, J., Prehistoric Ireland, 1951, fig. 263. 2 Esperandieu, No. 7616. 14 PROCEEDINGS OF THE SOCIETY, 1957-58.

chid an n describepointede b facee y Th ma sbeins . d a g without expressiond an , here they are clean-shaven. No neck is indicated. The head probably dates from the later La Tene period.

(iiiPouquelaie,a L ) Guernsey.1 A three-headed bust set on a gabled plinth, carved out of grey granite. The heighe monumenth f o t t is somfeaturee ins5 Th e1 .reminiscene ar s e th f o t Gaulish heads, but the Guernsey tricephalos is distinguished by the fact that there are three separate heads conjoined here, rather than the usual three faces on one headgenerae Th . l treatmen headse th e lacf o th f texpressio, k o e facth t d nan that they closely resemble Gallo-Roman works of the 1st and 2nd centuries, would point to some date immediately prior to, or during the Roman occupation of Britain for this tricephalos.

(iv) Risingham (Habitancum), Northumberland.2 A fourth British tricephalos which belongs in fact to Section II (Heads in Relief), but which it will be convenient to include here, is figured on a slab dedicated e Fourtbyth h Mounted Cohor f Gaulo t s " Numini(bus) Augustor(um) coh(ors) Till Gal(lorum) eq(uitata) fec(it)." "Te e augusspiritth th o f o st onese th , Fourth Mounted Cohor Gaulf o t s made bird-heade o this.Tw " d terminals enclose beardeda , pointed human face VictorA . y stand globa holdd n so e an wreatsa h or tore cranA . bira e( d which figure Celtin i s c mytholog fora r metas mya fo - morphosed women, and in connection with war) proceeds to the left, from three berries, while Mar depictes si d wit hgoosa e below him. Above either pelta s ai neckless human head, that to the right being heavily bearded and having prominent lefe th t beino t e gon ears beardede th , , moustache tricephalicd dan .

(v) .3 representatioe Th deita f no y having thre four eo r heads comes from Viroconium (Wroxeter). It seems to have formed part of the capital of the basilica. It is fashioned from a block of grey sandstone, and is 20 ins. long and 9 ins. high. unclassicas i t I l bot conception hi executiod nan s clearl i d wore a nan y th f ko native artist, representing a local god. The work is fragmentary, the lower part of the three faces being damaged, and it is not possible to tell whether the faces were bearded or not. The eyes are strangely slanting, as are those of a similar, singl t independenebu t head whic founs hwa d nea multiple-headee rth d monument, and may represent the same god naturalistically. This monument reveals a strengt vigoud han r whic perhapy hma comparee sb d wit htricephaloe thath f o t s from Condat.4

(vi) Llandinam, Monmouthshire.5 A tricephalos woon i d from Llandinam Church, Monmouthshire, carriee th n so early tradition. Three heads are carved in high relief from a block of wood, conjoine thao ds t ther foue ear r eye thred san e nosescentrae Th . l fac beardeds ei .

1 Kendrick, T. D., The Archceology of the Channel Islands, I, , 1928, pi. V. HistoryA 2 of Northumberland, , Newcastlxv e upon Tyne, 1940, pi. . facin80 . gp 3 Atkinson, D., Report Excavationsn o t Wroxeter,a 1923-7, . pi60 . 4 Esperandieu, 1316. Hemp, W. J., A "Grotesque" in Llandinam Church, the Montgomeryshire Collections, MV, 11, 1956, 5 pi. IV. E HUMATH N HEA PAGAN DI N CELTIC RELIGION5 1 .

The tricephalos apparently came to Llandinam Church without any provenance or history t hunI . g frosoute mth naia n hsoute o l t eas walth f no h yo l s aislei t I . to attach a date to this carving, and in no way does it resemble the medieval three-headed beings which were sometimes use depico dt Trinitye th t .

(vii) Ireland.1 A three-faced or four-faced head occurs on an ivory pendant now in the National Museum of Ireland. The pendant is shaped like a horn, terminating at one end in a duck's head, surmountea three-heade y b e othe d th en t r a d d male, having flowing moustaches beind an , g reminiscen little th e f ee th thea th f ko n do duck from Rotherley. e hea e bird'e stylth th s earldi f Th t o es2 ybu hea s di reminiscent of Viking art, and the object thus cannot be of very early date. The combination of stylized duck-heads, with human masks, is directly in line with the early Celtic tradition, the Torrs pony-cap providing one example. One may speculate whethe personara s thiswa l ornament wortalismana s na r useo , n di religious ceremonies. A three-headed figure occurs on another horn—the shorter golden horn from Gallehus thin I s. case ther three ear e separate head necksd san , these being conjoine neckse base th th f eo t . da

These then are the only tricephaloi as yet known from the British Isles which are likely to be early in date, namely, one from Scotland, two from Ireland (one of which is unlikely to be early), three from and (that from Llandinam being problemati froe Channe e date)mo on t th s d ca ,an l Islands Risinghae Th . m ston cleas eha r Gaulish connections Guernsee th , y head likewise would see mderivo t e artistically from GaulIrise Th h. head in stone has affinities with the Roquepertuse sculptures, and the Scottish head may also have Gaulish associations, in symbolism and mode of repre- sentation, if not of actual origin. The continental Celts may then have brought this religious concept with them when they came to Britain and Irelan pre-Romaa t da n e Risinghaperiodth d an , m altar would thue b s perpetuatin gtraditioa n already know Britainn i idee triplicatiof aTh o . n is essentially fundamental to early Celtic thought and outlook. Deities are conceived of as having three heads or three faces; sacred animals are given three rather thahornslatea o t ntw rA .stage e vernaculath , r tradition describes mythological or super-human beings as occurring in trios, bearing same th e name t witbu , h individual epithets, bein birthge bord on an t ,n a sometimes (lik three eth e son Uisliuf so ) meeting wit same hth e fate, usually decapitation. These triple characters thewely nlma reflect another aspect of the same fundamental, pagan concept of the magical or sacred power of numbee th r three. Apart from being triplicated the head was also duplicated, the power-of- the-twin concep doubo n t t underlying this mod representationf eo . There are several janiform heads from Britai d Irelandan n e mosth , t striking examples coming again from Ireland. 1 Catalogue Royale oth f Irish Academy, Dublin, 1857, 339. 2 Infra, p. 34. 16 PROCEEDINGS OF THE SOCIETY, 1957-58. . Janus2 , la Heads.

(viii) Boa Island, Fermanagh^ (PL IV). A fine Janus-faced statue, probably datin a TengL fro e mth periodn i t bu , common with several other religious representations from Ireland, all of which show early features and possibly date from the Iron Age, lack of chronologv makes more tha approximatn na e dating hazardous figureo bactw e e ar so kt Th . back bodiee th , s squar atrophiedd ean , clearly subordinate headse th o dt , with small, strap-like arms and traces of a waist belt, in the style of the statue-menhirs. The huge heads are triangular, the beards pointed, eyes enormous, noses long and narrow. The statue is strongly reminiscent of the "Maponus" head from Corbridge thid an s, would perhap guida earln a e sb o y et lattere datinth r g,fo which is perhaps pre-Roman both in art style and in execution.

(ix) Kilnaboy, Clare.2 This unusually-shaped Janus-faced monument has the heads placed chin to chin rathe rmor e thath en i usual bac baco kt k position neckse decoratee Th ar . d wit htore-lika e neck-band (wor gody warriors)d nb an s facee typicalle th ;ar s y Tena L stylen ei , their stereotyped "expressionless" feature reliefw lo .n si (x) Corbridge (Corstopiturn) .s An impressive Janus-faced head comes from Corstorpitum. It is unfortunately damaged, but sufficient remains of the lower part of the face to indicate that it e typicath wa n i s l triangular traditio e earlth f yno Celtic metalwork, broat da the brows, narrow and pointed towards the chin. It measures 8 ins. high by 8 ins. wide (a fairly typical measurement for such heads), and is fashioned from a coarse type of gritstone. The top is slightly curved, with a suggestion of hair arrange bandn di s running from bac eyebrowfronte o kt Th . prominene sar e th t brow furrowed eyee large deep-setTh d sar e. an nos e straighs Th ei . therd an t e are traces of side whiskers, a feature common to several heads from the region of Hadrian's Wall. The mouth is small, and from what can be seen of it, drooping. impossibls i t I whethey sa o et r bearthera s not r lowedee o wa th , facre parth e f to being completely destroyed. The reverse side of the stone is occupied by a second, smaller figure, more crudely worked tha firse nth t headfigure Th thas a ei . f o t man, and what remains of his body indicates that it was very much subordinated to the head. The chin is broad, the eyes huge, the nose short, and a moustache is clearly depicted. The mouth is small and severe. As is the case with the head on the revers e monument th sid f eo d witan ,h other heads from Corstopiturn described in Section 16 infra, one eye is markedly smaller than the other. In view e naturoth f f theseo e monuments s cleai t i ,r that this featur hardln ca e e yb fortuitous n explanatioa t bu , n doet comno s e readil o mindt y . Additional evidenc hely elucidato pet ma t soma t ei e future date t meanwhilbu , possibls eit e significanc mattea s ei speculationr rfo . Richmond 4flap o commenttw - e th n so like protruberances which come down from the side of the head to about ear-level, taked an s the mrepreseno t classicae th winge e f o th t t l dMercuryha . This seems 1 Henry, F., Irish Art, London, 1940, 46. 2 Baftery, op. cit., fig. 261. 3 Richmond . A.I , , "Two Celtic Head Stonn i s e from Corbridge, Northumberland, Darkn "i e Ag Britain, London, 1956,-11 ff. • ' ' . ' . cit.,. Op *12 THE HUMAN HEAD IN PAGAN CELTIC RELIGION. 17

extremely doubtful monumene th , t showin signo gn s whatsoeve influence th f o r e of interpretatio Romana, s ther i inscriptio y r ean no n which might poin thio t t s equation wit Romae hth n god. Moreover examination a , nativf no e representa- classicae tionth f so l Mercury fro generae mth l regioWale th f ln o reveal s thae th t wings of the petassus, where diverging from the classical mode of representation appear as almost horn-like protruberances, where Mercury seems to be becoming confused with his native, horned counterpart. It is also open to doubt whether this feature is really deliberate at all, or whether it is not simply an illusion created by the projecting pieces of stone on either side of the head which form part of the broad upper part of the first head. The extreme irregularity of the "flaps" would indeed suggest this to be the case. The head shows traces of horns. This head conjunction i , othero tw nr s o wit fro e same mhon th e localit associatee yar d wit hnativa e shrin whicn ei indigenoun ha s god, "Maponus worshippeds "wa d an , whose accepteculs wa t high-rankiny db g officerRomae th f o s n army dedicaA . - tion to this god is recorded from .1 The name survives in Welsh vernacular tradition as Mabon.

(xi) Kent.2 A janiform hea recordes di Horsely db illustrateo ygivewh d an s, sit Kens it s ta provenance wit furtheo hn r details difficuls i t I mak.o t t precisea e estimatiof no this head from Horsley's drawing. It consists of the head of an old man set bac baco kt k with younthaa f o t g man, the whol basea n neco o en , k being indicated. The young man has long hair and the old man wears some kind of helmet.

Another group of heads belonging to Britain, and being distinguished by some tangible attribute, is that depicting a horned or antlered god. Such a god-typ s well-knowei n fro mCernunnosd Gaulgo e th , furnishin e besth g t exampl thif eo s concept. Several monuments representin gsimilaa r deity are known from Britain and Ireland, but here, however, we shall be dealing maie th nn i with representationheaa s a d d alonego e .th f so

, 3la . Horned Pleads.

(xii) Neiherby, ? The first head in this group is the most dramatic artistically, and illustrates the cult of a horned god, in some of his aspects a warrior god, which "was widespread in Cumberland, eastwards along the Wall and S. into —roughly throughout the territory of the . The head is some 8 ins. in height, fashioned from red sandstone, square rather than triangular in shape, with a powerful, seemingly clean-shaven chin, flattened features, the eyes half-closed, the whole aspect grim and war-like. From the top of this impressive head grow two ram-horns which curve roun dowd dan n toward earse sth . From Bourbonne-les-Bains, C.I.L., xin, 5924.

1 2 Horsley, John, Romana, London, 1732, v. Kent, One may perhaps compare a full-length janiform figur unitee same ar th e en n dfroo ma m d Reimsol n a f ,o wherd younface an a f eth n eo gma body. (Esperandieu, p. 3666.) 3 In Tullie House Museum, Carlisle. I am indebted to Mr Robert Hogg, Keeper, for kind permission to publish a photograph of this head. 2 VOL. XCI. 18 PROCEEDINGS OF THE SOCIETY, 1957-58.

(xiii) Carvoran 1 (PL IV). An impressive head worked fro mlarga e piec f stoneeo e bac th ,f whic ko h seem o havt s e formed para building f o t thio s , s heahavy dma e been uses da something in the nature of an antefix. There is no neck, and the features are in fairl reliefear e w large ylo coarsd Th sar .ean placed unusualln ean a n di w ylo position. Horns grow from the top of the head. The eyes are huge, the nose wedge-shaped, the mouth straight, a moustache indicated by pick-marks. The head is fashioned of pale yellow sandstone 10^ ins. in height. It is reminiscent of cat-earee twth f oo d uniquheade th n so e Caerleon antefixa discusse infra.8 2 . p n do

(xiv) Lanchester.2 Small figure consistin impressivn a f go atrophien a en heao t dse d body, carved in the round. The head is horned, is bearded and has side-whiskers and the expressio phallice severes n b i figur e y t ,thTh bu ,ma e. worn conditioe th f no stone makes this problematic. Some circular attribute is held in the right hand.

(xv) Alnwick, Northumberland.3 Bearded, horned head, with usual large eyes, straight, expressionless mouth and wedge-shaped nose. The height is 7J ins.

(xvi) Chesters (Cilurnum).4 A horned head, fashioned from a block of stone, having a very thick, elongated neck which was probably set upon a base. The head is typical of the heads describe n thii d s section, having large eyes, straight nose d firman , , slightly drooping mouth, givin expression ga withdrawf no n severit featurese th o yt .

(xvii) Corstopitum.5 A very impressive head has large eyes, well-defined eyebrows, moustache, beard and sidewhiskers. The mouth and nose are straight. The head is crudely fashione neck o backe n seemd th s an , t havha o ,da st e been originally fashioned stano expressione baset a n Th do . generad an , l thicknes strengtd san necf ho k are reminiscent of the antefix, discussed on p. 26 infra. The head

shows trace havinf so g been horned. (xviii) Benwell () .e While it is not directly relevant to this study to trace the cult of a native horneBritainn i d monumene dgo on , perhapy ma t includee sb d here. Thia s si fine head fro little mth e templ eAntenociticus d dedicatego e th grounde o dth t n i s 1 Collingwood, R. G-., Roman Inscriptions and Sculptures belonging to the Society of Antiquaries of . Newcastle upon Tyne, 1926, No. 300. 2 Collingwood Catalogue, No. 273. In the collection of the Joint Museum of Roman Antiquities, Newcastle upon Tyne. Formerly part3 of the collection of Roman stones owned by the Duke of Northumberland at Alnwick Castle, it did not appear in the Catalogue published in 1880. As most of the Roman stones came from High Rochester or Risingham in Redesdale, it may have come from one of these sites, but this is not at all certain. Romae th n I n Museu t mChestersa .

Romae th n I 4 5n Museu t mCorstopitmna . 6 Collingwood Catalogue Simpso. G 302. . P Richmondd ,No . nan Roman e . A.,I Th , Fort Hadrian'sn o Wall Benwell,t a Archseologia ./Eliana, 1941, xix ff7 .,3 E HUMATH N HEA PAGAN DI N CELTIC RELIGION9 1 .

of Condercum House, Benwell, Northumberland foreara lea s gA d .werm an e found nearby muse w , t conclude tha heade th t , whic 12s hi | ins. high, originally formed part of an impressive statue with naturalistically-proportioned limbs. The head is of native workmanship, but classical influence is evident. The sculpture is of a good artistic standard, and the expression, although milder and less intense than is usual in the case of the purely native heads, is serious and self-contained. The face of the god is youthful and dignified, with the usual large eyes therd tracee an , esubstantiaar a f so l tore abou hainecke e s th tri Th . depicted in native fashion, and is plentiful, strands of hair falling from crown to brow in great profusion. The most interesting feature however, is the fact that two thick strands of hair turn back over the brow to form a decoration which one ca fairle nb y confident represents atrophied horns. This featur mors ei e clearly illustrated by a small human head in bronze from Thealby (v. p. 33 infra) where the hair is combed down on to the brow in native fashion, and a pair of atrophied horn e likewisar s e demarcate e browdth flan .o t For other example f thio s s feature see sections on Antefixa and Metalwork. The god wears a tore. Fragment thref so e altars have been temple founth n dAntenociticusf i e o n o , on whicf eo mistletoo htw e berrie carvede sar ,godnam e e closth th f .o e o t This feature bring o mint se threth d e berries from whic e cranth h e e -walkth n o s Risingham slab (v. p. 14 supra). The tore was a badge worn frequently by Celtic mose th f t o persistene on s deitieswa t t attributeI . Gaulise th f so h horned god, e followinTh . g horned head would see o mhavt e been directly connected with his cult.

(xix) (Corinium).1 A small male head carved from oolite, having curled haid hornsan o r T . support the equation of this Cirencester deity with Cernunnos comes a relief in local stone of a horned god with curled hair. He grasps a large serpent below the head with either hand. The serpents are ram-horned, have their tongues projecting, and raise their heads towards that of the deity.

(xx) Tanderagee, Ireland.* Although horne exteny dt figur godan no paga n o o t eo tsd n Irish monuments this is not surprising as such monuments are extremely limited in number. A strange figure, showing very early features perhapd an , s dating fro Iroe e mth n Ag consists of a large head superimposed on a small, badly-proportioned body. The mouth is wide and grotesque, the moustache boldly drawn. A line which may o indicatt e edga helme th ef o e t comes well down ove e browth r . From this helmet, or from the actual head project two horns.

(xxi) .3 A mosaic pavement is decorated with the impressive head of a bearded god, having protuberances fro s heamhi d -whic describee har Wheeley db lobstes a r r claws. This would identif figure yth e with Oceanus. They would seem however to be more in the nature of stag-antlers and here we may have yet further evidence for the cult of a horned god in Britain.

1 In the at Cirencester. 2 Henry . cit.,Armagn I op , . pi36 .h Cathedral.

Wheeler, B. B. M., Verulamium, Oxford, 1936, pi. XLI. 3 20 PROCEEDING E SOCIETYTH F SO , 19.57-58.

Gorgon Heads. Another grou headf po s includee whicb n hca thin di s sectio hornen no d heads, as most of them are horn-bearing or have traces of horns are the head sclassede b whic y t verhno ma , y satisfactoril s Gorgoya r Medusno a heads. Thi misleadina s si g term so-callee th , d Medusa head portrayes sa d by native artists in Britain having little in common with their classical counterparts questioe Th origie .th art-for d f nno an e Gorgo th f mo n head in Britain is one which merits an independent study, but one or two observation perhapy sma made sb thin ei s context. The neckless Gorgo r Meduso n a head, although havin o obvioun g s religious significance in classical times shows traces of a long ancestry amongs Indo-Europeae th t n peoples, wherl probabilital n ei t yservei a s da solar symbol, and later, under classical influence, it became crystallised int ostaringa , serpent-wreathed symbo f evil-avertino l g power t woulI . d be placed upon such objects as shields, temple porticos, eaves of buildings, etc. in much the same way as the Celtic t£te coupee was exhibited, sharing apotropain witi t i h c powers t havinbu , g lose divinth t e associatiof o n the Celtic heads. The Celtic heads were sometimes represented as having horns, and horned heads were not infrequently brought into the context of sacred serpente cul f thermath o t d an s l waters Meduse Th . a heads were likewise sometimes associated with thermal springs d invariablan , y with serpents. The assimilation of classical with native imagery is not therefore a complicated process native Th .e could clearly make concessione th o t s classical -type while retaining the symbolism and cult significance of god-in-heae th d form makind an , representatioe gth n acceptabl Romao et n and native alike. Deep-rooted native cults could thus be thinly masked under the image of Medusa, only the individuality of expression and vigour of execution betraying the non-classical origin of these heads. One or two more oth f e interesting example so-callef so d Gorgon head Britain si n wile lb described in an attempt to illustrate the above comments.

(xxii) Bath (Aquoe ).1 The Romano-British temple at Bath was connected with the cult of a native goddess Sulis, equated with Leucetiusd go witd e an , hth , equated •with . The temple, built probably during the Flavian period, was associated with deities of healing; their names implying an earlier association with solar cults. Gorgon'e Th sshiele heath Sulis-Minervf n ddo o pedimene th temple n ai th f o t s ei the finest example of the blending of native and classical imagery. The head is male, bearded and moustached, and its ancestry can be traced directly to the human heads which are so prolific on La Tene metalwork. The furrowed brow and two-dimensional features are typical of many examples of Romano-British 1 Richmond, I. A. and Toynbee, J. M. C., "The Temple of Sulis-Minerva at Bath," Journal of Roman Studies, 1955, XLV, pi. XXIX. THE HUMAN HEAD IN PAGAN CELTIC RELIGION. 21.

heads in stone, as is the expression of the face.1 The convention of the writhing serpents which here spring from the hair and are entwined in the beard and moustache is classical, but the connection of serpents with human heads is found deeply rootenative th n di e tradition. (xxiii) s.2 A head, which undoubtedly represents the fusion of the classical Gorgon head wit e nativhth e concep figurepedimens e i t th n tombstona o d f o t t Chesterea . typicas i Celtie t th I f co l heads, being bearded having moustache native-typd san e hair snakee th , s framin facee gth . Lik Bate eth h head males ii t . (xxiv) Caerleon (Isca Silurum). 3 Head carve centre relien di th larga f n i feo e stoneeyee largee Th sar e . th , nose wedge-shaped, mouth straigh foreheae severed th an t d an ,d frowninge Th . hair is ray-like and from it sprouting horns emerge. The beard is entwined with serpents. This constitutes a crude native version of the Gorgon head, and comes fro aren ma a wher antefixe eth rice nativaar n hi e head imagery. (xxv) Mary port (Alan-no).* A striking exampl t&tea f e o couple with serpents Gorgon-typ,a whict no s hi e head but which clearly illustrates this fundamental native concept of a head associated with serpents founs i ,ston a n di e from Maryport, known localle th s ya Serpent Stone. It is 51 ins. high and has a serpent on one side, 2 ins. broad and 45 ins. long. It is horn-bearing. On the other side is a neckless human head, the eyes seemingly closed, the mouth half-open, the whole aspect fierce and barbaric. Two serpent ford se fac an mreaframa th p er u rwhic fo e s moustachehi d dan probably bearded. Below the chin is a tore. Portions of a second, broader serpent were found nearby, the whole closely associated with two burials. According to local traditio e ston nphallid th eha c associations. Here the have nw ehumaa n head, constitutin ga fin e exampl e Celtith f co e tite coupee tradition, associated with serpents (one of which is horned) and fish. The horned serpent would suggest a connection both with war and with some aquatic cult, this supposition is strengthened by the presence of the fish. The local cult described in native terms on this impressive stone may not then have differed fundamentally from that practised at Aqua? Sulis where the local deities of the thermal waters, Sulis and Leucetius were equated with Mar Minervad san , bot whof ho m presided over war, wherd 'an e ' decapitated'eth serpente ' th hea d dan s were welde Celtia y db c artist into the classical image of the Gorgon. la, 4. "Phallic" Heads. Ther ampls ei e evidence, bot Celtin hi c iconograph Irise d th han n i d yan Welsh vernacular tradition veneratioe th r humae fo , th f no n head. Many powers were attributed to it. It was symbolic of deity. It may, at some 1 Attention may be drawn to a head recently placed on view in the Yorkshire Museum, , which ressembles the Bath "gorgou" head, but the stranded hair has no serpents entwined in it and it is altogether more native in concept. Bichmoud and Toynbee, op. cit., pi. XXXVI, 2.

32 Richmon Toynbeed dan cit.. op ,, pi. XXXVI. 3 , 4 Bailey, J. B., Netherhall Collection, Kendal, 1915. The male head rises to a strange point on top, which is somewhat reminiscent of the head from Killavilla, Ireland (38) and it may represent some sort of cucullus. 22 PROCEEDING E SOCIETYTH F O S , 1957-58.

period emergence prioth o t r e odistinctiva f e Celtic peoples, hav somd eha e solar symbolism, as the substitution of the human head for the solar disc on early solar barks indicates. Moreove clearls wa t yi r believe e Celtth sy db to be the bodily member in which the life force was situated, and possession of the head gave the owner access to the powers of the living man. This "decapitated" head could avert evil, and foretell events. Cond an alternative form of cend, means "head" and "wisdom." The apparent lack of phallic imagery in Celtic iconography has frequently aroused comment t doeI . s occur examplr fo , connection ei n wit Britise hth h horned god, and with a British god of the Hercules type, but such monuments e comparativelar y rare. Ther , howevereis , sufficient evidenc o suggeset t that the "decapitated" head set upon a stone pillar was, amongst other things, a phallic symbol, in which the apotropaic powers of both head and phallus became united into a potent whole. The fact that the head was conceived of as the centre of the life force, capable of continued independent existence after the death of the body, points to this conclusion. The implication is then that the usual Celtic symbolism had come into operation here, the male organ being represented symbolically rather than naturalisti- cally. Such symbolism would help to explain the comparative lack of phallic monuments, and would emphasise the fact that any attempt to appreciate or come to terms with the religious and superstitious attitudes of the Celts must take into account their preference for oblique ra.ther than direct expression. In order to give weight to this suggestion, one or two monuments are described which would seem to substantiate it.

(xxvi) Pfalzfeld, St GW.1 A stone pillar, originally over 6 ft. in height, shaped like a phallus, the glans restingrounde th n go . Originall surmountes wa t yi heaa y ddb bearing some sort of stylised horns. On each of the four sides is a typically Celtic triangular human face surmounte leaa y f db crow n which also probably double r hornsfo s . This unique monument dates from about the 5th or 4th century B.C. It shows Etruscan influenc e decorativth n i e e motif's. Phalloid tombstone e knowar s n from Etruria, without the addition of heads.

As this pape primarils ri y concerned with evidence fro Britise mth h Isles, attention "will onl furthedrawe e yb on o nt r exampl headf eo association si n with phallic symbols from Europe enigmatio Tw . c monuments come from Jutland.

(xxvii) Tommerby, North-west Jutland. 2 A earlpillan a n yi r Renaissance churc t Tommerbha shape fonth a s n ei t yha ophallusa f sidee th , s decorated wit hspiraa l patternhollowep cu e th ,d froe mth Powell, T.G.B. Celts,e Th , London, 1958 pi. 60.,

1 2 Detail f thio s s a photograppilla d an r h were kindly provide Pastoy db r 0stergaard, Glejbjerg, Denmark. E HUMATH N HEA PAGAN DI N CELTIC RELIGION3 2 .

top of the glans presumably for liquid offerings. Also from southern Jutland 1 comes a similar phallic pillar, in this case having the sides adorned with human masks from one or two of which a phallus grows from the top of the head. The datin f thesgo e object problematis i s f mano o y t tw thefro e cbu myar this area which show affinities wit Celtie hth c material.

(xxviii) Broadway, Worcestershire* Oolitic stone shaped like a phallus, on the glans of which is a crude human face carve reliefn di e glanth , s thus servin facheade e th eTh r .closelg fo y resembles some of the faces on the impressive Gloucestershire reliefs. This combination of head and phallus from Broadway would thus seem to be a stage between the naturalistic representation of the actual phallus and the symbolism of the neckless

pillaa n heao r t whicdse h woul dphallia seee b mo ct symbol. (xxix) Eype, Dorset.3 Large pebble, now 1\ ins. high in the shape of a phallus engraved with a human head e featureth , s reminiscen f certaio t n monuments from Europer fo , exampl e janiforeth m statue from Holzgerlingen, Wurttemberg e headth d s an , from Roquepertuse. Below the face are four interlaced circles. The lower part of the stone is broken.

(xxx) Corbridge (Corstopitum). V) L (P 4 A crude head carved in one -with the base, much -worn, but apparently phallus- shaped originally. The head, -which is also the glans of the phallus, has no ears, larga e mouth, emphatic moustache, narrow nose widd an , eexpression i eyes d an , n resemble monumene sth t from Tanderagee, Ireland discussesupra.9 1 . p n do

The above examples thus furnish some direct evidenc suggestioe th r efo n thae "decapitatedth t a pilla "n o hear t stonse d e constitutes a phalli- c symbol.

. Heads16 Without Attributes. The heads included in this section, while having a cult significance, are t distinguisheno sucy b d h individual attribute s hornsa s , multiplicitf yo faces, etc. They consist of simple heads which may have functioned as small shrinessimply ma r yo , have serve decorato dt e templth e f somo e e

native deity. (xxxi) Killavilla, Ireland (PI5 . V). A male head, some 16 ins. high, fashioned from a lump of limestone with an unworked bac basuncovered s kan e wa base largea th f eo t da r boulder whics hwa being moved. One of the most impressive heads described in this paper, its lean, 1 From Borkaer, South Jutland. Details and photograph from Pastor 0stergaard. 2 Ant. J., xxvin, pi. XXIV, e and /, and 166 for Professor Hawkes' interesting comments. a Ant. J., xvi, pi .323 d LIXan , . 2 , Romae th n I 4n Museu t mCorstopituma . 5 J.R.S.A.I., 1945, 265. 24 PROCEEDINGS OF THE SOCIETY, 1957-58.

hollow cheeks, lipless mouthslir fo t , staring eyebroad san d flattened nose lint ki with such heads as those on the Aylesford bucket (p. 31 infra), the Corleck tri- cephalojanifore supra),3 th 1 d . (p an ms heads from Roquepertuse likels i t I y. dato t e fro earle mth y Iron heaAgee Th d. lacks ears featura , t ini'requentleno y found with cult heads, and the abnormally pointed crown may represent some sort of cucullus. (xxxii) Camerton.1 This important hea founs trenca dwa n di h foundation e closth o et Romaa f so n hous t Camertonea , Somerset f finO e . white limestone s somi t i ins,4 e . high. It reveals all the features of Celtic cult heads having arresting eyes, no lids, deeply demarcated ears, long upper lip and a short nose. There is no neck. An unusual feature is the hair, which is represented by small squares all over the top of the head. Camerton is some 45 miles from Eype where the "phallic" head described on p. 23 was discovered. (xxxiii) .2 A sandstone head, som inse9 f ins.4 hig .d acroshan heae founs sth dwa n do flooe shrinea f th o r , having gravel claa step d y heae e sflooran th flaTh s di t a .t obviousls bottome th wa t d a back an ,d y meanan ,viewe e b o t t d fro frone mth t alone. The neck of the head is thick-set, and clearly a native deity is represented. The eyeballs, which are prominent, are in relief and surrounded by an oval depression, the nose is very straight, the mouth small, ears deeply indented, almost slot-like. There is no hair. It is possible that the slots for ears were intended

insertioe foth r cervinf no equinr eo e ears. (xxxiv) Caerhun (Kanovium) .3 A head crudely fashioned fro mluma f sandstonepo , foun graven e o dth n i l courtyard of the Roman fort. The features are nearly flat, being indicated by a series of incisions. A deep line on the brow defines the hairline, and the hair is show seriea y parallef nb s o l incisions whic fron hru m just abov foreheae eth o dt just abov nos necke e nape th lones Th eth narrowi d f e.o g an eyee th , s widd ean moute th h straigh firmd an t , givin expression ga severitf n o facee th .o yt (xxxv) Castlesteads.* A human head, roughly fashioned from a block of red sandstone 8-Jr ins. long and 5 ins. wide was excavated at Castlesteads Woods. The features are crude, and no attempt at a naturalistic outline of the face was attempted. The features are indicate straighy db t lines moute th , hlargee fashioneth slitso d , tw an , y db square eyes hav pupile eth s demarcate straigh e slitsy db Th . t hairline acrose sth brow, side-whiskers and beard are indicated by pick marks. (xxxvi) Carlisle.^ n impressivA e head worked from local sandston s recoverewa e d froe mth Roman road near Appleby. It is not easy to date this head but it would seem to belone e perioRomath th f o o t gdn occupation d Century, 3r perhap e th . o t s 1 Ant. J., xvi, 205 if. and 206 for illustration. 2 Archfeologia, 58, I, 150, fig. 7. 3 Baillie Reynolds, Kanovium, 1938, 214. 4 Haverfleld, Excavations on the Roman Wall, Kendal, 1903, 348. 5 In Tullie House Museum, Carlisle. THE HUMAN HEAD IN PAGAN CELTIC RELIGION. 25

A later date would seem more probable tha earln na y date mann I . y wayt i s resembles the magnificent horned head from Netherby (p. 17 supra), and likewise has features in common with the steelyard weight from Carlisle (p. 33 infra), and the pottery head from the same area (p. 17 supra). The head is unusually large, the skull powerful, the eyes and the mouth rendered in a manner similar to the Netherby head e eyeth , s narrowed e cheekth , s flattened somewhae th , t military appearanc e smalth f leo bear moustached dan s toning dow e barbarinth c aspect of the features.

(xxxvii) Corbridge (Corstopitum).1 A male head carved from a sandstone block, measuring 7 ins. from crown to chin, and approximately 5f ins. from ear to ear. There is a deep cup in the top heae olibationsr th f dfo haie rTh .come poina o st t abov broe nosee eth Th w. is shallow and prominent, the eyes large and the pupils indicated by cuts. There pupillefo e th ttw eyen nose e si ar lon s Th enarrowi nostril.d e gth an d e sar an , demarcated. The mouth is small and slightly drooping, and the deep furrows whic fron hru m nos mouto t e emphasid had tho t s e austerit e featuresth f yo . The cheeks are framed by side-whiskers, and there is a small pointed beard. The fac s triangulai e shapen i r shared an , s several features wit e heahth d froa mBo Island, Irelan infra).6 1 . d(p This headt actuallno f i , y dating fro mperioa d prior to the occupation of the Romans, can hardly have been fashioned at a time when Roman influence was strong in this area, since it owes little to classical art con- ventions, but springs directly from La Tene traditions. The left eye is considerably larger than the right eye; this feature has already been commented on in connection with the Corbridge janiform head discussed supra p. 16.

(xxxviii) Corbridge.2 Another head from Corstopitum resembles head No. xxxvii in dignity of expressio austeritd nan f executionyo ins 7 s heighn i .i t againd I .an t s comi , - pletely native in treatment. The features are in fairly low relief, the eyes large, nose lonnarrod gan w with little attemp defino t t nostrilse th emoute Th .s h i straigh severed an t severite th , expressiof yo n emphasise liney db s unde eyese rth .

Corstopitum has thus yielded up five examples of "decapitated" heads, all native in workmanship and cult significance, each reflecting an individual aspect of the symbolism of the human head. One head is janiform, one is horned, another is fashioned on a phallic-shaped stone; of the two "simple" libationsr heae fo culth e p thus df i to o tholloTh a se .s headsth ha w n i e on , well-documente thin di s are athroughous i nea t Wale i s th r (a l generae th t l region)l thesal d e headan , havy sma e been associated wit flourishine hth g cul Maponusf o t . These head evey sma n ante-dat Romae eth n occupation northe oth f , their similarit Irise th ho y t head s makin earln ga yo datn y eb means improbable d thian s, would provide some evidenc a nativ r fo e e tradition of stone carving, although strictly limited, before the influence of Roman iconography. 1 Richmond cit.,. op , pi. Ill. ,A 2 Romae Inth n Museum, Corstopitum. 26 PROCEEDING E SOCIETYTH F O S , 1957-58.

(xxxix) Lustymore, Ireland.

A large triangular head1 from Fermanagh resta smal n o sl square bodye th , arm whicf so weae attenuatedhd ar kan squattine ,th whiln i leg e e eth sg ar position characteristi e Gaulisth f co h Cernunnos beare Th d. is pointed e eyeth , s large, the nose long and narrow. The head seems to be without ears. The top of the head is damaged, thus it is not possible to discover whether the monument originally functione shrinea s da . This impressive head, whic similas hi somn ri e . xxxviwayNo o it s datee supra,b n d ca wit hreasonabla e degre f confidenceo e somewhero t e abou Christiae ture th th f t no n era.

(xl) Chesters (Cilurnum).2 Head, much worn, but of similar type to No. xxxviii supra. Carved in fairly low relief from a block of stone, it measures some 9 ins. in height. No hair is indicated, and the face appears to be clean-shaven. The eyes are large, the mouth thin and straight, the nose flat and rather wedge-shaped. No ears are depicted.

(xli) Dumfries.3 This head, fashioned of local pink sandstone is slightly less than life-size. It is female mosd an , t probably represent wore a Celtigoddessa s th f ko s i c t I . t cleano rartists i whethe t I . t formei r d parbusfull-siza a f r o o tt e statuer o , whether it originated as an independent monument. The thickness of the neck may indicate that it had become detached from a base or pediment. The head- dres remarkables si , resemblin gclose-fittina g bathin hairnetr o eyee p gTh ca s . are large eyeballe th , s flattened moute th , h lon slightld gan y drooping nose Th e . and chin are damaged, and above each nostril there is a line curving inwards to form a spiral. The face is serious and dignified.

(xlii) ().* Male head fashioned from local chal d showinan k g trace f havino s g been coloure ins8 d s i red. hight I . , arid date centurd s2n fro a e yms th i A.D t I . remarkable piece of native workmanship, made outstanding by the prominent eyes and utter sombreness of the expression of the incised features. The abundant hair is combed down onto the brow in points. The ears are very elaborately portrayed. The head was discovered during the excavation of the site of an extension to the Bon Marche buildings in 1934. A fine carved tablet of Mercury,

his Celtic consort Rosmert anothed aan r goddes founs swa d nearby. (xliii) Towcester, Northants.5 A female head, showing traces of classical art influence, but nevertheless completely native in concept and power of execution, dates from the 3rd century and was originally employed as an antefix. It undoubtedly represents a native goddess, having both religious and apotropaic importance. It is carved in one nece wit paro th modelledn bases s kf hi it to d stylan ,e hairdressinf Th eo . g shows classical influence, but the Celtic love of pattern has caused the artist to fashion a 1 Henry cit.,. op , pi. 4a. 2 In the Roman Museum at Chesters. 3 Toynbee, J. M. C., "A Roman (?)Head at Dumfries," Journal of Roman Studies, 1952, XLII, 63 if. Transactions of the Bristol and Gloucestershire Archaeological Society, 1934, 56, pi. VIII. 4 5 Kendrick . D.T , , Anglo-Saxon Art, 1938,. piIX . E HUMATH N HEA PAGAN DI N CELTIC RELIGION7 2 .

single stran haif do r int oserpent-lika e spira eithen lo r cheek, emerging from behind well-definee th d eye e enormouearse sTh ar . compellingd san pupile th , s indicated meany b deef so p holesbroe Th wfurroweds i . nose th , e lon narrowd gan , nostrils indicated e moutth , h drooping e chith , n broad entire Th e.e hea basth s df i eo broad, and the artist has ceased to model it below the chin, which forms part of the base.

(xliv) York (Eburacum).

This unusually larg eheightn 1 d i head . bottoe ol founs ft th n wa 2 a ,t , da f m o wall in the Castlegate and purchased by the Museum in 1789. It is carved in one with the base, which is triangular in shape, the back of the head is unfashioned, and it seems that it either functioned as an antefix or as a head to be seen from frone th t only somn i , e religious shrinemales i longs t I ha ,. , copious haire th , front hair rising inthorn-liko otw e peaks remaindee th , haie th rf formino r g thick strands below the ears, or rather to •where the ears should be were they represented. certain I n respect t resemblesi sizstylxlin . d I s si e .an headt No ei xlii . d sian No remarkably impressive, being e chibroanarrowind th nan t da towardp gu e th s brow, the cheeks sunken, all the features combining to impart intensity to the large, fairly close-set eyes. The mouth is firm, rather long and slightly drooping, somewhat reminiscen e Dumfrieth f o t s hea thin di s e noserespectw th s no ,i s a , worn at the tip, the nostrils indicated by an upward-sweeping curve. The brow is slightly puckered mosa , t regular feature with thes whice headsone h addand , s intensit expressioe th o yt n whic invariabls hi y sever penetratingd t ean no s i t I . easy to attach a date to this work, but it is possible that it is fairly late, possibly centurd 3r r eve yo centurh n4t y A.D.

Apart from constituting complete monuments, heads also figure in other cult contexts, for example, on reliefs, as anteflxa and on pottery objects having ritual importance.

2a. Heads in Relief.

(xlv) Lanchester.2 Relief in yellow sandstone, burnt red at the top left-hand corner, on which is a male head, having a neck, but with no trace of a body. The stone is 9 ins. by 8J face ins5 s elongatei Th ey . b d sombrdan expressionn ei , having wide eyes, usual puckered brow, small straight mouth, lackin earn hairgd i san .

(xlvi) Wallsend ().3 Fragment of a stone shield on -which is a circular face in relief. The nose is wedge-shaped, the eyes compelling, and the cutting of the nose and brow in one gives an over-hanging effect to the forehead and imparts a scowling intensity to the gaze. The mouth is small, straight and severe. The hair is close-curled and extends down either side of the face to give the effect of side-whiskers. The Yorkshire th n 1I e Museum, York. 2 Lapidarium Septentrionale, 714

treatment of this head resembles the circular masks on the diadem from Wilton- by-Hockwold, No. Ixxii infra. Here again it seems that we are witnessing the amalgamation of native and classical concepts, namely the fusion of the tete coupee and the Gorgon head in a context where both would be equally applicable. The fragment, -which is fashioned from brown sandstone, measure f ins s15 12y .b ^ ins. by 3| ins.

(xlvii) Charterhouse-on-Mendip, Somerset.1 A rough ston whicn eo hhumaa n hea bees dha n worke relieffacn e di s ei Th . almost circular eyee th , s huge earo n hai ,r shownse o rar face th clean-shaven,s e i , the mouth small, the expression fixed and penetrating. It may be rioted that No. Ixvii also comes fro same mth e locality.

(xlviii) Birrens, Dumfriesshire.2 A relie whicn o fsmall a s hi , horned head having pronounced, austere features. The letters —IADI visibllowee e Iar th possiblstonen s e ri e o hal th t I f o f. e that this head was connected with a cult similar to that testified to by the presence of several relief armedf so , horned gods fro generae mth l Carlisl ehornee areath y b ,d bust from Moresby, and by the horned head from Netherby, No. xii supra.

(xlix) Edinburgh.3 Nationae th A n torsoi w l Museumno , believed an , havo dt e formede parth f to Camelon Collection t publishedno t smala bu ,s ha ,l human type heath ef d o unde r discussion draw e stomach relien th headless ni w figure n ha o no f Th d s i e. an s lost both arme head legsTh an sd . e stomacdrawth y perhapn o n ma h e b s parallele sucy db h monument s thaa s t from Entremont, wher malea e torss oi decorated by a '' decapitated'' head from below which two serpentif'orm arms emerge.4

26. Antefixa. Ther severaeare l antefixa which testify furthe nativto r e predilectionfor the symbolism of the human head in matters of superstition and magical e contexth n practicei f antefixa o ts i t I ., pottery objects generalld yan metalwork that British religious traditions seem to have lurked longest, undisturbed and perhaps less liable to classical influence and the confusions brought about by interpretatio romana. As these tiles are native in imagery thes a yd weran e attache eavee th buildingf o sdo t amuletss sa ,f o the e yar particular interest here. Only thos whicn eo hhumaa n hea figures di d will be discussed here, which means that some of the most striking tiles from the point of view of native religion must be ignored in this context.

1 , British Museum Catalogue, London, 1951 , fig. 53 ,.25 2 National Museum of Antiquities, Edinburgh. C.I.L. VH. 1079. 3 National Museu Antiquitiesf mo , Edinburgh. 4 Lambrechts . cit.,op , pi, figII .. .7 E HUMATH N HEA PAGAN DI N CELTIC RELIGION9 2 .

(1) Caerleon (Isea Silurum).1 Triangular-shaped antefix, the decoration in high relief. A central rayed male face is surmounted by cat's ears. A six-spoked wheel is figured above the head, four-pointeo tw d an d star drawe ar s n below. This tile common i , n with some seven others l veral , y simila eaco rt h other, is comparatively earl daten yi , having been fashioned somewhere between the years 70-110 A.D. Several other tiles of similar type have neckless heads draw then no mhign i h relief, frequently radiate, some having cat's ears. These head e associatear s d with pine trees, wheels,

1 Caerleon Report on Excavations, 1926-29, 20 ff. The cat-ear thesn so e interestheadf o e sar . Cats sometimes figur Gallo-Roman eo n monumenta n si funerary context, where they symboliseee b mo t somf co e chthonic power. (v. Le Chat dans Statuairea l Qallo-Romaine, Revun i e archselogiqu Tese e d t ch e t i centre-est, Janvier-Juin, 1958 ff.8 ) ,12 They also e Gundestruplat e occuo th t on f ey o n o r ma p cauldron (Muller, Sophus grande L , vase Gundestrupe d n e Jutland, Nordiske Fortidsminder, Copenhagen 1890-1903, 62 if., pi. VII). There is, moreover, a mytho- logical Irish character know s Cairbrna e Cinn-cait, Cairbr cat-heade th mentiones i f eo e H e .th n di Coir Anman, a composition dating in its present form from the 12th century, with archaisms that indicate earlier compositions, which discusses the epithets attached to heroes and mythological beings, and their appropriateness. Material of considerable mythological interest is embedded in this work. The relevant entry, in translation, is as follows: "Cairbre Cinn Cait cat'e 'oth f s head', cat'a "s sincs wa head t ei ,fore tha th shapr m o s i t f eo gods therr hi O .n e o t thawer s cat'o ca twa etw a s ears upon him sai s poet,e a dth , namely Bochaid ua Ploinn: Thu Cairbrs swa cruee eth l who seized Ireland sout nortd han h two cat's ears on his fair head cat'a througr sfu earss hhi . (Coir Anman, Stokes Irische Text, Leipzig and Windisch, 1897, 384, para. 241). Again, in the story of Morann Mac Main's Collar, Cairbre is referred to as the cat-headed king of Ireland. ' "Now Cairbre, Morann's fathecat'a d s ha rsnout useeverd m thaboro s n dhi an t , yso o wa nt t have blemisa theo hkilles e nh d them. ..." (The Irish Ordeals, Irische Texte, 206, para. 12). O'Rahilly (Feil-sgribhlnn BoinMhic Neill, Dublin, 1940, 103) disagrees with the acceptance of Cairbre Cinn-cait as meaning cat-head. He holds that the form is cattchenn, standing for an earlier cotutchenn, meaning hard-head, this having something to do with his reputation for tyranny. He goes on to say, however, thahardls i t i t y necessar streso yt face sth t ("which wil sufficientle b l y obviou thoso st o ewh have some knowledge of ") that Cairbre Cattchenn so far from being an historical character ultimatels i , ydivina e personage, ancestor-deit e Erainth f yo n (102). Thus while allowing that Cairbr supernaturas ei l in origin, O'Rahilly regard fantastis sa explanatioe cth n given Coire inth Anman epithes hi r fo t Cat-head. Whil l vernaculaessentias i al t e i us o lt r reference pagao st n cultd san divine personages wit utmose hth t discretion traditioe ,th sucf no beinha g having animal morearo n s sei fantastic than the belief that the supernatural being should be capable of taking on animal form. When one bear minn si e numeroudth s representation divinf so e beings from Gau Britaid an l n having such animal part hornss sa , antlers, hoove animad san l ears sucd an ,h figures fro vernaculae mth r tradition Labhraids a h Loingseach wit horss hhi e earssurprisins i t i , g that O'Rahilly should take exceptioa o nt native explanation for a name, an explanation which is better attested in the superstitious tradition of the people than many of O'Rahilly's own religious hypotheses. Again, in Sanas Cormaic, a glossary dating from the 9th century, a method of divination is described in which the flesh of a red pig, a dog or a cat is chewed, then an incantation is sung over it, and it is offered to "idol gods." (v. Himbas Forosnai, Sanas Cormaic, ed. Stokes, W., 1862). The pig and knoe w havewo g t do e playeth importann da t rol Celtin ei c cult practices face tTh tha. flese f th t ho the cat is also included here imparts to it some status as a beast of divination. These examples from the Irish vernacular tradition, in which much of genuine mythological significance is to tie found, depict a mythological being, referred to as a god, having cat's ears. It is not of course suggested that there is an actual connection between the Irish character Cairbre Cat-head and the enigmatic being wit cat'e hth Caerleose earth n so n antefixa t onlbu ,y that suc htypa supernaturaf eo l being would not be unparalleled in Celtic mythology. 30 PROCEEDING E SOCIETYTH F O S , 1957-58.

pellet-like objects and crosses. The cross in association with a cult head has already been describe Scottis e notee th th n n so di h tricephalos. Their similarity to each other makes it more convenient to group them all together under this single number, than to attach a separate number to each example.

(li) Caerleon.1 Antefix having a mask-like human face drawn on it in high relief. The chin is pointed, eyes enormous, as are the ears, and the hair is indicated by a straight line drawn acros e browth s , formin a cap-likg e coiffeur, comparable with head No. xxxiv, p. 24 supra, from Caerhun. Pine trees in groups of three surround the head. This tile dates from somewhere betwee yeare nth s 105—200 A.D.

(lii) Silchester (Calleva).* An antefix, made fro mlocaa l moul baken di d clay, dates fro earle md th y2n century A.D. On it a male head is drawn in high relief, having a complex set of symbols. Ther thica s ei k fringe-like arrangemen browe th haie n th ,o r f beino t g depicte firy db m straight strokes from belo e crowe foreheadwth th o nt , above which two strong bull-horns come from the centre of the head and lie horizontally above the fringed hair (cf. the atrophied horns on head No. xviii from Benwell). There are human ears in the normal position, while on top of the head are ears which unusuan appea a felinee b d o t an r,l crest-like decoration between these secondary ears. The lips are thick, and below the chin are traces of what may have been a beard, or writhing serpents, in which case the blending of native and classical concepts would once agai manifeste nb e antefiTh .s triangula i x n i r shape. The eyes are, as usual, enormous.

(liii) Dorchester.3 A tile from Dorset has a similar male head in high relief on it, with large eyes, horne r cat-earedo d d wreathedan , . Several other fragmentary examples from Dorcheste neighbourhood ran locaf o frol moulde e lal mar manufacturee on d ar ,an . All these antefixa are fairly early in date.

2c. Miscellaneous Pottery Objects. A further sourc evidencf e o huma e symbolie th th f r o enfo e heacy us dma be found in the so-called "face-urns." These urns are too numerous and the study of them too highly specialised to allow of more than a brief reference here. Two examples, which may be taken as typical are briefly described.

(liv) Oxford* A crude face-urn with three vertical handle humasa n nose correspondino gt fourte th h handle, this being stampeflankeo tw y db d circle same eyesr sfo Th e. toobees ha ln use indicato dt ebearda . 1 Caerleon Report on Excavations, 1926-29, 20 ff. 2 In the Beading Museum. 3 Journal of Roman Studies, 1956, XLVI, 142, pi. XVI. 2 , 4 In the Ashmolean Museum, Oxford. THE HUMAN HEAD IN PAGAN CELTIC RELIGION. 31

(lv) .1 A slightly different representation, more than a mere approximation to a human face being aime t herea d frog , ju moccur a Alchestern o s , whero tw e well-defined menacind an , g male face depictede ar s .

Two further fragment f pottero s y f valuwhico e thin ei har s stude yar described below.

(Ivi) Carlisle () 2 (PI. VI). Fragmen larga f o te pitche r vaso r e foun n Castli d e Street, Carlisles i t I . decorated by a rudely modelled male head, the thin, drooping mouth giving an expression to the face -which is reminiscent of the expressions of heads No. xii . xxxvNo id froan m this fragmene areaTh . fashiones i t lighf do clayd d re t an , therinterestinn a s ei g corone disd an tc ornamen heade th n .o t Ball claf so y with incised circle and dot for eyes and iris have been used, and there are raised bands with notched chevron r side-whiskersfo s , bear d moustachedan . This seemo t s represen locaa t l deity.

(Ivii) Richborough.3 A fragment of pottery is ornamented with the head of a horned female in low relief probabld an , y represents some local goddess featuree Th . s a,nd expression comparee b y ma d wit . IxihiNo fro same mth e site.

. Metalwork.3 Bronze miscellaneoud san s metal objects constitut importann ea t source for Celtic religious symbolism. Although smal n sizei l , these objecte ar s frequently early in date, based on superb traditions of native metallurgy, and some are of major importance here.

(Iviii) Aylesford, Kent.* It was customary to bury buckets and cauldrons, often elaborately decorated, gravee th n Celtf i so noblf so e birth, thus implying that these objects themselves had a ritual significance. The most dramatic bucket escutcheons in the shape of human heads come from Aylesford, and date from the 1st century B.C. They are probably Gaulish imports and furnish a fine example of the Celtic tete coupee concept. The face is withdrawn and self-contained in expression, finely modelled and dignified, this air of dignity accentuated by the impressive helmet with its high crest, and by the huge, closed eyes, the long geometric nose, sunken cheeks and straight, firm mouth.

1 In the Ashmolean Museum, Oxford. I am indebted to Mr Robert Hogg, Keeper of Tullie House Museum for this photograph and for kind2 permissio publiso nt . hit 3 Richborough Report, iv, pi. XLII, No. 347. ' Leeds, CelticT. . E , Ornament, Oxford, 1933 fig, . ,39 .16 32 PROCEEDING E SOCIETYTH F O S , 1957-58.

(lix) Brough-on-Humber ().1 A bucket, recovered from the grave of a British noble, hooped in iron, with bronze escutcheons modelled as human heads. The heads are male, necks short firmd heade an th ,broad e sar , jaws prominent, lips pursed, noses broa straighd dan t browd an s jutting haie thics r i Th combe d . kan d straight dow nati.vn i e fashion. The chin emphasisee sar smala y db l beard. (Ix) Welwyn (Herts).2 Three heads, which functione s mounta d a meta n o s l bowl, fro mburiala ; La Tene in style, resembling the "decapitated" heads, with hair combed down on to the brow, eyes seemingly closed, and having well-defined moustaches and beards. (Ixi) Aldborough ().3 Bronze terre forn mala ti f mo e bust; fro heae sidee th m th df s o sprin g great horns, whil projectioea centr e heae represeny th th dn f nma i e o thira t d hornr o , may be associated with the leaf-crown, now almost destroyed, which originally emerged from the back of the head. The face is moustached, this giving added severity to the drooping mouth; the eyes are enormous with large pupils which must have been enamelle timee ferocite on gaze th ,t da th ef yo emphasise e th y db furrowed brow. A date shortly before 71 A.D. has been proposed for this work. (Ixii) Richborough, Kent.* Fittin bronza f go e attachmen basia r tnfo handle decorated wit hhumaa n face, surmounte bull-horny db s which terminat knobsn ei . Although knobbed horns are found especially in early Iron Age contexts, this is a feature which survived the conquest, and this fitment may date from the period immediately after the Roman occupation basie Th whico nt . handle h th attache s ewa likels di havo yt e had ritual significance. (Ixiii) Richborough.5 Bronze fitment consistin pieca f gtubinf o eo g covered with palmette patterns. One end is closed by a terminal composed of the masks of two women set back to back, Janus-fashion. A necklace divides tube and neck. The faces are identical e onlth , y difference being tha f positiono t face on e, looking upwards. The modelling is fairly crude. The women wear fillets in their hair. Artistically these probablfemalese ar o wh ,y goddesses, resembl female eth e bace bustth kn so three-hornee oth f d bull from Maiden Castle, Dorset.6 (Ixiv) Marlborough (Wilts.).

Vat, 211 ins. high7 , diameten nearli . ft y2 r which must have contained some sort of alcoholic liquid for feasts of a ritual nature. The sides are decorated with bronze heads, reminiscen e Gundestrue headth th n f o so t p cauldro sacred nan d beasts believes i t I . havo dt e been imported round abou yeae e B.Cth t0 Th r5 . "gode eyeth f so " head must have bee t witnse h cora enamelr o l . 1 Ant. J., xv, 459. r CyriSi 2 l Fox, Pattern Purposed an Celticn i Art, Cardiff, 1958, . pi33 . 3 Richmond, I. A., "Queen ," Journal of Roman Studies, XLIV. pi. n. 2. 4 Richboroug. h Report17 . No , pi, x , X . 5 Richborough Report, iv, pi. XLV, No. 172. • Wheeler, R. B. M., Maiden Castle, Dorset, Oxford, 1943, pi. XXXI. 7 Fox cit.,. ,op . pi34 . E HUMATH N HEA PAGAN DI N CELTIC RELIGION3 3 .

(Ixv) Old Carlisle * (PI. VI). Bronze steelyard weight in the shape of a male head. Two ins. high, it is filled with, leaweighd face dan almoss ei Th s . somoz t square5 shapen ei e th , broad chin heavily bearded, the thick hair extending down the cheeks to blend wit beard e moustache hth Th . thics ei drooping d kan nose th , e wedged flaan t - shaped. The eyes are heavy and narrowed (a feature found in several examples of heads from Cumberland), the eyebrows pronounced. The lips are thick, and giv impressioe eth beinf no g slightl heae yth pouteddf o hai e p thics rTh i to .n ko and divided into strands. Two of these may represent some sort of atrophied horns t unambiguous t thino bu s ,si .

(Ixvi) Furness.* A bronze mask, having typical Celtic features, and turned into a weight by Vikings who tore it from its base (British Museum).

(Ixvii) Charterhouse-on-Mendip, Somerset.3 Bronze mask, having extremely large eye-sockets a droopin, g mouthd an , curiously arranged hair, which is treated in a manner which is reminiscent of head No. Ivi.; that is in a chevron pattern.

(Ixviii) Stanwick, Yorkshire.* Bronze bowl with mas highly-stylisef ko d huma n, reminiscen it fac n eo e th f o t treatment of some of the faces on the Gundestrup cauldron.5 Dating from about t centur1s e th y A.D.

(Ixix) Thealby, Lines.

Bronze handle mouncentre 6 th whicn f i teo humaa s hi n face. Eyes narrowed or closed, having side-whisker d shorsan t beard, mouth smal d straightan l , hair arranged in native fashion, two atrophied bull-horns above the brow. From the same trench came two bronze escutcheons in the form of bulls' heads from which an eagle emerges. Probably 1st century A.D.

(Ixx) ?Ribchester (Ribble), Yorkshire.'1 Triple-headed bronze mount, consisting of three flat-backed heads. A hawk emerges from a bull's head, the horns of which are knobbed. On the back is a human head, wit htypicala , rather sombre expression stylised an , d treatmenf o t e e bird-heahairth Th . s dominanti d d thian s, comple s cleaha x r symbolic significance. This mount may be compared with that from Thealby (No. Ixix supra).

Tullin I 1 e House Museum, Carlisle. Journal of Roman Studies, 1957. , piXV . 2 Henry, op. cit., pi. 49e. Leeds . cit.,op , fig. 296. British Museum Guide Iron-Ageo t Antiquities, 141.

54 Op. cit., note 62. 6 Ant. J., xv, pi. LXXI. 7 Ant. J., 1954, pi. XXI 225d Van . 3 VOL. XCI.

34 PROCEEDINGS OF THE SOCIETY, 1957-58. (Ixxi) Rotherley, Dorset.1 An object of bronze in the form of a goose with a human head on its back. The hair runs in parallel strokes from brow to crown and the face appears to be bearded and to have side-whiskers. The eyes of the bird and of the man are enormous ford an m, deep cavities which were originall t wityse h some substance suc s glasse functioha Th . f thio n s objec s unknowi te symbolisth t bu n s mi not unique. One of the most important pagan sanctuaries in Gaul was at Roquepertuse portice Th s decorate. owa d with decapitated human headsl al , male, and none seemingly over 40 years of age. Horses are figured, and the whole portico was surmounted by a large goose in the round. The dominant position occupied by this bird is indicative of its importance in the Roquepertuse cult, and it is associated with human heads, as is the little bronze goose from Rotherley. Again, the janiform head from Roquepertuse has a great beak surmounting it and forming a division between the two heads. Human masks are frequently used with stylized duck-heads on La Tene metalwork. Birds and heads are associated again in the traditions of the Irish war/raven goddesses followine Th . g entry occur Cormac'n si s Glossary: 2 three "Machath f eo Morrignase on , s tha i crow a t i s ti r .o ; Mesrad Machae, Macha's mast afte n ,heade tharth me f s theiti so r slaughter." Thu deae headse th sth df o woul d see havmo t e been this battle goddesse's due. Archaeological evidence has revealed that in Gaul at least one altar to a native war deity under the guise of Mars had decapitated heads placed under it indicating that thes beed eha n deitoffering.renderee n a th s yo a t p d3u Again legene th , f do the head of Bran the "Blessed Raven" is one of the most convincing pieces of vernacular evidenc veneratioe th talisman a r s ehumae fo a e th f nus o ns . it hea d dan The wonderful head of the blessed raven would fit well into this argument, but althoug e namhth e Bran incontestably means "raven" convincino thern s ei g evidence from textual sources to show that he was a raven god in origin, although attempt como st thio et s conclusion have been made.4 thun showe ca b s t I n that althoug e littlhth e Rotherley bronz a mos s i e t unusual object, the association of a decapitated head with a goose or raven, both

birds of war, is not unknown in related Celtic tradition. (Ixxii) Wilton-by-Hockwold, .5 A bronze crown for priest or cult figure was ploughed up together with five sheet-bronze diadems in Norfolk, a little to the north of the Ouse, in 1956. Both diadem crowd sparticularlan e nar y interestin than gi t they cast somee lighth n o t accessories of cult. It is significant that this ritual crown, which was placed on the head of the deity or his priest, is decorated by medallions on which male "decapitated" heads are figured in relief. These heads are very impressive, one resembling the Gundestrup heads, and others having the eyes closed in typical

1 Pitt-Eivers, Excavations Rotherley,t a pi. CIII. 2 Sanas Cormaic, . Stokesed , xxxv. 3 An altar from Apt is inscribed Marti Vectirix Reppavi f.v.s.l.m. The names of the dedicants are Celtic. Unde altae th r r eigh ninr o t e male skulls were placed showin gdedicatioa humaf no n heado st the war-god in Roman times. (C.I.L. xn, 1077). 4 y. Krappe, A. H., "Les Dieux au Oorbeau chez les Celtes," Revue de I'Histoire des Religions, 114, 236 if. 5 Journal of Roman Studies, 1957, 211. E HUMATH N HEA PAGAN DI N CELTIC RELIGION5 3 .

tete couple fashion n eacI .h cas haie eth closels i r y curled e curlth , s continuing dow cheeke nth foro st bearde mth . Here the have nw twofolea d stressine th f go importanc heade th f e,o heads decoratin grituaa l objecplaceheade e b th o n t td. o

(Ixxiii) Farley Heath, Surrey.* A priest's sceptre found in the remains of a Romano-Celtic temple, consisting of a narrow strip of bronze attached to an iron head. The decoration consists of animals, human figure symbold entire an th s d e san desig clearls ni y Celtice Th . deity is depicted in full, naked and phallic. A neckless human head, with generous moustaches, wear helmea s t with cheek-pieces belod an , w this heawheela s di . associatioe Th humaa f no n head wit hwheea alss li o foun Caerleoe somn do th f eo n supra.9 antefixa2 . p e se ,

The first part of this paper has been concerned with a description of the various ways in which the representation of the human head is used in Celtic and Romano-Celtic contexts, as an independent feature, or to decorate objects having a cult or magical significance. Using the Scottish tricephalos as a starting point, we have attempted to show how the cult of the head was deep-rooteda , ancien endurind an t g featur Celtif eo c superstitious practice, howd collaps e time th an th f , eo righo t Romaf eo p u t n political influencn ei Britain, this feature persist d remainan s s firmly rootede n spitth i , f o e impact of Roman native cults. We can trace the history of this superstitious veneratio heae th df n o fro magnificene mth Tena L t e metalwork, where eth human head occurs with impressive frequency, through the early examples of stone carvin nativy gb e artists finae th lo t ,triumphan t blen Romaf do n technical skill with native imaginatio splendie th n i d antefix from Towcester, where the two traditions have united to make an impressive work of art. It is clear then that in Celtic and Romano-Celtic contexts, the human heas deeplwa d y symbolic n somI . e instances e heath , d undoubtedly represented some specific deity othen I . r instance havy ma e t stoosi a r dfo more abstract concept of divinity and supernatural powers. It may be noted that monuments in the form of heads are not accompanied by the specifinamy an f eo c labellin e godTh . religiouf go s monument classicaa s i l feature, and the Celts, like other primitive peoples, were superstitious about the use and abuse of personal names. A recurrent formula in the early Irish vernacular traditio "tongs ntongai a di thuath o o usd m " "I sweay b r whoy b peoply d mm go e th swears. individuaname o n Th f "s eo i d go l used in these oaths. It is, therefore, no cause for surprise that even during the Roman occupation in Britain, when names were being attached to native monuments and dedications were being made to native gods, this universal Celtic symbo f divinito l d apotropaiyan c power should remain nameless. e seconTh d par f thio t s paper will consis n attempa f o t o discovet t r

1 Ant. , XVIH J. ff1 . 39 , 36 PROCEEDING E SOCIETYTH F O S , 1957-58.

whether this well-documented iconographic featur tracey lefts a an eha n si vernacular tradition which, although compiled some centuries aftee th r advent of Christianity to Ireland, does contain fragments of earlier, pre- Christian superstitious belief Welse ritesd an Th s .h traditio laters ni d an , has been under stricter ecclesiastic censorshi e casepth tha;s witnwa e hth Irish material, but even so, it too does occasionally bear convincing testimony earlieo t r custom beliefsd san .

PAB. TII

The Human Head in Irish and Welsh Vernacular Tradition. Little suggestiv f actuao e l worshi s evepwa r allowe penetrato dt e eth early literary tales, and the most we can hope for is an occasional oblique reference to some belief or ritual practice. When the occasional fragment of evidence does get through to us, the material remains are invaluable when they can augment and substantiate it. No statement is anywhere made tha Irise tth timWelsr y ho an e t hbelievea d their god manifeso st t themselves e fora heada multipl th f mr o o n i , e head. Ther e howevear e r several reference charactero st s having multiple head marvellour so s heads. Occur- ring in literary contexts, the terms in which they are described strongly suggest traditions at one time current of a kind similar to those documented e iconographyth y b . Sinc e Scottiseth h tricephalo e focath s li s poinr fo t this enquiry, traditions about multiple-headed beings will be examined first.

. Multiple1 Heads.

(a) Ellen Trichend.1 The Cave of Cruachan, believed in Christian times to be a gateway into Hell, was traditionally associated wit otherworle hth pagaf do n tradition serieA . f so malevolent beings used to emerge from it periodically, and one of these was known as in tEllen treohend, three-headed Ellen. This creature used to devastate the whol f Irelaneo d until finally slai singln ni e comba Amairgeiny b t . O'Rahilly 2 identifies this Ellen with Aill^n mac Midgna of the sid of Pinnachad, who used to burn Tara every samain until destroyed by Finn. The name he takes to be a derivative of Ailill, sometimes spelt Elill, standing for Aillill, meaning a spirit or an elf. This identification is open to question, but in this otherworld three-headed being, possessing great powers of destruction, we do seem to have an echo of some tradition similar to that which caused the Celts at an earlier period to represent their gods in tricephalic form. 1 Revue Celtique, xin, 448, 15. O'Rahilly . F.,T , Early Irish History Mythology,d an Dublin, 1946, 300. 2 E HUMATH N HEA PAGAN DI N CELTIC RELIGION7 3 .

(b) Garb of Glen Rigel A two-headed being, the heads being set on a single neck, figures in the story entitled "Aided GuilCarbadc ma l a ocus Aided Gairb Glinne Bige." Cuchulainn meet e two-headeth s d Gar n singli b e combat e figh s fiercTh i td bot. an eh contestants are wounded. Cuchulainn finally cuts Garb's double head from off his nec impale d sharpenea kan n o t si d stake.

(c) Cimbe Cetharchend.2 Several slight references to a four-headed mythical figure occur in the Irish tradition, but no explanations for his name are offered. Loch Cimme was tradi- tionally named after him, pointing perhap origis locaa hi s o nst a l deit somf yo e sort. He is referred to in the Book of Invasions in the following terms:

"This is the advice which he adopted: against Ross he set Rind; against Conall Oernach of many pledges he set Cimme four-heads."

It may be noted that in this tradition Conall Cernach opposes Cimbe Fourheads, while Conall Cernach's father, Amairgein traditionalls wa , y repute havo dt e slain the three-headed Ellen. A reference to the naming of the loch after Cimbe also occure Bookth n si of Invasions:

"... Loch Cimme, from Cimme the four-headed, son of Umor, was it named."

2. "God" Heads.

(d) The Head of Sran the Blessed.3 That Bran "Raven deita s origin yi "wa evidens ni t fro qualitiee mth d san powers attributed to him. He is described as being so enormous in size that no house could ever contain him. Whe goee n h attac o st k Irelan avengd dan e eth ill-treatmen s sistehi f ro t Branwen, messengero t r d ashe an k sr comhe o t e interpret for them the meaning of the unfamiliar objects they have seen approaching their land. They ask her "what was the mountain that could be seen alongside ships?e th " "Bendigeidfran brother,y m , saide "sh , " coming by -wading. There was neve shia r whicn pi mighe hh contained.e b t " "What walofte sth y ridge lake eacn th e o d hridge?e an sidth f eo " '"He, saide "sh , "looking towards this island; he is angered. The two lakes on each side of the ridge are his two eyes, one on each side of his nose." AfteIrise Welse th hth rd havhan e virtually exterminated each othea n i r fight, seven Welshmen alone escape, Bran amongst them, mortally wounded. He command companions shi t ofhea s takd fcu hi dan t o ewitsi t h the thein mo r travelsprophesiee h d thingan , w sho s wil wito g l h futuree the tellth e mH n i s . them: "In Harddlech you will be feasting seven years and the birds of Riannon heae singinth dpleasand s a youwilo gt An e b l. s t eves companwa a t u ri yo o yt

t Gwalea d Penfrn An si wil u fourscor. e ob lyo me t bes a n o t s ewhe yearswa t ni ; Revue1 Celtique, xrv, 398 ff. edited from the Book of Leinster. 2 MacAlister, R. A. S., Lebor Gdbala Erenn, IV, Dublin, 1941, 70, 110. 3 Branwen Uerch Lyr, Pedeir Keinc y Mabinogi, ed. Williams, Ifor, Cardiff, 1930, 29 if., 40. 38 PROCEEDINGS OF THE SOCIETY, 1957-58.

and until you open the door towards Aber Henfelen, the side facing Cornwall, bidy eheae ma thereth u dd yo wittmcorruptedu an , hyo t fro time u Bu mth . e yo have opened that door, you may not bide there; make for London to bury the head. And do you cross over to the other side." And then his head was struck othee th r r fo side t theou ,d off thest an y,se eheae seventh dd witan , h themd an , Branwe therd eighte an e. nth . the h. y passe fourscore dth e year thao s t they were not aware of having ever spent a time more joyous and delightful than that. t morno s e irksomwa t I e than when they came ther fellos coulr hi tel ey no wy dlb an that it was so long a time. Nor was it more irksome having the head with them than when Bendigeidfra d beeha nn with them alived becausAn .f thoso e e fourscore calles year Assemble wa dt th si Wondroue th f yo s Head. The talismanic qualities of this supernatural head are illustrated by the following statement: ". . . they came to London and buried the head in the White Mountburies whed Three .owa e th f An t dni on . e tha s Happwa t y Con- e Threth f cealmentso e e Unhappon d an , y Disclosures disclosed s whewa t ni r fo , no plague would ever come across the sea so long as the head was in that conceal- ment." This legend, appearin medievaa n gi l Welsh tale uniqus i , containinn ei g verbal verificatio l thaal s implicif i t e materiao n th n i t l representation e humath f o sn head. Bra decapitatioe stor e originn th clearli th n s i f d yd o go an , ya probablns i y latea r rationalisatio originan a f no l cult legend abou wonderfua t l supernatural head. In this head are combined all the powers with which the human head was accredited by the Celts. It is apotropaic (it averts danger of invasion); it is prophetic (it reveals future events); it is divine (it presides over the Otherworld feast at which the wonderful birds of Riannon, the Great Queen, British equivalent of the Gaulish horse goddess , sing ceaselessly). The name Beridigeitfran, Blessed Raven, carries, a Christian connotation. It is probable that originally the element Pen was present, the original name being Bran the . . . Head.1

Heade Th ) o(e f Per c Maile gaMa l Duin.z An oblique referenc actuao et l worshi humae th f po n head occurIrise th hn si story Cath Almaine, preserved in the 11th-century Yellow Book of Lecan. The Leinstermen fight a battle against Fergal Mac Maile Duin, who is slain. Fergal's head takes i Cathalo nt . Cathal honoure heade dth , washecausine b o d t dt an gi braided and combed smooth, and a cloth of silk to be put about it, and seven oxen seved an n wether seved san n pigs were cooke placed dan d before Fergal's head. The head blushe presence th Munsterme e n di th f eo opened o t nan d s eyedGo it o st give thanks for the reverence and great honour that had been shown it. Although concerned with characters living well withi e Christianth nd eraan , although the episode is given a Christian slant (the head giving thanks to God), this strange episode suggests earlier traditions of the actual worship of heads evidence placine th y animadf b g o l offering face s befortTh thagives i . t eit i t na s honoufeasit n i t r suggests memorie traditionf so f divinso e heads e (sucw s ha have seen in the case of the head of Bran) presiding at otherworld festivities. r goddeswa e s alsi sTh o associated wit e heah th f Ferga followine do th n i l g verse:3 t midda"A Alienn yi , contendin e cattlth f Bregar eo gfo e red-mouthedth , , sharp- countenanced Badb uttered a shout of exultation round Fergal's head."

1 Bromwich, B., "The Welsh Triads," Arthurian Legend Middle e inth Ages, . Loomised ff4 . 4 ,

3 2 Revue Celtique, , para. 54 24 . 14 .p , . cit.,Op 54, para. 14 . THE HUMAN HEAD IN PAGAN CELTIC RELIGION. 39

e thre oe th f on eBade s IrisTh bwa h war/fertility goddesse frequentlo wh s y appeare n biri d d form supra4 (se 3 r furthe . efo p r detail f theso s e goddesses' association with decapitated heads). e Head(f Th f Bonn)o Bo.1 Another accoun e continueth f o t d existenc a decapitate f o e d hea s alsi d o contained in the story of the battle of Almu. A young man, Donn Bo, famed for his sweet singin battle slais th decapitatedd gn wa ne an i promise d ha e H o dt . sing that nigh Fergar fo t lors matteo hi ld n r where th e f theo e y shoulOn . dbe victorious Leinstermen goe t ont battlefiele ou so th orden di tako t r eheaa d back to the feasting-place as a trophy. As he approaches, he hears the severed heads entertainin deae gth voic e dfielde kin Donth Th f eo n . gi o so nsweeteB r thae nth othey voican f r eo head . Whe warrioe nth r approache heads lifo s t hi heate th , d checks him saypledged s i an t ,s i sin o dFergat r gfo l alone that night. Nevertheless he liftheade takind th s an , buildinge t gbaci th o kt ,pillars a i place n t I o .t si aske singo d t t turndarknes I n wale i .tha s facs th o i s lit t o sing d ei t san o s sweetly that all weep. It is finally taken back to the battlefield. In this tale then, a decapitated head is placed on a pillar, it entertains at a feast singd an , s with such powe listepoweo e rth wh thanf s l o rweep ha al t t I . speechf movement,o d an , turning away fro e audiencmth whico et t hdoei t no s wis significans i singo h t t I . t tha thin i t s same talhave ew e another examplf eo decapitatea d head treated with utmost reverence, decke adornedd dan finese th , t of animals cooked and placed before it as an offering, at which point it speaks and thus indicates that it is capable of life independent of its body. It seems undeniable therefore that this tale, dealing with historical characters in a historical context attractes ha itselo dt f elements belongin purela o gt y pagan society, which show genuine traces of the veneration which was at one time accorded to the human head. e Head(g Th f Lomna.) o 2 The story of Lomna's head furnishes another example of a head speaking at a feast. Lomna was Finn's fool. By chance he came upon Finn's wife in the act of committing adultery with a warrior called Coirpre. Finn's wife begs Lomna betrao t faithfu s t i masters e no yhi h doehero d t t . l an ,Coirprbu ,s so e decapitates Lomn revengn ai betrayale th r takee efo H heae . sth Lomn f do a away with him. Finn finds the trunk, and placing his finger in his mouth, divines that it belongs to Lomna. He sets out to find the head. He comes upon Coirpre cooking salmon a cooking-house n i Lomna'd an , s hea s stucdspia i n tko besid e him. Coirpre, when dividing out the fish, omits to offer any to the head. The head speaks. This happen t outsidheae s pu twicth thir s e ddoore i d eth Th edan . distribution is made and the head speaks from outside.

3. Speaking Heads and Prophetic Heads. (h) The Head of Conaire Mor.s This legendary king of Ireland has many links with mythology, tradition making him the son of a bird/man, and a woman who was daughter of Etain, time on emythicae •wifth f eo l Midei LeithBr f ro . Conair trappes ei d into breaking 1 Op. cit. 2 Sanas Cormaic, , v. 34 Ore treith. 3 "Togail Bruidne Da Derga," ed. Knott, Eleanor, Dublin, 1936, from the Yellow Book of Lecan. 40 PROCEEDINGS OF THE SOCIETY, 1957-58.

all the prohibitions placed upon him when he was about to become king, and he is slain in a hostelry which is treacherously attacked by his three foster-brothers. The forces of the otherworld are heavily weighted against Conaire. He desires water durin Cechc attacke g Ma th t wate t d goe ge himr an ,o rfo st , buwatee th t r suppl Irelanf yo bees dha n magicall supernaturaa s i e yH drie . dup l being himself, and he ultimately obtains water for the king, but on returning with it, finds the king already decapitated. He pours the water into the mouth of the severed head, and the head speaks.

e Head(i Th f Fothad) o Canainne.^ Fothad Canainne was one of three sons born at a single birth and all named Fothad. Ther e manar e y example e Iristh hn i straditio f triplno e birthse th , same name being given to each son. According to tradition, Fothad's other name was Caindia, "Fair God." Fothad Canainne neve t dow feasa sa r o nt t without decapitated heads before him. He possessed a war-band, and waged constant war against Ailill Flann Be f Munstes followerso ehi d an r . Fotha s mordwa e shapely than Ailill, but AililTs wife was more beautiful than Fothad's wife. He makes an assignation with her, and abducts her. Ailill pursues them with his warrior Fothad san decapitatedds i falld an s . Ailill's wife lift Fothad'p su s head, t singi d lonsa an g extempore poe hermo t , lamentin stats ginstructinit d ean r ghe returo t n hom takd ean e Fothad's material possessions with her.

(j) A Prophetic Head in a Barra Tradition.2 The tradition that heads sometimes have supernatural powere b s alsi s o ot found in local legends current at the present time. The following tale was obtained islane th Vatersayf n di o , near Barrasummee th n i , 1956f o r , fro mwomaa o nwh is unusually gifted in local legends and song texts. A Barra man was going to bury the heads of his three sons who had been slain. He was carrying the heads in a sack. As they passed a standing stone in the Cliait district, one of the heads spoke and made prophetic statements. One of these was to disclose the way in which the deaths should be avenged. The father was to go to a certain township in the island, where he would find a man whose daughter was pregnant to the son whose head was speaking. She would give birth to a son, who would avenge the deaths of the three brothers. The father did as the head advised, and in due abous daye borns wa y tOn wa , fourtee•whe.n bo cours e so nth e eth n year agef so , he was helping.at a fanking of sheep. The murderer of his father and uncles was also present. This man asked the boy to go to the well and get some! water for him to drink. -The boy did this but on his return deliberately let the cup fall and it broke. The man then went to drink directly from the well. The boy bene h s t a dow d followe drinko an nt m hi d dre wa shor t sword d whicha e h conceale sleevs hi decapitate d n dei an murderere dth welle hea th e lef s .n i tTh d wa . interesf o s i nott o I t onles thaywa whet ti heaclose n i n th s di e proximite th o yt standing stone that it speaks and makes prophetic statements. There may have bee earlien na r tradition, confuse transmissionn i d , tha heae th t d was takeo nt the stone pillar and placed upon it and that it spoke.

1 Reicne Fothaid Canainne, Todd Lecture Series, xvi, FianaigecM, . Kuned o Meyer if.4 , This is a synopsis of a story told in Gaelic by Miss Nan MacKinnon (Nan Eachainn Fhionnlaigh) of Bail2e Vatersay. E HUMATH N HEA PAGAN DI N CELTIC RELIGION1 4 .

. Vengeful4 Heads. o accounttw r o f decapitate o se Theron e ar e d heads with powerf o s speec movementd han , actin antagonistin a n endangerind gi an y cwa e gth living.

) Suibhne(k Heads.e th Geilt d an 1 grimmese Onth f eo mosd tan t dramatic accounts of decapitated heads behaving aggressivn a n i e fashion occurIrise th hn si tal Suibhnf eo e Geiltrelevane Th . t passages are as follows: "... when Suibhne came to the centre of Sliabh Fuaid he stopped still there, and a strange apparition appeared to him about midnight; even trunks, headless and red, and heads without bodies fivd ean , bristling, rough-grey heads without bod trunr yo k among them, screaming and leaping this way and that about the road. When he came among the heare mh d them talkin eaco gt h other thid whas san i , t they were saying: 'Ha s ei madman,' said the first head, 'a madman of Ulster,' said the second head, 'follow him well,' sai thire dth d head, 'ma pursuie yth long,e tb ' sai fourte dth h head, 'unti t reacheli s the sea,' said the fifth head. They rose together towards him. He soared aloft in front of them . . . great now was the terror, the crying and wailing, the screaming and crying tumuld e headan th s f then sa o aloudt di aftey m e werhi th r, e clutchin eagerld gan y pursuing him. Such were the force and swiftness of that pursuit that the heads leapt on his calves, his houghs, his thighs, his shoulders and the nape of his neck so that the impac heaf o t d agains clashine t headth againsl d al f an ,gside topo e e tth th treef sso d san of rocks, against the surface and the earth . . . nor did they cease until he escaped from them int filme oth y cloudsky.e th f s"o "Once as I passed over Sliabh Fuaid on a black, dark, gloomy night, on a hill I beheld five heads having been cut off in one place."

Although this extraordinary episode occurs here purel literara s ya y fantasy, there is sufficient evidence to show that these "five bristling, rough-grey heads" with their power of speech and independent movement, have many antecedents in Celtic tradition, and represent in this context the crystallisation in a literary settin ancienn a f go t Celtic potence belie th decapitatee n th i f f yo d head.

(1) Colann gun Cheann.2 An interesting paralle heae th do t lepisod Suibhnen ei Geilt occur locaa n si l tradition concernin gsona g supposedly mad spectra y eb e which too fore kth f m o a headlesss body (colan cheann)n ngu . Like Suibhne e Headlesth , s Bodo ywh originally belonged to Trotternish in Skye, had a dislike for his fellow creatures, and because of his wickedness was banished to the mainland, and took up his abode near Arisaig. When the inhabitants there discovered his evil habits, they kille decapitated dan d himthee H n. becam ehighla y malevolen dangeroud tan s spectrer abovai usee e narroeH a floao th dt . n i t w pass throd an hea,s whi t da traverso t d peopl ha , killineit o ewh g the thiy mb s means youn A finalln . gma y caugh descendine th t gswords poine hea hi th onl d f n o td o an ,y returnee th o t t di Headless conditioe Bodth n yo n tha ghose th t t went bac Trotternisho kt .

1 Buile Suibhne, ed. J. G. O'Keeffe, London, 1913, 122, 130. 2 This tradition was recorded in Kilmuir by the School of Scottish Studies, and by F. Tolmie in the Journal Polk e oth f Song Society, 1911, xvi, 186-7. 42 PROCEEDINGS OF THE SOCIETY, 1957-58.

(m) The Knight of the Red Shield.1 In one version of the Scottish Gaelic tale of The Knight of the Red Shield the following episode occurs: "One looked hither and one thither, and they saw a head coining in a flame of fire, and another head singing the song of songs. A fist was struck on the door of the month toota king,e oknockes d fth butto o him f htheran wa n d o s t an ,golf nde ou o wa silve r do r on the coat of the king but showered off him with shame. The head did this; three years after each other, and then it went home."

This is an example of the persistence of an earlier tradition in a purely folklore context.

CONCLUSIONS. iconographie th , up m Tosu c evidenc whico et h attentio bees nha n drawn above, supported by the vernacular tradition, points unequivocally to the objechean e a cul th s f dworshipf a to o t symboa s a , divinitf lo regenerad yan - tive power, and as a focus of superstitious belief in the period just prior to and durin Romae gth n occupatio Britainbeef s no ha n t showI . n that heads figure as monuments in themselves, and that they also appear as decorations forming part of more important objects having a ritual significance such as bowls, vats, bucket d cauldronsan s . Thes e depicteb head y a n ma sdi variety of ways; they may have two or three faces or heads; they may be horned apparentlr o , y equated wit classicae hth l Gorgon head; their powers of regeneratio repellind nan g evidemonstratee ar l contextn di s where they are clearly linked wit phalluse hth face triangulare th e;b itsely ma f , square, entirely circulare b y base broar narro d o brow,e th ma e an t th t d a I t w .a moustached and bearded, or clean shaven. The head may have abundant hair on it, or none at all. Ears may be exaggeratedly portrayed, or lacking altogether. The head may have a cup in the top to receive libations, thus testifying to its function as an independent shrine. These examples serv indicato et e varioueth s way whicn si h this native concept manifests itsel religioun i f s iconograph earln a t yya period througd an , maze hth f eo miscellaneous evidence, and the variety of representational traditions, clear patterns emerge e -weighTh f .evidenc o t e enable o concludt s u s e that, equally -with Gaul, Britain and Ireland shared in this fundamental Celtic concept of the human head as a symbol of the life-force, and of the powers of the other world. It is clear then that this Scottish tricephalos constitutes further evidence for -what is already firmly attested, and it is likely that it is early in date. Found in Scotland, but small and therefore easily portable, it may have been brought ove thio t r s countr y Gaulyb s escaping fro e Romansmth . 1 Campbell, J. F., Popular Tales of the West Highlands, Edinburgh, I860, II, 469. E HUMATH N HEA PAGAN DI N CELTIC RELIGION3 4 .

Again, it may have been fashioned in Britain by Gauls or by native Britons, or it may have found its way over to Scotland from Ireland at an earlier or later date. The attention drawn in recent years to several very impressive tricephaloid monuments from Denmar apparentlf o kl 1al y early date provides yet another possible place of origin, if not of actual manufacture for this thin i head sd casan , t coulei d well have reached Scotlan Vikinn di g times. These are speculations which the geographical identification of the stone from whic tricephaloe hth fashiones si hely vitae clarifyo pdt ma th t l poinbu , s i t that this little three-faced shrine provides furthe rcule prooth t f statuo f s of the human head in Gaul and in the British Isles.

Since completing this paper attentioy m , bees nha n drawn Roberr byM t Hogg Tullie ofth e House Museum, Carlisle anothero ,t typ e heath ef d o wit h whice ar e hw concerned, whic excavates hwa autume th n di 195f no t Brougha8a mWestmorn i - fragmena s landi t canopieI a f . o t d tombston nativn ei e style whic recoveres hwa d from the cemetery site of the Roman fort at Brougham. I am indebted to Mr Robert Hogg for kind permission to publish this photograph and for this infor- mation (PL VI). I am indebted to the Carnegie Trust whose generous support makes this programm f researceo h possible amongsd an , t othere Keeperth o t s s e followinoth f g museum r kinsfo d co-operation, photograph d permissioan s o nt publish photographs: Mr R. B. K. Stevenson, National Museum of Antiquities, Edinburgh; Dr David Smith, Joint Museum of Roman Antiquities, Newcastle upon Tyne r RoberM ; t Hogg, Tullie House Museum, Carlisle r WymerM ; , Reading Museum; Pastor Hogsbro 0stergaard, Grlejbjerg, for much information photographd an Danisf so h material particularlm a I . y gratefu r WinifreD o t l d Templ manr efo y profitable discussion informationd san Professoo t d an , r Stuart Piggott for generous help and co-operation in all aspects of my research. Plates III and IV were kindly photographed, on my behalf, by Miss K. M. Dickie, P.S.A.Scot. r MalcolM d man , Murray, Departmen f Prehistorio t c Archaeology, Edinburgh University. Other photographs are reproduced by kind permission of Mr Robert Hogg, Tullie House Museum, Carlisle; Dr Joseph Rafferty, National Museum of Ireland; Mr W. Bulmer, Corstopitum Museum, Corbridge; Dr David Smith, Joint Museu f Romao m n Antiquities, Newcastle upon Tyned an , Mr R. B. K. Stevenson, National Museum of Antiquities of Scotland.

1 0stergaard . H0gsbroK , Trehovedetn E , Gudl Kuml, 1954 ff5 .,5