NEWLY DISCOVERED SCULPTUEED STONES FROM PAPIL, PETEY B . R MOAR, CORRESPONDING MEMBERd an , JOHN STEWART, M.A., F.S.A.ScoT. Read March 25, 1944. • At Papil, Burra Isle, Shetland, where Goudie, in 1877, brought to light the famous Papil Stone, now in the National Museum, there have been discovered Aprin i , l 1943 three th , e ornamente sculptured dan d stones whose photographs and drawings are appended. The stones, which were found during grave-digging operations, were underground e mai,th n slab being foun feefee2 6 d2 t td frofro an wese norte mmd th th en t hChurce sidth f eo h of Papil. It was standing upright in the soil, supported by two socket stones, facind an g bein p southto ge burieth , deptinchesa 0 2 o d t f ho . Thers ewa no sign of other slabs which might have formed the sides and ends of a cist, but three other socket stones, besides the two supporters, have at various times been found in the vicinity. About 4 feet away was found the second slab depicted round-headea brokee f o th , p nto d stone wit interlacen ha d cross in relief. The present churc t Papilha sinc e ,t beewhicus no e n s i 1920 hha , only dates from about 1815, according to the New Statistical Survey. It was built fro materiae mth olden a f o rl towered church know t Lawrence'sS s na . The tower is described as a "steeple" by ministers of Burra in 1654 and in 1794, and from the statement that it was "five or six stories high" it is assumed to have been a round tower like that of Egilsey, , or perhaps Abernethy and Brechin on the mainland of . It is said havo t ethref o bee e e non "steeple d kirks Shetlandn "i othee beino th , rtw g situated at Tingwall, the centre of the Norse archdeaconry, and Ireland, . Edmondston, writing in 1809, laments that it has been pulled dow builno t parise dth h church. This older churc t LawrencS f ho e stood north and west of the present building, and a third primitive building is traditionally repute o havt d e been situatee nortth n ht o d S sid f o e 92 PROCEEDINGS OF THE SOCIETY, 1943-44.

Lawrence's Church e onlTh y e . towere tracth f o e d building consistf o s small bits of lime and mortar, not a stone being left. There is no reason to believe that the 1943 slab was found in its original position any more than e 187th 7 stone, which when first notice Goudiy db lyins ewa g e soutth f ho church, decorated side uppermost, and had been used past living memory maro t famile kth y burial Baptise placth f eo t missionary.

THE STONES . PAPI1 . L CROSS SLAB. d VI.(Pisan ). V . PI. V, 2, with the supporting socket stones shows this stone as it is state havo dt e been discovered, though ther s somei e doubt aboue th t actual position of the raised socket stone on the right. The propped-up position, howeverp whicn to i onle e sockete y th t th hwilth i y wa s t d i , lfi an , of this socket stone, unlik e rest th es heavil i , y weathered e slas Th bi . rectangular f close-graineo , d sandstone fee3 , inche4 t stope lonn th a ,t ga inch less at the bottom, 1 foot 10|- inches high, 2 to 2^ inches thick, and weighs about 1^ hundredweight e ston fron e th s bee f Th eha o t n. dressed smooth edgp bees bace to eha th ,e n k th carefull roughld an , yso rounded. s beeha Eac nd chippehen distanca r inche0 dfo 1 f eo s e fro th fooe mf th o t slab, to a depth of an inch or a little less, leaving two projecting shoulders or tenons above of approximately 13 inches long, which fit the sockets of the supporters. These tenons have been ground away, back and front, for e bluna th f m o o t t wedge betterordet n stonei ,e fi o materiae Th rt .th , l of which is not found in Burra, but can be obtained on the Mainland a few miles awaygooa n i d s i ,stat preservationf eo onle th , y flaked portions being part of the top edge', and the only figured part damaged being the head of the third monk. The decoration in the centre of the cross is somewhat weathered d coulan , d onl followee yb feeliny db g wit e finger-tipshth . e Th . e desig buldoublth n i f ks o ni e relief, many parts being raised above eth others, as, for example, the hoods, hands of clerics, leg of monk riding, hind leg of horse, hems of robes, and decoration in the centre of the cross. The only parts incised are the satchel and straps, the reins and bridle, and the eyes and fingers. The background has been pecked away to an average depth of one-eighth of an inch. Although a photograph gives that im- raiseedgesperfectle y s i pression th t wa t da bu o , n crosye n i th ,leve s si n o l the whole surface. The space between the central boss on the cross and the surrounding ornament is recessed, but much worn. The spiral is entirely in relief. . STON2 PAPI . No EL INTERLACED) V. CROSS L (P . This important fragment, which sheds more light on the nature of the Papil settlement than any other find there, was discovered by Mr Peter Moar while making investigations abou Crose tth e s desige Slabth n Th i s .n i NEWLY DISCOVERED SCULPTURED STONES FROM PAPIL. 93

natur Irisf eo h recumbent monument stone evidentls th eha d san y formed part of a cross-slab with an ovally rounded top. The cross is in low relief. The greatest lengt s 16hi f inches, breadth 12f e thicknesinchesth d an , s varies fro mrighte materiae th f e lefTh 1n o th t inc.^to n brows hi o l n sandstone carvine th d mucs gi an , h better tha photogrape nth h indicates. The ston d beeha en e samfounth t ee Monksa dtimth . as e ' Slab, about 4 feet away t littlbu , e attentio beed nha n . paiit o dt

STONE No. 3. PAPIL CROSS STONE (SOCKETED). (Fig. 1.) This stone, which is shown in position on the right-hand side of the Cross Slab, in PL, V, 2, has two sockets at right angles to each other, one fitting behine on int crosse slae d dth oth b an . These two sockets are centrally placed with regard to v?/ur*1' e stone ende th th f so , unlik e otheth e r socket stone, where it can be seen that the mortise is neare narroe rth w end t musI .emphasizee b t d that the photograph is intended to reconstruct the nature of the stones and slab when found, attempn a t no buil o t s i whatevep du d an e th r origina havy righe ma el th beenn t O edg. f eo the Cross Socket Stone, at a point 7 inches from the bottom centre, the side had been ground away to a depth of approximately half an inch all the way to the top, except at a point 17 inches from the bottom centre, where a flat- topped circle or boss of 2^ inches diameter and hal incn a f h higbeed hha n stone leftTh s ei . approximately 5 inches thick here and the boss is centrally placed crose Th s incisei s . a o dt depth of one-eighth of an inch, and the dagger- lik e incisio th appearanc o t ne "tailindu s ei g spirae out.Th "l decoration abov e crosth e s extend curvature th f narroe o o st th d t ea w en s lightli e stone th d yan , incise d almosan d t indecipherable stonee Th ., lik s neighboureit , is 24^ inches long, is of sandstone, and, with the exceptio e circulath f o n r side boss e decoth , - ration is incised. Fig . Rubbin1 . crosf go socken so t stone. e lefTh t socket ston d originallha e y been beed longerha n t reduce bu e require, th o dt d lengt chippiny hb e th t ga broader thiendn O s .ston sockee e slabe tenoon th th filles i tf ,y no b d e secondth , whicfronn i , PIn demonstrates i hV i t. e e naturth sth f o e 94 PROCEEDINGS OF THE SOCIETY, 1943-44.

sockets, and there is another shallow unfinished socket on the side farthest away from the slab. The socket in front in the photograph does not fit the tenon of the slab, leaving the one at present engaged there for a hypothetical side slab. The unfinished socket does not seem to have been beginnine th crossa f t thigo bu ,s sidbeed eha n toole somo dt e extents a , o alte it fe outlin e shouldeth rth o d t bose an rs effect achievee th n o d extreme right side of the decorated socket stone. The other three socket stone e similaar s naturen i r , with three sides f the o s foun plainsocketee e mwa on dOn .d som dan e years ago, about 10 feet presene nortth f ho t find, wit s foothskula it t .a l Fitting ints oit three sockets in signpost fashion were rectangular flat slabs. One of these churchyarde stils th i n li squara , abouf eo inche4 1 t inches2 sidd esan thick. Two corners have been chipped away, leavin gtenoa abouf no foo1 ty b t 1-| inch, -which can be fitted into some of the sockets. Evidently these stones have been used at the same time and for the same purpose. The difficulty is unusuae th l number discovered—five. Some minor points which wil apparene b l t fro illustratione mth e b y sma remarked. upon. At the foot of the Cross Socket Stone are some marks resembling an Ogam letter. These are apparently accidental and made by som enature too th narro a n f i le o w chisel. Whil l thaeal certais i te th f no design abov crose ethith n so s ston shows plateee i th n t nmusi i , notee b t d thae latteth t s incompletei r s referenca , o figt e1 wil . l showe Th . design cannot be followed by feeling with the finger-tips. Although there appears in the photographs to be some design on the book satchel on the Monks' Slab, ther definitels ei y nothin actuae seee th b n no go l t stone. Sculptured stones now found in Shetland can be listed as follows:— complet3 e slabs—Burr Papir ao l Stone; Stone; Papil, 1943, . 1 ornamented Socket Stone (Papil, 1943). 2. Interlaced Cross Stones—Papil, 1943; South Whiteness (see Proc. Soc. Ant. Scot., vol. Ixxi . 369)p . . incise1 d cros slan o s b (early type). Papil Churchyard (see Goudie's Antiquities of Shetland, p. 49). 2 Ogam slabs—Lunnasting t Ninian'S ; s Isle (bot Nationahn i l Museum). 2 Ogam fragments—St Ninian's Isle. Discovere Goudiey db t lostbu , . 2 Ogam fragments—Cunningsburg Museum)n h(i . 3 Runic Stones 1—At Cunningsburgh (in Museum). 1 "Pictish" Symbol Stone—Sandness (figured by Low in 1774). 2 pairs of fragments with spiral ornament—Uyea; Museum (locality unknown). Now in National Museum.

1 A atone Low records at Eshauess, Northmaven, seems to have been a seventeenth-century memorial (Ancient Monuments Inventory). NEWLY DISCOVERED SCULPTURED STONES FROM PAPIL5 9 .

The findine presenth f o g t stones probably raises more problems than it solves, for to discover parallels we have to go far afield. Only three Scottish rectangular upright slab knowne sar t Murthlya , , Meigle Dulld an , , all in Perthshire, and two of these appear to have had the sculpturing in relief in a recessed panel in the upper half of the stone, as at Papil. There , howeveris o resemblancn , n topii e r stylo c f treatmento e knowA . n exampl sarcophagua f eo s fro t AndrewsmS supposea d ,an fro e dmon , Orkney, had grooves to receive the sides, after the style of a window-sash or easel, and were radically different in artistic style both from each other and from the Papil stone. There is no parallel for the socket stones, and these so far add something to our knowledge. Figures of monks with the . same kind of peaked hoods appear in a fragment at St Vigeans (Romilly Alien, E.G.M.S., Bressa e 240). p th n formed i , yan r Papil stone t Maugholda , , (Kermoden IslMa f eo , Manx Crosses, pi. 67), and, more imperfectly witr o , h a difference, on a few other Scottish stones. The crooked crosier, which accordin Coffego t y only became commo Ireland'aften i sevente rth h century, is found at Bressay, Papil, Brechin, Meigle, and on the afore-mentioned St Vigeans stone t boo,bu k satchels, thoug hcommoa n enough piec monkisf eo h equipment confinee Shetlanar e , th o dt d examples. It is in general design, however, that the stone stands out on its own. The long-shafted Latin cross, wit s flarehit d arm almosd san t square base, is unique. TherScottiso n s i e h e parallelnearesth d tan , approaca o ht resemblanc s i presentee y Irisb d h recumbent slabs from Inishcealtra, Galway, of twelfth century date. These, however, lack the slightly flared ends, and the base is in each case a trapeze, with the narrower end at the shaft. It seems, however, that it is to Ireland that we have to look for parallels re-entrane Th . t C-spira Papie th n lo lsla painstakingls bi y made by rule-of-thumb methods. The sculptor makes the first five volutes with same th e numbe turnsf o r ; the seeme nh realiso st e tha wile t havh t no l e e ninrooth er mwhicfo apparentle h y aime resulth ss thre d i tat an ,e imperfect volute witlase a d on httura manyso an nto . Ther similao n s ei r spira Scottisn i l h examplesparalleles i t i t t bu Castledermot,da , Kildare, "where the monastery was founded in the ninth century (Crawford, Carved Ornament from Irish Monuments, pi. xvi). Modified C-curves, like those e centruseth n e Irisdi e see e th desig b e crossh n ni th n examplesf ca no , , both as borders (Crawford, ibid., Durrow Abbey Cross), and as terminals to the arms in cross slabs; but their use to offset a central boss is a tour-de- force of the Shetland sculptor. The style of the horse (which, with due respect to a national journal, is not an " pony," for Iceland, if discovered, had had scant time to evolve a breed) is the best, of its kind on sculptured stones e nearesth , t Scottish approaches bein t Aberlemnoa g , Angus (E.G.M.S., pi. xxxv), and Meigle, Perthshire. The two odd wedge- shaped insertions and imperfect spiral, the pictorial accuracy and perfection 96 PROCEEDINGS OF THE SOCIETY, 1943-44.

of the design, the skilful work in relief at three levels all indicate an original' genius" who owed more to his own native skill than to technical training. There samth s eei accurac restfulnesd yan s whic earliee founs hi th n di r Papil stone, and in mentioning this a hypothesis may be put forward on artistic ground bird-headeo stw thas alonei e t th tI . d figures with axes, pickin bottoe skulla th t ga n i ,m t contemporarsectio lattee no th e f rnar o y with the rest of the sculpture, but have been added after the stone had been set up, and by an inferior artist. If this is the case, the stone, before becoming a museum piece in the capital (like the Stone of Destiny), may have in turn marked the graves of Christian Celt, heathen Norse, and Baptist! The simple incised cross with wedge-shaped ends which is found on the socket stone seemwithoue b o t s t parallel e mainlanth 1 01 s f Scotlando d , though there is something similar in lona (E.G.M.S., p. 400) and again in Eilean na Naoimh (ibid., p. 402). But it is old in Ireland, and appears on e Kilnasaggarth t Stone (Armagh), whic e modifiehth alss ha od C-spirals in the centre design of the' cross on the Monks' Slab. Petrie gives an example from Clonmacnois, doubtfully dated to the ninth century (Christian Inscrip- tions, vol. 1, pp. 27—8), and further examples from Aran, Limerick, Inish- murray and Inishcealtra. The spiral on the socket stone appears to be completely paralleled, if we accept what the rubbing seems to indicate, in a more elaborate form at Tihilly, County Offaly, not far from Clonmacnois (Crawford, Carved Ornament, s alsi ot piI e nint.foun th xvi . n h5) o ,d century cross of Bealin, Westmeath. It may be added that Kermode, among his Manx crosses, figures a stone with the wedge-ended .incised cross, with Anglian sevent eighto ht h century runes. This diminutiva cros s sha e Chi-Rho appendage, but it is probable that the spiral above the Papil Cross also marks it as of reasonably early type. With regard to the Papil Interlaced Cross, there is no great difficulty findinn i g almost exact parallels Scotlandn i t no , , wher onle eth y e crosth f so t typlona s ei a (E.G.M.S., . 386) p t agai bu , Irelandn ni . Practicalle yth same cross is found at Clonmacnois, County Offaly, and is figured in Craw- ford's Irish Ornament, pi. xiii, and in Petrie, vol. i, p. 44, where from the inscription a date of 916 is thought possible. In vol. ii, p. 43, Petrie gives another example from Inishcealtra, Galway, and he has a third from Lema- naghan, Offaly. .Irish antiquaries, who have inscriptions and records to go upon, regard this fairly common typ crosf eo tents sa e hb centuryy ma t I . noted tha Whitenese th t s fragment (Proc. Soc. Ant. Scot., vol. Ixxi .369. p . ) gives us another Shetland cross of this Irish type. The conclusion must be that a community of Irish monks had established itsel Shetlann i f t leasda thousana t d years agothad an ,t this community could hav linkw efe s wit mainlane hth Scotlandf do t surpris t thino s Bu si . - ing. The historical trend of communications has in general been south and west, as evidenced by Stone Age collective tombs, which in many cases must

NEWLY DISCOVERED SCULPTURED STONES FROM PAPIL. 97

be derived from the Boyne culture of Ireland, by the distribution of brochs and by the Norse settlements. Ogam writing, of which there are but 9 exampleeastere th l al n n scountiei Scotlanf so d Morae soutth f ho y Firth, 2 each in and Orkney, 1 in , and 3 in south-west Scot- land, is represented by 7 examples in Shetland. There are 300 examples in Ireland, mainly in Cork, Kerry, Waterford, and Kilkenny, 26 in Wales, 5 in Devon and Cornwall, and 7 in the Isle of Man. No arguments can be advanced for either Pictish or Columban missions from this, for Ogams were prior in general to monkish communities; least of all can one argue any reasonable insular inscription in any language, Pictish, Gaelic, or otherwise, relativele th t bu y numerous inscription botn si h earl latd yean Ogame th n si Isles afford another exampl historie th f o ec sea-routes along which ideas peopld an e moved. The evidence 'of dedications in Shetland cannot lead us far with regard to the early church there. Though Orkney is sometimes mentioned in the monkish record labourscene e th th f s e o sServanusa f so , Moluoc, Kentigern, Drostan, Colm, Brendan, Dotto, Bathan, Caemha r Conranusno , Cormac, and other half-mythical personages, and Ninian is credited with having sent preachea "tho rt e islands tha afae tar r off," Shetlan nowhers di e mentioned, nor have the various references been put to that critical examination without which the valuelesse yar dedicatioe Th . Papif t Lawrencno S o lt e doet sno help us at all. Groudie, in reporting the previous Papil stone, mentions a Lawrence born 619n i , whose nam saye eh associates si d with Laurencekirk. As St Laurence's Day was celebrated in Shetland of old on August 10th, Old Style, this cannot be the Lawrence given by Goudie, if he ever existed, and is not merely a garbled version of Laurence of Canterbury, Augustine's successor diedindees o i 619 n t i I Spanis ,wh e d. th h roaste s martyrwa o d ,wh o ngridiroa Emperoreige e 258th n th i f nn o i , whor r fo Valerian mn ma e th ; the Escurial ("gridiron") is named, who became a popular Norse saint. Known Shetland dedication followss a e ar s : Holy Root JohS , d6 n 6, St Mary 6, St Paul 3, St Matthew 1. Saints: Olaf (King Olaf Haraldsson of Norway) 7, Magnus (Earl Magnus Erlendsson of Orkney) 5, Margaret 2, Bartholomew 2, Nicholas, Gregorius, Sunniva, Hilary, Barnaby, and Lawrence 1 each. Of Celtic saints St Ninian is commemorated in St Ninian's Isle (locally until recently St Ringan's Isle), and a St Ninian's Chapel at Norby, Sandness, which seems to rest on nothing more definite than the evidenc Ordnance th f eo e Survey map t ColmS . s give i ea sixteent n ni h century reference to Cunningsburgh, and St Columba seems a later tradition churcfoe rth Olat S t Hillswickf fha o suggestiono Tw . Goudief so Columbt ,S a for Clumlie and St Levan for Levenwick, must be rejected outright, as they secureo n resn o tr basis than their resemblanc placo et e names t wilI . l be seen that the evidence of dedications for a Celtic church is practically negative. 7 VOL. LXXVIII. 98 PROCEEDING E SOCIETYTH F O S , 1943-44.

Pre-Norse historical. reference e invariablar s o Orkneyt y , wite th h exception of a nebulous Thule, which appears to vary from Iceland and Shetlan Saxoe th Norwayo 9 dt n 44 chief n I .s Ochth Ebissd a an sai e aar d havo t e lai rulee d th wastOrkne f 5 o r Orkneys,e 56 e e th sain s b yi i o dd t an at Inverness and giving hostages to the Pictish king. , In 579—80, 682, and 709 the isles are alternately attacked by the Scots and Picts. These, with well-enough attested reference greao t s t plagueshistoricae th e ar , l sources apart from monkish records of saints, until the thirteenth century "Heims- kringla" refers simply to "a great going to Shetland" in 872. Viking raids on Britain appear to have begun as early as 784, and Br0gger and Finnur Jonsson evidencn o , e sligh itsel n t i bott bu fh linguisti archa3ologicald can , would postulat epeacefua l Norse emigratio thin ni s century thema , e well- developed in Br0gger's Ancient Emigrants. f thiI granteds si , therdifficulto n s ei lata n ei y datingr stonesou r fo '. For evidence of a religious community living under what is likely to have been more troubled .conditions in Norse times we can take Clonmacnois in foundes Ireland nintwa e tentt th d 548n I di hn . an I h. centuries wa t si sacked three times by the Irish, twice by the Danes, and once by the two combined perio e s greatesth it t s thif do ye , swa t artistic magnificencet I ; survived eight other plunderings by Irish and Norman before its final destructio n 1552ni . lons firswa at sacke s thre n 795i wa d e t i time; s destroyed in the ninth century, when among other eventful things King Kenneth MacAlpin was buried there, the shrine of Columba was removed for safety, and at one time the island was temporarily abandoned. But in tente th h centur producins wa t yi g crosse peacefun si l obscurity. Forty- Man7 11 f xo crosse fivt ou e s listed by Kermod Scandinaviane ear e th d ,an Bressay Ston Shetlann ei d almost certainl . yNorsis e Man th f yeo settlers in Ireland and Scotland embraced Christianity very early, certainly much earlier than 995 officiae date th th ,f eo l conversio Earf no l Sigurd Hlodverson of Orkne Norwegiae th y yb n king, Olaf Tryggvason adopteo wh , simple dth e expedient of seizing and threatening to kill the Earl's son, and then carrying the boy off as a hostage lest his father should waver in faith. Therreasoo n s ei caso nt t doubt s Br0ggea , r does (Ancient Emigrants, "papae . 61)th p n ,o " name Shetlann si d ("papa Norsd Ol en " i signifie n da Irish e twelft priest)Dicuin th o d r han l no ,centur y Historia Norwegiaen i their reference theio t s r voyages, their white robe theid an s r bookse Th . name "Papil," which may be translated as "priest's dwelling," occurs in islande th Yellf o s , , Burra Fetlard immediate l casean th ,al n n si i , e vicinity of, or at no grea,t distance from, church ruins or sites. , namee Papd th three f an ,s o aar e islands addition i d an , n there are some shore names and one or two field- names beginning with the word s variatiooit r n "pobi." Eac examinee b h s merits o t it nam s n do ha e , as some names, especially kill names, are derived from Norwegian '' pappe''— NEWLY DISCOVERED SCULPTURED STONES FROM PAPIL9 9 . a breast (cp. Paps of Jura). But Bragger's wholesale stricture is quite unjustified, for -we have a large residue of names which do commemorate early priests thesef o e Papid .on ,an s li With regard to the Norse use of "Pettr" (Pict) he is almost certainly correct. Pict names were always give Shetlann ni e Norsth prey o et db - historic ruins r lonelo , y places wher humao en n ever dwelt. Pettigersfield in , for example, a hill slope with one or two "rickles" of stones, investigate 1938dn i , provide dpieca e of Neolithic pottery certai,a n collective tomb, and a probable Neolithic dwelling. Jakobsen's so-called Celtic place- names (Place Names f Shetland,o . 175-207pp ) prove almost invariabln yo investigation on the spot in particular instances to be disguised or Norwegian words, for Shetland maintained a close contact with Norway, with individuals holdin gmigratind lanan frod n di man go bott h countries, beginnine th seventeento e t th p f u go h century Shetlane Th . d dialect, rich in thousands of Norse words, has no Celtic, if one omits a certain common Norse- Gaelic vocabulary, exemplifie n Henderson'i d s Norse Influencen i Celtic Scotland, and a few borrowings through Scotch. Certainly the linguistic argument supports Br0gger's theory tha Norse th t e settlers came into an almost empty land. In these days, far removed from pestilence and folk-wandering onln ca y e surmisw , t whaea t becam numeroue th f eo d san well-organise moro d ew no dn ca peopl e builo e w brochs wh eth t t Bu . than surmise about those early father sougho wh s t lone isles where they could worship their Lord in peace.

ACKNOWLEDGMENTS. Mr John Peter Moar, of the County Surveyor's Office, Lerwick, has provided the careful scale drawings, and made the necessary measurements and rubbings without which this description would not have been possible. Major Magnus Shearer, Provos . Sinclairf LerwickN o t . J r ,M Count; y Clerk, Lerwick; Dr George Duncan, Aberdeen; and Dr W. Douglas Simpson, Aberdeen, have given assistanc variouen i s ways.