16 PROCEEDING SSOCIETY OTHE F , DECEMBE 19028, R .

1.

NOTES E EAHL'ONTH " S BORDLANDR B(O U T ORPHIRA ) , , CALLED ORFJARA IN THE SAGAS, AND ON THE REMAINS OF THE ROUND CHURCH THERE. BY ALFRED W. JOHNSTON, F.S.A. SOOT. e Orkneyingth n I a Saga s mentionei t i 1 d (A.D. 1127-28) that Earl Harald r s bitOrfjaran livei hi t dn a 113I .6 :— " Earl Pall had a great yule feast, and prepared for it at his 16,, which is called Orfura (lorfiara, Flateybook) . . Ther . . e were large farm-buildings (husa-baer) there d thean , y stoo slopinn do g ground a slopd ther s t an ,a ewa e the back of the houses. And when one came on the brow of the slope Orridafjord abovs e wa other t eth (i.e.,i n r o side, beyond t i lie sn i Damisey); . . . . There, in Orfjara, was a large drinking hall, and there was a door near the east gable, on the southern side wall, and a splendid church stood before e hal th descendee l on door e d on an , ds e churcha froth e halt th mo t lBu . entered the hall a large slab was to the left, and inside many large ale-casks, facind an e out-doogth a room. Sags e wa rTh a " relato t goe dispute n th eo s e o Sveinnsbetweetw e th n, resultin e slaughteth n i gf Svein o r n breastrope, after which Sveinn Asleifarson was taken by his accomplices " into the room s therfacin wa e out-door ee th gh pnlle d an d, out throug a skja-window.h 2 There Magnus Eyvindarso horsa s enha ready saddle d tooawam dan hi k y behino Orridafjordrt n housee o d th d an s . Then they took ship d Magnuan , s brought Svein Damisey.o nt " 3

In order to identify the site of the Earl's bii in, or called Orfjara, t i wil e instructivb l o compart e d residence an a lisf eEarls o s t bu ' s mentioned in the Saga, with a list of biis, bulls, or bows and bordlands of the Earldom, enumerated in the Rentals of the Earldom.4 With regard to bordlands, Captain !\ W. L. Thomas writes:—" The Earls of Orkney must, from an early period, have had mensal farms, and these are marked in the Old Rental as ' bordland,' literally table-land; 1 Icelandic Sagas relating to the Settlements and Descents of the Northmen in e Britisth h Isles. Edite Vigfussoy db d Dasentan n . Publishe e Masteth y f b do r Rollse th . Orkiiey Saga,. 70 , 69 caps , 58 . transparena 2 s i Skj ) (r a t membrane stretched acros swindowa d usean ,d instead of glass. 3 MS. translation by Ion Stefansson, Ph.D. 4 Peterkin's Rentals. EARL'OEPHIEE T NOTEA TH U N SB SO . 17

thu e Bulth s , B61, Nf Orfero . , wher e Earlth e s usually dwelt,1 was- bordlaud."2 Colonel David Balfour writes:—" Bordland, N. Bord, mensa, cibus. The guest quarters of the King or Jarl, and therefore exempt from skatt." s We find the faring about the Islands in 1137, col- lecting his land rents, when he would undoubtedly have resided at Ms- bus and bordlands. * As the islands were frequently divided among several Earls at the same time in accordance with udal succession they must have had their separate headquarters, for which purpose their bus and bordlands would probably be utilised.

Bordland Earldoe th f so m landed estate, which paid no skatt. Earls' Residences, llth and 12th Centuries. From Rentals 1503-1595. From the . [Earls' Palace Birsa.] Earl Thorfinnr (d. 1064), after his pilgrimage N.B.—This is not included in Romeo t , resided almost alway t Byrgisa s - e e th rentalth s a , herad, wher e builh e t Christ Church whole of Birsa was (chapter 37). clmrchland 1503n i , . Ear lByrgis-herat Pala . lII d 1137 (chapter 60). Earl Bognvaldr II. at° Byrgis-herad 1155 (chapter 108). [ Castle.] N.B.—Mentione n Charterdi s Earl Rognvald t a Kirkju-vag . I r r 1046 of Earldom. (chapter 34). , Bu of. Burwic Soutn ki h Ronaldsay. , including its Bu. Orphair—Bu of Orphair, with Earl Haraldr liveo wh ,d mostl Caithnessn yi , Midland, Hovvthd an , bs Orfjarn diedi hi t da a 1127 (chapter 58). Orakirk. Earl Pal s bi Orfjar. n livei hi lII t a d a 1136- 37 (chapter 69-71). Earl Roguvaldr II. at his bd in Orfjara 1154 (chapter 103). Earl Haraldr Maddadsod b hidinn e ni th t ga in Orfjara 1154 (chapter 103). OnlEarlso ytw , PalHaraldrd an l mentionee ar , s dwellinda t Orfjaraga a i t i ; 1 f Earo u l b Kognvaldr a spoke s a f no d Ear an , l Haraldr Maddadso n hidini s nwa g there at that time. Proc. Soc. Antiq. Scot., 1884, p. 254. 3 3 Oppressions Sixteenththe of Century Islandsthe in Orkneyof .and Glossary, s.v. Bordland. 4 O.S., chapter 71. 2 VOL. XXXVIT. 18 PROCEEDINGS OF THE SOCIETY, DECEMBER 8, 1902.

Bordland e Earldoth f so m landed estate, which pai skatto dn . Earls' Residences, llt12td hhan Centuries. From Rentals 1503-1595. From the Orkneyinga Saga. Netherlyking in Saudwick. Westrey— Swartmeill, Was- buster Bi d Rapnessf o ian , . Earl Rognvald t Hreppis-nea . II r n 115i s 5 (chapter 107), . Sanday—Gryndleithf o s Bu , ? Earl Rognvaldr I. in 1036 fared first to Brugh, Halkisness, Tofts, those bus which his father Brusa had Walls, Lopness, and Tres- owned (chapte tole ar d r e tha26)W t. Earl ness. e northernmosth Brus d ha a e t th par f o t islands (chapter 22). Stronsa Holland y— an d; probably Clestraid an n Musbuster s thea , y paid no skatt. ? Papa paid no skatt, Earl Rognvaldr I. killed at Papey in litlu, and probably bordland. s yul hi d gon wherer ha mal fo ee h e t 1046 (chapter 34). Earl Rognvaldu b n i 1136-3 a . II d r 8ha called Knarrar-stadir (chapter 81). This is supposed to be the modern Knarston near Scalpa in . In the Rental 1503 Knarstan t OlaviS n ei s paris s describehi d rege,o aspr i.e., Kingsland formed an , d part of the landed estate of the Earldom. It paid skatt, and was not bordland.

The Saga mentions the bti in Orfjara and the bu called Orfjara. The Kental of 1503 mentions the bull of Orphair. In one instance in e Kentath f 150o l e spellin3th f occursw Bur-ray.o Bo gw ,Bo viz. e th , e Kentath n s I lusei 164w d Bo 2throughout .e forth Buem s i .i n Murdoch Mackenzie's Charts, 1750; and Bu' is now adopted by the Ordnance Survey e pronunciatioTh . s uniformli n y 600d appearan , s always to have been so. The question is whether bull is derived from O.N. B61 or Bu. The Saga always uses bu, which, although a generic term, appears to have become associated with place-names in Orkney and Shetland.- However, Bu, as now used in Orkney, is still a sort of generic term—it is always the Bu of .such and such a place, precisel s use e a Sagae sam -th n y n th Orknedi I en wa . i y d Shet yan - stils acattler i lan1 sl fo l a dB6 b6 used n Icelandn .placei pe s a a , r - fo , name in Orkney appears in the termination bister — bolstadr. The II in EARI/E 9 1 NOTETH N S O SORPHIRT B UA .

bull appears to be a Scottish addition. In illustration of the Scottish influence on Orkney place-names, so far as the letter I is concerned, take vdgr pronounced voe in Orkney; in Scotch this becomes wma then wall, as in vdgur, wawis, walls, and Kirkju-vdgr, Kirkwaw, Kirkwall. The true words are preserved in the folk-pronunciation Waas and Kirkwaa,. In the case of Hr61fsey, the I has been absorbed in the foregoing long vowel, and we now have Rousey. In the same way as the Scotch pronounce gold gmod, a mistake y b n contra-analog e occasionallyw y find /old fowr fo d (i.e., foged).1 Boll, seea d pod, becomes bow, wheren i s a , many Scotch words e doublth , e changes i II a centra dy b int d o- w?An analogy the Orkney bii becomes bull. This Icelandic b& is still in use in some Scottish place-names, meaning a house or village, e.g., the Bow Ballingshaw.f o o of Bo Fife e th , 3 The earldom landed estate in Orphir parish consisted of a compact distric f threo t e tuns, viz., Orphir, Midland Houtond an , m , ru wit e hth r faro Orakirkf mo . l describeThesal e e Rentalear th n di f 150so d 3an 1595 as bordlarids of the old earldom, i.e., the Norse Earldom, paying no ekatt. This estat s separatewa ef Swanbisteo n e dtu th fro e n mo th r •east by a tongue of the Common and by a large tract of rough uncultivated land, and the Fidge of Piggar, stretching from the Commo ne coastdowth o .t n Fro e namth m e Grin = a( d house nea a e grinsurroundinrth r gato dn i e n wall)tu g , mentioned along with Hangaba n i 1503k , ther s i presumptive e evidence thae Commo th e timt on e t a nextende d e coastdowth o t .n There is also evidence that the Common behind also extended alone coast e hildowp th gth to l o ,t n dividin northe estate th gth n eo - west from Peterton. We gather this from the fact that Orakirk, which s situatei e norte shorth th f Houton o ho t n e do s quoylandi , , i.e.,n a enclosure from the Common, and the place between it and Houton is called Midquoy. Tha quot old Orakireviden is y an is kt fromits

1 Mtmoires de la Societe Royals des Antiquaircs du Nord, 1845-1849, p. 225. Munc Scottisn ho Irisd han h local names. 2 Jamieson's Dictionary of Scottish Language, s.v. Bow. 3 Ibid., s,v. Boo. 0 2 PROCEEDING E SOCIETYTH F O S , DECEMBE , 19028 R .

pennyland valuation e earlTh y. dat f thio e s valuatio s i losnn i t antiquity. Late valuedo s rt quoyno .e sar whole Th e estat Common,e bounde s easth e enorty e th wa b t th d n hdan o , e seafroe soutth th .my s b hn dividewhic o a wallwa d y t i hb an d, There are no walls separating the three tuns from each other, merely recognised boundaries—the bur f o Too-giln l between. Houtod an n Midland, and the ridge of a brek, called Glensbrek, between Midland and Orphir. It formed an ideal estate. Midland and Houton, with the only haven in that part of the Mainland—where King Hakon moored liis galley on his return from his expedition to Scotland—and Orphir, with broad land farmingr sfo , extensive meadows, hill-pasture, peat-groundd an , a good fresh-water stream,—sheltered by a range of hills on the north) (including the indispensable ward or beacon hill), facing the south, and wit hsafa e land-locke frontn i a d,se stocked with salmon trou othed an t r fish. Kerlin-skerry, belonging to the Bii, used to be noted as a place for seal-hunting. With regard to the name Orphir, as applied to the present parish of that name, we do not know when ecclesiastical parishes were formed in Orkney, nor do we know whether existing civil districts were adopted for that purpose. The Saga does not mention any ecclesiastical or civil district. Byrgis-herad does not necessarily refer to a civil district, as herad is used for any district, valley, or country, bordered by mountains, or withi same nth e river basin. s noticeabli t e nameI th f parishen o i es n Orknei s y that thee ar y mainly taken fro e dedicationmth churchese th f e namese so th th r f o ,o tuns in which the churches are situated. The parish of Orphir consist a grou f f o o tunsspd take an s ,nam it s e froe tiith mn of Orphir in which the parish church and the Bu of Orphir are situated. The Saga merely mentions the Earl's bii in, or called Orfjara, so that- name th e Orfjar clearls i aunie th t s yi e Homestead thath n f tu o t e Th . of Orkney topography. The original tun by enlargement, and sub- division through udal inheritance, became a group of farms. The site NOTES ON THE EARL'S BU AT ORPHIR. 21

of the original tun, however, remained the principal farm, the bii or head enlargee th f o u db tun. The Bii of Orphir must have been the original farm from which the tun took its name. The tun of Orphir is described in the Rental of 150 consistins 3a e bul th f Orphai lo f go r ninepenny groua land d f pan o , surrounding farms called the threepenny land of Orphair. presene Th t Bi f Orphio i situates e e Hop i rheae th th ' th f o edo t a d Bii, also called the Bay of Orphir. Between the farm buildings and the shore is the parish church and churchyard, and in the churchyard are the ruins of the Bound Church of Orphir. Immediately to the north, outsid yare eth foundationde th wall e ar , f extensivso e buildings recently excavated. The ruins and the church stand on a gentle rising ground e south-eas th fooe f th o t t a t declivit Midlanf yo ease d dth en tHillt A . of the site and of the ruins the ground slopes down abruptly to the stream e SagTh a. states thahousee th t s stooa slope n o d, whicy ma h eithe re declivitrefeth o t r f Midlanyo e risindth o Hilgt r grouno l n do which the ruins stand. The Saga description is extremely loose and brief t takeI . s Sveinn "away Orridafjordr,o behindt n housese o th d an " .and mentions a brekJea or leiti, a slope behind the houses, from which Damsey coul seene hala db fd , milean bot f whicx ho ssi e foud har an r •distant respectively. Beve Th . Alexander Pope states, iri 1758, that larg deed an ep founda- tions were found undergroun e Bowe th landth f ,o n si nea de th r •church, but that there were no local traditions as to the Earl's Palace.1 Mr George Petrie states, in 1861, that the immediate neighbourhood of e Bounth d Church abounded with numerous trace f ancieno s t buildings e remain believee th Earl' th e b f so so dt Palace.2 Some years agoe th , writer pointed out to the Bev. W. Caskey, incumbent of the parish, e spoth t wher e Earl' th eu woul B s d have stoo n relatioi de th o t n Bound Church, in accordance with the description in the Saga. At that time there wer o indication n ee localit th y ruins d an f o yf an ,o s Torfaus, Hist. Orkney. Translate Revy db . Alexander Pope . 107p , , foot-note.

1 Achasological Journal, vol. xviii., Round Church of Orphir. 2 22 : PROCEEDINGS OF THE SOCIETY, DECEMBER 8, 1902.

previous excavations was unknown. In 1899 the grave-digger came across foundation e north-westh n i s t e churchyardcorneth f o r , lying from west to east. As this would correspond with the south wall of the Earl's Bii, Mr Robert Flett of Bellevue, Hon. District Secretary of e Vikinth g Club, mad excavationo tw e e eastwarth o t s n lini d e with supposee th d wall, wit e resulth h agais t wa than t i tstruck n 1900I . ' e write th r • Flet M d ran t mad a eserie f excavationso s , tracine th g buildings to their eastern extremity. In 1901 they made a careful excavation at that part where the doorway would be, in accordance with Sage th a description, wit e resulth h t that suc n openina h s founwa g d (ses excavateda e r walle planfa Th o s ,, . ,fig measure1) . s abou fee6 13 t in length westere Th . n extremit t beeno n s tracedyha t probabli ; y ter- minated at the west end of the churchyard, as no trace could be obtained outside. Beginning at the west end of the yard there is a continuous length of about 104 feet of dry built random-coursed wall, 4 feet thick, without footings or scarcement; this terminates eastward at the supposed doorway, oppositt possiblno s church e excavato et th ewa t I . e immediately eas f thio t fino e othest th d r jam f openingbo , owine th o gt present church roa fee6 d t wallt Bu 4e |. th e incheeast n th o ,o t s othe e roar th sidd f o ewalls foun e wa retur,th d n wal f anotheo l r building in the same frontage line. This latter building has walls' 4 feet thick t builbu , t with mortar d thereforan , e probabl f lateo y r date. The jamb of the opening showed no signs of a door frame or fastening. In a line with the jamb of the door on its north side, and standin ends foun1 wa 01 g, d a largpar f o te flat ston 5 inchee s thick.1 e bottoe foundatioTh th f mo f thio n s wal s levei l l withe th tha f o t Round e Churchdoorwayth n I . , alongsid e stonth n endf o eo e d ,an 2 feet 4 inches above the foundation level, was another flat stone 5 inches thick ,s sidelyinit n , o gwhic havy ma eh bee e thresholth n d of a door, or otherwise a portion broken off the stone on end. If it was the threshold, e Bithe th e flooi th f nmuso r t have bee 2 nfee 4 inchest - above that of the church, which would accord with the Saga statement 1 Is this the slah mentioned in the Saga. ? Fig . Pla1 . n and[sections, Orphir, Orkney. 4 2 PROCEEDING SOCIETYE TH F O S , DECEMBE , 19028 R .

that one went down from the Bri to the church. • As the foundation crose o th fdoorwae se westh wal th f o to t labous yi t level wit e flath ht e doorwaystonth n ei , probabl e flooth y e spacrs thi th level wa sd e an , "between the stone on end and the cross wall would be the ale-room mentioned in the Saga. If the stone on end is in its original position, then the fact that its lower portion is irregular shows that it was probabl e socketh y t underground n additionaa s i d an l, proof thae th t flat stone in doorway was the floor level. The fact that the wall above the level of the flat stone, and above the foundation of the cross wall, has iallen down, while below it remains intact, appears to show that the

Fig. 2. The two halves of the midrib of a Bone Comb of the Viking period found at Orphir.

lower portio s buriewa n d foundation belo e floowth r levell alonAl .g the north or inside of this wall the stones have fallen down, and are mixed up with quantities of bones, ashes, and oyster shells. The only articles foun halvee o ornamented th tw wer f e so eth d midri a bon f o be comb of the Viking period, shown in fig. 2, which were lying near the firs e Boune th taps th f ste f o edo p Church a roun d dan ; handlf o e deerhorn or bone from the westmost excavation of the wall of the Bii. tilp churcd U l ol 182 e h9th road r 'maseo , y gate,' crosse e streath d m immediately to the east of the clrarch, and then branched into two sections going round outside the north and of the church- yard, so that the wall now excavated was under the north road. NOTES ON THE EARL'g BU AT OEPHIK. 25

The whol coverew e no sit s i ed wit 5 feeh t dee f debrio p s above the clay upon which the foundations of the church and Bii are built. The Eound Church whicf o , l thahal t remain smala e apsd th lan es s i part of the Avail of the nave 011 each side (see fig. 1 and fig. 3) is undoubtedly one of those twelfth century churches built in imitation of e churcHole th th yf h o Sepulchre .

Fig. 8. Apse of the Round Church at Orpliir. With regare founde churche th th o d t f e Kevo r th , . Alexander Pope states, in 1758, that some thought it had been built by Earl Hakon after his return from Jerusalem addd an ,s that " Hakoii t seemedi , , chused seat.Orphis hi r " fo r Dr Joseph Antlerson states that Hakos hi n d "ha residence at O^hir,"1 and that " he seems to have resided " 2 there, and probably built the church. The late Mr B. H. Hossack, in his recent

1 Seotland4n Early Christian Times, Edinburgh, 1881, p. 29, foot-note. 2 Orkneyinya Saga, Translation, p. xov. 26 PROCEEDINGS OF THE SOCIETY, DECEMBER 8, 1902

work, Kirkwall e Orkneys,th n i apparently foundin r AndersonD n o g , states that Hakon built the church. Thus the surmise of one writer become face anotherf sth o t l thesAl .e statement Hakoo t s sa n livinn gi Orphir and building the church are mere inferences. We have no proof that Earl Hakon was ever even in Orphir. He may have visited Orphir, among his other bus", when o'n circuit -•collecting rents and skatts. We are not told where he lived, and merely know that he died in the islands. Earl Pall was living in Orphir in 1136, when the church was first mentioned, and his father, Earl Hakon, was the first and only Earl before that visitetimo ewh d Jerusalem f EarI . l Hakon buil e churcth t e h probabl afte o pilgrimags s hi rd ydi Jerusalemo et , betwee winterw fe na s afte murderee h r t MagnudS . 112c 1116n si d 3 an ,whe diede nh . The Eev. Francis Liddel n 179i l 7 suggested thae Earl'th t s Palace stood at Oback in the tun of Tuskerbister, in Orphir parish, as answering to the Saga description.1 Oback is on the slope of the hill from which Damsey can be seen. Possibly he was misled by a tradition which says that the Earl of Caithness was slain at Oback after the battle of Summer- dal 1529.n ei Liddelr M 2 l suggests tha ancientruine n a " f Th so t tower, ocirculaa f r form aboud 0 feean , n circumference,18 i tt Swanbistern i " , was probabl residence yth f Sveinreo i breastrope, from whic e placth h e took its name " Suenobister." This ruin, however, is the remains of one of the many pre-Norse broughs in Orkney and Shetland. From measure- ments taken by the writer in 1879 and 1901, the internal diameter is about 30 feet, which, with the 12 feet thick dry built walls, gives a cir- cumference of about 170 feet. Somehow or other, after this suggestion by Mr Liddell, Sveinn got locally confused with the Earl, and in a MS. map of Swanbister, in 1847, we find the Brough marked as the " ruins of Earl Sweyn's Castle." 3 The transition from " Earl Sweyn's Castle " o "t Earl' s Palac thes e "nwa 1 O.S.A., vol. xix. p. 394. Peace's HandbooTc to Orkney, Kirkwall, n.d., and J.. T. Calder's Hist, of 2 Caithness, 2nded., 1887 . 111-112pp , . late th e y Jameb p sma Johnsto. MS 3 f Qoubistero n , Orpliir e possessioth n i , n of James Johnston of Coubister, Orphir. NOTES ON THE EARL'S BU AT ORPHIR. 27

an easy one. Already 1842n i , ministee th ,parise th f o rh mentione th s " Earl's Palace " at Svvanbister.1 In the advertisement of the sale of the estate of Swanbister, iu 1844, it is stated, as an attractive feature of the- property "s n ancien,i thawa t i t t dayresidence th e sNorwegia th f o e n Earls of Orkney, the remains of whose palace are yet in existence."2 The new proprietor of the estate, buying it on that understanding, we- are therefore not surprised to find Professor P. A. Munch writing in 1845-49, that the inhabitants still show the ground of the Earl's seat at Swanbister. mattea s f A facto 3r , this spurious traditio s nevenha r become folklore. The Brough at Swanbister is called by the inhabitants- e "hillocth ' Brecknay,ko " fro neighbourine name th mth f eo g farm. traditiono Thern e ear s now moro n , e than there wer 1758n r ei M n ,i Pope'Earl'e th site f sth eso timeo Palacet s a , . Professor Munch located Orfjara at Swanbister primarily because he found there a large flat tract of land which was sometimes submerged by seae Orfjarad s derivativth an ,it r o , e Orfyri saye s ancientlh , swa y used oconsiderabla f e exten flaf o t t land t covereebba d y ,an dr t floo da d an d n prooi fo island refertw n o Norwayi t ss callew no ,d Oiferso, but anciently Orfyrisey, both land-fast during ebb. But, as will be shown late , Orfjara s derivativon r it r o , e Orfiri s soleli , y applicabl reee th f o et o r neck of land which connects a tidal island to the mainland, and conse- quentl onle yth y parallel names which Professor Munc adducn ca h e are those of two such islands. The proper designation for a low ebbing: shor s i ut-firi,e neuter (Cleasby fern., wrong) a lon, g foreshore where the water recedes far and leaves shallows running out from the beach.4 trace lanf Th o t d referre Professoy b o dt r Munc s callehi e Fidgth d f o e Piggar, from O.N\ low-lyinfit= g meadow beside water, which exactly describes the place. Swanbister is described in the Eental of 1503 as uclal land payin t mentioneno g s i skatt d s havinan ,a d g formed parf o t 1 N.S.A., Orphir. 2 Advertisement in possession of the writer. 3 Memoires deSocietea l lloyale des Antiquaires du Nord, 1845-49 . 252p , . 4 Dr Jon Stefansson, at present engaged on an Icelandic-English Dictionary for the Clarendon Press. 28 PROCEEDING SOCIETYE TH F O S , DECEMBE , R19028 .

the old Earldom landed estate. It was completely separated from the Earl's estate in Orphir. Professor Munch found that the site shown him, viz., that of the Brough, in relation to the adjoining site of a chapel, did t corresponno d wit e Saghth a description a o roo n r d therms fo an , wa e chapel between the Brough and the shore, so he accordingly accepted the site of the chapel as that of the church mentioned in the Saga, and placed the palace in imagination to the north, suggesting that the Brough might lave been one of the out-buildings. Professor Munch says that Orphir seem formen si r time havo st e been e commoth ne wholnamth f o e coast from Houto o Waulkmillt n t bu , •quotes no authority. It has already been shown that Orphir, excluding the parish name, is alone applicable to the tun of that name (in which Round e Bith an i d Churc e situated)ar h e inhabitantth , f whico s e ar h still sppke theiy b f rno neighbouring parishioner ' Orphi e th rs a folk.s ' t mos A name th t e could only have include e Earl'th d s estat f Orphireo , Midland d Houtonan , o whict , h latte s attachei r tidada l island, called e Holth f Houtonmo n Orfiriseya , o whict , e musw he tth loor fo k •origin of the name. Although Professor Munch visited Orphir, and consulte e Rentalsth d e makeh , mentio0 e Roun11 sth f o nd Churchr o , the Bii of Orphir, bordlands of the old Earldom. e SagaTh , before relatin e murderth g d Sveinn'an s s flight, appropri- ately describe e relativth s e positio e hous th d churche f th an eo n d an , islan f Damsedo o whict y h Svein f nwhico fled broughl e al ar h, t into the story. It says, behind the houses there was a brekka, a slope, or leiti, elevation a e horizoth n no n hidin viewe gth , from which Damsey coul seen,—the db e inference being that thi se roa th elevatio dn o s nwa by which Sveinn fled. Professor Munch says that it could only refer Ware th do t Hill t thabu , t Damse ye see e b coulth nt s a frono d , mit Keely-lang hills intervened. The Ward Hill is the highest hill on the mainland, and is a fjall, and not a mere brekka or leiti. The straight track from the tiin of Orphir to the Bay of Firth, by which pedestrians still go, is through the moor, and after a long, tedious ascent the sla hollor ko w betwee e hill nth f Lyradal o s Keely-land ean reacheds gi , NOTES ON THE EARI/S BU AT OKPHIR. 29s f Firto y h Ba whe(Orridafjordr e th n Damsed an ) y suddenly break into- view. The complete change of scene is striking. This can'only h& the elevation mentioned in the Saga, and it is part of the range of hills- which Professor Munch said prevented Damsey from being seen from the Ward Hill. No one acquainted with the locality would ever think of going to Firth by way of the top of the Ward Hill, or expect to get a vie f Damsewo e slayTh k fro mentione. mit e nearesth s i d t point, to Orphir from which Damsey can be seen, and it is on the direct and- shortest rout Firto et whicy hb hfugitiva e woul. dgo e followinTh ge derivatioremarkth o t d significancs a san n e of. the- word Orfjara are founded mainly on a correspondence with Mr Eirikr Magniisso n StefanssoCambridgf no J6 r D d nean :— Orfjara, derivative Orfiri, Icelandic, neuter, is solely applicable to the- reef or neck of land connecting a tidal island (Orfirisey) to the main- land. It is derived from or = out of, a negative prefix, and fjara = (1). low water, opposit higo et h water flod foreshore) (2 ,e par th f the o tr o -, beacwaterw d y lo covereonl dr t han a y , t higa d h water. Orfirisey would thu island,sb meaeb f o n" t "i.e.,ou showing abov e waterth e - line at ebb-tide. There is no difficulty in the way of the term Orfjara extending from the appendage to the adjacent parts of the mainland. The Holm of Houton must therefore be the original Orfirisey and the adjoining land Orfjara. But how came the name to be restricted or transferred to the present Bii and tun of Orphir, and the name Orfirisey discarded 1 o solutionse wholThertw Th e ) ee ar (1 distric.e Earl' th f o st estate- from Houto Orphio t n havy ma re been originally called Orfjara thed an -, name afterwards restricted to the Earl's Bit; or (2) Houton, adjoining the tidal island havy ma e, bee e originanth l e OrfjaraEarl'th d s an Bi, i first erected there, and afterwards shifted to its present site, taking the name n eitheI wit thesf . o hrit e cases, whe e nam t th nattache go e tho t de abode peopld an , longeo en r understoo exace dth t meanin f Orfiriseygo — which must have got lost very early in Orkney, considering that the same took plac n Iceland—thei e term Orfirisey became meaningles peopleo st ^ O S PROCEEDING E SOCIETYTH F O S , DECEMBE , 19028 R .

and Holmr took its place. Orfirisey, just outside Eeykjavik, has for a, long time gone under the name Effersey as though it were named after some person called Effer, which show utterlw ho s y unconscioue th s Icelanders themselves became of the sense of the old name. Even this island figured for a time as H61mr. Professor MacKinnon of Edinburgh University writes that therOrfiriseyn e ninte ar Hebridese r eo th n i s , where the name is changed in Gaelic to Orasa, the / and second r being •discarded—-f aspirated and r merged in s. The name on the maps appears wrongly as Oronsay. There is also an island Orfasey at the sout f Yelho n Shetlandi l . Orfiriseyjar e occurDiplomatariumth n i s Islandicum, i. 597. In Norway, Orfyrisey occurs in the middle ages, and Offerso in three places in modern Norway. Houtoi s i probabli e Icelandith y c Ha-tiin r higo , h tiin, whic- s i h descriptiv hile place th t lie li th n sides f so ea eo . This nam founs ei n di many places in Iceland, but is not recorded in any Saga relating to curiouNorwaya s i t I s .fac t that Hd-tunse almosth l al n Icelanti e dar small tenement lane s th f awithi d o manorian o r no mair o l n estated an ; what seems tolerably certai , thais n t whe e manoriath n l abods wa e -erected, Ha-tiinoldee th s r abode,wa evet i f n,i becam inferioe eth r house ;and remained so ever after. Even Midlan t withouno s di t significance. f o Fro e e nammon th f o e s farmit s Myre, Icelandic Myrr a swamp, , whic s stili h l descriptivf o e e placeth t i woul, e unsuitablb d r farminfo e g purposes d Midlanan , d would be an appropriate name for this unprofitable land which divided the b& from its Hd-tun or out-bu. The Norse term Medalland could only be given to a place which lay between two localities that had distinct names. Midlan 'mentiones i d s earla d s 1263a y , when King e quitHakob there,s y e nwa certaima thao e 1s w t n tha t thaa t t time, .an n 1136di , Orfjar s solelwa a y applicabl e Earl'th o st e Bii. Another important inference that may be drawn is that Midland implies a connected distric f threo t e places e Earl'th .) s (1 Thirefey o t ma sr three farms, (2) the tripartition and re-naming of the original district 1 Hdkonar Saga, Knlls Edition, chapter 328. NOTES ON THE HAUL'S BU AT ORPHIR. 31

oe middlf th Orfjara ) (3 e r placo , e betwee e originanth le Orfjarth d an a shifted abode Earlth f .eo hopew no ds bees i thaha nt I t i tprove r do , thae Earl'in th tu B s called Orfjara, mentione e Orkiieyingth n i d a Saga s i identica, l with the present Bu of Orphir—which in 1503 was described as bordlands of the Earldom—and that the foundations recently excavated at the Bu of Orphir and the ruins of the Round Church are those of e Earl'th s d kirkBi an o whicit , h they answe n everi r y particulas a r describe Sagae th n .di With regare preservatioth o dt e Roun e ruinth th f f o sdo n Church: presene th f o te eas d parisTh en t h church e usuath , whicf lo barn s i h - type, stands on the western half of the site of the nave of the Round Church e foundationth , f whico s e buriear h d underground. proposew no s i o pult d t I l dow e paristh n h churc d builan ha d new one further west, excavating the foundations of the round nave and preserving the ruin as an ancient monument. If the ruin of the Bound Church is to be preserved, something must be done soon, vaultee th a vers a n e apsdi y th precarious roo i ef o f s state, owino t g the action of the grass roots eating away the mortar and disintegrating masonrye th .