THE CLAVA CAIRNS1 by IAI . NWALKERC , M.A., F.S.A.SCOT. THIS pape divides ri d into three parts discussio:a generae th f no l affinities, sucs ha they are, of the Clava ; a discussion of their geographical setting; and some remark thein so r possible dating.2 n discussinI g possible affinitie cairnse immediatels i th e f o son , y e structh y kb uniqueness of the group as a whole compared with other groups of megalithic tombs. There are five outstanding features: firstly, the dual element of ring- and passage-grave in a group which seems otherwise completely indivisible; secondly, the us f corbellineo r roofingfo e passage-graveth g s (and paralle thiso t le equall th , y distinctive open chambered non-passaged construction of the ring-cairn); thirdly, surroundine th g circl free-standinf eo g stones; fourthly orientatioe th , f passageno - grav ring-caird ean n alike toward . quarter Se W sth fifthlyd an presenc;e th ,p cu f eo marks both on stones forming part of the cairn and on boulders in the general area of distribution. The nearest parallels to the ring-cairns are in the superficially analagous struc- ture Almerin si southern ai n Spain,3 though Blance4 believes that som thesf eo e latter were used as ossuaries rather than places for collective burial. Passage-graves in general can be shown to have an Atlantic-Irish Sea distribution, from Iberia north- wards,5 though the Irish Sea area is certainly equally an area of gallery-grave distribution. The use of corbelling to roof the structure is a which can similarly be Iristracee area th hSe an di alon Atlantie gth c seaboard: Rudh Dunain a ' Skyen ni , Barpa Langass in North Uist, Macs Howe in , and Achnacree at the southern end of the Great Glen, have this feature; so have the Boyne tombs in and . WalesN Sergenta L , Jerseyn ei , lie Longu lid eean Car Brittanyn ni tombsd an , Pavie Alcalth d n aan i a cemeterie Millares Lo Portugan se i th s d cemeteran l n yi Almeria.6 e burde1Th thif o n s articl originalls wa e y parsurveya f o t , nominall countiee th f y o f Nairnshire o s , Moray and Banffshire, in the and Bronze Ages made for an M.A. thesis at the Department of Pre- historic Archaeology, University of Edinburgh (Univ. Edin. Dept. of Arch, t^th Ann. Rep. (1960-1), 6-9). The Beakers and bronzes from the three counties shown on Maps I and II are taken from my thesis catalogue; Beakere th s fro adjacene mth t are Inverness-shirf ao e have been noted from various volume thesf so e Proceedings and the bronzes from the same area have been taken from the same sources, with the addition of several unpublished one whicr sfo h informatio . StewartC . E . nI exten ,M M.A.thanks y dm Mr , PH.D.o st , F.S.A.SCOT. 2 I am most grateful to Miss A. S. Henshall, M.A., F.S.A., F.S.A.SCOT., who had been studying Clava cairns independently as part of her survey of Scottish chambered tombs (Henshall, A. S., Chambered Tombs of Scot- land, I (1963) , 12-39) not on'y proofr f°he r ordee lettinn si se rewrit o e t r g m sectio e eth thn ni e for coma f mo - r muc studyn fo t how bu hel,r discussio d menhe pan n o t n durin writine originale gth th f go . Whero n e specific references are given, the relevant sources will be found in her work. I am also most grateful to Dr . BlanceM . B , M.A. . Coles, M F.S.A.SCOT. . J ,r M.A.D d , F.S.A.an , , F.S.A.SCOT. permissior fo , quoto nt e from their unpublished theses. Distribution map basee sar d upowit p O.Ssanctioe e nhth th ControlleMa e . th f no H.Mf o r . Stationery Office. Crown Copyright reserved. 3 LeisnerMegalithgrdbere V.d Di an , . G , der Iberischen Halbinsel, Sudenr I:De (1943), Taf. 1-7. * Antiquity, xxxv (1961), 193. 6 P.P.S., VII (1941), 1-49, cf. figs. 2 and 16. 6 Piggott, S., Neolithic Cultures of the British hies (1954), chap, vn passim. 8 8 PROCEEDING SOCIETYE TH F O S , 1963-64 e surroundinTh g circl f orthostateo a featur s i s e difficul paralleo t t othen i l r groups of chambered tombs: New Grange, one of the Boyne group, has such a feature Bryd an ,Angleseyn i Cellu Dd i , another Boyne tomth f bo e groupy ma , have had a similar circle.1 Kercado, a Breton passage-grave, also has a circle of

'54321 0

FIG. i. Map I. All the Clava cairns marked on this map are noted by Henshall either in her catalogue, tex r introductiono t , wit exceptioe hth possible th f no e exampl Inverfarigaif eo e gth d (se . ian i)eo p two possible example t Windhilla s , Easter Ross, P.S.A.S., LXXXVIII gratefu(1955-6)m a I o t . l 85 , Dr A. A. Woodham, M.B.E., B.SC., F.S.A.SCOT., for information on the latter. The two Beakers from the Gulbin Sand Findhore s th (immediatelf o . nW knowestuary e w froe th b no mo yo e nt t ar Aberdeen) - shire (see P.S.A.S., (1961-2)v xc , 305-6.

orthostats,2 as as a stone set on top of its cairn, as was apparently formerly the Millares Lo e r th sfo f Grange., tombso w 40 w casNe d fe - , t Nos1 e9 a 6an A , 3 7 . exampl e- hav e some suggestio free-standinnof g stones beyon cairdthe n kerbThe . Aberdeenshire Recumbent Stone onlCirclee th ye othesar r primarily burial monu- ment havo st ecircla orthostatsf eo theid an , r connections wil discussee b l d later. The orientation of the Clava cairns has its closest parallels among the Recumbent . ScotlanNE f thossd o d an souther n ei n Ireland discussiod ,an thesf no e Archaeologia, LXXX (1930), 205. Piggott, S., op. cit., 200. 31 Coffey, G., New Grange . . . and other Incised Tumuli ian Ireland (1912), 7-8, 9, 12. 9 8 THE CLAVA CAIRNS two groups will follow later. However, work by de Valera and 0 Nuallain1 has recently shown that all the wedge-shaped gallery graves of Co. Clare face between WNW.d an . majorite S th , . ThiW y sd betwee shoulan . comparee dnb SW d with, for example, de Valera's work on gallery graves in the N. of Ireland,2 where the orientation is much more varied, but the majority face between E. and NE., particularly betwee ENEd an . .n E Sufficient informatio orientatioe th n no n of the Breton allees couvertes, a type which must be ancestral to the wedges,3 is lacking but while there can be no connection between gallery-graves and Clava t anothecairnsye e b ,r Atlantic-Irisn y thia hin f ma so t a traditione hSe th f O . Almerian passage-graves planne Leisnere th y db s - Llan onlAtalaya o l ye tw o d a6 Millares Lo - fac d 5 es4 an SW. rese th :t faceastere th e compassne halth f o fe th , majorit thef yo m approximately SE. faca , t which Blance confirms. 4 studo bees N yha n orientatiomade th f eo Boynf no Grangw e Ne tombs e e th t a : cemetery,5 New Grange itself faces approximately SE., while the primary chamber at Dowth faces a little S. of W., the secondary chamber being parallel. It is uncertain where the chamber is at Knowth6 - it does not however face N. as Coffey7 suggested. A newly discovered passage-grave close by, Knowth Townland, faces SSE. In the 8 cemetery 13 of the 24 tombs remain in a state where the orientation

ca thesnotedf e o n b d e onlan , y one, Loughcre9 hal. face, W wf compasS e e o fsth th s- remaindee th f O . 10 . W °N r fiv S.° e, 20 facthre . . S eE . 10 eE d ° eac e an S. . on ,h E

and only two the NE. quarter. At Carrowkeel on the other hand, though in some 0

cases the orientations are not certain from the descriptions1 , the prevailing orientation facee on s - S. . , N Carrowmoreanothet seeme A b . o st rSE orientatioe 11th verf no y few of the 63 cairns is certain: of those which are, most face the E. half of the compass , SSW twod 59 facebue an d ,. t on Nos sS Fourknockan 6 5 . s I 1orientates 2i d approxi- mately NNE. Neither Fourknocks Iln or III is a chambered tomb13 but the former is rituaa l site with Loughcrew ware typicae th , l Boyne culture . A recently surveyed possible passage-grave cemetery at Bremore comprises a 14 large moun foud dan r smaller ones latte e signo passaga seen e :th f b so n rn i n eca formee t facthath y bu n eti ma r approximately NNW. (rathe Hartnets ra tha. nNE t suggested) ,15 Of isolated examples approximate orientations are Slieve Gullion16 SSE., Bryn Celli Ddu17 NE. and Barclodiad y Gawres13 NNE. The passage at the Mound of the

1 Valera, R. de, and 0 Nuallain, S., Survey of the Megalithic Tombs of Ireland, I: Co. Clare (1961), 106 and diagram at end of volume. Hartnett's remark (Antiquity, xxxvi (1962), 74) that this shows a 'tendency . . . towards the summer rather than the winter position of the setting sun' seems erroneous.

P.R.I.A. (C) 60 (1959-60), 9-140, figs. 2 and 3, PI. XXXV. 32 Valera, R. de, and 0 Nuallain, S., op. cit., 115. Blance, B. M., The Origin and Development of the Early in the Iberian Peninsula (1960) (un-

publishe4 d Ph.D. thesis, Universit Edinburgh)f yo , 113I , . 5 Coffey, G., op. cit., 3, fig. 2, and 46, fig. 26. 6 P.R.I.A. (C) XLIV (1943-4), I3i~3> '39- 7 Coffey, . G.cit.op , , 61. 8 Antiquity, xxxvn (1963), 226-8. 9 P.R.I.A., ix (1864-6), 355-79. xxi ) . cit1(G 0op x. (1911-12), 311-4?- 11 Wood-Martin Rude. e G.,W Th , Stone Monuments of Ireland (Co. Islande Sligoth d oan f AMU) (1888), passim. 12 P.R.I.A. (C) 58 (1956-7), PI. LXIV. 13 ibid., 266 and Antiquity, xxiv (1960), 115. J.R.S.A.I., c (1960)x , 79-3 refsd an i. P.R.I.A. (1956-7)8 5 ) (C , 264. l11a P.P.S., iv (1938), 241 . 31 , .No fig , 2 . 15 17 Archaeologia, LXXX (1930), 185, fig. i. 18 Powell Daniel. d E.G an , . . E.T , G ,, Barclodidady Gawres (1956), fig. t end)18(a . 0 9 PROCEEDING SOCIETYE TH F O S , 1963-64 Hostages, Tara, is on the E. side.1 Tibradden, it has recently been suggested,2 is probably a nineteenth century folly. othef O r tombs mentione givinn di g comparisons with Clava cairns approximate orientations are Macs Howe3 SW., Achnacree4 SSE., Rudh' an Dunain5 ESE. and Barpa Langass E. Of the tombs belonging to the Macs Howe group in Orkney (other 6

54321 0

LEGEND O e flaax t azor mould V halberd X unidentified D mould ' T sectioneiHrmlet C midribbed a

FIG. 2. Map II

than Macs Howe itself) wher orientatioe eth ascertainede b n nca , the approxie yar - mately Wideford Hill7 W., Vinquoy8 S., of Papa Westray9 SE. and Cuween10 Ojuanternessd an occurrence botTh . hE Boyne-likf eo t motifear t Holsa Papmf o a 11 Westra Edad yan y Church, 1ruine2a d site whic almoss hwa t certainly another Macs I P.P.S., xxi (1955) 6 Riordain, 16d , 4Tara:an Monuments P. e . HillTh S d ,e ed (y th . n 1960)o , first pag supplementarf eo y note. * P.P.S., xxin (1957), 221 and Antiquity, xxxm (1959), 141 and 283. R.C.A.M. (Orkney &c), n, No. 886, pp. 306-13, fig. 381, and refs. 43 P.S.A.S., ix (1870-2), 409-16, PI. XXIII. 5 op. cit., LXVI (1931-2), 183-213, PI. VI. * R.C.A.M. (The &c.), No. 224, pp. 73-75, fig. 137. ' op. cit. (Orkney &c.), No. 410, pp. 156-9, fig. 223, and refs. 8 ibid., No. 216, pp. 54—55, fig. 104, and refs. 9 ibid., No. 544, pp. 186-9, ng- 26ci, and refs. 10 ibid., No. 340, pp. 97-98, fig. 150. II ibid., No. 411, pp. 159-66, fig. 245, and refs. 12 ibid., No. 225, p. 65, fig. 119. E CLAVTH A CAIRNS Ql Howe-type tomb confirm suggestioe sth n plausibl botn eo h architectura culturald lan 1 grounds that the Macs Howe group must have a connection with the Boyne group. The scrap of pottery from with the face motif must fit into this tradition 2 thoug t standi s uniqus ha i t si Britainn ei . interestins i t I noto gt e that Newall3 believe sarsee sth n structur t Stonehengea e to be connected not with the midsummer sunrise as popularly assumed but with the midwinter sunset in the SW. Cunnington suggested that there might have been a 4 recumbent stone facing just W. of S. in one of the timber circles at Woodhenge. e outeTh r enclosur Bleasdale th f eo e Circle wit s Collarehit s burialn it dur d sha 5 entrance facing SSW. orientatioe Th prehistorif no c monument subjeca s si t deservin morf go e serious study than it has hitherto been given, but in many cases it must have reflected noth- ing more than a tradition - a tradition that may very well have been as little under- stood as that of the orientation of Christian churches and burials.6 The presence of cup marks both on stones forming part of the structure of Clava isolaten cairno d sdan boulders within their general are bees aha n fully coverey db Henshall — in particular one should note the convincing argument for regarding them as representing aspects of the builders' religious life unassociated with the actual rites of burial or tomb construction, and similar evidence is found among the Aber- deenshire Recumbent Stone Circles. A recent excavation of a Recumbent Stone . CorkCirclCo n ei note cupso surroundee dtw on , sub-wedge-shapea y db d groove, 7 recumbente th n o . Piggott alreads ha y note occurrence dth markp cu structuran f so eo l stonef so 8 chambered tomb Walesn i s Isle f Man th o e,possibld an , y Cornwall; a the e yar componen Boyne th f o et art-style occud an , r elsewher Irelandn ei Brittanyn i , d an , the Evora region of southern Portugal, as well as in Scandinavia and N. Germany. Cup-and-ring marksothee th n r o ,hand , belon traditioa o gt n outsidee thath f o t Boyne tombs and probably originate in NW. Spain and N. Portugal.9 There seems only one certain example of cup-and-ring marking on a stone incorporated in a Clava cairn — a kerbstone at Balnuaran of Clava NE., where there is also a wavy channel. The small stone circle at Balnuaran of Clava has one stone which has, among severa marksp cu lcup-and-rine on , g mark. lase cup-and-rind tTh an survep cu f yo g mark Scotlann si d appeare lons do a gag as 1882, when Romilly Alien publishe corpus.s dhi same 1Th 0e volume contained Jolly's survey f suco h Invernese stoneth n i s s district thesd articleo an , etw s still 11 remain the basis of any comments. Jolly noted12 three stones uncovered in 1881 in opening up new land near the farmhouse on the farm of Miltown of Clava, im- mediately S. of Balnuaran of Clava, and on the lands of which are the remains of two Clava cairns firse Th t . stoneexcellenn a n i , t stat preservationf p eo cu e on d ,ha I Piggott, . cit.S.op ,, 232-56, 324-36. a ibid., PI. XII. ,4 Atkinson , StonehengeC. . J . R , (is. 1956)ed t (an9 . 1960)8 , ded . 96 , 43 Cunnington, M, WoodhengeE. . (1929), 14. 6 Ant. J., xvni (1938), 154-71, 168, PI. XLVI. interestinn a r Fo 6 g discussio churcn no h orientation . cit.op , e xxxvse , i (1956), 205-13 refsd an , . . Cork7J Hist. A.S., ser2 . LXIV (1959), 1-27. 8 P.S.A.S., LXXXVIII (1954-6), 197-8. • J.R.S.A.I., LXXVI (1946), 62, 75. « P.S.A.S., xvi (1881-2), 79-143. II ibid., qoo-q6. 12 ibid., 336-9. 2 9 PROCEEDING SOCIETYE TH F O S , 1963-64 wit hsurroundina g ring, another witsurroundino htw g groov a ring d san e running from the outer ring, while two unringed cups appeared to have connecting , or in one case, an unconnected groove. It was possible to see that the cups had been excavated wit hshara p instrument somn i d e an case, s subsequently smoothed over. secone Th d stone, actuall yfragmena largea f o t r additioblockn i d ha , numerouo nt s cuplets largo tw ,e cups partly surrounde irregulan a y db r shallow groove, while eth third stone, again part of a larger block, was cupmarked on both sides and had 'an artificial groove' joining two cups on one side. Channels joining cups are noted on a few other stones scattered within the Clava cairn regiondifficult e vien i th t f wo tellinbu f , o y g whether the naturae yar artir lo - ficial, they are not noted here - Jolly himself seems to have been aware of the difference between artificial and natural grooves, though he apparently grouped them both-together. Two cup-and-ring marked stones were found at low water mark on the beach at Cummingston, between Burghead and Hopeman,1 and another three in the Roseisle district S. of Burghead,2 while a slab originally3 found at Clackmarras, near Elgin, has S-spirals of typical Boyne type on one face and one cup-and-ring mark and man markp y cu another n so additionn I . , anothe rcrescenta facs ha ebelod an , w this was originally a figure seven-like carving. In the most recent publication of this,4 parallels were drawn wit crescene hth t amon themotife e gth th n t a sunpub - lished Boyne tomb of Fourknocks I, Co. Meath. In fact this latter forms the mouth o statue-menhifa r like figure traditioa f o 5 n well know Atlantin i c passage-graves, and though Hartnett, the exca.vator of Fourknocks I, described the Clackmarras stone (under the name of Strypes) as 'of statue- type', the remarks by Young originae th n i l publication stonee th f 6o , tha crescene tth figurd an t e seven below were Pictish , correcte seemm o t , though Gordon7 believe same th crescenf e o s th e b o t t technique and in the same style as the spiral - Pictish and/or early Christian carvings markep cu n t o unknown.d no slab e sar 8 Cup-and-ring marks, therefore, are relatively rare9 in the Clava cairn area, but apar markp t cu fro e sm th which , take themselvesy nb scarcele ar , y diagnostic, only the Clackmarras stone shows any connection with Boyne art. The distribution of these 'artistic' manifestations, however, coincides with tha f Clavo t a cairns, witha slight peripheral scatter, and a similar coincidence occurs with the distribution of Aberdeenshire Circles, whil well-populatee eth d Laic Moraf ho y wit concentras hit - tion of Beakers and bronzes (Maps I and II)10 in the first half of the second millen-

. cit.1op , LXVII (1922-3), ii, note Enigmae Davidsonn di Th , oM. f . th,J e Cup and Rings (1939), 17. 2 Davidson, J. M., op. cit., 15-18; ibxd., 18—22; Nature, xxn (May-Oct. 1880), 122 noted in P.S.A.S., xvi

(1881-2) Davidsonn ,i 123d an , . M.,J . cit. op , , 17. P.S.A.S., XLI (1906-7), 172, n.i. . cit.op 4 , LXXXIX (1955-6), 444-6, PI3 . 4 LXIV d an 2 , P.R.I.A. (C) 58 (1956-7), 222-3, PI- LXXV.

Reliquary6 and Illustrated Archaeologist, 3 ser. in (1897), 41—47. 6 7 P.S.A.S., LXXXIX (1955-6), 445^6. 8 e.g. op. cit., LII (1917-18), 99-102 and refs. referencee Th 9 P.S.A.S.,n i s xvi (1881-2), 12 bouldero 3t t Califesa r HilGarded an l n Moor seem erro- neous, and descriptions of them have not been found. I must join issue with Mrs Stewart, op. cit., xcn (1958-9), 80, where she says that ' and moulds on th10e southern shores of the Moray Firth have a mutually exclusive distribution to short-necked beakers'. seee caseb e nn th A frot .sca , thimno II s si mapd an sI THE CLAVA CAIRNS 93 nium B.C s nearli . y clea f suco r h markings, e areexcepth ar peripherafo t e th o t l Clava distribution and the isolated Clackmarras stone. While the European dis- 1 tributio p markcu f s fairli sno y widespread d veran ,y probably mor thao s e s ni obviou t presentsa connectione th , s wit Boyne hth eAtlantie tombth d san c route should once mor notede eb . From this survey of the distinctive features of the Clava cairns, therefore, the nearest parallels lie along the traditional passage-grave Atlantic route from Iberia, a deduction which, n facti , adds onl detain yi previouo t l s observations. More specifically some similarities with the Boyne tombs of Ireland must be admitted, and Almeriae th n . tombSpaiSE f nso see havmo t e rather closer connections t musi t tbu , emphasisee b d are o than ther s ai n i t e know exiso nt passage-grava t e culture with enough parallels from which one can derive the Clava cairn settlement in toto, and thi st besid musse remarkable e b etth elacd unitparallelf an ko , yof s forgroue th , p n Scotlandi s markedlit d an , y complementary distribution vis-a-vis e Orkneyth - Cromarty tombs immediately on its northern border. This problem, however, is further discussed when dealing with dating evidence. Certain minor features of the religious ceremonies attached to Clava cairn burial may be noted. Firstly, the beliefs seem to have attached no importance to objects of us r materiaeo l hereaftere valuth n ei , thoug t Corrimonha staie yth n immediately abov 'silhouettee eth 'buria e hea representh y f do lma meaa t t offering opinioe Th . n soie oth fl analystperipherae th n o l chambee stainth n so r floo t Corrimona r s ywa

that they wer2 e causecuttine th y db g throug slighf ho t irregularitie land ol d e th n so surface, but Henshall suggests that they might represent vegetable remains. Secondly, no particular attention seems to have been paid to siting the tomb. Thirdly, there havmay e been little differenc outlooin e k between crematin inhumingand gthe deceased: both inhumation and were used,in the passage-graves, and cremation was used in the ring-cairns. In the one reliable inhumation burial, the bod croucheds ywa , wit facee hth , perhaps significantly, directed dow passagee nth , crouchin e t onlth No s yi whic . thif almosn go SW hsra e buriadu t l unusual among primary chambered tomb deposits articulate t seemn i thae a t b f o to bu ,st d bods ya opposed to the common chambered tomb practice of depositing bones previously mortuara lai n di y house. Fourthly tombe th , s see havmo t e been very markedly elece foth r t who judgo t , e fro amoune mth labouf o t r involvesometimee th d dan s very close placin tombse th f go , must have lorded ove sizeablra e community. While this is a generally admitted fact about chambered tombs, the apparent refusal to bury bodiemoro chambea tw er n s i o tha e sizaverage n thanf a on rth e o f to e Clava cairn, is unusual. Fifthly chambee th , r floor t Corrimona s Culdoicd yan h suggested that ture beed th fha n removed before erectin tombe gformeth e tha d flooe th an ,th tf o rr had been strewn with fresh sand at intervals and consolidated by stamping, which implies that certain rites were carriesomr fo t e dtimou e afte funerale th r . Thin si turn implies that the final blocking of the passage was not immediately connected

ver1m a y gratefu Alastair M o t l r MacLaren, F.S.A.SCOT. confirminr fo , severar gfo thid lsan references . alse H o confirm1 similae th s r concentratio Recumbene th n i t Stone Circle area.

P.S.A.S., LXXXVHI (1954-6), 204. 2 94 PROCEEDINGS OF THE SOCIETY, 1963-64 wit buriale hth infilline chambee Th .th f go passaged an r blockine welth s a ,s a l g of the entrance, shows that the lack of secondary burials was quite deliberate. At Corrimony the burial had been covered with flat slabs, among which were fragments of charcoal. This charcoal and that at the other sites must have been brought in separately casefro e remaine mon th , n Culdoichi s- distributios it , mors - nwa ere stricted tha ncrematee thath f o t dcrematione bonesth d an , s themselves must have taken place elsewhere. The cremation at Kinchyle of Dores also appeared to be under slabs, and a paving, possibly not primary, was noted at Aviemore. Lastly, there is the ritual use of white quartz, a rite which has a wide use in time, t whic bu soms hha e noteworthy parallels. Thousand piecef so s were foun t Corrida - mony behind, betweenkerbstonese fronn th i commof s d o t wa an , t I t .Druidn a - temple and noted at Gownie ami Sands of Forvie. It was found in the chamber at Achnacree and in the forecourt and over the mound at Rudh' an Dunain. At Bryn 1 2 Celli Ddu it was found all over the site, occasionally, for example in the passage, in deliberate piles. Childe thought the Boyne tombs might have had a quartz cap- 3 4 ping. Wood-Martin noted that virtually every tomb at contained 5 quartz associated wit burialse hth thad ,an t rock crystal alss o6wa used, both therd ean in other Irish tombs. At Kintraw7 at the southern end of the Great Glen, a cairn wit hfalsa e portal uprightsflankeo tw y b d , like thaminiatura f o t e Aberdeenshire , showed evidence of having originally been covered wite hth material alss owa vert I . y commo vicinite buriale th th n ni f yo , whiccrea s -hwa mation in the larger compartment of a trapezoid . A single large orthostat stands on the W. side of the cairn. Very interestingly the false portal faced SW., and the largest kerbstone portae . sidesth W t lA d . ther havsy an wer. ema S e e e th als n oo bee nrituaa l blocking largr fo , e quantitie carbonisef so d wood were found between the uprights. approximatele Th y . diamete12ft r stone circle nea Clyde-Carlingfore rth d cairn at Clach na Tiompan, Perthshire,8 contained quartz. A feature of these Perthshire circles is that one of the stones is often disproportionately large, with the rest gradu- ated down to the smallest diametrically opposite - at Clach na Tiompan the W. orthostat was the largest, but the position in other circles is not constant. Cam Ban, Glen Cochill, Perthshire,9 a large kerbed cairn with a surrounding circle of standing stones had quantities of white quartz incorporated in the cairn material, suggesting that the name perhaps preserved a long folk memory. The primary burial had been in a short cist with a massive capstone, and fragments of a long-necked beaker were ciste founth . n di At the Druids' Circle near , N. Wales, Griffiths10 found a large amount of quartz ranging from tiny chips to boulders up to 18 in. long. The entrance thio t s circle, whic jusban a s hdiamete i n s i t. kha oveft contiguou0 d 8 r an r s with, . cit.1op , ix (1870-2), 414-15. . cit.op 2 , LXVI (1931-2), 201, 209. 3 Archaeologia, LXXX (1930), 209. Childe * , PrehistoricG. . V , Communities of the British Isles (1940) (1957 6 , 6 reprint). 5 Wood-Martin, W. G., op. cit., 86. • ibid., 87-88, 23. 7 Unpublished preliminarr ;fo y note Discoverye sse and Excavation, . cit.op , d 1960, 1959,an . 5 7 4- P.S.A.S., LXXXVHI (1954-6), 124.

8 op. cit., xci (1958-9), 81. 10 P.P.S., xxvi (1960), 303-39. 5 9 E GLAVTH A CAIRNS but on the outside of, the orthostats, faces SW. The primary burial was a cremation Enlargen a n i d Foo 'secondaries'e d th Vesse f o e l on urn d ,Enlargen an ,alsa n oi d Food Vesse l a flaurn td ogivaha , l riveted . diamete1A 0ft . r circle nearby containe even da n denser concentratio f quartno z anothed fragmentsdi o s rd an , stone ring with outside bank. Another circle 40 ft. in average diameter, with a distinct ring-cairn like appearance, also contained quartz, though in much less quantity. As wil e noteb l d below a Recumben, t Stone Circl southern ei n Ireland also contained quartz, as possibly did an Aberdeenshire example, and into the same tradition must e artificiallfalth l y rounded ball f naturao s l chalk foun t Foura d - knocks I.1 It is interesting that Hartnett mentions their occurrence in megalithic sites, 'usuall passagf yo e grave type' Ireland,n i , including tomb t Carrowkeesa d lan Loughcrew face tTh . that such objects see occumo t roughln ri 2 y circular structures, whether megalithic tombs, cairns or stone circles, may not be coincidence. furthee On notede r b poin y t Corrimon:a ma t y non-structuraa ther s ewa l stone embedded in the passage. The main chambers at Bryn Celli Ddu3 and Carrowkeel F had upright standing stones of no structural function, and a central post-hole 4 with evidence of decayed wood was found at Fourknocks I5 though the excavator though t mori t e likel havo yt e bee supportinr nfo rooe gth f tha takinr nfo grituaa l pole. Centrally placed standing stones are a common feature of Los Millares, Almeria: Blance thinks they wer t eroo ritualno f d support an e Leisner, th s a s s 6 worte b thoughty h ma suggestin t I . g thauprighe th t t stones centrae noteth n di l areas at Balnuaran of Clava Centre and Sands of Forvie could belong to this tradi- tion t Kintra caire A . th f no w suggestin p post-hola ther to s e ewa th n gei parallels orthostate th o t s Grangatow pKercadd Ne ean o already noted. Boyne tombs often occu cemeterien i r Clavo d Grangs w a sa Ne cairns e e Th . group and are sited in unassuming positions, rather like Clava cairns, but Loughcrew, Carrowmore, Carrowkeel and Fourknocks cemeteries for example have outstanding positions.7 o definitivn t ye Ther s ea s survee southeri th f yo n Irish Recumbent Stone Circles recent bu , t excavation t Drombegsa Bohonaghd 8an takee b basia y s na s9ma for discussion. These circles vary in diameter from 8 to 36 ft. and more, and are characterised by a flat-surfaced block-like boulder, usually the lowest stone of the circl standineand g almost invariabl westerthe yin n semicircl monumentthe eof . Diametrically opposite stan dpaia portaf o r l stones tallescirclee e th th , f o t, usually set in its circumference but sometimes radially with their inner edges touching the circumference. The orthostats are more often slabs or pillars than boulders; some- times ther centraa s e i cup-marked e stoneb e morr lo y th boulderf e eso ma on d .an , encloses i e ditcha On y db ; none surrounds cairn r tumuliso . 1 P.R.I.A. (C) 58 (1956-7), 231, 235-7. 2 ibid., 235. 3 Archaeologia, LXXX (1930), 183-4. 4 P.R.I.A. (C) xxix (1911-12), 326-7, PI. XIII, 4. 6 op. cit. (C) 58 (1956-7), 208, 211-12, 213, fig. 4, 251-2. 6 Blance, B. M., op. cit., i, in. 7 P.R.I.A. (1956-7)8 5 ) (C , Piggott19d . cit.8an op ,, 193-4S. , - 8 J. Cork Hist. A.S., 2 ser. LXIV (1959), 1-27 and op. cit., 2 ser. LXV (1960), 1-17. 0 op. cit., 2 ser. LXVI (1961), 93—104. 6 9 PROCEEDING SOCIETYE TH F O S , 1963-64 Parallels between these Circles and the Recumbent Stone Circles of NE. Scotland were noted as early as igog1: in the same year2 one was noted orientated S. and in I9303 five all facing between S. and W. were noted. In 1931* one facing just S. of SW. was excavated revealing paving in the circle near the portal and through the portal d 'a an grea, t number quartzitf o ' e stones were also found, chiefly aboue th t portal. igsgn I orientatee 5on excavates wa . dN Oy db Riordai reveao nt trencheo tw l s at right angles to each other and originally taking timbers, the excavator thought to take an upright post. Beside this was a with two standing stones to the N. and under the cairn material was a circle of radially-set stones with the three biggest, represented by socket holes, on the N. lying in a straight line. 6 Riordain suggested that the Aberdeenshire Circles might have been in part ancestral to this group. Fahy's findings at Drombeg resulted in a fascinating reconstruction of the ritual whild an e recountesome t neeb i t f eo d no d her t doe i t beas e a knowsno n ro n Clava tradition compactee sth d floosingla f o r e laye stonesf ro levelline th ,site e e th th ,f go strippin s turit f go fro e entirmth e immediate vicinity e portaexcepth d r an lfo t immediately presencoutsidee th d fleckf an ,eo charcoaf o s lattere th al e t n a li ar l, least partially paralleled amon Clave gth a cairns. Fahy believe removae dth turf lo f havo t e been purel leve o ysitet e th l , however pavine th d havo gt ,an e bee t rituanno l but the covering for the well-concealed burial. The burial, which the excavator considered dedicatory, was a cremation in a pot of the Lough Gur Class II ware family, normally dated to the Early-Middle 6 Bronze Age but C 14 dated here to 13 B.C. ± 140 (with a proviso that it might have been earlier but not before 500 B.C.). The stratification virtually precluded it being perioo tw a d monument t Fah prepares ,bu ywa alloo dt slighe wth t possibilite th f yo circle predatin e buria th gd pavin an l g though also suggesting that Lougr Gu h Class II ware might be long-lived. The orientation Fahy noted was intimately connected with the midwinter sunset, being to the SW. and the tops of the three stones to the N. of the recumbent form a very accurate upward slope before the lower recumbent is reached. The profiles of orthostate twth f oo s suggeste fertilitda y cult. In the excavation of a stone circle and a place less than 50 yds. from 1site C 4 th e a datobtaine s ± 12 A.Df 9 0ewa o 22 . d from beneat moune he th th f do cooking positivplaceo n t ,bu e evidenc conneceo t t these features wit Recumbene hth t Stone Circle was found. At Bohonagh besides the circle there was a hut site, as well as a and an isolated cup-marked slab. Points in common with Drombeg were the stripping of turf from the site, the charcoal flecks outside the portals, the upward sloping top of the only flanker to the recumbent to survive, and the cremated central burial, thoug thin hi d undes casan n smala re ur wit o l hn mound e Th .

1 op. cit., 2 ser. xv (1909), 105-8. 2 ibid.. 59 , . cit. ser3op 2 , . xxxv (1930), 70-85. op. cit., 2 ser. xxxvi (1931), 9-19; plan in J.R.S.A.I., XLV (1915), 316-17. 54 op. cit., 2 ser. XLIV (1939), 46-49. 6 P.R.I.A. (1953-4)6 5 ) (C , 297-459. THE GLAVA CAIRNS 97 orientation was a little W. of S. but there appeared to be no continuous stone floor. The isolated cup-marked slab could have come from the Circle but there could be no certainty about this — the dolmen had a cup-marked capstone. As neither the t sit house ehu th her r e ecirclno t Drombee a occupatio y an d gha n refus conneco en - tion coul worts i provede db t i ht notinbu , g thalace f sucth tko h material could conceivably mean that if neither were habitations they might be ritual. lighe th presenf o tn I t knowledge therefor mose th esaie tb thas dthai n n i tca t their orientatio particulan ni othen i d r an rpart f theio s r constructio rituad nan l these monuments appear to have affinities with the Clava cairns and their namesakes in NE. Scotland, though the C 14 date presents an impasse. The Recumbent Stone Circles of NE. Scotland differ from their southern Irish namesakes in several ways, the most important probably being that they are primarily sepulchral. So many of the known examples have been so severely robbed, even virtually completely destroyed, almoss thai t i t t impossibl obtaio t e accuratn na e idea of their structure. Coles' surveys made between 1899 and 1906 are still the only published accounts of these monuments, for only a brief resume1 of Keiller's survey was ever published. The central feature of these Circles appears to have been a ring-cairn with a kerb inside and out. The most obvious features, however, are the orthostats and the eponymous recumbent block e formerth : , unlike those connected wit e Iristh h h Circles, are always highest on either side of the recumbent. Keiller maintained the latter blocke a passaged : this seems unlikely, y representhougma t i h a rituat l entrance. In some cases it is linked to the cairn by a kerbed construction, and plat- form fronn i s f the o te als mar o known, one2 possibly with quartz fragment. it n o s Keiller note ofte w e orthostatdho th n f quartzo e d alsar san o, noted causeways leading from certain orthostats outwards, includin apparene gth t linkin Druidsf go - field, Crookmore and Old Keig. Again he noted their unassuming positions. In view of HenshalPs accurate summary of these Circles, it is proposed here only to discuss certain aspects of them: their orientation, their dating and their relation- ship to the Clava cairns. The orientation is in a state of some confusion. A practical difficulty is that the recumbent is very often not tangential to either the arc of the orthostat caire th r nso kerb botd an ,h these latte oftee ar r n sub-circular. Further, the centre of the 'circle' is very difficult to estimate, and may be quite irrelevant to originae th l intention builders e assumptioe th th f so n O . n that such blocks must have been raised for a specific purpose, witness the platforms in front of them, and there- fort reasonablse e y accurately hers i t i e, assumed thastone th t e directly facee dth required direction, whic nearls hi y always confirme positioe th flankerse y th db f no . Of those plans published by Coles, c. 70% lie between SSW. and S. (including three which fac fractioea remaindee S.)f o Th divide. e nb E n rca d equally betweee nth approximate direction sSSE d WSW.an . . - onlexamplee ,SW yon Bourtred Ol , e Bush outsidy la , e these limit faced . an sKeille dE r maintained this monumens twa 1 Keiller ,Megalithic e A.Th , Monuments of North-East Scotland (1934), private reprin lectura f o t e given to the British Association at Aberdeen. ibid., 16, Keiller says Castle Frazer but gives no references, and I have so far been unable to trace the exact2 details. 98 PROCEEDINGS OF THE SOCIETY, 1963-64 no Recumbena t t Stone Circle t thougbu , h par Colesf o t ' interpretatio unlikelys ni , identificatioe th n itself seems reasonable. Keiller maintained, however, tha invariabln a t e . quarteorientatioSW e rth o nt was quite wrong, and noted 16 recumbents facing between 348° and 62°. Of these, nine are completely at variance with Coles' plans: Keiller gives for Yonder Bognie 355°, Coles a fraction E. of S.; for Esslie the Greater, 359° 41 mins. and SW. respec- tively; for Aquorthies, Manar o° and SSW.; Loudon Wood, Aikey Brae and Cairn- ton, Strathbogie, 9° and SSW.; Newcraig, Daviot, 22° and SW., and for Tomnagorn, 22° and SW. In the case of Loanhead of Daviot, Kilbridge-Jones' plans agree with 1 Coles', not Keiller's, and Childe's plan of Old Keig2 agrees with that of Coles. Rough plans publishe Petey db 188.n ri Loudof 5o n Woo Aiked dan y Brae3 also agree with Coles' orientations. Until, therefore, another complete survey of these monuments be made, some uncertainty remains on this subject. Apart from early and unfortunate diggings, only three have been excavated. At Garrol Wood, Kincardine, Coles4 foun a centrad l crematio a stone-line n ni d hollow, three other pitsn si , anothe deposia slab a d n o charcoarf an , o t somd an l e fragment 'undecoraten a f o s d urn l withial ' centrae nth Keig,d e l th ringOl l t 5al A . abundant pottery on the site apcirt from a few minute Beaker sherds belonged to the clas 'flat-rif so m ware' which Childe terme Keid dOl g . ColesJ war d ean 6 following Powell reclassifies 7ha Coveses da a ware t LoanheaA . Daviotf do 8 however quitea numbe possiblo Beakef tw o r r o e e rFoo sherdon dd Vessean s l sherds were found about the site, including the central area. Further, a Pygmy Cup in the form of a miniature Food Vessel was found in a burial secondary to the circle. This excavation came too late to receive more than a footnote in Childe's first work on Scottish pre- history which date Late th s e f thes o theiBronzn o d ee ren circlesassociae Ag th o t : - 9 tion witKeid hOl g ware. A fragmen Beakea f o t r bracer associated with sherds 'o reddisa f h colourd ha '

been found years Ra,ynebeford Ol t e:a further excavation t Loanheasa Daviotf do '

2 1 11

reinforced the evidence 10 that the site dated from Beaker times and that it must have been the centre of an area of prolonged activity into the Late Bronze Age and after. Childe recognised thi ig38n si I94Od an thougd manifeste an th n hi f 1944oo e "h 13 14 seemed less certain, subsequent editions of Prehistoric Communities of the British Isles have retaine earle dth y dating. Henshal sincs ha l e noted that four rims from Loan-

P.S.A.S., LXIX (1934-5), Pis. I and II, facing p. 196. 21 op. cit., LXVIII (1933-4), 381, fig. 9. 3 op. cit., xix (1884-5), 375~7- P.S.A.S., xxxix (1904-5), 190-203. . cit.4op , LXVII (1932-3), 37-5 LXVIId 3an I (1933-4), 372-93. ,.„,., ., 56 Coles, J. M., The Late Bronze Age in Scotland (1959) (unpublished Ph.D. thesis, University of Edinburgh): see Univ. Edin., Dept. of Arch, isth Ann. Rep. (1959-60), n-ia, see now P.S.A.S., xcm (1959-60), 44. 7 . Arch.PowelE . G M.L., l. I (1955)'Th T , 6 w e no Comin, e 17e Celts'se ,th . f 105-2e go pp , Th f 4o Prehistoric Peoples of Scotland (1962) (ed. S. Piggott); 116. P.S.A.S., LXIX (1934-5), 372-93-

8 9 Childe, V. G., The of Scotland (1935), 176, n.3- 10 P.S.A.S., xxxvi (1901-2), 530. cit.. "op , LXX (1935-6), 278-310. 12 op. cit., LXXI (1936-7), 401-5. 1S Univ. Land., Inst. of Arch. 2nd Ann. Rep. (1938), 36-39. " Childe, V. G., Prehistoric Communities, 100-1. 15 Childe, V. G., Scotland before the Scots (1946), 8, 41, 50, 103, 104, fig. 23. THE CLAVA CAIRNS 99 hea Daviof do t classifie Keid Ol gs da war e have fluting simila Lyleo rt s Hill warn ei Ulste d Grimstoan r n warsherde th Yorkshiren ei f o s t leasa o appea d tw t an ,o t r have come from primary positions. In 1951 Piggott suggested that the open 1 carinated bowl tradition originated in the Whitehawk ware of SE. England, spread nort Yorkshiro ht thencd ean Ulstero et , wit similae hth r bowls from Eastertof no Roseisle, Moray coming from Yorkshire. Without going into the matter in detail, though it is hoped to discuss the site at Easterton of Roseisle at some future date, I should prefer to follow Atkinson and 2 Corcoran suggestinn i g tha movemene th t t took plac reversen ei t leasa r to , thae th t 3 NE. Scottish material comes from Ulster. Fluted decoratio Irisn no h Neolithic potter commos yi Lylen no s Hill stylee th , thir Case'f do s style Westerf so n Neolithic war precedino e there,tw e t rarth bu gn e i 4 styles, Dunmurr Ballymarlaghd yan succeedine th n i d gan , style, Limerick (Lough Gur I/la). Case, using the recent C 14 dates for the Irish Neolithic5 suggests a florui thir e mostf th o t f do millenniu mLylee th B.C r s fo .Hil l style t Easa : t Finnercy Aberdeenshiren i Atkinson found Lyles Hill ware wit hscraa cord-ornamentef po d Bell Beaker whic6 Dutcn ho h evidence canno before tb sherd e 220. th ec f o s0 e B.COn . in question from Loanhead of Daviot has a lug, a feature very rare on Irish Western Neolithic wares, and it is quite possible that the Loanhead of Daviot material repre- sents a local version, but until much more is known about the Scottish non-Megalithic Neolithic, more Circles are excavated and C 14 dates are available little can be said with certainty. However, the resemblance between the false portal of the Circles and that at the Lyles Hill cairn has been noted, and the preliminary studies of the distri- bution of Tievebulliagh axes7 show that the Great Glen route from Ireland to NE. Scotlan uses Neolithidn dwa i c times. Much must remai discoveree b o nt associn di - ation with this trade. That there exists some connection between the Recumbent Stone Circles and Clava cairns mus admittede b t exace th t t linbu , k remain mysterysa e vien I th . f wo consistency of the orientation of Clava cairns between WSW. and SSW. one is entitle o suppost d e thae (assumedth t ) similar consistenc e Circlee th b f o o yt s orientated betwee meay ma n n. SSWS som d ean . differenc ritualn i e Henshals A . l notes, only the ring-cairn element of the Clava cairns is found on the other side of the Grampians, in Kincardine and Aberdeenshire. Henshall notes the impossibility of separating on distributional grounds the ring-cairns and passage-graves: the only hin differencf o t tha s formeei e th t r ten occuo dt r ove widea r r area tha lattere nth , and this suggests that the former may have lasted longer or eventually become more popular, which might explain the lack of passage-graves on the farther side of the Grampians. Henshall has already noted features among the Clava cairns in Strath Spey which suggests they represen latea t r movement from Strathe Nairnon d an , 1 Piggott, S., Neolithic Cultures, 116-17, 170, 182, 271-2. 2 Arch. N.L., 6, i (1955), 14; see now R. J. C. Atkinson, 'Fishermen and Farmers', pp. 1-38 of The Prehistoric Peoples of Scotland (1962) (ed. S. Piggott); 14-17. 3 P-P-S., Xxvi (1960), 135. 4 op. cit., xxvii (1961), 174-233. 5 Antiquity, xxxiv (1960), iu-i6. 8 Arch. N.L., 6, I (1955), 14; see now R. J. C. Atkinson op. cit. in S. Piggott op. cit., 18-19, 22. 7 Ulster J.A., 3 ser. 15 (1952), 31-60 and unpublished map in National Museum of Antiquities of Scotland; . AtkinsoC . J se. cit.. eR nop , 19. 100 PROCEEDINGS OF THE SOCIETY, 1963-64 could mention her possible eth e Clava cair t Lowena r Easo r t Lagmor cone th t -ea fluence of the Spey and Abhainn where the cairn material has spread beyond the sur- viving orthostats, suggestin e latteth g r mus e verb t y neae kerbth r . These late features are very marked among the Kincardine and Aberdeenshire ring-cairns. Thi itseln si f raise majosa r problem Recumbene sizth e f th eo n o : t Stone Circles alone impossiblis it , suggeseto t they develope thesof ed out lat e ring-cairns. Hen- shall suggests tha Circle th t e builders opposes (thaa ring-caire , is tth o dt n builders) if they penetrate , werd W. fro ee mosmth t likel havo yt e come alon innee gth r edge of the coastal plain, and then by Keith to the middle Deveron valley, where they settled, and on by its tributary, the Bogie, to the Urie and Don. From the distri- butio Circlee th f no s this seem moste sth likely rout settlerr efo s coming fro aree mth a Clave ofth a cairns t therbetweep bu , ga indees ei g nearese bi nth d a t Clava cairns coastae th n Deverone o l th plai d wils n an note e A .lb d more fully later, Clava cairns very markedly avoid the coastal plain, except in the immediate area of the head of the Great Glen, and there are no examples of any sort farther along the coastal plain, let alone anything that might be regarded as typologically linking Clava cairns and e Circles.th t seemI cluey thay an s ssa f safi t exiso t e shoo t t linke wth s between 1 these two types of monument, they will be found in the NE. where new field work and study are urgently needed, and not in the Clava cairns area. A study of large scale distribution maps to find out exactly what areas are occupied by these circle- builders, and how they compare with the distribution of Beakers and also of early metal types must be a first preliminary. The differing opinions of orientation must be resolved, and so must the problem of what exactly was the primary interment, for cremation is totally foreign to the Beaker tradition. Henshall suggests that the ring-cairn builders reached the NE. by going up Strath Abhainn, over the notorious Lecht, and into the Don valley. The Lecht, however, is the second highest motoring road in Scotland, and generally the first in Britai blockee b o n t snow y db . Wit exceptioe hth sitee th t Sand sa f no Forvief so , all these ring-cairns are in or accessible from the lower reaches of the Dee valley: by continuin Stratp gu h Abhainn instea f turnine Conglaso d th p gu s e Wateth o t r n traversLechtca e e narroth on e , d roadleswan s defil f Gleo e n Buig, only just touching 1600 ft. and reach the Gairn, a tributary of the Dee, but it is tempting to suggest that the route used was that through Glen Feshie, up the river of that name which join Spee sth t Kincraigya dowd Geldie an , nth e Dee e Burth .o n t e Thith s si most direct route from the Spey to the Dee and nowhere exceeds 1850 ft. Periodic attempt havto s a roae d constructed have been made ever since General Wade projected suc roaha linbarracko e dt th k t Rathvena s , Strath Spey, with that a t Braemar.2 vere th y n thorougO h geographicae studth f yo l backgroun Clave th o dat cairns

. Cole1F s (P.S.A.S., XL (1905-6), 198-201, 205-6) classifie stone th d e circl t Innesmillea , Urquhart, Moray, as a probable Recumbent Stone Circle, and Childe (op. cit. 104, fig. 23) marked it thus on his map showing the distribution of the Clava Cairns and Recumbent Stone Circles. However, this ascription is erroneous — Coles over-enthusiastically misquotes his references. Scotse Th 2 Magazine, (Sept6 , 75 . 1961), 537-4 Salmondd 4an , WadeB. Scotlandn i . ,J (1934), 175-84. esp. map on p. 178. I IO E CLAVTH A CAIRNS settlement which Henshall gives, only a few points need be made. Firstly, the almost total lack of settlement by Clava cairn builders on the fertile Laich of Moray must be emphasised, and this will be discussed when offering some dating evidence for the Clava cairns. Secondly, there is a reference to what may have been a Clava passage-grav than ei t compendiu referencesf mo , Chalmers' Caledonia.^ situatios It n above the Farigaig where it joins Loch Ness, would fit in with a movement coming up the Great Glen and finding its way over to Strath Nairn, for the Farigaig valley leads directl uppee th o yt r reache Nairne th f o s. Lastly, ther vera s ei y interesting reference to the presence of alluvial gold in Strath Nairn: in 1869 a native of the

district returned from the Kildonan gold fields in Sutherlan2 d and found gold in the Daltulich Burn near its junction with the Nairn, enough it was said 'to make a finger-ring'. Attempts to locate this burn from various sources have failed, but neighbourhooe th t n musi i e b t f Maindo f Daltulico s h 737418)farH m(N . This is abou mileo tw t s upstream fro Balnuarae mth Clavf no a grou cairnsf po nead e ,an rth f thao td stretcNairen e . th N f nho which seegreatese th s t concentratio f thesno e cairns. The discovery was later confirmed, but the quantities were too small to make exploitation practical. Its occurrence has probably nothing to do with the Clava cairns satisfactorf i t bu , y way identifyinf so g various gold lode founde sar t wouli , d be worth bearing this area in mind. Piggott has already3 fully discussed the Iberian relationships of the Boyne tombs, and to this one should add the comments of Hartnett on his excavations at Four- knocks I.4 There are numerous related features both in the construction of the tombs t motifar n i s d alon an Atlantie gth c coas fullese Iberiaf o tth t t bu ,parallel s occun ri the Almeria, Badajoz and Algarve regions of southern Iberia, where Blance's recent work5 has added greatly to the detail and understanding of this area. On the other hand, the spiral motif, so characteristic of Boyne art, is totally absent from the S. Iberian objects, and is rare both there and elsewhere in Iberia in rock carvings and paintings. The possible origin of the spiral will be noted later, but its rarity in Iberia suggest s- unles arguesone developes tha was Boyn it tthe dein culture itself- that there is more than one tradition present in the Boyne culture. The poppy-headed pins of the Boyne culture, though markedly bigger, have good parallels at Vila Nova de Sao Pedro and pestle-hammer beads occur in the Peninsula,e th f o t . aparbu SW t from thes stonf paralleld o ean e e us basin o t s r sfo

holdin6 g remains in some Iberian tombs, the Iberian affinities of the Boyne tombs so strongly suggested by art and architecture are totally absent from any comparison of material culture. This phenomeno featura s ni virtuallf o e whole yth e tradition of chambered tomb settlemen Europen i t : apart from tantalising glimpsed od d san

1 Chalmers , Caledonia,G. , I (1807) , 94-95, n.(j) reference th : e given, Phil. Trans. of Edin., , n14-15ii , , cannot be satisfactorily identified. Bain, G., History of Nairnshire (ist ed. 1893), 576 (and ed. 1928), 460; quoted in Matheson, C., Moray 2

and Nairn (1915), 109. Piggott3 , S., Neolithic Cultures, chap. VII. 4 P.R.I.A. (C) 57 (1955-6). 220-8, 239-40, 249-62. Blance, B, M., op. cit.: see Univ. Edin., Depl. of Arch. 12th Ann. Rep. (1959-60), 6-10 and Antiquity, xxxv (1961)5 , 192-202. 'J.R.S.A.I., LXXVII (1946). 65 , 102 PROCEEDING SOCIETYE TH F O S , 1963-64 traits there is no material to suggest great folk-movements such as might have been expected. Child lass hi t bookn ei 1 suggeste concepe dth 'megalithif o t c missionaries' to explain this sprea chamberef do d tomb western si northerd nan n iden a , a whic hI believ contaio et nconsiderabla e amoun truthf o t . However, ther littls ei e evidence that the newcomers ever came in numbers large enough to settle by force seafarerss eve thed A nha . y so wishe ,o f d thoso o . dt N e thad an t . settleW e dth Britai havy nma e been conten seae ofd th fan . liv o t y t eb Blance's examination of the Almeria culture of SE. Spain in particular, throws some interesting light on the problem of chambered tombs. Probably c. 2700-2600 B.C., there impinges on the indigenous population of southern Iberia a settlement from the eastern Mediterranean bringing with it among other things the use of ring-cairn like structures, eithe communas ra l ossuarie collectivr o s e tombs formin Almerie gth a culture. About this time, copper was introduced to Iberia by the first onumbea f settlementf o r s beginnin e 'Colonialgth ' period e earliesTh . f theso t e settlements, all of them heavily defended, was probably Mesas de Asta on the Gulf of,Cadiz othee th ; r sites, from Almizaraqu eastern ei n Almeri Vilo at a Nov Sae ad o Pedro upriver from Lisbon art; a little later. The founding of Los Millares, Almiz- araque and other sites in Almeria ended the development of the Almeria culture and introduced the passage-grave. It is to this period that the objects carrying Boyne-like art motifs belong. The arrival of these settlers in Almeria possibly caused an expan- sion W. of the Almeria culture, which seems to have played an important part in the formation of the Portuguese dolmen group in S. Portugal and SW. Spain, where the fla tAlmeri e idolth f so a cult developeue r d int decoratee oth d schist plaques whose triangula d lozengan r e liko motifs e e thosar s Boynn ei e art, though her face eth e motif is exceedingly rare. However, this last motif is common on clay plaques at Vila Nova de Sao Pedro. An examination of the Los Millares passage-graves suggested that certain of them, particularly of the earlier phase, were never roofed by corbelling. Llano de Atalaya l , alreada6 y onl e passage-graveo noteth ytw f beins o d a e gon s planney db the Leisners orientated SW., is such an example, being in fact an Almeria culture grave with the passage adopted from the newcomers. alreads Asha y been noted Mesae th , Aste sd a colony probably dates fro 2700. mc - 2600 B.C., and the other colonial settlements must date to before 2400 B.C. A C 14 date of 2345 ± 85 B.C. was obtained from an upper level at Los Millares, and all the site pre-Beakere ar s , there being evidence suggeso t t that Vil aPedro NovSa oe ad was actually destroyed by the Beaker folk. On the Dutch C 14 evidence,2 Bell Beaker sNetherlande werth n ei 220y sb 0 B.C. That Vil aPedro Novs Sa e owa ad founded before 2400 B.C. is suggested by the fact that while it has several links with the Cyclades the typical Cycladic spiral pattern is absent, and this implies a date before the beginning of the Early Cycladic II period, usually dated to c. 2400 B.C. Recen 1C t4 Iris e dateth hr sNeolithifo c have suggeste dfloruia t o2200-160. fc 0

Prehistorye Th Childe , G. . ofV , European Society (1958), 124-34. 81 Palaeohistoria, iv (1955), 3-27 and Antiquity, xxxiv (1960), 17, correcting the Ci4 dates, op. cit., xxxn (1958), 261. THE CLAVA CAIRNS 103 B.C. for the Boyne culture. Bearing in mind therefore, the dating suggested by the parallels betwee poppy-headee nth d Boyne pinthosd san e from Vila Nov Sae ad o Pedro, and the occurrence of pestle-hammer beads and cup-marks in the £vora region t wouli , temptine db regaro gt Boyne dth e cultur originatins ea aree th a n gi of Portugal Ee settle.th Mediterraneay db n colonists e manneTh . whicn i r e hth spiral becam integran ea Boynle parth t traditiof to ear t clearno s ni : Blance suggests that each of the Iberian colonies had a different source in the eastern Mediterranean; this together wit previoue hth s eastern Mediterranean settlement which resulten di Almerie th a culture, suggests tha firse tth thir te halth df fo millennium B.C. witnessed seriea independensof t voyages fro Aegeamthe Iberianto woul:it quitdbe e reason- abl supposo et e tha latea t r such movement fro Cycladee mth beginnin e th t sa f go Early Cycladic II brought with it the spiral pattern. The rarity of this motif in southern Iberia suggests that these voyagers, probably headin thin gi s case directly Tague foth r s region, move immediateln do y with procesthos e wero th en wh ei f so moving up the Atlantic seaboard to Brittany, Ireland and ultimately Orkney. Why this migration took place will very probably never be known: if the spiral motif cannot arrive until after 2400 B.C havy . thema e t narrivei time Vile th eth t ada Nova de Sao Pedro settlement was becoming beset by Beaker folk. This would confirm MacWhite's suggestion1 that 'the art of the Irish Passage Graves received some fresh stimulus fro e Mediterraneamth n which partly by-passe e Iberiath s n peninsula'. Typically the Boyne culture shows very little connection in its material remains with the area of its assumed origin. The settlement, whether by missionaries or refugees from Vil aPedroo NovSa e ,d a must have been numerically very small, and an impoverished material culture is only to be expected. Loughcrew ware, which is the recurrent pottery type in the Boyne tombs, has already been fully covere Piggott,y db recentld 2an y Case furthes 3ha r commente pooa s di rupot I . nit quality ware, the pots semi-globular in shape, with profuse overall decoration, both of these features being completely foreign to the Western Neolithic pottery tradition in Ireland. Case indeed suggests that it was the work of natives who had no previous knowledge of pottery. A comparison of the distribution of the Western Neolithic pottery and Loughcrew ware4 shows them to be almost mutually exclusive, and the idea of a handful of settlers arriving among a people still in economy seem versa y probable explanatio origie th f Loughcref no o w ware. Certain change religioun si s tradition undoubtedly took place exampler ;fo , none of the decorated round idols are known from Ireland. On the other hand, we have the transference of the decoration to stones in the tombs themselves, even, as for example at New Grange, to the extent of erecting non-structural stones in front of the dry-stone walling to carry the decoration. These stones may have originated in featura e note examplr Millares dfo Lo t linina ea s- stonf go schisr eo t slabs hiding lowee th r chambere wallth f so . J.R.S.A.I., LXXVII (1946), 65.

1 2 Piggott, S., op. cit., 202-4, PI. VI. 3 P.P.S., XXVH (1961), 185-^6, 206, 210-13, 217 under the name of Carrowkeel ware. 4 ibid. . 215 cf , 216d , an fig , 7 fig. 2 .28 104 PROCEEDINGS OF THE SOCIETY, 1963-64 The one surprising feature of the Boyne culture is its apparent total lack of con- tributio introductioe th o nt metallurgyf no , despit undoubtee eth d fact that copper metallurgy thad closen an ,i t d moulds beed ha ,n introduce ColonIberie o dt th y ab - ists. Indeed, Blance believes; that the Colonies were in fact founded by metal prospectors Irelandn I . , there seem havo st e bee thougho nn metaf o t l prospecting in the mind of the new settlers, who in this particular instance may have arrived as refugees rather than missionariee mosTh t. probablse r spe e explanatioe th r nfo metalless economy of the Boyne culture is again that so few settlers arrived that prospecting for ore, which would have necessitated journeys into southern Ireland, was impossible. However face th ,t that these some settlerus ed metasdi suggestes i l face th t y db that the decorated flat axes of the Irish Early Bronze Age included passage-graves motifs. e non-appearancTh 1 thesf eo e axe Iberian si decisiva 2 e seemm o et s point in rejecting the idea that contacts with the Iberian metal-workers still existed by Early Bronze Age times in Ireland. Alevidencee th l , therefore:, point numericalla o st y very small settlemene th n i t . MeatCo h are f eastero a n Ireland, where tomb plannin d constructiogan d nan artistic manifestations are at their best. Why the 'megalithic religion' of the parti- cular sect spread so far and became so popular, inspiring such expenditure of effort is something which archaeology perhaps can never explain, but is a feature of the entire megalithic traditio musd nan t represent some important psychological force. thin I s particular case reasonabls i t ,i supposo et e tha introductioe th t Neolithia f no c economy, building techniques arrivae th , sea-goinn i l g craft fro unknown ma n land and no doubt the use of a few copper all combined to mightily impress the Mesolithic people among whom they settled - the occurrence of cemeteries of tombs suggests quit econsiderabla e native population. Because of the total lack of the material culture of the Clava cairn builders tomb types are practically the sole positive feature which can be used in discussing the cairns in the light of Irish and Iberian evidence: if we accept the validity of the parallels drawn between Clava ring-cairns and those of the Almeria culture, and between Clava passage-grave e Colonistsd thosth an sf o e , the e e origith nth f no Clava settlement Almerian musi e li t type,o whertw se exiseth t together. ring-cairne Th s indicat timea t loneno g afte arrivae Colonistse th rth f o l , when the Almeria culture still existed, though the evidence particularly among the earlier Los Millares passage-graves that roofs wer t corbelleeno d suggeste th t a timsa t eno very beginning of the Colonial phase - on the other hand, the lack of side-chambers among Clava passage-graves mean fairla s y early dat e Almeriatermn ei th f o s n examples. The cup marks on present evidence suggest some contact with S. Portugal. face Th t tha Clave th t a cairn builders went directlGreae heae th th f do to y t Gle n seems extraordinary t presumablbu , y they avoide Iris e area dth hSe a becauss wa t ei already relatively well settled, which suggest movemene sth t took place later thae nth beginning of the Boyne culture. The total lack of Boyne tomb art among the Clava

cit.. 1op , iv (1938), 294. 2 loc. cit. THE CLAVA CAIRNS 105 cairns shows there can be no direct connection between Boyne and Clava cairns.1 geographicae Th l evidence fro Clave mth a cairn area itself seem conflico st t with this evidence, however previousls A . seee b nn froy ca distributio e noteds mth a d an , n map, the Clava cairns, except in the immediate area of Inverness at the head of the Great Glen, totally avoid the fertile Laigh of Moray. The fertility of the soil and geniality of the climate of the Laigh have been proverbial for centuries, and it seems extraordinary that a Neolithic community should prefer to settle in less favourable areas. It is all the more remarkable when one considers that the structure of the tombs, their occurrence in cemeteries and the social organisation implied by their being used privilegea onl r yfo d person must reflec well-organisea t d community. There seems only one possible reason for this: the Laigh was already relatively densely populate time Clave th eth y db a cairn builders arrived. I am, however, less confident than Henshall when she speaks of the 'considerable Neolithic population' on the coastal plain - I certainly cannot visualise a population dense enoug preveno ht organisen ta d settlemen aree newcomersth y ab f to . Wooded conditions would certainly exist in the Laigh, but the climate some 4000 years ago would have been markedly drier and milder than now.2 The first evidence for a major occupatio Laige th f nho comes wit Beakee hth r settlement thid followes san ,i d a marke y b d concentratio f earlno y bronze type d mouldsan s , whic s limitehi d rigidly to the coastal plain. assumee on f I a Dutcs h Beakee origith r nfo r settlemen . ScotlandNE f o te th 3 Dutch C 14 evidence suggests this must have occurred c. 2000 B.C. in which case one must assume Beaker settlemen Laige th 190y n hb o t 0 B.C , thereforeIf . arguee on , s that the Clava cairn builders were kept off the Laigh by the Beaker settlements, a date of 1900-1800 B.C. would be the earliest possible for the former. This, however, seems too low to make an Almerian origin plausible even if, as we have hinted above data , e later tha e star nth f Boyn o t e culture migh inferrede b t e outOn -. standing problem is that if the Recumbent Stone Circles are indeed Neolithic in origin, then the Clava cairns must be earlier still, for one can hardly derive them from the former. The lack of early bronzes in the Clava cairn area (cf. Maps I and II) is important: the restriction of bronzes to the Laigh continues throughout the entire Bronz quits i e t i Agee d certaian , n thaClave th t a cairn builder longeo sn r existed by the beginning of the Early Bronze Age in the Laigh which, I would suggest,

t proposeno s i t 1arguo I d t e abou disturbine th t g €14 dat 328f eo 0 (Grn-igSz 5 +7 lir efo ) Cam. Giot emphasises the undisturbed nature of the tomb and the association of the wood from which the dating sample was taken and the pottery (Bull. Soc. Preh. Franfaise, LVI (1959), 292-3) and Case (P.P.S., xxvn (1960), 220) derive Iriss shi h Western Neolithic pottery from Brittany. This beint seemi , gso s incredible that corbelled passage-graves should not have appeared in Ireland until so much later. The lie Carn date is far earlier than Blance's dates for the Iberian passage-graves, but the earlier part of her chronology was ultimately based on that of Troy I, and recent work by Mellaart (Anatolian Studies, ix (1959), 131-62) has inspired his radical reassessment of Anatolian and Eastern European chronology (Antiquity, xxxiv (1960), 270-8) which suggests highefar a r chronology than hitherto generally believed (fo criticisa r Mellaart'mof s chronolog Antiquity,ysee xxxv (1961), 276-80, also op. cit., xxxiv (1960), 293. gratefum a I 2 Messro t l. Durn. Gran Macaulae E R . th d sS f to an y Institut Soir efo l Research, Aberdeen for informatio thin no s point. e mos3Th t recent argumen mads r thi fo twa sProfessoy eb r Piggot s lecturhi n i te 'Abercrombd yan After', C.B.A. 'Problems of the Bronze Age' Conference, London, December 1960; see now S. Piggott, 'Traders d Metal-workers'an . 73-10 Prehistorice ,pp Th f 3o Peoples of Scotland (1962) (ed . Piggott)S . . 6 I0 PROCEEDING SOCIETYE TH F O S , 1963-64 starte . 165dc 0 B.C. This effectively dispose suggestionse th f so 1 that these cairne sar connected with the introduction of metallurgy. Here one would like to know what length of time might be in order for the initial settlement, subsequent expansion and final demise of the Clava cairn builders. It may also be noted that all pottery types are likewise absent from the Clava cairn area. A possible reason for the lack of metal Boyne th n i e cultur alreads eha y bee nClave suggestedcase th f th e ao n i cair t nbu , builders, assuming an Iberian origin, the explanation cannot be so simple, for copper deposits exist at Bona on the W. side of Loch Ness.2 Among the in the Food Vessel cemetery near Burgie, Moray, one contained 3 a cup-marked slab, and another one may have, while an urnless cist had quartz pebbles on the floor. Two other urnless cists were of proportions that might be called megalithic. A Food Vessel cist: also of megalithic size, at Loch of Blairs, Altyre, near Forres, Moray4 contained pebbles 'partially covered with a white limey deposit' which may be in the same tradition as the quartz. This would suggest that Clava cairns existed prior to the Food Vessel period. A white pebble was found with one of the Beakers from Achindown, Nairnshire.5 t DruidtempleA discovere s cisa , wa t caire d th 1 r 1f 4n5o o whicyds. .NW h con- taine d'smala l vas urnr eo ' reddis outside th bluish-blacd n ho e an k inside 'plaia f ,o n barrel-shaped pattern', 4 or 5 in. in height, with mouth, middle and basal diameters Si) Si and 2f in. respectively.6 If it were a Beaker, it might suggest that the cairn was built subsequently to this burial, as one might have expected the latter to have been placed Claveithee th n ari cair r welno l awa e complementary Th fro . mit y distri- butio Clavf n o Orkney-Cromartd aan y cairn hardln sca interpretee yb t presentda : the suggestion that the Clava cairn builders arrived up the Great Glen first because they chos southere eth nMorae sidth f eo y Firth nee t followdno ther fo ,y kept ofe fth fertile Laigh whic firse hth t settlers might have been suppose havo dt e settledp Cu . marks, admittedly unusually large, on a stone at the Orkney-Cromarty cairn at Conti noccurrence Mainth d san quartf entrance o th t za elsewherd ean t Kilcoea y West7 suggest that these particular tombs Blace , botth n kho Isle, were constructed after the arrival of the Clava cairn builders. 1 First suggeste Megay db Hardyd wan , P.P.S., iv (1938), 292. Piggott, S., Neolithic Cultures, 261, among others, has also suggested this. " P.P.S., , figxvi70 . i. i .(1951)p n o , sit 3 69 e,4 8 P.S.A.S., L (1915-16), 201-40. * op. cit., LXVI (1931-2), 404-5. T. Inverness Scientific Soc. and Field Club, I (1875-80), 188. 6 P.S.A.S., xvin (1883-4), 355-6 an . InvernessdT Scientific Soc. and Field Club, n (1880-3), 370. 7 Unpublished Discoverye se ; Excavation,d an Scotland 1956, 23—24.