IV.

RUDH DUNAIN A ' N , SKYE WY B . .LINDSA Y SCOTT, D.S.C., B.A., F.S.A.ScoT.

SITUATION OF CAIRN. As the situation of this cairn is somewhat surprising and may thro we distributio th ligh n o t f populatioo n n Neolithii n c timet i s seems desirable to deal with it before proceeding to a description of e cairth n itselfe promontorTh . f Rudhyo n Dunaia ' e wesnth lietn o s coas e ape a trianglf Sky th o f t o x t d moorlandea f uneve o an e g bo n , some eight square mile n extenti s , broke many b n y small cragse Th . base of this triangle is formed by the precipitous line of the Cuillin Hills, its south side by the Sound of Soay, and its north-west side by a loc f se Brittleho e th . Just withi e brackise apeth th ns i x h lochan, Loc a h'Airden h , which fallt inte Sounou sth o f Soao d y througa h channel 100 yards in length which fills from the sea at equinoctial springs (fig. 1). A neck of land, 200 yards across, divides the head of the lochan from the small bay, Camas a'Mhurain, on the Loch Brittle promontorye th sid f o e . Apart fro ma nai'ro w strip alon e shorth g e nea e hea th rf Loc o d h Brittle e onlth , y par f thio t s area capablf o e cultivation is a shallow valley running wesWsouth-west for something less than half a mile to Loch na h'Airde. Owine formidablth o gt e e barrieCuillinsth f o r , beyond whice li h Coruisk and the precipitous shores of the sea-loch of Scavaig, the only land acces o thit s s t areinvolvinno a g serious climbins i g throug s northerhit n corner from Glen Brittle. This lond narrogan w 184 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932.

glen, with high hill botn so s sides hit s itseli , f eas f approacyo h only b ya singl e pass fro s heamit d ove Drynoco t r e hea th f Loc do t ha h Harport. No s accesi r o Budht s n Dunaia ' n easie y sealanding-placb A r . e 3 miles to the east of the point is used in fair weather by the fisher- men of Soay, but it offers no shelter and is impracticable in any sea. The head of Loch Brittle is shoal and sandy and attractive enough for flat-bottomed boats in off-shore winds, but it is unapproachable in

\fc \ RuuH n Duta ' i n ai 0*t .A ft f1 §i« V Caro*aa'MMir*;J\ > r»rl"

Rudhf o p Dunainn a 'Fig Ma . .1 , Skye. (Scal in.=110e1 0 ft.)

winds between sout d westan h . Camas a'Mhurain e smalth y , ba l already mentioned north-eastwar f Rudhdo n Dunaina ' s rockyi , t bu , can be used by small boats in calm weather. The outfall of Loch na h'Airde, though its sea approach is encumbered with rocks, is sheltered e south-wesonth $plee f th Sgei o ty b t r Mhor, whic w s joinei hlo t a d water to the shore. It has, in fact, been used at some past period as a boat port and boulders have been moved aside to make a runway, but t wouli dangeroue db approaco st swelly e highean Th n .hi r levef o l the land in the Hebrides in late times, as evidenced, for example, by the position of the chambered cairn of Geirisclett, in RUDH AN DUNAIN CHAMBERED CAIRN, SKYE. 185 1 North Uist, would have improved the shelter provided by Sgeir Mhor,

n windi s unusabld n whici swa an t i h e resor1 t might perhaps have been had to Camas a'Mhurain, but access to the promontory by sea can never have been satisfactory. The triangular area described is now wholly uninhabited; Glen Brittle has eleven inhabited houses, but their number is diminishing. In the eighteenth and earlier nineteenth centuries there was a larger population, and remains of houses and cultivation extend both up the a mil gled e an nalon e southerth g ne lochshorth f . o e Untie th l middle of last century Rudh' an Dunain was farmed separately from Glen Brittle by the family of Macaskill, and the ruined two-storied farmhouse e shallo stande heath th f do t w sa valley running dowo t n Loch na h'Airde. The grave of the last Macaskill, tacksman of Rudh' e churchyarDunainn a th n i s i , f Kilmoruydo , Loch Eynort. Therknowo n s ei n evidenc f ancieno e t habitatio n Glei n n Brittle. Apart froe cairth m n e subjecwhicth s f i hthio t s paper e onlth , y recorded evidence of ancient settlement in the triangle of moorland of which Rudh' an Dunain is the apex is the " galleried dun" of the same name.2 This massively built structure defends the neck of a very small triangular promontory e verticath , l sea cliff f whico s h command outfale th f Loco l e boaa h'Airdn hth td porean t described aboves A . s limiteit f o dy class an s bee r neitheha , nno , excavatedit r s datit , e musmattea e b t f conjectureo r t primathero n bu , s ei facie reasor nfo datin t otherwisgi ee Earl thath o ynt witIroe broche hnAg th s which closelo s t i y resembles.3 Attention has been drawn to the situation of Rudh' an Dunain cairn because it appears remarkable that a chambered cairn should be found in this exceptionally remot inhospitabld ean e cairne spot Th s wil.a , l be shown below, was an elaborate structure of its class and was used in Neolithi n Beakei d an cr times. Ther a smal s i e l cair t Kraiknisna h on Loch Eynort, 3 miles to the north of Glen Brittle, which was excavated by the writer, and found to cover a pentagonal cist containing beakero a flintw d t an sbutto n scraper.4 This cist might possible yb 1 The evidence has been collected by Dr J. G. Callander : Proc. Soc. Ant. Scot., vol. Ixiii. p. 319. 2 Roy. Com. Hist. Mon. (Scot.), Outer Hebrides, Skye, d Smallan Isles, 483. No . 3 For a discussion on " galleried duns" see A. O. Curie, Antiquity, vol. i. p. 296, and Roy. Com. Hist. Mon. (Scot.), Outer Hebrides, Skye, and Small Isles, pp. xxxv and xxxvi. It is tentatively suggeste e Commissioth y db n tha simplee th t r for heagalleriestanf e my o th anf dma t o d . a n ddu evolutionary series of increasing complexity of which the broch is the culmination. So far as it can be investigated without excavation, however, Rudh' an Dunain appears no less complex than a broch, althoug , fro naturis e msitet s hth i it f ,e o les s extensive distinctive Th . e feature brocf so h structure are identically reproduced, and the manner and quality of the workmanship present no features which would differentiate it from the finest Hebridean brochs. Professor V. G. Childe comment thin so s subjec Proc.n i t Roy. Soc. Edin., not, . vol77 e par45 . . 1 p . . i t • Report in Man, October 1929. 186 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. regarded as transitional in type between the chambered cairn and the Bronze Age short cist. These two cairns provide the only evidence of habitation before the Iron Age in the mountainous area lying between e relativelyth fertile land f Strathairf o Locso e easd hth an t o t d Bracadal northe generae th Th o et . l distributio f chambereno d cairns e Hebrideinth s does not, however, suggest that they were ordinarily place t greada t distances fro dwellinge m th wor e livinge th th k d f so an , of constructing them must have involved a substantial number of men. It seems necessary to suppose therefore that this region^ which has never supported more than a sparse population in modern times, was moderately thickly inhabited in the Neolithic and Beaker periods. This s consisteni t wite vieth h w commonl e ydeterioratio th hel f o d f o n climate and with the evidence for the growth of peat since that time;! free f ito ds coveringrounw lo e f peatd go th e soi immediatelf th ,o l y to the west and south-west of the Cuillins would support a far greater numbe f domestio r c animal doessw thano . t ni

SITE OF CAIRN. e cairnTh , t showwhice Ordnancno th s n i hni e Survey, standt a s the highest point of the neck of land between Loch na h'Airde and Camas a'Mhurain latn i , . 57 °longd 9' an 58 . ." N 6° 18'45" heigha W. t a , f o t some 30 feet above the sea. The site is approximately level, and is now covered with rough grass and heather over the solid basalt. Although higher land lie spoine betweeth f Rudhd o t an Dunaihn t a 'ni viee th , w fro caire mth wid s ni included ean e Cuillin north-easte sth th o t s e th , coas north-west e f o Skyt th o t e islando m et th Bu d f Eigo d san , gan seaward. It is customary to remark on the view from cairns, though I confes I evidencsknoo n f wo e tha deae th t d have ever been regarded, or are anywhere now regarded, as being gratified by an extensive prospect.

CONDITION BEFORE EXCAVATION. The cairn stand o at s heighfee1 1 tf o abovt e solith e d rocn o k which the tomb is built. The northern side is disturbed, but is probably t substantiallno y denuded; fro e highesmth t point, whic s northi f ho chambere th f o centre caire d th th t slopei ,n f an e o s awa shalloa t ya w angle and the south side has obviously been largely reduced. A wall joining Loch na h'Airde and Camas a'Mhurain and skirting the cairn

1 E.g. in the report of the excavation of CaUanish, Proc. Soc. Ant. Scot., vol. iii. RUDH DUNAIN A ' N CHAMBERED CAIRN7 , 18 SKYE. is presumably built from cairn material; one substantial slab is incor- porated in it and two others lie beside its base. Except on the steepest parts of the northern face there is a thick growth of turf and heather, and that this has long been so is evidenced by the infiltration of soil into the roofed antechamber and vestibule. Before excavation seven slabs of the peristalith appeared to heights varying up to 2 feet through the material of the cairn, but there was no visible indication of any othe e structurer th par f o e boundart Th e cair. th s indefinitef i no y , but a circle 78 feet in diameter approximates to the apparent present periphery superficiao N . l evidence appear encirclinn a f so g ditch. EXTEN EXCAVATIONSP TO . With the kind consent of Macleod of Macleod, excavations were wify d myselm an e carriey n Septembeb i f t dou r e 1931chambeTh . r was found to the south-east of the centre of the somewhat dubious circle mentioned in the last paragraph ; its roof was fallen in and it was excavate p downwarddto fro e e solimth th do t srock . Opening chambee antechamberth n a f o s t rwa ou , perfecroofen i d dan t condition; this was excavated from the chamber. Opening out of this was a vestibule,1 also roofed and in perfect condition; this was excavated from the antechamber. The vestibule communicated through a portal, of which the lintel had slipped slightly forward and one jamb had tilted, with a forecourt defined by a crescentic fa

Photograph by Mr W. L. Coats. Fig. 2. Eastern part of Chamber showing Antechamber partly excavated.

The roof of the chamber has been formed of large slabs of a fine- grained basalt, which splits alon ga remarkabl y plane surfacs i d an e possibly a variet f basalo y t found elsewher n Skyi ed name an e d mugearite. Therevidenco n s ei f corbellingo e largA . e slab lies over panee stretched th 3/P. an d P l, .4. an 4 P sd inte bodan e th o 3 th f . yo P cairn for a distance of at least 3 feet; its outer edge does not oversail the west wall of the chamber, however, and it was probably designed merel spreao yt e weighe roof dth moderate-sizeth A f . o t d slab over- e paneth 2/P . d P l probabl.3 an lyin 3 . gP y serve same dth e purpose. The antechambe s approachei r d froe chambemth a doo f y o b rr 190 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. which the jambs are pillars roughly square in section set diagonally dooe Th r . leadin3) d (figsgan fro2 . antechambee mth vestibule th o rt e s similaha r jambs alst diagonalle y finse o dr Th e . 6) yd (figsan 5 .

Fig. 3. Doorway into Antechamber from Chamber. stone walling of the antechamber will be seen from figs. 4 and 5. e antechambe th rooe f o Th f vestibuld an r s composeei f heavdo y slabs stretching into the mound on both sides. The antechamber roof slab overlapslabo tw s e forminth s e vestibule th e roo f th go f ; theso tw e e samth t ea leve lasjoie d ar tan nl neatly. Attentio e drawb y n ma n e descendinth o t g height f chambero s , antechamber vestibuled an , . RUDH' AN DUNAIN CHAMBERED CAIRN, SKYE. 191 The vestibule leads directly into the forecourt through a portal now partially ruined. The portal is most clearly seen in fig. 9, a photograph taken whe e portanth l arefulls awa y cleared e heavth ; y linte s seei l n

Fig. 4. South Wall of Antechamber. fallen forwar d downwaran d d froe eastwarth m d rooe f th sla f o b vestibule. A large block, 3 feet 3 inches long, is shown in fig. 8 in the centre lying below the lintel in the position in which it was found; it was necessar removo yt t efroi m this positio o examint n e portath e l area s originaIt . l purpos obscures ei possibls i t i ; e tha t formei t d part 192 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. osuperstructura f lintee th o elt founs suci Mediters e somha n di th f eo - ranean tombs with which compariso s madi n e below, notable th y Giants' Tombs of Sardinia.

Fig . Nort.5 h Wall Antechambef so Vestibuled ran . Between chambe antechamberd an r , antechambe vestibuled ran d an , vestibule and forecourt are septal slabs or blocks standing on the rock floovaryind an r g betwee foo1 inche9 n inche3 d t an sn height i s . There is a similar block in the vestibule just east of the septal block dividing this fro antechambere mth roce t resti th ; s kn it floorso d an , most probable purpos o formt s i e, wit e vestibule-antechambeth h r RUDH DUNAIN A 1 N CHAMBERED CAIRN3 , SKYE19 . septum groova , holo et antechambere ddooa th o t r . Grooves, rebates, r holeba sd havan e been e rock-oufounth n i d t tomb d navetassan f o

Fig. 6. Antechamber, Vestibule, and Portal from Chamber, looking east. Balearie th c Islands, where they certainly serve holo dt placn di e stone or wooden door structures.1 The principle of the arch, which serves very well to secure the stability of a circular structure such as the chamber, is not available maintaio t e forecournth t facade thid s beean ha s, n pressed outwards W . HempJ . , Archceologia, vol. Ixxvi . 121—60 pp .Antiquariesd an ; Journal, vol. xii. .pp 125-351 . 3 1 VOL. LXVI. 194 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. e pressurb yth e cairn th f o .e Immediately e portasoutth a f s i o hl •panel of which the lower courses are in good preservation (fig. 7). Beyond this is a large orthostatic block, Q. 1; this has fallen forward, s heait s bee ha d n an d liebrokean s f beneathof n t .i (fig . It.8) s original height was probably abou fee6 t nort inches6 the t of h T.o . e portath e firsth l t pane s completeli l y collapsed; beyond n a tha s i t orthostatic slab, Q. 2, 6 feet in height, complete and tilted slightly

' tlsSHR''*"' ' ^' S^'C' ^ ^'T^llfc* ' fit" "::' ••*"<'

Fig Portal7. . , beginnin southergof n hal Facadof f Prostrateand e Slab, looking south-west.

forwar o succeedintw d d 10)e (figsan Th . 9 , . g4 . orthostatsQ d an 3 . Q , are 5 feet and 3 feet 6 inches in height; the first is vertical and the second tilts slightly forward (fig. 10). The intervening panels, Q. 2/Q. 3 and Q.3/Q. 4, are composed of large blocks of roughly rectangular section, but owing to the pressure of the cairn this walling is pressed outware uppeth d r dan course e fallear s n d 10)panee (figsan Th . 9 l. beyond Q.4 stands vertical and to what may be its original height fooo 1 fe nex inches9 tth tt succeedinbu , g orthostat, which stoot a d the extremity of the horn, is missing. The panel between this gap e nexth td followinan g orthostat , stand 5 . a heigh Q foo1 ,o t sf to t 1 inch in a single course; no stones are fallen, and if a second course ever existed it must have been stolen with the orthostat at the point RUDH DUNAIN A 1 N CHAMBERED CAIRN, SKYE. 195 of the horn. Q. 5, the last orthostat reached by the excavations, stands vertical and 3 feet 4 inches in height. Some 4 feet beyond, another orthostat, about 3 feet 2 inches in height, protrudes through the cairn and is apparently vertical. The forecourt facade thus appears as a structure crescentic in plan, wit e e e portaconcavitycentrth th th h f t a o el , curving s rounit t a d extremities to form rounded horns and continuous with the peristalith

Fig . Porta8 . l area looking west, partly cleared. which presumably surround e monumentth s n elevatioI . e facadnth e rise heighn si t fro hornportae centree e mth th th t o lowes t i l ,bu r than orthostate th norto t s, soutd althougit han f ho s possiblei t s i hi s a , suggested above, tha t originalli t y carrie dsuperstructura e which raised its height at least to that of these orthostats. The stability of the facade must have depende pressure th caire n th do f neo upo s outenit wels a r l notee b o s t dinne it s thati n t ro i ,sides d a whilan , e e beyonth 4 . dQ orthostats and walling showed no signs, to the limited extent of the excavations f beino , g pressed outward facade forecoure th th , f o es wa t tilted forward and in part fallen. As this could only have occurred when the forecourt ares cleareawa f cairdo n material t followi , s that such clearing had occurred at some period subsequent to the original 6 19 PROCEEDING E SOCIETYTH F O S , FEBRUAR , 1932Y8 .

•K;•• '~s,^c "'•" * f~-„* 8*'Mwfelj, *K- J^-SSw* -&ff - -, ^

Fig. 9. Portal and beginning of north half of Fa9ade, looking south-west.

Fig. Norther10 . n hal Forecourf o f t Fafade, looking north. RUDH1 AN DUNAIN CHAMBERED CAIRN, SKYE. 197 construction. Numerous blocks simila thoso t r f whiceo e panelhth f so facjade th e constructear e ground-leven o e dli r restinanothee o l on n go r at distances up to 3 feet from the facade (fig. 10), and they can hardly owe their presence here to the ravaging of the cairn for building material. They would, moreover, have provided particularly valuable building material, and would hardly have been left behind by people stealing stone f thesi thed yha n been exposed t occurreI . thae m to dt they might have belonged to one of those stone ramps external to the peristalith, the existence of which has been discovered by Mr W. J. Hemp.1 The block t Rudha s n Dunaina ' , identifiablhowevero n n i e li , e order, and nearly all flat on the ground. The only probable inference seems thae forecoure b th to t t ares deliberatelawa y cleare ground-leveo dt l for some purpose connectee tombth f o . e d us wit e hth Attention shoul prostratcallee e db th o dt e fee 4 slab , inche9 5 t . ,S n si length and 1 foot 3 inches in width, which lies across the forecourt, hut t quitno t righa e t angle s axisdistanca it t o a fee,t 5 s f t eo froe mth portal (fig. 7). This slab is not uniform in thickness, but it is wedged uthao ps flas tit t upper surfac horizontals i e . There seems little doubt, therefore, that its position is not accidental, and that it served some original purpose connected wit e tombhth .

CONTENTS OF THE TOMB. The upper part of the contents of the chamber consisted of cairn stones, earth broked an , n roo entirels f wa slabs d yan ,steril e dowa o nt height of 3 feet from the floor. From 3 feet to 1 foot above the floor wa a layes f browo r n earth mixed with fallen stones, which wile b l "e referre beaketh s a o rdt stratum. " From foo1 t above floor-level down solie th do t basal tlayea floos f blacro rwa k eartextremeln a f ho y slimy character containin gsmal w onlfe yla stones; this wil referree b l s a o t d the "neolithic stratum. dividine Th " g plane between thes strato tw e a was sharply marked. The same strata were found at the same levels in the antechamber and vestibule, the latter of which was filled to its roof and the former to within about a foot of its roof; the black earth was, however, less slim ychambere thath n ni . No y objecan f o t vestibulee founs sorth t wilwa i tn de convenien i b d l an , treao t te th t chamber and antechamber together. Plans of these showing the finds in the beaker and neolithic strata respectively are on PI. VIII.

1 E.g. at Belas Knap long barrow, Gloucestershire (W. J. Hemp, Trans. Bristol and Gloucestershire Arch. Soc., -vol. li. p. 268), and at Plas Newydd Chambered Tomb, Anglesey (report not yet published). Such a ramp supports the peristalith of a chambered tomb which I have recently been excavating in Anglesey. 8 19 PROCEEDING E SOCIETYTH F O S , FEBRUAR , 1932Y8 .

BEAKER STRATUM. The only pottery e beakefounth n i dr e beakestratuth s t a rm wa greatee Th . rfig 11 par. f thio t s vesse scatterey la l north-wese th n di t cornee chambeth f heigha o r t a rf abou o t foo1 t inches6 t above th e floor e remaininth ; g fragments wer t discovereeno e pary th f an o t n di area excavated. It has been restored at the British Museum. It is a large vessel, 7'8 inches in height, with a distinct foot, a round body, a somewhat high a waiststraigh d an , t slightly expanding e neckpastTh s i e. buff in colour, coarse, and mixed with a large quantit gritf yo somf so e dark grey stone. The waist and foot are plain; two similar bands of decoration cover the neck and the body. These bands are defined by three parallel lines above and below— fou rlowe e lineth t rsa e uppe edgth f eo r band—and divided int oseriea f panelso s setverticay x b fouf si so o t r l lines. Each pane s decoratei l a serie y f b dshoro s t diagonal lines extending inwardd an s downwards fro s righmit t and-left side. A series of similar diagonal lines hangs fro e lowemth re lowe edgth f ro e band of decoration. The rim is flattened but .LIA 1 I t thickenedno s i decorate d an , d with Fig . Beake11 . restoreds ra . parallel diagonae l edgeth s line l it n Al s.o diagonal lines except e uppethosth n ei r band of panels are impressed with a comb; the remainder of the decora- tio inciseds ni . e Alboneth l se beakefounth n i dr stratum were human. Their conditio s extremelnwa e majority th e tracebad th d f f bon,an o so y e found consiste moro n f eo d than white slime mixed wit e soilhth A . detailed report by Miss M. L. Tildesley of the Boyal College of Surgeons' Museu sucn mo h fragment possibls wa t i recoveo es t sa r Appendiin si . xI These include fragments of three skulls, and twenty-one teeth of a fourth individual remainine th ; g bones identifiable were consistent with belongin o thest g e four persons. These a weryoun ) (1 e g mana ) (2 ; broad-headed person aged about thirt o thirty-fiveyt , probably male; (3 youna ) youna g) adult(4 gd adultan ; aged about eightee twentyo nt . The leg bones probably associated with the broad-headed skull, showed e flatteneth d shafts whic e frequenar h n skeletoni t f thio s s period. BUDH' AN DUNAIN CHAMBERED CAIRN, SKYE. 199 The only other objects beakefoune th n di r stratum were white quartz pebbles; four rounded pieces of pumice, which occurred both in the chamber and the antechamber; traces of charcoal; and a roughly shaped poin f daro t k green chert 1'2 inch long. NEOLITHIC STRATUM. Fragment f pottero s f Windmilo y l Hill type were found scattered t varyina g depths withi e stratue th easternth n mi e n th par f o t

Fig. 12. Neolithic Bowl—conjectural restoration.

chamber and in the antechamber. They do not admit of restoration, t appeabu o represent r to bowls parttw f o s; fragment e samth f eo s vessel were foun de chambee antechamber th botth n d i h an r m a I . indebted to Mr Stuart Piggott for the conjectural restoration of one sectiom e othe ri e th drawin th e f rno r th fo f vesse go d t an figa l 2 .1 at fig. 13. It will be seen that the first is a round-bottomed, vertical- sided bowl, som inchee7| diametein s inche6 heightin sand r , wita h " ledge heavilm "ri y bevelle outsidee composeth s i o t t dI .a fin f edo dark grey paste containin n placegi s pocket a whitis f o s h substance considerabld an y pockmarke a resul s a d t apparentl f thiyo s substance dissolving outee Th .r surfac s burnishedi e e onlth yd decoratioan , s ni 200 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. close, diagona l e rim e beveflutine th seconth f Th .o n l o gd vesself o , which only a small part was found, has a straight side and a "ledge" rim slightly s probabl bevellee outsidwa th d o t dyan a round-bote - tomed bowl generally similar to the first. The paste is fine, grey-buff in colour, pockmarked like that of the first vessel and showing similar pockets of some foreign matter varying in colour between white and e outeredTh r. surfac s considerabli e y shaled t offwherbu , e complete shows sign f burnishingo s . Thersigo n f decora no s ei - tion, and two holes found at the same distance below the rim in two non-contiguous fragments were probably bored after firing. Trace f bono s e were founa numbe t a d f pointo r s in this stratum, but the only certainly human bones sufficiently preserved for identification were fragments of the left half of a mandible of a middle-aged indi- vidual, probably male, B.8 n additionI . a negativ, e f whico impressio e bon w completeld th hja eha a f no y decayed was found in the slimy black earth; the attempt to preserve it was unsuccessful. Two other deposits of bones were found. The first, B. 14, con- sistea singl of d e fragmen cannothe of tn bona of e sheep or goafc, completely burnt. The other, B.ll, was of special interest, as it occurred alongside of and under the foot of one of the orthostats of the chamber, P. 3, in such a manner as to suggest a foundation 3 curve deposit. P e foo f so Th t . upward Fig. 13. Section otf its southern side, leaving a small cavity between

a ^self e bone an(e paneTh th ^ 2/P. s . P l3 wer. e found a distanc uo t inche6 p f o e s inwards fro e facmth e s i unlikel t i d y an tha, 3 . t P ofthe y reached suc a hpositio y b n accident. They were unburn d consistean t f skulo d l fragments which cannot be identified with certainty; a piece of the cannon bone of a young ruminant, probabl ya calf e scapul a piecth ; a youn f f o eo a g animal, possibly a calf; and the head of the humerus of a bird, prob- ably a water-fowl. In Miss Tildesley's opinion (see Appendix I.) the skull fragments are probably not human. Dr Wilfrid Jackson, who has also been good enough to examine them, reports as follows regarding them sorrm : a " yfragmentaryI o s the e yar . This makes t veri y difficul identifo t t y them with absolute certainty I hav. e com- pared the skull fragments very closely, and there is the possibility that they migh humane b t . They appea agreo t r e very closel n texturyi e and in the bit of suture which is visible." RUDH DUNAIN A ' N CHAMBERED CAIRN1 , 20 SKYE. The only other objects e neolithifounth n i d c stratum were char- coal ovao tw ; l scraper f flinto s , 1'2 incd T0han 5 inch respectiveln yi greatest dimension; seven points and chips of flint and two of quartz showing 110 secondary workin somd gan e very minute lozengea d an ; - shaped object of quartz, l2 inch long, which might have been an

incompletely worked arrow-head- , but might possibly have been natural. scrapere Th s were made from flat beach pebble f flinto s , wit a hcon - siderable amoun f polisheo t d white crust remaining.

FORECOURT. The material of the inner area of the forecourt, as defined by the prostrat s examineewa , slab5 . S ,d fully t I consiste. a lowe f o dr stratum, approximatel a yfoo t thick f blaco , k earth—not slimy like that within the tomb—and above this brown earth. The lower stratum containe smalo dtw l fragment f pottero s y belongin same th eo g t vessel; the paste was coarse and reddish yellow in colour, and each fragment was decorated with parallel incised lines. The other finds were in the brown earth and consisted of three rounded pieces of pumice and a naturally shaped implement of a fine-textured dolerite, 7'3 inches long. This implement showed battering on its thicker end and also on a flat- tened surface near the point. The remaining material of the fore- cour examines wa t de cours s excavationonlth it f n yi e onlo th yd an , objects found, apart from a minute trace of charcoal at one point, were a number of blocks and large pebbles of white quartz. It is to be note f thesdo thao e tw tpiece f quartso z were e materiafounth n di l e cairoth f n vertically abov e forecoureth t facade.

COMMENTS. Structure. Crescentice Th Facade.—While e crescentith c forecours i t present in all horned cairns, the crescentic facade is found in Britain i nexample w onlfe ya n Pembrokeshirei s e Isl f th Man,o e , "western Galloway, Arran, Sutherland Caithnessd an , . The great burial chambe f Pentro r e Ifan, Pembrokeshire, oncd eha a crescentic setting of orthostats in front of it, described in Fenton's Tour, p. 560, as "seven stones that doe stand circle wise, like in form to the new moon "; no record appears to exist of the form of the mound.1 exampln e lonMa Isle th f ego Th s e i cair f Ballachrinkno f whico , h

Wgiver s . Hemp1.J thiM e ha no m s referencewh , , tell thastandine falleo o sm tw ttw nd gan orthostats of this fagade still remain and that near by there is an unrecorded tomb, Garn Turne, which has an irregular, deeply concave setting of orthostats of which five are still standing. 202 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. plaa partiad nan l elevation Reliquary, e existh n i t 1884-5 . 165p , , pi. xix. 1 On this evidence, which appears fairly trustworthy a e cair d th , nha long,, tripartite (presumably segmented) chamber opening directly upon a semicircular forecourt defined by a facade of orthostats set at intervals round it. Two smaller orthostats set transversely to the axis of the chamber appea havo t r e forme portala d . The Gallowa valleye Cree yLuce th th example th er n i ef o s o e li l sal Mid-Gleniron,e ar d an 2 Cairnholy, Boreland.d 3an bese Th t 4preserve f do these, Cairnholy, shows a facade of orthostats, ascending in height to e centreth , very simila thao t rt Rudha t n Dunaina ' t differinbu , n i g that the two highest pillars seem themselves to have formed the portal (thoug linte o remains)hn w no l , wherea lattee th n ri s tom portae bth l was a separate structure set between the highest orthostats. Probably same th e for portaf mo l existed othe e presene cairnso th th t rtw a n I .t state of all these three cairns the exact shapes of their forecourts cannot be seen. All are long cairns with oblong (long cist) chambers e generaoth f le Arra typth f no e cairns. Ther alss ei orouna d cairn t Cairnholya which show e samth s e e portalonth gs e a lcairth f no

same name5 , but there appears no adequate reason to assume that this formed part of a crescentic facade. Amon e longth g cairn f Arrano s definite evidenc a crescenti f o e c

facade existEasd an Cairn tsi n witBennan nBa ca h d fainan r certaint6 y be inferre Giantse th n di ' Grave Moinechoilld san . These appea havo t r e resemble Gallowae dth y example tha n si centrae th t l orthostate th f so facade themselves formed the portal. The shape of the forecourt seems to have been approximately semicircular, the facade joining at right angle e peristalithth s , which, afte a shorr t distance, turned agait a n right angles along the more or less parallel sides of the cairn. The chambers are in all cases of the oblong or long-cist type. In Sutherland, the southern of the two long cairns in line with one another at Coille na Borgie, Strathnaver, shows a crescentic facade at

each end. That at the north 7 end, in the centre of which is the portal, ap- pears fro plae sketcd mth n an h give havo nt e been semicirculare th d an , portal would seem to have been a distinct structure. As is noted below,

indebtem a I 1 thir dfo s referenc r StuarM o t et Piggott examines tell o thae ha se m wh , h t d e potterth y from this tomb preserve e Manth n xi d Museu d ascertaineman f o d s i tha t i t neolithic type. Roy. Com. Hist. Mon. (Scot.), Wigtown, No. 261.

2 3 Ibid., Kirkcudbright, No. 228. 4 Ibid., No. 362. s Ibid., No. 287. 8 Professo . BrycBooke H th . T rn ei of Arran, . 33-155volpp . .i . Proc. Soc. Ant. Scot., vol. xviii 228-33. . AndersonJ .pp d an , , Scotland Pagann i Times, vol. i .

pp.7 260-3. RUDH DUNAIN A ' N CHAMBERED CAIRN3 , SKYE20 . internae th l structur f thieo s toinh closely resembles Rudh Dunainn a ' . Ther mora s ei r les eo s round cair t Achanyna eastern i 1 n Sutherland which show a continuous s kerb roun e peripherth d y from which branche a shallos w crescentic settin f slabo g f whico s e axith h s coincides wit e assumeth h d positio e passageth f o ne chambe th ; s i r bipartit oblongalsd n inferree ean ca ob t I . d from Anderson's account of his excavation of one of the long-horned cairns at Yarrows, Caith- ness, that this had a facade rising in height to the portal. This fa

with those long barrow3 s of western England and which show

a sharp incurvin e enclosinth f o g g wala tru r falso o et l e portalt bu , 4 in these the forecourt is narrow and cuspidal and not in the shape of wel y , the a crescent.be ma l y s a represen , If 5 a developmentt e th f o same idea it is quite a separate development. s morit n eI perfect stat f preservatioo e n Rudh n Dunaia ' n shows better than any other chambered tomb in Britain the crescentic facade risin heighn gi portala o s panelt tt alonit I .s masonrf eso ha y joining e orthostatith c pillarst showi d an s, more clearly thaothey e nan th r relation of the facade to the portal on the one hand and to the peri- stalith on the other. The close similarity with the crescentic facade of westere th n Mediterranean wil noticede s beeb l ha n t I argued. that

this form originated in southern Spain6 , where examples are found at Los Millares d thencan , e e Balearispreath o t d c Islands, where th e feature is shown in some of the navelas; to Sardinia, where magnifi-

'. Roy. Com. Hist. Mon. (Scot.), Sutherland, . 447No . 2 J. Anderson, Scotland in Pagan Times, vol. i. p. 238, and Proc. Soc. Ant. Scot., vol. vi. pp. 442-51. 3 G-. Coffey, New Grange, p. 86. E.g. Hetty Pegler's Tump, Rodmarton Belad an , s Knap, Gloucestershire Caped an , l Garmon, Denbighshire4 . West Tump, Gloucestershire, show versa y shallow crescent. . LeedsT . E ,y AnnalsB s6 of Archaeology Anthropology,d an . 29-40 volpp . .ix 204 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. cent facades rising to an impressive portal are seen in the Tombs of the Giants; and finally to Malta, where the so-called temples show massive crescentic facades. Despit distance eth e betwee Millares nLo s d Pembrokeshirean e absencth d f geographicallo ean , y intermediate forms,1 it is difficult to resist the conclusion that the British facade derives ultimately from tha f southero t n Spain e facTh t . thae th t Britis e wesh th r nortsite o te aln o a lar s h se coaste closd th an so t e renders such disseminatio Atlantie typth e th y eb f cno route plausible, and the fact that the Spanish examples belong to the Copper Age in- volves no anachronism considering the late arrival of copper in the north. It is only proper to point out that, unless we can derive Wind- mill Hill pottery from the pottery of the Portuguese ,2 these similarities of tomb form between Iberia and Scotland are not paralleled by any corresponding similarity in tomb furniture. In e lates e th ligh th f to t researche f Professoo s . ChildeG . V r 3 this difficulty is one which may possibly have to be accepted as extending to the of north-west Europe as a whole. If Windmill Hill pottery represents a British development of a premegalith type common to western Europe, the megaliths are left without any distinctive pottery; although, as Professor Childe points out, there is an important connection as regards decoration between Beacharra pottery and certain types from megalithic tombs in Brittany. On this view we are left to draw the uncomfortable inference that the megaliths are the tombs of raiding chiefs from the Atlantic coasts of e Continentth , who n settlino , n Britaini g , adopted alik e wometh e n pottere e countryth th d f yo an . Other Structural Features.—In its general internal structure—vestibule, antechamber, and chamber ascending in height—Rudh' an Dunain is in marked contrast to the narrow oblong chambers of Galloway, Arran, and Argyll. It resembles certain of these, however, in the absence of a passage and the presence of septal slabs. Except in its large, polygonal main chamber littls ha et i ,resemblanc commoe th o et n Hebridean type of a single chamber with a long, low passage and little or no sign of antechamber. Its closest analogy is again with the Sutherland long cairn of Coille na Borgie mentioned above and with the neighbouring

t woulI possible db citmegalithio e t eth c structure Annaclochmullif so Newblisd nan s sa Iris1h e accountformsth t bu , s available (summarise "W"n di . Borlase C . Dolmense Th , f Ireland)o leave some doubpurpose plane th th d t thesbotf o san et o s ha e monuments. argueds A example r . LeedAntiquariese T fo , th . E n s i y b , Journal, vol. vii. 456-62pp . . 2 V. G. Childe, " The Continental Affinities of British Neolithic Pottery," a paper which I have had3 the privilege of studying in advance of publication. It will be published in the Archaeological Journal, vol. Ixxxviii. RUDH' AN DUNAIN CHAMBERED CAIRN, SKYE. 205 long cairn of Skelpick. Both these have three chambers increasing in

size and apparently als1 o in height, but they have also fairly long passages. (As has been pointed out above, it would be possible to regard the vestibule of Rudh' an Dunain as a very short passage.) The method of construction at Rudh' an Dunain, which I have compared to half timbering, is found at Skelpick, and apparently also in the main chamber at Coille na Borgie. This type, which appears to me to represen a distinct t structural method s elsewheri , e rare2—thera s i e Welsh exampl Capee th n eli Garmon tomb already mentioned—although of cours walliny dr e s widelgwa y use o filt d l interstice n buildingi s s whose stability depended upon their orthostats alone e typicaTh . l Caithness cairn is essentially a dry-walled building, with orthostats set transversely to give rigidity. Rudh' an Dunain therefore corresponds most closely in structure with these Sutherland long cairns, but the presencf septao t i l n ei slab s seemunique b o st e among Scottish cairns of the large, polygonal chamber type. The septal slab, however, appears in Ireland in a group of cairns on Carrowkeel Mountain, Co. Sligo, which contained food-vessels and had chambers of complicated, but principally cruciform, plan.3 The recess in the southern wall of the vestibule has a close parallel recese th n i juss* t withi e doonth r check e passagth f so f Maeeo s Howe, Orkney, but does not appear to be found elsewhere in Britain. A vague analogy exists with the small recesses on each side of the passage just withi e portanth t Brya l n Celli Ddu, though thes r HemM e p regards as the vestiges of an antechamber,5 and with the square recesses open- ing off the passage in Stoney Littleton long barrow, Somersetshire. Example f smalso l chamber recesser so passag a give e f b sof n n eca fro m chambered e westertombth n i s n Mediterranea e wellnth area-d an , known tomb with crescenti cs Millare a facjadLo roun s t a ha eds recess in the left-hand side of its short passage just within the portal. Nothin recesfoune s th awart g wa t n Rudhdsi a no Dunainem n a a ' I d ,an y evidencoan f e from other example o indicatt s y rituaean l purpose. It is possible from their presence behind the portal that they represent merely some feature in the nature of a guard chamber inherited from

1 Proc. Soc. Ant. Scot., vol. vii. pp. 273-4, and J. Anderson, Scotland in Pagan Times, vol. i. p. 263-4 with plan. 2 Mr "W. J. Hemp has suggested to me that this method of construction could be considered typica thosf lo e dolmens whic show hwno only separate orthostatic pillar t couli f sassumei e b d d tha spacee tth s between these were originally fille panely b d masonryf so . 3 Proc. Boy. Irish Acad., vol. xxix., Section C, pp. 311-47. Stuarr M 4 t Piggotthas very kindly show a large-scal e nm e plaelevatiod nan f thino s from " Notice of Runic Inscriptions discovered in Recent Excavations in the Orkneys, made by James Farrer, M.P. Printe privatr dfo e circulation, 1862." 5 Archceologia, vol. Ixxx. p. 191. 206 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. domestic architecture. I have mentioned above (page 192) the possible inference from the position of a block of stone in the vestibule that this, wit e vestibule-antechambehth r septum, forme groova d supo t e - port some ston r timbeeo r door structur antechambere th o t e . Apart from this no evidence existed at Rudh' an Dunaiii that the portal was blocked otherwis estonee cairne thath th y f vernsa b o ; y large rounded ston s indeeewa d found just withi t e suffiportalth nno t s -thibu ,wa s ciently much larger than other e cairstone th - o justif t nf in o s n a y ference tha t t hai fallet dno n there from above. Attention must also be drawn to the prostrate slab in the forecourt which, being chocked level, may be assumed to have served some ritual purpose. Sino e closeth e investigatio e forecourtth f no f chambereo s d tombs had hardly been attempted before Mr W. J. Hemp called attention s importancit o t s studhi f Bryy yo eb n Cellt surprisinno i s Ddui t i , g tha analogieo n t thio s t quoted e forecoure sb th sla n n I bca . f tha o t t tomb, however, about 10 feet from the entrance, there was a structurel some 10 feet by 7 feet defined by two short parallel stone walls with lina f post-holeeo s joining thei noticee rb inneo t ds i thar t i ends td an , this, like the prostrate slab at Rudh' an Dunain, was set slightly askew with the axis of the monument. If conjecture may be permitted where nothin s knowngi , both these features represented arrangementr fo s the use of the relatives of the dead who may be expected, on the analogy of primitive modern practice, to have watched the tomb for a prescribed period after the burial. The Burials.—The facts available for the consideration of the period and method of use of the tomb are: (1) the definite stratification of the burials in a lower black-earth stratum marked by neolithic pottery, and uppen a r brown-earth stratum marke beakey db r pottery tha) (2 ; t there were not less that two burials in the neolithic stratum and not less than e beaketh fou n i rr stratum tha ) e onl(3 ;th t y skull complete enough to allow of classification was broad-headed, and was from the beaker level; (4) that the beaker was of a typically British shape common in north-east Scotland (cf. page 208) tha) (5 ; t e sampartth f eo s neolithic bowl were found in the chamber and the antechamber on opposite sides of the septal slab; (6) that a deposit of unburnt bones was found beside and under the foot of one of the orthostats of the chamber in e neolithith c level tha) (7 ;t fallen walline flooe th gth f o rblockn o y sla forecourt below the stones of the cairn in such a position as to make it virtually certain that their presence was not due to a modern clearing of the forecourt in quarrying the cairn for stones. 1 Archceologia, vol. Ixxx. pp. 191-6. An ox's carcass was found in a pit in the centre of this possiblt no structures ascertaio ewa t t i t bu ,n certainly tha ancienn a t this swa t burial. RUDH DUNAIN A N CHAMBERED CAIRN7 , 20 SKYE. 1 There is an almost complete absence of record of stratified deposits in other Englis d Scottishan h chambered tomb comparo t s e wite th h stratified deposits in dwelling sites, such as the ditches at Windmill Hill, which showed Windmill Hill pottery under Peterborough pottery under beakers. There is a fair amount of evidence of the prevailing dolicho- cephalic characte e neolithith f o r c inhabitant f westero s n Scotland. Ther alss ei ogooa d dea evidencf o l e that vessels deposite megalithin di c tombs were normally deposite dt strew wholno d nean abou sherds.s a t 1 With these premises the inferences suggested below are drawn in respect of Rudb Dunainn a ' e degreth ; f probabilito e y attachin eaco gt h must be judged fro e factmth s e precedinlistedth n i g paragraph. The tomb appear havo st efoe beerus successivn ni e burials2 during a period long enough to account for the thickness of the neolithic and beaker strata. During that period "beaker folk" entered Skye from mainlane th becamd dan e sufficiently absorbed int native oth e population alloo t theif wo r being burie nativa n di e tomb. Their arrival, however, involved some change in burial customs, as evidenced by the difference between the black earth of the neolithic and the brown earth of the beaker stratum, though exactly what this differenc s obscurei s wa e . When later burials were made the vessels belonging to earlier burials were throw t fro chambere n ou mforecoure th th d an , t arecleares awa d s floo it f cair o o rt n material—no e propedoubth r fo rt performancf o e e deade rituath th f .o l Finally d e subjecremarkth an , o t t s made below, a foundation deposit was made at the foot of one of the orthostats when the chamber was erected. s unfortunati t I e thae skulth t l fragments include thin di s deposit certaiadmio n f o tn determinatio o whethet s na r the e humayar r o n animal. Even, however, if the latter were the true view, it would be difficult to interpret the deposit as one of food made in connection wit hburiala , becaus e skuleth l fragments wer einches6 placeo t p du unde e orthostatith r e bonec th pilla d s an identifier d belong mainly to parts of a carcass which are not edible. Despite the complete absenc f recoreo f foundatiodo n burials under chambered tombs,e th 8 probability that such sacrifices were made is not negligible; since the

Compare, however, Thurnam's statement that the sherds in the chamber of West Kennet long1 barrow, Wilts, were deposite separatn di t represenno ed heapdi d t completsan e vessels (Archceologia, vol. xxviii. p. 417). Mr W. J. Hemp has stated the argument for supposing that some chambered tombs were not used2 , or at least were not intended, for successive burials, in a recent paper in Arch. Camb. (December 1931) . 253p , . t BryA 3 n Cell deposia theru s burnt!a Dd ewa f o t , human ear-bone unburnn a , t piec hazelf eo , charcoad an l fragment slab-coverea n si d pit, three feet outsid wese eth chambee t walth f o ld ran exactl centr e monumente th th t yf a e o . Thiclearls swa somn yi e sens efoundatioa n deposit con- nected with the erection of the tombs. (Archceologia, vol. Ixxx. p. 196.) 208 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. bases of tKe orthostats are normally some feet below the floor of these tombs and have very rarely been reached by excavation the negative evidence proves nothing e practicTh . f buryino e ga huma n beinr go an animal under, or actually clasping, the post of a house, or under a wal r foundation-stoneo l e s spreab i , n l oveca e al world th d r dan shown to be of great antiquity.1 It still survives in modern Greece, wher ecocka , ram r lamo , s buriebi d unde e foundation-stonth r a f o e building; in Transylvania, where human shadows are buried and the owner of the shadow is expected to die within forty days;2 and Mr Stuart Piggott tellthae m sn Hampshiri t recenn i e t year a parsos n was hel a mason tali dy b k n whil e latterth e , after ascertaininy b g discreet inquiry tha parsoe first-bora th ts nwa n son, buil shados hi t w inte foundatiooth a churchyar f o n d walle foundatioTh . n buriaf o l two women beneath the wall of a hut at gives evidence of the practice in Scotland in late Bronze times.3 There were cremations in all but six of the thirty-two holes so far excavated in the " Aubrey" circl t ,a e s highli t i yd probable—thougan 4 t i hdoet no s appear previously to have been suggested—that these represented sacrifice f humao s n beings -whose souls were intendee th o holdt p u d wooden pillars which these holes contained. same Th e5 purposy ma e explai n e buriale foosom th f th standin o f tt o ea s g stones, whether isolated stones or members of stone circles, and particular attention may be called, since these belong to a chambered tomb,, to the crema- e foo f tionmonolithth o t t inmosa e s th f o st circle surrounding .e Pottery.—In our present knowledge of British neolithic pottery it is not possible to say more of the two neolithic bowls than that they are of Windmill Hill type and that they do not show the decorative motives of those vessels classified by Professor Childe as the Beacharra secone group th holee n i dTh t likel s. vesse m no belori o e yt e ar wlth be rivet holes and may well have been bored for purposes of suspen- sion; being made after firing they are not likely to have been for decorative purpose n somi s a es neolithic vessels fro e soutmth f o h England. Judges shapeit a e laty beake b df th ,e o s Britisi r h type particularly common in N.B. Scotland. Its decoration is apparently unique, the nearest parallel in Abercromby being No. 290, probably V. G. Childe, Skara Brae, p. 142. 21 Sir J. G. Frazer, The Golden Bough, abridged edition, p. 191. Dr J. G. Callander, on the other hand, argues that Skara Brae dates from the Early Iron Age3, Proc. Soc. Ant. Scot., vol. Ixv. pp. 103-14. 4 Lt.-Col. R. H. Cunnington, Wiltshire Archceological Magazine, vol. xliv. p. 338 (July 1929). That originathin a s si l belief from which foundation burial Frazer. s G aris . J shows er i , Si y nb The5 Belief Immortalityn i e Worshipth d an Dead, e oth f vol. 446p . i ..

Archceologia, vol. Ixxx. pp. 201-4. 6 RUDH DUNAIN A ' N CHAMBERED CAIRN9 , SKYE20 . froia Ross and now in the Scottish National Museum. This shows the same arrangemen s doea e tRudh th s n Dunaia ' n beake f simplo r e decoratio rectangulan ni r panels arrange horizontao tw n di l bandd an s shows also the same edge decoration. Metopic decoration is, of course, fairly common on British beakers and derives from central Europe, being unknown on Spanish or Breton beakers. Edge decoration as on e Rudhth Dunain a ' n beake rars i r Englann ei d outside Yorkshire,t bu

e influencth o t f t uncommoe o eno du s i e b t i y n Scotlani n ma d dan 1 neolithic y accordinglpotteryma t I . inferree b y d thae maketh t f o r this beaker reached Skye fro mainland.e mth 2 Quartz.—The presenc f whito e e quartz neae e surfacth th r f o e mound over the forecourt facade shows that pebbles and blocks of this rock were spread t merelovee mounno th rd yan ddistribute d roun e peripheryth d quarto e neolithiN s founth . wa zn i d c stratum within the tomb, and that found in the beaker stratum, though probably a funerary deposit, might possibly have fallen in with the material of the cairn. Pumice.—The finding of seven lumps of pumice in the beaker level in the chamber and antechamber, and in brown earth in the forecourt, raise questioa s n which doe t seesno m previousl havo yt e been discussed. These lumps varied in greatest dimension from If inch to 3 inches and showed no signs of use. Dr H. O'Neill Hencken has very kindly given me the following information: " About thirty years ago the late George Bonsor found a piece of pumice in a passage grave on St Mary's Island Scillyn i . Wit t hweri e t likpiecepo e f e earliesthoso sth f eo t Bronze Ag Cormvaln ei Brittanyd an l . Bonsor subsequently tole thae dm h t had also found pumice in some megaliths in southern Spain, and that he suppose people dth e use t insteadi f soapdo I don'. t know whether e eveh r published these finds. Scille Britise Th th "y n fini h s di Museu m and is a small rounded lump, pierced near one end. I have myself found a large piece in a chambered tomb in North Uist at which I have been working; this had a flat face and had presumably been use r rubbindfo g down woo r skinsdo numbeA . f referenceo r e th o t s findin f pumico g sandhiln ei l sites wil foune b l Erskinn di e Beveridge's North Uist, but slag is also found in these sites, and in the absence of . KendricD . T r kM 1 kindly inform beakere thate th Britise s m f th o , n si h Museum, threer o , possibly four, from Yorkshire have some ornament inside the lip, and one from the Thames at Mortlake. bees ha nt I argue2 Misy b d s Margaret Mitchel Antiquityn i l (March 1932) thanecessars i t i t y to postulate sea-borne invasion fro e south-wesmth o accoune "t tB th r beakere fo tth n o s western Scottish seaboar more th r e wels northerlfo d a s Ca d l ' grouan A y ' Lewis n pi , Uistd an , Skye. samClare . agreI D "th . en ki G numbee . witJ r hAntiquityM f ro tha lattee th t r groupo t , whic e Eudhhth Dunain a ' n beaker belongs, derive fro Englise mth beakehA movemeny rb p u t ease th t coas Britainf to . VOL. LXVI. 14 210 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. microscopic examination this appears not always to be distinguishable from pumice.1 I am indebted to Dr H. H. Thomas of the Geological Survey for the full investigatio e origith f f thio nno s pumice whic s publishei h n i d Appendix II. It will be seen that the pumice is basaltic and probably of West Indian origin, carried to Hebridean shores by ocean currents; it is definitely not of Mediterranean origin, and it is not necessary to suppos s broughe thawa t y humai tb t n agency. Present information does not allow of a conclusion as to whether it was designed for the r some fo dea th r practicae o d f o magica e us l l purpose e Nortth ; h Uist example suggests the former and the Scilly example the latter, though the mere fact of being pierced for suspension is not conclusive against domestic usee Rudh Th .n Dunai a ' n finds, being unworked, are more suggestive of a magical use and, though those in the forecourt havy ma e bee e chamben th e occasio throwf th o a n latet o rf nou no r burial, it is more likely that they were not a funerary deposit. The most, probable view e thaseemb to t spiece f pumiceo s , like quartz pebbles and shells, served as receptacles for souls and therefore, at a later stage of development, as charms. The fact that pumice is full of holes may be significant as allowing the entry of a soul without e necessitth f breakinyo e lumpgth was a , s frequently done with quartz pebbles. It may also be significant that pumice floats, having regard to the widespread desire to provide a boat for the dead to enable him reaco t s ultimathhi e resting-place moderA . n Serbian ritual performed on the anniversary of death may throw some light on this. The wife or daughter of the dead man takes wet pebbles from a river, places them on the bank with food upon them and, when the soul is attracted to the stones by the food, makes circles round them to enclose it in e storiesth . Thes e theear n place smaln do l planks with lighted candles and sent floating dow streame nth purpose Th . clearls ei facilitato yt e soul'e th s departur wates y laseb it t o t rhome . I desire, in conclusion, to acknowledge with gratitude my indebtedness o Macleot f Macleoo d r permissiofo d o excavatt n o t e cair d th e an n e Counci th e Societth f o lf Antiquarie yo f Scotlano s r approachindfo g hi mthao t r Macraet M end e o tenant ;th , f Gleo t n Brittles hi r fo , interest in the work and the loan of tools; to Miss M. L. Tildesley of e Royath l Colleg f Surgeonseo ' Museu e e bonesreporth th r n mo ;fo t to Dr J. Wilfrid Jackson of Manchester University for a supplementary e Imperia. RileL th repor . f yH o bones e somr n th o ltD f Collegeo o t ; e

. CallandeG . J r D * r cite snumbea f findro pumicf so e from earth-hous eEarle siteth f yso Iron froAged an m, broch Hebridee th n si Orkneyfrod e inhabitesn an mon a d an , Firtde sitth f hn o ei Forth suggeste H . s usrubbinr efo g down bone.—Proc. Soc. Ant. Scot., vol. Ixv . 350p . . a H

______4——-_ -,

-a _ s * 2 C je» 0 V w 3 c; j x. O 03-1 t- C u •- .r ° x V ..J _C: ^ ..^ c a

H O O

03 C I fc J Proc. Soc. Ant. Scot.] [VOL. LXVI.

Section alone Line X.Y ml

C h a An t « c t| am t.. r Vestibule Portal F o recou v t

EUval'.on o*' Pocl-at at.t( For«tour|- Pacode (Q/ reat o r

W. LINDSAY SCOTT. PLATE VII. Proc. Soc. Ant. Scot.] [VOL. LXVI.

Excavation Plans of Chambe& r

ifi, Bi.bt, «fe. aas ttfp.%)

W. LINDSAY SCOTT. PLATE VIII. RUDH DUNAIN A N CHAMBERED CAIRN1 , SKYE21 . 1 of Scienc Technologd ean chemicaa r yfo l repor certain o t n boneso t ; . Thomae H Geologica th . H f o sr D l e e Survereporth th n r o tfo y . O'NeilH pumice r D l o Hencket ; f Harvarno r informatiodfo n about other finds of pumice in megalithic tombs; to Mr Stuart Piggott for drawings of the neolithic bowls and for other assistance; to Mr W. L. Coats of Glasgow for two photographs of the chamber, one of which is here reproduced . finallJ d . an ; W o Professot yr . ChildM G d . V ran e Hemp for their very valuable assistance on many points arising in the preparatio f thino s paper. Sir Reginald Macleod of Macleod has approved of the pottery being preserved in the National Museum of Antiquities at Edinburgh.

APPENDIX I. REPORT ON BONE FRAGMENTS. By M. L. TILDESLBY, Curator of Human Osteological Section, Royal College of Surgeons' Museum.

NEOLITHIC LEVEL. Site B.8.—Four fragments comprising left hal f mandibleo f teeto n ; h lost pre-mortem, all three molars up and all worn. Middle-aged indi- vidual, probably male. Site B.ll.—The bone fragments includ a fragmene e uppeth f ro t e cannoth en f do n bonyouna f eo g ruminant, probabl ya calf . Frag- ment of scapula of a young animal—could be a calf, but fragment too imperfec o determinet t . Hea f humerudo f birdo s , probabl a ywater - fowl. Skull fragments which canno e identifiedb t t probablbu , t no y human. Site 14.—Burnt fragment of the lower end of a cannon bone of sheep or goat. BEAKER LEVEL. Site B.2.—Fragments of the skeleton of a young man, including parts of skull, thigh bone, upper arm bone, rib. Site B.I.—Fragments f humao n long bones only e identifiablth , e fragments being from the legs. Bones from Sites 1 and 2 may belong to the same individual. Site B.12.—Parts of human leg bones. Also heel bone and fragment of pelvis. Tibia platycnemic; individual adult, probably male. Site B.13.—Fragments e fragmen on f skul o d an l t eac f humeruho s and radius. The skull was that of a broad-headed individual, aged c. 30-35, probably male. 212 PROCEEDINGS OF THE SOCIETY, FEBRUARY 8, 1932. Site B. 5.—Parts of thigh bone and upper arm bone and many small unidentifiable fragments. Femur platymeric. Bones from sites B. 12, B. 13, and B. 5 may all belong to one individual. Site 5.3.—Skull fragments of young adult individual, together with the enamel crown of four upper molars, viz. 1st, 2nd, and 3rd molar o nmolad 3r right left n d o r .an , Site 4.—. B Twenty-one teeth—includin l twelvgal e molars—o youna f g adult aged c. 18-20. Only the enamel crowns remain in many cases, but a root of one 3rd molar remains apparently intact, and still some- what open, showing that thi t completels no toot s hwa . yup Site BB.l.—Some fragment f verso y much decayed bone with chalky deposit1 on outside.

APPENDIX II.

REPORT ON PUMICE. By HERBERT H. THOMAS, Sc.U., P.R.S., Petrographe e Geologicath o t r l Survey. I have examined the sections and specimens of your material from Skye and find that it is definitely a pumice of basaltic character, and e samth f eo type l al . t consistI brownisa f so h glass with abundant vesicular cavitiese Th . glass contains microlite f greeniso s h augit felspard an e d there an , ear occasional small rounded phenocrysts of basic placioclase. These facts, and the general character of the glass (low refractive index) indicate •without questio artifician a t nno ls thabasia i slat t i tc gbu volcanic product suchs A s presenc. it , f greao s ei t interes t seemi s a t indicato st e a volcanic eruption of considerable magnitude at or about the period of sitee th . e source th s probabl i Ao t st i , e that volcanoes were erupting basic material in the Mediterranean, Iceland, and the West Indies, but the characte pumice th f ro quits ei e different from Mediterraneae thath f to n volcanoe f Strombolio s , Vesuvius e othe th d Etna f an r, O seaboar. d sources Iceland is the less likely, because a southerly drift from 1 Dr H. L. Kiley, A.K.C.S., D.I.C., D.Sc., Lecturer in Chemistry, Imperial College of Science and Technology, reports as follows on this chalky deposit:— "The white deposit on the bones from Skye contains alumina and phosphoric acid together with small quantities of ferric oxide, silica, water-soluble sulphate, chloride, and organic matter. It is probably a basic aluminium phosphate which has been formed by the slow interaction of the calcium phosphate of the bone with aluminium silicate present in the soil in which the bones were buried. This latter coul e presenb d t eithe kaolinises s claa a rr yo d felspar fro granitee mth . Fossilised bones have been reporte contaio dt n ferric phosphate, whic probabld hha y been formed in a similar manner." RUDH1 AN DUNAIN CHAMBERED CAIRN, SKYE. 213 Icelandic water s i contrar se knowth o t yn directio f currentso n , wherea a drifs t fro e Wesmth t Indies wouln accordanci e b d e with what we know takes place. I therefore incline to the view that the pumic f Weso s i et Indian origin d thaan , t you e conb r - y sitma e temporaneous with some great paroxysmal eruption in that region, during which much pumiceous material was ejected into the se.a arid carried eastward prevailine th y b s g wind currentsd an s .

MONDAY, Uth March 1932. THE HON. LOUD ST VIGEANS in the Chair.

A Ballot having been taken, the following were elected Fellows:— ROBERT BLACKWOOD, J.P., 9 Oxford Street, Dundee. CECIL JERMYN BROWN, M.A., Buccleuch House, Melrose= . WILLIAM M. CROOKS, J.P., Ardmere, Monifieth, Angus. ROBERT DUNCAN, M.A. Strathmartin4 29 , e Road, Dundee. ROBERT JAMES LARG, J.P., "The Hollies," 63 Clepington Road, Maryfleld, Dundee.

The following Donations to the Museum were intimated and thanks e Donors:voteth o dt — (1) By JAMES R. DURHAM and PATEICK M. THOMSON. Fragment f hand-mado s e Pottery a piec, f Pumicefragex o si -d an , ments of rude Stone Implements, found by the donors in a "burnt mound " at Weisdale Voe, near Heglibister, Tingwall, Shetland.

(2) By JAMES 8. RICHARDSON, F.S. A.Scot. Scraper of white Quartz; rim fragment of a Vessel of Pottery, show- ing part of a lozenge design, and a Disc of Bone, found by the donor on Quendale Sands, Dunrossness, Shetland. (3) By Sir GEORGE MACDONALD, K.C.B., F.B.A., LL.D., D.Litt., F.S.A.Scot. Cast of a triangular Bronze Relief of Jupiter Dolichenus and other subjects from Heddernheim, Germany, now in the Museum at Wiesbaden.