A CHEARDACH MHOR, DRIMORE, SOUTH UIST
by ALISON YOUNG, F.S.A., F.S.A.SCOT. . RICHARDSONM . K d an , , M.A., F.S.A.
o describT e siteth e adequately some indicatio environmene th f no necessarys i t . The western coastline of South Uist consists of a superb sandy beach, defined on the mora eas y r lesb teo s continuous ridg dunef o e , whic t Drimorha e rise fee7 o s2 t t machai e levee th th f lo r coastaro l plain. Eas thif o t chaisa shallof no w lochs with occasional scrub-covered islands stretches from nort southo ht , betwee machaie nth r anrocke dth y hills whicf o , highese Mhorn hth Be s i ,t 2,035 ^eet above sea-level. S machaie o th fla s i t r thaIsle Barraf th teo , som south e mile0 th e2 o plainls t i , y discernible Scurrivan nortDu e th d t han a l, poin thaf o t tidentifiede islanb n dca . The sandy coastal plain includes a stretch of dunes, covered with coarse grass, persisting from the shore on the west to a broken line of low banks which suggests beacd ol n ha line t appearI . s probable tha nearba t ysitee norte burth th , f ho o nt now draining eastward, has changed direction, probably due in part to the shifting sands of the machair. The dune, in which a number of sites were located, alternates wit aren ha a whic evers hha y appearanc beinf eo g water laid seasonaA . l lagoon of varying extent forms immediately to the east of the seaward ridge. Though cultivated in part at the present day, the machair is mainly grazing land, and in spring and early summer a flora, chiefly white clover, silver-weed and orchis in variety, provides a dense and brilliant growth. The mound known as a Cheardach Mhor,1 the Big Smiddy, is 300 yards from the shore, nowhere higher than 6 ft. 6 in. above the machair level, and covers approximately three quarter acrn a ef so wit h irregular hummocks, showing plainly ovan aa s l hillock abov e flaeth t grazing land which surround lone th g , it saxi s according with the prevailing south-west wind. foue Baseth r n markindo g peg surveyore sth lefy Ministre b tth f so Worksf yo , whose contour plan provided the basis of fig. i, two lines of pegs were driven in at right angles, and a grid was laid out. Work started on the north-east quadrant on 23ry 195 continued dMa 6an perioa weeks. x r si d fo f do 2 surface moune Th th brokes f rabbid eo d ol wa y ntb burrows from which, during the preliminary survey undertaken by Mr P. R. Ritchie, fragments of pottery and bone refuse had been recovered at various points. Large stones heaped on the surface indicated fairly recent robbing of some underlying structure, and, though odd stones protruded from hollows mainly on the south side of the mound, no wall-
1 For the bibliography of other sites referred to in the text see p. 169. 2 This excavatio undertakes nwa behalMinistrn e no th f fo Worksf yo programme parts th a f o rescuf eo e work anticipatin thif o se stretcmachaie us e th f rockegth a h o r Stewar s a rM o tt range e r thankt du Ou e . sar . CrudenH , F.S.A., Inspecto Ancienf o r t Monument Scotlandr fo sco-operatios hi r fo , n throughoute ar e W . also indebted to the following for their several reports: Mrs F. L. Balfour-Browne, Dept. of Botany, British Museum (Nat. Hist.); Dr A. D. Clarke, Dept. of Nat. Hist., Royal Scottish Museum; Mr K. W. Elliot, Geologica . Lacaillel O Survey . A r . Oakley, P M ;F.S.A. . K r , D ;Dept Geologyf o . , British Museum (Nat. Hist); Mr James Phemister, Geological Survey; and Prof. G. M. Yonge, Zoology Dept., Glasgow University. FIG. GheardaciA . h Mhor, area excavated A CHEARDACH MHOR SOUTH UIST G ^
IA C*>) Uprightts Slabs oor
i7\ FFORECOURT1
5_____0 10 0 2 15 FEET H H H Fro. a.A I Phase d an sI A CHEARDACH MHOR, DRIMORE, SOUTH UIST 137 ing was visible at any point. The nearest available natural stone is half a mile distant. No reason can be suggested for the earliest choice of the site, now almost hal mila f e fro nearese mth t sourc surfacf eo e water, for, though water occura t sa depth of 22 inches today, it is unlikely that this was as high when the earliest struc- tures were first built; the foundations now stand in waterlogged sand. Excavations revealed five distinct phases of occupation, with the robbed remains o wheelhousa f lowese th t ea moundete levelth n I . d sand, whic covered hha e dth earliest building, four later phases were defined. Stones brough e sitr th efo o t t buildin wheelhouse gth e were used, probably severa foue l timesth r d succeedinan , g occupations honeycombe mounde d addeth dan o dt . These later living horizons d theian r midden tips became distorte blowy db n sanfurthed an d r confusey b d stone-robbers, so that the uppermost level (Phase V) cannot be considered as strictly stratified. Fortunately the central area of the primary building lay under a variable depth of blown sand which, though tunnelled into by later inhabitants, remained sealed traco N roofinf . eo g material coul identifiede db r werno , e fallen stones from the upper courses found in the bays. Corbelling, for which the incurve maie oth f n walwidenind an l g pier sole ar se evidence, must therefore have been removed by later builders, which argues that the original sand-blow was probably much deeper, and the aspect of the mound may have changed materially before the present growt turf ho f stabilise. dit The series of trenches £1-7, Fi~4, Cn, and Ci2, Di2, Fi2, revealed little more than tips of peat-ash with unrelated settings of stone, seldom more than a single course high and some material from temporary occupation levels; the lower layers in the main were sterile down to the water-logged sand which formed the 'natural'. This confirmed the evidence provided by a trial trench which had been dug beyond the limits of the mound, where water occurred at a depth of two feet belo pure wth emachaire santh f do madd an , e further digging impossiblee Th . water level, however, varied with the rainfall.
PHASE I, THE WHEELHOUSE A stone linte Trencn i l , whilhF6 e obviously moved fro originas mit l positiod nan set on edge, was the first indication of the wheelhouse, confirmed by uncovering in Trench F7 what appeared to be the inner end of a pier. It was, in fact, the south wall of the entrance passage (figs. 2 and 3; PI. VI). structure Th eirregulaf proveo e b o dt r shape lowee th , r coursecurvino tw f so g walls forming a forecourt outside the entrance on the north-east, faced away from weathee th r quarter prevailine th , g wind being fro south-weste mth lone Th g. axis of the main structure was 40 ft. and the inner area was divided into bays by eleven radial piers whicf o , h nine remained l ruineal , varyinn di g degrees e floof Th o .r centrae th l area measure . froft m 2 . d2 Stonewor 8 Piero t westere s2 th n c ko n ar outee oth f r walbeed ha l n entirely destroye robbing y foundatione db th t bu , f o s one pier . 10,No , were traced belo robbee w. th long in 6 rston ,a . leve ft d e2 an l possibly a re-used kerbstone, was deeply embedded at a point where another pier A CHEARDACH MHOR S.W. F7 F6 B F9 F8 A iiiiiiin minium inn H iiiiiiiiiiiiimMiii|i|iMi'|;Mi iiiiiiiiiii|iji|ii iii;;;;;;;!;i,:;-;;-; li; ;;;;
PHASE is,,,,,; hll .Hi; "H' uli. .i"i -V ""Illllllllliiiiiiniiiiiiiiiii^llillllllllllllllWI11111"
5r4r^Ud!li!ll*I- "" ' •> N\\\ \\vVv \^<>X ;x\\\\^\.x ^ >\vv^ xXx N x Robbed walf o l PIER 9 \\\ ,\ > .\\\\\\\\\\'\x\ \\\\\\\\\x wheelhouse •^"- Quern projected Forecourt Wheelhouse wall projected.
N£. F2'
8 .;.;.;.;.;. Humus 9 Blown sandI Clean sand v 2 Subhumus 10 Occupation PHASEI* Uv.-:!-.-.:"i.'::.W Water oqged '" '" ' "' sand 3 Peatash PHASET b&bA Stained sand 11 Robbed area 4 Dirty sand 7 Occupation PHASEET 12 Blown sand 10 FEET 5 Interleaved peatash & loam PHASEIH 8 Sand flecked with charcoal 13 Occupation PHASEI FTG. 3. Section of site A CHEARDACH MHOR, DRIMORE, SOUTH UIST 139 would normally occur. The robbing, though extensive, was localised and the almost vertica lintruders e sideth f so ' trenches were clearly defined agains blowe th t n sand. Nowhere was a sufficient number of large stones left to warrant certainty as to the method of roofing the bays, but from the nature of the piers these were probably corbelled. The upper stonework, where visible, would provide a convenient source of building materia succeedinr lfo g occupations maie Th .n wall varie thicknessn di , at widest 3 ft. and, where it stood to any height, showed a tendency to corbel in- wards. Thoug aree hth a enclose wels dwa l levelled lowee th , rstructuree parth f o t , t least a originalls wa , y backed agains sande th t , - whicir accouny e th hma r fo t regular shape poine , reinforces andon t wa outside a ,, behinth 5 n y y do eb dBa boulders piled roughly against well laid courses.1 Here, however, there were suggestion wallinge breaa th f e possibilito sn ki Th . secona f yo d entrancs ewa considered but no evidence at foundation level could be traced, either from the interior, which at this point was particularly rich in occupation refuse, or on the outer periphery, where the likelihood of a passageway into a second building was envisaged. The upper courses of stonework were disturbed and rebuilt, but the resultin t characteristigno wals wa l f wheelhousco e masonry e disturbancth d an , e most probably related to secondary walling of Phase IV at a higher level (see . 157)p , thoug possibilite hth blockea f yo d doorway canno disregardede tb . Notably large stones had been used to build the lower courses on the south, and part of the south-east arc was stained on the interior face by peat-ash, which must have piled againsp u maie th t n wall. Bed f greeso n clay localitye occuth n thid d i r an sha , been used in places as mortar; there was, however, no evidence of clay foundations and the walling was laid on sand. Clay was also used to secure a large stone set in entrance th wheelhousee th f eo bouldee . th Measurin . firmls in - rwa 9 yem . ft g2 2 bedde r hal heighs dfo it f wedged an t d with stones, som f theeo m worn hammer- stones fragmenA . f pottero t clae founs th y ywa n infili d l together wit hpartla y polished gouge-like implement made of antler (fig. 7, 9). The purpose for which this ston carefulls edifficulwa s i t y se assess o t tt woul i s a , d appea impedo rt e traffic single th f eo entrancet ou d an t seem I .n i s likely therefore tha servet ti structurada l purpose, such as the foundation for a support to carry some form of roofing over the paved passage connecting the wheelhouse with the forecourt. THE FORECOURT The two curving walls of the forecourt were asymmetric and had been extensively robbed sealeo pavin e wheelhouslevee e Tw th .d th th f layerslt o gn a i e on , e passage, the other divided from the first by sand and interleaving peat-ash, contained pottery, a bone needle, a pin and half a polished bone pin-head, as well as a broken bone quern handle ston p (figrotar a Nos, to f 7 .eo , ie 14),d 6 .y uTh an hand. - mill wit hpartla y perforated handle-hole e loweth n ri levey la , l d (fig an , i9 . PI. VIII, 4).
This is borne out by the fact that tips containing wheelhouse pottery occurred on a level with the remainin1 g walling on the north-west and above the level of the robbed walling of the forecourt to the east. claf o mortas ye a us reportnotee e s ri th 2Th Bac d n Foshigarrydi n an s o Mhic9 29 . p Connain, . 42 . p 140 PROCEEDINGS OF THE SOCIETY, 1959-60 BAYE E PIERTH SD TH SAN The inner terminals of piers Nos. 2, 4 and 5 (PI. VII, 3), still standing to a height abovo s e wheelhousr o th e . ft 4 f o e floor, showed feature n commoi s n wite hth nearby sit f Kilpheder,o e which was, however n aislea , d house with lintels joining oute e piere th th o rst wall. But, wherea Gheardaca t sa h Mho survivine th r g piers, though carefull maie th y no built wall p u t , wer t bondet theseno ye , e din showe d no sign of a straight joint (see fig. 4). We must therefore conclude either that the
FIG . Pie4 . r a hous neves et wa beei r originallyd o aisledns ha , or , , tha completa t e rebuile th f do piers must have taken place uppermose Th . remainine th f o t g piere stoneth s swa larges oversailed an t lowee dth r courses wit hcorbellina inche3 feet.f 4 ge o n si Th 1 facing stones of piers had been dressed and trimmed, and worn hammer-stones were built intwallinge oth , doubtless discarded mason's tools heaa ; f chosepo n beach pebbles including some used hammer-stones, lay in Bay i. This bay, defined from the Central Area by a kerb running from pier to pier, was sterile, apart from the heap of stones and a hollowed-out whale vertebra 'cup' which was found at the bas f Pieeo (fig2 r , 19)8 . ; this singulapies wa r havinn i r ga protrudin g footing- ston t flooa e r level. Fro e water-loggemth d foundation e maith f no s wallinn ga interesting sher s recovereddwa . This fragmen encrustes wa t d wit a e cas f hth o t sea worm, and its presence, 300 yards from the beach, was curious.2 alss Baowa kerbedy2 antlen A . r quern-handle, wor polished endne an on ,t da was found in this bay. Bay 3 showed signs of considerable disturbance. The wheel- house walbeed ha l n partially remove mucd dan Pief ho demolishe3 r d towarde sth main wall, though the lower courses, including the kerbing, remained untouched. A clumsy, haphazard structure filled the upper level of the bay, using Piers 3 and 4 wall e temporara th sf so y shelte wheelhouse Phasn Th i r . eII e occupatios nwa covered by a level of blown sand and the two periods were well demarcated. A dressed stone, robbed fro e wheelhousmth used secondary an edb y buildersd ha , fallen on and crushed sherds from which the complete section of a pot belonging
inware Th d corbellin radiae th f go l pier t Kilphedersa becomes apparen. heighe ft th 4 t f a to t 1 Dr Yonge, who kindly examined the sherd and identified the encrustation, says that the sea worm woul2d choose a flat surface in open beach conditions for forming the cast. We are left with the query: did occupatioe th d wheelhousee en th brina f n na o se o ge t th , flooding ove machaire thith s r swa sherr o , d left on the beach by some previous inhabitant, and, found on the seashore by the builders of the wheelhouse, subsequently brough sit e accounn th eo interlace e o t th f o t d pattern? A CHEARDACH MHOR, DRIMORE, SOUTH UIST 141 to the Phase II horizon was restored. Though the upper levels of Bay 4, also kerbed toward centree sth , showed sign disturbancef so wheelhouse th , e floosecurels wa r y sealed. Bonded into the angle of the wall, a thin slab of stone formed a shelf.1 A bone quern-handl polishea d ean d bon wern epi e recovered from this bay wels ,a l as numerous potter notabl s quantite yth wa sherdsr 5 typed fo y ysherdf an e o Ba . s recovered, mainly localise westere th n di n half; household refuse, noticeably deeper precedine thath n i g bay, covere thicda k laye peat-asf ro clayd han . Bone remains were remarkable, not only on account of the quantity but for the prevalence of longbones and jaws of sheep, which were thrust vertically through the occupation level into the sand. A cache of 32 ox teeth was found buried against the footings of Pier 6. In a pi2t dug into the floor, a heap of beach pebbles lay, apparently collecte somr dfo e purpose; these average ounce4 o t weightn d3 i s . Small lumps of prepared green clay, stil conditiont l we plasti lowee e th th f n csri o levels, werf eo even size and free of stones, apparently raw material prepared for some specific use. Bay 6 was notable for a heap of razor shells lying against Pier 7 and under paving stones in the angle of the wall lay a closely packed collection of limpet shells.3 The occupation level in this bay was not so deep as in the two immediately preceding t foobu d refus s abundantwa e trimmeA . d scapula, which could have been used as a scoop for shell-fish, lay in a crevice of the wall. Here it may be noted that oyster, mussel, razor and large whelk shells were found throughout wheelhouse levels. Sherds of an almost complete pot stood mouth downwards against Pier 7 containe7 y PI, d i)Ba X ..an 3 d(fig3 man, 6 . y pottery sherds, and, unde heaa r p of particularly large limpet shells lowee th , r stonshalloa f eo w rotary quers nwa sane founth dn o dbelo livine wth g levePId .an l VIII2 (fig , . 9 Outsid. ,3) e eth bay a crude setting, slightly dished in the centre, could have been used for steadying a pot. In Bay 8, the wheelhouse occupation, sealed below robber levels, yielded many fragment f almoso s t complete pots, suggesting that this potter , like yis eth vessels from Bay 7, contemporary with the abandonment of the house. Sherds were also recovere t wheelhousa d9 froy e 1y m0th Ba Ba ed levelan 9 .y BotBa n hi blown sand had been much disturbed and the main wall was in part removed in the latter bay. A boulder with a worn circular depression, 6 in. in diameter, probably a pivot stone, was built into the surviving wall in the angle formed by Pier n (PI. IX, i). The outer end of this pier included the broken half of a saddle-quern (%• 9, 3)- THE CENTRAL AREA This was sealed by blown sand and the floor was a level, trodden crust, greasy in places, with impacted peat-ash in varying depth. A bronze fragment, square in section, probably a broken ring, lay on the floor as well as a flat ring made of yelloantlea d wan r glass bead. Pottery sherd vern si y fragmentary condition were also foun t flooda r level, but, wit exceptioe h th whal o tw f eno vertebra post-holes Compare a shelf or seat mentioned in the Galson report, p. 191, fig. 5. Appendie 2Se 1 bonesrepor. e a 169 p r th , xI fo ,n .to 3 The localised heaps of shells suggest that they were used for some purpose, possibly to be pounded for backing pottery. 142 PROCEEDINGS OF THE SOCIETY, 1959-60 described below o artefacn , r occupatioo t n materia s recoverewa l d beloe wth trodden surface. centrae Th l hearth, Heart pebblhs iwa , e edge madd d an claf eo y burnt brick red; Hearth 2 had been built in part over it (PL VII, i). The upper hearth, which was lined with flat stones set in clay and also pebble edged, was entirely free of ash. irregulas wa t I formn ri , constrictin narroa o gt w nec fish-taia fannind o kt an t lgou shape, with flat slabs which could have been used for baking. Hearth 3 was eccentric- 1 ally placewese VII L th Hearthf o . t(P Thio d t 2) , 2 sd sstonalss an i owa e lined and pebble edged. Somewhat bag-shaped, it narrowed to the north, ending in a ston t extensionse e e heart encrustes Th . hwa d with peat-ash, which sprea- dir regularly outside the stone edging and contained refuse from iron working.
south-wese th n I fouc ar t r post-holes were uncovere t witdse h wedging stones. 2 In one instance a spade-shaped whale bone, n in. by 8 in. and ij in. thick, ap- parently also used as a wedge, was driven into the sand at an acute angle. Several of the post-holes were originally identified by the clean white sand filling the hole, which examinationn o , , reveale packine dth g stones. These holes wer averagn eo e 3 in. in diameter but the depth could not exactly be ascertained in the wet sand of the subsoil. The function of these slight post-holes was not immediately apparent. The possibility of their having held supports for a verandah type of covering, stretching fro pierse envisagedms th wa , interio e spae th apparentlf s Th .n o rwa o yto grea sucr tfo h slender posts havo t e bee nframewore useth s da k fo completra e roofing. A patc whalf ho e bone found almost flush wit e floorinhth immediateld gan y soute th Hearthf ho o t prove2 d sexcavatio n an ido trimmee th e b do nt vertebra diameten i o whalea . f in . dee9d ^in 2 an r,. p ft iwit centrae hth l bony tissue hollowed vertebre outTh . beddes awa clawedgen dd i yan d with seven large stones fourteed in.9 an ,y b . ranginn in smalle8 . ft gio t fro r. stonesny in mb . 10in , including worn hammer-stones welsmalles a a ,s a l r whale vertebra whic beed hha n trimmed (PL VIII, i). On the opposite side of the hearth, a second vertebra of slightly different proportions . deep diamete n in beei d ,. 2 ha ,ft . in ft id simian r - larl ys als treateowa n firmlclad d i wedge t an dan y se y d with stones, together wit hportioa spinoue th f no s process, whic beed h ha VIII nL saw(P f , n2)of . Buried wit e seconhth d vertebr a hamme s awa r bure madd broth an rf w eo n a tin f o e antler. These carefully constructed post-holes could have held supportn i . in 9 s diameter, and, with the supplementary post-holes already described, may have carried the rafters of a roof, possibly thatched with heather or with the reeds which grow freely in the near-by lochans.3 The clay settings of the vertebrae were con- siderably below the floor and when excavated were practically water-logged. These must have placn i bee t ense befor e constructioth e f Hearthno , 2 d s an i though the top of the whale bone was flush with the living level. The hollowed-out portions were mainly filled with white san showed dan signo dn decayef so d wood, 1 Compar bare eth n heartTighe th t ha Talamhanta farm-hous Allasdalee th t ea , f Islo Barra f e9 o 8 . p , repore PIth d . an tVI.2 . Appendie 2Se . 172p , .xII 3 Post-holes comparably sited were on either side of the lower of two hearths in the aisled farm-house, Tigh Talamhanta, see p. 84, and fig. 3. The diameter of the central court, 22 ft., appears to rule out the possi- bility of a corbelled roof as recorded by Capt. F. W. L. Thomas, R.N., for the aisled house of Uisnich (P.S.A.S., in (1880—i), 124 whicf o ) centrae hth l areonls adiameterywa n i 1. 5ft . A CHEARDACH MHOR, DRIMORE, SOUTH UIST 14.3 suggesting tharooy an tf supports which were originally contained must have been removed whe wheelhouse nth abandoneds ewa e vertebraTh . e were identifies a d representing both the common Rorqual and Sibbald's Rorqual, most likely from stranded animals e carefuTh . l constructio e post-holeth f o n s demonstratee th s ingenious availablf o e us e material. The evidence of the animal bones, sheep, ox, pig, supports the pastoral nature wheelhouse ofth e inhabitants, though pon onls yi y foun post-wheelhousn di e levels. The most interesting of the bone finds is the evidence of a small dog, compared by Dr A. D. Clarke to 'the Highland terrier', from which have descended the Skye, Cair Aberdeed nan n breeds (see Appendi . lyo). p presence , xI deerd Th re f ,eo 1 no longer found in South Uist, may be noted. It was a matter of regret that time and weather did not permit of further in- vestigation on the south-west of the mound, where there was the possibility of locating another structure possibls wa t I e. tha buildina t barn-byre th f go e type existed; on the other hand the absence of any steading would have supported the view that featura thessecone e th ear f o ed clas farm-housef so , with free-standing hign wallso t h se ,ground , suc s thosha e identifie t Clettravala d Nortn i h Uisd an t Tigh Talamhanta, Isle of Barra. THE WHEELHOUSE FINDS ) XI ThL P e d Potteryan 6 d (figsan 5 . It was fortunate that part of the wheelhouse, from which much of the pottery described below was recovered, was undisturbed by later occupants. It is also an important feature in the study of the pottery that the stratification of an earlier and later group within the wheelhouse was well defined. The earlier was recovered from below the living levels of the bays, the later lay immediately on the surface of the floor level, as if abandoned at the onset of some disaster. The pottery from the wheelhouse falls into two main groups. The earliest, sofwhica f to pastes hi plais ha , n inturned rims associated with incised decoration. Sherds Nos. i and 2 are closely paralleled at Foshigarry. These were stratified in the lowest levels of Bays 5 and 7. A sherd, No. 6, with a worn rim, is decorated with herring-bone whicr fo , h many comparison quotede b y sma , mainly small fragments and from similar sites. This ornament is also reported from Shetland.
unusuan A l typ incisef eo d patter2 presens n i wheelhousee th t a t sherdo Tw .n si well prepared paste have curving incised lines with fine 'stitched' infill suggesting cop e leathea th f yo r prototype associatee b rim o n N sca . 9 ,d Nos d witan .h8 this small group, nor are the fragments of sufficient size to indicate more than a globular type of pot. No exact parallel can be suggested. The sherds are from the sand blow, immediately over the floor of the wheelhouse, and from the forecourt. A carelessly executed ladder patter ; ncomparabl7 appear. No n so e materiae b y ma l quoted from Foshigarry, Clettraval and Buaile Risary (see North Uist). No. 19, decor-
1 In this context mention may be made of the second century bronze figurine of a dog closely resembling an Aberdeen terrier, found in Coventina's Well, Carrawburgh, figured by Prof. J. M. G. Toynbee in Art in Roman Britain (1962), . PI62 .
Dun Cuter, Clettraval, Jarlshof, Foshigarry, Dun Nighean. 3 144 PROCEEDINGS OF THE SOCIETY, 1959-60
Fio. 5. Pottery, Phase I (1/3) ated with the simplest form of applied fillet, was found in the clay foundations of the monolith at the entrance. Other early type f decoratioso n includ sherdso etw , wit hraisea d boss Nos. 17 and 18. One of these No. 17, is also stamped with the head of a small shouldered pin and comes from a wheelhouse midden tip, the second lay in a disturbed ashy 1 level with later material. Typologically belonging to the primary horizon, it seems reasonable to include it in this group. Comparable sherds with boss ornament are from Tigh Talamhanta, Foskigarry and Dun ladhard. A small group of rims, Nos. 13-15, can be compared to others from Coll, Eileen Maleit Jarlshofd an (sherds fro broch)e mth r HamiltoM . n figures similar material from sites in Orkney and Shetland. The main characteristic of these sherds is the pinched and thumbed-on clay fillet placed in the angle of the everted rim. Part of another vessel, No. 20 from Bay 5, with clumsily incised linear decora- muca tion s hha , sharper carination materiale tha th rese f nth o t uniqu e th , - eex ample of a rather debased type. 1 For a discussion of pin-stamped sherds and comparisons see the reports on Tigh Talamhanta, p. 92 and Cuier, . 312p . Fro evidence mth pin-stampea f eo d sher Peninsula. e KdB . Ey . froStevenso e R mth r M , n deduced that this decoration persiste Islande eighth-ninte th th n di o st h centuries A.D. (P.P.S., xxi (1955), 282). A CHEARDACH MHOR, DRIMORE, SOUTH UIST 145
33
FIG. 6. Pottery, Phase I (1/4) 146 PROCEEDING E SOCIETYTH F O S , 1959-60 latee th rn I ware Phasf so , sharpleI y everted rims preponderate thoug hshora t thickened rim of unusual type occurs on the restored pot, No. 33. One rim, No. 31, is fluted round the inside as though by two finger-tips (another similarly grooved t illustrated)no s i idiosyncracn a , y also recorde t Foshigarry,da Beag,n Jarlshofd Du an 1 (Phase II). innee A th t occur doubln r po joino t sidi e bode s m seth a sth ri e f yo e angl th f eo on certain sherds, e.g. No. 29, and may be due to the use of a turn-table, a feature also seen on others from Clettraval and Tigh Talanihanta. The angles are, for the most part sharp and the interior smooth, though in the larger sherds, for instance No. 28, a finger-channelled vessel, with applied ornament, the smoothed neck unit terminate interioe th roughla n do n ri y folded join ove ring-builde th r . A numbe potf o r s with everted rim e decoratear s d with finger-channelled arcs between the applied decoration and the neck, Nos. 28-32. This type of ornament, less well executed presens i , t Tigha t Talamhanta. varianA thif o t s decoration occurs in Phase IA (see fig. 10 No. 42). Bases include saucer-shaped units upon which the pot is built up coated both within and without with clay. A large base sherd, No. 41, shows the smoothing or 2 parin lowee th f gro ring-building , whic 34 sign s . slipa hha f nickes sNo o ;i , d with the finger nail on the outside of the foot, an unusual feature in Hebridean pottery, while No. 40 is finger pressed round the outside of the foot. An interesting featur thif eo s later groubase th es p i decorate inside th y n eb do potter'e th s givethumb7 3 pattere . sunflowesa Th No . n no r effect, while Nos5 3 . dividee ar 6 widy 3 d b d e shalloan w intersecting grooves with deep thumb printn si each quarter. No. 35 shows a more angular type of base. Amongst the material in Nationae th l Museu Antiquitiesf mo , base sherds with comparable decoration have come from site Orkneyn i s , Shetland, Caithness e Oute wels th d Inne a ,s an a rl r Isles. The majority of the pottery consists of globular pots with sharply everted rims and various forms of the applied decoration characteristic of the Hebridean wares,
sitew n Orkneyknowfe i s a t a n , notabl e brochyth f Ayr o sd Lingrow an e d an , 3 reported from broch sites in Caithness. The ring-building of many of the larger pots is lumpy on the outside and the build-up is in some cases further masked by applicatioe th thia f no coatin clay f termegthice o o b to ,o kt paste dslipa Th .e varies, som wels ei backee b l preparey dma witd dan h pounded sherds largee Th . r pots show coarse grit backin whicn gi h both rolle shattered dan d quartz occurd san paste th e includes particle mica f skilse o Th l. evince pottinn di remarkabls gi e when e mosth t obvious mean f firino s g woul e providedb sherdw peat.y db fe e ar sA * abrademaie potterth e nfriablth d n i survive s t dconditiont an yha ebu we e dth f so the machair amazingly well. No trace of a kiln was found in the area explored. Hamiltor 1M n notes this flutin pottern go y from ShetlandsOrknee th d yan ; seejarlshqf,. 65 . p . Tighcf 2 Talamhanta, fig. 2 . d 5.1an 3 Foun t Bacda Mhic Connain, Clettraval, Cuier, Foshigarry, ladhard,n Du Kilpheder, Tigh Talamhantad an other sites. growtw lo islann e ho Th 4 d site Soutn si h Uist somn i , e instances obviously plantedmaie th nn i scrus i , b willo alreads ha s wa y been noted t seemI . s possible that with fewer hea stocf d o islan e havky th dma e carried a greater amoun willof o t w presente thath t na . A CHEARDACH MHOR, DRIMORE, SOUTH UIST 147
16
FIG. 7. Objects of bone and antler, Phase I (1/2) Bone) 8 d (figsan 7 . e bonTh eantlerf o tool e sar , whale bon unidentified ean d bone. Many animal bones, roughly split as if for the extraction of marrow, were amongst the food refuse, these doubtless served a turn as a makeshift tool, for bone would be a practical substitut r wooa sparselefo n di y afforested countryside. Unworked antles wa r found in the sandy floor of the bays as if for storage. Amongst the antler finds the wor mentionede nb picky . 21ma ,No , implemenn a , t commonly use Hebridean do n othed an r Norther halcleverle a n f th sitesf o d an y, devised rin whicr gfo paralleo hn l has been found, No. 12. Three implements, Nos. 13, 14 and 23, two of antler and one of bone, with one end trimmed and worn, may be noted; similar artefacts are mentioned in the Kilpheder report, where Mr Lethbridge suggests their use as handles for rotary querns. Whale bons use varioun ewa di s ways, fro e hollowed-oumth t vertebra post-holes of the Central Area to the shaped and handled 'cups', Nos. 19 and 20. The use of these last artefacts, recorded from various sites in the north and west is problematic. A people who made and used quite sophisticated pottery would hardly tolerate such an ungainly drinking vessel, but, filled with animal or 148 PROCEEDING E SOCIETYTH F O S , 1959-60
2O
FIG. 8. Objects of bone, whale bone and antler, Phase I (1/2) A GHEARDAGH MHOR, DRIMORE, SOUTH UIST
FIG. 9. Querns: 1-3, Phase I; 4, Phase II; 5, 6, Phase III (j/8) 150 PROCEEDING SOCIETYE TH F O S , 1959-60 fish fat, these would serve as lamps and could provide a feeble light. Another use has been suggested for these so-called cups, that of a pivot to carry the framework o doorfa t thosbu , e foun Cheardaca t da h Mho fragilro appea to sucr e efo b h a o rt function. Of the pins from the wheelhouse all, with one exception, were of the head- less type, and only one, No. 11, is fashioned with any care, consisting of half a polished head, oval in shape, bearing traces of an iron shank, square in section. While bone needles found in Phase I levels would be useful for thonging, they would be clumsy for any fine material and, in this context, it may be noted that no true spindle whor r fino l e needl s recovereewa d from, this horizon unusuan A . l bone find has been identified by Dr Kenneth Oakley as a flake of fossil bone, perhaps, as Mr Lacaille has suggested, deriving from the gravel beds at Daliburgh, a distance of over 12 miles, where fossil bon alreads eha y been observed. Glass A yellow bea vitreouf do s paste diametern i . e , sancm foune ith s th n dwa ,o n di wheelhouse floor. A similar bead, unstratified, was also recovered. These have many widparallel a typ e s th e eha d distributionsan . Metal e slagTh noted from Heart suggesth3 s iron working, but, apart fro mfraga - mentary bronze ring, No. 16, no metal artefact has survived the permanently wet condition of the wheelhouse levels. Stone ) (fig9 . Two notably flat quern-stones are both of stone typical of the Lewisian Gneiss. Thes similae ear type,n i r with oblique, central perforation ovae Th l. handle hole of the upper stone, No. i, though only partially perforated, appears to be complete. This is from the forecourt, the second, No. 2, possibly a lower stone, is from Bay 8. A broken saddle quern, No. 3, was used in building Pier 11. A pivot stone was found built into the main wall at a point where this had been much robbed. This was made fro mbouldea f irregulao r r shape wit ha shallow , well-made circular hollow, 6\ in. in diameter (see PI. IX, i). Forty-one hammer-stones were recovered from the wheelhouse levels. Some showed signs of use at both ends, and a few over one side. These are fashioned from water-worn stone varyinf so g type rockf so . Five whetstone e alssar o recordedf O . piece8 2 pumicf e so th e fro mounde l level mfroe al th wheelhous e ar mn si th 2 2 , e and mostly found in the bays. Summary The sequence of style in the pottery would account for a long occupation of the wheelhouse developmene Th . maie th nf o tfeature toolee th , d applied banda s ha , surprising continuity, and persists in the later occupations. Characteristics of two phase Jarlshoft sa , both represente sealee th n ddi occupation wheelhouslevele th f so e bays, imply intercourse beyond the Hebrides, or perhaps a common origin. A CHEARDAGH MHOR, DRIMORE, SOUTH UIST 151 PHASE IA Sealed by the blown sand, but about 2 ft. above the level of the wheelhouse floor, rectangulaa r hearth, Heart , stonh4 e line pebbld dan e edged undey la ,collaps e rth e owala f l which, when standing, must hav within eru na distanc. fro ft t m2 i f eo hearthe Th . ,2) whic , , PI3 IX . d h(figs an yielde 2 . d iron slagstratigraphicalls wa , y related to a double wall on the east, well built on the outer face and containing in its core iron slag. e doublTh 1 e wall backed agains poorla t y made lin f singleo e stonework which trace e adjoinine coulb th t n ddi no g trenches linteA . l stone. ft 4 , by i ft. 3 in. and 8 in. at its thickest, probably removed from the adjacent wheelhouse, awkwarn a t a y dla angle ,heart e closth o et h (PI. VII, 4). Successive level peatf so - ash and sand had been tipped beyond the return of the double wall on the north side. Fragmentary pottery sherds of this phase were from a thin and introdden occupation level t penetrat. no Thi d sdi e beyon wheelhouse dth e entrance thers y wa e an r no , evidence of it to the west of Trench F^. Though successive tips suggest more than a temporary habitation, the middens of this phase were singularly lacking in bone or other debris and relatively few sherds were recovered from them. It seems possible that the occupation was of short duration and though, judging from the material recovered, not far removed in time from the earlier phase, the inhabitants were concerned wit wheelhouse hth e merel sourca s ya f buildineo g material havins a , g
8 W4 ^9 r4 SO 44
53 54
FIG. 10. Pottery, Phas , Phas44 ; ; 43 45-57 e, II eIA :42 , Phase III; 58, Phas (1/4V eI )
1 See Appendix II, p. 1 72. 152 PROCEEDINGS OF THE SOCIETY, 1959-60 create ddriea r building site tha surroundine nth g machair. Wallinge coulb t dno wese traceth t n havy wherdo ema t beeei n robbelatee th ry dinhabitantsb e Th . slight evidences of Phase IA walling were in a ruined state when they were in turn buried by the blowing sand. THE PHASE IA FINDS Pottery (fig. 10) This includes nothing distinctive, mainly devolved wheelhousee typeth n i s same paste relatee whicb staga n o hdca t e little later tha primare nth y structure. sherdw fe e s froth f mO this horizon eight joi foro nt uppee mth t wit rpo hpara f o t finger-channellea d patterappliee th d ndan betweedecorationm ri . e 42 nth . No , This compares poorly wit executioe hth archine th f no g pattern wheelhouse th n so e pottery and is nearer to a debased form found at Clettraval and Tigh Talamhanta, both aisled farm-houses. Three base representede sar , includin whice gbeeon s hha n saucer-shapea n builo p u t d unit. Bone and Stone (fig. 13) A bone trimmed to a gouge-like form, No. 47, was recovered from the occupation level, together t antlerwitcu ha , partly strippe cortexe th f do . Three hammer-stone piece pumicon f eo d frosan e ear m this horizon. SECTION D.E.
vwvv^w.NV Hearth 3 v^-x
1 Humus ?A Occupation PHASE! 2 Subhumus -12 Blown sand H 3 Peatash PHASEI 13 Occupation PHASEI 4 Stained sand
FIG. ii PHASI EI blowe Th n sand which covere wheelhousee dth , doubtless shifting wit weathere hth , entirelt no d ydi concea walline wese lth th tn gwheo n fresh arrivals foun usefua t di l quarry r buildinfo f gtheio material e r on temporar n I . y occupations e inth -, habitant thif so s period use primare parda th f o t y structure maie Th .n wheelhouse A CHEARDACH MHOR E-5
PHASES
-— Wheelhouse Phase I ggg Collapsed wall over heart f hIf Phas e IA HI Phase JJ E6 DM Phase ffl •I Phase If I—I PhasI e
HEART H5 /s ^ ^. / / Q.uern 4/
D12
FFFT FIG. 12. Phases II-V 154 PROCEEDING SOCIETYE TH F O S , 1959-60 wall was lowered in Bay 3 to effect an entrance from the west, and stonework was carelessly piled against the upper courses of Piers 3 and 4, using wheelhouse material to form a shelter backing against the blown sand (figs, n and 12). The upper part of the wheelhouse main wall formed one side of the entrance passage, of which the outer wall was built of shoddy Phase II stonework with footings at a higher level than the wheelhouse foundations. This occupation was dated by pottery with characteristic decoration. Sufficient sherd d beesha n preserved unde fallea r n slab for the reconstruction of a well-made pot (fig. 10, 44 and PL X, 2) with applied decoration on the shoulder. The irregular pattern of the ornament compares with materia Scurrival,n l Du froCuier n d mDu an both galleried e Isl dunBarraf th eo n si . In the entrance passage a broken quern-stone was found (fig. 9, 4). Trace irof so n working occurre blowe th n dni sand immediately overlyine gth occupation horizo f thio n s period d thereforan , e very probably relate o thit d s phase. Also from this level were recovere. D . whicg remaine A d th do r a hD f so Clarke1 compare modera sizn si o et n retriever, breea 2 d introduced perhape th y sb newcomers, inhabitant e gallerieth f so d duns. Bones identifie belongins a d a o gt stocky breed of pony, roughly twelve hands in height, came from Phase II levels werg frod , sheepi ean md Ox als p an thos. oV Phasef representedeo o t I sII . PHASE III This period shows little evidence of true structures, but a foot or so below the modern turf level short stretches of curving walls occurred, similar in character, seldod an m more tha courseo ntw s high 12)d (figsan . 3 .Associate d with these walls, midden tips of peat-ash interleaved with dirty sand contained pottery, iron slag,
bone and shell refuse, mainly mussel, scallop and oyster. No built hearth was3 traced, but the only charcoal recovered which retained any form in the wet con- ditionmachaire th f so , fille shalloa d w e curvin holloe bacth th f ko t w a g waln i l Trench F$; the charcoal was identified as willow. A quern-stone was found in this horizon. Structure f thio s s phase were nowher complets ea e shelteth s ea r within wheelhouse th e piers wher primare eth y occupatio seales nwa blowy db n sand. Pottery comparable in type to the rough ware from these levels has been found northern i n Iriss beeha hn d sitesrecoverean d elsewhere from well planned dan
constructed duns5 4 which supports the opinion that the occupation is temporary, perhaps seasonal. Squatters of this phase may well be coeval with the builders of the numerous island duns in South Uist, who must have maintained an agricultural link wit e mainlanhth d base thein do r strongholds e identitTh . f thesyo e casual inhabitants of the pasture lands of the machair may prove to be the herdsmen of the island duns. PHASE III FINDS Pottery (fig. 10) finee Th r ware thif so s phas ring-buile ear fragmentaryd an t . Three sherds (not illustrated) have coarsely incised decoration. Characteristic of Phase III are 1 See Appendix II, p. 172. 2 See Appendix I, p. 170. 3 See Appendix II, p. 172. 4 Ballintoy. b e.g. Dun Cuier, and Dun Scurrival. A CHEARDACH MHOR, DRIMORE, SOUTH UIST 155
t
39 ^g 0 y4 V7 4 IP !4 4 V^TT>6 4 '
FIG. 13. Objects of bone, whale bone and antler: 47, Phase IA; 24-46, Phase III (No. 40 1/5, the rest 1/2)
fragments of coarse cooking pots in harsh, grey paste which tends to crack at the lon ge ring-building jointh f o s , Nos. 45-57 e outeTh . r surfac s roughenei e d dan encrusted with soot, the inside clean. These wares are closely comparable to material from Dun Cuier, where the larger pieces recovered from the peat-ash of the domestic hearth showed an irregularly made, wide mouthed pot, or with short rim fingered i56 PROCEEDING E SOCIETYTH F O S , 1959-60 ont bodye onle oth Th y. vessel fro mCheardaca h Mho whicf ro h sufficient sherds were recovered for restoration shows a weak shoulder and a slightly flaring rim. One rim from this horizon, No. 50, is incurving, but it is too fragmentary to identify as a bowl form. Bone (figs . d 1413an ) Of the ten bone pins recovered, Nos. 24-31, four are nail headed, a type possibly used for moulds in casting bronze pins and found at Dun Cuier, Isle of Barra. Three f simpleo e ar , round headed type wite on ,h characteristic swellingo shanktw f O . lona s needlesg ha 'eye' e e otheron th , , 3 Nos3 , , ver32 . y roughly made coms i , - parabl o needlet e s fro n Cuierd Foshigarry,mDu an also from Ballintoy d othean r northern Irish sites. A hook with perforated head, ingeniously devised from an comparee b y ma thre, o dt antler 39 e. brokeNo , n implements from Foshigarry with similar perforation whalA . e bone , witartefact40 h. roundeNo , d oblonan d g perforations possibls i , ytethera . Whil remaine eth thif so s phas fragmentarye ear , either broken or unfinished, the workmanship indicates a marked advance in the manufacture and use of iron tools. Stone (fig. 9) Hand mills of this phase, Nos. 5 and 6, in common with the other querns from thi discardea s sitd ean d quern built int structure oth Kilphedee th f eo r aisled house, hav obliqun ea e perforation insteae hour-glasth f do s type recorded from many island sites. Twenty-three hammer-stones were recovered from these levels. Onl whete yon - stone is recorded from this horizon. Four lumps of pumice were from the occupation of this phase.
FIG. 14. Antler tool-shaft, Phas I (1/3eII )
Metal (fig. 13) While no metal finds can be associated with Phase III, the broken remains of a crucible, probably of triangular type, can be related to this horizon, No. 46. Residue of bronze working remained in the crucible, and clay and bronze had fused on the lip. This was recovered from an otherwise sterile pit filled with peat-ash dug in Phase III through the Phase I levels. The small fragment of a second crucible comes fro blowe mth n sand abov Phase eth I robbing. eII 45 . No , A CHEARDACH MHOR, DR1MORE, SOUTH UIST 157 PHASV EI e dwellinTh thif go s phas buils holloa ewa n i t wmoundee scoopeth f o t ddou sand . 3) , PI d d 1IX .(figs2an an 3 . A semicircula t wal hu f robber o l d wheelhouse slab t verticallse s y intd an o backing against the sand, showed traces of horizontal upper courses but nowhere more tha remaineo ntw situ.n di This structur somewhaa d eha t dished floo dirtf o r y sand, visible in section as a discoloured band 4 in. thick, but with no tangible trodden surface. One wall of a curving entrance passage remained, doubtless designed to keep out the blowing sand which, with the high winds, is a force to be
4 5 9 4 8 54 2 FIG. 15. Object metalf so , bon , Phasstoned 49 ; e 50-54an , eIV :48 , Phas (1/2eV )
reckoned with in the islands. Of double construction, this wall used stone settings preceedine oth f g phas backinge pars th e a f outee o t Th .r passage walth f d o l eha been mainly demolished t coul bu ,traceentrance e e db th t de buildera Th . s were most probably responsible for the disturbance and partial rebuilding of the wheel- house wall over which the entrance passage ended. To the east of the semicircular structure a T-shaped wall built of vertically set slabs was consolidated with smaller material. In an angle of this wall a stone-lined post-hole measuring 6 in. by 8 in. internall firmls ywa y bedde e santh dn d i , 4)(PI .IX . This characteristic typf eo building occur t Jarlshof,sa wher r HamiltoeM n relate structure sth Earle th yo et Christian period.1 From this horizon little occupation refus recovereds wa e e Th . finds include a finely made bone pin, a trimmed whale bone disc, and a cast bronze Jarlshof, p. So, Passage House I of Phase IV. 1 158 PROCEEDING SOCIETYE TH F O S , 1959-60 pin stainee founth n di d sandy floor level relate T-walle th o dt , (fig 49)d an . .158 4 ,
Beside1 s closely allied material fro parallels mha Irelandn anothepi n si e th , r from Doune Hill, Relugas Findhorne th n o , regioa , n wher interestins i t ei noto gt e eth remains of early settlements. This brings the structure, built into the mound rather , intit tha oa perion no t earliedno r than seventh-eighth century A.D. ,a datin g further supported by the recovery of a sherd (fig. 10, 58), not strictly stratified, but from a disturbed level near the entrance passage to the Phase IV hut. This has . RadforbeeR . nA simila . s identifieda C r importeo M rt y db d wares known mainly from Irish sites. The date ascribed by Mr Radford to this sherd is seventh-eighth century A.D. 2
EVIDENC LATEF EO R OCCUPATION, PHASEV Above the sand blow into which the Phase IV hut had been inserted there was much disturbance. Surviving stretche f straigho s t walling, seldom more thae non course high, occurre thesn di e uppermost level depta . belot sin a onlf r iho o o w y7 the humus (fig. 12). In a thin level of black greasy earth at the base of one of these walls (17 ft. long and 2 ft. 6 in. wide, in places three courses high) sherds of a plain pot were found caked with sooty material in which were embedded the fragmentary remain a doubl f o s e sided composite comb (fig. 15, 51). Fro e sammth e stratum cam iron ea n knif whalea (figd ean .) bon 1550 , e shovel together with other frag- ments of pottery which included a sherd with applied boss (fig. 5, 18) and another (fig. 10, 49) comparable to Phase III cooking pot wares. Considering the incon- gruit f thesyo ee disturbanc sherdth d an s e create successivy db e buildin t thiga s level, material from this horizon cannot be described as stratified. Fragment f workeo s d steatite occurrin e subhumuth n gi s included a part f o lamp of a type ascribed to the Viking period at Jarlshof (fig. 15, 54). Part of a human jaw was recovered from the sand blow at this level. The mound, by now a landmark in the machair, had undoubtedly been used and re-used, and a fragment of late medieval pottery from the ground surface, dated by Mr S. H. Cruden to c. A.D. 1600, indicates the continuing use of this dry site.
• CONCLUSIONS outcome Th thif eo s excavation reveal storysa t continuousno , t witbu ,h significant stage habitationf so . Ther ampls ei e evidenc Phasn edescribini r fo eI g that occupa- tio s pastoralna hazardouo landfalA to . t no lr ligh fo s t craft, though subjeco t t vagarie f win so tided dan , would make occasional deep-sea fishing possible, while proximite th shalloe th o yt w waterSoute th f hso Ford, where ther oystee ear r beds and an abundant supply of cockles, would provide an alternative source of food; shell refus plentifus levell excavationewa e al th t f a slo . Onl primare yth y occupation, however, indicates more than temporary quarters.
1 This pin has been fully discussed in Ant. J., xxxvin (1958), 92. 2 C. A. R. Radford, 'Imported pottery from Tintagel, Cornwall', in Dark Age Britain (1956), 59. A CHEARDACH MHOR, DRIMORE, SOUTH UIST 159 Amon e findth g s from this phase e yelloth , w vitreous beads equate wita h second-century dating, perhaps rathe re extrem lateth n i r e west lonA . g pottery sequence, however, includes simple rims with incised decoration of lattice, infilled ribbon, chevro herring-boned nan wels a ,applies a l d rondels, characteristie th f co earliest wheelhouse pottery. Links with Foshigarryd Jarlshofan I I also poina o t t long tenurwheelhousee th f eo , possibly covering several generations. Similarities between a Cheardach Mhor Phase I and Jarlshof Phase II, though tenuous, are unquestionable. A direct connection implies intrepid seamanship, and, moreover, boat building on a scale for which probably, then as now, the necessary material would be lacking. The building technique of the wheelhouse as well as the equipment accords well with comparabl outstandinn ea sited san g characteristi practicae th o t s ci e us l which whale bon puts ewa . The first phase ends for reasons which can never be known with certainty. These could be climatic, possibly a temporary break in the coastal dune and the resulting necessit changa f yo grazinn ei g lands. Whateve reasoe rth leavinr nfo sitee e gth th , wheelhouse walling was standing when the sands covered it. The occupants may have left precipitately, which the almost complete cooking pot standing mouth downward suggests7 supporty y an t Ba bu n ,swhale'i e useth n di s vertebra post- holes did not remain long enough in situ to rot away, for the holes were filled to approximately floor level with clean sand. The second occupation of the site, Phase IA, cannot be far removed in time from e origina th tha f o t l builders, though f thesi , eknowledgy peoplan d eha f o e primare th y structures surprisins i t i , g thahorizoe th t thif no s occupation doet sno extend beyon wheelhouse dth e forecourt. Their , howeversis shadowa , y tenurf eo site th e with poorly built walls which fell befor sane eth d covered them. Pottery of the third occupation, Phase II, equates with that from the Galleried neighbourine Dunth n si g Isl Barraf eo . For Phase III, comparative material is again quoted from Barra, and from Northern Irish sites. The pottery is not of very high quality and is undecorated. Improvement in iroe th n tools use r shapindfo g bon alss ei o obviousiroe th n d slaan ,g materias a l well as the triangular crucible, with its residue of bronze, suggest more than a purely e e closessiteth pastoraTh .f o t e comparisonus l e witCuiern ar r whics hDu fo ha fifth-seventh-century A.D. datin s beegha n suggested. For Phase IV there is parallel construction in the Far North. By this date there seems more likelihoo direcf do t contact. Fro sixte mth h century A.D. onwards early Christian missionaries wandered the seas to the limits of the known world. East coast and Irish connections, implied by the bronze pin and the sherd of imported ware, to which a seventh-eighth-century date has been assigned, are further links evidence th n i movemenf eo lany sead b t dan . Fro datine mth thesf go e findd an s from that of the closely similar hut construction at Jarlshof, to which an early Christian dats beeha e n assumy givenma e e w ,tha a Cheardact t Phasa V I e h Mhor falls into that period. l6o PROCEEDINGS OF THE SOCIETY, 1959-60 That the mound remained a desirable building site is shown by the material recovered from the greatly disturbed upper levels of Phase V. The knife, the comb and the stone lamp are related to later occupations. The long stretch of single course wallin t higga h level, under whic knif e comd hth an e b were found, coule db the result of Norse penetration. sherA glazef do d pottery fro surfac e moune mth th f eo d seventeentdatee th o dt h century points to activities on the site, probably still of pastoral nature, in late medieval times.1
CATALOGU FINDE TH SF EO
PHASE I THE POTTERY Figs 5 and 6 1 Much abrade sherm ri d d fro ma vesse heaviln i l y gritte pasted dre , incised with lattice patter paneln i s above applied wavy band. . Fro7 y mBa 2 Weak abrade fairln i ym dri goo d reddish paste with coarsely incised lattice pattern. Cf. Foshigany, fig. 24. 23-26; Tiree, see Coll, p. 174, PL 6 and 10. From Bay 5. 3 Fragment in burnt dark reddish ware, thin uneven potting, finger-pressed inside, brushed on outer surface over which is incised what appears to be a pattern of alternate hatched triangles abov abraden ea d applied band . Foshigany,Cf . fig. 24, 13. From Bays. 4 Sher hardn di , rather harsh grey paste, outer surface burnt, decorated with deeply cut incised pattern above a slashed, pushed up band. Cf. Foshigany, fig. 23, 19; a rougher versio fros ni m Tigh Talamhanta, . Fro72 m d , 7 1level an 64 fig , 7 . s disturbed by robbing. fragmento Tw 5 burnn i s t reddish ware with incised pattern, possibl. y3 . similaNo o t r Cf. Foshigany, d figs19, 1 an 2d ,23 12.fig an i4 ; 2 .Tigh Talamhanta,d an 7 7 fig , 8 . Ayre, N.M.A. No. L 1946.69. From Bay 4. 6 Fragment of small pot, the short rim, much abraded in fine red paste decorated with vertical rows of incised herring-bone pattern. Cf. Foshigany figs. 23, 7 and 24, 7 and 11 ;; Cuier,Clettraoel,4 , IX figL P . 12, 108; Nighean, Tiree Coll,e se . ,174 p VII L P , I and X. From Bay 6. 7 Sherd of globular pot in reddish paste, decorated with double lines filled with vertical strokes; cf. Foshigany, fig. 24, 15; Clettraval, PL IX, 2; Buaile Risary, see North Uist, p. 210, second plate. From wheelhouse midden. 8 Fragment in dark paste incised with decoration similar but much finer than No. 7 with additional 'stitched' pattern. From blown sand over floo centraf o r l area. 9 Fragment in dark paste with lightly drawn lines and 'stitched' decoration. From blown sand immediately over wheelhouse floor.
1 Fuller discussio islane th f ndo pottery wil founarticle n b la n 'Thdi n e o e Sequenc PotterIrof e eo nAg y Hebridese inth ' forthcomin Councie th n Britisr gi fo l h Archaeology Research Repor. 6 . No t A CHEARDACH MHOR, DRIMORE, SOUTH UIST l6l 10 Fragment in yellow ware with large grits, smooth outer surface, with lightly scratched, uneven herring-bone pattern. From blown sand over central area. 11 Small sherd in dark brown paste with incised lattice pattern. Cf. Foshigarry, fig. 24, 24; Dun ladhard, fig. 13,3 and 6; Kilpheder, fig. 7, i; Clettraval, PI. IX, 5. From Forecourt. 12 Fragmen gritted re n i td ware with incised lattice pattern. . 7 Fro y mBa 13 Short rim in crumbly red paste, with inner bevel and applied band in angle of neck. r thiFo s grou . Coll,pcf . 174 p PI , I .; Eilean Maleit NorthJarlshof,e d se an 8 Uist,20 . p p. 46, fig. 25, sherds from the broch. From Bay 5. 14 Rim in well prepared paste and smoothed surfaces, with double bevel, the upper slightly concave, the applied band is thumbed into the angle of the neck. Cf. Clettraval , fig , typ59 5 . . ep ib; Tigh Talamhanta, d 2 fig; an , iCno 5 . ca Comhdhalach e se , North Uist, p. 204, second plate. From midden. witm hRi 1 5ridg t loweea r angl bevelf eo toolee th , d ban applies d i nec e th ko dt angl e and slightly slashed with the nail. The paste is reddish, finely gritted and burnt on outer side. From forecourt. 16 Fragment in well prepared buff paste, heavily burnt inside, smoothed outside, decorated with cordo triangulaf no r section. From central area. 17 Sherd decorated with applied boss similar to No. 18. In addition half of a stamped impression made with a small shouldered pin is also present on the same sherd. For analogie Tighe sse TalamhantaCuierd figd an an 7 ., PI fig.X 92 I VII.. p 12d ,Ian 109 and PI. XXVII. From midden. 18 Sherd with applied boss squeeze finger-pressed an n do centren di . Thi founs swa n di disturbea d bas e latelevea th f eo t rla wall (see Phas . 158)p probabl t , eV ,bu y derived from the wheelhouse horizon. Gf. Tigh Talamhanta fig. 8, 75, 76; Foshigarry fig. 24, 21; Dun ladhard fig. 13. 19 Sherd with thin rouleau of applied decoration. From monolith pit. 20 Fragment with sharp carination in brown burnt ware with huge grits, and smooth surface very coarsely incised with an irregular pattern. This is the only sharply carinated vesse a paralle s l ha fro e siter l mth beeno , n found amongst Hebridean material. From Bay 5. 21 Large vessel with everted rim, folded ovesmoothed an r d inside base th ,narros ei w e pastflatd Th an gritteds .ei , with smooth outer surface slightly pinburnd kan n i t places. From wheelhouse midden. 22 Abradeglobulaa f o t witm dri po r h slight fol insiden o d vern i , y fine reddish paste. From Bay 4. 23 Parsmala f witt to lpo h short bevelled rim, slightly concave finn i , e paste, buff inside, black outside, a cordon round the girth is represented by two incised lines filled with diagonal strokes, the lower line finely incised, the upper coarsely hand drawn. From Bay 4. 24 Large fragment with sharply everted bevellefolded an innee dm th dri fla n r po sid e smoothed away e past Th pinkiss .i e d burnhan placesn i t pottine th , s unevengi , thoug surface hth smootheds ei . From wheelhouse midden. 25 Rim of weak shouldered pot with concave bevel, in burnt red paste with large grits, surface abraded. From Bay 7. 26 Short rim with long inner bevel on pot with swelling shoulder, in fine reddish paste with burnt surface. From forecourt. 27 Parsmala f o t l vessel with bevelle pinkisn i m dri h paste with large grits, smoothed l62 PROCEEDINGS OF THE SOCIETY, 1959-60 outside, inside abraded, decorated with a flattened applied plain cordon at maximum girth . JarlshofCf . Clas ; Foshigarry,, figss22 66 .II. d p 2, an 1 fig , 14.23 . From Bay 7. 28 Vessel with bevelle pinkisn i m dri h paste with some large grits inside th , fingers ei - presse d corkdan y looking. Ther folded-ovea s ei walle insidth e r th sd fla n ean i p are uneven in thickness, the surface is smooth above the applied band and very roughened belo burntd wan . Double finger-channelled arcs spring fro appliee mth d band. . 7 Fro y mBa 29 Everted rim with double bevel and folded over inner flap, in hard pink paste with some large grits, rather rough outer surface and much finger-pressed on inner side. The finger-channelling above the applied band shows no distinct pattern but the vessel closely resemble . . 7 Fro 28 y . mBa sNo 30 a Parvesse f o t l with bevelle hardn i m , ri dligh d pastere t , finger-pressed inside, smoothed above and below the applied band. Triple finger-channelled arcs spring fro bande m. th Clettraval, Cf . ; Tigh9 d PI an Talamhanta, . 7 VIII , 5 , . 3 , 86 , fig85 , 8 . . 7 From y Ba 31 Parvessea f o tweln i l l prepared, gritty reddish paste e surfacth , e smoothed above brushed an d belo appliee wth d band finger-pressed an , flutea s ha d dm insideri e Th . beve doubld an l e finger-channelled arcs spring fro carefulle mth y applied bandr Fo . e fluteth d beve Foshigarry,e se l . . Jarlshof,d 4 Fro, 3 fig , y an 66 m .fig3 35 . Ba , p . 25 Another fluted rim, much o abradefragmentarto d an d r illustratiofo y s froi n m Bay 7. 32 Vessel wit abradem hri d away weln ,i l prepare pasted dre , much exterioe burnth n to r witd an h remain folded-ovea f so r flap visiblinsidee th n eo . Double finger-channelled arcs spring fro mcarelessla y applied band. From forecourt. 33 Reconstructed vessel, very irregularly made, the rim is short and upright, swelling inside th showind n ean o g folded fla clayf po ; five ring-building visible e sar th n eo inside, smoothed away on the outside. Thumbed-on ribbon decoration is found at the greatest girth. Found at base of Pier 7. 34 Small base fragment in dark paste showing finger-nail impressions round the foot. From midden. 35 Slightly dished bas n coarseli e y gritted brown ware with decoration apparently similar to that on No. 36 below. From midden. For bases with comparable decora- tion see sherds in the collections of the N.M.A. from Dun Beag, Beveridge Collection, broch of Lingrow, West Howe, Burra, Cumlins, Sae Breck, broch of Keiss, Nibster, Roa Whitd dan e brochs. 36 Very thick base fragment, the underside burnt and scaled away, in coarsely prepared hard paste. The inside is divided into quarters by intersecting, finger-channelled grooves within a concentric groove and has deep thumb-prints, presumably in each quarter. From disturbed levels at the back of the wheelhouse wall. 37 Flat bas shortn ei yelloy dr , w waredecorates i t inside I . th n edo with overlapping, shallow, finger-tip impressions surroundin gcentraa l thumb-print. From midden. 8 Bas3 roughn ei , buff paste with large grits, burnt inside. From forecourt. 39 Base in well prepared pinkish paste, burnt inside. From Bay 6. 40 Base with splayed foot, in fine paste, though with large grits, and finger-pressed. From midden. 41 Dished base, roughly worked brownish paste, surface brushed. The inside shows ring-building. From midden. A CHEARDAGH MHOR, DRIMORE, SOUTH UIST 163 BONE AND ANTLER 8 Figs.and 7 i Roughly made, blunt headed bone pin, the point polished, the other end trimmed. From forecourt. 2, 3 Bone pins from Bay 7 and midden. 4 Blunt headed antler pin, highly polished . Galson,Cf . . 200p , fig d 14, 1.9 3 .an From Bay 4. 5 Needle made from a thin strip of antler with round perforation at the widened head ; though carefully rubbed down, the cortex is still visible. From Bay 7. 6 Short, double pointed bone needle e oblonth , g 'eye s reamei ' botn do h sides. Cf . Broch of , figAy44 re.. 11p , ; Foshigarry, fig. 19, 18. From Forecourt. 7 Bone needle of similar type to No. 6, but smaller and the point at the head is broken. From midden. 8 Antler awl with marks of knife trimming and rounded point. From the central area. 9 Gouge-like implemen antlerf o t poine th , t highly polished. From clay settine th f go 'monolith'. 10 Whale bone peg. From central area. 11 Half of a polished bone pin head with socket for square shank retaining traces of iron. Cf. Foshigarry, fig. 8, 3 and 4, the latter of whale bone has a square shank hole; Broch of Burray, p. 237, fig. 210; see also P.P.S. xxi (1955), 282, R. B. K. Stevenson, 'Pins and the Chronology of the Brochs'. From the forecourt. 12 Half of an antler ring, semicircular in section, cut and polished on the rounded side; the cortex remaining on the flattened surface has been rubbed down. From central area. 13 Wor grooved nan d antler quern-handle. From wheelhouse midden. 14 Bone quern-handle, worn into a biconical shape, one end trimmed, the other end broken. From forecourt. 15 Parbona f o t e toggl r dooeo r sneck with straight perforation found wit hfragmena t of bone with hour-glass perforation . Freswick,Cf . PI. XLIX . . 7 Fro2 y d m, Ba an i 17 Perforated antler tool made from the burr end, broken at junction of second tyne, the brow tyne has been smoothed off, the surface rubbed down. From forecourt. 18 Ox scapula shaped to form a knife-like implement, perhaps used for scooping out shell-fish. From a crevice in the wall of Bay 5. 19 Handled 'cup' or lamp of whale vertebra, with the cancellous tissue hollowed out to a depth of 2 in. Cf. Foshigarry, p. 342, where 6 are recorded; Broch of Burray, p. 237, fig. 209. Fro iy .mBa 20 Handled 'cup' similar to No. 19, the centre scooped out to a depth of i J in. From beside Hearth i. 21 Pick from shed antler trimmed and worn, the haft, from which the cancellous tissue has been removed, may be broken. Cf. Foshigarry, p. 338, fig. 21, 1-3; Galson, fig. 9, 27-28; Bac Mhic Connain, fig. 16, i. From midden. 22 Shaped piece of whale bone, neatly perforated, possibly part of a drinking vessel. From forecourt. 23 Hollowed antler quern-handle, grooved with use . Kilpheder,.Cf . 187p ; Foshigarry, fig. 13; Galson, fig. 9, 2 and Bellochban, see North Uist, p. 230, second plate. From Bays. 164 PROCEEDINGS OF THE SOCIETY, 1959-60
PI. VIII, I and 2 Two whale vertebrae with the processes partly sawn off, hollowed out and set in clay below floor level, apparentl diametern servi o yt . post-holess in ea 2 . ,ft s i . i . 9i ^in No . deep and, includin e remainingth g portio e spinouth f no s process . . 3acrossft ^in 2 , . . diametedeepn i r in identificatio . 2 ft i s Fo . i . ft d i2 . ran speciee No th f Appendie no sse x . 170 p smal , A I . l trimmed whale vertebr diameten i . clain foune s a9 th y rwa n dsettini g of post-hole i. Mot illustrated A disc, the unfused epiphysis of a whale's vertebra, with crude perforation, diameter 3 in., is from Bay 5. Part of a similar disc but unperforated, diameter 6 in., is from the wheelhouse midden. Cf. Foshigarry, fig. 12. Whale bone peg, . 4long!in , showing striations, possibly mad rubbiny b e g with pumice. From central area. Distal end of a femur halved, with shallow hollow 2 cm. deep scooped out of the cancellous tissue. From blown sand over floor of central area. GLASS Not illustrated Yellow bea vitreouf do s paste diametern i ,m ic . Tigh Cf . Talamhantap. 104, Appendix A. From sane th d immediately ovecentrae flooe th rth f o r l area secondA . , similar bead is unstratified. METAL Fig. 7 16 Fragmen bronzf o t e ring, squar sectionn ei . From central area. STONE Fig- 9 1 Flat upper stone of a quern with oblique central perforation and oval handle hole, partially perforated. The stone is from the banded granitic and micaceous horn- blendic gneiss layers, typical of the Lewisian Gneiss. A broken quern-stone built into Kilphederfabrie e th th f co aisled houssimilaa s eha r perforation. Fro forecourte mth . See also PI. VIII, 4. 2 Possibly the lower stone of a quern, also very flat, with oblique central perforation and of similar stone to No. i. From Bay 7. See also PI. VIII, 3. 3 Saddle quern, broken and built into Pier 11. (Reconstructed) , PI.i IX Pivot stone with circular hollow 6| in. in diameter, and a secondary hollow 3^ in. in diameter, and \ in. deep. Cf. Foshigarry, p. 303, Midhowe, p. 500, fig. 43, 3, where mentio mads sockeni 3 2 f eo t stone varyinf so g size; Bac Mhic Connain,. 50 . p Not illustrated 41 hammer-stones bolste8 , r shaped, 10 oval roughl5 , y triangular spherical4 , flat4 , , pea3 r shaped, i almost square broken6 d ,an . stonThesf o e ee ar fro drifte moccurth r o , - ring in dykes in the island, the majority of the Lewisian Gneiss complex. Some are used t bota h end dowd sidee san non . 22 pieces of pumice. A CHEARDACH MHOR, DRIMORE, SOUTH UIST 165
PHASE IA THE POTTERY Fig. 10 42 Ring-built sher reddishn di , well prepared paste burnt outside, decorated with finger- channelled parallel strokes above the applied band. The latter has been thumbed on and nail marks are visible along the upper edge of the pattern. Cf. Clettraval, PI. VIIId Tighan 4 , Talamhanta, . Fro 86 fig , m8 . midden. 43 Large sherd, thin walled, in reddish, finely gritted paste, heavily burnt outside, inside unevenly smoothed. From midden. BONE Pig- 13 47 Bone trimmed to a gouge-like form. From midden. STONE Not illustrated Three hammer-stone piece on pumicef eo d san .
PHASE II THE POTTERY Fig. 10 44 Reconstructed pot, well made goof o , d paste, with smoothedd evertean m dri , inner bevel, interior showing folded over flap of clay. The applied decoration is of irregular pattern same th en i , rather flori t frod po m styla Cuier,s ea fig. 12, 105. From under a large boulder on the occupation level of Phase II structure erected in Bay 3 of the wheelhouse. STONE Pig- 9 4 Part of a quern with hour-glass perforation. From the passage leading into the Phase II structure.
PHASE III POTTERY Fig. 10 45-57 These sherds are mainly from vessels in harsh, grey paste, heavily gritted, the inner surface light brown, the outer 'crepey' and more or less burnt, with sooty material adhering, comparable to cooking pots from Dun Cuier, fig. 7, i-io. In additio illustratee th o nt d sherds fragmentary rim f 1s1o othe rsame potth en si ware were recovered, all from the peat-ash midden tips of this phase. Nos. 45-48, 51, 53, 54 and 56 were from the peat-ash middens in Ci2, Fi, F3, F4 and £4. Nos. 50, 55, 57 were from levels disturbed by the robbing of wheelhouse walls in Phase III, in G8, J8 and Jg. No. 52 was from the sub-humus and No. 49 fro mdisturbea d leve Phasn i l e V(a). ; PROCEEDING E SOCIETYl6TH 6F O S , 1959-60 BON ANTLED EAN R Fig- 13 24 Elegant antler pin, with curved shaft and swelling end not truly a head. Cf. Ballin- derry, fig. 26, 123. From midden in £3. 25 Bone nail-headed pin, with slightly swelling shank. From stones of curving hut wall inFs. 26 Roughly made bone nail-headed pin, with straight shank. From midden in stained sand .65n i , . 27 Well made bone nail-headed pin, highly polished, possibly a pattern for a mould. Cf. Cuier, . 321p , where comparable materia quoteds i l . From same. leve26 s a l 28 Fairly well made, round-headed bone pin with swelling shank, typical of broch and wheelhouse secondary occupations. From midden in F4- 29 Roughly made round headed pin. From same level as 27. 30 Roughly made bone pin with flat expanded head. Cf. Cuier, fig. 14, 26, 27. From midde F$.n ni 31 Coarsely made bon witn epi h squared head . Coll,Cf . plate facin . e Col132gp Th l. pin is perforated and the Uist pin may be unfinished. From same level as No. 30. 32 Long-eyed bone needle. From midden in £5. 33 Roughly made bone needle, with neatly reamed eye. Jarlshof,Cf . . 51p , 4 fig , .28 fro aislee mth d house; Gabon, . 200p , fig , 12.9 ; Cuier, fig. 14, 39; Foshigarry, fig. . 1920 , From midden. 34-36 Awls made from splinters of bone. Cf. Foshigarry, fig. 18 the lower row. From blow< . n san n Gd i Jg gan . madl 3Aw e7 fro mmetatarsaa l bone. From midde stainen i d sand. H5 , 38 Possibly spindl f weleo l polished antler, with poin t eithea t r end, ova sectionn i l . From midde F5n ni - 39 Broken antler hook, square in section, head perforated. Cf. Foshigarry, fig. 15, 2. From same level as 38. 40 Possibl ytethera whaln i , e long. bonein 2 ,2 , with round perforatio diameten i n . iin f r and oblong perforation 3^ in. across; one end is nicked as though to take a cord. From peat-ash midden disturbe rabbia y db t burro found wan littla d e beloe wth modern ground surfac F4n ei _ 41 Broken latc r snecho whaln ki e bone. From midde. F8 n ni 42 Unfinished spindle-whorl, made from hea femurf do . From. samH8 n i e, leve37 s a l 3 Broke4 n spindle whorl made from hea femurf do . From same. leve38 s a l 44 Worn quern-handle made from the metatarsal of an ox. From midden in Fi.
Fig. 14 44A Antler tool-shaft with roughly squared, transvers e face perforationon , n o . in f , tapering to \ in. by -f§ in. at the other. Dr Clarke notes: 'basal 170 mms. antler cut off just below brow tine, which has also been cut off- distal end broken, proximal end trimmed and smoothed', possibly with pumice stone. From the blown sand over the robbed wall of the wheelhouse and probably therefore of Phase III. A CHEARDACH MHOR, DRIMORE, SOUTH UIST 167 illustratedNot Two roughly made bone pins, one with a round the other with a flattened head and the shank broken; two roughed-out bone pins; three broken points of pins; part of an ox femur with stone inserte heae th thougs dprocesa e n d i th n hcrackini f so marrowr fo t gi , the shaft shows hack marks; a handled whale bone club, 10 in. long with blade marks which a beatersuggest s a dolphio e tw ;us s n vertebrae, possibly use s gaminda g pieces. From midden tipblowd san n sand. METAL WORKING Fig. 13 45 Fragment of a circular crucible. From blown sand above wall of wheelhouse robbed in Phase III. 46 Part of a crucible, probably triangular in shape, showing residue of bronze on lip. From Phase III pit. STONE Fig- 9 5 Part of a quern, the handle hole showing hour-glass perforation. The stone is a banded biotitic hornblende feldspar rock suc mas ha y occu basin i r c portione th f so Lewisian Gneiss. From robber levels. 6 Parquera f o t n with hour-glass perforation e ston feldspathia Th s .ei c hornblende- bio title schist or sheared gneiss; this could have originated from the crush belt which occupies the eastern side of the Outer Hebrides. From midden F5. Not illustrated Part of a quern in very fragmentary condition. The stone is similar to that of No. 5 t morbu e fro s basici msimilad a an , r level. 23 hammer-stones of which 6 were oval, 4 bolster-shaped, 2 triangular, 2 spherical, r pear-shaped, i almost square, i flat and 6 were broken. i whetstone and 4 lumps of pumice.
PHASV EI POTTERY Fig. 10 58 Wheelmade sherd of reddish paste, burnt on the outside and with the outer surface regularly rilled. Not strictly stratified but recovered from a disturbed level near the entrance passage to the Phase IV hut. For photograph and analogies see Ant. J. XXXVIH . als(1958) abnorman Cf oa . 92 , l sherd from Tigh Talamhanta,. 88 fig, 8 . BONE Fig- !5 9 Finel4 y made bon n witpi e h squares brokeni p . Jarlshof,ti Cf de . . end126th p , , . Fro3 , figm59 . stained sand floor level near T-wal Phase th hutf V eo lI . Not illustrated Broken disc, 4 in. in diameter with central perforation, made from the trimmed epiphysi whale'a f so s vertebra. From belo stoneT-wale e wPhase th th th f sf o eo l IV hut. l68 PROCEEDING SOCIETYE TH F O S , 1959-60 METAL Pig- 15 48 Cast bronze pin, 5 in. in length with part of a plain bronze wire ring remaining in the perforated head. The unpierced sides of the square head have horizontal grooves at top and bottom of a lozenge shaped panel, the lozenge formed by chamfering the four corners conventioa , n commo groua o n t similaf po thirn i pins sd examplan , e t verno y definitely rendered Ant.e notse a . xxxvJ thir n n eo pi sFo . m (1958). 92 , From close to the T-wall of the Phase IV hut, in stained sand.
PHASE V Among the sherds from these disturbed levels are No. 18, which probably derives from a wheelhouse midden, and No. 49, which should belong to the Phase III occupation. (a) From blackthe horizon below stonesthe Lateof Wall BONE Pig- '5 51 Fragments of a double-sided bone comb perforated for suspension and with the teeth aftet platee cu cross-barth rd san beed sha n assembled decoratioe Th . witns i h incised cross-hatched lines and strokes on end. METAL Pig- 15 50 Iron knife with V section, much oxidised. (b) humus Fromsub the BONE AND ANTLER Fig- '5 52 Bone spindle-whorl decorated with five concentric lines incised on two faces and four lines incised round the girth. Cf. Ballinderry, 2 fig. 22, 627. 53 Bone 'dress-fastener'. Gf. Freswick, PI. XLVIII, 11-14. Not illustrated Antler tool handle, 3 in. long. STONE illustratedNot 1 hammer-ston d i piecean pumicef eo . (c) From peat-ash below subhumus unrelated to any structure STONE Fig- 15 54 Part of a lamp in stone identified as a 'phlogopite-talc-schist', the most probable source being Shetland. Cf. Jarlshof, p. 117, fig. 54, 19. A perforated fragment of similar ston alss oewa recovered from these levels. illustratedNot 2 hammer-stones and i whetstone. A CHEARDACH MHOR, DRIMORE, SOUTH UIST 169 BIBLIOGRAPH SITEF YO S TEXE REFERRETH TN I O DT
Ayre P.S.A.S., XLVIII (1913-14), 31 Bac Mhic Connain P.S.A.S., LXVI (1931-2)2 4 , Ballintoy Irish Naturalists' Journ., (1934)v , 109 Ballinderry 2 P.R.I.A., XLVII (1941-2), i Beag P.S.A.S., LV (1920-1), no Burray . AndersonJ , Scotland Pagann i Times (1883) Clettraval P.P.S., xiv (1948),6 4 Coll Erskine Beveridge, Tiree,Colld an 1903 Cuier P.S.A.S., LXXXIX (1955-6), 290 Foshigany P.S.A.S., LXV (1930-1), 299 Freswick P.S.A.S., LXXIII (1938-9), 7i Gabon P.S.A.S., LVHI (1923-4). l85 ladhard P.S.A.S., XLIX (1914-15)7 5 , Jarlshof Hamilton . C.,R . Excavations,J Jarlshof,t a Shetland, H.M.S.O., 1956 Kilpheder P.P.S., xviii (1952), 176 Lingrow Anderson, J., Scotland in Pagan Times (1883) Midhowe P.S.A.S., LXVIII (1933-4)4 44 , Nighean P.S.A.S., LXXXVI (1951-2), 196 North Uist Erskine Beveridge, North Uist, 1911 Tigh Talamhanta P.S.A.S., LXXXVII (1952-3), 80 Unival P.S.A.S., LXXXII (1947-8), i
APPENDIX I
REPORANIMAE TH N TLO REMAINS Clarke,D. A. byDept.Dr of Natural History, The Royal Scottish Museum
Three animal groups of economic value are represented - mammals, birds and molluscs. Fishes were clearl significanco n f yo onls boneo ea ytw s were recovered. Molluscan shells commoe th f wero l ne al edibl e specie s- scallop , oyster, winkle, razo whelkd an r . Bird bones were few and mainly inedible. Of the six species identified - gannet (?herring-)gull, shag, puffin, swan and raven - all but the raven are aquatic, most are marine and all still commo regione th n least i a , t until very recent times (Harvey-Brown Buckleyd an . A . ,,J VertebrateA , E. . T Fauna of the Outer Hebrides (1888)) e particulaTh . r swant specieno s swa determined e owinlac f comparativth ko o gt e materiae smalth t e singll th bu sizl f eo coracoid recovered suggest Bewick'a s s swan Cygnus bewicki rather tha nwhoopea r swan cygnus.C. Of the mammals, seal, probably grey seal Halichoerus grypus, is represented by a juvenile humao tw humeruleve w d nlo an lfragment a t sa s were recovered from upper levels. Whale occurre a numbe t da f levelso r , mostl s specificallya y indeterminate, rather amorphous chunks of typical spongy bone. These had been broken off large limb bones r vertebrao l centr animalf ao t leas sa lessesiza e f th teo r rorqual Balaenoptera acutorostrata. 170 PROCEEDING SOCIETYE TH F O S , 1959-60 One smaller vertebra might be assigned to the pilot or caa'ing whale Globicephala melaena as being the most likely, though the size and shape are compatible with several other species. The two large vertebrae used as post-holes (see p. 142) represent either the common rorqual, Balaertoptera Sibbald'a physalisr o ) (L s rorqua . musculusB l (L) r both,o . Only thre boneg edo s were recovered l froal , m moderatel levelf w o t they lo e sbu yar interest as representing two distinct sizes of animal. A partial radius, length 160 mm. and distal width 29 mm., together with a nearly complete left ramus of a mandible indicate a dog roughly the size of a modern retriever and larger therefore than the breed represented e almosth y b t complete skeleton recovere t Jarlshoda f (Platt, M. L, Ann. Scot. Mat. Hist. (1933), pp. 17-24); this was recovered from the blown sand over the wheelhouse, ante- dating Phase III partia.A l right maxilla with five teet fros hi mvera y much smalleg rdo comparabl a Skysizn ei o r cairt eeo n terrier. This came fro wheelhouse mth e midden levels. The value of this single item is enhanced by the discovery in the nearby excavations at a Cheardach Beg of a very small radius and ulna and another similar, ulna. The radius, length 86 mm. is little more than half the length of that of the Jarlshof dog (144 mm.) and is small even compared with the length (97 mm.) of the radius of a Skye terrier used for comparison. It is thus clear that at an early date two breeds of dog were on the island, largea r animal, perhap huntingr sfo smala d lan , working terrier comparable with and, surely, ancestra anciene th , lto t broken-haired Highland terrier from which have descended the Skye, cairn and Aberdeen (Scotty) terriers. deed Re r remain se twelv werth f o e n founlevelte n di s distinguished and mostn ,i , bone as well as antler fragments were recognised. The most frequent pieces were tips of tines usually broken off at a notch cut, or sawn, into the outer hard layer of bone but usually showing no other sign of working. In only one case was there clear evidence that the antler had been cut from the skull and not seasonally cast. By far the greatest amount of material was referable to the domesticated animals, pig, horse. , sheeox d pan remaing Pi s wer levele e th t onlrecovere f sbu smalo n y i o tw l quantitiet d frobu l mal s communitvalue s it th d o et an difficuls yi assesso t t . Nontuske th f seo showe beadine dth g usually characterising wild pig and all the remains are assumed to belong to a domestic breed. It is not clear why pigs were kept, for Highland prejudice against eating pork seemvere b yo st ancient. Wit exceptionw h fe remaing pi l sal s recovered were froe mth head regio mayd nan , perhaps, reflect some totemisti r religiouco s significance possibln yi parallel with the ritual underlying the traditional English dish of a boar's head at a major religious festival. Horse remains were numerically less frequent than thos f otheeo r domestic animals excep tproportionatele dogth t ,bu y large numbe intacf o r t bones makes possibl comea - parison with known breeds. Measurement available th f so e bones were compared with corresponding ones from Shetland and Exmoor ponies and a modern Highland riding pony. In all cases the measurements of the South Uist breed fell into the range covered by those of the Exmoor breed and intermediate in size between the Shetland and the modern pony Soute Th . h Uist remains would therefore see represenmo t typicata l stocky Celtic pony of about twelve hands and, no doubt, closely related to the recent Island breed. Sheep remains were very numerou mand san y intact bones, particularly metapodials, have been examined. The mean measurements of length, proximal and distal widths of metacarpals, metatarsal radiud s an give e sar n in Tabl ewitI h mean measurement thesf so e bones from four present-day breeds. A CHEARDACH MHOR, DRIMORE, SOUTH UIST
TABLE I Comparative mean bone dimensions (millimetres') A Cheardach Bone Mhor Shetland Mouflon Soay Blackface
Metacarpal length 114 117 116-5 124 I!5 Proximal width 20 22 19-5 23-5 27 Distal width 22 24 22 24-5 29
I2 Metatarsal length 124-5 125 9'5 '35 116-5 Proximal width 18 20 17-5 20 23 Distal width 21 23 21 23 27
: Radius length 37'5 139 140 143 158 Proximal width 26-5 31 28-5 31 39 Distal width 25 27-5 24-5 30 33
From this tablcleas i t ei r thae Soutth t h Uist sheep differ from Blackfac Soad an e y sheep but are closely similar, in these recorded dimensions, to the Shetland and Mouflon type. Only six intact horn cores were recovered, varying in length of outer curve from 53 to 100 mm. The markedly oval section of the bases of the cores had greater diameters d lessean . r diametermm 0 4 o t fro. respectivel s 0 mfro3 mm m3 shortese 2 th 1 o r 5t yfo t and longest lengths. In size and appearance they closely resembled cores from Shetland sheep and support the conclusions drawn from other skeletal parts. Wit e exceptiohth e exceedinglth f no y numerous second firsan t d phalanges, intact bones of ox were considerably fewer than those of sheep. They were sufficient however to show that the South Uist oxen were of comparable size and proportions to a skeleton oRoyae a Shetlanf th n i l x Scottisdo h Museum, differin horne gfore th onlth f m o n yi cores, som f whico e h resembled e Celtithosth thougx f o co e h measurement f theio s r changing proportions with age revealed them to be fundamentally different. Direct comparison with horn cores of Highland cattle was not possible but it seems likely that the South Uist beasts were related to this ancient breed. Highland cattle show marked sexual dimorphism in the shape of their horns and horn cores, only the females possessing the typical long horn curving outwards and upwards; males have shorter downward tending horns. The bulk of the excavated cores show the forward and downward curvature male oth f e Highlan whilx do esingla e cor characteristicalls ei y curved likfemalee eth . Though the vast majority of bones are from small, lightly built beasts, one, a damaged calcaneum, vers i y muc reflecy h largesizmature a ma f th etd o ran e bull othee th , r remains therefory ma assumee eb largele b o dt y derived from bullocks. Although very tenuous, the evidence may be interpreted as indicating the existence on South Uis t thia t s tim cattlf eo Highland e ol simila e th o dt r Black Cattl r Kyloeeo f so which, in general, only bullocks were slaughtered. Note. completA e catalogu materiale bonee th th f eo d s themselvesan , preservee ar , d Royae inth l Scottish Museum, Edinburgh. 172 PROCEEDINGS OF THE SOCIETY, 1959-60 APPENDIX II REPORT ON THE IRON SLAGS by K.Elliot,W. Geological Survey Greatof Britain e sixteeTh n specimens submitte r examinatiofo d e divideb y dnma into three groups: (i) fused gneiss partly composed of vesicular glass charged with iron ore, (2) specimens composed of glass containing numerous grains or fragments of quartz and feldspar, and (3) slags- somn I e sample glasse sth , containing grain quartf so feldspard zan , occur association si n with slag origie glasse th Th .f nyo specimen uncertains si unlikels i t I . y that they represent fused rocks. Though some of the specimens might represent fused ' flinty crush-rock' or pseudotachylite e absencth , f signo e f crushin o se grainsth n i g, including composite grains, of quartz and feldspar renders this unlikely. It is possible that the specimens may represent fused grittexturae th sandr so d san l evidence supports t possibli this s I . e that a sand or grit might have been used in the construction of a hearth or oven or alternatively as a moulding sand?1
Phase I (1vesiculaA ) r slag compose densa f do e granular aggregat irof wite o nor h interstitial olivine. Locally the olivine occurs as subhedral crystals. From central area, Hearth 3. (2) A slag similar in composition and texture to No. i. From Bay 2, near Hearth 3.
Phase IA (3) A slag containing well developed skeletal crystals of olivine with much iron ore in a dense glassy matrix. The iron ore also occurs as inclusions in the olivine. From core of wall related to Phase I A. (4) A slightly vesicular slag similar in type and texture to No. 9, the more massive slag is margine vesiculay db r sla r glasgo s containing grain r fragmentso quartf o s d zan feldspar. From cor walf eo l relate Phaso dt e. IA (5slaA ) g simila . Frotyp 6 n i rtexturd . eman No Hearto et . h4 Phase II (6) A slightly vesicular slag composed of rather elongated, often skeletal crystals of olivin skeletad ean densa ln i iro ee turbinor d patchily devitrifie dedgee baseth t sA . specimee oth f n contact seee sar n with vesicular material containing grain quartf so z and feldspar in a glassy base which is locally devitrified. From blown sand imme- diately above occupation horizon. PhasI eII vesicula(7A ) r slag compose skeletaf do elongated l an iro e nor d skeletal crystal olivinf so e with an undetermined mineral of lower refractive index interstitially. Some brown glass occurs interstitially. From peat-ash midden. (8vesiculaA ) r slag containing olivin s subhedraa e euhedrao t l l crystals, commonly acicular, with skeleta densa n li iroee devitrifienor d matrix oftes i e iroe nTh nor . concentrate aggregates da irregulan si r patches. From peat-ash midden.
Note specimenn o s s other tha slage nth s have been omitte reasonr dfo economf o s spacen yi e Th . complet1 e report is obtainable at H.M. Geological Survey, 19 Grange Terrace, Edinburgh 9, Ref.: 6i/232/E. DRIMORE, SOUTH UIST 173 (9) A slag composed of elongated skeletal crystals of olivine and skeletal iron ore, perhaps with some dense opaque glass. The slag occurs as a massive patch surrounded by vesicular slag containing numerous grains or fragments of quartz and feldspar. Also enclosed in the vesicular slag are patches of pale coloured glass, of low re- fractive index, containing many grain f quarto s d feldsparan z e feldspaTh . s i r commonly rather altered to glass. From peat-ash midden. (10) A vesicular slag or ferruginous glass altered to a translucent reddish brown oxide of iron. From peat-ash midden.
SocietyThe is indebted Ministry the to of Works for meeting the cost of the illustrations for this paper