II.

A NEOLITHIC DOUBLE- OF THE STALLED TYPE AND LATER STRUCTURES ON THE CALF OF , CHARLEY B . . CALDERT . SS , A.R.I.A.S., F.S.A.ScoT.

The Calf of Eday in the Orkneys is a small uninhabited island separated fro e maimth n islan e water f Edayth do y f Calb o s f Sound. Near Caithnes ss souther Geoit n o s n sidgroua s ei sevef po n prehistoric monumentse yard0 lyinth f 25 g of so t fro 0 m5 shore (fig. 1). Three of them are marked on the O.S .s "Era Map s o a d tw ,e Hoxiseson d an " "Standing Stones," but recent investigation has proved that eace representon h a Stons e Ag e chambered cairn. The largest of these, whose © SITE F O . true characte s previouslwa r y unknowd an n I«ON AGl which coul e identifieb d de standin th onl y yb g STRUCTURE CHAMBERED stones s connectei , e otheth f dro wit struco htw - CAIRN tures by a later wall. Together they constitute the most northerly and most prominent members soo rf.tr e grouoth f p (fig . Writin.2) e yeath rn g i 185 5 H————I————I Farrer recorded this particular comple sita s exa Fig . Cal1 . Edayf o f : having standing stones within the line o Monumentsf fo p anMa y Ke . "Ancient Rampart of stones and earth some- what resemblin e letteformn i th g d havin S e ran "th t eaca gd en h "foundation-stone towers.f o s " 1 1 Proc. Soc. Ani. Scot., vol. ii. p. 166. 116 PROCEEDINGS OF THE SOCIETY, JANUARY 11, 1937.

THE WALL (1). Little chang s takeha e n place since then t whabu , t Farrer calla s rampar s mereli t a ygrass-covere d stony ridge e debriformeth y sb d of a drystone wall. It is so ruinous that the original thickness is not ascertainable e sprea th fee6 t s tdi bu ,wide . Ther e e traceth ar e n o s surface suggesting thae lineth t , whic s definiti h a lengt5 8 r f fo eo h yards s extendeha , d southward e shoreth o t s. Wit e exceptiohth f o n a short length at each of the ends, which return in opposite directions, walthe l runs north-eas south-wesand t northerThe t (fign2). . extremity curve a semi-circl n i s e aroun e south-westerth d n a shallo half o f w

'• STONY ^ AREA C5T.CALDE.R "^^__'X

Pig . CalfofEday2 . : Pla Walf nAssociateo d an l d Structures.

depression, roughl fee 0 diameteryn 2 i t , before losing itsela dilapi n i f - dated stony moundcoursits walIn the e. l surmount othetwo s r mounds, these being presumably the "towers" referred to by Farrer. Its association with the mounds is the principal reason for the inclusio wale th thin i lf no s account moritselfo n i n ridge , f eo th as , s ei importanc edikd tha ol e remain y nth whic an f t o resembleshsi e Th . wal s certainli l y later tha e moundnth s which were already heapf o s ruins when it was built, and its relation to them is purely incidental.

E CAIRTH N (2).

Toward e ridg ereco th e souther f tw th eso t d slabsen n ,d SS2an t (fig stoo, 3) . d paralle eaco lt h othe feer5 inche6 t s apart juse on ,t breaking e othee surfacth th d r risin an ea heigh fee2 o gt inche6 f to t s above ; it both, however, were reduce weatheringy b d . From their position and general appearance it was felt that they did not fit into the categor f Standino y g Stone o whict s h thed beeha yn assigned. This A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 117

DATUM :±^fc*

SEQUENC ^2I F ^ 5O E CONSTRUCTION BEa8888$^] PLAN O 5 IO 15 2O 25 3O 35 4OFEET I I I I I I I I I | I I I I I I I I l|l II I I I I I I | I I I I I I I I I | C.S.T.CALDER

Fig. S. Stalled Cairn, Calf of Eday: Plan and Sections. 8 PROCEEDING11 E SOCIETYTH F O S , JANUAR , 1937Y11 .

idea led to a trial investigation which confirmed the view that they belonge structura o dt e underneath, from whose walls they projectes a d partitions. Subsequent excavations disclose e existencdth longa f eo , neolithic, chambered e stallecairth f o nd variet e lay-ouyth , (fig f whic3) o .t h was identical in many respects with the cairns that have recently been placed on record by Mr Walter G. Grant and Dr J. Graham Callander. In addition to four of the type excavated in Rousay, four

others are1 supposed to exist there, but the cairn now being described firse isth t exampl examinee b o et d outside that islan recenn di t times. Eighty-two years ago, what seem o havt s e bee e stallena cairth f no d type was excavated in the Holm of Papa .2 Our monument, however, is distinguished in its internal arrangement from the Rousay group by the occurrence of a most unusual combination of o differentltw y designed , lyinchambersB n juxtapositioni d g an A , ; one is long and high and the other round and low, but both contain stall-like compartments. a singlWhethe t no e r chambeo r r was contemplated originally is a point that cannot be answered satisfactorily s certaii t i nt bu thae ,complete th t d design allowed for the incorporation of both. This is made clear by a glance at e finisheth d outline, wher e sweeth e f eaco p h side wall a run n i s sweet unbroke e othee cairth th n o r f t n o curv wit d heen froe mon an obvious allowance for the presence of -the second chamber. This contention is further strengthened by the fact that no outer wall s beeha n buil s para f teithe o t r chambee linf th contaco e n o r t between them Othe• . r cairns containin r moro o e tw gchamber e ar s t unknownno thin i st casebu , , -wher o distinctw e t type f chambero s s can be assigned to the same period, the chronological value of such a relationship is important. Three successive stage f operationo s e apparentar s t whatevebu , r lengt timf ho e elapsed betwee t lonno n g s them,enougwa t i o brint h g the last stage beyond the limit of the late Stone Age in . The lon e gfirs th chamber te builts b par wa o t , ;t A chambe, , whicB r h abutted agains s e massivbacit tth e next kd th walls ean ; outewa , r wall encasing both chambers was the last. The addition of the latter create e double-walth d l feature whic s beeha h n observen i d e Rousath y n othei cairn d r an stype s elsewhere e inneTh .r -walf o l chambe , measurinA r g fro m4 fee o 3te t sideJd th thic an t s a k

1 Proc. Soc. Ant. Scot., vol. Ixviii. pp. 320-50; vol. Ixix. pp. 325-51; vol. Ixx. p. 407; another cairn of the type (Blackhammer, Rousay) will be recorded later in this volume. 2 Ibid., vol. ii. p. 62. A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 119 about 1 foot more at the inner end, was built of drystone masonry faced on both sides with horizontal courses of rubble. That of B, which was mainly 4 feet thick, was of similar construction but of inferior workmanship. This was most noticeable in the facing of the passage and in the very rough rubble of the outside face where s exposed wa e positio t e outei Th th . f re o ninne th face f ro s walls wa s though verifiewa s a t dr necessaryfa onl o s y e partth , s which wer t examineno e d being assume d indicate a e an plad th y b nn o d

Mg. 4. Stalled Cairn, Calf of Eday: south-east corner of Outer Wall.

broken line. What remaine e outeth f ro d casing wall varien i d thickness fro e mwesterth fee0 about e 1 sidea t th n o fee7 t st a t end. t I showe d well-built horizontal stonewor e facth e n withouo k t any projecting footings (fig. 4). The stones of the lowest course appeared to have been selected for their larger size and squareness e anquointh d s seeme o havt d e been finely adjuste o e suit dth t convexit e sided th ends t an somf sA o y. e place a battes r was evident, but at others the wall-face was vertical, a circumstance whice north o a pointhrustth t t hn e a ,O t migh du .almos e b t n i t line with the outside of the back •wall of chamber A, a slight change in the character of the masonry was noticed. The stonework, however s welwa ,l e facbondeth e d showean d o morn d e signf o s 120 PROCEEDING E SOCIETYTH F O S , JANUARY 11, 1937.

interruption than would be caused by a temporary stoppage in its erection. caire Th n measure fee6 fees6 7 t2 t lon wided wallinge an th ; g survives to a general height of 2 feet on the outside and to 4 feet in places within e lonth g chamber accounn O s followinit . f o t e groune slopth gth f eo d one end is 6 feet higher in level than the other. The main axis lies north-eas d south-westan t , and, unless thera specia e b e l significance in this particular orientation, it is difficult to explain why the cairn should have been built lengthwise with the slope when, by building across it, a more level foundation would have been obtained. The entrance-passage of chamber A, measuring 11 feet long and 2 feet 2 inches wide, is situated in the middle of the higher end, but it is now without lintels and is much destroyed. There is no evidence eved thaha r t i beet n seale threr . o Onl dup o e ytw course masonrf so y remai e nortn partni th f ho s side, whil large on e e solitary l blocal s ki that is left on the south. The inner end of the passage passed through a gap between the inner edges of a pair of upright stones erected opposite each other to form the eastern end of the long chamber, but the southern slab is now missing. The chamber (fig. 5) extends almost half-way through the length of the monument; it measures 23 feet long and from fee7 o t5t wide, whil s originaeit l heigh beed ha tn considerably greater than that of the tallest upright stone still existing. This attains a height of 7 feet measuring from the floor to its -weathered top, and it is

identifiable with the taller of the two stones, Sj and S2 (fig. 3), which alone marke positioe dth structure th f no e befor excavationr eou s began. A built wall, with an inward overhang of 10 inches in its present heigh fee3 inche6 f t o t s terminate e chamberth f e o innee Th sd th . en r latter is divided somewhat irregularly into four communicating compart- ment y threb s e more pairp opposit u f slab o st se s e each othere Th . compartments 'range from 4 feet 6 inches to 6 feet long and from 5 feet o about fee7 t t wide. Each upright sla s fixebi d edgewise intod an , projects from, the side walls in a manner reminiscent of a row of stalls on either side of a central gangway or corridor. The intervals between e inneth r edge f eacso h pai f stoneo r s vary fro feem2 inche9 t fee3 o tst 6 inches. Ther vera s ei y slight inward corbellin e uppeth n gi r courses of the side walls. All the upright slabs are broken on the top. They range fro minche3 foo 1 inche5 o tst thicknessn si lattee th , r size being that of the tallest stone. Against each side of the partition slabs, small stones set on edge, the highest rising 1 foot 4 inches above the floor, have serve t onl dno s packin ya r rigidits supportga fo t ybu r ston fo s e shelves singlA . e shelfs assumedi t i , havy ma ,e been provide r eacdfo h A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 121 stall sinc e presencth e inneo tw f suc s certairo ee compartth hwa n ni - ments brokeA . n slab whic d beeha h n a useshels foun s a dwa fn i d position, though e nortlooseth hn i , stal f compartmeno l d an , 3 . No t leanin edgn go e agains anothers wa t i t , brokepieceso tw n ni . This, pre- sumably, had been the slab originally in the corresponding stall on the south side. In compartment No. 4, each of the stalls contained a shelf, 2 feet wide and 1-J inch thick, one end being built into the back wall, and the other supported by a slab on edge close to the upright. Both

Fig. 5. Stalled Cairn, Calf of Eday: Interior of Chamber A, looking west. shelves were in such a shattered condition that they could not be preserve n situi d after clearance t righA . t e bacangleth ko t swall , betwee e shelves e th enda thi t th i nf s edgen o so t wa , sla t bu , se b had been dislodged. It may have been a low divisional stone similar to one, underneath the shelf in the north stall, which divided the space intirregulao otw r partseastere Th . n part, marke plann o ,Y d was y boxeanotheb n i d r slab covering hale front th e fspac th ,o s e enclosed measuring 1 foot 10 inches long, 1 foot 8 inches wide, and 1 foot 4 inche a smals floo sit y highla n lr O quantit. f brokeyo d badlnan y decomposed bones of a human being and of an otter, which are reporte n later o dslaA .b construction , measurinZ , 2 feeg t long, 1 foo0 inche1 t s highd fro an m,9 inche 1 foo inche5 o tt s s wide, 122 PROCEEDINGS OF THE SOCIETY, JANUARY 11, 1937.

forms, along with the remaining upright at the entrance and the side wall of the compartment, a sort of recess which contained shell and bone refuse mixed with peat ash; opposite, on the south side, there were slight indications of a similar structural feature. Their purpose t apparent no e chamberher s th e floo f wa eTh o r . s wela ,s thaa lf o t the passage, sloped in keeping with the foundations, and from the entrance to the back wall the declination amounted to 2 feet 6 inches. The floor consisted of a layer of blue clay lying on the natural buff- coloured clay subsoil. It may be mentioned that the whole foundation e cairoth f n seeme blue th res o edt n layero t e materiath , f whicho l t i , is understood, does not occur nearer than the south end of Eday. In this connexion e cite,b tooa dnoticy a chambere ma f , o e d cairt a n Clady Halliday, Tyrone, wherein it is stated that "Underneath the

1 whole monument was a ritual clay EJ3 BLOWN SAND floor which had been laid after erectioe th northostate 6th f d san peristalith." A latea t r period, whee th n chamber had become roofless and ruinous, the original clay Fig. 6. Stalled Cairn, Calf of Eday: Section, on line A—B, of Chamber A, snowing Occupation Level. s floocovere a wa fillinr y b dg of blown sand. Thiwedges swa - shaped in section, increasing in depth from zero at the entrance to 2 fee e inneinche6 tth t ra s end, whil t i emaintaine a dfairl y level surface throughout (fig. 6). On top of this there was an accumula- tio f refusno inchee4 froo t generan msi 3 l thickness t witbu , h pockets o mucs foof1 a pea s h tinmose ha dee as th t n pi t compartment. This layer represented an occupation floor probably of the Early Iron Age. As well as the peat ash it contained, amongst peaty humus, broken pottery, animal and bird bones, and a large quantity of shells, mostly o f s coverelimpets a wa growt l y b dAl f pea d o heather.h an t d an , throughout the respective layers there were many fallen stones. The smaller chamber, B, which measures 10 feet long by almost 7 feet wide, was set with its main axis at an angle of 45° to the back wall of the larger, A, immediately adjoining the latter in such a manner as to leave the entrance just clear of its south-west corner e purposTh . f thio e s singular arrangement might have been concealment, whic s successfullwa h y accomplishe y placinb d e th g entrance in a position most unlikely to be looked for or found. Indeed, even in an exposed condition during the excavations, its 1 Congress of Archaeological Societies. Report for 1935, p. 39. A NEOLITHIC DOUBLE-CHAMBERED3 CAIRN12 .

situatioo unexpectes s e laswa th n t s de featur b wa tha t o i t te discovered thad an ,t only afte a closr e search e entrancTh . e passage, which is somewhat irregular in shape and wider than normal, was deliberatel s e whola lengtit well-laith y f y r o b ehfo close dp u d infilling of stones, with some especially large blocks placed transversely outside its mouth. Moreover, it never extended outwards beyond the e entirth inne t et onl bu chamber no t wally i d furthes an , wa r r masked by the thick outer casing wall (fig. 7). The filling in of the passage

Pig. 7. Stalled Cairn, Calf of Bday: Outer Wall, and Chamber B, looking through Entrance-Passage. y havma e precede y somb d e e outetim e th e layincor th ef th r o ef go wall, but under the circumstances it is more likely that both were done durin e samgth e operation. This entranc fee 5 lengtn i s te i frod han m 3 to 4 feet in breadth, but, owing to the disappearance of the roofing slabs, its height is uncertain. It may be judged, however, to have bee t a leasn 2 feet t 8j inches, e heighn whicuprigha th f s i o ht t slab leaning apparentl situn yi agains e right-han th e wal n th t o l d side nea mouthe correspondiny th r An . g slab havthay ma et existee th n do other sids beeha e n removed e passagTh . e continue e reath f o r o t s a centrachambe s a , l B rcorrido r throug o compartmenttw h s with stalls on either side. The stalls are contained between three pairs of uprigh thoss ta e sam slaby chamben th ei en wa si , witA r h intervals 124 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

of 1 foot 9 inches to 2 feet 3 inches between their inner edges. The middle pair are free-standing, with packing or shelf-supporting stones at the base, the two next the entrance complete that end of the chamber, but those at the other end are set against the face of the back wall. These uprights have little floor- or wall-hold and they have depende r steadinesfo de e roofing-slabsweighth th f n o o st , whic o longen h r exist. Eac ha recesse s stalha l d concave back wall in contrast with those in the other chamber which are straight.

Fig . Stalle8 . d Cairn, Cal Edayf o f : Interio Chambef o r r B, showing north Stalls.

n respecI f heigho t e differencth t s i more e pronounced e slabth , s in chamber B being only from 2 feet 2 inches to 2 feet 9^- inches high. e northerOnlth n i y n stale floo f compartmenth o lr s beeha 1 n . No t left undisturbed, and it consists of a single slab or shelf, 2 inches thick, restinclathe y doubton gNo (fig othe the ,.8). r stallbeehad sn furnished with similar slabs. Behin e mid-partitioth d e soutth n hno ther indicatione ar e s that tha e wall-fact th par f beed o t ha en rebuilt, and on that side also, in the inner stall, an intrusive upright slab leans agains e wall t th formtI . structurao n s e lorigina parth f o t l word kan may belong to the later reconstruction, though, alternatively, it may only be a fallen stone. e chambe e flooth inche0 slopa f Th 1 o r s f eo ha rs from fron baco t t k A NEOLITHIC DOUBLE-CHAMBERED5 CAIRN12 .

s composei d an f clao d y lik e lon th tha gf o tchamber e heighTh . t from floor to roof may be reckoned as having been at least equal to that of the tallest upright, 2 feet 9^ inches, but the height may have been increase a littl o t de e thicknesoveth fee3 n eke-stony ra b tf o s n o e tope th . Suc expedienn ha t woul necessare db y wit shortee hth r uprights, y case e chambean but th , n i , s beeha r n exceedingl comparison i w ylo n with chamber A. This lowness, along with other characteristics, brings this particular structure into close affinity with the two other cairns maie oth (figp f ne chamber ma grou. 1)Th y e . ke th p f e noteo sth n do latter, however, are partially sunk in living rock below the surface of the ground. An infiltratio f sanpresens no dwa t throughou e oute e corth th f tr eo wall of the chamber, but the absence of blown sand in the chamber itself seeme indicato t d e tha a low-set t s stilroowa lf intact whee nth sand-drift occurred. The interior was filled with black peaty earth mixed wit a hquantit f peao y t ash, suggestin a glate r occupation. Outside the north wall of compartment No. 2 three slabs, W, set loosely d reachinan d oen ng e cladowth y o nt bed , were arrange n box-likdi e fashion wit e roughth h wall-fac e chambeth f o e r closin e ende Th gon . slab scontacn i wer t eno t with eac gape h th othe sd betweean r n them, as well as the space inside, were completely filled with the stones from the core of the outer wall, a fact which made it appear that the forma- tiomerels nwa y fortuitous doubo N . t s alsthie casoth swa e wito htw or three single slabs leaning against the outer face of the inner wall, where they had probably been dumped in reserve during operations and afterwards filline lefpars th a tsavf o gt o ttrouble eth f theieo r removal. The yield of neolithic relics from the cairn, though falling short of expectations e triaraiseth ln do digging s considerablewa , . Portionf o s at least thirty-four vessels of clay have been identified by Dr Callander, whs describeoha e potterdth y from thifrod e othean smth r structures f thio sd paper en s collecte e wa th . t i t Nearla df o durin l al ye th g preliminary excavation from the original floor of compartment No. 2 of chamber A, in the small area within the dotted lines marked X on plan; the other compartments, cleared later, were disappointingly barren. It seems that the whole lot had been gathered together and thrown a heap n i ,s foun wherwa dt i e belo e sanwth d amongst black, greasy, peaty humu d peaan s t ash. Along wit e potterth h y were founo tw d leaf-shaped arrow-heads brokene possibla on , d an , e thir damagede don , a short knife o scraperstw , d severaan , l unworked flakes l madal , f o e flin e presenct Th (fige mixtur th . f fir .9) o en s attestei e e ewa th y db peat' ash, carbon, small pieces of calcined bones, and the calcination 126 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

of all the flints, while soot and ash adhered to the pottery. Two axes of sandstone were found on the shelf above the skeletal remains in

. FigStalle9 . d Cairn, Cal f Edayo f : Flin , Scrapers8 t d Implements fros an i , 7 m7 ; , 3 , 1 . occupation level; 2, Knife; 4 and 5, Arrow-heads; 6, a Flake, (y.) compartmen e large11)d Th an .4 (figs r 0 . 1 measure. No t d 6^ inches long and the smaller only 2f inches long. Mr Walter G. Grant has

128 PROCEEDINGS OF THE SOCIETY, JANUARY 11, 1937.

thate m tilp pointe u ,lo t now t associatioe dou th , f stonno e axes with Scottish neolithic burial structures is of rare occurrence. Only two instance beed ha sn recorded beine ax ge foun on ,hornea n i d d cairt na Lower Dounreay, Caithness, and the other in a cairn at Clachaig, Arran. To e thesadde b o other o et tw d havsw whicno e r GranM h t found recentl n cairni y t Blackhammea s t Lairoa d an ,r Rousay moro tw ,e recovered from cairns which I have excavated this year—one at Hunters- quoy, Eday d anothean , r near Caithnes se Cal f Edayth Geoo f n o s — juso tw t describede th d beino lase an tw Th t g. fro ma singl e cairn, totae th l numbe localitief o r sucf so h finds therefore amount seveno st . A small flint scraper of indeterminate period was found in the thicknes e broken-dowth f o s e nsout th wal n he entranceo l th sid f o e - passage. From the domestic level the yield included, from compartment No. 3, shards of a decorated pottery vessel, and, from No. 4, fragments of plain pottery, a calcined flint scraper (fig. 9, No. 7) and a broken implemen f splio t t bonee latteTh .r measured 3-J inchet si lond an g was smoothed and rounded at one end. There were also skeletal remains of animals—ox and sheep—and of birds—cormorant, shag, sea-eagle, black-headed gullherring-gulld an , bonee Th .s have been examiney db Miss Platt, whose repor appendeds i t . addition I e abovth o net relics from chambedeero tw fina - f , o dA r horn tines and some animal bones must be included tentatively as being within the domestic period. These were found by the proprietor, Major Harr . HebdenH y d previousl,ha M.C.o e spacg dowwh th , du yn e i n

betwee e exposenth d uprights inside ,Slin e th d fron Sn 2f i th ,an xe o e f o t wall-face which at this part had been torn out. The actual depth of the find was not noted precisely at the time, but the shell-layer had been reached. One tine, measuring 8f inches long, still remains in possession of Major Hebde t Carricna k othere Houseth d , alonan , g wit bonese hth , •was presented by him to the Museum. The objects of the latter level belong to the Early Iron Age, which may e perioe comparativelth th als f e odo b relicw yfe s foun chamben i d r B. These consiste f smalo d l bit f calcineo s d bones, three flake f flinto s , two of which were calcined, and broken pieces of thin-walled plain pottery. The pottery fragments were found throughout the debris e floodow th f th' o ro et n chambe d alsan or amongst the stones which fille e passageth d . Piece f eighso t different vessels were noted. Outside the northmost corner of the cairn there were found pieces of plain pottery, and from other parts round about, a rough stone, measuring 9^ inches by 6 inches by 3^ inches, with a square dish-like A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 129

cavity pecked out on one side, two small angular stones, each with a pecked-out cup-like hollow on one face, and a few broken rude stone implements. These may also date from the late occupation.

STRUCTURE 3. The third structure of the complex lies 33 yards from the last, under- neath the north-eastern return angle of the ruined dike first described. Prior to investigation only the top of an upright stone, A, was to be seen. It measured 1 foot 10 inches wide by 5 inches thick, and stood 2 fee inche6 t s abov e dike e surfacth line norte eth n f i th th ,eo f o ho t e grass-coverew lo a f o a hollop to d we mound th scoope f o t . dou Lac k of time account r onlsfo ypartiaa l excavation being made t evebu , n this extende e digginth f o t o dg hale areth f a before anything like a coherent scheme coul e madb d e out (fig. 12). The mound had accumulated ove n e ruina th rf o s oval building which when complete UNEXCAVATEO AREA

might have measured approxi- UPRIGHT SLAB STUMP SLABF O S S fee8 3 n crosi matelt- y b di s 1 4 y SLABS ON EDGE MLARTH RED CLAY s coula ameters r made db fa s eA . PIT G KNOCKING STONE , e chambeouton , r onls conwa y - H BUFF CLAY / tained within the building. The outer face of a wall, discovered just unde e d turfbeeth ha r ,n reduce s lowesit o dt t coursesd an , Fig. 12. Structure 3, Calf of Eday: Plan. measured 1 foot 8 inches in height above the natural clay. It followed the western curve of the oval outline for fully half its circumference, beyond which it could not be traced. The inner face of this wall was much broken down and parts were awanting. At the highest it stood only 3 feet 6 inches above a floo s uncoveredra whichr fa o s , , consiste f well-fittino d g paving-slabs clae th lai y n founs subsoildo wa d buils t I tha .slae wa t th tedgebA - wise into and flush with the wall and that it rose to a height of 5 feet inche3 s abov dee o irregulafloore d eth pan Tw . r recesses , Rd R2an ,x openine chambeth f of g r were constructe e maith n i d wall, -which varied from 7 feet and 8 feet 3 inches in thickness behind them, to 14 fee inchee 6 solit th dn i s between e stumpthio Th tw n . f slabso s , B, remained in the western recess but appeared to have stood clear of VOL. LXXI. ° 9 130 PROCEEDING SE SOCIETYOTH F , JANUAR , 1937Y11 .

the wall-face, which here was totally destroyed. The northmost slab had been kept erect, partly at least, by a packing-stone fixed on edge on either side. The surviving west side of recess R was extended

1 foot 6 inches by a thin slab placed on edge and measurinx g 1 foot 9 inches high d anothean , r measurin feeg4 inche3 t s long foo1 , t 10^ inches high and 2^ inches thick was placed at right angles along the front. Three thin slabs als edgen oo measurind an , g fro mfee2 inche9 t s feealond 4 frone en o tt glongth recesf o t o t d , slineen wern o Rt sese

conformin2 g to three sides of an octagon. None of the slabs had a secure floor-hold and the latter three leaned over precariously; all are indicated on the plan by the letter C. On the floor at D a slab had suffered badly through use as a hearth. It was much shattered and fire-fractured, but it did not appear to be different from the rest of the pavement nor havo t e been laid speciall fireplacea s ya . Impingein s northerit n go n a mas edg s f reddiso swa e h clay, B, coverin n arega f somo a fee3 e t diameten i overflowind an r g fro msmala l box-like arrangement formed of four, thin, loose stone e floors th roughl n . o Thred ybe lai n ed o fee t wes thisf o t smala , l pit , measurinF , feeg2 t foolon1 inche2 y tg b s wide incheothere 0 1 th beed . d ha sat , n an soute sunth deptd a t hen o ka t h stone-buila ofood 1 f inchefloorha e 4 t t th I n .s ti pavinglinina d gan - a cover slar fo b , whil a loose e flat stone leaned agains d almosan t t covered its south end. The pit (fig. 13), "which had a layer of fine buff- coloure de bottom th cla s haln i y wa f, fulf wate o a ld ther s an r wa e small heap of pebbles in each of its two western corners. Before the mid-frontal sla f recesbo s R mortara ther s ovaln ewa , whica G ,d - hha

shaped hollow measurin2 g 10 inches by 8 inches by 4 inches deep. In the recess itself was a thin layer of local buff-coloured clay, H. Both morta d claan ry appeare separatee b o dt d fro e pavemth d a floo y b r thin layer of black peaty soil. It may be mentioned that the recess was also paved with stones very closely fitted together. e interioe chambeTh th f o r r e fireplacnortth f s filleho wa ed with an accumulation of black peaty earth distributed through a mass of debris, but the area on the west, particularly in the recess, was solid with peat ash containing innumerable fragments of plain flat-bottomed potter f redyo , buffd blacan , k e colour'higheth n i r; leve e boneth l s of a large animal were found. The amount and variety of pottery from so small a space seems to me to be satisfactorily explained only by concluding that the building had been or had become a potter's work- sho thad e fragmentan pth t e heath pn i srepresented discarded pots. Thus also would the presence of different clays be accounted for. Further, possibls i beed t i ha n t eusereceptacla pi tha s e da washine th tth r efo g A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 131

of the clay, the coarse raw material being rubbed against the flat stone at its south end and the pebbles being freed and pushed aside. The pottery would hava s ea beee hearthe us nth e firbakee n Th o eth . n di potter's workshop, however y havma e, bee a nlat e adaptatioe th f o n chamber, and a few of the details would seem to support this idea. For instance t constructe e heartno th ,s wa h d speciall a fireplace s ya d an , alse wall-facoth f receso e s R2 appeare o havt d e been tort befornou e

Pig. 13, Structure 3, Calf of E day: Interior of Chamber showing Pit, Mortar, and western Rec 333.

the peat ash and pottery had accumulated and the stumps of the uprights, beed Bha ,n covere drubbishe oveth y b r . The ravages of time, as well as the damage caused by haphazard digging, has apparently robbed the mound of so much of its character that even after this latest investigation ther feelina s ei f uncertaintgo y which may be dispelled only by a complete exposure of all the features. The evidence establishing its domestic nature is seen in the numerous stone implements 1abou d foun buildingan e n th tdi , whic suggestive har e of the Early Iron Age. addition I mortae th o nt r these includ following:e eth —

Two small boat-shaped saddle-querns whicf slightls i ,o e hon y broken. r typesFo som f 1 o e implemente ofth Proc.e sse Soc. Ant. Scot., vol . 104p . , flg, Ixvii, , fig13 .97 . 18 . .p p. 106, flg. 21, and p. 107, fig. 22. 132 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

Eac s roundei e hunde th n ro d side concavely peckinb th e d an g y rubbed face shows small pit-marks. The stones measure 12^ inches by 5 inches by 2f inches, and 9^ inches by 4f inches by 2 inches. An exceptionally large rude stone implement whic founs hwa d with a few pieces of pottery in the pit (fig. 14). It weighed 10 Ibs. and

j Inches. . i . StructurPig14 . , Cale 3 Edayf o f : Stone Implement. measure inche 3 inched8 1 y 2y b sf sb inches e stonedgee th Th f eso . were flaked. Half of a flat-sided heart-shaped stone broken through the middle obiconicaa f l perforation which measures 2^ inche diameten i s e th n o r outside and 1-J- inch in diameter in the middle. The object measures 6 inches across the break, 5 inches along each side, and it is 2 inches thick. Three club-like implements with their sides convergin o bluntt g - pointed butts and all shaped by chipping, measuring 11^ inches by A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 133

3^ inches by 2 inches, 10 inches by 4 inches by 1-|- inch, and 8^ inches by 3f inches by 1^- inch. Another club-like implemen f ovao t l section, tapering more sharply to a point at one end than the other so as to give a grip for the hand. e surfacTh s finishei e d with small pit-mark e stonth ed measurean s s 9^ inches by If inch by If inch; five broken tools of similar type. Two flat triangular objects with chipped bases; each measures 6 inches long and 4^ inches across the base, and 1|- inch and |- inch thick respectively. Four pounder f usuao s l type, worke t eacda f theho e endmo ar tw ; peckedsidese th n , o probabl givo yt betteea r grip. A spherical hammer-stone measuring 3^ inches in diameter. A rude cleaver made by chipping a flat stone all round and shouldered fo handlea r t measureI . s 8^ inche 5^y s-b inche inch1 y sb . A rectangular knife-like implement chipped on one edge and on one end; measuring 1\ inche inche4 y |-sb y sinchb . Anothe f similao r r shap t chippee facebu ed rounth an e sn o dth d edges; measuring 7j inches by 3j inches by 1 inch. Seventy unworked flakes varying from 2-| inches in diameter to inche8 6y ^sb inche crosn si s diameters. A scraper, made of a flake chipped round part of the edge on the smooth side, measuring 3^ inches in diameter. Six flat unworked discoid stones varying from 3^ inches to 6 inches in diameter; another, worked on one side and measuring 4f inches in diameter pot-lida ,e b whic y . hma Nearly all these objects have been made from water-worn beach- stones. A scraper, a worked piece, and four flakes of flint. Eight piece f pumice-stoneso showino tw r o ge signon , wearf so . A fragmen wale steatitia th f o lf o tt wit marce po perforatioha th f ko n s brokeit f acroso ne edgesson .

STRUCTUR. 4 E The fourth construction (fig. 2) at the end of the dike need only be alluded to briefly as it was left practically untouched. It lies between 30 feet and 40 feet south-east of the last in a stony grass-covered mound which is now so low that there can be little structure left in it. In the middle foundatione mounth th f e o e dar n oblon a f o s g chamber which is set with its main axis lying east-north-east and west-south-west. The western end is missing but the chamber has measured at least 134 PROCEEDINGS OF THE SOCIETY, JANUARY 11, 1937.

14 feet long by 8 feet 6 inches wide, within walls 3 feet thick. The masonry has been reduced to 1 foot 6 inches in height at most and it res'ts on 1 foot of black earth above the natural clay. There was no sign of a paved floor, but only a very small area was examined at the south-east corner. Appearing above the surface just inside the broken northere th f o d n en wall four slab foro t s medgn a o werthre o t es ese e sides of an open-ended box, two of the slabs being in alignment on one side. No relics were found t judginbu , g from what little evidence ther, eis probabla erectioe datth r efo als s nEarli e oth y Iro ncirculae AgeTh . r depression adjoinin "wely contemporare gb ma l e fac vien yth i tf wo tha t latee th r supposed dike make detousa r roun. dit

I have to express my gratitude to Major Harry H. Hebden, M.C., who readily consented to the examination of the monuments and gave me n entirela y free han thein di r excavation e SocietTh . s indebteyi o dt him for the presentation of the most important relics to the National Museum I shoul. d also lik acknowledgo et appreciatioy m e f otheno r privileges accorde r GeorgSi , edme Macdonald having defrayed parf o t the travelling expenses and our Society having met the expenses of the workmen, thereby enablin undertako t e gm a more e extensiva d an e more exhaustive exploration than would otherwise have been possible. Thanks, too herewite ,ar h expresse . GrahaJ r D mo dt Callandes hi r fo r report on the pottery and to Professor Alexander Low and Miss Margery . PlatI their fo t r report humane th n so , animal bird dan , bones. Finally, I should lik mentioo t e indebtednesy nm friendy m o t s, Bailie . WM . Ritchie, Johnstone, for valuable assistance in the course of the excava- tions and to all the workmen whose enthusiasm and unsparing labour helpe muco ds lighteo ht tasy nsupervisionf m k o .

REPORE POTTERYTH . J GRAHAN O TY B . M CALLANDER, LL.D., F.S.A.ScoT.

While excavating the two monuments on the Calf of Eday, just deacribed r CaldeM , r foun considerabla d e quantit f potteryo eachn yi . The pottery fro e longmth , stalled cairn belong distinco tw o st t periods, the Neolithic and Early Iron Ages of Scotland, but the shards from the other building, which he considers may have been used, at least ulti- mately potter'a s a , s workshop, belonlatee th ro gt time . y Thisma t i , be mentioned, is the period of our brochs, earth-houses, and crannogs, A NEOLITHIC DOUBLE-CHAMBERED5 CAIRN13 .

and many of the kitchen-middens in the , all of which may be considered roughly contemporary with Romano-British times.

THE STALLED CAIRN—CHAMBER A. Neolithic Pottery. This pottery, like all the rest which was recovered, is so terribly smashed up that it has not been found possible to restore, and that only partially, the mouths of more than four vessels, though at least thirty- four are represented amongst the shards. After eliminating the pieces of rims ove Ibs7 f 1 brokero . n pottery remain t althougbu , h thers ei suc a hlarg e quantit o wort y k upon o e sharde s edgeth th ,e f o ar ss irregular e siz e variatio f manth ,smallo o th es d s yi an , n colouni f o r single pot greato s , tha piecet w onlfe sya have been fitted together. Still, from the curves on many of the fragments it was clear, even though there had been no tell-tale rim pieces present, that we had to deal with round-based vessels which pointed to a Neolithic date.

Carinated Urns. wald an l fragmentm Ri . 1 bowa f harf o so l d brown ware largese th , t piece measuring 3j inches long and 2y§ inches deep (fig. 18, No. 1). About if inch below the rim, that is bevelled downwards on the inside, mouldingw lo isa under whic wale hth l curves int oroundea d basee Th . space between the lip and the moulding, which is slightly convex, is occupied by reversed hatched triangles bordered by a single horizontal line above. The marginal line and those forming the triangles have been lightly e incisesteadth y yb da sharp-pointe pul f o l d tool t thosbu , e fillin trianglee gth s displa ystab-and-draa g pattern mad unusually eb y shor lighd wale an t tonls Th li stroke y. ^1) . incs (figNo h , thick.15 . d walan l fragmentm 2Ri . a bow f f gooo so l d hard ware n parti , s black with a skin of brown colour on the outside, and at other places red e outsidbotth n insidd ho ean e wit. 2) hblac. a No k, cor18 ed (figsan 6 1 . The upper part between the lip and the shoulder, "which is 1^ inch high, is concave and entirely occupied with decoration, and the lower part turns in towards the base in a fairly quick curve. The vessel has been about 9|- inche diameten si moute th aboud t ha r an J t 10 inchee th t a s e rimth f ,o thoug p shoulderto he roundedTh . s flattisi , n placesi h , and measures -fg inch in thickness. The lower part of the vessel is also YQ inch thick. As for the decoration, there are three lines encircling the 1 The diameters of the mouths of the vessels are measured externally. 136 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

immediateln ur y unde e brim th re space th , s between them being filled with short lines, those above slanting from lefrigho t thosd t an t e below e oppositth n i e directioforo t ms herring-bona a o ns e pattern. Under-

Pig. Fragment15 . Carinatef so d Neolithic Pottery from Cal f Edayo f . ($.)

neat e panele for ar hth f reverse m o n i s d truncated triangles hatched with straight lines alternately slopin n oppositi g e directionse th l Al . line e sharplar s y incise d havan d e been mada pointe y b e d tool (fig. 16). d walan l m fragment3Ri . f abouo s t three-quarter e uppeth f o rs part of an urn of hard red clay, which fit only in one or two places. But herring-bone foth r e pattern slopin opposite th n gi e directio scheme nth e

138 PROCEEDINGS OF THE SOCIETY, JANUARY 11, 1937.

of decoratio e samth s thas a eni e incisete last th seeTh .n i nd lines are generally made by a steady pull of a pointed implement, but at one place the point has been lifted at the middle of the line so as to form two fore . m3)Th .shor No als , vers oi t18 oneyd . 1similaran ,s No (figs , 17 ., thoug e carinatiohth s morni e pronounce rathed an d r neare e lipth r. The diameter at the mouth has been about 8 inches, and the wall, of which a height of only 3J inches survives, at most is J inch thick. 4. Fragment of the wall of a vessel of dark ware, measuring 2^ inches inches2 y b , showin gsmala lcarinatiowale e parth th f lf o to below d nan , lattee th r being f inch thick. Doubtles same bees th f ha eno t for si m lase uppee th t th s two a n keerdecoratee s O i sid. th t li f eo d with vertical incised lines (fig. 15, No. 2). 5 to 9. Rim and wall fragments of four, probably five, carinated urns of har ward re buf er o fwit decorationo hn vessele Th . s hav l beeeal n of large size, possibly having a diameter at the mouth up to 13 inches. somn I e piece wale sth -fgs i l inc^d hincan h d thickan , 2 .. FigsNo , 17 . , shoe 6 rim d fore uppeth wd th an sf m an o , Nos , r 5 e 18 part , th 4 . f o s wall, and fig, 17, No. 2, the vesicular character of the ware. 10 to 13. Small fragments of the carinated part of the wall of four different urns, three of dark-coloured ware and one buff-coloured on the outside. Three have been decorated above the keel with oblique incised lines and one seems to have been plain. f brigh o . Fragmen 14 m tri red-colourea f o t d ware measuring only 1^ inch long by 1 inch deep at most. The top of the rim is flat and measures -fg inch thick. Immediately under the lip is a transverse line with vertical lines depending from it all incised with a narrow pointed tool (fig. 15, No. 3). Although so little of the urn remains, the ornamenta- tion which it bears points to its having been of the carinated variety.

Urns Carinated.not d walf haro an ldn m fragmenur buff-coloure Ri n . a f 15 o t d ware measuring 5J inches in length and 3 inches in breadth. The wall is vertically convex on the outside and shows a deep hollow moulding under the lip (fig. 18, No. 12). The top of the rim is flat and bevelled downwards towards the inside; it measures ^ inch in width. The diameter of the mouth seems to have been at least 10 inches. What brie ornamenteth e mf ar o p to e remaindth walwite d th an hlf s o finger - nail markings set obliquely. The nail has been inserted into the damp clay and a small piece dragged out, leaving a vesica-shaped hollow side e on e th e naimar e steepethae th n th i lothee f th k o nd th n o an rr A NEOLITHIC DOUBLE-CHAMBERED9 CAIRN13 .

bottom (fig. 17, No. 3). There is a small perforation 1^ inch below the brim. 16. Rim and wall fragments representing rather more than half of the mouth of a vessel of very fine hard dark ware with a tinge of brown in places and rubbed smooth until it shows a gloss. At least only 2^ inches of the wall below the lip is preserved. The urn has measured about 8j inche diameten i s walle mouthe th th ,d t whica a ran , s hha rounded lip measuring ^ inch thick, increases to f inch at a depth of 2^ inches from the rim. There is a very faint suggestion of a hollow moulding jus. 11) tNo (figp .unde, li 18 .e th r fragmentm Ri . a large 17 f o s , plain, coarsely made bow f daro l k. brown ware extending to about one-third of the circumference of the mouth. At one place a piece of the wall measuring 4^ inches in depth remains, but for the most part there is less than 2 inches. The rim is slightly rounded on the top with a gentle slant downwards towards the interior and a distinct curve on the outside. It projects both on the exterior and interior, in some places showing an undercut roll (fig. 18, No. 7). It has measured about 13 inches in diameter at the mouth, the rim being 1^ inch broad and the wall y^ inch thick. 18 to 30. Thirteen vessels, each represented by a small solitary fragment of the rim. None of the pieces measures more than 2^ inches d 2ian inchep li alonn depthi e s th g , f brigho excep d e re ton t colour whic 2s hi ^ inche d 3^an s inche greatesn i s t lengt d depthhan . The surviving shards show that the upper part of the wall in every one d beeha n vertical e thicknesth , s varying from •§ inc o 3^5ht - inchn I . s innes it roundedi e p r li on edg e n i bevelle,s th ei f o p dseveto e nth down-wards n foui t , i projectr s s slightli t i e y on outwards n i d an , flat (fig. 18, Nos. 8 to 11 and 13 to 15). Several of the vessels a-re buff- coloured, others dirty grey and black, and two bright red on the exterior. One of the smallest pieces, which is only lj^ inch long and If inch deep, has a hole bored through it -| inch below the lip, and the largest fragment has a perforation ^ inch down. Only one of the shards from these thirteen vessels gives an indication of the shape of the base, and this is the last one mentioned. The lower part of this fragment curves in distinctly, so that there is no dubiety about the bottom having been round (fig. 18, No. 13). It is more than possible that the other pots had the same peculiarity, especially the thicker pieces. Amongst e largth e collectio f urnno s e chamberefounth n i d d cair t Unstana n , also in Orkney, which is really of the stalled variety, there are fragments of several with vertical walls whic cae hw n claim were rounded below, although ther s onlevessei e yon f whic o l h enough remain shoo st e wth Fig. 18. CalfofEday. ({.) A NEOLITHIC DOUBLE-CHAMBERED1 CAIRN14 .

ingoing curve at the base. None of these thirteen Eday pots is orna-

mented f thoso , y neithee an wit e har r vertical wall1 s from Unstan.

All the above pottery was found in a single mass in the Neolithic layer innee footh e f th rwestero t e edgt a th f eo n divisiona lsoute slabth n h,o side of stall No. 2, at the place marked X on the plan (fig. 3). But a large handful of small broken fragments, the largest measuring only 2 inches by 1^ inch, was found in the middle of the same compartment at its eastern end. This potter s devoiwa y f ornamentatioo d d theran n e wer fragmento n e f rimo s s amongs . it tClos o thit e s depositn i t bu , e adjoininth g a rathecompartment s wa r , smalle1 . No ,r quantitf o y pottery n whici , h were single smal piecex differenm si ri l f o s t vessels, one bearing ornamentatio anothed nan r coming fro ma carinate d bowl broken off in the middle of the keel; in addition there were pieces of two decorated carinated urns broken off above and below the moulding. Among thi t werslo e fragments from four urns which matche e coulb t dno d in the big selection of pottery from compartment No. 2, so this number has to be added to the previous total of vessels identified, bringing it up to thirty-four at least.

Carinated Urns, 31. Very small rim fragment of a vessel of fine, dark brown ware, measuring only 1^ inch long, f inch deep, and ^ inch thick. The top roundeds i p li jusd e oan ,th tf belothree ar wt ei incised horizontal lines which have encircled the urn. 32. Small rim fragment of buff-coloured ware with a reddish tinge, of good texture, measuring l|^ inch by !•§• inch. It shows a piece of keee th l wit e outehth r skie undeth n n scalero f side distance dof th , e d flas i an t m measurinri p li e e th betweef th o d p gan to | - t n i e inch Th . measures -fg inch in thickness. 33. Small wall fragment of a vessel of fine, dark brown ware, glossy in places, measuring 1§ inc 1y h b -fg inch showind an , gsmala le parth f to uppee keelTh .e carinatior sidth f eo decorates ni d with incised vertical lineswale s onlTh i l . y yg- inch thick. 34. Several fragments of fine, pinkish ware, with a black core, show- ing small portion e carinatioth f o s n with e walpartth lf o sabov d an e below it. The largest piece measures only 1-^ inch by 1^ inch, and the wall is only -j inch thick. What remains of the ornamentation consists of a

horizontal line with vertical lines, dependinkeele th s , a r gfa s froa t mi 1 Proc. Soc. Ant. Soot., vol. Ixiii. p. 45, fig. 14, No. 1. 142 PROCEEDING E SOCIETYTH F O S , JANUARY 11, 1937.

l incisedal impossibls survivesi m .ri t i e Nond th an ,f completo et o e decoratioe th nhorizontae betweeth d an t ni . l lin4) . e (figNo , .15

Generally the pottery is hard and of good quality, but the firing is uneven. Buff-coloured pieces often sho e samwth e tinl throughal t ; reddish pieces may have a skin of that colour on the outside only, or on the inside as well, the remainder being black; and in some single pots e r walblace parto burnth ar ld f ko sre t throug d throughan h e Th . thickness of the walls varies from ^g- inch in the larger vessels to no more than YQ inch in the smaller. In much of our Scottish Bronze and Iron Age pottery crushed ston bees eha n addeclae varyinth yn i o dt g quantities before the dishes were made, but this practice does not seem generally to have found favour wit Orknee hth y potteStone th f ero largee Ageth n rI . pots crushed ston hardln eca detectede yb thesn i t e bu ,som e small pieces of stone, sometimes water-worn, are to be seen, as if clay containing very little grit had been selected. The clay from which the smaller vessels have bee ne appearanc th mad s eha f havineo g been washed. Somf eo pottere th slightls yi y vesicula texturen i r thid san , appliese eveth o nt fines thinnesd an t t thice shardsth kn I shard. e cavitieth s muce ar s h larger than in the thinner pieces. The vesicules are quite apparent in figs. 15, 16, and 17. A few of the shards have a gloss on the surface which is not a slip but which was probably formed by rubbing. Three pieces have a hole bored through the wall not far from the lip. All these peculiarities are to be seen in other Orkney ware of the Neolithic period—on urns from RousaUnstae th d nyan cairns. Whe compare nw forme eth decoratiod an s vessele th f no s from Eday with those from the other places in Orkney mentioned, we shall find that there are also striking resemblances. Wide-mouthed, shallow, carinated bowls, with or without ornamentation between the keel and the rim, and cylindrical pots with plain vertical walls and rounded bottoms prevail. Only two of the Eday vessels, of which enough survives to giv n accurata e e ide f theio a r form e differenar , n shapi t e froy man recorded from the other Orkney sites. These are No. 15 (fig. 18, No. 12) and No. 17 (fig. 18, No. 7), the first showing a deep groove under the lip, secone th d dbroada an , projecting, rounded rim. Regardin decorae gth - tion, hatched reversed triangle a favourit e ar s e motive d thougan , h only one solitary vessel from Eday, No. 1 (fig. 15, No. 1), exhibits the stab-and-drag process, it is an interesting piece because of the lightness delicacd e an linesth f y.o Also , show onlvessele 15 . yon s No ,finger - nail markings (fig. 17, No 3), but this style of decoration is rare on the other sites. A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 143

Considered as a whole it would be impossible to mistake this collectio f pottero n s comina y g froy placman e outsid e Orkneth e y Islands.

When Mr Calder made his trial examination of the cairn, in 1935, and discovered this hoard of Neolithic pottery it was expected that a complete excavatio e monumenth f no t woul rewardee b de disth -y b d cover f moryo e vessel humad san n remain periode th t f thio s bu ,s hope was doomed to disappointment, as very little more pottery was recovered d skeletaan l remains were n pooscanti d r an ycondition . However, if the results in the matter of finding such relics were not what was desired, there arise e questiomuco th s y h nwh broke n pottery should be concentrate spote on .n di e knoW w fro r WaltemM . Grant'G r s excavation f stalleo s d cairns in the island of Rousay that these tombs had been entered after the time firse oth f t burials, possibly when later ones took place t woulI . d appear that pottery previously deposite d receiveha d d scant attention from newcomers, as not only did vessels get broken but fragments of them were tramped inte flooroth . This happene e Midhowth n i d e cairn, where human remains were very plentiful and pottery scarce.1 It may be that something simila d happene ha re Cal f th Edao f n i dy cairn. Certainly the pottery found in the cairn must have been heaped together befor chambee rooe th f eth o f r collapsed, greateande th s ,a r par mosf o t t e vesseloth fwanting s wa s e t potterlooki f mos,i th f s beed a o st yha n thrown out of the chamber altogether. It might be suggested that the vessels had been removed from their original positions to make e receptioth wa r yfo f lateno r burials f thiI .s t wouli wer , so ed seem indicato t grav e ewandeada e th th reverencer f er o tfo ,furnish fo t no f i ,- ings thabeed ha tn deposited wit have hw indicatioo n ethemt Bu . n of how many bodies were buried in the cairn, nor of how they were placed, although more than thirty urns were represente shardse th y db , and so we cannot offer a satisfactory explanation for such treatment potterye oth f . IronPottery.Age Afteburiae rooe th f th ro f l chambe disappeared rha interioe th d dran had been partially filled up, judging from the pottery found at the higher leve chambee th l beed rha n occupie peoplIroy e db th nf o eAge . While the Neolithic floor, following the slope of the site, slanted down from east to west, the Iron Age layer was nearly level, so that while it Proc. Soc. Ant. Scot., vol. Ixviii . 339p . . 1 144 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

was onl inche6 y s highe rs 2tha^ e e wa earlieeasfee nth th t i tt t a r higher at the west or inner end of what originally was the burial chamber. fragmente Th f pottero s y foun thin o ds occupational leve neay la l r e structureth f o innee edgee th d Th r.en s were crumblf o verd w yan yfe the pieces could be fitted together. The ware was hard, and contained a considerable quantity of crushed stone mixed with the clay. It was

. PotterIroe Pig19 n.Ag y from Cal f Bdayo f .

blac colourn ki t unevennese owinth bu , o gt firine th e exterio f gth o s r was black placesn i , d brownre . d Theran , beed eha n onlvessele yon , which seems to have been bucket-shaped with a straight, sloping wall. Fro msingla e survivin fragmenm gri moute th t h appear havo t s e been about 8 inches in diameter. The rim has a gentle downward bevel towards the inside and measures f inch in thickness, while the wall is -fg inch more. All the shards are decorated by incised filled triangles and lattice designs, irregularly and carelessly drawn (figs. 19 and 20, No. 1). Fragment f anotheo s r vessel f sofd waro ,re t e containin a gsmal l admixture of broken stones, were found outside the north wall of the cairn fee2 , t unde presene th r t surfac grounde th f eo , opposit northe eth - A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 145 west corner of the burial chamber. The pieces of pottery are too small to give any indication of the size of the pot, which had a slightly convex

Pig . Section20 . f IroPottere so nAg y from Cal f Bdayo f . (}.)

•wall, TT inch thick, devoi f ornamentationdo . Onl vere yon y small piece of the rim, which was rounded on the top and projected outwards, was recovered. VOL. i,xxi. 10 146 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 . THE STALLED CAIRN—CHAMBER B. Iron Age Pottery. -This chamber, though certainly built during Neolithic times, had suffered so much damage later on that no relics, even pottery, of the period were t 3discovere^Bu Ibs . . it weigh n di f shardo t s were found distri- buted throug hthicknesa f abouso fee 2 tf falle o t n stone soid an sl with which the chamber was filled. Fragments of eight vessels were counted, piecee nonth f eo s being ornamented. The pottery included:— 1. A rim fragment, 4^ inches long and 2^ inches deep, and a few f hard"walo t l, po pieceblack a f o s, thin ware, which seemo t s have had a globular body with an upright rim rounded on the d bevelle e insidan . th p m Unde6) n ri to . eo d e (figNo th r , .20 is a shallow moulding. The mouth has been 6 inches in diameter, and the wall is ^ inch thick. smalo fragmentm 2Tw .ri l f anotheo s r vessel, which seem havo t s e been almost simila n shapi r texturd e lasan eth t o et (fig , .20 No. 5). 3. Rim and wall fragment of fine, hard, black ware with a brown exterior. The rim is flattened on the top and is everted in a fine graceful curvemoute beeTh s .h ha n 7^ inche diameten si r e walth Y$s i ld incan . h 4) thic . kNo , (fig20 . 4. Small rim fragment of good, hard, grey pottery, the lip everted d bevelle e topan th ,n measurindo g -J inc. h 3) thic . kNo , (fig20 . 5. Smal fragmenm ri l a vesse f o f lighto l t brown, vesicular ware, the brim everted and with a slight roll on the outside (fig. 20, . 2) . No 6. Piece of the flat bottom and fragments of the wall amounting to less tha nquartea globulaa f verf o o r t y finepo r , hard, brown ware wit hblaca k wal e core ^s Th li inc. hvessee thickth d l an , has bee t leasa n inche8 t n diametei s r internally aboue th t widest part and about 5 inches across the base (fig. 20,'No. 8). 7. Small basal fragment of a vessel which resembles the last in form, but which has been smaller. The pottery is hard and black with a brownish exterior (fig. 20, No. 7). walw fe l 8fragmentA . f daro s k ware wit exteriord hre , measuring f inch thick. 9. A double handful of small pieces of very coarse, crumbling ware, which contains a large proportion of crushed stone. Amongst bas e flat-bottomea sharde th f eth f o smala o s st i lbi d vessel. A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 147

POTTER'S WORKSHOP—MOUND No. 3. Iron Age Pottery. Althoug e potterhth y from this structure differs entirely from that found in chamber B of the stalled cairn which has just been described, we hav same assigo eth t eo t period nt i Iroe ,th n Age. sharde Mosth f o ts from chambe werB r f gooeo thid nharan d texture, while those from e coars ar thickd Moun 3 mand ean . an , dNo y contai nmuca h greater amount of crushed stone among the clay. Again the pots have been thoroughl vers i yt i y smashe doubtfud an , completa dup f i l e sectiof no the wall of a single vessel could be pieced together although there are 42 Ibs. weight of fragments. Most of the ware is black, sometimes with outee th r surface burn througd tpiece re w red throughe d fe sar ha ,an , d occasionallan y the e buff-colourear y r yelloo d w wit r withouo h a t black core. The walls seem to have been nearly vertical, and none of them shows any traces of ornamentation. There is nothing to suggest vessele th f o stha y wertan e other than domestic. The thickness of the wall fragments varies from ^ inch to Y§ inch, and the thickness of the bases from -| inch to 1-j inch, but few pieces indeed are of the smaller dimensions. Though there are twenty frag- ment man w f s impossiblrimi o sho t yi sy pot sa e represented ar so t e . These pieces show suc hshora t lengttham ri tf honlo thren yi e cases si it possible to ascertain the probable diameter of the mouth; these three fragments seem to have come from vessels measuring 5|- inches, 7 inches, and 8|- inches. Basal fragments of at least ten pots, all flat, measuring fro mshada e less tha inchen3 5o st ^ inche diameten si countede b n ca r . have A t beesw eno n abl correlato et thef o piecemm y ri ean wit y s han we cannot ventur suggestioy e heighe vesselsean th th f o n t o I t s . na som e wals curveth e ha l niceln i d y toward other n i e base t th s t i sbu , curves slightly outwards forming a splayed foot. Figs. 21 and 23, Nos. 1 to 4, show the form and thickness of some of the rims and bases. A fragmen splayee coarsf th o f t o t depo baswarea f eo , wit hwala l \ inch basa a thicd lk an diamete nearlf o r inchesy5 , parseeme b t o t s of a "waster." The clay is burnt very red all through, and among the broken stone which has been added to it is a piece measuring -|- inch by \ inch by \ inch, the largest that I have ever seen in a Scottish prehistoric pottery vessel. Althoug e firsth t t ha glanc e this seem mosa s t disappointing collec- tion of broken dishes, some of the pieces demonstrate very clearly how several of them were fashioned, and the general appearance of the shards requires some explanation. Fig. 21. Sections of Rims of Iron Age Pottery from Calf of Eday. A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 149

makinn I sepulchrae gth domestid an l c vesselStonee th f so , Bronze, and Early Iron Ages in Scotland, generally it was the practice to work up the clay from the base to the lip so that the wall was homogeneous, wheel-turnen i s i t jusi s a t d pottery. But havI s ,a e already pointet dou r Proceedings,ou n i vol. Ixiv . 195p . , fig , anothe.2 r metho occasions dwa - ally employed Bronz e botth Irod n hi ean n Agesclaformee s yTh wa . d

. IroPotteryPige 22 .n Ag , showing metho buildinf d o vessele th p gu , from Cal f Edayo f . (\.) into strip bandr so thesd san e were superimposed edgewis anothere on n eo . The upper edge was squeezed into wedge-form usually rounded at the apex, and the lower side of the higher strip was pressed on to the arris of the lower so as to form a groove on the under side of the higher. The clathes ynwa workeds dowa o bots n ninsidt o e hth outsidpo d e ean th f eo foro t overlan ma botn po h sides. This naturally mad weaea k joint, whed an n pot t fracturesgo d they often brok t thiea s part. Three th f eo Eday pots have been thin buili sstripe p wayu tth clayf d so ,an , measur- ing from the apex above to the bottom of the hollow below, are !•§• inch, 2j inches, and Z\ inches in breadth (fig. 21, Nos. 3 and 5, fig. 22, and fig. 23, Nos. 4 to 9). Some pieces must have been broader. In a food- vessel from Kilspindie, East Lothian, which measured 3J inches in height, there were two strips above the basal one, so they must 150 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

Fig. 23. f BaseWald o i an } ls 4 Portion (o 1t s showing metho f buildindo p gu the vessels (4 to 9), from Calf of Bday. (£.) A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 151

have been less than 1 inch broad. In a cinerary urn from Long- niddry, East Lothian, a strip if inch broad was noted. Professor Childe discovered fragments of pots built up in the same way at in Orkney, and Dr A. O. Curie at in . In all the pieces from the Calf of Eday, and in those from the two East Lothian sites, the bands of clay had been laid horizontally; but in an Iron Age vessel found in the Bishop's Loch crannog, near Glasgow, they were built up spirally, and the junction between them formed a bevel from the outside to the inside of the pot. The pottery does not look like a haphazard lot of broken dishes such wouls a d have bee dwellina genera n n i n period e i th e f us glthero s a , e are so few fragments of fine, thin-walled vessels. It looks as if the shards retaine d beeha dn selected becaus f theio e r thicknes d coarsenessan s , and that for a specific purpose. Outside the principal buildings of the broch of Midhowe, in Rousay, what might be called a stone chest was found nearly fulf brokeo l n pots. These evidentl d beeyha n specially inletlaia e se Stanchnd th ,e aside th s a ,a Geo s onl,i yard6 y s distant, sure b ey tha ma t woulti e w utilise e dd b duma an s dbrokea r pfo n dishes and kitchen-midden refuse. These shards were believed to have been kep r beinfo t g pounde dwitx dowmi h o fresnt h clay batc whew ne hna of pottery was being made (Proceedings, vol. Ixviii. pp. 483 and 508). Possiblhavy ema similaa e yw r collectio Cale Edaythif f th i o f t snd a an , explanation is correct it strengthens Mr Calder's claim that the building wa potter'sa s workshop occurrence Th . wastea f o e r amon sharde gth s may be considered further evidence in support of this.

I am much indebted to Mr A. J. H. Edwards, Assistant Keeper of the Museum, for assistance in sorting out the pottery.

E REPORSKELETATH N O T L REMAINS. y ProfessoB r ALEX. Low, M.A., M.D., F.S.A.ScOT.

Among the fragmentary human skeletal remains from the stalled Cale cairEdaf th o f identifiedn e no y ar : Lowe representew rja piecy db e of bod righf yo t sidfragmend ean f lefo t t ramus o cervicao tw tw ; d an l thoracic vertebrae; of hand, a scaphoid, piece of third metacarpal, and three middle phalanges of fingers; fragment of head of tibia; of foot, an imperfect right talus and a left navicular. 152 PROCEEDING E SOCIETYTH F O S , JANUAR , 1937Y11 .

E REPORANIMATH N O TL BONES. By MARGERY I. PLATT, M.Sc., Royal Scottish Museum, Edinburgh.

This collection of animal remains found by Mr C. S. T. Calder last summer (1936 lona n gi ) stalle de remot cairth n no e e islanCalth ff do of Eday, Orkney s extremeli , y interesting, sinc e relicth e f domestio s c animals form an insignificant part of the remains and undomesticated specie sdee d equalle sucre ar rs ha y sparse conditioe Th bonee . th f no s is as usual very decayed and fragmentary. A bone implement formed from the metatarsal of an ox is markedly different in texture from the majority of the remains. It is hard, heavy, apparently well impregnated with mineral salts suggestd an , originalls s thawa t i t y derived froma different locality tha e caire boneth nth nn i s itself. Throughoul al t e excavatiopartth f so e bonenth se givf fireo Th e.e evidencus e th f eo various specie e recordear s d below with referenc o that e e t th par f o t structure from which they are taken.

Neolithic.—Associated with the human bones occurring at the original neolithic level are the fragmentary remains of an otter (Lutra lutra) including the left ramus of a lower jaw; parts of the skull (a fragmentary, parieta d threan l e teeth) a lumba; r vertebre remnanth a d f an o at thoracic vertebra incompletn a d an ; e right femur.

Iron Age, Chamber A. Shell-layer, \st Compartment.—A few small fragments of an ox (breed indeterminable hornea d an )d sheep occur here. Although from the title it is evident that shells probably of many kinds were present only those of winkles (Littorina littorea) were sent to me. Remains of birds were very plentiful t leasA . t five cormorants (Phalacrocorax. c carbo) are represented, and although most parts of the skeleton are included wind an gbesg e bonele ,th te preserved,sar frod an m these some e numbeth ide f o af bird o r s present have been calculated core On - . morant -was especially large, the others of average adult size. The coracoid of a sea- or white-tailed eagle (Haliceetus a. albicilla) also occurs. From small Slab Construction, 1st Compartment.—Here s includei da single mammalian remain—the fragmentary jaw-bone of an ox. The remainder are all bird bones, excluding innumerable fish bones of minute size which wer l probabiliteal presenn i d e refusesoi th an d l n an i yt, represent the stomach contents of the birds. The relics indicate the A NEOLITHIC DOUBLE-CHAMBERED CAIRN. 153

presenc o cormoranttw f eo s (Phalacrocorax a sha. carbo)c d g an (Phala- crocorax aristotelis).a. Shell-layer, 2nd Compartment.—Two fragments of mammalian bone occur, probably of bovine nature, and the remains of three cormorants. Shell-layer, d Compartment.3r —Aremainw fe f sheepo s , mostl- yim mature, and an ox bone occur here. There was also included a limpet shell (Patella vulgaris), e man probablth f yo comprisine yon e shelgth l layer. Bird remain e agaiar s n most numerous t leasA . t eight cor- morants (Phalacrocorax carbo)c. were present a greate, r black-headed gull (Larus marinus), a herrin g gull (Larus . argentatus),a a sha d g an (Phalacrocorax aristotelis).a. Shell-layer, 4th Compartment.—In this section the hard ox-bone implement occurs. A few other bovine remains are present, together wit molae hth r toot sheepa smala f hd o lan , long bone (probably foetal) of the same species. There are the relics of six cormorants (Phala- crocorax f ver. carbo),c o yo largtw e sizefoud an , r medium-sized adults.

Chamber B. n thiI s sectio na numbe f sheeo r p remain l froal s m young animals, including small horn cores, occur. There is also a terminal phalange and a bone from the sternum of a red deer (Cerms elaphus). There s onli aviae yon n remain e wing-bonth , merlia f eo n (Falco columbarius cesalon). Lastly, several bones from the head of a cod (Gadus virens) are present. Potter's Workshop; from small Pit F. Here occur again many bones of the cormorant, representing two calcinew birdsfe sheeA d d .an als e bonepx ar o o f evidentso .

Potter s Workshop; 2 feet 6 inches above Floor. In addition to a few bovine remains, a single fragment of the metacarpal of a red deer (Cervus elaphus) also occurs.

t I wil e seeb l n fro e foregoinmth g thae specieth t f paramouno s t importance occurring in the domestic level of the cairn and the potter's workshop is the cormorant. The remains of twenty-eight birds could be counted and probably there were many more originally. Other bird n suci s t happear no grea t bu t, number d thereforan s f minoo e r importance. It is possible in consequence that these primitive islanders were usine cormoran e flesth gth f o harticln a e sams f dieta tth eo en i , 154 PROCEEDING E SOCIETYTH P SO , FEBRUAR , 1937Y8 .

way tha t KildanS t othed san r inhabitant Westere th f so n Islands have been known to subsist comparatively recently on the solan goose as their staple food.

wisI recoro ht sincery dm e thankCalder M kindlr o st fo r y forwarding animae th e m l boneo t identificatior sfo reportd nan .

III.

EXCAVATIO F TOBRO N S CAVE, KIRKCUDBRIGHT. BY S. V. MORRIS, F.S.A.Scot.

(To be printed at the end of volume.)

MONDAY, Sth February 1937.

SIR GEORGE MACDONALD, K.C.B., President, in the Chair.

followine Th gAddrese replth o yt Majests s Hi sen Kine o t t n yTh go his Accession to the Throne has been received:— SCOTTISH OFFICE, WHITEHALL, S.W. 19th January 1937. SIB, I have had the honour to lay before The King the loyal Address of the Society of Antiquaries of Scotland on the occasion of His Majesty's Accession to the Throne, and I have been commanded by The King to convey to you His Majesty's deep appreciation of the sentiments of loyalty and devotion to which it gives expression. Sir, ,am I Your obedient Servant, (Sgd.) WALTER E. ELLIOT, Secretary Stateof Scotland.for