154 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929.

1 : • • I. ON A JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. B . YHEWAJ T CRAW, P.S.A. SCOT.

The villag f Kilrnartino e , Argyll, standapee triangulaa th f xo t sa r area of low-lying ground some three miles in length. At its base is the level stretch of Crinan Moss, and on either side rise the rugged hill f Argyllso . This delta is bisected by the Kilmartin Burn, which flows to meet Rivee th r Add t geologistbu , s tel s tha e valletimu e l th as ton te ywa traversed by a much larger river, when a lobe of ice blocked the Pass of Brander e waterth d f Locdischargee so ,an hAw d into Crinan Loch.1 The course of this river may still be seen at parts, and further evidence of early conditions e forexist th f gream o n i s t bank f gravelo s a , hundred feet above the present sea-level. These extend along either e valleyth sidspea d f a timeo e an ,f ko when Knapdal Kintyrd ean e were e waveislandth f Locd o s san h Crinan rolle whero t d e Kilmartin villag standsw eno . When the sea receded to its present level, the district now traversed Crinae th y b nKilmartie th Cana d an l n Burn must have becom- im n ea portant thoroughfare thio doubtlesT s si . face th tparo e than sdu t n i t of Scotland are there to be found remains of greater interest, dating from early prehistoric dow mediaevao nt l times. These monuments have frequently formed the subjects of papers in our Proceedings, and have been elsewhere described.2 Dr Christison dealt s givee witdistrict e fortha th n accounth hf na d e o s th an , f o t work done by our Society at Dunadd, when by the kindness of Colonel Malcolm, C.B., of Poltalloch, our museum was enriched by the addition of ove object0 r30 s found durin e excavationsgth cairne d theiTh .an s r excavation have been described by Canon Greenwell and Dean Mapleton, whil r JosepeD h Anderso gives nha accounn n a urne th sf o tdiscovered . The standing-stones have received the attention of Mr Romilly Allan, Professor J. Y. Simpson, and Dr Christison. A very remarkable alignmen f cairno t standing-stoned an s s extends from Kilmartin throug e wholth h e e valleycoursth f o e; severaf o l these f moro cairn e ar stha n usual intereste largth en I cair. t a n Kilmartin f thiGlebenorte o sth d linefouns t h en t necklacea , je wa , da , whic he traced b canno f thio w s.p no tcairn to Fro e mth , whics hwa

1 Memoirs of the, Geological Survey—Sheet 37, p. 3. r BibliographyFo 8 Appendie se , . xA JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 155 only partially excavated d whican , h originally measure fee0 n i 11 dt diameter and 13£ feet in height, one can see the next three in the serie n exaci s t alignment secone Th . d seem havo st e been untouched. The third was unfortunately recently removed. The fourth, at Nether Largie, contains a megalithic segmented chamber which has been figured by Professor Bryce.1 In this cairn Canon Greenwell found the beautiful e Britis(non th ur wn i h Museum) whic s bee hha ofteo ns n figurea s a d type of neolithic pottery. The cairn is situated 1200 yards from No. 1 of the series. Continuing southwards the line passes a standing-stone, having close to the east a remarkable group of standing-stones, and to the west circla f twelveo e stone smala n i s l wood. Beyond Hi Cruie nth line passes over the site of a cairn which contained three cists. In two of these the side slabs were grooved for the reception of the end slabs fro d thire s takean ,mth dunique wa nth e slab bearing sculptured representation f axe-headsso f whico , h r museum.ther ou a cas s n ei i t 2 Continuing e linth ,e passe n successioi s n three buriaa l sited an s standing-stone withi a distancn f abouo ea halfd a mild tan , an e eventuall t Islandada yd crosseAd e dsth Bridge . This remarkable line, 4J mile n lengthi s , probably mark e coursn earlth a s f yo e trackd an , n faci s ti closely followed to-da a road y b y, except wher t i passee s throug e groundth h f Poltalloco s h, House. E. JE t I pointy b . . N s being parallee e Kilmarticourse th th th f o l o t elal no t Burd an n ridge valley d districte noticeabla an sth s i n t i sI . e fact thaaxee th t s of practically all the Bronze Age graves in the district point in approxi- mately the same direction. Another lin f standing-stoneo e d buriaan s l sites extends froe mth modern cemetery north of Lochgilphead to Dunadd, a distance of about 2f miles. About half a mile north-east of Poltalloch, the low-lying triangular area above referre boundes i wese gravea o t dth y b tn do l bank some

1 Proc. Soc. Ant. Scot., vol. xxxvii. p. 60. 2 Proc. Soc. Ant. Scot., vol. viii, 378. parp , . t2 Opinio divides ni d whether thes moulde ear r so mere representations lattee Th .r woul correce dth seee b tm o interpretationt similas ,a r carvings occur on the slabs of a burial chamber in Brittany. There the shafts of the axes are also shown, thed accompaniede an yar curioua y b s figure resemblin heae rakea g th f do . This figur alss oewa carved on one of the Kilmartin slabs, suggesting to Dean Mapleton "large Ogham letters." See Proc. Boy. Irish. Acad., vol. viii, 1861-64, p. 398. There can be no doubt, however, that the figures in question are the debased representations of boats with rowers, such as have been found carve bouldern do . Madsen'P n Denmark i s. A e sSe Alfbildninger. af Danske Oldsager og Mindesmaerker: Broncealderen II. (1876), p. 49, pi. xxxvii. The same motif appears on knives of the Bronze Age in Denmark (Ibid.: Broncealderen I. (1872), pi. xxiv.). Mr George Coffey describes ha d these "ship-figures Grangew Ne t "a , near Drogheda draws ha d n an ,attentio o nt their similarity to Scandinavian rock carvings. He does not seem to have known, however, e Scottisoth f h exampl t Poltallochea Trans.e Se . Roy. Irish Acad., vol. xxx . 30-3pp . 7 (1892), also New Grange, Coffey, p. 60 (1912). 6 15 PROCEEDING E SOCIETYTH . F O S . , JANUAR , 1929Y14 . -,.; . 50 fee yard0 t 60 n length—thhig i sd han e 100-foot beac f pre-glaciaho l e norttimesth o h T thi. s bank endpromontora n i s y abov Norte eth h e hamleLodgth d f Slockavullino tean a.somewhas s souti it d t hen A . t similar promontory, roun e bas f th dwhico e h curve a smals l stream. The crest of this southern promontory has been used as a gravel pit for many yearsknows i t I s Brouc. na Drummin ha n (thee brath f eo

' '''

Y0 4 W 0 3 0 0 2 10

Fig. 1. Plan of Gravel Pit.

elder bushes), but the old name is said to have been Kill y Kiaran : (the cel r burying-grounlo t Kiaran)S f do . e Ooccasio nth e visi th f member o f t no e Ancien th f o s t Monuments Commission to this district in April 1928, their attention was directed a shor e pito t e edgeth t. th f cis o Thit beed a t ha sn opene 191n i d 0 by Professor Bryce, but was found to contain nothing save a few fragments of bones. A slight examination led to the discovery of the ends of other slabs projecting from the gravel, and Sir lan Malcolm

Donalr inf.x E 1M d Campbell, Poltalloch. KiaraIrisn a s h Sainn wa live o dwh t A.D. 515-548. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 157 decided to excavate the site. In the third week of August I went to Poltalloc helo ht p wit worke hth . A plan of the gravel pit is shown in fig. 1. Some 11 yards to the north stand smonolita fee5 h inche6 t s abov e grounfee2 th e d t an d by 1-Jt foot at ground-level. Its major axis points north by east (fig. 2). Abou yard0 1 t s farthe re trace b norte fainn th dca ht remaina f o s mound whice promontory th s bee f ha h of n s onlt i raise cu t yI .o t d som inchee6 elevation tracee si b fee4 n width2 n di ca t y fronb d man , steee nea edge th th pf r easte o ban t curveth I .o kt s wit hwesterla y

Fig . Grave2 . t fro Northe Pi l mth .

cours r abouefo yard0 7 t s untis losi t ti l when turning south, som0 e2 yards to the south-west of the monolith. The area cut off by this mound measures about 100 yards north and south by 70 yards east and west. The work of examination was begun at the east side of the gravel pit, where flag-stones could be seen projecting from the side of the excavation, and where bones had been found some time before. Here four full-length stone-lined graves were found in an approximately easwesd tan t position. They contained human remains relicso n t .,bu No pointe1 . d east-south-east covee th , r stones bein foog1 inche7 t s belo formes e surfacewth wa f thret I o d . e slab eacn o o s htw sidd ean end slabs, being covere threy db unpaves e slabswa measuret d I .dan d 8 15 PROCEEDING E SOCIETYTH F O S , JANUAR , 192914 Y . 5 feet 6 inches in length internally by 15 inches at the east end, 18 incheshoulderse th inchewese 2 t 1 th a s t d a send an ,t containeI . da skeleto muca n ni h decayed condition e remain th e skul, th f l o sbein g ease th t t enda . e soutth n parallehy o No la , side a distancit 2 . t inches8 a o ,t l1 f eo . The west end had been damaged in removing gravel from the pit. Thi rathea s t a grav s r loweewa r level, bein feeg2 t belo e surfacewth . There were two slabs at the north side and one at the south side, a slab at the east end, and two remaining cover stones. It measured 8 inche seas widinchethe t 11 endat e smiddle and ,wid the heaeat the , d having evidently been placed at the west end. alignmenn i d an , 1 t. witNo , f hit o e easd closy th Nola en t o e3 t . slightla bu t a t y higher level, bein foog1 inche4 t s belo e surfacewth . This grave seemed to have been damaged in the making of grave 4, which had cut into it at the east end. The slabs had been displaced, but the west end slab, two slabs at each side, and one cover stone remained. Part of a skull was found at the west end. No. 4 was at a lower level, being 1 foot 9 inches below the surface. It did not lie parallel with the other graves, its axis pointing east by formes northwa f thret I o d . e slab t eaca s slabsd hd en sidean o , tw , three covers e dimensionTh . s were: lengt fee5 h inches8 t , widtt ha west end 18 inches, at shoulders 20 inches, at east end 10 inches. The skeleton was sufficiently preserved to show that the body had been s bacplaceit kn o d wit e armth h e wes sd hea th extendean to t d t a d each side. This grave was placed about 5 yards from the edge of the slope to the east and about 38 yards from the enclosing mound to the north. e graveNonth s pavedf e o deptwa s Th f eac.o hd bee hha n from inches4 1 1o 2t . There were in all probability more graves at this spot, but as these evidently date from medieval times, operations were not carried further e oldeTh .r e placenamth f o ,e Kil Kiarany l s suggestivi , f o e Christian burial skulle Th .s were sen Professoo t t r Bryc r examinaefo - tion,1 the other bones were reburied. Work was next begun at a spot 6 feet south-west of the cist C found in 1910 (fig. 3), at the south end of the gravel pit, and some 35 yards south-wes e lonsouty th b tg f o hgraves t thiA .s spoe proth t - montory ends in a slight knoll, sloping steeply to the east, south, and west horizontaA . l slab could her e seeb e n projecting froe th m grave escaped l ha face t I d. earlier interference fro face mth t thaa t larg trem el ee of considerably ove a century'r s growth grew directly Appendie Se . xB T NECKLACJE E FRO MCISA T POLTALLOCHA T , ARGYLL9 15 .

on the top of it. It was decided to open the cist A by removing the end slab, and to do this a considerable amount of roots had to be cut away. When thibeed sha n founs donwa dt e i tha slae th tb originally visible was not the cover of the cist, but a small slab placed close to it, the actual cover being directly behind it. When the end slab of the s removedcis s seewa t wa n t thai ,s formee cisth twa t f fouo d r slabs of schistcoverina and slab, end g sslab werThe . e placed betweenthe ends of the side slabs, and at the north end a shallow perpendicular

Fig . Cist.3 Gravet sa l Pit.

groove had been chipped in the west side slab. The end slab was not place thin di s groove coupla t inchebu ,f eo s outsid. eit The cist pointed north by east, it measured 3 feet 1 inch in length, foo1 inche8 t widthfoon 1 i s inche4 d t unpaveds an , depthn si wa t I . . The cover was 24 inches beneath the surface of the ground. The interio s abouwa r t one-third ful f san o graveld l dan , whics hwa bound int ocompaca tclosel masthe sby y matte treeelm d. rootthe sof To clear this out was no easy matter, as every care had to be taken to avoid damaging any urn or other relics the cist might contain. Only one person could work at a time, and it took six and a half hours e contentbeforth l al e s were cleared out. The best tool was found to be a piece of strong wire, bent at right angles, with which the roots were slowly combed out. The sand and 160 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929. gravel were passed throug e fibrou th t ha sieve bi d f san , of root t cu s by bit as they were freed. The contents of the cist were a jet necklace, a flint knife, fragments of partially incinerated human bone teeth,d smalw an s fe la 1 piecef o s charcoal, and small lumps of ochre.2 The bones and charcoal were all found in a circular area in the middle of the north half of the cist. The beads and the knife were found in the south half of the cist, and more to the west than to the east side. This was the part most filled with roots, which extended almost to the top at the south-west corner. After clearing out cist A,] the remaining part of the knoll between the edge of the gravel pit and the top of the steep bank was examined by cutting narrow trenches acros . Durinsit g this work no sig therf no e having bee ncaira stonef no s possibls i spofound e s ot i n th t wa t ebu , that stone y havma se been removed whee th n groun s undedwa r cultivation. s foun wa fee 1 ) 1 d4 t Cis d (figsB tan 3 . inche3 1 y sla sout belot I f e cis. ho wth A t surface, and pointed slightly east of true north larga covee s Th e.rwa slab, measur- ing 6 feet 9 inches by 3 feet 5 inches ; it was about 5 inches thick, but tapered con- siderably towards the edges. It was much to e cisto th lon ,r projectingfo g southwards s muca fee2 inches4 s ta h .f o Thid en s e slas onlth bwa feey2 inche3 t s wided an , Pig. 4. Cist B. Showing groove in side sla t north-easba t corner. a small slab lay over it at the south-west e cornecistth ,f o wherr e largth e e slab little more than covered the cist. A large number of markings made by a sharp-pointed tool were noticed on the under surface of the cover neae nortth r h end, apparentl o reducyt s thicknesit e t thia s s part. Slab beed sha n e th lai f do horizontallp to e e leveth th f o t l ya e northth cist n o , eastd wesan , t e sidescisTh t . measure 3 feed t 9 inches lengtn i h internall d fee2 inchee an nort3 y t yth b d t hen a s 1 foot 9 inches at the south end. It was 1 foot 7 inches deep, and was paved with 69 small, flat, water-worn stones. The cist was very symmetrica slabd formen n i l s e beinth , g fitted betwee e sidnth e slabs. Appendie 1Se x B. 1 Mr George Bond reports that the charcoal is that of the oak. Mr G. W. Tyrrell, A.R.C.S., Ph.D., reports that some fragments consist entirely of ochre, others are quartz pebbles, with ochreous stain and impregnations. He suggests that the occurrence in the cists is fortuitous. As explained below, however, I think the ochre was purposely laid in the cist. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 161 The d groovelatteha t eaca rd sen h simila o that r cisn . i tThes A t e were from. lIo ft j inch wid ^y eb inc beed h ha deepn d madan , e with a tool similar to that used on the cover. In fig. 4 the groove in the north-eas tseee corneb ny directlma r yfoote abovth f -e righo eth d ten rule. The groove on the south end of the east slab was not continuous, but had been made for only 3 inches at the top and the bottom; as the slab at this end was slightly concave, a continuous groove had not been thought necessary. The comparative narrowness of these grooves, slad bface en th placee groovee t th d th tha s n an t dneithei a t wa , d en r a distanc t a t f J eo bu inche2 inc o ht s outsid , suggesit f o e t thae th t grooves may have been made for the attachment of a lining of wood.

Grooved slabs were recorde Deay b d n Mapleto cisto tw s n nadjaceni t 1 to that in which the slab with engraved axe-heads was found.2 In that instance the side slabs were stated to be " grooved to admit the end slabs." As the site of these cists is some 600 yards east of the gravel- probabls i pit t i , e that they wer esame madth ey b epeople.

The interio filles e cisth wa t d f o rwit h san graveld dan . This woul3 d see o havmt e bee t therpu n e befor e cove s originallth e wa r y placed,4 covee th s r a closely fitte ciste th d, leavin aperturo gn whicy eb h gravel could enter; neithe s there pavemenspacy th wa r an en i e t beloy wb whic hburrowina g animal could find admittance. On the pavement and chiefly in the south half of the cist were found fragments of unburnt human bones and teeth, a piece of flint (fig. 5), small pieces of charcoal, and some fragments of ochre. An urn e food-vesseth f (figo ) 5 . l type stood nea e north-easth r t corner, slightly tilted towards the corner. It was in a damaged and fragile state, part e sid oa th distinegratefe n whici d y brokela ha h f of nd e statth n i e

interior. It would seem likely that the tilting and breaking of the urn S1 o far as I know, this is the first record of a stone cist lined with wood. Canon Greenwell records the use of wood in rare cases in Yorkshire (not accompanied by stonework) in the for splia hollowe d f mo an t d tree-trunk alse H o. record e bottoth s f gravemo s laid with slabs of wood, the sides also having been lined, in other cases planks had been laid over the body.— British Barrows, p. 13. ' Proc. Soc. Ant. Scot., vol. viii., part ii., p. 378. Canon Greenwell records the cover stone of a, cis t Eglinghama t , Northumberland, "exhibitin e almosgth ta uniquf o e eus feature th f o e tool upon it, in the shape of a groove cut with a sharp-pointed instrument round its narrower end doubo n , t wit e objechth f facilitatino t s beingit g dragge e mordth e e sideasilf th eo p yu the hill, upon the summit of which the cairn is placed."—British Barrows, p. 418. . CrawforS . G . recentls O dha r M y Scille describeth f yo Islee cison da s n wheri t e there were groove side 3 th e n slabsi s allo"o t end-stonee wth fittee b o dst more securely."—Antiquity, vol. ii . (Decembe9 41 . p r 1928). The unusual feature of a cist having been filled with soil at the time of burial was observed in two4 other instances in the Poltalloch district. At the Glebe Cairn, Kilmartin, a cist was found half-filled with gravel, it contained a necklace above the urn (Proc. Soc. Ant. Scot., vol. vi. p. 340). At Duncraigaig, in a cist nearly filled with gravel, the urn was found on the surface of the gravel {Ibid., . 347)p . 1 1 VOL. LXIII. 162 PROCEEDINGS OE THE SOCIETY, JANUARY 14, 1929. occurred when the wood lining decayed and the weight of the sand d gravean lslabn d againsminutforceA ur en .e e th dth te fragment of bronz founs ewa cleanin n e intersticedi e san th th n dt e i th gou f so stones forming the paving of the cist. An examination of the gravel below the pavement showed that it had not been disturbed. Before leaving the site a further examination was made of cist C, whic openes hwa n 1910 di e s cisriddled e santh wa t .Th n t donli bu , y a piece of ochre was found.1: This cist was formed of four slabs and slabd a coveren se beinth , g fitted betwee e sidth ne slabs o sign f , o n grooves in the slabs could be found. It measured internally about 3 feet 4 inches by 2 feet 1 inch, and was about 1 foot 10 inches deep. The cover was 12 inches below the surface, and the long axis pointed slightly to the east of true north. A short trench was cut from the monolith at the north side of the gravel-pit, for four yards to the south. A layer of small stones 9 inches e soidees foundth l pwa t t beebeneat bu no , nd disturbedha h . This layer, form a rins g roun e stoneth d , causin a gsligh t e mounth n o d surface e gravel-bedTh . e monolithth s a doet e extenr th no ,sfa o s d soil bein loama f go y character.2

THE RELICS. knife Th - pointeea s (figi ) 5 . d flak f ligheo t brown 2s i flint£ t incheI . s in length and shows signs of wear along one edge. The piece of flint (fig. 5) from cist B is roughly sectoral in form, it measures f by f inch, and shows chipping along the curved edge. The urn (fig. 5) is composed of light reddish-brown clay of fine texture and is somewhat rudely made. It doet stanno s d evenls baseit n ,yo bein incheg5 side s higon e t a h and half an inch more at the other. It measures 6^ inches across the mouth and 2f inches across the bottom. Two mouldings, f inch apart, encircl e urneth , the upper being Z\ inches belo e lipwth . e exterioTh s entireli r y covered with ornamentation consistinf o g some forty row impressionf so s which encircl urne eth . These impressions have been made wito toolstw h , each impression having been- made separately. One tool with a point like that of a penknife has been pressed flat int f V-shapeclae o oth mako yw t ro ea d impressions 9o t 5 , n incha o t ; immediately below this zigzag lin s anotherei f inverteo , d Vs. Eight of these double lines encircle the urn, being approximately 1 The presence of ochre in cists has been previously recorded. My own experience leads me to think tha t wouli t muce db h more often foun carefullf i d y looked for largA . e sla f ochrebo , wite sidon he much hollowe rubbingy b d s foun cisa t wa Chesterknowes,a n ti d , Chapelhill, Cockburnspath in 1913 (Hist. Ber. Nat. Club, vol. xxiv. p. 181). 1 A short account of other sites examined will be found in Appendix F. T NECKLACJE E FRO MCISA T POLTALLOCHTA , ARGYLL. 163 equidistan incf paid p h emphasiseto s an taparti r e Th . beiny db g more deeply impressed and by having the Vs wider apart. Bach pair of rows is separated from the next by 2 to 5 (usually 3) rows of impressions made with a tool having a point like a bluntha f o tt lead pencil, the dots being from ^-^ 8 to 10 to an inch. Where the impressions have been made obliquel e effecth y t resembles that a mad y b e twisted cord. The inside of the lip is decorated by two rows of hoi-izontal thumb- nail impressions, the impres- secone sion th bein w n si d ro g reversed;1 on each side of thithree sar e row f dotteso d impressions decoratioe Th . n musn ur te ohavth f e occupied some considerable time, there being between 500d 6000an 0 separate impression. it n o s The necklace (fig. 6) is of jet or some allied substance such as lignite or cannel coal. It consists of six plates, one triangular piece of the type usually called a pendant and 110 fusifor r barrel-shapemo d beads. • The plates are ornamented with rectilinear designs forme f rowdo f smalso l cup- shaped punctulations. Some of these still retai whitna e Flintsd an n .FigUr . 5 . chalk-like substance which must have made the design a much more striking e featurth f o e necklace in its original state.2 terminaTho etw l plates bea lozenga r e Thumb-nail impressions similarly placed on the inside of the lip were found on an urn at Greenhill1 , Balmerino, Fife. Another urn in the same cairn had V-shaped impressions and rows of producef doti s teetse a comb.—a th f hy o d b Proc. Soc. Ant. Scot., vol. xxxvi . 635p . . examination A 2 necklacee th f Scottise no th n si h National Museum show same sth e fillino gt exist in the punctulations of a necklace from Lunan Head, near Forfar. Similar encrustation 164 PROCEEDINGS OF THE SOCIETY, JANUARY 14,. 1929. in double outline containing a saltire. At the base they are pierced with three string-holes which emerge at the back of the plates. The right terminal e wearer' (i.e.e platth th n eo s right side s piercei ) t a d the point with two holes placed on the median line, passing through from back to front. The point of the left terminal is pierced with one hole •which enters at the back and emerges at the point. The second

_3 Inches.

Poltalloce FigTh . 6 . h Necklace. plates bear a lozenge in double outline, and are pierced from edge to edge with thre e f whice broadeo holes th e ht on a , emergero tw s a s edgee fronTh t. plates beao triangletw r n doubli s e outline, placed apex to apex; they are pierced with four holes passing out as seven. may be seen in the impressions forming the design of several Bronze Age urns in the museum, and has been recorded on the urns of the Stone Age in Denmark. The plates of a necklace found t Assynta , Sutherland, wer e"e saicuriouslb o dt y studded with gold spots " (Arch. Scot., vol. iii. p. 49). The punctulations of this necklace, however, which is in the National Museum, appear to be filled merely with earth. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 165 The so-called " pendant s triangula"i r wit ha slightl y concave base, two holes on a line at right angles to the base pass from one flat side to the other (fig. 7, a). 7 The beads var lengtn yi h from T ¥ inc l-o ht ^ inch. The necklace appear almose t quitb o no st f eti complete.1 As the beads and plates were slowly extracted from the mass of roots in which they lay, a note was made of the approximate position in which the larger pieces were found. It was impossible to do this with much accuracy, the mass being so dense, and the conditions of excavatio o unfavourables n e worth , k havine donb o et g froe mth farther end of the cist. Every bead and every perforation of the plates was threaded with a tiny rootlet. When the notes were examined later e positioth , n whici n e plateth h s were found corre- sponded wit usuae hth l constructio t necklaceje f no s much more closely than was expected. The relative position of two plates was reversed, this might be due to conditions of excavation or to the previous action e rootse terminaoth f Th . l plates were much closer together than as usually figured, this was also at first attributed to the pressure of the roots. Seventy of the beads were found before the first plate appeared: s thoughthiwa s e e actioo havt th th t f o eo nt beee du n roots carryin e smalleth g r beads away fro e plates positiome th Th . n e "pendant,oth f " however mors wa , e difficul explaino t neay t la rt I . e terminalsth r froe fa wes mth t e cistt tneano th ,e sidno ,f th ro e fron e necklacet th par f o tt I coul t .hav no d e been carried theres a , e oppositth n e growt i e root th s eth f directionwa sho e characteTh . r e piecoth f e itself doet suggesno s s beet ha tha a npendant t i t A . pendant is naturally the most valuable constituent of a necklace, upon which most car s bestowedei . This piece, like other similar piecess i , not only devoid of any ornamentation such as is found on the plates, s disfigurei t bu y beinb d g pierced witholeo tw h s which thers ha e been no attempt to conceal. e necklac t firsa Th s t wa reconstructee d accordin e accepteth o gt d Of some 52 necklaces of the plate type on record from the British Isles (36 of these bein1 g from Scotland), more than halmere ar f e fragments onld an ,y abou dozea t n examples have been found in a condition even approximately complete. Although this must in part be due to faulty excavation, carefully examined cists have been found to contain no more than a few plate beadsr so . Wit hmateriaa nature th f jetf o le o t mus,i t have been seldom tha necklaca t e survived till the death of its owner without loss by breakage. That the ornament was frequently wor s cleani r fro e extenm th e bead f weath d o plate t n an so r s from constant frictione Th . e necklaceth l finesal f s thai o st t foun t Balcalka d , Tealing e countth n i ,f Angu yo s (Proc. Soc. Ant. Scot., vol. xiv. p. 260), but the Poltalloch necklace must be given a high place in the first half-dozen t containI . fusifor0 s11 m bead Balcale againsth s sa n i 0 k128r tnecklace14 (o )0 12 d an , from f Burglethao tm foun,Da Raffordt a d , Morayshire (Proc. Soc. Ant. Scot., vol. xii . 298).p . Three other necklaces have over 100 beads, and six more have between 50 and 80. Five English necklaces have between 50 and 80 beads. 166 PROCEEDINGS OF THE SOCIETY, .JANUARY 14, 1929. plan, while an investigation was made of the evidence upon which this reconstruction was based. The evidence was found to be peculiarly slight. The facts relating to the discovery of Scottish examples have been collecte Grahar M y db m Callaiider; Englise th J h record more sar e widely scattered. The jet plates of a necklace found at Assynt, Sutherland, were figured in 1824,2 but no attempt was made at arrangement, the

I in.

Fig. 7. Necklaces and Beads. plates in fact being supposed by Hibbert to have been suspended from a girdle. In an account of a necklace found at Pitkennedy, Aberlemno, county of Angus, in 1858,3 the small triangular piece4 was thought to "bn ear-ring.ea " Ten years earlier, however, in 1848, Thomas Bateman had published his Vestiges of the Antiquities of Derbyshire, showing an arrangement

1 Proc. Soc. Ant. Scot., vol. 1. pp. 201-40 (1915-16). 2 Arch. Scot., vol. iii. p. 49. 3 Proc. Soc. Ant. Scot., vol. . 78 iii . .p * This t beetyp no beaf neo s dealdha tBecr witM kClassificatioy e h b ( "Th Nomenclaturd nan e of Bead Pendantd san s "—Archceologia, vol. Ixxvii. 1) . p . JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 167 necklaceofjet s thasinchas t e been accepte correctdas employeand , d in all our textbooks.1 In figuring a necklace found at Cow Low, near Buxton, placee h e smalth s l triangl backe t strunth bu t ,e a g through

as a bead2 . In Ten Years' Digging in the Celtic and Saxon Gravehills of Derby, Stafford, d York,an publishe 1861n di e sam,th e writer places the small triangle as a pendant in a restoration that is obviously fanciful (fig. 7, b). It has since been usually figured in this position in

necklaces both of3 the plate and of the disc type, but in an account of a necklace found at Holyhead, Anglesea, it is described in 1867 as "a triangular object the intention of which has not been ascertained."4 In 1870 a triangle was figured as a pendant in a necklace from Tayfield, Newport, Fife. There is no evidence in Bateman's records to show

tha beade tth s were actuall5 y positiofoune th n di n figured. A similar absence of evidence exists in the reports of the Scottish finds, in only two of which does the finding of a necklace in an approxi- mately complete condition appea o havt r e been recorded wity an h care. The Balcalk pendant (c. 1879), referred to above, is stated to have been found below the central part of the necklace. The second careful record is that of a necklace found at Burgie Lodge, Rafford, Morayshire, t beadinje 7 1913s10 "f , ove wheo wer 0 t 4 r nou e e positiofounth n i d n apparently occupie e breasth y r neckdb o t seemingl d an , y som f theseo e closely retained their original position, as two groups of four beads each forme star-shapea d d design." Unfortunatel positioe e yth th f no small triangl thit n recordedei sno fins dwa . Englise th n I h discoveries remaine th , s have been more fragmentary, and the details of the relative position of the beads are similarly meagre. Canon Greenwell records at Weaverthorpe, in the East Riding of York- shire, a "triangular pendant" of jet found at the middle of a necklace of graduated discs. He also describes a " pendant" of rather a different

form at the back of6 the neck at Groodmanham, also in the East Riding.7 thin I s case there wer otheo en r beads t PainsthorpA . e Wold, York- shire. MortimeR . J , r foun skeletoe dth youna f no g person, "the head pointe e northsmala th s behindd o dt lan wa , triangulat i r pendanf o t 1 "It iscoursef ,o , almost impossibl rearrango et egroua beadsf po , often more tha nhundrea d in number, in the exact order in which they were originally worn ; there are, however, frequently several peculiarly formed plates found wit beade hth s which seem susceptibl beinf eo g arranged particulae on t inbu r order thao s , t appeari t s probable tha mannee th t whicn i r h som f theseo e necklaces have been reconstructed ... is not far from being correct."—The Ancient Stone Imple- ments, Weapons, Ornamentsd an f Greato Britain, . 457p , John Evans (1897)Munror D . , how- ever, seem havo st e preferre e reconstructioth d e necklacth f no e from Mount Stuart, Buten i , whic terminae hth l plate broughe sar t togethe back.—e th t ra Prehistoric Scotland, 212. p , fig3 .13 (1899) alse Se o . Proc. Soc. Ant. Scot., vol. xxxviii fig, (1903-4)0 66 .4 . p . . « Vestiges, p. 92.3 3 Page 25. 4 Archaeological Journal, vol. xxiv. p. 257. 6 Proc. Soc. Ant. Scot., vol. viii. p. 412 (1870). • British Barrows, p. 198 (1877). ' Ibid., p. 297. 8 16 PROCEEDING E SOCIETYTH F O S , JANUAR , 1929Y14 . jet" beado ;n s were found figure Th e. show trianglsa type th ef eundeo r consideration.1 evidence th , up em T froosu m actual discoveries, pointmeagr, is t i ss ea strongls a triangle th o yt e having bee ntoggla r fastene eo bace th k t a r pendana s a havins frontfacit n e ti f o tTh g. been accepte pendana s da t might bias subsequent records. The lack of ornamentation and the fact of the holes being usually unconcealed, support the fastener theory.2 Used as a fastener the piece would be attached to the point of one of the terminal plates, and would connect with a loop attached to the point of the other terminal. The holes in the Poltalloch necklace are peculiarly suited for this reconstruction. The point of the left terminal has a single hole; to this the loop would be attached. The right terminal has two perforations corresponding to those of the "fastener," as we may now call it. Other necklaces vary somewha e fastene th e for f th d o m n an i rt in the perforation of it and of the terminals. Usually the fastener is triangular, with one hole near the base or with two placed close together, either parallel to the base or at right angles to it. The finest example, from Balcalk, has a V-shaped hole at the base. The Poltalloch fastene e onlth r ye seem b exampl o t s e havin a gconcav e baseA . necklace foun t Kyloea d , Northumberland, uniqua 192n 3i s 7ha e billet- shaped fastener (fig. 7, e), which Mr Parker Brewis designated a " toggle- shaped piece," but which he described and figured as a pendant. The "pendant" of a necklace found at Lunan Head, county of Angus, is

described as being cubical in form with rounded angles; it is now lost4 . The metho f piercindo terminae gth l plate also varied hole Th e . some- times entered at the back and emerged at the point, and sometimes one or two holes passed directly through the plate from front to back. Structurall lattee yth r metho probabls dwa beste yholee th ,th s beinn gi line with the point. This distributed the strain, the cord being threaded through plate and fastener in a double figure of eight. The signs of wear e Poltallocth n o h fastener clearly show this method f threadin,o g (fig, .7 a (?/)). That the fastening was always a weak point, however, is testified

1 Forty Years' Researches Britishe th Saxond n i an Burial Mounds of East Yorkshire, . 128p , fig. 323 (1905). Professor Macalister, however, describes a triangle, found in a cist at Oldbridge, County 2 Meath, wit hnecklaca discoif eo cylindricad dan l beads "s ,a pierce d wit hhola e running froe mon broad face to the other, so near the base of the triangle that the pendant would hang apex down- wards doubo n . e Therb t than eca t this objec s suspendewa t d fro necklacee mth thad t an ,i t was an amulet."—Ireland in Pre-Celtic Times, p. 193(1921). This triangular type has also been e Continentfounth n o d . 3 Archceologia ^Eliana, vol . (4t.v h series) (fig.)6 2 . ;p , Proc. Soc. Ant. Newcastle, seriesh 4t , vol. iii. p. 94 (fig.). 1 Proc. Soc. Ant. Scot.; vol. xii. p. 294. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 169

by the number of terminals that are broken and sometimes redrilled at the point. consideree b Wha y e trianglth ma t f a do s e finaa us l e prooth f o f fastene e presencth s i r f distinco e t signe e edgth f wea th o f s o et a r base on either side of the hole (fig. 7, a (x)). This was caused by the friction of the loop. It is most clearly seen in the fastener from Tayfield, Newport, Fife,1 but it is also present in these from Poltalloch; Blinmill, Rothie-Norman, Aberdeenshire; Mound an 2 t Stuart, Bute,l al 3 e Scottisith n h National Museum. These cord mark e alsar so very clearly shown on the figure of a lozenge-shaped "pendant" pierced at one end. It was found by Mr Mortimer4 at Painsthorpe Wold, Yorkshire s possibli t I . e tha a semicirculat r bead wit hgrooveo "tw s across the flat top," described but not figured by Canon Greenwell,5 owed that feature to the wear of the loop. It is described as the "central" bead in a string of 124 discs of jet. If, then, we accept the triangle as a fastener, we must find another arrangement for the string of beads. Here the evidence from Burgie Lodge comes to our aid with 40 beads placed in front and two groups obead4 f s formin a gstar . This clearly point a constructio o t s n like that shown in fig. 6, and suggests that the 70 beads found together at Poltalloc t bee t firsno a n d s te roots shiftehha thoughtwa th s y a ,b d , but were actually in situ. The " star" formation of the Burgie Lodge necklace needs three string f beads o se fron Th . t plates, however e piercear , r sevefo d n beads. This must have required an arrangement somewhat similar to that suggested in fig. 6, where, at either side, each of two pairs of beads is connected with a single bead, the latter being connected with the middle string. This construction is of course of no structural value to the necklace, and must be regarded as an artistic development, the fringe of seven beads below each front plate having a pleasing effect.8 Further evidenc "starthe of e" formatio founis n signin d obliquof s e wear at the points of some of the beads. The genera le necklac reconstructeeffecth w f no o ts ea s strongli d y

1 Proc. Soc. Ant. Scot., vol. viii . 411p . . a Ibid., . 203 p vol . . vi . 3 Ibid., vol. xxxviii . 66 . p . 4 Forty Tears' Researches, . 126p , fig. 312. ' British Barrows, 330. p . • With some necklaces (Pordoun House, Kincardineshire, Proc. Soc. Ant. Scot., vol. 1. p. 212; Pitreuchie, Angus, Ibid., vol. xli . 65p . ; Tayfleld, Balgay, Newport, Fife, Ibid., vol. viii . 411p . ; Kyloe, Northumberland, Arch. JEL, Lowseriesh w 4t Co , Buxton ,d . 26volp an . ; v . , Derbyshire, Vestiges, p. 92) there have been found from one to four triangular plates resembling terminal plates, but much smaller, and having one hole at the point and two at the base (fig. 7, /). These have in some cases been thought to belong to another necklace, probably that of a child. It seems likely, however, that thehavy yma e been use r connectinfo d e extrgth a e beadth f o s fringe with the main strings. See Appendix C, Details of the Construction of the Necklace. 170 PROCEEDING E SOCIETYTH F O S , JANUAR , 192914 Y . suggestive of another ornament of the Bronze Age, the gold lunula or crescent.1 This striking similarity appear e confirt b onl no so t y- matory of this reconstruction of the necklace, but also to provide the explanatio f wha o ns bee ha ta nmuch-dispute d question—the origin d meanin an e ornamentatioth f go e lunulce.th f o n Before considering the connection between these two ornaments,

J.M.CiowlJZS^ Fig. 8. Map of Distribution of Jet Necklaces and Lunulce.

it may be of use to examine the distribution of each. This is shown on the map (fig. 8). The necklac f crescentieo c form f musvero e yb t early origine Th . first developmen simplthe of et strin the chilshell the of g or d of s savag grado t shellse s i eeth , wit largese middlee hth th n i t : thit sa once forms a crescent-. A long process of evolution, however, must take

1 Although the designation "crescent" is in some respects preferable, "lunula" is here used, the ornament having been usually referred to and indexed under that name. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 171 place before such an artistic ornament as the plate necklace is produced. The use of the spacing bead is a notable step in this development.1 Wha almosy ma t t certainl consideree yb prototype th plats e da th f eeo necklace is found in Denmark, as a relic of the Stone Age. This is an amber necklace consisting of two terminals and four spacing beads all pierced for five strings of cylindrical and discoid beads (fig. 7, d).2 The figure show a ssmal l piec f ambeo e e th f ro attachee on o t d terminal havy ema t beei s; ntoggla smale th eo l t simila trianglee us n ri s t necklacesje r oou f reconstructione th ; , however conjecturals i , , Amber necklaces with spacing beads have been found in Wiltshire, the amber, if not the finished necklace, having probably been brought fro e Baltic mth e platen ambe a Th f . o s r necklace r figureRSi . y b d Colt Hoare3 are graduated, the terminals are not triangular, but show six holes at the smaller end (fig. 7, g). It is in jet, however, that the plate necklace has attained its highest development. No plates of jet seem to have been found on the Continent, and the records from Ireland are extremely meagre.4 In Britain their range extends from Orkney to Cambridgeshire, but while e nortinth h they greatly excee numbern di secona s d typ f necklaceo e consisting chiefly of small discs of jet, the latter preponderates in the south r JosepD . h Anderson states that frequentlo "s the e yar y found with Bronze Age burials in Scotland and so rarely in any other part of Great Britain e thacharacteristi b e sai b to t dthey e ma yth f o c Bronze Age in Scotland." A list of those found in Scotland has been

r GrahagiveM y nb m Callande s papehi n ri r mentione6 d above. They

are found associated with urns of 0 the food-vessel type, and also with beaker urns, indicating use at the beginning of the Bronze Age. The English examples are chiefly from Yorkshire and Derbyshire, remarkable ar d an whew ylarge fe nth e numbe excavationf ro thesn si e countie considereds si . Mos f theso t e English plate necklaces have been extremely fragmentary featureA . , especiall f thosyo e from Derbyshire,

1 It is found at Ur (fig. 7, c), dating from 3500 B.C., when it was made of s\x>ne*(Archceologia, vol. Ixxvii. p. 14, fig. 15, A3, 61—" The Classification and Nomenclature of Beads and Pendants "—

Horace C. Beck). 2 Afbildninger af Danske Oldsager og Mindesmaerker, A. P. Madsen (1869), pi. xlii., No. 33. e AncientTh 3 History f Southo Wiltshire (1812), pi . e alsiiiSe o. Archceologia, vol. xliii.

p. 504 (1865). 4 Ireland Pre-Celticn i Times, Macalister (1921)2 19 . .p , . ' Proc. Soc. Ant. Scot., vol. xxxv. p. 270 (1900-1). r CallandeM • e followinrth informf o e g m stha t have s knowledgcomhi o et e since eth publicatio s listhi f a necklac: no e from Angus plata , e from Orkne 1928n yi , several plated san beads (apparently representing three necklaces) fro Spottiswoode mth e collection (Berwickshire) (Proc. Soc. Ant. Scot., vol. Iv 20). necklacp .A . e foun t Wesda t Morriston, Berwickshire 184n i , 6 (Hist. Ber. Nat. Club, vol. ix. p. 49), and a necklace in a cist near Pluscarden, Moray (Proc. Soc. Ant. Scot., vol. Iviii . 239)p . . 172 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929. is the frequency with which beads of an alien type, such as discoid beads, r conicao l studs wit hV-perforationa , have been foun associationn i d . This would suggest that the source of supply was far removed, arid that breakage replacee b o t d d sha wit h whatever substitut availables ewa . o recordtw e s Th from Cambridgeshire contained respectivele on y

and two plates. Of five from Derbyshire one has the" plates made of 2

bone, but simila1 r in type to those of jet. The others all contain to some extent beads of an alien type. Yorkshire supplies six examples, e mosth t complete containing seven beads.o Northe platen t Th sbu -

umberland necklaces, three in number3 , all come from near the Scottish Border, one in fact being found north of the Tweed. The Welsh example s fairli 4 y complete, with ,the additio a conica f o n l stud. These f courso stud y somn i e ma s e cases have been dress-fastenersy ma d an , t actuallno y have forme dnecklacee parth f to . The evidence from the discoveries of plate necklaces would seem to indicate that the type originated iri Scotland, being developed probably fro a mScandinavia n prototype, thae souther s taketh t iwa o tt n n part of the island, but that very few found their way to Ireland. If there was a single source of manufacture, it was probably situated somewhere in the vicinity of Forfar. Lignite could be got from the adjacent county of Fife, if not nearer. With regard to the distribution of the lunula, this has been else- where clearly set out6 (see map, fig. 8). The vast majority have been found in Ireland, from which we have more than 60 examples. England supplie , fro4 s m Cornwall; Wale , fro1 s m Carnarvond an ; Scotlan r probablo n com5 d Te e . y6 fro e Continenmth t (Brittan, y6 Belgiu , Hanovem1 , Denmar1 r . Thek2) e consideree ar yth e b o dt earliest, as well as the most numerous, of the gold ornaments found in Ireland, and are admitted by all authorities to belong to the period of the flat bronze axe. The evidence, which is entirely derived from one discovery, point thio t s s early t cannoi date t admittee bu , b t s da conclusive.6

Appendie 1Se x D, Englis t NecklaceshJe . 2 Prom Windle Nook, Hargate Wall. See Vestiges of the Antiquities of Derbyshire (1848), p. 89 (fig.). 3 From Middleto Woldse th Forty e n ni Se . Years' Researches, . 353p , fig. 1017. 4 From Holyhead, Anglesea Archceologicale Se . Journal, vol (1867.7 xxiv25 ,. p .fig.) . 5 See "The Distribution of Early Bronze Age Settlements in Britain,"'by O. G. S. Crawford, Geographicale inth Journal, vol . 18 xl4pp . and-304 (1912) Bronzee Irelandn i Th :e Ag (Coffey), pp. 47-55 (1913); Armstrong's Catalogue of Gold Ornaments, p. 12 (1920). 8 The facts are these: in 1864 a labourer came home from a day's digging near Harlyn Bay, Cornwall, wit golha d lunula roun e calf eacdth o f alsd h ha oleg e founH . flaa d t bronze eax somd an e other objec objectr o t s whic e threhh wplace awayth t e A beewherd . ha n e diggineh g• were some stones, thought at the time to have formed a place of concealment and since con- sidered to have been a cist. Although the likelihood is that the lunulce and the axe were JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 173 Muc bees hha n writte attempe th n n i explai o t t ornamentatioe nth n and use of the lunulce. Most authorities are now agreed that they were used as collars, and not as a head-dress as was formerly supposed.1 M. Salomon Reinach2 ascribed to Ireland the origin of all lunulce in north-west Europe. Although M. Joseph Dechelette3 has expressed the opinion thae Danisth t h lunulce were mad n Denmarki e , being copied from Irish models, M. Reinach's view has been accepted by most authorities, including Mr Reginald Smith, who in 19204 supported the religious origin as discussed by M. Camille Jullian. Mr Smith takes the

view that "the most obvious interpretation lunulcee th f "o thas i t the5 y are to be taken as lunar symbols, being connected with the worship of the moon. "Taken at their face value the crescents represent the moon. suggeste H " s tha restrictioe th t ornamene th f pointe no th o t st of the lunulce, for which "no explanation seems to have been given," s sacrifice origiit th d n ni bullsf ha o e horne ,th f whicso separatee har d foreheade byth . Professor Macalister says tha tArmstronr "M g seems incline o adopt d t this hypothesis I confes t t i seembu , e s m thas o t t needlessly speculative." e doeH 6s not, however, wholly renounce th e religious explanation: after remarking that " the decoration is dis- tribute remarkabla n di unexpected ean d way, concludee "h s thae "w t must infer thadecoratioe th t t leasa s s muca t nwa h magica s ornaa l - mental." Mr Armstrong in 19207 suggested that "possibly the crescent-shaped form may have been influenced by the amulet composed of two boars tusks placed together."

deposite dmean o togethern y t b proves s certainno i t s i i ,d t I tha constructio.e tth cista s ,nwa t cleai no s i r r tha relice th t s wer actualll eal . it Case n yknowi e sar n where periodl urnal f so s e Bronzoth f have eAg e been same founth en i d buria l moun sam e eved th an en n i d; cis t have been found a food-vessel and the remains of a beaker belonging to an earlier interment. It may be added that little importance can be attached to additional evidence taken fifty-two years after d seee sit ha e eventth n eth o afte e findwh , th r froe s on mwer e made e ArchaeologicalSe . Journal, vol. xxii. p. 275 (1865); Antiquaries Journal, vol. i. p. 294 (1921). What is a much more convincing association was the discovery of a lunula and a gold ear ornament in the same burial mound in Moray, at Orton, near Fochabers (see Anderson's Bronze d Stonean Ages, Proc.d . 65p an , Soc. Ant. Scot.,.vol. , 1868) viii28 . .p . f Golsuco s di h rare occurrence in Bronze Age burials in Scotland, that it is highly improbable that these objects belon differeno gt t interments r ornamenea typa e f o eTh s .founi t bronzn i d e associated with perforatea d stone axe-hamme conicaa d t an bear lje d with V-perforatio t Cowlamna , Yorkshire (British Barrows, p. 222). At Largetreany, Donegal, what is considered to be a fragment of a lunula s foun wa 187n i d 7 wit hoarha d which included tore othed san r gold ornamentse Th . details of most finds, however, are meagre : " with bones," "under a large stone," or "in a bog" being all the information available. 1 Mr Leeds, however, considered that the gorget theory was inadmissible, and the lunula was a woman's head-dress. See Ant. Jour., vol. i. p. 138(1921). 2 Revue Celtique, 1900, pp. 75 and 166. 3 Manuel d'Archceologie, vol. ii. p. 353-4 (1910). 4 Antiquaries Journal, vol. i. p. 131. * Journal des Savantes, p. 153 (Bordeaux, 1911). Archceologye Th « of Ireland, ' (1928)Catalogue6 6 . p . f Goldo Ornaments, (1920)2 1 . p . 174 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929. n describinI e decorativth g t eneck je motive e e plate-th th f f o so s laces r CallandeM , s pointeha r t thaou d t they follow " those seen no other periode relicth f so , axesmald an s l oval knive f bronzeo s , lunulce of goldd drinking-cuan , p urns." r Coffe M d 1carrie ha y e comth d - parison a step farther in 1909: "The centres of the lunulce are plain, the exact reason of which is not quite evident; the way in which the ornament is gathered to the ends and spaced by bands reminds us of t necklaceje e e plateth th f so s ornamented with triangle lozengd an s e ornaments which are ascribed to the end of the Stone Age or the early Bronze Age."2 The analogy, however t beeno ns closelha , y applied, nor'ha t beei s n use o helt d p e rearrangemeneitheth t n necklaci rje n e i th r o ef o t tracin e histore lunula.th gth r Coffef o yM y himself seem havo t s e departed from the idea; in his subsequently published Bronze Age in Ireland, n 1913i 3 e omith , l referencal s t necklaceje o t ed merel an s y states that " the centres of the lunulce are plain, the exact reason of which is not quite apparent." Returning, then, to the examination of the lunulce, we find a general resemblance between their ornamentation and the design of the neck- laces closeA . r comparison shows that some lunulce (fig ) bea9 . a r striking resemblance whic n hardlca h e attributeyb d eithe o mert r e coincidence or to a general similarity of type in all the ornamentation Bronze oth f e Age. The three plates which are placed at either end of the necklace are represented in the same position on the lunulce by three groups of ornament of similar proportions, which have as an edging a dog-tooth design resembling the points' of beads. A design of oblong figures separate narroy db w spaces along eithe re fron edgth f te o e parth f o t lunulce suggests beads as closely as was possible for an artist limited to f straigho e us te linesth . Lastly e fastenin fore th th ,f e samemo th s gi . The e fronreasoth r e tlunulce fo nth par f o t being left plain, except fo bead-lika r e edging s probabli , y becaus e lunulce,Proe th us -n i eo t .fessor Macalister's words, "no curvilinear patterns are used."4 Whether any pigment was used to decorate this space, it is of course impossible

1 Proc. Soc. Ant. Scot., vol. 1. p. 215. 2 Proc. Roy. Irish Acad., vol (1909) 2 idee . xxvii.25 Th a . . p quote . , 10 Pro Sectioy . db -No , nC fessor Macalister seems to be somewhat different: "This curious distribution of the decoration has been ingeniously explained by Mr Coffey as pointing to an origin for the idea of the lunula in suspensioe th chaino tw beadf f nso o s roun necke dth , which would hang loos oped efrontan n ni , but would mass togethe shoulders.e th n ro Irelande Se " Pre-Celticn i Times, . 140p . ' Page 48. e ArchaeologyTh 4 f Ireland,o (1927)4 6 . p representatioa . e b AVha y ma t f fusiforno m beads, however founs i a bronz, n o d e bracelet found wit a necklach t Melfora e d figurean t Dy b dr Anderso Scotlandn ni Pagann i Times . 57)(p . T NECKLACJE E FRO MCISA T POLTALLOCHA T , ARGYLL. 175 now to tell, but the finding of ochre in the cists and of white matter in the punctulations of the necklace plates, shows that the use and effect of colour was well understood early in the Bronze Age. Similarly, if e fronth te lunulce parth beed f o tha n covere ornamenn a y db f neto t - work or lace, no trace of this would remain. A close examination of the details of the lunulce, shows distinct sign decadencof s e fro typmthe e most clearly resemblin plategthe sof e necklaceth . This decadence suggest arrangemenn a s foun i t r classes as follows >~ CLAS . (ThSI e Plate Type) whic(fign i , e group.9) th h f ornameno s t approximat t e necklaceplateje th e f foro sizsn d th ei ean o m.t CLAS . (ThSII e Narrow Group Type) (fig. 10), with groups much narrower, usually four in number, sometimes five, on each side. In some examples the dog-tooth is much exaggerated (Nos. 15, 16, 17, and 18). CLASS III. (The Lozenge Type) (fig. 11) whicn i , e e pointth hth f o s dog-tooth meet, forming rows of lozenge-shaped figures. In some examples from this class the groups are broad, in others they are narrow mose formee th Th e t . elaboratelar r y decoratee th l al f do lunuloe (Nos. 36, 37, 38, 39, 40, and 41). No. 35 shows a transition form between this clas Clasd an s s II.pointe e dog-toot,th th f so h in only the front and second groups meeting. CLASS IV. (The Vertical Design Type) (fig. 12), differs most widely from e necklaceth , havin e ornamength t arranged vertically insteaf do horizontally. Transition forms o arrangementcombintw e th e s (Nos. 50, 51, and 52). For a time the front group is retained (Nos. 53, 54, and 55), but eventually all evidence of the plate origin is lost. Some lunulce are quite plain, or have merely a border of engraved lines. The great majorit f lunulceo y fall readil r otheyo intf e theso ro on e four classes. e decadencTh e beadeth f o ed border, which become a continuous s design o_f lines and zigzags, approximately accompanies the decadence of the ornamented groups. Thus of 5 Irish lunulce, with the edging most closely resemblin Clasn i g e 21)d beadsar an I.,, 3 , s20 , (Nos7 , 6 , .5 although that class containIris2 5 hf o lunulce st onlou 9 y having grouped designs. Eve n Clasi n s IV., however n occasionaa , l survival e bead-likoth f e border appear decadena n i s t form. Comparing now, in the light of this classification, the Irish lunulce with those from other countries, we find that only 9 out of 52 decorated 176 PROCEEDING E SOCIETYTH F O S , JANUAR , 192914 Y .

Fig. 9. Gold Lunulce. Class I. (J.) = National Museum of Antiquities of Scotland, Edinburgh. D = National Museum, Dublin. L = Liverpool Museum. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 177

Fig. 10. Gold TJunulat. Class II. E=Natlonal Museum of Antiquities of Scotland, Edinburgh. B.M = British Museum. = National Jluseum, Dublin. B = Belfast Museum. L=Liverpool Museum.

VOL. LXIII. 12 178 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929.

Fig. Gol11 . d Lunulcn. Class III. (J.) BM = British Museum. D = National Museum, Dublin. T = Truro Museum. JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL 179

J.H.Craw 1928 . GolFig12 .d Lunulce. Class ContinentaIVd an , l examples. BM Britis= h Museum = D Nationa. l Museum, Dublin. 0 18 PROCEEDING E SOCIETYTH F O S , JANUAR , 1929Y14 .

Irish lunulce can be placed in Class I. In Scotland 5 lunulce have been found; two of thesel have only a lined border, but a sixth of unknown locality should probabl added,e yb 2 being fro ma Scottis h collectiof no antiquities. These three or four Scottish lunulce with group ornament all belong to Class I. Even though this number be small upon which to base a theory, taken in connection with the decadent type of the great majority of Irish lunulce, it strongly points to the origin of the lunula being in Scotland and not in Ireland. When it is considered that the plate necklace from which the lunula is copie Scottisa s di h ornament almost unknow Irelandn ni Scottise ,th h origin of the lunula would appear to become much more than a sup- position. The much larger number of lunulce found in Ireland is the natural e ricresulth h f golo t d deposit f thao s t country l DoradE e th f pre, o - historic times.3 It is upon this greater prevalence alone that the theory e Irisoth f h e lunulaorigith f s no basedi similaA . r argument would place the origin of the rabbit in Australia, or might suggest to some future archaeologist that golf originate n Americai d significancee Th . , however, of an earlier type of ball being found in Scotland would sugges e inferencth te drawb o t en fro e facmth t thae decorateth t d Scottish lunulce belon o Clast g . I sWhe n more have been foune th d case will be clearer. As we have seen, Class I. consists of 3 or 4 Scottish lunulce and 9 Irish. In Class II. are 1 Welsh and 17 Irish examples. Class III. has 3 English (Cornwall) and 15 Irish lunulce. Class IV. is represented only in Ireland. Nothing is to be learnt from the distribution of the different classe Irelandn i s . Whe e Continentacome th nw o et l lunulce, the difficulty seems to increase. One of the Danish examples and that from Hanover are entirely plain, the second Danish example has merely an edging of engraved lines. There is no information available con- cernin frog3 m lefe Brittanyar t e wit w othee ho th ,s fro3 r m Brittany and 1 from Belgium, which all have group ornaments. They are shown . 64 d an , 63 , fign i , Nos62 ., 12 61 . The proximity of Brittany to the coast of Cornwall, and even to that of Ireland, would support the generally accepted theory of an Irish origin. On examination, however, we find no evidence of the features peculiar to Classes II., III., or IV. in these lunulce. They do approxi-. mate to Class I. in having broad clearly defined groups of ornament.

1 From South Side, Coulter, Lanarkshire, Proc. Soc. Ant. Scot., vol. iv. p. 291; vol. 1. p. 16, fig.; vol. Ivii. p. 162, fig. » Lanflne Collection (Ayrshire), Proc. Soc. Ant. Scot., vol xxxii. p. 240, fig. s See Appendix E, Gold in Scotland and Ireland. T NECKLACJE E FRO MCIS A T POLTALLOCH TA , ARGYLLi 18 . As in the case of the Scottish lunulce, we have the disadvantage of there bein gexample w sfe o r comparisonfo s e samth t e A tim. e w e must indicate what these examples suggest. It is unlikely that the likeness to Class I. is purely fortuitous, three possibilities seem to be indicated. Irisn (aA ) h connection which lasted only durin perioe gth whicn di h Class I. was being made in Ireland. (6ScottisA ) h connection. (c) A Continental origin for the type, spreading first to Scotland, the Ireland,o nt 5'and from ther Waleo t e Cornwall.d an s 1 This would involve the existence of a Continental amber necklace, approximating more closele Scottisth o t t ynecklac je h e thayey ntan e founth n o d Continent, having plates simila n typi r thoso t e e foun Wiltshirn i d e (fig g)., 7 . This hypothetical necklace woulprototype th e morde b th f eeo perfect Scottish jet necklace, and of the lunula of all countries. Again we can but await further discoveries to suggest a solution. Our thanks are due to Sir lan Malcolm for having placed the neck- lace and other relics on loan in the National Museum. The success of the excavations is due to his enthusiasm in the quest; no efforts were spared by him in giving every assistance that could be thought of, bot supplyinn hi e necessargth y personalaboun i d an r l interesd an t encouragement. I have als expreso ot thanky m s mano t s r informatioyfo held nan p variouin sauthoritie waysthe BritisTo .the sof h Museum Royathe , l Irish Academy e Belfasth , t Municipal e MuseumLiverpooth d an , l Museum, for permission to reproduce figures of lunulce in their custody. To Professor Bryce for his report on the bones; to Mr G. W. Tyrrell, A.R.C.S., Ph.O., and to Mr George Bond for their reports on the ochre e charcoal r Grahath M n o o T d m. an Callande s unfailinhi r fo r g help .in many directions e informatioth r fo s wel a ,s a l n derived fros mhi pape n Scottis. o CurirO r advic. t fo eA necklacesje h n r o e M o T . several points, and to Mr A. J. H. Edwards for help in the examination of necklaces and lunulce in the museum. To Mr R. C. Bosanquet and . EdgaJ M. r mucR r fo r h useful information lastlyDonald r M An o .,t d Campbell, Poltalloch, for local information, and to Mr Snelgrove for carefu enthusiastid an l cwore helth excavationf kn po i .

1 A Continental ornament which seems to have a common origin with the lunula is the bronze " diadem" found in Denmark. Several of these are figured by A. P. Madsen in his Afbildninger af Danske Oldsager og Mindesmaerker : Broncealderen I. (1872), pi. xxxi. The method of fasten- s beeha ng differentin e pointe orescenth ,th f o s f pit o .bein 6 xxxi t . offgcu No .. most closely resemble lunula,e sth havin groupo gtw f ornameno s t eithea t edged en r d with dog-tooth orna- ment e fronTh .t par f thio t s "diadem," instea f beino d g plain s ornamentei , d with rowf o s design suggestive of strings of discoid beads. 182 PROCEEDING E SOCIETYTH F SO , JANUAR , 1929Y14 .

APPENDIX A. BIBLIOGRAPH ANTIQUITIEP O Y THN SI E POLTALLOCH DISTRICT. ALLEN, J. ROMILLY.—"Notes on some undescribed Stones with Cup-markings in Scotland," Proc. Soc. Ant. Scot., vol. xvi. pp. 110-19 (1881-2). ———— "Notes of Sculptured Stones at Kilbride, Kilmartin, and Dunblane," Proc. Soc. Ant. Scot., vol. xv. p. 254 (1880-1). ——— "Note a Cis n to s with Axe-head Sculptures near Kilmartin, Argyll- shire," Journ. Arch. Assoc., vol. xxxvi. pp. 146-50 (1830). ANDBRSON, JOSEPH, LL.D.—"Description of Sepulchral Urns, exhibited by Col. Malcolm, C.B., of Poltalloch," Proc. Soc. Ant. Scot., vol. xxxix. . 232-4pp 4 (1904-5). BRUCE . J COLLINGWOOD.—, Incised Markings n Stoneo n Northumberland,i Argyle, d otheran places, from drawings made n 186i d 186 3an y 4b direction of Algernon, Duke of Northumberland (1869). BRYCE , M.A.H. . T ,, M.D.—"O e Cairnnth f Arran,o s " Proc. Soc. Ant. Scot., vol. xxxvii (1902-3)6 3 . p . . CHRISTISON, DAVID, M.D.— e Standing-stoneth n "O d Cup-markean s d Rocks, etc., in the Valley of the Add and some adjoining Districts of Argyle," Proc. Soc. Ant. Scot., vol. xxxviii. p. 123 (1903-4). ———— e Fort " Th f Kilmartino s , Kilmichael-Glassary d Nortan , h Knapdale, Argyle," Proc. Soc. Ant. Scot., vol. xxxviii (1903-4)5 20 . p . . ——— "e ReporSociety'th f o t s Excavation f Forto s s e Poltallocoth n h Estate, Argyll, in 1904-5," Proc. Soc. Ant. Scot., vol. xxix. pp. 259-322 (1904-5). GREENWELL, Rev. WILLIAM.—"An Accoun f Excavationo t n Cairni s s near Crinan," Proc. Soc. Ant. Scot., vol. vi., par (1866)6 t33 ii.. .p , MACGIBBO Ross.—D NAN The Castellated d Domestican Architectiire of Scotland (1887-92). MACLAGAN, Miss C.—The Hill Forts, Stone Circles, d otheran Structural Remains of Ancient Scotland (1875). MAPLETON, Rev. R. J., M.A.—" Notice of a Cairn at Kilchoan, Argyllshire, and Contents,s it " Proc. Soc. Ant. Scot., vol. vi., par (1866)1 t35 ii.. .p , ——— "Notic Artifician a f eo l Islan Locn di h Kielziebar," Proc. Soc. Ant. Scot., vol. vii. p. 322 (1867). ———— " Notice of remarkable Cists in a Gravel Bank near Kilmartin, and of Incised Sculpturing f Axe-heado s othed an s r Marking e Stoneth n so s Cists,e oth f " Proc. Soc. Ant. Scot., vol. viii., par (1870)8 t37 ii. . . p , ——— "Not a Bon f o ee Cav t Duntroon,a e " Proc. Soc. Ant. Scot., vol. x . p. 306 (1873). ———— "Notice of the Examination of a Cist on the Banks of the River Add t Crinana , Argyllshire," Proc. Soc. Ant. (1880-1)3 Scot.,10 . p .vol. xv . SIMPSON, Prof . Y.J . , M.D.—"On Ancient Sculpturing f Cup o sConcentrid san c Rings, etc.," Proc. Soc. Ant. Scot., vol. vi., Appendix (1866)1 . p , . STUART JOHN.—r D , The Sculptured Stones' of Scotland (1867). JET NECKLACE FROM A CIST AT POLTALLOCH, ARGYLL. 183 THOMAS, Capt . W, R.N.—P . L. . " Dunadd, Glassary, Argyllshire Place th , f eo Inauguration of the Dalriadic Kings," Proc. Soc. Ant. Scot., vol. xiii. p. 28 (1878-9). WHITE , , n.ft.—ArchaeologicalP. Capt . T . Sketches n Scotland,i Knapdale (1873-5). WILSON, DANIEL, LL.D.—"The Kiliuichael-Glassrie Bell-Shrine," Proc. Soc. Ant. Scot., vol. (1883-4)9 xviii7 e als. p .Se o . Archceologia Scottica, vol. iv. pp. 117-25 (1827). In 1915 the County Council of Argyll compiled a List of Ancient Monuments and Historic Buildings Countythe in of Argyll. APPENDIX B. REPORT ON BONES FROM CISTS AT POLTALLOCH, ARGYLL. By Professor T. H. BRYCB, M.D., F.S.A.ScoT.

GRAVES WITH BODIES EXTENDED. bonetrune e limbd th Th f an ksentirele o sar y absent fro depositse mth A . fragmentary skull with a portion of the mandible belonging to it, a small part e frontaoth f l regio secona f no d numbea skull d an ,f teet o r h alone remain describede b o t . Skullrepresentes i . I righe th e face y td e loweb th sid th f , eo r e parth f to frontal bone—the right half of the sphenoid and the right temporal bone. The vault occipitae th , l whol e lefe regionbase th th e tth f ed ar sided o an , ean absent. The characters of what remains of the forehead and face, as well e palated teethath s an e lowe th ,w , suggesja r t thae individuath t a s wa l womane relativelth t bu , y larg e mastoi th siz f eo d process fairle th , y thick upper orbital rims rathee ar , r male than female characteristics. nasio-alveolae Th r height, indicatin heighface e gth th e f abovo t mouthe eth , e righTh t . s orbi6'i cm 4 t measure 3'y b 1s . cm.3'8cm , yieldin orbitan ga l index of 81 '6. This figure indicates a low broad rectangular as distinguished from a circular eye-socket. The teeth on the right side of the upper jaw are all in place save the central incisor crownt uniformle no Th e . sar ye molar wornth e f first O th ss wa .a , s expectedbe to , shows most weathirdthe ; rwisdoma , , very little. Thiis s explained by the fact that the corresponding tooth in the lower jaw had been e genera t riseimpacteth no wano o t nt d lf e roome o ha levetth ddu t f I o .l bite and shows therefore an unabraded crown. The lowe s ramrit angled jas an i wha s broken away e symphysiTh . s i s shallow and the mental tubercle is not prominent. So far as the anatomical features go they point to the probability that the individual was a woman. The teeth sho wconsiderabla exaggeraten a t no t ebu d degre f weare o Th . person womana f i , s stilwa ,l probabl e earlieth n i yr perio f middldo e age. 184 PROCEEDINGS OF THE SOCIETY, JANUARY 14, 1929.

Skull II. The second cranial fragment is a small portion of the frontal bone includin glabelle nosee th e rooTh gf th . o ts slightla i y prominentt bu , e thth e f fragmeno o smalx to se o permit s li e a judgment th f o to t s a t individual. Three lots of loose teeth were included in the deposit sent for examination. A number of these fitted the alveoli in the upper and lower jaws of Skull I. Of the teeth remaining, two with worn crowns (a premolar and a molar) belon aduln a o gtt setthir pathologicaA a . s di l specimen, wit hsingla e much- thickene dcrowa fanand gn lik eeye-toot thaan of t h encrusted with tartar. The rest belonimmaturn a o t g e set. Four with roots broke belof nof e wth neck are probably the four six-year-old molars of one individual. All the others are rootless, consisting of only the hollow crowns of unerupted teeth still contained within the secondary alveoli. Even in the absence, therefore, of any bones belonging to a child, it must be concluded that with one of the adults a child of some seven or eight years was buried. CIS. TA deposie Th t from this cist consist f comminuteo ) s(1 d fragment skule th f lso e smalleth f o bone r d lonan s g e boneskeletonth f so l havinal , e ordinargth y appearances of bones which have been burnt. None of these fragments yield any data regarding the age or sex of the individual; (2) of larger fragments of the long bones which see mhavo t e escaped never o , havo rt e been exposee th o dt actio firef no . They includ portioa femue ) e(a on f rno correspondin abouo gt t the middle third of the shaft; (6) a piece of the other femur consisting of the neck and about one-fourth of the length of the shaft, but without the head e bonebone oth s beef Th eha . n s lonspliit gn i t axis, openin e marrogth w cavity, and only one section of the wall has been preserved ; (c) the upper (d)d ulne an halon varioua ; f o f s splint-like fragment othef so r long bones. The broken surfaces of the bones have a chalky appearance, but the bone is soft and crumbles when touched. The bones therefore do not have the characteristics of bones deposited after cremation. The conclusion must be either (a) that the deposit represents two burials, one after cremation, the othe ordinarn a r y inhumation r (&o ,) tha incineratioe th t vers nywa partial so that the extremities escaped consumption. Nine teeth, including five molars (fou e lower)r on uppe d , an rhav e been preserved. At least one of the molars is a wisdom tooth with closed fangs, indicating that the individual had reached the twentieth or twenty-second year of life. The crowns are only slightly worn, the first molar, the earliest eruptede b o t , alone showin degrey gan f abrasioneo persoe Th .who o nt m they belonged must therefore have die earln di ye enamee lifeth Th f . o l teeth is rather chalky-looking, but the fangs are intact and do not seem o havt e been expose e smalth s o firelt d A fragment. e skulth f l o sbone s presen e deposith n i t t have been burned conditioe th , teete th hf n o describe d tends to support the idea that the cist contents represent two separate inter- ments. The presence of the necklace suggests that one of these was that of a woman, but the anatomical evidence as to sex is not determinative. All T NECKLACJE E FRO CISMA T POLTALLOCH A T , ARGYLL. that ipossibls ti i y s thasa fragment e o et th lonf o s g bones presen quity tma e well have been those of a female.

CIS. TB e deposiTh s veri t y scantye skul th par o presents f i lN o t. t fourteebu , n teeth have been preserved. They are nearly all imperfect, the roots in e majoritth y being broken off. The relativele yar y smal sizen i l , suggesting probabilite th y tha e individuath t a woman s wa l . Amon e teetth g h occur two third molars or wisdom teeth, the fangs of which have fully "closed." This means that their developmen completes wa t thad an ,t consequentle yth individual was over twenty-two years of age at the time of death. The crowns of the molars are only slightly "worn, and from these we may conclude that the person was still young. The rest of the skeleton is represented by some eighteen small fragments only. Thre f theseo e portion ear e shaf th f f o to s e femorath on f eo . muco Theto he yar broke weathered nan o permidt f o t any estimate of the length of the bone, so that no evidence is forthcoming o staturet e fragments a Th . s show thae femuth t r wa f abouo s t average thickness. A small fragmen burnf o t t bon probabls ei ywanderea r fro othee mth r cist. APPENDI. XC DETAIL THF SO E CONSTRUCTIO NECKLACEE TH F NO . The beadnecklace th f so e vary muc formn hi , length thickness.d an , 1 Some are almost cylindrical in form, others are prominently barrel-shaped. One bead is longer than any other, and must have been placed to connect the outer and middle stringfront thithe As s at . s cylindrica beaof is d l type othethe , r cylindrical beads would seem to have been used to connect the strings, leaving barrel-shapee th d string e beadth r sfo s themselves. The connecting beads become shorte thes a r y approac frone th h t platet sa either strine sideTh .g beads become thinner towards each side e shortesth , t bein innege useth r r dfo string , an longese outere dth th r .fo t signe Th weaf so r roun platee holee th dth f so show that very thin beads have been used for the fringe, those between the plates being thicker.

particulare 1Th e three beadth th f f eo so smos t complete necklace Scottise th n si h National Museu s follows:a e mar —

No. 1 Necklace. of .1ft •^ •*i • H r& • A Beads. inch. inch. inch. inch. inch. inch. inch. inch. inch. inch. inch. in ch inch.

Tealing . 140 1 3 4 7 11 6 10 25 16 23 19 15 Poltalloch 110 1 4 10 12 10 26 15 7 12 7 6 Bute . . 99 2 4 5 6 12 26 13 17 8 5 1 APPENDI. XD i o TABL ENGLISP EO H JET NECKLACES WITH PLATES

O ^ o Z c U' 3 #3 OJ Q a Fastene r o'

Numbe r o Plates . o^ £. 0,s=w5 re JB s> a.c p a= H & a. d- CO »£ CQ p a W 02 O CAMBRIDGESHIRE. o 1 Soham Fen. 2 With unburnt skeleton and British Archaeology of the Cam- wN=( socketed axe. Museum. bridge Region, Cyri, 55 . lp Fox (1923). H

Burwell Fen. 1 Remaining par f necko t - Cambridge Ibid., p. 55. -LOO1 QCd4 lace also found. Museum. Cam. Ant. Soc., vol. i.,Rep. . 13 xiv. p . 2 DERBYSHIRE. a 3 WindJe Nook, 6 76 2 With female skeletoa n i n Sheffield Vestiges Antiquitiesthe of 1846 1848 Margate Wall. cist in a barrow. The Museum. of Derbyshire, (fig.)9 8 . p . plates are of bone. Cat. Sheffield Mus., p. 59 (fig.) 1899. 4 Cow Low, Bux- 6 40 1 2 1 With female skeleton in a Do. Vestiges, (fig.)2 9 . p . 1846 1848 „*• ton. cist in a barrow, also 2 Cat. Sheffield Mus. small plates and a ring. •h-oJ 5 Middleton Moor, 6 54 1 348 10 In cist, with skeleton of Do. Ten Years' Digging in the 1848 1861 Arborlow. female and child, and a Gravehills Derby,of etc.. cow's tooth. p. 24 (fig.). Sheffield Mus. Cat. (fig.). T NECKLACJE E FRO MCISA T POLTALLOCHA T , ARGYLL7 18 .

i-H S CD 8S »-H . S 1 S OJ 1 S i-H 1 I—1 1 S CO •H iH 1 1 I—1 1 !§ s 3 g JS o >8 e» 3"* •*„ •5 1". ^ C| 1 » ti * « w w K -=- £1»: S § Q, .& ^ S cX s *> bi i5 -g k ^^ (^ V, S- S a P.^I g ^ "^ of •§•- M 0.1 '|S I ^ ^1.- 1 OK iS "> 1* §7«5 P. 0[) 03 O r^ Pq P- ^-f3 ^c«« wS^^^ -a "^85 SSL" • • °* o ._, wV^^i- SJ* — S^§ •e§1 1^1 os§ -g ts «o 1> .SS-^ ^ *"* ^ i** M

Researches, (%.) . firing , p . 228 0s|, &S GQ fi^'Cr) !^r 4 t^ ^ Mortimer' s F British Barro ^ Bateman , Te n

c5 6 6 p P P Museum . Museum . Museum , Museum . Museum . Edinbur g Yor k Driffiel d Sheffiel d Scot . N a Newcastl e M a) cc co ** 0> s 4J r M = to .sfi "§'S oSS % ^ t« !B C 3 cS 03 B-g fs0) l0>) -ST3 ^1 tf« '3'fl an d elaborate . ( 1 i s o f bone ) equa l n an d al l pierce f urn . strings . Thre e skel ( 2 female) , rud e fli n an d 2 ova l beads . skeletons . barrow . " Ver y bea u a* Wit h a skeleton . I n a barrow , als o 1 tub e Wit h huma n remain s t-H I n a barrow , wit h 3 I n a tumulus . I n a tumulus . I n a disturbe d barrow . I n a barrow . I n a cairn , 8 squaris h p I n a cist , wit h food-v e

CM to O) i i— 1 S

iH 1——1 i

rH

1-4

Co. Ca. + m •^ ec. § S 8 8 8 g o

t>s. 00 1O tH !M :M e~ ^ m !«>:. CD 10 a) in ffl % fc o -u> a C8 H P4 S -rfo a 1 =3hj ~ IE O M -15

w 1 LAND . g B 1 _CoJ ™8m3 >H -S S ">-, NORTHUMB J Berwick . tures . Grindlow . Wolds . Hig h Coc k Loochto n For d M Calai s Wold . r Mlddleto n i Pickering . Ove r Ha d If w

APPENDIX E. GOLD IN SCOTLAND AND IRELAND. The chief source of gold in Scotland is in the Lead Hills, where it is found mainl alluvian yi l deposits t alsbu ,o occasionall quartz.n yi 1 Gol alss doha been worked at Kildonan, in Sutherland ; and occurs in small amounts in very many localitie l oveal s r Scotland s founi t Shetlandn I i d. , Caithness, Sutherland, Inverness, Nairn, Moray, Aberdeen, Perth, Angus, Fife, Stirling, Bute, Lanark, Peebles, Selkirk, Dumfries, and Kirkcudbright. The Scottish prehistoric gold ornaments were in all probability made from native gold, as has been pointed out by Mr Graham Callander.2 Irish gol founds i d chiefl Wickloe th n yi w Mountains amoune Th . t minedn i prehistoric times must have been very large ; and Ireland has been called the El Dorado of prehistoric times. The search for gold may have been one of the chief cause migratiof so peoplf nwesto e th o e.t Dr "W. Eraser, M.R.I.A., in 1897, arranged Irish gold relics into three classes accordin theio gt r analysis: Clas r cent . containinpe sI f o 3 , 2 o t 8 g1 alloy Clasd r centan ;pe s; 2 .Clas 1 III . o .witt sII consistin0 h1 f almoso g t pure gold.3 One Irish lunula analysed by Mr J. W. Mallet in 1853 was included in this arrangement and was placed in Class II. The locality is unknown, it belongs to the type with vertical design (see fig. 12, No. 56). opinioMe rth Mallef o ns thatwa t little information regardin geographie gth - cal source of the gold could be obtained by analysis, as the metal had not been used in its natural state, but had been artificially alloyed.4 Analysis of Scottish and Irish Gold :

Scottish Gold, Crawford. Irish Gold. Irish Lunula.

centr Pe . centr Pe . Per cent. Gold .... 86-60 91 -92-36 87-67 Silver .... 12-39 6-17- 8-85 11-05 Etc. .... 1-01 •14- 1-50 1-28

1 "The Regali Scotland,f ao . SJ . . Brook"A , Proc. Soc. Ant. Scot., vol. (1889-90)xxiv0 9 . e .p Se . also Historical Notes on the Occurrence of Gold in the South of Scotland, P. Dudgeon, F.K.S.E., F.S.A.Scot p showinma e goldfleldA th g. f Scotlano s d wil foune b l Jour.n di Boy. Geolog. Soc. Ireland, 1869. 1 Proc. Soc. Ant. Scot., vol. Ivii. p. 157 (1922-3). 1 Jour. Boy. Soc. Ant. Ireland, vol. xxvii. p. 53 (1897). • See Trans. Hoy. Irish Acad., vol. xxii. p. 313 (1853). T NECKLACJE E FRO MA CIS T POLTALLOCHA T , ARGYLL9 18 .

APPENDIX F. EXAMINATIO OTHEF NO R SITE POLTALLOCHT SA . While makin above gth e described excavation e gravel-pith t a s t near Pol- talloch, three other sites were examined slighd an , t excavations were made. Though nothing of importance was found, it is right that a short account should be given of the work done. bees ha n(1 t I state) d above tha e gravel-bankth t whicf o , e Brouehth n ha Drummin gravel-pit forms the south promontory, ends in a northerly direction i somewhana t similar promontory abov Norte eth h Lodge. Ther hers ei e distinct evidence of a curved mound having been thrown up to cut off some 50 yards of e pointth . This moun s aboudi yard2 fee4 3 t lengtbreadth4 n n i i st d han . Externall yt risei feet sinternallt 3 bu , onls i yt yi 1| foo heightn ti feA w. yards to the south a circular cup-like depression lies near the top of the steep bank. Within the fort, if such it can be called, a cist built of four slabs and a large cover 8 feet by 4 feet 8 inches is exposed at the edge of the steep bank overlooking the lodge. The cist measures 4 feet 4 inches by 2 feet. It is 2 feet 2 inches deep unpaveds i d cove inchee 8 an 1 Th s r.i s belo surface w groundth e e th th f eo d ,an major axis points nort easty hb . beed Thaha n t tpreviousli y opene shows dwa n by the presence of a modern pocket-knife in the interior. Nothing more was foun riddlinn do soile gth . shoro Ontw r te leve eo pit th ln s o grounwer g edu d above shord an , t .flag- stones were found. It is possible that these may have been laid adjacent to cist covers, as at the gravel-pit: a further examination would settle this point. (2) Close to the left side of the road to Nether Largie lies a stone circle in a smal lcirclcentre e woodth th foun s f cisen a eo I wa .t d many years agoA . narrow trench was now cut across the north side of this circle. The surface was found to be closely covered with a layer of boulders some 6 to 9 inches in diameter t subsoiseee th no hav m o t t d lbu di e, been disturbed. (3) Examinatio nexs nwa t mad buriaa f eo l mound publiclose th eo t c roat da Ballymeanoch f theso . o eTher tw mound e e pare ear th whic n th ki n i se ar h standing-stones. The mound standing farther from the road was examined and described by Canon Greenwell.1 There is no account of any excavation at the other ; it has the appearance, however, of having been broken into. An opening was now made at the centre of the mound which is some 90 feet in diameter and 3 feet in height. It consisted chiefly of earth, but a large number of boulders irregularl ywitht placeme s . dwa Beneat h these depta feett 3 a ,f layeha o , f ro charcoal covere originae dth l surfac e groundth sigcisa f o conr e o f o N tno . - struction of any kind was found so far as the excavation went. Proc. Soc. Ant. Scot., (1866)8 34 . vol.p . .vi 1