EXCAVATIONS IN PASSAGE-GRAVES AND RING-. 173

XV.

EXCAVATION PASSAGE-GRAVEN I S RING-CAIRND SAN S E CLAVOTH F A GROUP, 1952-3.

BY STUART PIGGOTT, B.LITT., D.LITT.HUM., F.B.A., F.S.A., F.S.A.ScoT., Professor of Prehistoric Archaeology in the University of Edinburgh. With an Appendix by ALLARD H. JOHNSON, B.Sc., A.R.I.C., Departmen Chemistryf o t , Wes Scotlanf o t d Agricultural College, Glasgow.

CONTENTS.

PAQB PAGE INTRODUCTION 173 EXCAVATIO RING-CAIRNF NO D SAN STONE CIRCL CLAVAT A E , 1953 188 EXCAVATION OF THE CORRIMONY PASSAGE-GRAVE, 195. 2 174 Clave Th a Grou Cairnf po . s 188 Ring-Cairne Th , Balnuaraf no The Site before Excavation. 174 Clava .... 188 The Excavations of 1952 176 Ring-Cairne Th , Culdoich 190 SUPPLEMENTARY EXCAVATIONS The Clava 192 AT DRUIDTEMPLE, LEYD AN S DISCUSSION .... 192 KlNCHYLE OP DORES, 1952 184 APPENDICES .... 199

INTRODUCTION. n pursuancI e excavatioth f o e n policy outlinee Societth y f b o dy Antiquaries of Scotland after the War, certain regional types of chambered tom beed bha n selecte excavatior dfo representatives na s within areas where evidence for their date and culture was as yet lacking. The chambered long cairns of Galloway were investigated by means of excavations of two tombs at Cairnholy in 1949,1 and the next group for consideration was that e Clavoth f a Grou passage-gravef po ring-cairnd san s mainly centree th n do valleys of the Ness, Spey and Nairn. Of the ten tombs of this group belong- ing to the passage-grave type rather than to the ring- series, that at Corrimony in Glen Urquhart was chosen for excavation in view of its apparently untouche daccessibilitys statit d ean . Excavations were therefore carried out on behalf of the Society of Antiquaries of Scotland at Corri- mony for three weeks in the summer of 1952, and two small supplementary

1 P.S.A.8., LXXXIII (1948-9), 103. 174 PROCEEDING E SOCIETYTH F O S , 1954-56. excavations described below (p. 184) were also undertaken at the conclusion of the main task. In 1953 two of the nine ring-cairns, the Central Cairn at Balnuaran of Clava and the Culdoich Cairn nearby, were examined, and the small stone e Clavcirclth f ao e Group also excavated.1 Field-word an k surveothee th n ry o site s withi Clave nth a Grou whola alss s pa oewa carried (figp show) 1 .ma distributioe e sth th outd an , botf no h type r cairf so fa o ns as at present known, with indeterminate and badly ruined sites, of which only a stone or two remains, omitted. The sites are listed in Appendix I.

EXCAVATIO COBEIMONE TH F NO Y PASSAGE-GKAVE, 1952. The Site before Excavation. The Corrimony Cairn, locally known as the Druids' Circle, lies on the level flood-plai Rivee th f nro Enric heae th Glef dt o k a n Urquhart between e 600d 800-footh an - t contours (Nat. Grid. Ref. NH (82)/381304)s It . position in relation to other members of the Clava Group is shown on the map (fig. 1, No. 6), but in general it may be noted that it is an outlier of the main concentratio thesf no e passage-gravesf o , . whicSE he th mainl o t e yli Loch Ness. Owing to its comparatively remote position the cairn has received little attentio archaeologican i l literature importancs it d an , s ewa first appreciate Professoy db r Child 1943.n ei 2 In 1952 the cairn appeared as a roughly circular structure, some 5 by diamete. highn i ft 6. 8 0ft , d consistinran mose th r t gfo par water-worf o t n stones with turf encroaching only roun p edgee footo a d th twor r n o tsO fo . of the cairn was a large flat stone lying loose, with cup-marks on its upper surface (PI. XXXVII), and free of the base stood a ring of eleven free- standing stones up to 6 ft. in height. Near the base of the cairn traces of kerb- stones could be detected, and at the SW. these appeared to be interrupted so indicato t s a beginninge eth entrancn a f s o passage-grav a o et Clavf eo a type. The whole site seemed untouche previouy db s excavation. The results of the 1952 excavations, together with a collation of the earlier accounts of the site however, show that its recent history was more complicated and eventful than had been thought. The first accurately dated episod movine large th th s e i f e g o cup-marked stone fro wese mth t

1 The Corrimony excavations were carried out with the sanction of the Ancient Monuments Depart- ment of the Ministry of Works, by whom the site had been scheduled, and with the permission of the Forestry Commission, owners of the property, a,nd of Mr Girvan of Corrimony Farm. We are most grateful to these bodies, and to their personal representatives, for their co-operation and help: the Forestry Commission further assiste wore dallowinth y kb excavation ga n camestablishee b o pt o tw n di of their cottages near the site. The Clava excavations were similarly sanctioned by the Ancient Monu- ments Department and the tenant of Culdoich. The work was under the direction of the writer, assisted Davir Woodhams M Mr y d b d Baird AUaran r r M , D , d Johnson studentd an , s fro Universitiee mth f so Edinburg t AndrewsS d han . 2 P.S.A.S., Lxxvm (1943-4) ; earlie26 , r reference ibid.,n si (1873-4)X , 644; XVI (1881-2), 314. EXCAVATIONS IN PASSAGE-GRAVES AND RING-CAIRNS. 175

CAUMS OF CLAVA TYPE

M 0 * A Y FIRTH

PASS A C I- C RAVE ® R 1 N G • C A 1 R. N

F • • I LAND A B 0 V t LuJ 600

MILES

r lis sitef Appendio te Fo sse . 1 Fig . 199.p ., xI 176 PROCEEDINGS OF THE SOCIETY, 1954-56.

s presene cairit th sid o f nt eo t positio e summith n no 1830,n i tthid an s

raises the whole questiol n of the original function of this stone, discussed in detail below (p. 181). To anticipate, however, we have reason to think that it originally formed the capstone of the burial chamber which, as our excava- tions showed, had either fallen in or been broken into, and then deliberately refilled at a date likely to be early in the 19th century, when the capstone would have been dragged to one side, and returned to a central position only after the refilling of the chamber. The plan f 187o sd 1882 4an , e earlgiveth yn ni account e cairth f no s referred to above, both show nine stones only in the free-standing circle, and latee th r plan mark composites a (fig 2 ) s2 d . nosan thes 1 . a to-day,e yar . This was at first taken to be the result of frost fracture of the original stones, t closebu r examination showed that thes stoneo etw s werf o e madon , eup two and the other of three slabs of stone which had originally no connection wit anothere hon . They mus recene b t t addition monumene th t o st se t bu t up before 1874, and there is a strong suspicion that their component slabs were in fact capstones robbed from the now roofless outer portion of the entrance passage (p. 180). Stones nos. 10 and 11 are clearly additions since 1882, and this was confirmed by their excavation, which also showed that nos. 8 and 9 had been reset in their present positions in the moved soil of recene th t bank boundin gdraia n running alonroae side th dgf th e o whic h skirts the monument on the SE. The site, in fact, was by no means as intact as surface observations had suggested.

The Excavations of 1952. The site was recognisably that of a of passage-grave type attentiod an , directes nwa fouo dt r main objectives e clearancth : f eo the cairn kerb, the examination of the entrance area, the excavation of the passag chambed ean r witcontainey han d burial deposits, anexaminae dth - e stonefree-standine tioth th f f no o s g conveniens circlei t I . describo t t e the excavations in this sequence.

(a) The Cairn and Kerb (figs. 2-4; Pis. XXXV-XXXVI). All stones of the structure were- local, the larger blocks being obtained from adjacent e cair outcrope bulth th f n ko d materiaan s l consistinf go water-worn boulder largd san e pebble sEnrice deriveth f o k d d be fro e mth a few hundred yards away. The removal of these stones from the lower slopes of the cairn enabled the upper parts of the kerbstones which sur- roundeexposee whole b th o t r et di fo circuit excavatiod an , carries nwa d ground-leved dowol o nt entranccuttine a th n i t a ld g ean opposite , thin so

1 P.8.A.S., xvi (1881-2), 314. EXCAVATIONS IN PASSAGE-GRAVES AND RING-CAIRNS. 177

:'C0 fc Y I UCHAMBEREN M0 D CAIRN GLEN UHQUHAD.T

0 8 t i t D A I E A

G.MET OK MOB!HN 3 STUUCTUIA( L

tAIR.»4, » XT I II* V UN EXCAVATED

5 10 15 20 2j 30 FEET

Pig. 2.

VOL. LXXXVIII. 12 178 PROCEEDINGS OF THE SOCIETY, 1954-56.

tjj E EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 9 17 .

NE.e th , wher esmala l area insid keralse s eth o bwa examined. Since eth whole cair s constructenwa f roundedo d boulders withou y admixturan t e of earth or other binding material, it was extremely unstable and could not be cut in section in the normal manner. Where it was possible to examine lowee itth , rcaire parth nf o tappeare formee b o dlarget f do r boulders than remaindere th mighs a , expectede b t . The main structure of the kerb consisted of a setting of some 45 stone slab sall w originall, no witedgn exceptione t o hon t ebu yse thn o ,e SW., forced outwards by the pressure of the loose cairn at all angles between a slight inclination and a completely recumbent position. It is difficult to estimate the original dimensions of this setting, but it must have been between 45 and 50 ft. in diameter, with a slight tendency to be elongated on NE.axis. e th SW ., Large longed ran r slabs wercontaininc ar ee useth n di g entrance th e than elsewhere t withoubu , t formal symmetr axiss it n .yo Behind this main kerb there seems to have been a ring of large boulders forming grounder e cairth r n sfo construction kerb-slabe Th . t s se wer t eno ustone-holesn pi t forwarbu , f theo d m deliberately built cairn material, with large grounders, continue t seemI distanca . r ft s d fo likel6 f eo y that the original appearance of the tomb after completion was that of a cairn rising from a ring of kerbstones held upright by an external platform or ramp 6 ft. wide and some 2 ft. high, above which the kerb-slabs would project for a foot or more. Had they been originally completely contained within cairn material it seems unlikely that they should have been forced outward suco st hunifora m entrancee degreeth t A .t couli , seee db n that this external ram beed pha n roughly finishe eacn o f h dof side , though this woulp ga d afterwards have been covere materiae th y db l blockin accese gth s to the passage. At all points examined behind, between and in front of the kerbstones, broken fragment f whito s e quartz were foun abundancen i d , deliberately strewn durin constructioe gth monumente th f no rougA . h coun mads wa t e of fragments found in clearing the upper part of the western quadrant of the kerb, and these amounted to between 700 and 800 pieces, representing onl ysmala l proportio totae th f ln o likel havo yt e been associated with this area alone. Entrancee Th ) Passaged (b an (fig .XXXV) L 2P ; . The area in front of the entrance was stripped to the old surface in a cutting 14 ft. square. Cairn material was found to cover the sand and loam subsoil, and there were no features or finds to suggest any forecourt ritual having taken place. Large grounders marke roughly-buile dth te endth f so ramp outsid massivo tw kerbe ed th e an ,upright t righsa t anglekere th bo st continue e passag th line f dth eo e beyon , witd it . betwee ft hspaca 3 f eo n them. 180 . PROCEEDINGS OF THE SOCIETY, 1954-56.

The passage proper opened on the line of the kerb with a pair of side- foro t ms a slightla o stones t y se sfunnel-shape d entrance, reducine gth width of the passage at their inner ends to 2 ft. Between them were the lower course whaf so t must have been original blocking material, obliquely crossing the passage, and in front of this on the southern side was the calcined and eroded remains of a rough bone pin, the sole artifact found during the excavation passage Th sroofless (fig . edistanc a wa . inward 5) .r ft sfo 0 1 f eso from the kerb, and it is probable that the lintels robbed from this stretch had bee ne 19t th use hn i d centur o makt y e boguth e s standing-stones nos. 1 and 2 of the free-standing circle. plann I passage ,th e widened beyon constrictioe dth n behin blockine dth g already mentione maximua o dt m widt s totait 4f h d o lft. lengtan , h from

Fig. 5. Bone pin from entrance passage, Corrimony. (Full size.)

the entrance through the kerb to that into the chamber was 23 ft. The walls consisted of massive uprights, in many instances deeply sunk into the sand subsoil, with roug heavd han walliny ydr g wit hsligha t corbelling above, carryin roofine gth g lintel . abovheigh a ft poin a t floore 3 t sa ef th A o t . . fro ft entrancee m1 th 1 passagee toward. d sidth f an ,SE eo e smala , sth l rounded block had been firmly set in the floor, projecting 9 ins. above the surface passage Th . e floocoveres wa r d with small rounded bouldera o st depth of some 18 ins., apparently deliberately placed there as a part of the final blocking of the tomb, and at the inner end merging with the recent in-filh'n e chambeth f go r described below.

Chambere (cTh ) XXXVIII)L (P . It could be seen from the partly, excavated passage that the chamber was filled with stones to a level above that of the lintel of its entrance, and therefors wa t i e decided tha t i must approachee b t d from above. This necessitated the removal of the large flat stone, with an estimated weight of som tonso etw , which sinc esummie 183 laith cairne d n 0th n ha o f to . When this had been done a slight depression was visible, and the removal of a very few cairn stones brought to light the upper edges of a circular corbelled vault completely filled with stone earth. d this san A s fillin removeds gwa ,

collapsresule e uppeth th e f t th seeo t s f no e ro e ns wa par th that i wa f o tt i t 1 heave compled 1Th yan x tas movinf ko stone removingd th ean chambee gth r fillin takes gwa n over My b r Thomas Fraser, contracto f Drumnadrochito r Fraser'r M . s personal interes t e worinth k con- tributed greatl succese undertakine co-operations th mose th hi o y ar t f r so e t fo gratefu w m d hi gan o .t l EXCAVATION PASN I S S AGE-GRAVE RING-CAIRNSD AN S 1 18 .

vaultin muco gdeliberats a s ha fielf o dp estoneti eartd san h thrown intn oa empty chamber, and the inclusion of fragments of 19th-century china and glass in the debris showed that we were dealing with an unrecorded episode, involving the refilling of the chamber after it had been broken into by chance or as the result of antiquarian curiosity in the past. Fortunately, as will be seen below e actuath , chambele flooth f o beerd ha rn left intac thin i t s operation. The chamber when excavated proved to be circular, with a diameter of s lowesIt 12 . "tft course consiste fifteef do n a blockmassivw n i lo t t sse ebu ring and sunk deeply into the sand subsoil, probably partly as the result pressure o th fsuperincumbene th f eo t structur cairnd ean : some settlement from the same cause had broken the. entrance lintel from the passage across its middle, and cracked other stones of the corbelling at this point. Above these low orthostats the walling was carried up in rough dry walling, over- sailin uppes it n gi r course foro st mcorbellea d vault maximua o t , m extant ins.6 . heigh,ft wher6 f e overhano t eth ins6 soms . . ft g beyonwa e2 e dth base of the orthostats. This corbelling did not begin immediately above the tops of the foundation stones, and there was in fact a perceptible back- ward inclination of the wall for over 3 ft. from the floor, so that at this point e diamete th e chambeth f s slightlo r wa r y greater tha ground-leveln no : above this point the inward corbelling immediately began and continued dozea r o nn foirregulate r r courses. Although roug appearancen hi e th , wor moss kwa t skilfully contrive rathee th f mako do t r e unsuitableus e stones available for the purpose. No direct evidence remained for the original height of the corbelled mannee structureth r fo whicn ri r vaule no ,finall hs th twa ye roofedon t A . poin corbex t si fiv r o el stone collapsed sha situn di ove modere rth n filling, and suggest that at least another 18 ins. could be added to the extant height, and this, continuing the curve of the surviving corbelling, would reduce the diameter of the top of the chamber, at a height of about 8 ft. from the floor, halo t f s baseanalogn thait O f o .t y with other chambered tombe th f o s same type in Britain and on the Continent, the final roofing is likely to have take fore flaa n th presence f mt o th massive slab d th f an ,e o e cup-marked slab on the cairn at Corrimony, some 8 by 5 ft. overall (PI. XXXVII), suggests that it was in fact the capstone of the chamber, supported on a cor- belled vault at a height of something slightly over 8 ft. (fig. 4). Its position on the side of the cairn before 1830 would suggest that it was moved from its original position, and the chamber broken into, during the first quarter of the 19th century, and the stone later replaced in what.approximated to its primary site afte e chambeth r beed size weighd ha rn eTh an fillef . o t din stone th e make t versi y improbable that suggestes a , beed 1882n di ha n t i , a stone of the free-standing circle dragged up on to the cairn.1 1 P.S.A.S., XVI (1881-2), 314. 182 PROCEEDING SE SOCIETYOTH F , 1954-56. (d) The Chamber Floor and the Burial (fig. 6; Pis. XXXVIII-XXXIX). The lowe rmodere parth f o t n chambefilline th f go r merged int olayea r of water-worn stones similar in all respects to those found covering the floor of the passage in its intact stretch, and can therefore be regarded as an

(OIUI M ON Y : STAINS OF CENTRAL BURIAL Fig. 6. (Scale i.) original feature. The removal of these revealed a floor of yellow sand, stained with iron-pa d veinre f swito d nan h patche gre f whitsd o yan e around the edge of the chamber. Approximately centrally, there was found an area of flat stone slabs, some 3 ft. 6 ins. by 2 ft. 6 ins., but with an indefinite irregulad an r outline, associated with charcoal fragments n placeI .e th s slabs formed a double layer, with charcoal between. EXCAVATIONS IN PASSAGE-GRAVES AND RING-CAIRNS. 183

When the sand surface beneath these slabs was scraped clean, it was recognised tha seriea t stronglf so y marked grey, purplish-blac whitd kan e stain thin si s region formed identifiable outlines which coul interpretee db d as those of a crouched inhumation burial, with its skull particularly well defined, s righlyinit n gto side wit heas hit d toward NWe sth . Phosphate tests (p. 200) showed a marked enrichment in this area. Such stains formed by the decomposition of skeletons in acid sand soils have been frequently recorded in Holland and elsewhere in northern Europe though not hitherto identified in Britain. The matter is further discussed below (p. 198) and in Appendi. xII With the recognition of this burial, the sand surface of the chamber floor was re-examined in detail by study in plan and section. It was found that originae th l surface groun e ofth d withi chambee nth r are beed aha n irregular, surface-lind ol e th d ean coul identifiee db grea y yb d humic layer formed over the whitish-yellow natural sand. The central burial had lain directly on this layer, and its stains merged imperceptibly into it, being nowhere more than a few centimetres thick. The status of the grey patches around edgechambere e immediatele th th f on so d ,an y abov centra e heae th eth f do l stain, is difficult to establish. By analogy they might have been regarded as organic material deposited upon the original ground surface, and those e chambeth y b r walls could have bee redistributioe resule th nth f o t f no earlier burials in the chamber to make way for the last body, represented by centrae th l deposit beneat flae hidentifiablth o t n slabs. t Bu 1 e shapes could be recognised in these areas, despite careful scraping and sectioning,, and the low phosphate conten . 204)(p t , contrasting wit valuee hth s obtained from the undoubted burial and the object above its head, show that this view canno maintainede b t available Th . e evidence implie a singls e crouched burial alone in the centre of the chamber. No grave-goods or other artifacts were found. The yellow sand floor between and partly covering the grey sand stains (excep centrae th n i t l area) coul seene db , sectionn whei t ncu contaio t , n thin layerirod re n f stainingso threo t foup r e ,u o number n i r , separatee don from anothe centimetra y rb . Thiso r eso depositio irof no n oxide, forming incipient "iron-pan, normalls "i resule yth consolidatef o t d surfaces within permeabla instance th chambee e th n layeri f eo d an r, floo t Corrimona r y strongly suggests that fresh san beed dha n strewn ove aree rth t intervalsa a , and consolidated by trampling. Free-standinge (eTh ) Stone Circle (PI. XXXVI). On level ground beyond the outer edge of the cairn an approximately circular setting of upright stones with somewhat irregular spacing surrounds

1 Fo discussiora n evidenc e ofth e for suc hsequenca chamberen ei d tomb burials P.P.S.,. ,cf rv (1938), 129-31; Daniel, Prehist. Chamber Tombs (1950), passim. 184 PROCEEDINGS OF THE SOCIETY, 1954-56. the site: where a diameter can be measured between stones in their original hav e stone1 w 1 standine s w positio e seenA sno th abouf s . i o ,ft g t ni 5 7 t remainine th f o 2 d modere , g47 an ar ) 11 n d additionan 0 1 , 2 s (nosd an 1 . hav) 9 ed (nosbeean 8 .n rese recenn i t t times. Excavations were made roun l stonebase al dth f eo sorde n i excep 2 r. tha no tstone-holee th t s could be examiney rituaan ld depositdan s recovered a furthe d an ,r s arewa a stripped betwefe , separaten7 stone d unusualln a an y s6 d b y large interval, searco intermediatn t a r hfo e stone-hole artifacto N . s werey foundan r no , burial similar so r deposits. Stones 3 to 6 form a consecutive series with intervals of 20 ft. between nos. 3 and 4 and 4 and 5, but 25 ft. between nos. 5 and 6. All are in stone-holes similar in type to those at, for instance, the Monument on Cairnpapple Hill in Phase II, and the stones vary in height from 5 to ins6 . ft 5.d surface abovol e stone-hole e Th th e. s contained packing- stones in the form of water-worn boulders and large pebbles from the river. Betweeintervan a excavationd ft.4 s an 3 , f wa no l7 stoned an s s6 wer e made to search for an intermediate hole. None was found, but an irregular are f cobbleao e sans th lai dn do subsoi founs wa l d approximately midway betweestoneso tw e .n th Excavatio surfacd ol o nt e level outsid caire eth n kerb opposit poine eth t where this cobbled area occurred showedd thadi t ti t continuno e unde caire th r n material s purposIt . obscures ei t doei t s bu , originan a appea e b lo t rfeatur e occupyin e centragth e widel th par f ro t interval between. 7 stone d an s6 Of the remaining stones, averaging some 5 ft. high but with no. 9 being 7 ft. high above the ground-level of its present setting, little can be said. No. 7 appeared to be in situ, though loose in its unpacked socket, but nos. 8 and 9 were set in the material of a modern bank bounding a small stream or drain skirtin circle gth t thiea s point analogn O . y wit othee hth r passage- gravee Clavth f aso Group e Corrimonth , y setting should have consistef do mora lesr eo s regular circle enclosin unfortunats i t tombe i gth t bu , e that modern interference with the setting on its "western side should have deprived us of evidence of the precise original planning of this part of the monument.

SUPPLEMENTARY EXCAVATIONS AT DRUIDTEMPLE, LEYS AND KlNCHYLE OF DORES, 1952. In an attempt to supplement the very inadequate evidence provided by the Corrimony excavations, it was decided to make small trial excavations within the chamber area of two completely denuded passage-graves of the Clava Group, at Druidtemple, Leys and Kinchyle of Dores, Scaniport. In bot hcairn-stonee th site l sal s had,been removed, leaving onl massive yth e EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 5 18 .

stones of the free-standing circle and the cairn kerb and passage, with a few stonechambee th f so eacn ri h instance.1 (a) Druidtemple, Leys Castle (fig. 7). e remain e chambereTh th f o s d cair e behinli n e presenth d t farf mo Druidtemple, nearl mile yon e NNE. from Leys hil Castle.w llo a t lie I n 2s o withi 400-fooe nth t contour, wit hnotabla e view northwards ove moute rth h of the Ness and the Beauly Firth (Nat. Grid Bef. NH (82)/684420). All the stones are massive rounded or columnar boulders: the kerb is almost com- plete, with 28 stones surviving on a diameter of about 40 ft. and an average e surroundinth f heighO f . somo tft g 4 efree-standin n gi . circleft 5 8 , diameter, 5 stones still stand upright and another 5 are fallen or leaning. The standing-stones averag . higft he5 e excepth e Wf th o r tha. n fo to t entranc cairne th o et , whic huga s hi eheightn i bloc . ft k9 . stone e Mospassage th th f f t o sfalle o e ear leaningr no lins it clear es t i ,bu , and 2 stones forming the entrance pair to the chamber remain in situ. An arestrippes awa d beyond thes edestroyee stonesitth e f th e o n so d chamber, w s founno wa surfacs d dt ol i wa d bee thaover g e d d ha eth ndu t an an , covered with 2 ft. of stony debris, among which flat slabs of a type suitable for corbels were relatively abundant d roundean , d white quartz pebbles were common smalw fe l A scrap. crematef so d bone scattered through this disturbed soil were the only finds made. ) Kinchyle(b of Dores, Scaniport (fig. 8) . This site, very similar to Druidtemple in its present denuded condition, smala lien o s l hill-crest between Scanipor. sidf SE eo e d Doreth an t n o s Loch Ness (Nat. Grid Ref(82)/623386.H N . ) Her stone5 e standine sar g

recumben3 d outee an th diametern n ri i t circle. ft 6 highese 6 ,3 th : t stone (5 ft. 9 ins.) is again on the western side of the entrance to the cairn. The kerb is reasonably complete, with 19 stones upright or fallen, on a diameter e passagth d ft. 0 os clearlan e3 i f, y visible. Three stonee chambeth f so r remain situ, n i excavationd 2 ,an s were made eastward thesf so e ovee th r original chamber area. There was little stone debris and the old surface was within a foot of the present turf. A small bowl-shaped pit was found cut into the boulder clay, 18 ins. in diamete ins0 1 . deepd ran , filled with dark soilnead ,an SW re s thith .wa o st scattea crematef ro d bones with darsurfacd ol kunder d e eartean th a n ho few flattish stones. No other finds were made. 1 Permission for these excavations to be carried out were most kindly given, at short notice, by Mr John MacBean f Messro , s Davidson, Scott <$e Co. behaln e Raigmoro ,th f o f e Estates, on which the Druidtemple sites lies, and by the Baroness Burton of Dochfour, owner of the Kinchyle of Dores Monument. 2 Arch. Scot., m (1831), 211; P.S.A.S., xvm (1883-4), 354. There are good photographs of the site in G. F. Browne, Antiqs. in the Neighbourhood of Dunecht (1921). 3 Arch. Scot., in (1831), 211; P.S.A.S., xvm (1883-4), 356. 186 PROCEEDINGS OF THE SOCIETY, 1954-56.

DRUIDTEMPLE, LEYS, (HAMBEHED ( A U N

o

STONES ^5TANHNC OFALLEN &.LEANINO O D I 5 K A ( E D :..;C A I K. N A H t A : ; : f 0 5 10 IJ 20 2J 10 011149674 i-m 1-1 l l l l fEET i i i———i i———i i———i i METH.ES

Pig. 7. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 7 18 .

KINCHYL 0 5RE . D C F HO AE M B E R E D C A I &N

S T 0 N E 5 '• ^ i I A N D I N G £) F A I 1. E N C.) D I S P L A t t D •:•:: C A I H N A H. £ A 5 0 5 10 '' If 20 25 0 I 2 t 7 I 6 f Ui ui 1-1 i —— . — i i ——— i FEET i ——— i r METIEJ 188 PROCEEDINGS OF THE SOCIETY, 1954-56. The result thesf so e supplementary excavation underr littld ou sad o e t - standing of the Clava passage-graves, but they do show that cremation was a burial rite alternative to inhumation among the builders of these tombs. The total absenc grave-goodf eo l siteal consistents si n si f unsatisfactori , y from the archaeological point of view.

EXCAVATION or RING-CAIRN STOND AN S E CIRCLT EA CLAVA, 1953. The Clava Group of Cairns.' The well-known cairns on the flood-plain of the River Nairn at Clava (map, fig , nos.1 . 15—21) for mremarkabla e grou t leasa f po t seven chambered and ring-cairns spaced over a distance of about a mile between the 350- and 400-foot contours (Nat. Grid. Ref. NH (82)/752438-760447)e th f o o Tw . group, at Balnuaran of Clava, are the classic passage-graves from which the type-name is derived; three sites (nos. 15, 19, 20) are ruined and of uncertain type, thoug havy (Maine ma eh on ) beeClavaf so 15 n. ring-caira no , n rather tha a npassage-grave e remaininTh . o cairnstw g , however, Balnuaran Centre (18 d Culdoican ) h (21) e gooar , d example e ring-cairth f o s n type withi Clave nth a Grou t largepa whicf o , h nine certain example knowe sar n (see map, fig. 1, and list in Appendix I, p. 199). The writer recena n i , t study, classifie e Clavth d a tombs into passage- graves or the Balnuaran type, and ring-cairns or the Gask type. Both

types shar e featureth beinf eo g surrounde free-standiny db g stone circles, 1 Gase th kn i typ t ebu t ring-cairndealinno e gar e witsw h passage-gravet sa t cairnallbu , s wit hmassiva e boulder peristalith within whic circulaa s hi r area enclosed by a kerb of stones, and averaging some 20 ft. in diameter. Ther entranco n s ei e passageevidency an manne e r th no f , e o roofin f o r e gth inner area. No recorded excavation taked thesf sha o n y eplac ring-cairnsan n ei t bu , the fact of their close association with the passage-graves of the Balnuaran type, bot distribution hi commoe th n i d n nfeaturan enclosine th f eo g free- standing stone circle, made it desirable to obtain further evidence if possible. Excavations -were therefore carried out in 1953 in the Central Cairn at Balnuaran of Clava and in that of Culdoich.

The Ring-Cairn, Balnuaran Clava.of This monument stands betweewell-knowo tw e nth n passage-gravet a s Clava, and in its present condition consists of a slightly oval cairn 70 by 64 ft. overall, within which a massive boulder kerb 60 by 52 ft. contains the main structure bulth f ko e (fig. 9). Roughly centrally within are n thia as si open 1 Cultures of the British Isles (1954), 257-62. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S . 189

UNC-CAU T CLAVA N A

A IT A I

MEHH

Fig. 9. 190 PROCEEDING E SOCIETYTH F O S , 1954-56.

naturae th o t l ground surfac enclosed an enearl a y db y circular settinf go stone slabs 21 by 18 ft. across. The whole cairn stands within a circle of standing-stones wit approximatn ha e . diameteft Thes 0 10 ef stonero s vary in height from 7 ft. 6 ins. to less than 1 ft., and to three of them, on the W., SSE.d an . ,E narrow "causeways smalf "o l pitched stones. ft ,8 froo t m6 wide, run from the boulder kerb of the cairn. These causeways are a unique feature among the Clava ring-cairns, no trace existing at the other surviving sites. Excavation was carried out at three points. The central area was completely cleared aren a ; a ("A") agains e cairth t aren a kerb ad an , ("B"junctioe th SSEe t th )a f n.o causewa y wit standing-stone hth e th f eo circle. The central area had evidently been disturbed. The natural soil lay immediately beneat presene hth t rough centr e turfwesd th f an , to e shalloa w pit appeared to represent a recent digging. Two stones on its edge suggested that some for f cis mr o stono t e settin beed gha n destroyed t nothinbu , g could be said with any certainty. The central area was blackened by charcoal, and there was a sparse scatter of cremated human bone. No artifacts were found except some small rough flakes of pebble flint (PI. XL). Area "A" was excavated because at this point of the cairn kerb there was a slight flattening, and three larger and flatter stones suggested the possi- bility of a feature comparable with a "false portal." Were such in fact intended, it would be a likely place for ritual offerings. Nothing was, however, found excep smala r fo t l patc surfaced charcoaf ol ho e th . n lo Area "B" was dug to examine the construction of the "causeway" at this point: it consisted of small boulders carefully packed to form a slightly convex strip. The Ring-Cairn, Culdoich. Afte e negativth r e resulte foregointh f o s g excavation, attentios nwa turned to the ring-cairn at Culdoich, at the other end of the Clava cemetery. Withi s bouldenit r kerb thi . sdiametern ft i cair abou. 7 s 2 ft nwa 7 d 5 t an , awa . stooy e solfroSW dth e me survivoth thi e originan o sth f o r l stone circle, a massive slab 12 ft. high. Within the cairn upright slabs delimited an area some 22 ft. in diameter and containing infilling to some height above e naturath l ground-level givino s d gan ,promis undisturben a f eo d deposit PId .an (fig XL)0 1 . . Excavation of the greater part of this inner area showed that this was in fact the case. The surrounding stone slabs were found to have stood to a height of up to 5 ft., and the old ground surface was covered by 2—3 ft. of stone d earthe originaan s Th . l clay surface itself e centrehadth n ,i a , roughly circular are f abou. o adiameteft 7 t r heavily impregnated with charcoa d scatterean l d cremated human bones. Other more sparsely EXCAVATION N PASSAGE-GRAVEI S D RING-CAIRNSAN S 1 19 .

IU N G - ( A I UN AT (ULDOICH

r-—----.-, EXCAVATE!) AREA:

"I FDCE UQ OR.R AF Y

:;i-Xi> PLAN

OLD GHOUN& SUI^.F^CE

™ Ft IT P 1 ^ ?, ! MET^H

5 SEC N 0 I T

MEUE5

Pig. 10. 19. 2 PROCEEDING E SOCIETYTH F O S , 1954-56.

scattered cremated bone fragments, unaccompanie charcoaly db , extended e enclosinth o t t g ou stones impressioe Th . n give thas nwa t charcoad an l cremations had been scattered on the surface, from which apparently the turf would have been removed thed ,an n allowe percolato dt impregnatd ean e clae th y unde t conditionsrwe artifacto N . s were found. The excavatio f bot no e ring-cairn hth s described abov singularls ewa y uninformative, beyond establishin face gth t tha crematioe th t n burial rite was used at both, in a manner comparable to the less abundant evidence at Dores and Leys. The Clava Stone Circle. West of the three main cairns at Balnuaran of Clava lies a small circle of contiguous recumbent stones, approximatel internan i . ft 2 y1 l diameter, with one of the stones on the ESE. bearing cup-and-ring and simple cup carvings (fig. 11). Excavation of the interior revealed what appears to have been a shallow grave for a now vanished inhumation, 5 ft. 6 ins. by ins6 .. 2ft Ove e fillinth r f thigo s a scattenea s s westerit rf wa o r d nen white quartz pebbles. No artifacts were found beyond a scrap or two of pebble flint. DISCUSSION. It remains briefly to discuss the implications arising out of the excava- tions described field-wore aboveth d an ,cairn Clavn e ko th f sao type carried out at the same time. In 1884 James Eraser published in these Proceedings a valuable study of the monuments of this type in the region of Inverness and Strathnairn, with a catalogue of 29 sites; other surveys were made by Beaton, Jolly, Cash and Childe between 1882 and 1944. Practically all

recorded sites were visited in 1952—3, and as a result a minimu1 m of 31 are included in the map (fig. 1) and list in Appendix I (p. 199). That this represents a minimum must be stressed, for in addition to the dozen "uncertain" sitest possibl no whicn i ,s i determin o et t hi e with certainty whether the surviving stones represent a passage-grave or a ring-cairn, others recorded by earlier authors and now vanished have not been mapped or listed. Certai Eraser'f no s sites lists hi s , n amonincludei m hi gy dcairnb s of Clava type, were found on examination to be cairns or stone circles of differing types, and are therefore omitted here. A passage-grave totan te f lo Balnuarae th f so n type have been recognised, one of which, that at Dalcross (fig. 12), is here planned for the first time. The opportunity was also taken of making new plans of the two fine passage- grave t Balnuarasa Clavf no a itself publish(figo t .d 13)an , , wit sanctioe hth n 1 Discussion with reference Piggottn si , Neolithic Cultures o referencefe Britishth o T ff7 . Isles,s25 . pp shoul addee db Anderson. dG Ston e ,som n "th O f eo Circle Cairn d Neighbourhooe san th n si Inverf do - ness," Arch. Scot., Ill (1831), 211-22. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 3 19 . (LAVA STON I HC ( E LE

D E T A V A ( X E N II

/• - . S' ^~r WHITE .QUARTZ

CUP- MARKED STONE

0 I 2 T 4- y 4 0 i——i i——i i——i i FEET i= METRES

Fig. 11. VOL. LXXXVIII. 13 194 PROCEEDING E SOCIETYTH F O S , 1954-56.

AD HC L 0 A T Cl 5S A S A P- E CV A I

QUARRY

FALLEN

Fig. 12. EXCAVATION N PASSAGE-GRAVEI S RING-CAIRNSD AN S 5 19 . 196 PROCEEDING E SOCIETYTH F O S , 1954-56. e Ancienoth f t Monuments Departmen Ministre th f o t f Worksyo photoa , - graph, taken during conservation work some years ago, of a cup-and-ring decorated kerbstone of the NE. cairn (PI. XLI). e distributioTh f botno h passage-grave d ring-cairnan s e Clavth f ao s see e (figb p n, n consistent ma fro.1) ca e types mth a , sis ; both typee sar present, though widely scattered e Blacth n ki , Isle-Beauly Firth-Glen Urquhart regionn mori d e an ,concentratio e valleye Nairth th n f ni no s and Spey. Surviving examples suggest a rough equality in numbers (ten passage-grave nino st e ring-cairns) sequence th d an , e formerl t forwarypu d by more than one writer whereby the ring-cairn was a local "degeneration" fro passage-grave mth e seems har sustaino dt . Their close associatio, nin for instance, the Clava group itself suggests that we are dealing with two tomb-types which were current at the same time: one is reminded of the association of bell-barrows with disc-barrows in the Wessex Culture cemeterie southern si n England. Unfortunately recene th , t excavations have given us no help here, except to show that although inhumation (and that of a single individual) was the rite at Corrimony, cremation was practiced at Kinchyle of Dores and Leys among the passage-graves, as well as Balnuaran Centra Culdoicd an l h amon ring-cairnse gth . The Culdoich evidence suggests that the central area of the ring-cairns wa t leassa t partly antiquity n fillei p difficulds u i t I imagin.o t t e themn i , the original state, as cairns in which neither kerb nor interior setting was visible. The most plausible suggestion would seem to be that after the final ritual performances had taken place in the open central area it was filled just short of the tops of the enclosing stones, so that the cairns would be more or less flat-topped structures in which both interior setting and retaining kerb would be visible. Of the roofing of the central area, unless wholla y b y hypothetical wooden structure, therevidenco n s ei e whatever. It has long been recognised that the passage-graves of the Clava Group were among the best examples in the British Isles of what Daniel and Powell have called "primary passage-graves, whicn "i h architecture, plannind gan ritual arrangements are closely related to the tholos with corbelled roof as seen in classic form in the Los Millares cemetery in Almeria. The same

authors also drew attention for the first time to the resemblanc1 e between Clave th a ring-cairn analogoud san s structures agai Almerian ni , consisting usually of a shallow circular excavation in the ground set round with stone slabs, but sometimes contained with a cairn. Such tombs, which contain collective burial grave-goodd san earlf so y type normalle ar , y only hale th f diameter of the Clava series, but on occasion equal them in size.

Outside the Clava area itself, comparable ring-cairns seem know2 n only areae i nt on Raedykea , s near Stonehave Kincardineshiren ni herd e an ,eth

1 P.P.S., XV (1949), 169-87. . LeisnerV d Megalithgrabere an Di ,. G 2 der Iberischen Halbinsel I—Der Suden (1943), pis. 1-7. EXCAVATIONS IN PASSAGE-GRAVES AND RING-CAIRNS. 197

cairns are about half the size of the normal Clava series. They do however shar feature eth e with Clav standingf ao somn i , e instance t leastsa , within a free-standing stone circle.1 The derivation of the north-east Scottish "recumbent stone" circles from the Clava tombs has, in general terms, been suggested more than once.2 In such arguments, the ring-cairns have taken their place as an intermediate e degeneratiostagth n ei n series fro e primarmth y passage-grave. While some connection presumably does exist between recumbent stone circles and the circle-surrounded cairns of both Clava types, the relationship is hardl e expresse b e whic yon n ca h n direci d r simplo t e form. Wite hth Grangw exceptioNe ee passage-graveth f no e Clavth , a e onltombth ye sar examples in the British Isles of such a free-standing stone circle around the cairn; continental examples (such as Kercado in Brittany) are equally few. Britise th s a h r S tombofa concernede perhapsar y ma e s on ,sugges t than ti the free-standing stone circle we have an element which can be related to the Late (or Secondary) Neolithic traditions of "Henge Monuments" in general, and that in the compound monument of passage-grave or ring-cairn within a stone circle one may have a fusion of traditions paralleled by the encircling of the Maes Howe or Bookan passage-graves with a ditch, or by the inclusion of a miniature burial-chamber of passage-grave type within the circle and avenues of Callernish. The may be a parallel, rather tha na derivativ e form resulting fro ma similaf o t se r circumstances. The presence of cup-marks on the Corrimony stone (p. 181) and both cup cup-and-rind san g carving t Clavsa . 192a(p ) call commentr sfo s i s A . well known, artificially made cup-mark knowe sar n amon mano gs y primitive communities widely separated in time and space that their significance as distinctiva e cultura vere b l ytraiy smallma t . However large th , e group of cup-mark stone th t Corrimonn ea so y whic have hw e seelikels nwa o yt have bee capstone nth corbellee th f eo d chambe XXXVII)L r(P com e b -y ,ma pared with the relatively frequent occurrence of such "decoration" on the capstone f megalithiso c burial chambers elsewher western ei n Europen I . Britain, examples are known in Wales, the Isle of Man and a possible site in Cornwall; Irelann i 3 componena s da Boyne th f o te art-styl probabld ean y elsewhere ;Portugan i * Evore th n li a region Scandinavi;n i 5 Nortd a an 6 h BalnuaraGermany.o tw e th Clavf n nI o 7 a passage-graves cup-marke b o t e sar found on the uprights of the passages and chambers and on the kerb; here parallels coul foune db Portugan di l agai mixed n 8an d with other designn si P.S.A.S., LVII (1922-3) . 20 , 2 E.g. Childey b , PreMst. Scot. (1935). 52 , Daniel, PreMst. Chamber Tombs (1950), 117. Piggott, Neo. Cultures, 211 flf.; Borlase, of Ireland (1897), passim. . LeisnerV d an . , AntasG dos Arredores jfivorae d (1949), E.g. Bransted, Danmarks Oldtid, I (1938), 194. Sprockhoff, Nord. Megalithkultur (1938), 141. Leisner. V d Concelhoan o ,. n Antas G . 8 . Esiremoze . d (1955); Biblos, XX (Coimbra 1944). 198 PROCEEDINGS OF THE SOCIETY, 1954-56. Bretoe th n chambered tombs. kerbstone On e north-eas e ofth t tom t Clavba a 1 is decorated not only with cup-marks but with wavy channels, in one instance .formin mannea ringa n i gp rrouncu comparable dth e wit decoratee hth d stone in the Clava small stone circle described above (p. 192). Both would come within the general class of cup-and-ring carvings which on MacWhite's showing Britis e seeth mn i h Islet traditio ar belon o st n a o gnt outside that Boyne oth f e tombs probabld an , y ultimately linked with northern Spain.2 It remains to comment on the technical aspects of the stains representing e decayeth d skeleto -thn ni ee Corrimon centre flooth th f o f r eo y chamber. Such '' silhouettes " or " ghosts'' of human (and on occasion animal) remains are best known from Holland, where the acid soils render the. survival of actual bone unlikely, and where the fine sands and loams have proved peculiarly favourable to the uncovering of such stains with the high degree of technical skil excavation i l n develope Giffen otherd va an ny d b than si t country. The character of the Corrimony stains—their greasy texture and blackish-purple colou particular—conformn ri s exactl circumstancee th yo t s of the Dutch finds.3 One point that should be noticed is that at Corrimony the skull, while preserving an entirely characteristic profile, has in the process of dissolution into a stain become symmetrically enlarged on both horizontal axes, so that measurements of its outline exceed the normal range of huma ne proces skullsth s f stain-formatioA so . n involve e transformasth - tion of a three-dimensional body into a virtually two-dimensional area of discoloration, distortion is not to be wondered at. We were fortunat haAdnn ei gsoia l scientist workin membea s ge a th f o r excavatio r Allarn M team d d an ,Johnson' s e results reporth hi n f o o st analyses (Appendix II, p. 200) carries the investigation of these "silhouette" phenomen a stag e further tha nr continenta ou tha f o t l colleaguess Hi . demonstration that significantly high phosphate percentages were obtained only froassumee aree th m th f ao d crouche damorphoue buriath d lan s object above its head, and that these were of an order to be expected from the former presence of organic remains, go far to confirm the observations made in the first instance from analogy. It is to be hoped that opportunities will present themselves thin i , s countr abroadd yan further fo , r wor thif ko s kind to be undertaken.

1 Pequar led Bouzictan , Corpus des Signes graves . (1927) . . , passim. J.R.8.A.I., LXXVI (1946), 59-80.

31 Examples have been frequently illustrated: e.g. Antiquity, vn (1933), 470, pis. i-ii; Clark, Arch, and Soc. (1939), pi. xvii; van Gelder et al., ed., Oudheidkundig BodemondersoeTc in Nederland (1947), pi. 9. EXCAVATIONS IN PASSAGE-GRAVES AND RlNG-CAIRNS. 199

APPENDIX I. LIST OF SITES ON MAP (Fig. 1). Abbreviations: PG, Passage-Grave; RC, Ring-Cairn; U, Uncertain. Reference volumeo t e ar s P.S.A.S.f so

No. TYPE. SITE. REFERENCES.

PG Cam Urnan, Black Isle xvi (1881-2), 478; LXXVIII (1943-4). 37 , 2 RC Culburnie, Kiltarlity xvi, 316; LXXVIII. 38 , 3 U Belladrum, Kiltarlity LXXVIII. 38 , 4 U Bruaich, Kiltarlity xvi, 326; LXXVIII. 38 , PG Carn Daly, Balnagrantach LXXVIII, 38. PG Corrimony, Glen Urquhart Present paper. PG Kinchyl f Doreeo s Ibid. U Stoneyfield, Raigmore xvm (1883-4), 361, no. 31. 9 U Culloden . 33 r o 2 Ibid.,3 . no 10 U Cullearnie Ibid., no. 34. 11 U Newton of Petty . Ibid.,35 . no 12 PG Dalcross Mains Ibid., no. 36; present paper, fig. 12. 13 U Urchany Ibid., no. 21. 14 U Cantraybruich . Ibid.,17 . no 15 U Main f Clavso a Ibid.,. 36014 . ,no 16 PG Balnuaran of Clava NW. Ibid., . 10no ; present paper, fig. 13. 17 PG Balnuara f Clavno a SB. Ibid., . 12no ; present paper, fig. 13. 18 RC Balnuaran of Clava Centre Ibid., . 11no ; present paper. 19 U MiUtow f Clavno a NE. . Ibid.,8 . no 20 U Milltown of Clava SW. Ibid., no. 9. 21 RC Culdoich Ibid., no. 7; present paper. 22 PG Leys Ibid., no. 30; present paper. 23 RC Daviot . Ibid.,6 . no 24 RC Gask . Ibid.,4 . no 25 RC Tordarroch Ibid., no. 2. 26 PG Croftcroy Ibid., no. 3. 27 U Avielochan E. XLIV (1909-10), 197. 28 PG Avielochan W. Ibid. 29 RC Grenish XL (1905-6), 245. 30 RC Aviemore Ibid. 31 RC Delfour Ibid. 200 PROCEEDING E SOCIETYTH F O S , 1954-56.

APPENDIX II.

Examination of Soil from Corrimony Chambered Cairn, Glenurquhart, Inverness-shire, with special reference Phosphateto Content. BY ALLARD H. JOHNSON, B.Sc., A.R.I.C., Departmen Chemistryf o t , Wes f Scotlano t d Agricultural College, Glasgow.

Methods Samplingof Analysis.and Soil samples were taken fro e exposemth d surfaces whe e excavationth s nwa completed, wit e primarhth y objec f obtainino t g chemicale evidencth o t s a e nature of the central stain. "Large samples" 3 ins. in diameter by 3 ins. deep were taken with a trowel. These comprised a series along the axis of the passage chamberd an ; several taken withi chambere nth , selecte r theidfo r variation ni colour; and three from surfaces outside the passage or chamber but previously covered by cairn material. In three cases, similar samples "were taken at the same places to a further 3 ins. depth; these are distinguished by the numeral II in the table f resultso s e largTh e. samples provided enough materia r variefo l r o d replicate analyses. The more numerous "small samples" were cores 4 ins. long and approximatel diameten i . in 1 yr taken wit ha piec f metao e l tubind gan recorde referency db tapa o et laid acros chambee sth r intersectin centrae gth l stain. They were used mainl phosphatr yfo e determinations. locatione Th botf so h serie samplef so showe sar fig, n whici 14 . bases hi n do Professor Piggott's plans t wilI conveniene . b l identifo t smale yth l sampley sb means of the axes OX, OY drawn from an imaginary origin outside the chamber. The figure also shows the concentrations of total "phosphoric acid" (P2O6) at the various points, expressed in milligrams per 100 g. of air-dry soil. The soils were air-drie t rooda m temperatur passed ean d sieve. througmm .2 ha No inclusions were noted, except small pebbles in some cases. All were similar in texture, and the mechanical analysis of a single sample gives an adequate idea of it. Such an analysis was performed in duplicate by the method of G. W. Robinson,1 clasild ytan being determine pipetty db e sampling. Apparend an t true densities of two samples were determined, using a density bottle. The true or particle densities agreed closely. Hygroscopic moistur dryiny eb t 100ga , °C. and loss on ignition by heating to constant weight in a silica crucible, were deter- mine seriea n do s representativ varioue th f eo s colour-types encountered. (Loss on ignitio measura s ni amounte th f eo organi f o c matter present. Knowledgf eo the moisture content allows quantities determined on an "air-dry" basis to be recalculate "oven a o dt n dry" basi f desiredsi differencee Th . s woul smale db l in the present instance. Incidentally, hygroscopic moisture and loss on ignition were found to be approximately in direct proportion; this confirms that the soils wer f uniforo e m mechanical composition.) Representativ H valuep e s were determined electrometrically, using a glass electrode. Strong acidity was indicate beloH d(p . w5) Total "phosphoric acid" (P2O5) was determined on the large samples, and, in view of the variation encountered, subsequently on all samples, by the method of

A. J McGregor.. 2 Soi s digestei l d wit ha mixtur f concentrateo e d nitrid an c A1 s described by C. H. Wright, Soil Analysis, 2nd ed. (1939), 39-44. 2 McGregor, Journal of the Science of Food and Agriculture, I (1950), 107-9. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 1 20 .

l« 309 / 226 " 336 \ 29«-"^363,————312 321r ^ 32cx 4 191 7 \--'"- . W |'64

K 183 AI(UNHER.A) 312

SMALL SAMPLES: "227 LARGE SAMPLES 1 . P,.OS PER iocu. SOIL) CORRIMONY PHOSPHATE DISTRIBUTION

Fig. 14. 202 PROCEEDINGS OF THE SOCIETY, 1954-56.

sulphuric acids, and, after dilution, phosphat s estimatei e d colorimetricalls a y "molybdenum bluealiquon a n "founi s t wa partd t I possibl. simplifo et y some detailtechnique th f so e without serious los precisionf so grae soilf mOn o . , i.e. a smaller amount than that usually recommended, weighed without further trituration after air-dryin siftingd gtransferres an wa , smala o dt l Kjeldahl flask graduate mare e necon k th t ka (15n di 0 ml.). This flask served bot e acihth d digestion and the subsequent dilution to standard volume. meae Th n difference between duplicates case th ninf eo n i ,e samples 2-s 3wa , r cent. pe range th , e being 1- r cent5-o 0t 8pe . Since determinatio f readilno y solubl r "availableo e " phosphoric acis dha found applicatio archaeology,o nt compariso1a thif no s with total phosphoric acid was made. soiFivf o l. ewhicg beed hha n groun agatn a n di e mortar were shaken mechanically for two hours with 200 ml. of half-Normal acetic acid, and P2O5 was determined colorimetrically in an aliquot of the extract. In seven cases the readily solubl e totae othee r centP th neas 2 th pe Of x l o n S,wa si 2 rrPi 2t O6bu , examined it varied from 1-3 to 11-9 per cent. Readily soluble P2O5, considered alone might sugges maximua t m phosphate concentration unde e skirte th th r f so cairn, behind the burial chamber! In view of these and other results,2 it is questionable whether this particular analytical tool, applie predominantle th o dt y acid soil Highlanf so d Britain, will contribute muc archaeologyo ht , except perhaps in so far as soil fertility studies are of interest in that connection.

Evidence for Phosphate Enrichment in the Centre of the Chamber. It is clear, from the high local concentrations of phosphate that were recorded, and the sharp concentration gradients, sometimes amounting to a doubling of the figure withi nmattea inchesf o r , tha e soireceives th tha l d increment f somo s e concentrated phosphate source highese Th . t concentrations (750, 730. mgr pe . 10 . soil0g )centrae werth n eo l stain, whic defines hpresene i th r dfo t purposs ea the area within the continuous line on the diagram. It includes the polygonal area from which sample , DC wer, sB e taken. Values belo occurre0 w20 d only outside the stain. Mean concentrations for the stain and for the north-eastern e chambeth hal f o f r outsid t eweri e calculated e differencth d an , e betweee nth mean founsignificans e b swa o dt t when examine "Student'y db s " tes<-teste tTh . rigorout no s i s unless applie unifora o dt m serie samplef so s distribute t randoda m withi areae nth s they represent presene th ;n i 3 t instanc t indicateei s odd abouf so t 300 to 1 against the observed distribution arising by chance, if the central area realitn i wer t eyno riche phosphaten ri average th n so e tha peripherae nth l area. For the reason just given that figure may be in error, but it gives qualitative support to the finding that the principal phosphate deposit is central. Although a pattern of high and low phosphate areas seems to be discernible in various parts of the chamber, contour lines cannot be drawn with any con- fidence. However, hig hstaie valueedge th th nf eo n suggesso t tha aren a tf ao enrichment rather larger than the stain itself may exist, and there is in fact no reason why the phosphate-rich area and the organic matter should coincide exactly, since the processes determining their preservation or migration are different. For the purpose of calculations described below, such an area has been

1 Reviewe Daunceyy db Advancemente Th , of Science, (1952)X I , 33-36. 1 Johnson Piggotn i , t and Powell, P.S.A.S., Lxsxni (1948-9) t Cairnholy(a 1 16 , , Kirkcudbright); Johnson, unpublishe t Glenlucd(a e Sands, Wigtownshire) varioud an ; s author sliterature inth soif eo l science. 3 C/. D . FinneyJ . , Introduction Statisticalo t Science Agriculturen i (1953), 146-8. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 3 20 . postulated, definearee th as a dwithi nlina e enclosin e centragth l stain ins6 , . from it. This takes in three further values (580, 226, 324). The mean is little affected. Ther gaia s statisticaei n ni l significance over that found above, when the area thus defined is compared with the north-eastern half of the chamber outside it. Such an area has been drawn on graph paper and its size estimated, but it has not been inserted in the diagram since there is no evidence of a definite boundar. it o yt possible Th e presenc recenf eo t bones, excret r otheao r source f phosphateso s must alway consideree sb investigation a n di n e sucvien th i s this f t hwa o bu , relatively intact state of the monument and the undisturbed character of. the chamber floor it will be assumed that they are absent. We have then a structure, typologically a burial-chamber, in which phosphates were deposited in antiquity, notably in one area. The presence of bones is a reasonable conclusion, although no recognisable fragment remains althougd an ; visible hth e centrae l th stai s nwa starting-poin investigatione th f to same th , e conclusion could have been drawn i absences it .

Estimation of the Quantity of Phosphorus deposited.

The weight of P2O6 in a given block of soil may be calculated as the product of volume, densit concentrationd yan placeo densite tw s soie Th t sha th a l . f yo been recorded pore Th e .sande spac th f eyo compactesoila n i , d conditions wa , probabl r cent pe lessr o 3 , y3 ; assuming this figure bule th , k densit s 1-73yf i O . more interest than the total amount of P2O5 in the soil is the amount added above that naturally present. Samples from under the cairn outside the burial-chamber or passage provide an estimate of the latter. The sample Q from under the peristalith to the SE. of the entrance gave a value of 168, and the sample R at the analyseo tw n i differenn so 5 e cair15 rea th d f nt o r an occasionsgav 0 e15 . More such results would be desirable, but the agreement of these figures with the lowest value chambee th n si satisfactors ri suggestived yan hundree On sixtd . dan y will be accepted as the basal or natural level for purposes of calculation. For the present purpose the north-eastern half of the chamber may be defined aree asth a abov X-axi e diagrame eth th n si othed an , r area followss sa : e centra(iTh ) l are f enrichmento a , enclose a lin ins6 y eb d . outside th e central stain, (ii) The half-chamber outside (i). Alternativel calculaty basie ma th e sn y eon o of : (iii) The central stain, half-chambee (ivTh ) r outsid. eit

Estimates of the total quantity of P2O6, and the corresponding increments of phosphorus above that natural to the soil, have been calculated for a layer of soil 4 ins. largdeepe Th e. sample t strictlno o d sy represen a 4-int e . layerth t bu , error introduced by grouping them with the others is not large enough to mattepresene th r rfo t purpose incremeny An . t belo wins4 . canno estimatee b t d numerically, but is probably small. The expected falling off of concentration with depth was shown by samples AII and LII. The body of a 70 kg. (11 stone) man contains about 630 g. of phosphorus.1 80 per cent, of which is in the skeleton. Taking a round figure of 600 g., and

regarding "mathia s sa n unit/' one-fifth uni founs i t d directly unde centrae rth l W1 . B. Fearon, Introduction to Biochemistry, 3rd ed. (1948), 43. 204 PROCEEDING E SOCIETYTH F O S , 1954-56.

stain; nearly half a unit under the stain or within 6 ins. of it—on the assumptions indicated above in denning the central area of enrichment; and about two-thirds ounia f t elsewher north-eastere th n ei nchambere halth f o f . Numerous burials, simultaneous or consecutive, in the chamber thus appear very unlikely, unless the bones were subsequently removed. As regards a burial in the centre sub- sequently undisturbed e detectioth , n appreciabla f o n e fraction, less thae nth whole s phosphatit f o , e content woul expectede db . (Some phosphorus-rich soil would have been remove exposinn di staie levellind gth nan s surface.git e Th ) findings are quite consistent with, and lend support to, the view that such a burial existed polygonae Th . l figure "abov heade esimilath s "wa appearancn ri d ean phosphate content to the rest of the central stain and most likely represents human or animal remains also. Distribution of Phosphates Outside the Central Area. phosphatie Th c distribution ove e chambeth r r floor could have occurredn o , evidence th e available, before, durin after gconstructioo e th r monumente th f no . There is no reason to derive it all from the central burial. Mechanical scattering would assumee havb lattee o et th f dri quantite wer th casee t eth bu , y involved seems too large. Ther definito n s ei e associatio phosphatf no e wit peripherae hth l stains. They probably represen horizontaa t l section throug humie hth c upper laye slightla f ro y undulating soil surface. The low phosphate content of samples G and H does not establish, with statistical significance, tha south-westere tth passagewarr no chambede halth f o f r differs from the north-eastern half. Of the three samples from the passage, N was fro ma roofles disturbed an s d areasaie b littl o ds n , aboueca . it tRabbi t bones were found not far away. L and M were sealed under cobble-stones and sugges generaa t l conditio passage th n ni e simila chambere thao th t r n i t , i.e.a scatter of phosphates. The lowest value anywhere on the site was LII, 3—6 ins. deep under the passage floor. It was 16 per cent, lower than the figure assumed above for the natural topsoil of the site, and lower, also, than AII or P II from chambere th topsois Wa . l remove constructinn di passagee gth ? General Note on the Site. Without assuming thae pHth t , under vegetation identicas wa , l with that under the cairn to-day, it can be said that this has always been an acid soil. Although deep, leve easild an l y worked probabls wa t i , y unsuite wheao dt d an t barley and, to a lesser degree, oats; most suited to rye if the choice were available. artificiae Th l embankmen large riverth e ed th moderf an ,o t n ditch very neae rth cairn, indicate that the site is liable to be flooded. It is surprising to find a chambered tomb on such a site. However, possible climatic changes apart, e change th e river-leveth n i s l woul lese db s violen s catchmenit f i t t area were densely wooded. EXCAVATION PASSAGE-GRAVEN I S RING-CAIRNSD AN S 5 20 .

TABLE I.—APPEARANC F SOIEO LE TIM SAMPLEP SAMPLINGTH EO T A S A S . MOISTURE , SELECTEF LosO IGNITIONN O sH p D D AN SAMPLES, .

Hygroscopic Losn so moisture. ignition. Sample. pH. As percentagef so oven-dry soil.

A. Slightly brown-stained at surface 4-5 1-7 2-8 A II. Light grey-brown .... 4-65 B. Grey-black at surface, fading with depth. Slightly greasy character before drying .... C. Pale, greyish, bleached 1-2 2-i D. Like B ...... 2-9 5-2 . LikE eC ...... 4-42 1-3 1-8 half-incp To . F h stained black 4-64 F II. Light grey-brown .... 4-82 G. Dark brown, fading with depth, mottlings of iron .... H. Brown stain just under surface . K. Shallow dark brown band at surface, iron ban insd2 . down L. Slight surface mottling, iron at about . . in 1 1-2 2-2 L II. Light grey-brown, iron stained at 6 ins...... M. Light brown stai t surfacena , strong . iro n• stai. . t n1a in N. Lik eM ...... 4-66 2-2 3-5 . BrownP , especiall t a surfacey o N . separate iron stain Q. Mottled by iron R. Light brown ..... 4-78 1-1 2-i (x = l,y = l) ...... 1-1 1-9

pH values were determined by Mr E. K. Schofield-Palmer, to whom thanks are due. 206 PROCEEDING SE SOCIETY OTH F , 1954-56.

TABLE II.—COMPARISO P TOTA NO READILD LAN Y SOLUBLE PHOSPHATE. Small samples designate feetn i distancy Y ;d b O larg, eOX sample lettersy sb .

Total P 0 Sample. 2 5 Readily soluble PaO5 (mg./lOO g.). (mg./lOO g.).

D 748 16-5 7,2 730 5-2 4,2 580 10-0 Gr 462 10-1 9,2 438 5-5 7, 1 363 8-5 2, 2 336 6-4 9,1 321 7-2 4, 1 226 6-7 12,1 197 4-4 H 189 11-8 13, 1 186 5-7 1, 1 158 9-1 R 153 18.2

TABLE III.—MECHANICAL ANALYSIS AND DENSITY. Sample R. Constituents as percentages of oven-dry soil; loss on ignition, 2-1 per cent. Coarse sand, 2-0-0- . ...2mm . 29-1 Fine sand, 0-2-0-0 . ...2mm . 52-45 Silt . , 12-8 . 5 0-02-0-00. . 2mm Clay, . unde.3- . r 0-00. 4 2mm Sesquioxide . s 1-1(los . solution5y sb . )

Difference from 100 . . . . . 1-5 sane Th d fractions were partly compose micf do a platelets.

Apparent, or bulk True, or particle Sample. density density Pore spac centr epe . (g. per ml.). (g. per ml.).

R 1-403 2-582 45-7 K 1-214 2-575 52-9

Calculated bulk density, at 33 per cent, pore space and true density 2-578 = 1-73. EXCAVATION N PASSAGE-GRAVEI S D RING-CAIRNSAN S 7 20 .

TABLE IV.—MEAN P2O5 CONCENTRATIONS OP DIFFERENT AREAS AND ESTIMATED PHOSPHORUS INCREMENTS.

f o . No Mean P O ± standard error Area. 2 5 samples. (mg. per 100 g.).

i. Central stain and 6-in. surround 16 410-7 ±12-11 * = 3-28, p = 0-002 . Elsewherii . NE n ei half-chamber 28 294-0 ± 15-9J (bu texte tse ) iii. Central stain . 13 416-2 ±13-71 £ = 3-13 =P , 0-00 3 iv. Elsewhere in NE. half-chamber 31 302-0 ± 15-4J (but see text)

Phosphoru) s(P Area Wtincr pe .h layer Mean P2O5 minu0 s16 incremen 4-inn i t . (sq. ft.). (kg-)- g.)0 .10 (mg r pe . layer (g-)-

i. 15-4 62-85 251 280 ii. 38-9 158-8 142 390 iii. 6-5 26-53 256 120 . 46-iv 0 187-8 134 440

Note.—Phosphorus (P) =P2O5 x 0-437.

ADDENDUM. Since the above paper was set up in proof, a report has been received from Dr F. P. Lisowski, of the Department of Anatomy, University of Birmingham, e cremationonth s from Leys, Dore d Culdoichan s remaine Th . s fro e firsmth t two sites each represented a single individual, while at Culdoich one male and one < female, bot f middlho e age, were represented r Lisowski'D . s full report wile b l . P . S publishe nexe th tn d i volum Proceedings.e th f eo