I. ARDLUI MEGALITHS AND THEIR ASSOCIATIONS; CROSSES AT LUIB AND ALLOWA SHORA D TYAN CIS T EDNAMTA , ROXBURGH- . LACAILLED . SHIREA Y B . , F.S.A.ScoT.

ARDLUI MONUMENTS. Loch Lomondside has not figured in any outstanding event in Scottish history t froye , m time immemoria t i musl t have occupie a dplac f o e considerable importance, for with Glen Falloch, its approach from the north, it formed a natural communication with the south and south-west froe greamth t east-and-west rout f Gleeo n Lochy, Strath Fillan, Glen Dochart and the Tay Valley.1 1 The Glen Falloch and Loch Lomond way was no doubt used by some ecclesiastics going to frod an m lona; onle famouyon s name, however s clearli , y indicated—tha e biographeth f o t r of Columba n "EaI . s Eodhnain d "Croian " t Eodhnain n uppei " r Glen Falloch, Adamnan is commemorated in a stream and a croft. Rowardennan, at the base of Ben Lomond on the east side of the loch, was possibly connected with the venerable abbot. (Vide W. J. Watson, The History Celtic e oth f Place-names of Scotland, . 270.p ) 326 PROCEEDING E SOCIETYTH F O S , APRI , 1929L8 . In the past, tracks existed along Loch Lomond and these were the predecessor e moderth f so n highway which take mose th s t even course

MILE i o 4 MILES i I Ardluf o p FigiMa Distric . 1 . t showing Archfeological Sites. . Chaistei1 l Grigoi Sit. 2 f Coir o er; e Earbh Steadin . Chaistei3 g; l Eab . Adamnan'4 ; s Croft; 5. Clach nam Breatann; 6. Mounds of Stones; 7. Site of Clach naro Breatann Clachan ; 8. Boundary Burn; 9. The Dun; 10. Cairn, Remains of Chapel and Cemetery; 11. Ben a' Chaisteil; 12. Stone Circle, Sculptured Hock, Stone Axe found; 13. Boulder; 14. Long Cairn; 15. Pulpit Kock. wese onth t side water'e closth o et sformen i edge s A .r times roae th , d passes through the few places where cultivation is possible, and, con- sequently, where there are settlements (fig. 1). Changing conditions ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 327 necessitated alteratio lineroutd e ol th sn e n i roads e taketh th t y nbu ,b have been followed closel generalld yan y ther s beeeha n little deviation. addition I land-routa o nt e alon westere gth n shore heravailabls ewa e the means of water transport giving small craft a passage to the Clyde e Riveth f o r Leveny b ywa e larg th , e effluene lochth f .o t Wit e exceptiohth e loweth f rno reachesomn i d e san rive r valleys, the Loch Lomond basin could not have supported a large population. Nort f Inverbeho r mor(o g e correctly, Lower Inveruglas) e westh tn o , sided Rowardennaan , n Lomone east Be th e foo f th ,n o o dtt a n steep and rocky hillsides exist. Only in a few places on the east and e loch chieflheae th th thers f i , do stretct y a ye an f lanho d that might have afforded scope for farming and of "which advantage is taken to this day, but principally for the grazing of cattle and sheep. There are more cultivated areas on the west side between Inverbeg and Tarbet as the hills are not so precipitous. To the south of Inverbeg the slopes gradually become less accentuated, and finally disappearing, merge into the plains in the lower Fruin valley, the flats of Kil- maronock parish, lower Strath Endrick on the opposite shore and e Val th f Leveno e . Fields, both pastur d undean e r cultivatione ar , found at Tarbet on the isthmus between the fresh-water loch and the sea-arm, Loch Long at Arrochar. Three miles north of Tarbet at Upper Inveruglas nearlr fo d yan , 1| mile sout f Ardluiho , als abour ofo t 2 mile f loweo s r Glen Falloch nea e embouchurth r e riverth s f i ,o e there pasture and arable land, although in the last-mentioned part onlnarroa n yi w alluvial stri eithen po e streamr th sid f eo . Glen Falloch and the upper reaches of Loch Lomond are in a region formerly politically interesting. Here was debatable land, the northern part of Strathclyde, wedged between Dalriada to the. west and Pictland to the east. Even to this day the county boundaries e peculiarar ; their arrangemen s evidenci t a lon f o ge successiof o n territorial disputes and settlements by charters since the union of the ancient kingdoms. Dumbartonshire extends for more than two miles north of the head of the loch, and on the west of the Falloch it is separated from Perth- e Allth shir t y Arnab e n which n easterla flow n i s y direction until crosseroae t th Inverarnand a y db , thence runnin ge las soutth t r hfo half-mile of its course and continuing to form the boundary until it joins the Falloch at Garabal Farm. The boundary between the o countietw s northward s thei e smai th n n river which, a sav r fo e peculiar irregularit south-ease th fiela o n t ydi f Inverarnano t , divides poina s a tr thenearlfa s ma y opposit e Benglath e s Burn. This, except- ing a few yards near its confluence with the Falloch, is the march with 328 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929.

Killin (Perthshire) and Arrochar (Dumbartonshire) on either side, to norte soutth d han h respectively. Dumbartonshire extend e easth tn so of Loch o Lomonmile e tw Allth sr t s fo dsouta r f Ardluo hfa s a i Rostan separatin t gfroi m Stirlingshire. Argyll marches with Dum- bartonshir e heighte south-westerth th n f o o es n extremite th f o y Grampians approximatel o mileytw s wes f Gleo t n Falloch. Neae th r mountain tarn, Lochan Arnan, Perthshire joins these two counties; thence northwards Argyl d Pertan l h meet along Druim Albaine th , ancient Dalriadic boundary. Survival f Regardo r Megaliths.-—fo Where thern absenca s i e f o e distinct natural features by which parish or county boundaries may be clearly defined, recourse is made to artificial landmarks to indicate the marches. Isolated boulder r heapo s f stoneo s e placear s t intervalsda , ann imaginara d y line running through thes e limitss takei eth s na . Particularly does this appl o sparselyt y populated district mound an s - tainous regions. In localities where agricultural operations are carried on, boundaries may take the form of field confines such as walls and ditches or hedges planted in the past to separate one area from another. It will sometimes be found, when an actual examination is made of the limit f territoryo s naturall, w thafe a t y placed stone e utilisear s d to serve as boundaries, and it is not unusual to observe that on some of these are prehistoric sculpturings.1 Stones still bearing upon their surfaces mor r leseo s distinct traces of archaic markings are usually of such size or shape as to arrest attention. Their striking appearance would make them easily recog- nised landmarks which, through custom, ultimately acceptecame b o et d f limitariao s a n value. Megalith e founar s d incorporated into march- walls whild an , e som f theseo e structures are doubto n , , comparatively modern, many go back to the mediaeval and have undergone repair at different times. Other boundaries, acknowledged formerly, have become obsolete. Near abandoned clachans and shielings old limits of terri-

toria othed an l r division2 foune b ruinen y i d ma s d dyke whosn i s e lines one seldom fails to detect the presence of huge stones. Frequently in e neighbourhooth e numerouth f do s deserted habitation e Highth f o s- land glen e enormouar s s boulders, and, althougo n hf o most e ar moment a few, s landmarksa , , certainl d filn importana ydi l t parn i t the past. Respect for megaliths was not solely restricted to those serving the The late Dr D. Christison mentions two Peeblesshire instances of cup-marked boulders buil1t into walls, one near Manor Kirk and the other near Castle Hill. (Proceedings, vol. xxiii. pp. 140-2.) Examples are to be found in the defining marches of the detached portions of counties incorporate2 d into the large shires in 1891. ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 329 purpos f denotino e g boundaries. Innumerable instances t onlno ,n i y these island t alssbu o abroadcitee b dn wherca , naturaea l landmarf ko this kinstild dan whichle moron o ,es fro mose mth t remote antiquity had attracted man, fulfilled the duty of marking meeting-places for the holdin f open-aio g r courts, elections, religious observance d othean s r popular trysts. As a practice, assemblies near natural megaliths have not altogether ceased. Still more deeply ingrained in peoples who have preserved primitive custom e habith f attendin o s ti s g meetingt a s standing-stones set up by human handiwork. In this connection it is almost unnecessary to refer to Brittany with its pilgrimages to un- dressed boulders, menhirs, stone-circles, alignment tumulid san . Wales furnishes some examples of a similar nature. The usefulnes e hugth ef o sisolate e dveneratio th ston d an e n i n which it was held did not escape the attention of the founders of the early Church in Scotland and elsewhere. The crosses which indicated ecclesiastical boundaries n additioi , n .to bringing befor e peoplth e e Christian teachings, were, doubtless, the successors of the simple or plainly marked megalith. Despite the general discontinuance of the regard in which large stones were held, tradition lingers round them and nearly all bear names suggestiv f someo e creatio f mythologyno ie t b ,deity , demi-god, fabular hero, man, woman (usually aged r legendaro ) y animale Th . namin f peculiarlo g y shaped megalith t ceas r no rocko s d e di swit h these freak f naturao s l causes t ofte stonee ,bu th n se artifichewth y nb e of prehistori n receivema c d appellation n mani w yno s cases garbled beyond recognition. Designations sometimes referre e practicath o t d l importance of the landmark's value in denoting territorial limit. In Glen Falloch I have had the opportunity of studying examples which come under the categories of the march-stone and the trysting- plac r religioufo e s meetings n theI . - first e whicserieon s hi s beara s name suggestin a recognise s g wa tha t i td boundar t merelno y f o y parishe r largeo s r local land-divisions f kingdomso t bu , . This claims i substantiate e distinctivth y b d e appellation " ClacBreatann,m na h " by which the boulder is known. Clach Breatannm e Chaisteils.—na th d an On accoun s whiloit f o t m importance, ClacBreatanm na h n (the e Britonsstonth f o e n Glei ) n Falloch s probabli , e mosyth t interestin l Scottisal f go h boulder-land- marks. Situated about 700 feet above sea-level on the western slope of the glen, about three-quarters of a mile north by west of the upper Falls of Falloch, it commands an extensive view on all sides but particularly of lower Glen Falloch to the south and the Fionn Ghleann e westth o . t Fro e Wesmth t Highland Railwa e a yglimpsb y ma e 0 33 PROCEEDING E SOCIETYTH F O S , APRI , ]929L8 . obtaine e hugth ef o droc k apparently perche e cresa hill th f o .tn do It can also be seen from many points on the roadway which runs parallebelod an e railway wth o t l . Clac Breatanm hna n (fig. 2) e ,solitar th however t no y s bouldei , r restin hillcresa n go t seemi t s when, viewed fro distancee mth a s i t i : grou f pilepo d schistose megaliths. Upon approachin e assemblaggth e one is struck by the impressive appearance it presents with its base, a conical knoll fee0 circumferencn i 18 t, aboud an efee2 1 t t high, situated in an arena-like depression in the grassy plateau. Such a landmark

Fig. 2. Clach nam Breatann from the south. coul t havdno e been ignore pase thad th an tn n d i this moro n , e distinct natural feature could have been chosen to indicate a march. The capstone surmounting the whole is inclined upwards, and, from e appearancth e breaksth f e stoneo eth , s upon whic t liehi s were part ot formerlyi f thad an ,t ther originalls ewa bouldee yon r onl probables yi . The o lespean ss i ktha fee5 1 6 ninchet s above th e bas th f eo e supporting boulders; so that, including the knoll of large grass- covered stones, the full height of this great landmark is nearly 30 feet. In length the uppermost stone measures 14 feet 5 inches. It varies n widti feet4 1 h .o t fro 0 Averaginm1 fee4 g thicknessn i t s girtit , h is over 50 feet. The longer axis is orientated' 21° south of geographical east. The name, " ClacBreatann,m na h " goe r bacsfa k intpase d oth an t relate o ancient s t boundarie d specificallan s e northernmosth o t y t ARDLUI MEGALITH THEID AN S R ASSOCIATIONS1 33 . limit of the kingdom of Strathclyde. Here also was the western frontie f Pictlando r t onle confineNo th y. t so f bu Dalriada,o s , cor- responding very nearl o whayt Argylw t ino s l (whose present march runs a little to the west of and parallel to Glen Falloch), probably joined with those of Pictland and Strathclyde at the boulder. Frontiers would vary on account of the constant wars \vaged among the peoples e threoth f e e realmsgeographicath t ye , l positio f Gleo n n Falloch, in the immediate vicinity of Druim Albain, inevitably made it a marchland. Maps indicating the limits of the ancient kingdoms can- not be strictly accurate because of the fluctuations, but cartographers agre showinn i e g tha divisione tth t abousme t here.1 Few works refe thio t r s boulde t latelbu r y Professor W . Watson'.J s Celtic Place-names has brought it to notice.2 Dr W. F. Skene states tha a battlt s foughwa e n betwee7 A.Di t 71 . e Dalriade th nth d an s Britons nea stona r e called " Minvirc cTighernachy "b ; thiauthoe sth r of Celtic Scotland suppose e Clacb Breatann.m na o ht s 3 Pinkerton, in An Inquiry into the History of Scotland preceding the Reign of Malcolm III, vol. i., part ii., p. 77, refers to a battle " at the stone called Mimro." Professor Watson, admittin e possibilitth g e stonth f eo y having served as a boundary, suggests that the name "Minvircc" or " Minuir agrey c "ma e wit Welse hth h " maen "— "stone,a "d iwr"an h "— "a roe-buck."4 The second part of the name may be compared with the corrupted designatio e R.M.Sth n ni . vanishew "Currierkno e th d r fo " holding, farther up Glen Falloch, which stood at the foot of Coire Earbh, " the rocky hollow of the roe-buck." 5 On the 6-inch to the Mile Ordnance Survey Map (Perthshire, Sheet XC.) bouldee th r figure s "a sClac a Briton.hn t iI s" describe Northe th n di British Railway Company's Guide f 1895o . ,128p s ,"tha e curious boulder near the Falloch . . . ' Clach-na-Breton' or'Mortar Stone,' the former name said to be derived from a clachan that at one time stood lattee th d rnea froan , peculias mrit it r form." That Clach nam Breatann was regarded as an important and well- 1 A parallel may be drawn with " The Auld Wife's Lift" on Craigmaddie Muir near Glasgow. At nearr o , , that natural trilith three parishes meet. 2 Git. supra,. 15 . p 8 Celtic Scotland, . 273volp . .i . 1 History Celtic e oth f Place-names of Scotland, 387. p . connection I 5 n with this place-nam eI hav e exchanged notes with Professor Watsone H . tell e thae pronunciatio m svien th i t f wo f "Coirno e Earbh" whic hI succeede obtaininn di g from three Gaelic speaker Glen i s n Falloch justifiably ma I , y mak e attractiveth e comparison between this name and Tighernach's "Minvircc." The place appears as "Corryard," 1598 (twice); " Corrieerb vel Currierk," 1602; " Tory ark," 1642, 1649; all R.M.S. ; "Toryark," 1640, and " Torzach," 1670, in Retours—truly extraordinary instances of the corruption of a place-name denotin site.e gon Onl e designationyon , tha f R.M.So t . 1602 s perfectli , y clea y givinb r e gth alternative rendering. 332 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. known landmark down to the sixteenth and seventeenth centuries is positively established n easili r ,fo y identifiable t somewhabu , t corrupted forms, the name appe'ars frequently in . the K.M.S. Apart from the association wit e bouldehth r itself e namth , e is los bornthan w i t no t e y thb e ruined steadin f Blacgo k Crof d manan t y remain f houseo s s near it beside the River Falloch and about f mile east of the megalith. Black Crofd remainan t f buildingo s e eas th e rivetn th o sids f ro e are no other than the " Clachinbretane" or " Clachinbrentane" of the charters. Between BreatanClacroae m th d na h d (abouan n t e \ th mil o t e e boulderth eas f o e sixteet ar ) n heap f fairlo s y large stonese On . of these heaps seem o present s t features indicatina g s i tha t i t burial cairn now" measuring 20 feet 5 inches by 16 feet 3 inches and 2 feet in height. The others have the appearance of having been pillaged of a considerable part of their stones to build a dyke passing in the immediate vicinity e possibilitTh . f thesyo e being accumulated land- gathere considerede db stoneo t s shora ha ss a , t distancee awath e yar stone foundation o smaltw f l o srectangula r houses wit e cornerth h s rounded on the outside. Considering that the district was in territory so frequently disputed, three other sites in upper Glen Falloch must be mentioned. Pinkerton, wholla whil t no e y reliable authority t sayye , s that eastern Dalriada possessed a chain of forts. Although there are no really fortified places e glenith n , certain names show tha t leasa t t three eminences were probably regarded in the past as having definite strategic value. Two of thes e " Chaisteils.ar ea rock s i ye heighOn " t 1| mile north-east of Clach nam Breatann and opposite the Allt Andoran, a left bank tributar Falloche th s situate i f yo t I . d between .the present highway road ol o Strathfillan.dt an e th d 1 Designate e large-scalth n o dp ma e as Chaisteil Rab, it does not seem to possess any remains of built defences; but commanding a wide view, it was no doubt a look-out. Two e north-easmileth o shora t s d tan t distance from Coiletter Farm,' e otherth , Chaistei Grigoirl- a mor s i , e imposing heigh d fros an tm it summit a more extensive territory can be observed. Like the neigh- bouring chaisteil o ,sign n ther e s ar etha t artificial improvements were ever made her o defencet e s already provide nature.y b d e On 2

roadd ol e , possesseTh f somo d e fine bridge t manbu s y severe gradients tracee b n dca , mostl1 y to the west of the modern one from north of Luss. Throughout it is comparatively wide and well engineered. Fro wese mth railwae t sidth f e o t Croiya t Eodhai s practicalli t ni y intact as far as Strathfillan where it joins the present highway, a mile west of Crianlarich. 2 The Andoran rises in An Chaisteil (2958 feet), a peak north-west of Ardlui. A quarter mile Benglae nortth f ho s march ther Lochaa s ei n a' Chaisteil near which toldm a , I ,ther ruinse ear . Out of this loch flows the Allt Criche (the March Burn). ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 333 mile north of the confluence of the Dubh Eas and the Falloch is a circular hill-top called the Dun, but more usually known as the Round Hill. Three side e precipitouar s e summith d s ovean i st r 1300 feet above sea-levelt e steepwes e slopeth no th tn e , O hosa ar sbut . o t t, occupying it, the height would be a wellnigh impregnable position as admirabln a wel s a l e observation-post t onlNo y. doe t i commans da view of Glen Falloch, upper and lower, but none could use the track- wa Ghleany yb Caorunn nna r fro o mo nt Dalmally without attracting the attention of watchers on the Round Hill. So far as I know, the hill doe t sho sno signy w an f adde o s d fortification. Thus Glen Falloch did not lack defences, and those it did possess had the merit that they require mako t d fron eai d m ma theo littl n mr eo serv s guardiana e s e marchlanoth f n whicdi h they were situated. Possibly never more than temporary muniments, these in a different locality and under other conditions would have been worth f lono y g occupatiod an n constructions of some kind would certainly have been built upon them. The Pulpit Rock.—One mile and a half south of Ardlui, about 75 yards west of the road and below the railway, is the enormous detached rock known as the Pulpit Rock and in Gaelic, Clach nan Tairbh (the Rock of the Bulls). It is 45 feet high and approximately 300 feet in girth at the base. It slopes to ground level on the north only but rises perpendicularl l otheal n yro sides. The Pulpit Rock derives its English designation from a large niche 4 feet deep, 7 feet 6 inches high and 6 feet wide, hewn out of the east face, about a hundred years ago, to provide a shelter for a clergyman from Arrochar who n accordanci , e with custom, visite e placd th ean officiated here four time a years Neive 1 Th (fig . Statistical.3) Account dealing with Arrochar parish does not mention the rock, nor is it referred to in The Old Statistical Account. Arrochar parish was created in 1658 from the disjoined northern half of Luss parish ; but the Pulpit

Rock continue o fulfit d s rolit l e long afte e tim th f rthi eo s parochial 2 re-arrangement, although that purpose mighe on , t imagine, should have come to an end with the belated building of Arrochar Parish Church in 1733. The Rev. H. S. Winchester, B.D., minister of Arrochar, informs me that the practice of holding divine service in the open air was long in vogue, and that it was not until well into the latter half of the nineteenth century tha t i ceasedt . Quit a enumbe f residento r e paristh f o hs remember the attendances, the most important being at the Communion

Service hel n Junei dpopulao S s thi. wa sr tha n suco t h occasions Maclean1 , Dumbarton, Vale of Leven, and Loch Lomond, p. 219. * Caledonia (New Edition), vol. vi . 908p . . 334 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. a brisk trad s e carrie e n footh bacd liquoi wa e th f an n d o kt o d a r boulder! To reply to speculative inquiry as to why a more convenient spot, such as a house, should not have been the place for meetings of a religious character after the fall of the ancient church, consideration must be given to some traditional significance borne by the huge boulde e e mindinhabitantth th f o t ro se locality— th f o s a significance, moreover ,origis whicit d remotn ni h ha e antiquity. The origin of the meetings at the place cannot possibly be a post-

''*"•"* •..•''• "'.""" ••'"• .*,,""«••* , . _•' 4 '• V

Pulpie FigTh . t.3 Rock near Ardlui.

Reformation one, but an extremely ancient and firmly established usage. Further, it was apparently intended never to allow the custom to die out, and to ensure this end the preacher's shelter was devised. No doubt the interesting institution was bound to disappear through time but it did so gradually, the recent building of a convenient place of worshi t Ardlupa i havin severew gno d this link wit paste hth . Referenc s madi e e late thesn i r e ecclesiasticanoten a o t s l site near Ardlui, but so far removed from others that consideration of its isola- tion supports the opinion that conventicles at the great boulder were indeed of ancient standing.1 The strong flavour of mythology in the Gaelic name, " Clach nan Tairbh,e Pulpith r tfo " Rock e traditios accountei ,th e n th i f r no dfo 1 Ut infra, pp. 343-4. ARDLUI MEGALITH D THEIAN S R ASSOCIATIONS5 33 • . Red Bull of England and the Black Bull of Scotland meeting in mortal combat on Ben Vorlich. So terrific was the contest that the rock on which they fought became detached by reason of the shocks it was subjected to by the onslaughts of the infuriated animals, and finally t i slipped dowe mountaie slopth th nf o e reso t n t permanentls it n yi present situation. Victory learne w wit,s northere wa ,hth n bull which, wit s crookehit d horn, dispatche s rival.dit e stor Th 1y ends wite hth statement that Clac Tairbn hna e largeshth "s i t boulde three th en i r realms"—an indication that the legend associated with the place may g oe tim bacth eo t kwhe n this countr s stilywa l divide intp e u d oth three kingdoms of Strathclyde, Dalriada, and Pictland.2 There appear se tallittlth e n ei itself t s wheconsiderei bu ,t i n d thae bulth t l figuree mythologth n i o sman s f yo y countrieso s d an , t frequentlwitme h e e onomatologih th n s i y f Scotlando y t i seem, s that the tradition provides interesting parallels a.nd it should not be omitted fro m notica districte th f eo . There can be no reason for supposing that such a practice as that of holding meeting a religiou f o s s characte placa t a re indicateo s y db prominen a naturat l feature d onean ,, moreover, whic d attacheha h d to it the pagan legends of the past, should have had its origin in days when such observances were looked upon with disfavour by the reformed ecclesiastical system. Clearly, strong regar n somi d e for r Clacmfo h nan Tairbh survived until almost the dawn of the twentieth century, for, combined with its practical use was the disguised continuance of long-established oldan d custom regardee eveb whan e ni y th ma s t da most prosai post-Reformatiof co n times. e Againamth n eni " Ardlui, e angliciseth " d for f "mo ar d laoigh," the " rock" or " height of the calf," reference is made to a bovine. Near the mouth of the Falloch, on its west bank and close to Ardlui Hotel a smal s i , l plantatio f coniferso e ncentr th f thin o I e .s wood is a large outcrop of schist which gives its name to the place. The association of the words " tarbh" and " laogh" no doubt bears on old- time tradition e localitth f o s y which have disappearede th sav n i e little-known myth of the bulls, and such legends, disguised in the recondite place-names, have lingere. don At the back of the boulder, 5 feet from the ground and above a tiny stream flowing from the west thence along the south Base of e rock th e weatherear , d o shallotracetw f o ws artificiall t parallecu y l

Campbell, Popular Tales of the West Highlands, vol. iii. pp. 131-2. 1 8 Unti e ancienfusioe th th l f o n t kingdom Strathclyden i e rocs th se nort wa kth d o han T . north-west were Dalriad Pictlandd an anorte north-easterd th han n i ; n parte th f wha so s wa t latter are many instances of the carvings of bulls. '33 • 6 PROCEEDING E SOCIETYTH F SO , APRI , 19298 L . grooves about 12 inches long and J inch wide and 3 inches apart. n angla f t aboua o e degreet 5 4 se tThes e sar e fro e verticalmth t A . this point a small chip of whitish flint was picked up. The fragment, over its greatest measurements, is '65 inch by '5 inch, and is nearly one-tent incn a thicknessn hf i h o t bearI . s trace f secondarso y working probabls i d an ypieca e broke implementn a f nof . Found elsewhere this flint worthmighe b t commentf tno y o beins it t gbu , obtained where there o nativn s i e materia e kind cominth dan f o l g fro ma distric t where, ascertainede b n ca s a discoveriee r sth ,o fa flinf so r othe o t r stone imple- ments have been rare, necessitate placing this specimen on record.1 Long Cairn at Stuckindroin.—Occupying an old site midway between Pulpie Ardluth d t fariRocan e th m s ki Stuckindroin. Fro remaine mth s and traces of buildings in the immediate neighbourhood it can be seen that here arose quit considerablea w e ho settlement y hars sa i t o di t t ,bu r bac fa e differenkth t ruins date. Grant f lando s , however, show that Stuckindroi e Macfarlane th s hel y e wa nb dthirteent th n i s h century. e vestigeNonth f eo f structureso s formerly dwelling r farso m buildings are of archaeological interest; but the field opposite the farm, between the road and the loch, contains an important antiquity. Wha tgroua seeme b f fou po o t sr irregularly circular burial-cairns, arranged in an approximate line running 69° east of north, appears, on careful examination of the ground, to be remains of a long cairn.2 The gradual demolition of the structure must cover centuries, and the nature of the destruction varied. Between each tumulus are irregular spaces, showing that s materiaa , requireds wa l mose th , t vulnerable parts were knowI s a onle r th , yfa otheo 1S r flints showing sign f workinso g foun Locn do h Lomondside are those I mention below. (a) Two small scrapers found at "Fingal's Tomb," Glen Luss (1927). (6) Barbed and stemmed arrow-head found on the island Inchmurrin and now in the National Museum. (c) Two worked chips found at Claddochside, Kilmaroriock, are now in the Glasgow Art Galleries. (d) It was reported to me in the summer of 1927 that a stone implement had been found on the island Inchgalbraith. I asked my friend Mr Henry Lamond, P.S.A.Scot., Luss, if he could obtain any information about the discovery. Mr Lamond writes:—" An angler picked up an axe-hea Inchgalbraithn do , wherlond ha g eI surmise e ruineth d d castl d originallha e y been built on the site of a prehistoric lake-dwelling. The angler invited a stone-mason to tell him what kin f ston o de axe-heath e s mad . dwa Wit of e s heav bloha hi f wyo n hammema e th r shivered it in pieces and remarked, ' It's flint.'" I hope later to give a complete detailed notice of the implements found on Loch Lomondside. t seemI 3more b o est than coincidenc evicinite thath n i t Clacf y o Tairb n hna h there should exist prehistoric remains. Taurian place-name frequentle sar y associated with such antiquities. To cit casea poinn ei t where this featur markeds ei woulI , d refe larga o t r e wood called Coilln ena Tairbh on the Poltalloch estate at Ballymeanoch near Kilmartin. Within a radius of a few hundred yards are standing-stones all bearing carvings, and a rock-outcrop with markings at Baluachraig. Ther e alsear o burial-cairns e demolisheon , d e excepcirclth r largf eo fo t e stones formin s outegit r margi anothed nan r openelate th e y Canob d n Greenwel 1864n i l . (Proc. Soc. Ant. Scot., vol. vi., part i. pp. 348-9.) ARDLUI MEGALITH THEID AN S R ASSOCIATIONS7 33 . attacked. Stones were use r dykedfo s and, doubtless, manye wenth o t constructio ruinew no de steadinth f no g buildings nea boundine rth g wall southe th o t ; these, again, were probably utilise presene th n di t farm- house, barns and sheds. Nearest the loch is the well-preserved easternmost mound, about 5 feet fee 1 diamete n 4 feehigi t4 d tbase18 s h an distanit d t ra ,an t fro wale mth l separatin e roagth d fro fielde t unevemth A . n interval three th e ar s other cairns; the one next the first mentioned is in good condition and nearl s high ya s immediat It . e neighbouo s s dilapidateo i t r t no t dbu great an extent as that close to the wall. In fact, this part of the structure is practically razed to the level of the surrounding ground— a state due, not only to the wholesale removal of the stones, but to the actio f wateno r dripping fro e overhanginmth g tree rainn i s y weather treae alsd th cattlf do o an t e sheltering unde foliagee rth . In the plantation to the west of the road (a little over 20 feet in width betwee s confininnit g walls e tracedoutline 0 b e 4 cairth ) r th n f nfo ca e o feet. Recent verge-cutting of the turf near the base of the wall on the west side reveal e presencth s rubble-stonee f someo th f eo e preth f -o s historic structure through whic highwae hth cuts y.wa Whe actuae nth l carriage-wa widenes ywa fea wy db feet some year mano sag y stones were removed whe labourere nth s were engage wore th t thikn a d o s point. The dimensions available indicate that this cairn was at least 244 feet in length, but it is likely that when complete, it would be longer, say 250 feet. At the east end, as has been noticed, it is 41 feet across, and, judging from its intact appearance, the burial is probably unchanged from what t measurei d originally here demolisheo S . othee th e r dar part s thal al t onesaycan , wit degrehany certaintyeof thais , t graduaa ther ewas l narrowing towards the western end where it is now reduced to an in- determinate outline about 20 feet in width. Dilapidated thoug thi, be s t hstructuri e importancf th o s s ei a r fa o es distric concerneds i t , becaus t provideei s evidence tha Neolithin i t c times man had availed himself of one of the few localities on north-west Loch Lomondside which, even regardee nowb n ,ca capabls da f supportineo g communitya . In the summer of 1925 a holiday camper ventured to dig into the easternmost mound whose artificial nature he recognised. Fortunately the work of the enthusiast was stopped in time, but not before I was able examino t stonewore eth k expose labours hi y db . Many large water-rolled stones of different sizes were used by the builder cairne th sepulture f th so , e being built regularly laye layery b r , wit heaviese hth t stonebottome th trenche t sa Th , t quit, whicno ed hha reache originae dth l subsoil, measure fee6 fee d 3 tfeed 4 deep y an t tb s wa , 2 2 VOL. LXIII. 838 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. fille aften di r examination stonee th l sal werd an , e replace nearls da s ya possibl thein ei r original positions. Boulder March-dyken i t Blarstainge.—a The high groune easth tn do sid lowef eo r Glen Falloch, nort Ardluif ho wil a deserted s i ,d an d region. Geologically the hillsides are full of interest, and in few places is there seee b vaso ns o t numbea t f immenso r e boulders, ice-carrie r brokedo n away from the rocky faces of the steep hills. A little-used pathway leads from Ardleish near the head of Loch Lomond, to Benglas Farm, nearly half a mile north-east of Inverarnan. About half-way, 312 feet above sea-level and a little to the north of the tiny Lochan Dubh which, wit outlets hit , form depressiosa n betweee nth Falloch valley and the western slopes of Cruach, is a huge, gnarled boulder of schist. It is 16 feet high, 11 feet wide and 11 feet thick at the base. By its conspicuous size and position near a small stream this large rock forms a prominent landmark. As such it must have been recognised lonwheo ge ag gle nth n was more populous t dictatei r fo , d the position and alignment of a march-dyke extending on either side of. the monolith to east and west. The wall terminates near a stream a short distance to the east of the stone. The march is an old one indicating the division between the lands of Ardleish and Benglas. Little more tha quartena nortruinee e mila th th f s ho reo i t d shieling of Blarstainge, at one time the home of a fair-sized community as the numerous small ruined houses testify s "BlaystaingA . t i figure" n o s the map issued with The New Statistical Account of Dumbartonshire, and in Macfarlane records it is referred to as "Blairstang and Stuckmud."1 e boulder t a serieTh actuall f no o f boundar f o si e , yon y landmarks, was indubitabl n recognisei y d march-land. Strathclyde was n thii , s district, but a narrow strip between its neighbours Dalriada and Pict- land. The limit between Strathclyde and Pictland, to the south-east of Glen Falloch, corresponded to the present boundary between Stirling- shire and Perthshire formed by the Glengyle Water and Loch Katrine. Beyond the head waters of the Glengyle the division is now sinuous and made still more complicated by the intrusion of Dumbartonshire on the east side of Falloch from Benglas and down Loch Lomondside as e Allth t fas Kostana r . After centuries these fixedmarchew no ;e ar s t whebu n Scotland was stile threth l f emado kingdomsp u e , sucn a h important and strategic natural feature as Loch Katrine was an admirable frontier and as a demarcation has withstood many changes. therefory ma t I e concludeb e d thae Glengylth t e boundare th o t y west of Loch Katrine is also ancient, but beyond that river the absence 1 James Macfarlane, History Clan e oth f Macfarlane, . 143p . ARDLUI MEGALITH THEID AN S R ASSOCIATIONS. 339 of such clearly defined limits necessitated some other marks which coul e easilb d y f thesrecognisedo Blarstaing e e th e on s A . e boulder provided an excellent guide. Stone-circle d Rockan Sculpturings t Inverarnan.—a Situated nearly two miles north of Ardlui on the east side of the Dumbarton-Crianlarich

INVCRARNAN FARM

KILLIN PARISH (PERTHSHIRE) _ - -——"STON E AX E FOUND

Fig. 4. Plan of Stone-circle and Cup-marked Rock at Inverarnan. road, immediately after that highway has entered the parish of Killin in Perthshire, is Inverarnan Hotel with the farm of the same name on the opposite side, a few yards to the north. yardw fe sa nortd Behinan h, farm e it wese dth f th o ,t o t tha , is t of the old Lairig Arnan drove-road, a year or two before the War I noticed a number of stones, and I concluded from their arrangement that here were the remains of a double circle (fig. 4). It was not until 1919, however s agai, thawa nI t abl pursuo t e investigationy m e t a s 840 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. this place, and as the region offered me great attractions I spent much districte timth n ei . Regarded from the view-point of the picturesque, the Inverarnaii circle occupie a singularls y beautiful situation. Immediatele th o t y north is a large \vood of tall larch trees. To the west is a small tree- covered hillock shuttin e risinth t g gou groun d formin valley-slope gth e of the south-western Grampians. Scarcely more than a quarter of a mile to the east is the precipitous and almost perpendicular western scarp of Benglas, wit falle hthaf th so t name descendin leapw fe sa fro n gi ma great height almost to the level of the River Falloch. Between the circle and tha Allt Arnan, to the south, is a narrow strip of pasture, and beyond the stream, on the Dumbartonshire side, is rough undulating moor. The site is peculiar as the circles are not on regular ground, and at least two erratic boulders to the west appear to have been incorporated int settinge oth . Excep innermose r thosth fo t n ei t ring e stoneth , o st a lowe n o r e eas e leveth ar t l tha e firsnth t mentione e otherar s a ds greao s o degreet a e nort t grounde th southd no o han Th t t . bu , , save in the raised centre, has become extremely marshy and even in summer moisture th e remains. The outer circle has beefee 2 n diameter i 10 e ntinne th 1 7 d r an , feet. Judging fro e distancmth e betwee e stoneth n s which appeao t r be placed at regular intervals, the larger circle probably consisted of thirty-one stones and the smaller of twenty-nine. In the centre of the setting grassa s si y platform where therhavy ema e bee nthira d ring, occupiew no s y i somb d t i e t boulderbu s thre f whico e h apparently served as corner-stones of a building dismantled long ago. The struc- ture, whose longer axis ran north-west and south-east, is now reduced to these stones (which from their size and appearance seem to have been taken from one of the circles), and a slightly raised irregularity extending for a few feet. Two isolated boulders of similar appearance and size lie almost on e nort th d soutan h hd 10 0an line fee8 4 , t e outenortth f ro h circle. Their position seem indicato t s e that e settingthe pare th yf ar o t . The boulders forming the circles are not very large. Those remain- average.frog in feee sam3 lengtn th i o t t height n ei d m 2 h an , ands a , is usual in such sites, some have, in the course of time, become displaced r havo e sunk deepl t d consequentlsoiyan we l inte th o y show little abov e groundeth . Stones have been removed gradually e boulderth ; s lef situn i t were doubto n , unwieldo to , permio yt f theio t r being easily removed, and those in the marshy places were doubtless left untouched accounn o f theio t r situation. The Inverarnan circles resemble remains of a small setting on the ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 341 sloping grounsoute th f Loco ho dt h Ardinning near Strathblanet A . Melgum Lodge, Tarland, Aberdeenshire, are vestiges of a stone-circle like examplth e e under review. n theiI 1 r dilapidated conditioe th n Raedykes Circles, near Stonehaven, show analogous features.*

Near the south corner of the fencing separating the pastur1 e from the farm-enclosure are the turf-covered remains of a small inclined banking buil f land-gathereo t d stones. This raised work, carried over e softh t ground fee2 heightn measures i w i ,t no t i ; s abou fee0 6 n ti t lengt d averagean h fee0 1 sinche6 t n widthi s s longitudinaIt . l axis runs in a line approximately north-west and south-east. When entire, the structure formed part of a roadway leading up to the building erecte e middle circlesth th n df i eo . A a distanct yard 7 e circlsoute 6 th f du sf eo eh o behind Inverarnan FarmAllthe t in ,Arna smala nis l island westerits , consistinnend gof an outcrop of the native schist 63 feet in girth. The crest of the rock

is 20 feet above the present bed of the stream 3 which, as far as I have learnt, although subjec o suddet t d heavan n y floods s neveha , r been e easterknowth f o ncoveo p n t portio to islande e roce th rth kf Th no . ifee8 s inche6 t s above grass-leve slantd an l s toward s basethisn it O s. part, near the main stream, an easel-like surface measuring 7 feet 8 inches by 5 feet 6 inches has been prepared. The irregular surface of the re- maining rock-face average inche2 s s abov e dresseth e d portione th n i , smoothing of which much labour must have been expended. This area, now spongeous and absorbing, much moisture through its striae, muca s hha weathered appearance. Despit raie eth n actio f centurieno s a large number of sculpturings can be detected. In common with Scottish prehistoric markings, cup-marks predominat thin ei s assemblage made up of thirty cups, three channels, and one half-moon. The cups, of diameter varying from f inch to If inch, in spite of their weathered, conditio a poo e mosn no rth r surfacet fo par e tar , nearly J inch deep as are the other sculpturings.4 The drawing reproduced from a rub- bing shows the group (fig. 5). e summeth n I f 192o r 5I onc e more examine e rocth d k wite th h object of checking my rubbings. The prolonged drought had reduced the volume of water in the Arnan, and the stony bed of the stream at e islandth e easterf o th , d whernen e normall e flo yth s rapiwi d an d

1 Proceedings, vol. Ixi. pp. 265-6. 2 Ibid., vol. Mi. pp. 23-25. 3 Since these notes were writte ncottaga bees eha n built betwee e bur th e circl nth d ean thus impairing the north and south line. . MaclennanA r M 4 , residin t Furnacga e Cottag Ardluiy eb , inform thae s wifem s hi t , before marriage in the 'eighties, lived at Balloch with an aunt. The older woman, who had been in servic t Ardluea i between 184 1850d referre0d an ,ha circulaa o dt r stone-settin rock-carvingd gan s at Inverarnan. 342 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. rather shallow, was dried up for the greater part on both sides. A dark greenish stone was distinguishable among the great variety of water-rolled pebbles exposed. I picked it up to examine it and saw at once that it was an axe in very fair condition. The implement is polished except near the biitt end, and is 3|f inches long, 2f inches broa 1d | dan inc h middlee thic cross-sectioth e n ki Th . n

u V . c

, IB .... K) INCHES Fig . Cup-marking5 . Rocn so t Inverarnanka .

is elliptica e e cuttinbutt t th Th ova t bu . la lg edge, unfortunately slightly chipped in one or two places, shows more signs of wear at one corner than at the other. The butt has been damaged but not sufficiently to impair the symmetrical appearance of the tool. Mr Peter M'Nair, F.G.S., Kelvingrove Art Gallery, to whom I showed axe-headthe nativ,of tell thais e me s it tgree n schis outcroan t pof which occurs a little to the north of Ardlui Station. Comparing this example with others I ,fin d tha t closeli t y resembles one foun numbea d t Livermeref a year o ro ag s , Suffolk. l - *• Evans, Stone Implements of Great Britain (1872 edition), fig. 59, p. 116. ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 343 Only three similar stone implements have previously been founn i d the Loch Lomond district, or at least, no greater number has been recorded. One of these, an axe-head, found on the shores of the loch at Claddochside, Kilmaronock, is now in the Glasgow Corporation Art Galleries and Museums. Another of the same type, picked up near Arden s presente wa ,e Nationa th o dt l Museu m1899.n i 1 In 1892 a stone axe (AF 409) from Glen Falloch was purchased for the National collection. Unfortunately exace th , t place s foun•wherwa dt ei is not known. In mediaeval times a number of small houses rose up on either bank of the stream, and until the latter half of the eighteenth century the place figured in documents relating to the Macfarlanes of that Ilk who sold their estate n 1784i e nams Th .e " Inverarnan" figure n mani s y unrecognisable R.M.S.e formth n i s , where referenc mads ei chartero et s confirming grants of land to the Campbells of Glen Falloch whose e nortth propert r Perthshirn ho o s ywa e Alle th sid tf o Arnane n I . latBrskinthe Mr e e Beveridge' Abers InversThe s and Scotland,of p. 38, is a complete list of the appellations collected, and these quoted are: "Inverinarren, Innerymeren, Inverintrane, Inverymerrain, Invery- nans (R.M.S. 1598^1649)." The place-name occurs as Innerymeran in Acts Parl., vol. v., Anno 1633, and in Betours 1640 and 1670, as " Innerymoran " and " Innerynnane " respectively. Another designation is " Innerintrane."2 -Round Cairn d Chapel-sitean t a Glen Falloche th Farm. n O — 6-inch to the Mile Ordnance Survey Map (Perthshire CL, N.E.), the hill- side to the north-west of the small tree-covered knoll screening the circle from the main slope of the mountain figures as "Meall-an-t- Sagairt" or "the Priest's Hill." This designation may relate to the stone-circle but bears more probably on the vestiges of a small rect- angular building e longeth , eas° rf northo 33 axit f t whico sse , s i h e presumeremainth a chapele b f e rui o s o s situateTh t dni . d about 100 yards east of Glen Falloch Farm, f mile north of Inverarnan, close e Falloch th e confluenc d th an e Dubo .t s th f hEa o e At this place is the nineteenth-century mausoleum of the Campbells of Glen Falloch. Between this and the ruins of the supposed chapel ia dome-liks e mound fee2 fee0 n heigh2 i n diameterti w t d no , an t . This is a burial-cairn, and while in good condition and covered with thick turf it is not as high as when first seen by me a number of years ago. Many stones, which were then grass-covered, have been take ne present th awa t ybu tenan s e tumululefth ha t s undisturbed, 1 The axe-head found at Inverarnan is now in the National Museum also (AF 684). 2 The Clan Campbell, vol. ii. p. 106. 344 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. d althougan h somewhat reduced s retaineha t i n ,appearanc a d f o e preservation n probinO . g wit a hcrowba a deptfee3 t a f rt o h from the surface a slab was encountered. This, when struck with the implement, rang hollow, probably indicating the position of the burial chamber. To the west a number of fairly large water-rolled white quartz pebbles are exposed. Here are also a few inscribed headstones of the seventeenth and eighteenth centuries. Scanty in its remains this site provides, nevertheless n interestina , g exampl f continuouo e s a regar r fo d place. Named " Stuckchaple " in the charters, it is evidence of the known nature spotth f ;o eGaeli c speakers always cal t "i l Stuc a' Chabeil," e " Rocth r Pinnaclek(o e Chapel.th f o ) " While the remains of the small building near the round cairn can- not be definitely pronounced to be those of an ecclesiastical foundation, deduction shows that ther reasos ei r assuminnfo g that they are. In the absence of any definite evidence of other church sites between t Fillan'S s Chapel nearl e nortn mileth d Lusyte o an ht s s nineteen miles '•• south s hardli t i , y conceivable tha o greas t a distanct e could have existed between place f worshi o se Gle th ns a pFalloc e th f o h past was comparatively populous. Because of the natural features, beyond a certain point to the south towards Luss, there could have been little or no agriculture, consequently there were no habitations and, therefore, little need for a church. If the Pulpit Rock, previously referred to, were resorted to at fixed intervals in mediaeval times, no doubt consciences were satisfied. The orientatio e buildinga surth t f es trueno i no guide t s i ,i , t bu , chapels vary greatly as regards what has become a much debated question. Indeed, numerous instances might be mentioned where ther s muci e h divergence from true cardinal e samsettingth en I . count I yhav e note e orientatioth d f mano n y ancient chapeld an s have been struck by the absence of uniformity. To find evidenc f post-Reformatioo e n buria a plac t a le like this i s pre-supposo provee t b y mann di ma s ye(a paralle l cases e existencth ) e of an earlier church with burial-ground attached.

CUP-MARK DUINISHT SA , CBIANLARICH. On the high ground above a sharp bend in the River Fillan, about a mile north-east of Crianlarich, are the remains of a hamlet figuring as Duinish on an estate map dated 1769 shown me by Mr Gordon Place in whose father's possession was the site until a few years ago. ARDLUI MEGALITH THEID AN S R ASSOCIATIONS. 345 Her r PlacM e e pointe a schistos e d m ou o t te boulder fee0 5 n i , t girt d averaginan h fee 5 gheightn i t n whico , e ahouhar t sixty large and well-preserved cup-marks. Probably the boulder bore a greater numbe f carvingo r s tha e stonth n s evidenea showw no t s signf o s having lost s originamucit f o h l surface. Nearb e severaar y l large pieces which have become detache r havo d e been removedw fe A . yards to the east of the boulder is a remarkable stone I wish to bring befor Societynotice e th th e f eo . Viewed from eas westo t t e monoth , - lit f schisho t appear triangla s a s e whose ape s beexha n injurede Th . longe e rston th axi s f orientatei eo s ° eas 3 df tru o t e t northa d an , grass-leve t measurei l fees4 t along eastd an west faces. Taken medially the full height fee3 tw 4ino s| inches t greates thibu ,swa r originally as a few inches are broken off at the top. The two faces and the side to the north are fairly smooth and the side has the added feature of tapering evenly upwards fro e groun8 inchema widtth 1 t f a o hsd to 13| inchese narroweTh . r south side, averaging 10 inches, is not quite regular on accoun f fissureso t t frobu ,m a distanc e th e unevenness is not apparent. The west face bears an assemblage of fourteen cup-markings varying from | inch in diameter to 2 inches. These sculpturings Fig. 6. Cup-marked Stone at l fairlal e yar deewelld an p preservedn O . Duinish, Crianlarich. e slopinth taperind gan g north sidfoue ear r cups each 1| inc diametern i h beino tw , g placed about midwae th n yo incline do almos surfactw t ground-leveld a t an e e illustratioth n I . n based on rubbings, an aspect of the north side is shown as if perpen- dicula o indicatt r e relationshith e p betwee e maith n ne grouth d pan marking e sidth en (figo s. 6) . Three-quarters of a mile west of Duinish, near the West Highland Railway, are the ruins of another shieling which appears on the Place estate map of the eighteenth century as Luibmore. Among the many large scattered boulders here is one on which are a few distinct cup- marks but the group is not remarkable. concludinn I g this sectio I nwis recoro t h indebtednesy m d e th o t s many friends who assisted me in making surveys of the sites described. To Mr Callander, Director of the Museum, and Mr Ludovic M'L. Manmosm a I tn gratefu r invaluablfo l r egoin fo advic d g an ovee r the difficult terrain with me. I am under a great obligation to PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. Professor VV. J. Watson for his help in respect of the place-names to which special referenc s beeha en made.

CROSS-PILLAR AT SUIE, GLEN DOCHART. On a low knoll in a field to the north of the road and close to the River Dochart at Suie near Luib is the burial-ground of the Macnabs of Inishewan gravee familw th fe e withif A o s yar . . a nrectangula r enclosure of dry-stone masonry in which is a mural tablet stating that the structure. dates back to 1759. There are some plain slabs without the low roofless building and also clear traces of foundations. e sepulchraNonth f o e l stones call r commentfo s e t closth bu ;o t e south-east e corneMacnath f - o r ben LUIB closur a leanin s i e g pilla f schiso r t grounde th firmln i t . yse From grass- leve stone th l e rise fees3 t 8| incheso t s irregulait whicp to r h still retaina s feature commo sucn ni h monuments, side thaeon beinf o t g higher thae nth other. Roughly quadrate in section and badly weathered, it varies in thicknes beard san sigo sn havinf no g been dressed. At the base the girth ifees2 inches8 t . Two faces, approximately easand t

WEST EAST west, are presented. Each bears an O incised cross, t rud no n for i et mbu ,1_ fFT totally devoi f symmetro d r simiyo - larity of position. Fig. 7. Cross-pillar at Suie near Luib. Measuring 9| inche 6y fb s inches, e crosth se wesbornth tn o efac s i e fairly well preserved, but while now only -J inch in depth it was prob- ably more originally opposite th n o ; e aspec symbole th t , although very shallow s largeri , , havin a gshaf t 10| inches lonarmd gan inche 8 s s across greatls i t I . y impaire haviny db shaft,ane gth darme parth sf o t deea cubroa d n i tpan d groove which allows rainwate floo t r w down easily, thus wearing the stone and almost obliterating the carving at its lower end which now appears to merge into the groove. Both sculpturing average th o alike t t ear sappearanc n ei cu e ar d ean widt inch1 f ho . The illustration is reproduced from rubbings (fig. 7). Writing in the Proceedings, vol. xxiii., footnote, p. 117, the late ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 347 Dr Joseph Anderso no believ t lead e on se thae burial-grounth t t a d moro n Suis e ethawa na plac f sepultureo f Macnabs eo vie n i t w Bu . place-name oth f e " Suie," derived from "Suidhe," "a seat" (an doubo dn t referring to one of St Fillan's places of contemplation),1 the existence n earla of y chapel her s i certaine . Evidenc f thio es i provideds , not only by the vestiges of foundations but also by the cross-pillar oa typf e which canno classee b ta memoria s da e e dead th Th f .o l monumen s carvinge placeit b d y s coevaa dan t ma s l wit e greahth t patron of the district. The position of the stone, while perhaps not canonical accordin e latee usagth th rf o o et gchurch s neverthelesi , s exactly similar to that of a cross-bearing boulder noted by me some t Blane'S t a year so Chapelsag , Lochearnhead. s ii t r likelNo 2y that the stone was moved from its original situation in the eighteenth century, as the builder of the family burial-enclosure undoubtedly chose the spot because of its long recognised sacred character. At the time of the erecting of a building on the chapel-site the Highlander evinced an even stronger regard for such relics than he does now.3

CROSSE CAMBUSDOOT SA BLAIBSTOND NAN , ALLOWAY. ThKird eol f Allowa ko annexatioe s builth n ywa 1516o n i t d ,f an no e paristh f Allowa169n ho i r e buildin0th Ay thao yt f o tg became dis- used s ariallowedwa falo t d l into ruin. Close associations wite th h

Ayrshire poet, however, 4 have made both churc d churchyaran h e th d resort of countless visitors. The neglect and decay to which the kirk seemed destined whe e purpos t th ceaseni usee r b r whic fo do fo de t h it was built was arrested little more than a hundred years later, not through any veneration for the sacred character of the site, but through the sentimental regard for Robert Burns which commenced to find expression not long after his death. Consequently, Alloway Kirk, roof- less though it stands, is in good condition. No architectural feature of any moment distinguishes it, and the plain rectangular building with its simple belfry is too well known to call for description here. Lately this early sixteenth-century edifice has come under the protection of H.M. Offic Worksf o e thato s , , independen f privato t e enterpris r goodeo - will, its preservation is assured in the future. Innumerable references have been e Kirmad th f Alloway o ko t e , but none makes mention elucidativ e antiquitth e sito n t th eo s f ea yo "which it was erected. The presence of a well known as Mungo's, 1 Celtic Place-names, p. 261. 1 Proceedings, vol. Ixi. pp. 134-7. * Vide supra, . 343-4pp , referrin Campbele th o gt l burial-groun Glen di n Fallooh. 4 Caledonia (New Edition), vol. vi. p. 503. 348 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. abou a hundret d yard se church th wes f o t , neverthelesso t lead e on s infer that the building set up in 1516 was not the first structure of ah ecclesiastical nature here. Mungo's Well is situated within the grounds of Cambusdbon School whose policies to the south-east are contiguous to the old burial- ground surroundin e churcth g h ruinse sprin issuew Th .gno s inta o concrete basin 1 foot 10 inches deep and 3 feet 1 inch in diameter. Its outflow serves to fill an ornamental basin- in a rockery constructed a few yards e south-easbeloth o wt e norther d closth an to t e n arcf o h the bridge carrying the Turnberry Branch of the London, Midland and Scottish Railway ove e Riveth r r Doon. To test the presumption that until the present time Mungo's Well might have been respecte wishing-wella s da s recentlwa t i , y clearef do deae th d leave debrid san s whic hencumbered fillean t i d effluencee dth . e rubbisth n I h removed were dozen f whito s e quartz pebbles showing that the spring had its votaries who dropped a pebble into its clear depths upon making a wish even after the well was restored and put into orderly condition when the new railway line was constructed some twenty years ago. Whether the practice of visiting Mungo's Well and wishing at its side was made in a serious spirit or only half- heartedlyd ol d perhap keepin . n an n ,amusinf a a o s p a y u sg wa g recognise dt clai custom no o decidemt o d I , . What particular virtue s ascribewa Mungo'o dt s Wel t Cambusdooa l t e knownno th t s i nbu , number of springs reputed to have been blessed by the great missionary of Strathclyd s certainlei y considerable face Th t .tha f t o leasa t x si t thes e neaar e r churche r church-remaino s s indicates that wells were associated with the cultus of Kentigern or Mungo. Probably the most interesting instanc e wel th f thio e ls i swithi e crypth n f Glasgoo t w Cathedral, the shrine of the saint being only a few feet away to the north-west. Alloway kirkyard is full of sepulchral monuments of the conventional types rangin datn gi e from e latteth , r hal f theo f ' seventeenth century o modert n tombstones o stonn t e bu goin, g bac o pre-Reformatiot k n times remains within the enclosure. It may be said that the collection of gravestones cover e perioth s d whic s elapsehha d since Alloway Kirk became disuse 1690n di . Recently, however monumena , t giving definite e ancienprooth f o f t association f Allowao s noticy s comm ha y o et e through the chance discovery made by Master James Galloway of Cambusdoon School. This was found in the small coppice within the propert f Cambusdooo y n here separated fro e burial-grounmth a y b d •wall, clos o whict e h stoo e relith d c almost covered with decaying arboreous refuse. When the heaped-up dead leave d stickan d s ha s ARDLUI MEGALITHS AND THEIR ASSOCIATIONS. 349 been cleared away the stone was seen to bear markings of an interest- naturg in e nea s head it rafted an , r removin thice gth k growt f mosho s and lichen on the surface the well-preserved carving of a cross of ancien uncommod an t n typs fullwa ey exposed excavation A . n made roune monumen th e bas th f do e o ascertait t s exacit n t dimensions revealed tha lowee th t r part, firmly fixed in the ground, was fractured and that the upper portion rested upo. nit The fine-grained sandstone slab, pitted and weathered, with a fragment broken p right-hanto e ofth f d corner, measures 4 feet 1 inch in length and 1 foot 6 inches in width at the head, tapering to 1 foot inc1 hlowee width t rea extremite th f yo small detached piece; the break, running almost straight across the width of the stone, occurs 1 foot 3 inches from the e thicknesfootTh .7\o t 5 inches( s d an ) the irregularity of the back suggest that originally this monumen d beeha a tn rude-back slab dressed down fairly evenly, data t ea possibl t remoteyno servo t , a n ei building. The design is formed of arcs struck from eight points on the circumference of a circle of the same radius, circumscribing bs outlineyit , £ inch wid deepd e an e th , geometrical rendering of the cross which measures 13£ inches down and across its equal armsarc e arrangee Th sar . s a o ds Fig . Cross-sla8 . t Cambusdoonba , Alloway. not to intersect, and the interspaces be- tween the arms are hollowed out to the depth of \ inch, corresponding in this respect to the enclosing circle. The hollows and incised circular outline surrounding the cross seem to give this the appearance of being relieved from the remainder of the stone (fig. 8). A featur notee b o s thatdt i e common i , n with other ancient crosses, usuae therth s eli absenc f striceo t adherenc symmetro et laye th -n yi executiod an e carvingt th ou f no . This characteristic doe t detracno s t fro e pleasinmth g appearanc e Allowath f eo y example. The type of cross, either simple or elaborate, is of rare occurrence in Scotland, and where found the site is invariably in what was the la-boure scenth f e o f missionarieo s s influence e Churcth y db f Candid ho a 350 PROCEEDINGS OF THE SOCIETY, APRIL 8, 1929. Cas Whithornn i a e AllowaTh . y cross, althoug ha surviva f typeo l s i , an addition to the limited list of the Strathclyde monuments of its class which includes the earliest sculptured stones of the Christian era in this country. Of rude d apparentlan r y more primitive sculptur s anothei e r cross carved on a large granitic boulder protected by a low surrounding wall of drystone masonre lan e th soute th f o .leadino ht y o Blairstot g n Mains, about midway betwee hige nhousee th mila hth road ed an , dan and a quarter south of Alloway. The recumbent megalit f irregulao h r appearance lies easwesd an t t bu t i mat y once have been uprightt I . measures 6 feet 2 inches in length. For a distanc 3 fee 4 f incheto e s froe mth western end it averages 3 feet 5 inches in widt beyond han d thi t i narrows s down roundea o t d endsurface preTh w . no e- sente bearind dan sculpturine gth fee2 s gi t d en froe e groundmth th side e d Th san . wese roundede th ar to t incisee Th . d cross t centrallino s y place thin do s stoned an , except for the head and arms, it is so shallow and weathered that in places the sculpturing is almost difficult to trace. It now measure fee3 s inche2 t lengtn i s h from the extremity of the head at the west to the end of the shaft which is still BLW5TQN. ALLOW fairly visible, but there are faint vestiges of the hollow for a-little way beyond this Fig . Incise9 . d Cros Boulden so t ra Blairston, Alloway. point. The upper portion consisting of head and arms, whose ends expand towards the sides of the stone, is more deeply cut and better preserved than the shaft. At their extremities the head and arms measure 10J and inche8 s respectivel taped inche2 yan o t rt theisa r intersection where the ye sam joie th shaf th nf e o twidth shafe Th t. gradually narrows down to If inch finally becoming indefinite. Probably the shaft was of uniform width originally t thibu ,s canno saie b t d with certaints ya the whole appearance of the cross is vague and disappointing (fig. 9).

SHORT CIST INHUMATION AT EDNAM, ROXBURGHSHIRE. On llth October 1928, Robert Mathewson employed by Mr J. Hamilton, Highridgehall, Ednam s completinwa , s day'hi g s ploughine th n i g ARDLUI MEGALITH THEID AN S R ASSOCIATIONS1 85 . north-east corner of a field known as the Haugh, when the horses' progress was arrested while taking the gradual rise from the flat near the River Tweed towards the Kelso-Coldstream road. The labourer noticed that, at a depth of about a foot, the ploughshare had fouled arid broken off a piece of a very large stone. Finding that the latter remained immovabl thad an ea cavitt s reveale e forciblywa th y b de removal of the fragment, he inserted his hand into the opening, and from the hollow beneath withdrew some bones which he recognised to be those of a human being. In Mr Hamilton's absence, the discovery was reported to the police . Davidsoe presencS th r t n Kelsoa D i f f thad eo o n an ,t tow e sointh l was cleared s see whewa n t i nthae stonth t e expose s entiretit n di y greaa s twa sla feeb4 inche9 t lengtn si fee s 3 wides incit y 1 t h b t ha t and 4 inches thick. This, on being lifted aside, revealed a well-made stone-lined grave at the bottom of which lay, on a bed of coarse river sand, pebbles and shingle, a number of decayed bones and two teeth later taken to Kelso Police Station. I visited the site four days after the find, and am taking this oppor- tunity of giving details of the antiquity accidentally brought to light. Mr Hamilton was good enough to accompany me to the Haugh, an extensive field bounded on the south by the Tweed and on the north by the highway. A hedge, the eastern limit, extending from the road rivere th o ,t separates Roxburghshire (Ednam parish) fro Berwicke mth - shire parish of Eccles. No indications of cairn or barrow structure exist at the site, nor is there anything to lead the inquirer to infer that there was a mound ove gravee rth . field-wallo Thern e ear s suggesting tha thein i t r build- g stonein s robbed fro ma prehistori c monumen d beeha tn usede Th . Haugh is profusely covered with small water-rolled stones such as can be picked up from the bed or sides of any river, so that in this regard no material is available to enable one to come to a definite conclusion. The grave, situated 90 feet above sea-level, is 103 feet from the north- east corner of the field and 55 feet south of the low hedge dividing Hauge th h fro roade mth . typicaa s i t I l shor fore trapezoida t th f cism o n i t e orientatioth , n of whose longer centra ldegree 3 axi1 s i s s eas f cardinao t l northe Th . sides and ends are extremely well constructed of slabs rudely but regularly dressee thicknesth o t d f 1£o s inchlone Th g. slabs measure 4 feet by 2 feet and 3 feet 2 inches by 1 foot 2£ inches internally, but s thea y overla e endspth inche3 , t leasa s t mus addee b t o givdt n a e estimatio f theino r full length endse Th .,t equa like sidesno eth e l ar , in size; that, at what for convenience may be called the north-east 352 PROCEEDING E SOCIETYTH F O S , APRI , 19298 L .

extremity, bein irregulan ga r quadrilaterafoo 1 inche0 1 p t to se th t a l in length and 1 foot 4 inches in height. Its opposite is 1 foot 9 inches long and 1 foot 7 inches high. The depth of the actual cist is 2 feet, and none of the stones forming it has been set vertically. At the top "widte th h acros s almossi t uniforml t bottofeeya 2 inc1 d t h man foo 1 t 9 inches peculiaA . r featur s thai ee shorte th t narrowed an re th f o r slabs formin t arrangeno e side s thay th g s topn sucwa si dit wa t ha , likuppee eth rthre s parit f eo t neighbours, should form perfeca t support e cover-stonefoth r . Care was taken, however, tha e loweth t r portion of the four slabs should be placed to form a uniform line at the bottom gravee oth f . Consequently ensuro t , should eli thae dth t rest evenly e substructureonth numbea , f flao r t stone s placewa s d p alonto e gth of wha s (viewei t d fro e south-westmth ) the right-hand slab. As I did not see the bones in-situ, I had to elicit what information I could as to their position when found. It seems that what little remaine e ciste south-eas th e craniuth . f th t o a f Fro do d y men la mt e dimensionth e cavitth f t i ywilo s l readil e understoob y d thae th t body \vhen interred was placed in a crouching position with the knees I coul s a d r learnbentfa o s ,, toward e soutth s e cisth th sid.f o e Apparentl o relicn y s were noticed amons ther wa e bone er th g no s tracy an f metal eo t t upome bu , n siftin e gravellgth y deposi whicn o t h reposed the skeletal remains I found a prismatic piece of mottled grey flint 1'8 inch long, '4 inch wide and '175 inch thick. One end and part e undersidth of e bear trace f secondaro s y working. Probable th y implemen a knife t s fro e worbu ,wa m th t n conditio e worketh f o nd seed edgeha n t i smuc h service. e boneTh se take b coul t nno dfro e custod e mpolicth th r f fo eo y detailed anthropological examinatio d reportnan t thebu , y consisted bonem d portionar an osd f e skullan partth f vertebrag o s f le ,o s e of aduln a t human being, probably male.