1. FORTS, CAMPS, AND MOTES OF THE UPPER WARD OF LANARKSHIRE. BY DAVID CHRISTISON, M.D., SECRETARY. (PLATES VI.-XV.) papea n I r rea thio t d s Societ n 1886i y I describe, d from personal observatio e existinth n g remain f prehistorio s c fort n Peeblesshirei s . Durin summee gth autumd an r s enablef 1889no wa I exteno ,t d y m d investigations westwards, so as to include the upper ward, or hilly portion, of the neighbouring county of Lanark. My measurements lack the advantage, and the check upon error, assistann a f o d whict ai woule hth d have afforded. They were generally e rougmadth y heb proces f pacingn o sseverai t bu , l cases, when tested with the tape, they were found to be sufficiently exact to give a good instancesw genera fe e objectsa th e lsiz f n th ,I o idee f .whe o a n greater accuracy seemed desirable, all the measurements were made with the tape. I have also checked my observations by consult- ing the Ordnance Maps, the Archaeological part of the first volume of The Upper Ward of Lanarkshire, written by the late Mr George Statisticaw Ne Verd an el d AccountIrvinOl f Newtone o g f th o s d an , Scotland. The Ordnance Maps are not so helpful as might be expected. The 6-inch Map, indeed s invaluabli , pointinr e positiofo ee th th t f ou gno forts; but the plans on the 25-inch Map, the only one in which the scal s largi e e enoug o givht e sufficient detail, have unfortunately been laid down with very varying degree f accuraco s d completenessyan n I . several instances, indeed n whici , e remainth h e quitar s e substantial, only the site is marked, or the existence of the fort is entirely ignored. The lack of system in the Archaeological Work of the Survey is further 2 28 PROCEEDING SOCIETYE TH F O S , FEBRUAR , 189010 Y .

illustrate Clydesdale th y db e remains being invariably called " camps," while those of the adjoining Tweeddale are as invariably called "forts." The accounr VerM ey b Irvingt , mainly intende o provt d e conth e - nectioe " campth f o n fortd s "an s with supposed Eoman roads, shows unconscioue th s bia f mino s d which writin a gpurpose r "fo " almost inevitably entail n matteri s f descriptiono s . Rude plan f eleveo s n forts, supposed by him to be Eoman, are given, which, as well as e measurementth f otheo s e rtext th fortn mosi ,n i s t instances differ considerably froe relativth m ee O.Md th planfroy an m n m.i s own. Here it may be remarked that in measuring old remains of e kinth d exactitud s hardli e y attainable, fro e difficultmth f fixino y g the precise limits of decayed mounds, apt to merge indefinitely into the surrounding ground. Other sources of difference in the results obtained, particularl measurementn yi f extremo s e dimensions e thaar , t one observer may include some exterior mound or trench which another y recko ma e interio e th nomits a on for f r to ro , from cres creso t t f o t the inner circumvallation, while another may reckon from the foot of the inner slope. Perhaps it would be an advantage if the interior of forts were measure commoy b d n consent from cres o crestt t s thesa , e offer tolerably fixed points. informatioe Th Statisticae th n ni l Account n generai s i s l either very defectiv r extremelo e y s agreeabl i vague d thee t an ever on nbu w , yyno surprise meeo dt t wit accuratn ha valuabld ean e description. Explanation of the Plans, §c. drawinn planse I th t I ,hav gou e followed mainl rulee yth s observed Peeblesshire papey th i nm n o r e forts. 1. The forts are placed as if the north were at the top of the page. 2. The scale for the plans is 120 feet to the inch, which was originally selected as the largest possible for showing the forts on a uniform r Proceedings. e pagescalou th f n so ei . Sections,3 when 0 feecomplete12 t e scale th t e als when ar ,bu ,o o n ramparte th e inch th 0 fee e 6 scalo .st th tf onl o shownee n o yar e ar , The sectiont quitno e sar accurat o scalet e , particularl o heightst s a y , but the accompanying figures give the approximate results. FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 283

heighe e rampartth 4Th .f o t s give s ntheii r greatest e heighth t a t side through whic sectioe hth n passes. . Trenche5 dottee sar d whe necessars i t ni distinguiso yt h them from ramparts. 6. The nature of the site is indicated by arrows, the steepness of slopes being represented by feathering from 0 to 4, the former representing a very gentl eclimbee slopeb stee o o lattee t s s th e ,d pa witon r h difficulty. approximate Th . 7 e height abov e neareseth t level groun gives di n ni figures beside arron a w pointing perpendicularly downwards. 8. All measurements are in feet. forte th ss havA names. o 9 en I ,hav e indicated sake themth f eo r fo , convenience name th somf y eo b , e neighbouring hill, house other ,o r object. (Platp ma ee XV.Th . )well-authenticate e give10 th l sal d forts known , includin f whico me w o t h fe o tracga littln r eo e remains. Thee yar numbere e figureth , ds54 froo beint m1 g place s nearla d s possiblya e over the localities, and corresponding with the figures attached to the e map th figuree texte n th ,fort th O .n s i s referrin moteo gt enclosee sar d circlea n thosd i an , e referrin rectilineao gt r fortrectanglea n i s . 11. Abbreviations. O.M. Ordnance Map. O.S.A Statisticad Ol . l Accoun f Scotlando t . N.S.A Statisticaw Ne . l Accoun f Scotlando t . The present inquiry includes all fortifications, whether of earth or dry-stone, which, although no precise date can be assigned to them, are commonly considere o havt d t leasa e t precede e feudath d l castlef o s stone and mortar. They admit of division into three distinct classes,— Motes, Kectilinear Camps or Forts, and Curvilinear Forts.

I. MOTES. As motes have been but recently systematically studied in England, and have been only incidentally taken notice of in Scotland, I shall make a few general observations on the subject before describing the Lanarkshire examples. s beeha n t I well established, particularl r GeorgM . y Clarkb yeT , F.S.A. (Mediceval Military Architecture n England,i 1884), thae th t fortress of the Saxons, at least as early as the ninth century, as described 284 PEOCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

Clarkr "a b mound,"ys M wa , ' " motte, r "burh""o r "e motou th ; f o a records," consisting of a truncated earthen cone, circumscribed by a ditch, usually connected with a base-court of a circular, oval, or horse-shoe form, enclosed by a bank and ditch. The mound was placed either in e end r on forme e court centro e ,t a th th enceintee r f th paro da e,o f o t . Olevee nwerth p to l e wooden buildings protecte palisadea y db whicy ,b h the outer bank was likewise strengthened. The number of these for- tresses was very great, and a considerable proportion still remain nearly perfect, excepting, of course, the wooden structures ; while many more can be traced, in greater or less detail, in spite of subsequent occupation by the Normans, who, in many instances, availed themselves of the advantages whic raisee hth d mounds offere strons da g defensible siter sfo their stone castles e remainTh . f innumerablo s e mote f preciselso e yth same kind .exist in France. De Caumont (Abecedaire d'ArcMologie, 1869) regards them as the fortresses of France in the tenth and eleventh centuries thinkd an , t probablsi e that havthey yema severae beeus n ni l centuries earlier. In the. Old and New Statistical Accounts there are numerous notices of motes, particularl e south-westerth n yi n countie f Scotlandso o s t bu , little attention have they attracted from archaeologists, thae worth t d occurs onl y r Proceedings,e Indetwicou th o n t xn i ereferenci d an e merely to brief incidental notices of two examples. In Muir's Ecdesi- ological Notes, e bes somth tf eo preserve d e south-westermoteth n i s n counties are described as unfortified, flat-topped, earthen mounds, with level terrace successivn i t scu e stages roun slopese t t dthith a Bu s si . variance wite descriptioth h e samth ef o n object e Statisticath n i s l Accounts, which represent them as simple mounds,, usually fortified by trencha terraceo n , s being mentioned; whil e drawingth e Grosy sb f eo the mote of Urr show both terraces, trenches, and ramparts. Evidently systematia c investigatio mucs ni h neede cleao dt thesp ru e discrepancies. noteworths i t I y tha nonn i t f theseo a description thers si mentio y ean n of a base-court,—so characteristic a dependency of English motes. Never- theless, in Lanarkshire there is undoubtedly a base-court at the Abington mot Dumfriesshir n . 285)i e(p d an , Moffae th t ea t mote. Terraced.motes do not appear to be confined to South Scotland, I am informed by Mr Proc. SOG. Antiq. Scot. Vol.XXIV Plate VI

MOTES ON THE CLYDE BETWEEN ABINGTON AND ROBERTON FOKTS, CAMPS MOTED AN F LANAEKSHIEE, O S 5 28 .

Clark that Beauforta te theron s i e, near r EicharInvernessM y b dd an , Ferguson that ther s anothei e r between Amblesid Hardknotd an e n i t Cumberland terorigie e th mth f o no t " s motA . e Scotland"n i , although sometimee b y ma t i s applied merel modera s ya n ter mf convenienceo t ye , fro frequence mth s occurrencit f yo e map th n ,e o eve n whe l tracnal f eo originae th l mot lons eha g disappeared d fro e applicatioan ,m th e th f no word to existing examples by the native peasantry, there can be no doubt that it is often a survival from ancient times, as in the instance of the Far f mMoteo describe nexn o d t e pageterTh m. also occur hisn i s - torical documents, as in the license (1430) for building Borthwick Castle, which conveys the right "construendi castrum in illo loco qui vulgariter dicitur le Mot de Lochorwort"; and in a grant of Dumfries Moit, dated 1521. . AUngton1 (Plate VI., lower figure).—A mile village nortth f n ho eo lefe th t e Clydebanth f ko . This fortification apparentl locao n s l yha O.Me markes i th t i .n name i dd "an , Camp," wels bui t i tl describes da a r VerN.S.A. e motM th en d ei Irvin an , g clearly s resempointit t ou s- blance smala n o , l recognisee scaleth o t , d great mote Englandf so e Th . fortifications includ base-courtea , wit hmouna d only slightly higher thari it in the south-east corner; the whole resting to the east upon the river. generae Th l elevatio base-coure th f no abous i t fee0 3 t t abov streame eth , the descent to which is steep. The mound or citadel is raised about 6 feet higher, and is cut off from the base-court by a trench not more than fee5 deptn i t h belo eascourte e levee th th w th tf n sid o lo moun;e eth d is defended by the river, and on the south by a tiny ravine, through which is drained a marsh, which covers the west or landward side of the base-court e latteTh s enclosei .r a singl y b de rampart, apparentlf yo earth d onl an , feey5 t hig t mostha , which, commencin e edgth f o et g a the river bank on the north, curves round to join the south end of the citadel-trenc ravinee th n fron I t ha f thi. o t s rampar norte th ho t t there are fain t wese tracee mars th trench a th to f t s o slightl s d hi ,an y hollowed, as'if it too were originally a much wider trench. The interior of the somewhat horse-shoe-shaped outer court measures on the O.M. about 240 by 180 feet. The flat top of the inner mound I made to be about 70 feet from . nort•• 5 froo southh3 t ' md an eas, westo t t . 286 PEOCEEDINGS OF THE SOCIETY, FEBKUARY 10, 1890.

Mr Irving mentions the remains of another mote within the grounds of Abington House, whic e "hlittla on perhap ee th highee th s s i p ru river," referred to in the N.S.A. 2. Robffrton (Plate VI., upper figure).—A hal a mild f ean norte th f ho e samth last en e rivermila o ,sidth d ef ean o ,sout f Eobertoho n village. The Clyde here makes a sharp, almost rectangular, bend eastward, the ground within the bend on the right bank being a low haugh. The left bank is high, steep, and rocky, and at the apex of the angle formed by the bend stands a farm called Mote (Moit in Font's " Survey," before 1630), nama e whic s justifieh i conspicuouse th y db , though small mound, sur- rounded by a trench, close to the house on the south. The mound has now a horse-shoe form, the centre having been dug out from the north. f therI ebase-courta eves wa r t wouli , d natiirally occup space w yth eno covered with farm buildings e viee farth wth m n I .buildings , &c., have been omitted. They stand on the rough bank to the right of the mote-mound. . Covington3 (fig. 1).—I e sequesterenth d somewhaan d t bleak cul de sac. of the Clyde, opposite Carstairs, an unusual number of objects interesting to the antiquary are concealed by an oasis of old trees, as protectior ifo f n froe destructivmth e moder th gaz f o e n " improver." village Th e scattered itselfan w t ,neawell-thatche d fe witdbu an s t it h d cottages and flowery gardens, half buried among trees, has an Old-World look. The church, although a mere barn in shape, retains in its south wall three deeply splayed Gothic windows with fine eth traceryn i d an , north wall a built-up round-headed doorway with a fine carved shield of Lindsayse th , dated 1659, inserte e moderth n di n masonry beneate th h arch outside. The ancient tower of the family, cruelly maltreated as it bese beens th ha ts i ,preserve d feudal northere castlth n ei n e reacth f ho Clyde; its well-proportioned, circular, and domed pigeon-house (H in the plan, fig. 1) is still nearly perfect; lastly, what concerns us most, the castle seems to have been built on or near the site of a mote. It e numbe o recordedth s e O.M. d th arrangemend s i an an n r ,o f o t e existinth g trenches seem incompatible wit a hScottis h castlf o e inferior size, although it is difficult now to separate the old from the less old. Their present conditio shows plane i sectiond th n an n ni s (fig. 1) . FORTS, CAMPS MOTED AN F LANARKSHIKE, O S 7 28 .

I! 41% r

^F g; =';.-.v.'.;;!i',''V'i'''''-Vi!'i:.'i''

' Fig. 1. Covington Moat. 288 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890. trenca 3 fees i r fee 0 to ) h1 deept2 r onl(A o wid d y8 , an etouchin g the castle wall on three sides, but absent on the north, where is the entrance. (B) is a second trench, from 5 to 7 feet deep, and flat bottomed, with obscure trace f sido s e walls, enclosin a gquadrangl e 150 feet square, somewhat towards the east side of which stands the castlee erectioTh . f farno m building s nearlha s y obliterate norte dth h side of this trench. (CCC) is a third trench prolonged southwards from the south-west comer of the last, turning to the east, then to the north, and lost at the farm buildings. It is perfect on the west side, but half shaved e higawath he easty y nearlb d th roa an ,n do y sida wholl y eb o ys e southth roat I inclosen o .d larga s e space, rectilinea wholee th n o r, but rounded at the angles, measuring about 450 by 350 feet, of which the castle, with its double ditch, occupies the north-west corner. On ease th t side this trenca sor f pendicls o t ha h e (D), lik gall-bladdeea r with its duct. At the south-east corner of the enclosure there is a small low mound (E). The rectangular enclosure (F) has no doubt been the terraca s gardeni ) e (G raisefeew . tfe dhighea r tha e gardee nth on n no side and the outer court of the castle on the other. Perhaps this may be a remnant of the mound of the supposed mote. The trenches are inhabitanquitd ol en a dry d t,an tole thadm o streat n thers r mo wa e spring at the present day from which they could be filled. The recti- linear e trenche th for f mo s againsi s e motth t e theory, althougr M h Clark mentions rare instances of the kind in England. 4. Biggar.— e BiggaCloseth o t r Burn a littl, e north-wese th f o t church; O.S.Ae calleth n di a .tumulus e N.S.Ad inth .an , . described as a large moat, 120 paces round the base, 54 at the top, and 36 feet high. It is a large artificial-looking mound, represented on the O.M. as a rectangle measuring 100 by 60 feet on the top, and referred to in The, Upper Ward of Lanarkshire, i. p. 361, as "the Moat-knowe." Vere Irving a motedoet cal t t describei no ls bu , s "artifician a a t i s l mound abou 0 fee3 t t high -with a fla, t topa rectangula f o , r form, measuring 91 feet by 50." The ground around it has been much altered by building and road making, so that exterior defences may have disappeared. 5. Carnwath.—The O.M. marks a mote here, and represents it : FOKTS, CAMPS MOTED LANARKSHIREF AN O ,S 9 28 .

as a circular mound, about 50 feet in diameter on the top. But Mr Vere Irving describes it as a remarkable cairn or mote at the west end of the village, somewhat elliptical in form, of considerable height, and surrounde deea y dpb ditch saye H s. ther a hollo s ei .thn wo e topd an , ascribes a sepulchral character to the structure. The writer in the O.S.A. says this hollow has been the entrance to a rude turnpike kind of stair, leading to the bottom, but filled with stones and rubbish; and besides ditche th , mention larga s e mound surroundin wholee N.S.Ae gth Th . . also notice e surroundinsth g holloe ditc mound,d th f han wo t remarks-bu , " it is said it was the entrance to a rude stair."' The Ordnance Map gives neither ditch nor outer mound. It is called a tumulus or moat in the O.S.A., and a cairn or moat in the N.S.A. Lanark.—A supposed Roman Castlehil e camth n po l her s describeei d Very b e Irvin elevaten a s ga d mound, partly natura partld an l y artificial, surrounde a ditch e y comparedb H . e Castlth s f ito Brambert >o e n i , Sussex, which is a mote. In the parish of Lesmahagow, about a mile and a half south-east of Dillar Law, a house is marked Moat on the O.M., and near it there is apparently a farm house marked Moatyett. Among other mounds, which possibly have been note e motesb y d ma , a small t a simplWolfe e on e Clyde (Culter Station), still existing, O.S.Ae th said n dan havo i t . e bee othero n cona "sightw n i t f s-a o t siderable distance : " also one at Bomflat or Bomphlet in Culter Parish, mentioned in the N.S.A. Couthallan.—On the O.M., at the ruined castle of this name, about a mile north-west of Carnwath, an oval space of 470 by 200 feet is re- presente s beina d g enclose y threb d e concentric mounds (?) marked " moats site t of.thaseeme I "m t o thesst muce ear h more likele b o yt remain ovan a f lI hav so fortt visitet eno bu , spote dth .

II.—KECTILINEAR FOHT CAMPSB O S . e subsequenth n thid I an s t class I ,shal l describ e fortth e s before making general remark themn so . 7. "Roman Camp" Clydesburn Farm.—Beginning from the south this e firsth t s i fortificatio e countys aboui th t n I mile3 i nt. s easf o t T VOL. xxiv ' . 0 29 PROCEEDING F THE'O S . "SOCIETY, FEBBUAKY , 189010 . .

Elvanfoo fro5 d m an tth e watershed which separate e waterth s s of' the Clyde from' those: of' the'Annan s aboui d tan , 1000 feet abov e seaeth . According to'Mr Yere Irving, it .was only discovered in'1858. On the 6-inch5 O.M. 'lais i dt :i regula; dowa s na r rectangl f 0 150o e 97 0y b feet; traceable all round,'but most distinction the north side. • Such is its present state, the better preservation of the north face being due to s beinit uncultivaten go d ground e sit .s on Th e wholi ; th ' e leveld an , in some' places marshy,.' although mostly under cultivation. It is tra- versed by two burns", one' cutting, off the south-east angle, which stands 20 feet above'the' rest of' the interior, :the'other, entering at 'the" middle of the north face, and emerging at the south-west angle. • The half of the north vallum, to the west of the latter, burn, stands on the foot of the slope of" Harleburn' Head-(1792) ground'fron o , fee0 2 t o t highe 0 m1 r than, "the interior commandes i t bu ,graduall e th y db y steepening slope'of the hill; the other half of the same vallum,'to the east of the burn, is on level ground, but is also closely commanded from the hill. The vallum is nowhere more than a foot or two in height, and from 12 1o widthn t 5i quantite Th . f stoneyo s half buried unde s turfd it r an , the size and adaptability of some of them for building, suggest that the original structure must have been of dry masonry, the materials for which lay close 'at hand,' as-the whole hillside'is" thickly strewn with large stone degrea o t s e quite' exceptiona soutn i l h Scottish hills n fronI . t of the mound is a trifling trench only 6 or 8 feet wide and one foot deep. . : d roadwayol n A ' , skirtin e hillsidth g e froe eastmth , approachee sth -nort enteo ht came f sidi rth s ea p along.wit burne hth t stoppin ,bu g short, •merges on a 'littl e plateau,—shore meany b p f stonedu so s embedden di earth, —which closely overlooks the camp, the road itself near this point being supported b y6-fooa t retaining wall, constructed of large' blocks. vallue Betweeth enormoun a d ms i an t ni s mas loosf so e stones, like eth wreck of a huge cairn, beside several nondescript mounds; and 100 yards west of the -burn, close in front of the vallum, eight or ten prostrate stonegreao n f tso sizes, savwhice feee7 on semicirclea s hti longn i e li , , and. may be the remains of a large stone circle. There are several irregular mounds hereabouts also. FORTS, CAMPS MOTED AN F LANARKSHIRE, O S 1 29 .

Altogether these remains deserv mora e e minute investigation thanI could mak stora n ei f winm o raind dan . 8. Castle Hill Plantation, Culter.—The site is 805 feet above the sea, on a gentle eminence near Culter Eailway Station, a quarter of a mil Clydee e th easabou d f 0 feeo an t, 15 t I faileabov fin. o t dit e d any trace of a fort here, but Mr Vere Irving describes it as "a rhomboid ofee4 fdiameter n 16 i t19 y 0b , defende singla y db e entrenchment,d "an representeo s s i e O.Mt i th n do . . Huntly9 Hill.—The cama sit f eo markes pi de O.M herth n y e.o b a dotted line enclosin ga nearl y rectangular space t •witbu , a slighh t e westercurvth n eo n side, measurin fee0 lengthn breadtn i g90 ti d an ,h s wides it e mil t on a narrowests e t 48 s it i 0 sout partt 0t a 54 I hd . an , of Cleghorn Camp, two miles east of Lanark, and one west of the Clyde, on cultivated undulating ground, only about 50 feet above the river. I could find no trace of mounds, and was assured by.an aged native that ther beed eha n s day nont hi bein bu ,n e i g ignorans time it th f e o t a t precise- position I did not examine the whole site. Probably it has entirely disappeared under frequent ploughing, as it was barely discer- nible at the time of the survey. The farm people, however, knew that a " Eoman Camp " had been there. This is probably " the small explora- tor ye Romans camth e opposit th f o pn o ,e Mouseth sid f o e from Cleghorn Lanarn i d an ,k Moor, e O.S.Ath f "o . 10. Cleghorn Camp.—This well-known " Eoman Camp," one and a half mile wese f Carstairsdu o t s village s representei , y Generadb y Ro l aa rhomboids f whicho , , however e lonth , g side e slightlar s d an y parallelly curved, measuring about 1800 by 134 plans 0t hi feebu n , o t called 180 e vallu 1260y texts b Th hi mgive.s n 0 i i n as entire from the slightly rounded north angle for 1450 feet southwards and eastwards0 70 ; also fro ease m th 0 t fee angl55 t r southwardefo d an ; fro e soutmth fee0 h 50 anglt r westward5 O.M.2 e fo th e n ,Th o . other hand ,a regula giveas sit r rhomboi onldof y 170 0100by 0 feet, with straight sides, marked by a dotted line only, except at the north angle wher e valluth e ms representei s runnina d fee0 g35 t southward fee0 68 t d eastwardan givey o entranceRo tw s . n theo s . north-west, and one on the north-east, each covered by a small, straight, detached 2 29 PROCEEDING SOCIETYE TH F O S , 1'EBBUAK , 189010 Y .

traverse in front. These traverses are not to be found in the O.M., and I did not observe them on the spot, but I did not search particularly for them. The greater part of the site is represented by Koy as cultivated, e northerth t bu n quarte s moorlandt a i stilr o ls remainsd an , , though covered wit hplantationa recentld an , deey b y p muc u drains t hcu . The groun boggs i dd peaty an yd ver an , y rough with heathed an r blaeberry e valluTh . m seem remaio t s n muc s representea h e th n i d t examinno d cultivatee di eth O.M. I t abous bu ,i d sitee t I parth t .f to 5 fee1 t wide d nowheran , e more than abou 4 feet t higher thae th n trifling, 3 or 4 feet wide, shallow, trench in front of it. The vallum is a pure earthwork, as shown in several sections through it. The O.S.A. mentions a small post or redoubt near the south-west came anglth pf eo tha t seems eithe havo t r e joine beer o t dni connected linea y wit.b t hi 11. Battle Dykes,- Car stairs.—Eoy's plan show sa rectangl 0 57 f eo by 540 feet inside, with a vallum and trench, each about 30 feet wide, on three sides, but with a mere trace of fortification on the south. The over-all dimension e completeth f so 0 footd65 wor ,y b k0 woul70 e db • without reckoning a space included by a line 20 feet beyond the counterscarp, possibly a narrow glacis, but which, in the absence of a sectional view unintelligibles i , . Strangely enoug e O.M ht th onl no . y reduce interioe sth 0 r 46 siz o t e by 440-feet, but represents the vallum as complete on the south side, where it is nearly entirely wanting in Eoy's plan. Moreover, on the O.M. vallue th , ms deficieni e nort th d two-third n han o t e westh tf so sides, and there is no trench. The N.S.A. states that it measures 6 square acres. " Though con- siderably injured by the plough and spade, the walls of ciroumvallation are still easily traced. The pretorium is visible, and the causeway to and from the camp, running in a direct line, can be traced for many miles." Mr Vere Irving says:—" The progress of modern improvement has, ingreaa t measure, destroye s rampartsdit smalA . l portio , howevernis , preserved on the side of the avenue at the back of the modern mansion- house "—a. position which doet corresponno s d with that e giveth n ni maps thareferrine o s b , ty "hdifferena ema o gt t work. FORTS, CAMPS D MOTE,AN LANARKSHIREF SO 3 29 .

This exhausts the number of undoubted rectilinear forts or camps, taky herp insignificano ema u e tw w t bu t work f doubtfuo s l character. 12. " Roman Camp t Sheri/cleuch.—"a Ten miles wes f Abingtono t , e Duneatouth p n water .- Thi s callei s a d" Roman Campe th n o " O.M., and represented, perhaps symbolically, by a very small rectangle. I could not reduce to any definite form the insignificant mounds re- maining. 13. Fort on HartMll.—A faintly marked oval of only 80 by 60 feet on the O.M., with a " font" 120 feet east of it. Mr Vere Irving describes s "a a t i rectangula feet7 8 r ,y fortificatio b whereo 0 e wes15 th ff to n corner has been cut off by a marsh. Its single rampart is slight and faintly marked. Near this is a curious font of very rude construction." I faile tracy fino dt f an fordmora r o efont o t t ebu , unlikel r moryo e difficult place to find either, on such a dreary extent of bog, far from human habitation, could hardly be imagined.

General Remarks on the Rectilinear Camps and Forts. Although the rectilinear works may conveniently be . treated separately fro e curvilineamth r ones t doei t , necessarilno s y follow that thee constructear y y differenb d t races t I was. , indeed, lone th g general belief tha l rectilineaal t r works might safel e ascribeyb e th o t d Romans. But that the Celts, or some other primitive race, might lay clai t mleasa somo t t f themeo clearls i , y proveoccurrence th y db o n f eo small number of rectilinear forts in Ireland, where the Romans never set foot (Archceol. Jour., 1867, vol. xxiv . 139p . , "On Roovesmore Fort," by Col. Lane Fox) t I become. a questios n then, whether rectilinear work primitiva y sb e distinguisheeb racn ca e d from thosa mor f eo e advanced race. I believe that, besides minor differences, the main distinction is to be found in the irregularity of the quasi-straight lines formerthe in , compared wit stricthe h t straightnes linethe the sin sof latterfirste factn Th i , . , have been laid down roughl th. eyby d eyean , e seconth d accuratel measurine th y yb g line. Thu e ruleth s d accuracy e " e th lineRoma oth n fi s n Camp st Cleghora " Clydesburd an n t na once strikes the eye accustomed to the irregularity of the lines, whether straigh curvedr o "te Britisth f o , h Forts."" Unfortunately this test cannot 4 29 PROCEEDING SOCIETYE TH F O S , FEBRUAR , 189010 Y .

be applied to the works at Castle Hill of Culter and Sheriffclench, believed to have been rectilinear in plan, as they have almost entirely disappeared. . .••.•''• e questioth t Bu n - still remains e thesar , e regular rectilinear works necessarily Roma I fea nr? thae earlieth t r archaeologist y havma se been too.read o accept y poiny t tsom ma ou thet eI s suchm a d an , difficulties which have occurred to me in the present inquiry. First, as to the'larger temporary." Camps," believed'to mark the northward marc f Agricbla'ho s army. General Roy, our best authority, only knef wo one-suc Uppee th n hi r Ward, at'Cleghorn looked an , vain di otherr nfo s about Elvanfoo d Biggaan t supplo t r y supposee linkth n si d chainn I . 1.858, however, the Clydesburn " Camp," near Elvanfoot, was discovered, o whichfas r s .inwa ,• favour of-the Roman theory t therBu • .e still remains the difficult o campye tw th thae s n th i diffet d n sizi ran e materials of construction, Cleghorn measuring 1700 by 1000 feet accord- e O.M.th in o gt , 180 134y b 0 0 accordin o Eoy't g s map d beinan , f o g earth; while Clydesburn measures 1450 by 900 feet, and is, as I believe, of stone s i wel .t I l ; also o remembert " e warninth - g give y -Db ne Caumont, who, referrin a contemporar o gt y accoun f Philio t p Augustus' camp at the siege of Tours, proceeds as follows :—" Ainsi 1'arm^e de Philippe-Auguste campait, comme les legions Eomaine s'entourann se t de fosses dont le vallum e^ait couronne de palissades en bois. Nous en conclurons que certains campements du Mogen-age peuvent offrir une grande resemblance aver les camps- Romains, et qu'ils ont du parfois etre confondus ave dernierss cce " (Abeceddire d'Archeologie, . 559)p . Passing to Roy's class of smaller and stronger permanent stations, or forts e grea th poiny , t t.1 ma ou differenct e between Ardoch, '.wits it h numerous small mound trenchesd an s , multiplie behine don 'other e dth , and such simple works as Fortingall or Battledykes, with their single massive rampar widd an te ditch: square redoubt n facti s , whics i t hi surel t extravaganno y o supposet t y havma ,e bee f comparativelo n y recent origin. - It is true we know little or nothing about mediaeval earth-works in Scotland, but we do know that field-works of some kind were thrown up, under the name of sconces, on the battlefields in the sixteent seventeentd han h centurie Britainn si thad an ,redouba t t called Soc. Antic}. Scot. Vol.XXIV Plate Vll

DISTANT VIEW OF BODSBERRY HILL AND FORT FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 295

Cromwell's Fort still exists at Eyemouth. It may be said that Battle dykes mus Komae b t na Roma becausn o s i n t t hereeroadi Bu ,. again, asky •wma e , e proowhath s fi t that this roa s Komadi I fea n1 r thaf i t not only its claims, but those of the numerous other alleged Eoman roads in Scotland were carefully sifted, they might be found to be of vaguese th mosd an t t unreliable character. Before passing from this subject, I may point out in a tabular form the great discrepancies between the measurements of Cleghorn Camp and Battledykes as given by Eoy and the Ordnance Survey.

Boy's Map. Roy's Text. Ordnance Map.

Feet. Feet. Feet. Cleghorn, 180 132y 0b 0 180 126y 0b 0 170 100y 0b 0 Battledykes (interior), 570 0 54 „ 460 „ 440

In the case of Battledykes the discrepancy is the more remarkable, as it is a comparatively small object, and is represented in both plans as perfectly well defined. When it is remembered that important argument n favou i sa Eoma f o r n origi e capacitye baseth ar nn o df o camps to contain certain numbers of men, these discrepancies assume a very serious charactere wholeth n t i tha O seem,e e studm . th t o f st yo Eoman camps, stations roadd , an Scotlann si d require recommencee b o st d de novo, with, alas ! greatly diminished materials since the days when it was first taken up. II, CURVILINEAR FORTS, District A. The left banJc of the Clyde and its tributary Glens, as far north as the pass by Biggar from Clydesdale to Tweeddale. On the highest feeders of the Clyde there are no forts marked on the O.M., exceptin e rectangulath g t Clydesbura e on r n already described. t sooBu n afte unitee th r d waters begi streae floon o nt m wn i southwards the first curvilinear fort occurs. 14! Bodslerry (fig. 2 and Plate VII.).—The isolated hill of Bodsberry (Boadsberry N.S.A.e th n i , ) rises very steeply fro e directiomth e th f no 296 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

•Clyde, which is J of a mile to the west, to a height of nearly 450 feet above the river, and 1312 above the sea. A deep and steep ravine cuts

300 Fig. 2. .Bodsberry. t ofi f from Lady Cairn Hill (1716 north-ease narrowee th th o )t t a d tr ;an ' north-west and south-east ends the ascents ' are 'also steep. The fort occupies the whole available -somewhat irregular top, and consists of a CAMPS, FORTS, AND MOTES OF LANARKSHIRE. 297

single rampart, rounded at the ends but tolerably straight at the sides, suio conformatioe t th t e comparativele groundth th t f no A . y inacces- sible sides the rampart is of trifling dimensions, whereas at the more accessible endbroas i verd t si dan e north-wesy th stony t A . t end0 ,4 feet down the slope, is an advanced rampart, which, beginning at the north-ease edgth f eo t ravine s straighi , t thed firsa t an nt curves round e south-westh e lostn b oo t t r faceIrvingM . somy b , e strange error, makes this fort hals propeit f r calle sizeRomaa H t si . n castellumr fo , which I can see no sufficient reason, and makes a Roman road pass through it, of which I could see no evidence. The singular basin-like well which he discovered (Upper Ward of Lanarkshire, p. 9) has been filled up. In the view, Plate VII., the fort is seen as a faint circle on the top of the conical hill above Newton House Woods. About 3|- miles down the Clyde from Bodsberry, and on the same e riverth sidcome grouf a w e, o o et f threpo e forts, withi e spacnth f eo a mile, near the village of Crawford, on the skirts of Southwood Rig (1556), each of them about 100 feet above the river, and 950 above the sea. . East15 Crawford.—Three-eighths a mil f o e north-wese th f o t ruined Crawford Castle conspicuoua n o , s solitary frof knolle of m th t cu , uniform slope of Southwood Rig by a trench-like natural cut, and pro- jecting int Clydee plaie th o th f no . irregulan " A 12y b 09 r 34 ova f o l

Sectio Easterf no n Fort, near Crawford. feet, fortified by a single rampart, which, on the side next the river, is very indistinct" (G.V.I.). As Mr Irving's figures agree closely with the dimensiont measurno O.M.e d th di e n I ,so thi s fortaddition I . n ramparte toth , whicside th e n quits h i o endo d etw san distinc e th t a t "298 PROCEEDING E SOCIETYTH F O S , FEBRUAR , 189010 Y . towards the hill, as a green smooth mound, but of which I could see no trace towards the river, there runs on the- side towards the hill, at a varying level of 5 to 10 feet below the rampart, a narrow ledge or terrace about 5 feet wide, which, with some kind 'of parapet now gone, no doubt formed a second line of defence" on this the weakest side. . . - 16. Mid Crawford (fig. 3).—The site is' about half a mile westward of Crawford Castle s i lastsimilae ruins th d , s moro an i onlt ,r t i ye

^^"SiifP1

Fig. 3. Mid-Crawford. detached from the hill. The single rampart, slight on the high south side, is substantial elsewhere, and appears to be of earth, although a few large and small stones are visible in breaks. The interior rises several feet higher than the rampart into two eminences of about equal size, and flat on the top, but leaving a space or roadway all round between them and the rampart. There is no evidence of these eminences having been fortified round the edges, as might be expected,-but there are faint traces of circular foundations on them. .Outside the fort, at the foot of CAMPS, FORTS, AND MOTES OF LANAKKSHIEE. 299

the west end, there are remains of two circular foundations, one of them 90 fee diametern i t . . West17 Crawford (fig. 4).—700 yards west-south-wes e lastth .f o t site peculiars ei Th littlA . e ravin furror e weso th twn i slop f Southeo - wood Big, bifurcates about 150 feet from the bottom, and thus isolates a small projectin r gfeet6 M 6 knoll .y b ,4 wit 14 n ova a hf o l p flato t Vere Irving say thif so s fort: " Being commande y highedb r grounn di the immediate vicinity, it is ill adapted for a military post, and seems

Fig . Wes4 . t Crawford.

rather to belong to the enclosures of the sepulchral class; this idea is further confirmed by the existence of a cairn of stones in its interior, beneath which two stone cists were found. In one of these was a capitally preserved urn, with a bracelet and two spear heads of bronze , (figure39).36 . " dpp Notwithstanding this sepulchral e "findth d "an close command from the hill (which is frequent in prehistoric forts), the remains strongly suggest an oval stone fort of dry masonry, a portion of which still continues e visiblnortth n h o e remain sideTh a .f o s transverse wall in the interior and of additional defences down the slope showe Plane trac so ar th cairna I couln f n .ne i e o dse . 18. Campswater d Platan (fig.e5 VII.).—Situate vera n i dy remote 0 30 PROCEEDING TEF O SE SOCIETY, FEBRUARY 10, 1890. CAMPS, FORTS, AND MOTES OF LANARKSHIRE. 301 spot among the hills, near the eastern boundary of the county, upon Grains burn, a northern tributary of the Campswater, 4-J miles above its junction with the Clyde. The site is on the east side of the burn, nearly 300 feet above it, 1250 above the sea, and 600 yards from the junction with the Campswater. Although not isolated, the position is strong, being protected on the north flank by Casan Cleuch, a precipitous ravine, 50 feet deep; in rear by an equally deep transverse cut in the hill, which separates it from the continuation of the slope upwards to a summit (1612 Fairburf )o gentle (1779)g fronth n Ei t i y d ebu b t an ; long slope to the burn. At the south end it is weaker, as a small triangular plateau t littl,bu e lower t thainclude forte no e nth s th i , n di works. The mode in which the ancient builders varied their fortifica- tions according to the nature of the site is well illustrated in this case. The comparatively weak south end is fortified by four concentric ramparts and trenches. These ramparts vary in construction. The outer and inner ones are ordinary, wide, rounded mounds; but the second, reckoning fro e outsidemth s narrowi , , steep d sharp-crestedan , , while the third has a flat or slightly hollowed top, 18 feet wide in the middle t gettinbu , g narrower toward e endsth s , affording ample space for a large body of defenders. Viewed from the triangular plateau out- side l foual , r e risinrampart on e seen b ge n th ,behin ca s e otherth d . Following the lines northwards to the west face, the outer rampart and trench and second trench are lost, but the other three ramparts and two trenches remain to protect the south side of the entrance. On its nort hsystew sidne ma e begins, consistin gn outea firs f o r t rampart, which runs only halfwa Casao yt n Cleuc h; secondly rampara f o d , an t trench, gradually lost on the Cleuch; and inside of all, a trench with a high scarp, which, howtfver, instead of a parapet, carries, a little with- drawn froe edgemth a ston, y mound, apparentl e foundatioyth a f o n substantia shora l r walltFo distanc. e extremth t ea e nort d therhen e has apparently been nothing but a slight wall at the edge of the pre- cipitous descent to the Casan Burn. Turning round southwards to the rear of the fort, we find another system. In the northern half of this o stee side slopto eth o admit p s i e f entrenchmentso t d insteaan , f o d them two 4-feet wide terraces have been cut, one below the other, which, 302 PROCEEDINGS!OF THE SOCIETY, FEBRUARY 10, 1890.

when furnished with some kind of parapet, would constitute, along with the wall at the edge above, a triple line of defence. A re-entering curve in the ground, with an easy slope, follows,-where probably there was another entrance, but all trace of fortification here, excepting the top mound losts ,i secone th ;third dan d ramparts of .the south system, how- ever, curve round, so that their ends completely flank the re-entering approac westhe o ht e Th ttop entranc. e (Plate VIII. wels )i l preserved; the ends of the ramparts flanking it ' are -not opposite each other. Instea thisf do ramparte th , side opposite on ear f s o trenchee eth e th f so other, and, owin o thgt e conformatio grounde th f bottoe no e th ,th f mo trenches on the south side are above the present level of the entrance, whil ee nort thosth beloe n ar h . I :o thine it w k there are tracef o s circular foundations in the interior, but only one semicircle—on the north sid f theo e . entrance—is quite indubitable; 'In e breakth n i s ramparts a few large and small stones are visible,' but no building. ' Returnin e Clydeth d proceeding'onlo gt an , y a'mila quarte d an e r e threenortth f ho . Crawfor de samfortsth e en river o th ,sid e f w o ,e encounter another grou f threepo , withi space milena th f eo . . Arlory19 d Plat an (fig e6 . IX.).—The forthen o t ' summi f thio t s steep conical hill, 1406 feet above the sea, looks down upon the Clyde, half a mile westward, from a height of 600 feet. The slopes on three side long'ane sar e eas th dgentlea t o steept t rbu , descen o f t onl o t 0 y8 fee0 10 t connect f Tewsgilo p e sitto th se le wit Hilth e ris hth o t el (1867). The triple circumvallation is well seen from the road over the pas o Campswatet s r (Plate e interioIX.)Th .s leve i d rnearl an l y circular, measuring from 130 to 135 feet in diameter, and contains low nondescript mounds. It is enclosed by a zone of tumbled stones, 25 to 35 feet wide, and 3 or 4 feet' high in the middle, with no visible remains of building groune Th ' . d falls gently from this distanc a " t ring,a d e"an from it varying from 18 to 40 feet comes the second "ring," consisting osligha f scarpa t f parapeo ,p whichto te ath t ,counterscarp w witlo ha , forms a trench, the bottom of which is from 6 to 9 feet below the top of the parapet. A second interspace, varying from 18 to 50 feet "in width, follows, which, where narrowest, on • the south and west, assumes elevaten e fora th f mo d platform with trifling parapet fron o st d rear an t , Proc 3oc. Antiq. Scot. Vcl.XXIV HateVffl

ENTRANCE TO OAMPSWATER FORT Proo. Soc. Antiq. Scot. Vol.XXlV Plate DC

ARBORY HILFORD AN LT CAMPS, FORTS MOTED AN ,F LANARKSHIRE O S . 303

but elsewhere merely slopes gently to the outer "ring." This closely resemble e middlth s esweepinn i one t bu ; g round from nort o eastt h , it diverges outwards so as to end at the north side of the entrance,

Fig. 6. Arbory. e samth e f souteo th full rin d n h0 feeo g2 yen n advanc i t e th f eo e entranceth sid f o e . Three other entrances pierc e outeth e r rings e probablar only d an , y ancient, becaus e openingth e s o througtw e th h 304 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

e mai th e cas f th n eo rings n entrancei t opposits a ,no e ar , e each other. structure th ringe o t th s f se o A ther be-littln eca e doubt tha e inneth t r one has been a thick stone wall, of which the large stones have been plundere e inferioth d an dr ones alone remain e outeTh r. rings have apparently bee f eartho n , perhaps faced t certainlbu , y strengthenen i d somwity wa he stones bots a ,scarpe hth counterscarpd san s have bulky stones, som f theeo mfee3 t long, r hallyino fn go embedde themn di d an , partlrowsn i t y.se 20. S. Coldchapel.—The site is slightly raised above a nearly level field at the foot of Arbory Hill, 400 yards south of Coldchapel Burn, e Clyde th eas0 d f onl15 o tan , y abou 0 fee 3 te singl t Th above . it e rampart was in excellent preservation about forty years ago, according o representetotalls w r Irving s no i e M O.M. s y th i d o t t n an , o d,bu destroyed peasan A spoe .th t n o tolte tha n enormoum da t s quantity of stones had been carted from the rampart, but that,none were found in the interior. A low mound still marks the ground plan, which is undoubtedly a circle, 180 feet in diameter over all, as the O.M. makes it, although Mr Irving unaccountably calls it rectangular. 21. N. Coldchapel (fig. 7, the small inserted view).—Half a mile north of the last, 300 yards east of the Clyde, 50 feet above it, and 850 above the sea, on the slope of Priestgill Rig (1361), where it becomes quite easy towards the river. It consists simply of a rampart, nearly circular in form, whic s beehha n much knocke de sakth abou f plunderineo r fo t g the stones concealed unde e greeth r n mound I foun. tapy e b dth e dimensions of the fort to be 306 by 287 feet over all, and 246 by 220 inside. From faint marking groune th n s o d I thin interioe kth bees rha n filled with hut circles, and this probably saved the fort from sharing the fate of the lastj as I was told it was found " ow'r hard to ploo'." 22. Shiel Burn Platd (figan e.8 X.), 1000 feet abov seae eth , lies2 miles north-east of the last, among the hills, on the Shiel Burn, \ of a mile above its junction with Wandel Burn, and a mile and a quarter from the Clyde. Its situation at the north end of an extensive, high-level, gently sloping moor, and on the edge of the steep ravine of Shiel Burn, s showi Platn nlone i oval n th . a g ef f X whics I o i o ,sidet e s hson ha been, as it were, cut off by the ravine. Here the defence has been to X

o 33 FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 305

o

13a

VOL. XXIV. 306 PROCEEDING E SOCIETYTH F O S , FEBRUAR , 189010 Y .

merel a slighy t wal r ramparo l t running irregularly alon e edgeth g . Toward mooe sth r ther massivea s ei , rude, earthen mound, wit h1a o 0t fee0 2 t wide shalloww no , , trenc marsa n fronti h d han , beyonde Th .

enclosed space measures about 150 by 80 feet. The .entrance is from e eastth ,ravine e closth o et . 23. Devonshaw (fig. 9).—One mile north-west of the last, 1000 feet 0 abovabov 30 e Clyde sea e th ed e th ,an , 0 yardwhic30 ss i h distant. fore broaa Th ms i d oval 16y f aboub 0,o 0 fee20 t t inside defencee Th . s are obscured by dilapidation, but the inner line seems to have consisted of a considerable rampart of earth and stones to the east and south, where e naturath l strengt a les f so s leasti hsubstantiad an , l wall f whico , h socket r stonesfo f somso e size remaine otheth n r o lowea , sides t A r . FOBTS, CAMPS, AND MOTES OF LANARKSHIRE. 307

level of from 8 to 15 feet an outer line seems to have been formed, consistin naturaa f go l trence eastth ,o ht whic h southward passes into a space 40 feet wide, defended by a low mound, again becomes a natural trench on the west, and passes on the north into a narrow terrace cut in

II A f \

Fig. 9. Devonshaw. steee th p surfacslopee interioe Th th f . eo irregulars ri , rising higher than the enciente, and besides natural rock there seem to be remains of mounds d perhap t circle. o entrancesan hu it , Thertw f n i o e s ar e , one from the south, the other from the north-east. 308 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

Between Devonshaw and Lamington, the U.S.A. places two forts—one at Hartside, " oblong, 90 by 60 feet, having a rampart on all sides ; on the north, west, and south it is protected by a rather steep brink, and f Starthop o deea ease op ny th p b tto ditch. e othere th Hil Th n "o , l (1399 feet) s describei , rina f earts go a dd stone an h s abou 0 fee6 t t in diameter. Another is mentioned at Braehead, a site which I cannot identify. None of these are in the O.M. 24. WMtehill (fig. 10).—From Devonshaw we have to proceed five miles down the Clyde before encountering the next fort. It lies quite apart fro hige mth h gentl e hillsth n eo , southwar ridgw d lo eslopa f eo

v,'1' "•, v ,l«v^;l..--'.. - 'v/;/''-,- i'"' ~\ \

\ /I &1. .-..., .'••>••• |i Mhltia.^"1"

Fig. 10. "Whitehill, Lamington.

called Whitehill, and is very conspicuous from a distance of several miles to the west, although, when examined, it proves to be ill preserved and difficul detail n e i plamako t tt Th n ou .e shows , however, that i t s beeha na larg e wor f unusuao k l form, consisting apparentl f threyo e ovals, somewhat compressed by being joined togethe n line i rt eac bu , h separately defensible, with double mounds in some parts, and additional ramparts thrown out at the east end. The total dimensions are about 550 by 110 feet, according both to the O.M. and my own measurements. Proc. Soc. Autiq. Soot. Vol.XXIV Plate 1X1.

NIS8ET FORT AND GLEN FOKTS, CAMPS MOTED AN F ,LANARKSHIRE O S . 309

The N.S.A. describes this as " three camps quite adjoining one another, all of considerable dimensions, the largest 210 feet long and 120 broad, wit hditca fee5 h1 feet0 wide.sit80 e t s eabovi Th "seae eth , 100 above the Clyde, and half a mile south of a bend of that river. e resemblancTh f thio e s for grounn i t d plaa Frencsitd o nean t h mote,

:.-•! ' v< .-' '. -x- .£C .-. 3:. i J &***< .^'^-^ ?f

ioa Fig. 11. Nisbet.

the Chateau d'Aulnay, figured in De Caumont's AMcedaire d' Archeologie, . 398p , although probably accidental certainls i , y most remarkable. We next encounter a remarkable group of six forts, within a radius of little more than half a mile, on Culter Water. . Nisbet25 Platd an (fig e1 VI.).1 . — Fully 1000 feet abov e seath e , terminae th f o p l to ridgespua e of no th r , three miles long, descending 310 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

northward from the summit of Culter Fell. The site is easily accessible up this spur from Nisbet Farm on the north, and is only cut off from the e ridgascenth f e o tsouthwar a sligh y e descentb dtth dipt bu ,s of about 250 feet to the Culter Water on the west, and 150 feet to Nisbe te east e th steepBure vie ar ,n Th o nw . . eastwards p Msbeu t Glen to the sources of the burn, with the steep slopes interlacing at e footd entirelth an , y covered wit e e finesth th h f f o heatherto e on s i , kin n Souti d h Scotland n shapI n ova a e f .forabou o s lth ei 0 t 21 t by 150 feet inside. The poor remains of three concentric ramparts can be easily enough made out on the north and east, and more particularly f e i therweste southos th i nth t t en i a middl,bu o , s evee wa on re entirely spac e goneTh e .betwee innee middlnd th an r e ramparts seems to have been a platform and not a trench, and on the long sides the defences assume somethin e terraceth f o g d type. Touchin e outeth g r line, but not included by it, there is a curious isolated mound. Inside, e nortth t ha end, ther a sharply-define s ei d circular depression fee6 5 , t diametern i e flath ,t botto f whicmo s strewhi n with small stones, like basdestroyee a f th eo d cairn towardd an , w e sout lo th thersd a h en s ei round rocky mound. Ther e alsear o remainrectangulao tw f o s r struc- tures. Ther three ear e entrances difficul s i w whicy t i ho sa r t o ho t t bu , many are original. This fort was measured by tape. Castlew Co . (fig26 . Platd 12an , e XII., upper figure)e southth n o -, east side of a little trough-like valley which opens on Nisbet Burn at its junction with Culter Water, and which, skirting the northern buttresses of the Culter Fell range, affords an easy pass to Kilbucho and Tweed- dale. The site is at the south-west end of an isolated ridge, a quarter of a mile long on the summit, 1040 feet above the sea, about 300 feet 0 fee10 t f abov froo a stee e valleyp y mth eb pdi entireld f an of , t ycu Blackhill seriea e firs f th ,o f teminence so s ending southwar Culten di r defencee FellTh .fairl e ar s y preserve e complear d plann an dxi , com- prising a central work, with additions pushed out to north and south. The interior is an oval of about 150 by 102 feet, and rises abruptly on south-ease th t int omound—partla y natural, probably t ver,bu y stony— 15 feet high to the inside and 8 to the outside. This completely con- ceals the interior, which otherwise would be much overlooked from the Proc. Soc. Antiq. Scot. Vol.XXIV Plate 211.

COWCASTLE AND SNAIP FORT FORTS, CAMPS MOTED AN F ,LANARKSHIRE O S . 311

slop Blackhillf eo describinn I . defencese gth t i wil, conveniene b l n i t meantime th north-wese leavth o e t t eou othee th t n sider threO . e sides innee th r rampar surroundes i t seconda y db , whictrenca s hha fronn h i t to north and south, but on the west is nearly continuous with the steep slope of the hill, although a narrow path probably indicates that the

ISO Fig. 12. Cow Castle.

second s carrielinwa e d round here stockadea als s a oe souther Th . n outwork is a crescentic space, 50 feet wide in the middle, fortified, at the very edge of the south-west end of the site, by a double 312 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

rampart and intervening trench. The northern outwork, of similar size, somewhas i t triangular s defendei d an , d toward e levesth l continuation of the ridge by a rampart and trench, but on the edge of the steep slope to the south-east merely by a trifling mound, perhaps the remains of a e connectiowallTh .e outeth f rno line f defencso north-wese th n o e t face with the rest cannot be precisely made out, but they are represented by two broad, straight terraces cut on the face of the slope to the valley. The entrance is at the north corner, where it is flanked on the west by a sharp bend of the terraces, which here become a double rampart with intervening trench, and on the east by the northern outwork. The passage seems then to turn sharply to the left between the inner and second rampart, and finally, turning sharply to the right, and flanked by a high mound which connects the two ramparts, pierces the inner one. Its course is marked on the plan by a dotted line. This fort was measured by tape. ridg e detachea north-ease th A s th ef i t o d d en tova 8 l 10 wor f o k feeb0 y7 t inside e singlTh .e rampar s weli t l preserve e nortth t ha d d soutan h ends. Perhaps ther s beeha e na terrace d defenc 0 fee1 e t lower on the north, but the plough has destroyed any additional defence that may have been on the other sides. . Langlocli27 (fig. 13).—Looking acros gorgee sth fee0 10 ,t deep, from Cow Castle to Blackhill, the double lines of a fort are plainly visible, as if engraved on the face of the hill, at about the same level as Cow Castle ; buapproachingn o t site littla founth s ,e i b e o moundt d detached from the hill by a slight dip of 20 feet. The work is an oval, of about 100 by 70 feet inside, encompassed by two lines of defence, the outer 10 to 15 feet d diverginbelo an e inne we th on rg considerable yth t froa t mi two ramparte endsTh .muce ar s h destroyed, particularl e nortth t hya end; but where better preserved there is much stone about them, and this has probably been a stone fort. A third rampart has been inserted between the other two at the north and south ends only. In the hollow between the two forts there is a " hut-circle;" on the way to the neigh- bouring Msbe wese t th fort tn o sid, f Msbeeo t Burn, ther anothers ei ; and about a mile east of Cow Castle in the pass to Kilbucho there is a third. 28. Lower Snaip.—Low down on the slope of Snaip Hill, 900 feet FORTS, CAMPS MOTED AN F ,LANARKSHIKE O S . 31J

abov0 10 ed Cultean a abovse r e "Watereth quartea , mila f ro e south-west from Snaip Farmhouse. Describe r IrvinM y s beinga b d g "defended by a single rampart of irregular form measuring 153 and 128 feet at its two widest points. O.Me Th ". make irregulan a t si r hexagon, measuring

^ .W.Wi/ii/M* JiiU'S'i/''^ • (

Fig. Langloeh13 . . about 175 by 170 feet, and gives a mere outline without details. It has been so knocked about that I could make out nothing definite about it. 29. Upper Snaip (fig. 14, and Plate XII, lower figure).—On the top of Snaip Hill, 1187 feet above the sea, 400 above Culter Water. The position is very strong at the north-east end, from its height above Culte e north-westh r n Watero d tan , face, becaus e grounth e d falls steeply for 300 feet to a ravine separating Snaip Hill from a lower height, crowned by Culter Park fort. But to the south-west there is e continuatio e th fee0 th 4 o t tr n ridgeo e o th 0 d 3 f an n, o f o p onldi ya south-east side it is separated from a somewhat lower part of the summit merel a narrow y yb , probably natural, trench e defenceTh .e mucar s h 4 31 PROCEEDING SOCIETYE TH F O S , FEBRUAR , 189010 Y .

decayed, but consist of an inner somewhat dumb-bell-shaped enceinte, probably originally a wall; going round from 10 to 15 feet below

Fig . Uppe14 . r Snaip. it is a terrace, partially provided with a parapet, but lost on the rocky north-east point d alsan ,o wher e hil e souts th esteei lth t hpa corner. FOETS, CAMPS MOTED AN F LANARKSHIRE, O S . 315

There are distinct remains of a third rampart at the two ends, and though a connecting-lin hardln ca e e north-wes tracee th yb n o d t side on the spot, it is distinctly marked on the hillside, as seen from Culter Park fort. On the south-east side there is no third line of defence, but its place is partially supplied by the trench already mentioned. A great part of the interior is occupied by two rocky mounds, a few feet high. curvew lo dA artificial mound runs acros e southerth s n one, from side to side of the inner enceinte. 30. Cutter Park (fig. 15).—About 1000 feet above the sea, on the

If v el

Kg. 15. Culter Park. top of a hill which faces Snaip Hill (1187) from the opposite side of the ravine already mentioned, and 150 to 200 feet above the bottom of the ravine. The two forts are about 400 yards apart. The interior is level, 316 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890. irregulan a s 0 i feet 12 d r , an y ovaenclose b f abou o l0 22 y subtb d - stantial mounds at the two ends, passing into trifling ones at the sides. t occupieI s only abou levee t th hal f l fo souts topit t h,a end. twentyA . - feet wide terrace, 10 to 12 feet down the hill, runs round the west and south sides, becoming a mere narrow ledge on the very steep east side, passind an g int trenchoa , with rampart beyonde nortth n ho , towards e unoccupiehill-tope th th f o d . den

Fig. 16. Hartree. . Hartree31 (fig. 16).—Quit e loth w n grouni ee broa th f do d easy pass or strath from .Clydesdale to Tweeddale, 657 feet above the sea, on FORTS, CAMPS MOTED AN F LANARKSHIRE, O S7 31 .

e westh e roat th sid df eo fro m Hartre hundrew fe o Biggar t ea d an , yards south-west of the ruins of Boghall Castle. From its situation in s beeplaie ha th n t n i long e e subjecploughramparth th d o t s ti an ,t almost obliterated excep s scarpit t , which, wit wide hth e trenc n fronhi t fragmene th n outed a f an o rt rampar norte th t ha t end e stil,ar l quite distinct. It is an oval, the east side of which has been shaved off by s situatei e roada d nearlth an n ,i d y level field, with scarcely an y natural advantage whatever e southerTh . interiore n th hal f o f , however, rises 6 feet above the rest, and nearly 20 above the trench to the south, where the ground falls somewhat. This elevated mass seems too large n thai d t an case wor , th e so k e shoulb y de ma artificialb t i o t t ye , perhaps be regarded as a mote rather than a fort.

B. Forts on the left bank of the Clyde and its tributaries, as far north as the westward bend of the river at Garstairs, vere th y n uppeI r par Lanarkshirf to e ther remarkabla s ei e absencf eo Clydeease e e wes th th balanco fort th tf T n sideo tn o . so n ,e te lyinn gi spac e mile8 th f eo s from Bodsberry Hil Devonshao t l w Hillt , therbu e ear three in the much more extensive mountainous region to the west, and while all but two of the former look down upon the Clyde, the latter quite ar e remot vallee th e f Duneaton yfroo i t mi n Water. Moreover, e threeth tw f f triflino o , g size, were probably rectilinear d havan , e been included by me in the rectilinear class (A), p. 293. . Blackhill32 smoothly-roundea 17).—Of d (figso an p 7 .to e nth d spur, 1100 feet above the sea, of Blackhill (1260), which stands within a semi- circular bend of Duneaton Water, 300 feet below. In two directions the ground fall e south-eassth fro streame fore o t t oncth ma th tt o et bu t, worke th somewhae ar s t withdraw e oppositth n n fro o edgee emd th an , side the site is connected with Blackhill by a wide, nearly level neck. viee wTh froe for s mth strikingi t e heighth whicn s a o ,t t hi stand s e centr circlappeara th f e highef o eb o e o st r hills, through whice th h streaa narro n i mwy forcesemicirculawa s it s r gorgee prospecTh . s i t hence much circumscribed, although the projection of the site into the 318 PEOCEEDINO SOCIETYE TH F O S , FEBRUAK , 189010 Y .

bend afford a vies w westwar p Duneatodu ne Watepeea th f d po an r distant Clydesdale north-eastward a circle for s f i abouTh to e 5 . 24 t feet diameter over all, the defences consisting of two " rings." The outer one is of earth and stones, and has a slight trench in front towards easo ytw e innee sidesth Th r. one t concentric—ano , e closs i th t o si et south-weste th n o oute e t leaveon r bu , considerablsa e interspace th n eo north-east,—has been in all probability a strong stone wall. It now

e^ loo Fig. Blaokhill17 . .

5 feeform3 t o zonea t swide 5 2 f ,smallis o , h stones, partly overgrown with grass "A ree,. circular "o r sheep-pen, showplane th ,n n i account s for a number of the larger stones, but a few still larger, three feet in length f these, o als e aboute li o, On apparentl. e inne th n situi yt ra entrance, admirably adapted fro s msmootit h regula ra casinfor r mfo g stone, is perhaps an evidence, and the only visible one, of building. The only entrance is from the north-east and passes straight through both " rings." FORTS, CAMPS MOTED AN F LANAEKSHIRE, O S . 319

Ther e fainar e t trace f traverseo s s on eac e passaghe th sidth f o en i e interspace, and more distinct remains of e anothesouthth n o .r There a goo s i d dea f shapelesso l , over- grown, stony debris within the inner "ring." othee forto Th Duneatop rtw u s n Water have been already noticed at p. 293. 33. Newton (figs. 18 and 19).— e hav W proceeo et mile5 d s north- ward before encountering the next fort. It is 3 miles west of the Clyde, projected from the south into the wide valley of the Garf on the summit (1048) of a spur of Dun- garvel (1675) positios It . showns i n - ap e Th , figwooe . A th .18 t da n i proac eass hi y alon e necth g k from the south, and in the other direc- tions the ground slopes gently to the e streamth fee0 valleyo t 20 t, d an , below. The two massive ramparts of this slightly oval fortolerable ar t y well preserved, except where the plough has shaved away the outer one on the flanks. They are appar- ently of earth and stones; the inner one, particularly toward northe th s - east, is flat-topped, and the outer one becomes a wide terrace to the south. e entranceso th Ther tw n i e d ar e an , interior two massive curved mounds which touch each O.Me otherTh .. 320 PROCEEDING E SOCIETYTH F O S , FEBRUAR , 189010 Y . most unaccountably gives onl ramparte yon r IrvinM . g rightly describes e twoth .

nM'.'p|W"/'/ "'#£ ' '>\ •&A,*&'\ ^. %~ ^ % ^\ \ ~ 2 '<-A ...cf .-«' ,, —--^ _-— •>^> ,v,^ t ? "^ ^-M^ •'•;> /= "3 S \i« \\\» * 5 S •?- ^ ~\\ ^% ? ? I'-Ss'-g 11^ \i -^ ^ '^r u .iv^, ^J Jp^ -^ -TV '-X17 -A ll ,,.^^\^- II V, ....V .O * >C .--.^ (

Fig. 19. Newton.

. Wliiteside34 Hill.—Strictly speaking, this fort belonge th o t s Douglasdal e hea th f ePonfeig o dt watersheda s i ht i Burns a ,t bu , as it is also close to the head of Garf Water, and commands a fine vie thers otheo a wn d s edowi r an for , nit t connected wit e Douglath h s waters, we may conveniently notice.it here. It is 2J miles west-north- west of Newton fort, on an eminence 1139 feet above the sea, com- manding an extensive view. Nothing remains but a faint mound-circle abou fee0 5 diametertn i t hald an ,f .a dozen large r IrvinstonesM t gbu , FOKTS, CAMPS MOTED AN ,F LANAKKSHIRE O S . 821

describe s havina t si g bee a nston e fort from whic e stoneth h s were remove builo dt dmarca h dyke. Opposite the Garf the close valley of the Clyde opens up considerably, and becomes more distinctively a dale; hence the remaining forts to the north are divisible into hill and dale forts. The former again may be subdivided into two groups, one occupying the north-east skirts of Tinto, the other situated on Pettinain Hill. t John'sS . 35 Kirk (fig. 20).—A t John'S t s Kir a littlk e glen con- tainin Kire gth k Burn penetrates Tintohundrew fe a ; d yards abovs eit

t John'S Fig. 20 s. Kirk.

opening in the dale it is joined by a " hope " from the left, which runs parallel with the hillside for a mile, thus leaving a ridge 200 to 300 feet high betwee e south-easth f thie o dalen n th s O d itsel .d en t an f ridge, close to St John's Kirk House, there is, I believe, a much- decayed fort, although no notice is taken of it either by the O.M. r Irving M e e fordefence th y s Th movali b d . r an o , s consisn a - of t x VOL. xxiv. 322 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

inner mound, par f whico t h only remains, enclosin glevea le spacth t a e west end, and an outer one down the slopes to the north and east, but t traceabl no weste d th leve an n , eo l d clost witi an innee o et he th on r at the edge of a steep 'slope which descends to the hope on the south. There are three breaks in the outer mound at its lowest or east end; prob- nearese on abl e ravine originae th tyth th s ei l entrance n froni f d o t an , it, on the ravine side, an additional mound has been thrown up. The

Fig. 21. Kirkhope. mound aboue high4 s ar fee0 r 2 o t . douba t 3 wid e Therd b ty ean ema o whethed marke e O.M.o I s th t quarry s a i n bu r t ,t o di thi no s s a ,si t thinno k tha e de'bri th quarrta f so y could assum regulae eth r oval form of these mounds I thin. k ther alsis e o pretty distinct evidenca eof trenc outee fronn th hi f ro tmoun ease th tt da corner . 36. KirTchope (fig. 21).—Continuing along the ridge for only 150 yards we encounter another fort, easily accessible on two sides from the ridge t strengthenebu , -thn o dlony e b othe go e hopslopetw th r eo t s an s beeds circulai ha dalen t d I defendeconcentrio an .r tw y b d c ramparts f whico , outee hth merels i r y traceabl ploughee th n eo d slopes, Proc. Soc. Antii}. Soot. Plate

EAST ENTRANCE TO FALLBURN FORT POUTS, CAMPS, AND MOTES OF LANARKSHIRE. 323

while the inner, well preserved only on the south-east and south-west, owes its partial safety to having been enclosed in a plantation. Druid's Circle.—Three-quarters of a mile further along the ridge, and 1083 feet above the sea, a double circle, with this title, is drawn on the O.M. diametee th ,e oute th f r o r circl e inneth e f beinro feet0 d g27 an , one 210 feet. Mr Irving says of it:—"The top of a small protuberance is occupied by a Druidical or megalithic circle, consisting of two irregular rows of stones, of the ordinary rock of the country, rough as they came from the quarry, and of no great size." The O.S.A. describes it as " a circle, surrounded with large stones, erected on end, close to one e distancanother th 0 yard 1 t f A so e . ther s anothei e r wall, nearly resembling the former. In this place a large mound of earth is erected. Probabl Sheriffa s wa t ysi Court ths a , e adjacent far s mcallei d Sheriff Flats." I could find no trace of these circles unless some holes in the groun indicaty dma e where e stonesomth f eo s stood. 37. Fallburn (fig. 22 and Plate XIII.).—The ridge of which we have been speaking merges in a shoulder of Tinto a short distance beyond the "Druid's Circle," about half a mile west of which lies the next fort, 0 fee 90 e tnorther abovth fooe f e seath o t eth t na , declivit f Tintoyo , where it suddenly eases into a gentle slope leading to the marshy plain below. The interior of the fort is dry and grassy, .but outside, the groun s i rougd h with heathe d marshy an e rfor s Th i mcircular. , e enceintth d an e consisto o ramparttrenchestw tw f d o s an s. This e besth tf o preserve e on s i d fort n Lanarkshirei s e rampartth d an ,s are unusually higd massivean h , particularl e southth o t ,y where e positioth s i weakene n e e descenslopth th y f eb o dt upo. it n e rampartTh e als ar o sentranceso tw a enlarge little d th e an t , a d square foundation, enclosing a space big enough to hold a few men e inneth f ro p rampare nortto th e e n hweso th th t sid f tn o o e entrance, together with three traverses in the trench below, forming three little pits, may have been intended to strengthen the defence of the entrance; at least, it is difficult to assign any other use for them (see e easals th . s 344)oute op tit d entranct en rA . o t f ei opens a p su give a better flanking defence (Plate XIII. In this view the rise of e Tintbackgrounth n oi omitted)s di interioe th n I r. ther uniqua s ei e 324 PROCEEDINGS 'THF "O E 'SOCIETY, FEBRUAR , 1890Y10 . irregular circular grassy mound ,inne e closth ro et rampart s onli t yI . 6 or 7 feet wide, and about a foot high, and has no opening. Although earthee thith sf o for ns i ttype gooa , d many stones sho n breakwi n i s the ramparts, and some large ones lie upon them. The O.S.A. mentions another circular camp in Covington Parish, at Castledykes," which has two subterraneous passages leading.from the east gate." I cannot find this name in the O.M.

Fig. 22. Fallbura. e nexTh t grou as pi remarkabl extensivn a n o e eeon eminence , crowned •by a semicircular ridge, which on the O.M. has no general name, but a lower summi f whico t h is.marked Swaite e Olth sd Hilln StatisI . - tical Account .(1794), however, s .ii sait d thae westerth t d an n FORTS, CAMPS, AND MOPES OP LANARKSHIRE. 325

highest part is called Westraw Hill, and the bend eastward and south- ward, Pettinain whole HillTh .e Uppee eminencth lase f th ro t s i e War de lef Hillth t e Clyde n ban th so d frof k o an ,m0 a heigh 50 f o t feet command a fins e e westwarvieth f wo f o de rive d benth an f r o d grea a countrw lo t e exten yth e hil f beyonTh lo t (Map . it d , ) fig25 . e likeneb a hugy o t dema ravelin f whico , e lonhth g gentle slopef o s upwards of a mile towards the river form the glacis ; the semicircular hillmilee o cresth tw , f s to long ramparte ,th elevaten a ; d marshy plateau, sheltere d haldan f enclose e crest e th interiore steeth y th , db pd an ; descent of 200 feet from the plateau to the Mill Burn, which separates it from Carmichael Hill, the rear or gorge. On this singularly-formed hill are four forts, three on the crest, one on the plateau.

of Figs. 23 and 24. Cairagry£e. 38. Gairngryfe (figs. 23 and 24).—On the summit of the Westraw division of the hill (1141), which is also its highest part (A on Map). The top of the hill is flat, so that no immediate advantage is gained from the elevated site. There can be hardly any doubt that this was stona y masonrye dr for f o t , consistin innen a d oute f an gro r walle th , 6 32 PROCEEDING SOCIETYE TH F O S , FEBRUAR , 1890Y10 .

former approximately circular, enclosin a gspac fee0 8 n i te o t fro 0 m7 diameter lattee th , r encircling thdistanca e t a inne o e f t froeo on r 5 m1 fee0 r three-fourth2 fo t e way t th divergin f bu , o s g nearin n froo t mi g the entrance, so as to leave a fore-court about 40 feet wide there. I have give e probablnth e original measurements e wallth e s ar sa t bu , almost wholly cast down e debrith , s encroache e interspaceth n so d san somewhat narrows them. In the outer zone not a trace of building is portioa e inne e th on f rwhan n i o seene t b tgrassa o bu ,seemt e b -o st grown wall about 12 feet thick can be made out amidst the mass of ruin on the south-west side; and at the entrance, where the wall apparently was thicker, there is a wall-surface about 15 feet long and 3 high, whic perhaps hi e originath s l eas e tentrance s th sidit f t eo A . inner end are the remains of a doorway, only 2 feet wide, well built of large stone 3 feesn lengthi t , which gives accesa circula o t s r enclosur fee6 e3 diametern i t . Thi sharsa t doorwaa p t anglse s yi o et passage th e throug partialle walle hth Th . y ruinou senclosuree walth f lo , 2 or 3 feet wide and in places 6 feet high, is well built of small stones of uniform size. This is probably a modern " Kee " or sheep-pen, although I do not know the object of the doorway being oblique with re- gard to the entrance. The structure of this part is shown in the enlarged plan (fig. 24). Something like foundatioeth wese th e tf no th sid f eo entrance can also be traced, indicating that the passage may have been 3 feet wide at the inner end, expanding to 6 feet at the outer end, but excavation is required to settle this point. It continues 6 feet in width through the fore-court, where the remains of its side walls can be faintly traced alsd an o, throug e outehth r e remainwallTh . f fouso r traverses cros interspace th s e betwee fore e wall t irregulath nth a t f so r intervals. Walking round the crest of the hill, after a descent of about 150 feet, we rise about half that firse heighth te o summith t e tpar f th o tf o t crest pertainin Pettinaio t g n Hill, about three quarter mila f eo s from Cairngryfe. e severaUpoar t i n l moundMap)n o B ( s, whics i t i h difficult to make anything of. They are of trifling width, and one of them seems to be part of an enclosure, 36 feet in diameter. About half a mile further on is the next summit (1049), crowned with a good-sized grass-covered cairn a littl, e southward from whic e remainn ar ho C ( s FOKTS, CAMPS MOTED AN F LANARKSHIRE, O S 7 82 .

Map), which might almost be taken for those of a small circular fort, 120 feet in diameter over all, but they probably mark the site of the " Hero's Cairn," described by Mr Irving as a large tumulus, in which, " when it was taken down a larg, e urn, surrounde y fivb d e smaller ones,s diswa - covered whole th , e being include stona n di e cist." Thi probabls si y also the "Cairn (site of)" marked on the 25-inch O.M. A short distance south-west of "Hero's Cairn," looking down from the e cresedgth f te o upo plateaue nth as ,i circula Map)n ro forfee0 D ( t,17 t diameten i r inside stone Th y. circumvallation, scarcely recognisabln ei some parts, is quite substantial in others, although nowhere more than 3 feet e grounhighTh . d outsid s marshyi e t withibu , s dryni , hard, rough, and completely covered with heather. This fort is noticed by s probablO.Me i th t I n . o y a detacheMt rno Irvings i dt workbu , , appertaining to the great fort of Knowhead on the plateau. . Knowhead 39 n Map).—Wo F ( e fourt th comd w an o no ht ee most important fort of the group, situated, not on the crest, but in e elevateth d slightlan d y hollowed plateau, whic s hali h f encircled by the crest. This plateau is, for the most part, a half-drained marsh which surround e fortth s , except wher e ramparth e t e cometh o t s edge of the plateau, and descends a little way down the hollow, through which run a slittl e rill, passing Knowhead Farm o joie t Mil, th n l Burn. This rill takes its origin by two heads in marshy ground within the fort, the interior of which is higher elsewhere and tolerably dry. In the unusually precise and accurate description in the O.S.A., it is stated that " the camp includes several springs of excellent water," in connection no doubt with this rill. The single enclosing mound is an irregular oval, which I made to be about 1200 paces round, but as surface th vers i e y roug pacee hth s were probably shortt . no Stillo d I , think it can be much under 1000 yards, although Mr Irving's diameters fee0 t69 maky b t somewha0 ei o90 f O.S.Ae t lessTh . . give e areth s a about 6 acres. The mound, although it rarely exceeds 2, and never 3 , ofee 4 r heightn tolerabla i t f o s i , y uniform breadth, averagin feet5 g3 , and although thinly covered with moss and heather, consists of little else than stones, as shown in many breaks and several accidental sections. e mosth r te stonefo par Th e tar s smallish t everbu , y herd theran e e 328 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890. large blocks t occurseemi d improbablo san n , e hypothesis thahave w t e her remaine eth massiva f so e wall, whic bees hha n cast dow spread nan d out for the purpose of plundering the large stones. Indeed, how else can we account for a great zone of stones, 35 feet wide, and only 2 or 3 feet high ? Probably the. remains were in better preservation at the

mile furt

Fig. 25. Knowhead. .end of last century, as the O.S.A. says "the wall seems to have been very thick and high.;" while, according to the N.S.A., in 1845, "the FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 329

traces of the camp, so well described in the O.S.A., are gradually disappearing." On the north-west there is a singular re-entering curve in the oval, opposite to which, and about 35 yards outside the line of the oval, is a detached circular work (G on Map), 150 feet in diameter, with a single vallum, 15 feet wide and 2 to 4 in height. This detached work is planted entirely in marsh, which even occupies the inside of it. There soms i e appearanc causewaa f eo y connectin t gwiti e maihth n word kan a fragment of mound halfway between. This outwork is also accurately describe O.S.A.e th n di , wit additioe th h n thas evidentlha t "t i y been connected with the large camp by a passage through the moss." Mr Irving describe single sth e entrenchmen Knowheaf o t d For beins ta g comparativela f "o y slight character, t thi "contrars bu si evidence th o yt e Statisticae oth f t seemi l e Accountsm s thao t e remain t d th t no an , e sar only considerable in themselves but indicate a work originally of unusual magnitude. He goes on to say: " From the dimensions, position, and general character of this fortification, there seems no room for doubt that it is the intermediate camp, which, from the well-known length of e dailth y marc a Eoma f o h n legion e westerth , n colum f Agricola'o n s army must have formed between that at Little Clyde, and their junction wit othee hth r divisio t Carstaira n r moro s e probably Cleghorn.t Bu " it is only about half the size of the Cleghorn Camp, and in structure and form there is nothing in common between it and what are commonly accepte s Romaa d n " camps"; besides a mer r efo , temporary halting- place, why should the Romans have constructed a formidable stone fort n thii s reasoo secluden r separatine fo nse dn spoca gI t ?thi s work froe clas th mf nativo s e fortse d Romansascribinth an , o t t t bu i g, undoubtedly in some respects it is peculiar and worthy of further investigation O.Me Th . .merel y marks " Camp (site of )t thi"a s spot. 40. Chesters (fig. 26 and E on Map).—At the extreme south-east creste 101d th an f , o 3 d feeen t abov e fourte seath eth s i ,h fort, com- mandin a fing e vie p Clydesdalwu d eastwardsan e , facing Tinto t o e south th d lookinan , g down upon Thankerton Moor fro ma heigh t of abov e slope0 feetTh 30 e .s e lon ar n frothreo g t mi e sidest bu , s i easil t i y accessible alon e ridgth g e froe northe O.Mth m Th . . 330 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

make t circularsi measurementy m ; slightle b s o t brough yt ovalou t i t, the internal diameters being 300 and 277 feet. It is enclosed by a double rampart with intervening trench. Although apparently of the earthen type, stone has also been used in its construction. Not a few

Fig. Chesters26 . , Pettinain Hill. large stones, som mucs a e s 4ha fee n lengthi t e abou li , e rampartsth t , but besides these loose stones, the scarp and counterscarp of the trench appea n severai r l places, where breake turth f n i soccur o havt , e been reveted with stone at the foot, the ramparts thus consisting of an upper FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 331

part of mixed earth, and stones, and a lower part faced and made steeper with stone. Such, at least, seems a probable interpretation, but excava- tio necessars ni o determinyt e whether this construction, uniquer fa o s , moderne t b aftea sI l know no al ry . entrancesma ,o Thertw e ear , which pass throug highea t ha r level tha separatee e trenchnar th d an , d fro. t mi by a slight mound. Mr Trying gives the dimensions of this fort accu- rately, but most inexplicably describes it as being defended by a single rampart equalle ,ar althougo ytw distincbote e ar hth h d uncommonlan t y well preserved alse H o. mentions callethas i t ti Forrest' n di e "p Th sma Chester's outpost." Other Remains.—Besides the "Hero's Cairn" and the mound-enclosures of doubtful nature already described t seemi , s probable that other pre- historic remain exis thismay on t s interesting hillfirsAt t. is sigh one t continually deceived by the appearance, on the inner slope of the crest and on the harder ground of the plateau, of partial stone circles or other artificial forms, which on close investigation turn out to be accidental, remarkable th du o whicn et i y hewa boulder sparsele sar y scattered over hille t som artificiath e bu b ; thesf ey o ema l afte remarkable ron alld ,an e grou stonef po s seem Map) n ssituates o i undoubtedl t H I ( . o ds e b o yt on a dry part of the plateau, near the rise to Cairngryfe. The singular triangular arrangemen e inneth f o rt collectio f stoneso n fusiford an , m figure of the outer set, is shown within the rectangle in the left-hand lower corner of the map. A little nearer the Cairngryfe rise there is a low hillock (I on Map) which has an artificially smooth look. Altogether, Pettinain e unusuallHilb o lt seemee ym rico t dn possibilitie hi d an s likely to reward systematic investigation. 41. West-side.—Passing to the group in the dale proper, this is the first fort met with coming from the south. The site is on Westside Farm, 2 miles east of Tinto Top, in a field about 70 feet above the Clyde, whic hals hi milfa e distant single Th e. rampart, although much flattened frequeny b t ploughing s stili , l quite distinct enclosed an , n ovasa l space of about 210 by 140 feet from crest to crest. It is cut off from the level field to the north by a trench 60 feet wide and 5 or 6 deep, measuring from the crest of the rampart. This trench probably merged originally in natural hollows, 15 to 20 feet deep, on the south and west. 2 33 PBOCEEDING E SOCIETYTH F O S , FEBKUAK , 1890.10 Y -

42. CastleMll, Symington.—A quarter of a mile south-west of the village gentla n o , e eminence, which, however, rises conspicuously above the minor inequalities of the dale, to a height of 150 feet above the Clyde, hal easte t milconsistI a fth .o et singla f so e mound, enclosing, as I made out, an oval space of 190 by 150 feet, but which the O.M. and Mr Irving agree in making circular. The mound is about 20 feet 5 wide s rareli d t an ,i e nort yn heightd th i exceed t en h4 A o t . 3 s feet abov inferiore eth onld fooan y1 t abov interioe outside.e th eth n I r ther some ear e curved mound f indefinitso e shape. Irvinr M g describe lef e hole camsa th t Clyde e th bann mth o n p i f ko , opposite Wolfe Clyde muct "o s bu , h destroye plouge th y dhb s thai t i t impossible eithex s fi dimensions. it s t foriit no rr t no ' mo s i t I " the O.M.

. BrownfordFig27 . .

. Brownford,43 Covington (fig. 27), overlook e formerlsth y important ford ove e Clydth r o Libertot e n fro ma a distanc0 yard d 40 an sf o e height of about 100 feet. It comes near the edge of a bank, which descend river-haughe th o st fore peculiarms Th i . , resemblin lettee gth r O, and it consists of a double rampart of low relief and no great width, wit distinco hn t intervenin g0 leavetrench throae e th widTh sa f o .t e • FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 333

entrance, contracted, however fee9 ,o t t fro h 0 yrectangulama 5 r worf ko trifling strength, which fills the space on the south side. On the north side spacth e e betwee e rampartnth s expandsgivo t s ea moro s , e room for the defence of the entrance. Little is left of this work but the ground plan. Cart tracks tha twels shava s plough a , sideo l it etw f so - ing, may also have modified the original plan. But as it stands there is much that is unique about it. There seems to have been another fort near this, halfway down the slope, which Mr Irving describes as having been so often ploughed " that s forscarceln it mca e tracedyb t appearsi ; , however havo t , e bee lopna - sided oval of 143 and 120 feet diameter." It is not in the O.M., and afte carefua r l searc hI coul d fin . traco it dn f eo

0. Forts in the hilly country north of the Pass by Biggar Water from Clydesdale to Tweeddale. This district comprise sa mas f comparativelo s hillsw lo yf whico , h the northern part may be considered as the final subsidence of the Pent- land chain toward south-weste sth south-ease Th . t corner, boundey db the Broughton Burn to the east, is in Peeblesshire, and its forts have already been describe forte pape y th f tha m o s n n o rdi t county. This corne welis r l defined fro mucmthe h large rdistric parthe the of t by t frontier of the two counties, which runs from Biggar north-east to Dol- phinton alon granga f heightseo e foo f whiche th Lanarko tt th a , n o , - shir eease sideth ys i ,pas s which contain e roath s d between theso tw e places. 44. Bizzylerry,~$o. 1.—Mr Irving speak f p "o sato campe th n "o of Bizzyberry Hill, consisting of " a single entrenchment of an oval form, 5 fee diametern 13 i t y b 5 ,18 with some faind ill-definean t d tracef o s outworks." I had some difficulty in satisfying myself that the mound roun conspicuoue dth f thio sp hillsto , 1190 40 2 d feean ta abovse e eth feet above the low ground, was of a defensive character; but granting that it represents a fort, I should be inclined to add to it a much larger space west e haunce I thinoth hile n th s o th a ,t l f tracekmoune a ho b n ddca roun norts d it wes d han t sides. Ther foro n t s ei marke dO.Me herth n .eo 334 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

45. Bizeyberry, No. 2 (fig. 28).—950 feet above the sea, 700 yards from the last, at the north-east end of the ridge, just before it drops 150 valleye e wholth fee o Th t thif t .eo s much-dilapidated rougo s for s i th with half-hidden stones and coarse heather and blaeberry that it is very difficult to interpret it, but it seems to have been of unusual strength.

. FigBizzyberry28 . . ovale th f , o approachin s i t I circulare gth , type d consistan , f severao s l concentric " rings," but it is not easy to trace them out, or even to determine their precise number e O.MTh . gives two, with faint indica- tions of another, but I think that there are certainly more, and that the followin prette gb facty y sma clearl y made out. The limited site rises slightly from the gentle descent upon it of the FOKTS, CAMPS, AND MOTES OF LANARKSHIRE. 335

Bizzyberry ridge, and if we regard the " ring " which girdles the top as rampare e th interioth f o tr citadel o r , the nothere o th ther tw n e so ear gentle descen ridgee th o t ,t there bein rooo n gmorer mfo ; e b thes n eca traced round the south flank of the hill and on the descent to the plain e steee easth th e lospt n t ar o a endtnort t ,bu h t flankbesideBu . s e eastheseth t n slopo , e ther concentria s ei c rampart between thed man the citadel, and another outside, further down the slope. Neither of these, however, can be traced more than a quarter or a third of the way round. Thus there are three ramparts to the west and five to the east. But these are perhaps not all. I think that there is evidence of another substantial " ring " on the top, within the citadel, and within that again the remains of a circular " house." whole ringe th Th f seo fad lose e ar awat d towardyan northe th se th , inner ones upo e summitnth outee th , r upo steee nth p slope trencA . h separate outere sth rampart ramparte weste fro ridge th Th m th .o t e s are only a few feet high and are very stony under foot. 46. Castlehill, CandybanJc (fig. 29).—A little more than a mile north-east of the last, 993 feet above the sea, 200 above the Candy Burn, which 0 runyard70 s s eastward d oppositan , e almosth e t obliterated fort, called Candyburn " Castle," on the other side of the stream. The ground slopes gently from the site except to the south and regulafora weste s i t Th . r oval appeard ,an havo st e belonge whao dt I t have ventured to call the terraced type, having two terraces on the south and west, wher e descene nortth eth d steeps hn an i to t onle bu , on y east, wher e slopth e s triflingi e slighA . t parapet runs roune th d top of the inner scarp, but a few remaining stones probably indicate that a formidable wall had been the original defence. There is no vestige of a parapet at the edge of the terraces, which are smooth and level, and are from 20 to 30 feet wide. There are two entrances—one to the north-west, the other to the south-east. On the east side of the former collectioa f stoneno s perhaps indicates some additional defencen o d an , s wesit t side commence stona s y lowee slopth o ert terrace, which ends near the south-east entrance. This entrance is prolonged into the interior by two foundations, having on the east the remains of one circular en- closure, and on the west, of three, touching each other, and one of them 336 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890. fee5 diameter4 n i t . Outside e south-westh n o , t slope closd an , e under the fort, there is another very regular circular mound about a foot high, enclosin ga spac e abou fee 0 n diameter i 10 tinteriore t th n I ., closo et the parapet on the south, seven large stones are arranged in the formal manner shown in the figure.

% '», S^^m^^^-^ '%>7^®w^//%////// l x -'^',(,Mri!l^.-^'^^

3OT Jr^/3 30 ; . . ^^- f

. CastlehillFig29 . , Candy bank.

AsMiUl, Meadowhead, DolpMnton.—Proceeding 4 miles further to the north-east, we next come upon this site, on a very gentle slight eminence groundw lo n i , , abou 0 fee80 t abov e seath e. Here about the half of an oval fort is marked on the O.M., but I could find no trace , neitheoit f thers ha r e bee withiy nan e lon th ng recollectiod Or r M f no Mackenzie proprietore th , . FORTS, CAMPS MOTED AN F LANARKSHIRE, O S 7 33 .

e Cairn.—littlth ef o summit e e Onwoodeon th f o s d ridge which form e boundarsth countye th abouf d yo an , a milt e laste soutth ,f ho ther cairna s ei , covered with gras d heathersan , abou fee8 t t higd han 5diameter0in . Eecentl ynumbea largof r e stones, forming a par of t circle at the foot, have been uncovered. The largest is 5 feet long, and another is not much less, but the rest are considerably smaller. 47. Chesters (Chesterlees on the O.M.).—900 feet above the sea; a quarter of a mile west-south-west of Dolphinton House ; almost quarried away; two fragments of a mound composed of earth and small stones remain apparentls ha ; y bee na lon g grea o ovaln s site t th ha naturae; l advantages. 48. Kier.—On a formidable position, half a mile north of Dolphinton House fee7 ,t99 abov seae eth , with a goo dneighbourine vieth f wo g valleys. It is unfortunately impossible now to distinguish mounds which maybe old froonesw mne , quarryinn i cas p u t g operations e 2 5-incth n O h .O.M . e " Camth p (sit confine"s ) i eof spacea o vere dt feet0 th , y4 10 n y o , 0b top of the hill, but my impression is that there has been a strong fort here, with two or perhaps three " rings." A little way down the slope to soute th h ther a smal s ei l platea r subordinato u e eminence, whichs a , Mr Mackenzie pointed out to me, has much the appearance of having been artificially levelled on the top and scarped on the sides. The descriptio e O.S.A th cama n f n i thio n .pi s paris s morhi e applicablo et Kier than to the other Dolphinton forts, although the size seems too great for the site. It is said to be on the top of a hill near the middle of e parishth o enclost , e four acres o havt , e been apparently strongly fortified with a wall of earth and a broad deep trench cut out of the rock, also with some strong otttwork, particularl cornere th n yo s where the hill was most easy of access. 49. Cocklaw.—Eeturning to Biggar, the next fort lies three miles dun Cocklawo e , nortit f o h, 1010 feet abov e seath e . This hill, although from the gentleness of its slopes it is not conspicuous, is neverthe- less higher tha neighbourine nth g eminences commandd an , extensivn sa e view all round. Mr Irving describes this fort as " defended by a single circular rampart," VOL. XXIV. Y 8 33 PROCEEDING SOCIETYE TH F O S , FEBRUAR , 1890Y10 .

and the O.M, represents three concentric rings of some kind close together remainine Th . g mound vere sar y insignificant e t seem bu ,m o t to consist of an inner, nearly circular one, with fragments of a con- centric outer one, 30 feet in front, to the south, but diverging from the inner one to the north, so as to enclose a crescentic space, 90 feet in greatest widthe interioTh .pretts i r y level t irregulabu , r with shape- less mounds slightla ; y hollowed space, however s kepi , t clea l rounal r d betwee rampartne e centra th thesth f d i els an a ,par beed ha tn masa f so habitations, but a clear zone had been left next the rampart to allow 0 27 s defencerooy it b r m 0 fo I mad 30 ovan .e interio a f eth o e l b o t r feet, but the O.M. reduces it to a circle 225 feet in diameter, while Mr Irving still further reduces it to 200. The O.S.A. also makes it a fee0 circl20 t f diametereo , enclose concentrio tw y db c circle f eartheso n dykes or mounds, 15 feet apart; to this description the N.S.A. adds two ditches, one for each mound. 50. Whitecastle.—A little above two miles west-south-west of the lastfee1 97 ,t abov 0 Clydee abov 30 th e dal e seaf eth d o eth , an ,upo n which it looks down, as it is near the western margin of the congeries of hills nort Biggarf ho . Irving says: "t appearI havo t s e bee ovaln na , 181 by 128 feet; although having been included in a plantation the line of its rampart cannot easily be traced." The wood is now thin, and the mound, although much destroyed, and altogether lost in parts, can be made out, I think, to have formed a circle 180 feet in diameter from crest to crest. There are also traces of a second " ring," 30 feet to the outside. This fort is not in the O.M. 51 Graigie.—A mile and a half west of the last, about 700 feet abov0 abov10 e Clyde seath e d th ean , , which e flowyard0 th 70 o st s , therwesit f o et stoo da fin e fort, thus describe Irving:—y db s wa t "I perfectly circular, and was defended by two entrenchments of remarkable height, the diameter of the interior area being 248 feet. It had only one gate, which face e eastth d . Since these pages were written (circa, 1864) every vestige of this fortification has been swept away by agricultural improvements. O.Me Th ." agrees wit descriptions hhi t bu , reduce diametee e O.S.A feetsth 0 Th 23 . .o t rdescribe havins a t si g "a double wal f earto l h wit a hdee p ditch between o stoneworkn d an , , FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 339

except in the middle, which seems to have been paved with freestone." 52. Newligging (fig. 30).—One mile east of the village of that name, and two north-north-west of Cocklaw, No. 48. It is on the north side of the South Medwyn Water, 900 yards from it, 250 feet above it, 950 broae th f do hil p n whico lto e t i t hquitth abov no t a e sead eth an , e nortextensivn viea th o westd stands n t o wh an t s bu ,tha ha o e s ,t i t one soute d foreast s Th an irregularli hm. y circula r two-thirdfo r s of the enceinte, but on the north there is a slightly re-entering curve, and to the east a straight line 140 feet long. There is nothing in the nature groune oth f t presenda accouno t t r thesfo t e peculiarities enceinte Th . e

Fig. 30. Newbigging.

consist a singl f o s e rampart, apparentl f eart o yd stones an h , much tampered largwitw fe eh a jstone e scatterear s e easd th about t n O . side ther soms ei e appearanc slightla f eo y raised platform fee8 1 , t wide, in rear of the rampart. On the west side, where the interior of the fort 340 PROCEEDINGS OF THE SOCIETY, FEBRUARY .10, 1890.

is helow the level of the exterior, there is a banquette about 9 feet wide. Possibl e resuly a modern-lookinth thi f s o i ts g slight excavatios it n ni rear, but I do not think it is entirely due to that. Within this shallow excavation, and the slightly raised platform on the east side, with which s continuousi t i e interioth , r rises int a omuc h flattened domew fe a , feet in height, whether natural or artificial it is difficult to say. Irving speaks of " an outwork on the east of a slighter character, extending to distanca f 15eo 0groune th feet. l I coulal d "t dbu , fintraco it n df o e close up to the enceinte has been ploughed. Diller Law ; Blackhill; Drafan.—Three forts in the parish of Lesma- hagow, quite detached from those hitherto considered. Irving says of them : " On the top of Diller Hill (1017) there is a detached camp of no great size, and another at Blackhill (952), with its ditch much obliterated by a thriving plantation thirA . d e faroccur th f Draffam o n o s n (624). It is circular, with a diameter of 70 or 80 yards." They are ignored by the O.M I hav visitet . e no locality e dth .

General RemarJcs Curvilinearthe on Forts.

. Form1 Classification.d an —A glanc plane sectionth d t esa an thesf so e works suffices to show that there is a considerable variety in their design, and an examination of the remains themselves reveals some variety in the materials employe thein di r construction reduco T . e these varieties ordero t , however difficults i , e hand, on becauseremaine e th , th e n ar so , in general too imperfect to afford the means of coming to safe conclu- e otherth r knowledg n ou o , sions d an f , eve o e bes th n t preserved examples does not go beneath the surface, and is therefore defective if not often misleading. generae Th reducee ltypes—th o b form y tw o ma sdt e circulae th d an r oval. Not that any one fort would be found, if accurately measured, to e strictlb y circularf thirty-ono t bu , e whos e accuratel b figur n ca e y made out, eleven approximate .so closely to the circular form as to be easily distinguishable fro e ovamth l y variationtypeAn . s from these type obviousle sar exigencieo t e ydu f siteso . Whereve site rs tolerth ei - ably levegentla r uniformn o fieldla o esmootha r n i o .slope,n s o a , r o , FOKTS, CAMPS, AND MOTES OF LANARKSHIRE. 341

regular, dome-shaped eminence fore pretta th ,ms i y regular ova circler o l . But where the site is irregular, the form may be slightly distorted in order to take full advantage of the ground, as at Culter Park (fig. 15) ; or may he lop-sided, as at Shielburn (fig. 8), where one side rests on the evee edgb ravin a n y f approximateleo ma e r ;o y rectangular with rounded ends, as at Bodsbeny (fig. 2). In a few instances, such as Newbigging (fig. 30), variations t easilno , y explained now havy ma e, bee origie ndu - e necessitnallth o t y f avoidino y g boggy ground before the dayf o s drainage. variatione Th s froe normamth l circula d ovaan rl types being thus e e natursite th th o usefu o n ,f t o ee l du classificatio e foundeb n ca nd on them, any more than upon the number and arrangement of the lines of defence, which are also much regulated by the site and the varying natural strengty takin b grounde t th f ghBu o int. o consideratioe th n ground plans, profiles, and materials of construction, so far as the latter are revealed in chance breaks or cuttings, the following classification may be adopted, which, if not truly scientific, may at least aid the mind in forming a clearer idea of the nature of these ancient forti- fications. a. Stone Forts.—Cairngryfe (p. 325) is the best example of a fort constructed entirely of stone. Although almost completely overthrown, the traces of building are quite sufficient to prove that the original structure was a strong little work of dry masonry. West Crawford . 299(p anothes i ) r example, with slight trace f buildinso g left. White- sid . 320)e(p , fro s histormit sized yan probabls ,i y another t nothinbu , g remains save a few stones. The great fort of Knowhead on Pettinain Hill (p. 327), with its two outlying fortlets, are also apparently entirely of stone, but fas surface-appearancs ,a a r e goes, sho traco wn f buildingeo . Here we may notice that stone was no doubt used for parapets, &c., to muca h larger extent appearsthaw no n t onlno , y becauso find a d e ew few large stones left herd ther ean n forti e s whic t firsa h t sight seem entirely of earth, but because it is. natural to suppose that, if stones were removed from these ancient structure r moderfo s n usesn i , many cases not a trace of them would be left. b. Forts defended l>y a single Mound.—Newbigging (p. 339), North 342 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

Coldchape . 304) (p lfiv d ean , f otherthio t i muse s t ar kinde s b t Bu . remembered that originall t havyno ey the bee ma ysimpleo s n , that other defence havy ma es disappeare e coursth f centuriesn o i de ; also thamateriale th t f constructioso uncertaine nar t evidentl,bu y vary, some being much more stony than others. c. Forts defended Moundsy b d Trenches.—an Hartree . 316(p d )an Shielhurn (p. 305) are examples with a single mound and trench; Kirkhope (p. 322), Fallburn (p. 323), and Chesters (p. 330) have two trencheso moundtw d san . !Nb e absenc tsaide th mucb n n i , f excavationsheca o e composith f o , - moundsthe tio of n . Judging from chance break themsin , theymay eithee b f earthro , with merel ynaturaa l admixtur f stonese o f eart o r ho , and stones purposely mixed, or in some cases possibly entirely of stone, over which a covering of sod has gradually been formed. Probably few of them retain their original height, form r steepnesso , t i seem t bu ,s certain that those of earth, or mixed earth and stones, were generally round-crested; in a'few instances, however, they were sharp-crested, and in others they were flat-topped, forming broad platforms, on which the defenders could stand, protecte parapea y db t bot fron n hreari d t A an t . Campswater (fig. 5) there are examples of all three kinds. I saw nothing e moundth f nature o o r th ramfbanquettea o y s f ean o e rea -f th o rn i parts, excep t Newbiggina t . 339)g(p , where , e spurihoweverb y -ma t i , ous, resulting partly fro nature e groundmth th f eo , partly from modern changes I thin t bu k; ther fais ei r evidence, both there t Norta , h Cold- chapelt Cocklawa d kina an , f f roadwayo do , n somi , e places slightly raised, in others slightly hollowed, situated in rear of the inner rampart for eas f lattecommunicationo o e tw r e caseth n si probabld an , y also t leave a clear space between the rampart and buildings in the interior. trenchese th Ao st thein i , r present state the vere yar y shallow doubI . t instancy an n ii f e the more yar feee3 thar to belon2 originae wth l level e groundoth f , althoug f courso h e the e severalar y feet deepere th f i , heighe ramparth f o te added b t . There may o doubtn , , have beea n considerable filling up in the course of ages, but as the trenches are rarely mor yarde3 thar o sn 2 wide t wouli , d seem that they coult no d have afforded nearly sufficient materia e moundth r fo l s associated with FOKTS, CAMPS, AND MOTES OF IANABKSHIBE. 343

them, and that the material must have been mostly brought from else- where. Where concentric ramparts occup ya slope , ther e oftear eo n n true trenches, i.e., hollowe th s betwee rampart e t belonth no e we th ar s level of the ground, but are merely produced by the mounds heaped up in front and rear. At Chesters (p. 330) there is some evidence of both scarp and counterscarp having been made steeper by a stone revetement, an t Arbora d . 304(p y ) havthey yema been faced with stone without being steepened t thi bu littls ,si e more than conjecture. d. Terraced Forts.—In e detaileth d description, frequent instances are given of the use of terraces. They commonly occur on steep slopes, wher t woulei difficule db impossiblr o t construco et usuae th t l ramparts and trenches n sucI . h case e terracesth fee w e onltsfe ar ywidea d an , must have been defended eithe slighy b r t wall r palisadeso s . Possibly they were only foundations for walls, intended, not for actual defence, t merelbu obstruco t y e ascentth t n otheI . r instances, howevere th , terraces are on gentle slopes and are 20 feet or upwards in width. In a single instance (Candybank, fig. 29) terracing constitutes the main principle of defence. e. Complex Forms.— In a numbee forte stonth th sf o re walle th , mound, the trench, and the terrace are found in combination. Thus t Arbora ye hav w (fign innea e) 6 . r "ring f stoneo " , wito tw h nearly concentric outer " rings," each consistin mouna f go trencd dan h ; t Blackbala n linne a (fig ) r 17 . rin f stonego , wit n outeha r non-con- centric ring, consisting of a mound furnished with a trench on two faces only t Cowcastla ; e a ver(fig ) y 12 . complex combination f moundso , , terraces trenchesd an ,t Campswate a d an ; r f ston(figo ) e.5 wall, mound, trench, and terrace. These are the best preserved and most pronounced examples t seemi d an ,s needles quoto st e other distinco s st whicno te har and reliable,

2. Minor Details.—a. Entrances.—Sometimes there is only one entrance, as in the small forts at Cairngryfe (fig. 24), Shielburn (fig. 8), Langloch (fig. 13), in the larger N. Coldchapel, and perhaps even in complee th x Cowcastle (fig. 12)n otheI . r case entranceso s thertw e ear , which either pass straight throug workse t hth Fallbura s a , n (fig. 22), 4 34 PROCEEDING SOCIETYE TH F O S , FEBRUARY 10, 1890.

Newton (fig. 19) d Chester,an s a (figzigza .n i g26)r o manner, s a , t Arbora y (fig . .6) Her e also there appea o t havr e been three entrance s outeo througe centratw th r e lineso th t ht l onle bu , on y citadel. Occasionally one side of the entrance is brought considerably in advance of the other. This is most marked at Arbory, but exists to a less degre t Campswatea e d Fallburan r ) (fig5 .n (fig. 22), where th e trenche side opposite eon ar n mounde so othere th th n .so Thero n s ei evidenc f thaeo t multiplicatio f defenceo n s abou e entranceth t s whics hi so marked a feature in the best-preserved Welsh forts, and to a lesser degre Peeblesshir e somn eth i f eo e ones. Almos e onlth t y evidencf o e a special strengthening differena of but t, kind Fallburnat is , , wherethe mound f the o e entrance e msth t wideon appeara d t an n,ou havo st d eha a small square structure on the top, besides three traverses at intervals yardsw fe oa ,f forming thre e trenceth pit n i hs immediatel n frontyi . I was inclined at first to regard these as doubtfully ancient, having never seen another instance precisela t bu , y similar arrangemen f fouo t r pitn so on entrance e th sid forf a e o o tt e calle dCastles,e "e Th far th f mo n "o Latch, Newton Hall, East Lothian describes i , figured d an Willia y b d m Waring Hay, F.S.A. (Arch. Scot., iii. 302). b. Traverses.—The general absenc f traverseeo checo st e ruskth f ho an enem l rounyal e enceintedth , betwee ramparte nth outee th f rso lines of defence, is remarkable. There are unmistakable remains of them, indeed, in the stone forts of Cairngryfe (fig. 24) and W. Crawford (fig. 4), and faint traces in the part-stone fort of Blackhill (fig. 17); but not, I think, in any of the others. Perhaps wooden defences of this kind were used. c. Interior Structures.—The remains of " hut-circles " or other circular structures are not so numerous or well marked as in Peeblesshire. At the best they are but faint surface-markings, easily overlooked, and perhaps only visible in a favourable light. I noticed them in Mid Craw- ford, Castlehill (Symington), Castlehill (Candybank), Campswatern I . other instances t Cocklawa s a , , Arbory, Blackhill, North Coldchapel, Devonshaw, Bizzyberr ther, 2 generaa . s ei yNo l stony roughness, percep- tible under the sod, or a visible, confused mass of low mounds, suggestive Soc. An% Scot. Vol.XXIV Plate XT

TORTS IN THE UPPER WARD OF LANARKSHIRE.

MILES 0 I 2 3 4 FOKTS, CAMPS, AND MOTES OF LANARKSHIRE. 345

of building. Newto Hartred nan e contain large massive curved mounds. The true nature of all these surface-markings might be ascertained by excavation. The circular base of what may have been a large cairn in Nisbet deserves special notice. In the interior of many of the forts, however, tracthero n buildingf s eo grouni e th d dan , seem retaio st s nit original irregular surface, sometimes with natural rock cropping outI . think few f any,i , have been artificially levelled.

3. Geographical Distribution.—A glance at the map (Plate XV.) shows that while the forts are numerous in the Yale of the Clyde proper, and on the Culter Water, which lies between Clydesdale and Tweeddale, they are almost absent in the numerous tributary vales and glens on the south and west. This fact would be much more conspicuous if the showep ma e wholth d e instea f onldo y abou e hillt th hal f yo fpar f o t Uppee th r Ward; becaus omittee th n ei d hal singla f t therno e s fortei . But, confining attention to the glens shown in the map, there are no e Elvanth fort n o ,s Gonner d Douglaan , s waters; only one, besides two small ones of doubtful character, on or near Duneaton Water, and onlconnecteo ytw d with Garpositioe e fortth f th o Water t sf no s A . with regard to each other, it will be found that while a certain number are isolated, other groupee sar d closely together. ease Thuth t n sido s e of the Clyde, near Crawford and close to the river, there are three withi a mila nmile d e an ;furthe , threon r e others withi a miln e; on the Culter Water there are six within a semicircle having a radiu f littlo s e above hal a mile f d Pettinaian ; n Hil s i occupiel d t leasa y tb four. Then, withihundrew fe a n d yard f eaco s h other, there are or were two at St John's Kirk, two at Brownford, and two on Bizzyberr e otheyth Hilln r O hand. , som e solitaryar e d Campsan , - wate exceptionalls i r y isolated amon e hillsgth , 4J miles froothey man r fort.

4. Elevation above Sea.—e th The highest sit thas ei f Arboro t y (1406 feet), next follow in succession Bodsberry (1312), Blackhill (1260), Campswater (1250); then four between 110 120d 0an 0 feet, Bizzyberry No. 1, Snaip, Whiteside, Cairngryfe. Then come seven which only 346 PROCEEDINGS OF THE SOCIETY, FEBRUARY 10, 1890.

slightly exceed 1000 feet—Newton, Cowcastle, Langloch, Nisbet, Diller Law, Chesters d Cocklawan , . d Betwee100an 00 feee 90 nar t sixteen—East, Mid, and "West Crawford, Shiel Bum, Devonshaw, Culter Park t John'S , s , PettinainKir2) kd (Noan ,1 . Knowhead, Bizzyberry No. 2, Castlehill (Candybank), Kier, "Whitecastle, Newbigging, Blackhill (Lesmahagow) e remaininth f O . g fourtee e lowesth ne Hartreear t , Brownford (No. 1 and 2), Craigie and Draffan, which are between 600 and 700 feet above the sea. From this analysis it appears, that, although a considerable number are higher than the present inhabited spots near them, nevertheless Arbory, the most elevated of them higheo n all s i , r tha e largnth e villag f Wanlockheado e e inhabith , - tant f whico s r BeddohD e e amonfounb e talleso t dth gd bes an tt developed of the Scottish people.

5. Local Elevation.—As in Peeblesshire, so also here, a number of the forts are situated on the last eminence of a range of heights before its final subsidenc heigha valleye n th o n r ei t o , projecting intvallee oth y from the flank of a range. Arbory (Plate IX.) is a typical example, except tha 0 fees heigh it 60 tgrount w f lo abovo t s e unusualli d th e y great. Others are on the top of considerable heights, which stand out more independently, such as Bizzyberry, Cocklaw, and Cairngryfe. Others, such as East and Mid Crawford, are on small detached heights at the foot of ranges, or like North Coldchapel (fig. 7) and Fallburn (Plate gentle th XIII.) n eo slopee ,ar whicn si h terminat steee eth p descentf so the ranges. Finally, there are or were at least eight, and possibly ten, senso n callee b n ei whic dn hilca h l forts theis a , r site completele sar y detached from the hills, in the dale of the Clyde. One of these—Castle- hill of Symington—is on the summit of a gently rising height, but e severaothersth f o e ,lexistin th suc s a h g Hartree, Westfieldd an , Brownford, together wit vanishee hth d Brownfor"Wolfclyded an 2 . dNo , t littlbu ed vieha d derivean w d scarcel strengty an y h from their position. It is unfortunate that the total destruction or serious decay of these dale-forts makes it very difficult to determine their true nature. The great width of the trenches of Hartree and "Westfield, although it may partly result from frequent ploughing, appears to differentiate them from the FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 347

ordinary forts, and possibly they were motes; but there seems no reason regaro t e othedth r dale-fort differins a s g essentially fro e hill-fortsmth .

. Defensibility6 e Forts.—th f o e alloEvenw wf i thae moundth t s were originally steepe trenchee d higherth an r d an ,s deeper than now, which no doubt was the case, yet the mere earth-works never could have been very n scarcelformidableca t i doubtee yb d an , d that they were crowned with stone wall palisadesr so e disappearancth , f whiceo h is of course easily understood. The general absence of traverses has already been noticed (p. 334), e wanth f communicatiot o t bu n throug e moundsth h , froline meon of defence to another, is also remarkable. Lieut.-Colonel M'Hardy, s suggesteR.E.ha , d thae outeth t r lines wert intendeno e b o t d defended hand to hand, but rather as obstacles to check the rush of an assault. He was led to this opinion very much by an inspection of the remains at Ardoch, where from their small size the numerous d somewhaan t cramped rampart d trenchean s s see l adaptemil r fo d direct defence. Such also was my impression on visiting Ardoch, ordinare th bun i t y curvilinear fort e concentrith s c " rings muce ar " h further apart d alloan , f wamplo e spac r defenceefo ; sometimes indeed the top of a rampart, which is not the inner one, is enlarged into a broad platfor r placemo d'armes. I think , therefore, tha n theii t r case each line was intended for direct defence, and that the garrison were probably abl retiro et e froe " rinmon g anothe"o t r through doors in the palisades or stone walls, which I suppose to have been necessary adjuncts to the defence. possibls i t I e that simple a singlfort f o se enceinte were intendeo dt protect cattle, &c. t nigha , t from thieve r wilso d animals, rather thao t n stand a siege. At the same time, this does not seem to have been the cas t Nortea h Coldchapel—a fort wit a hsingl w e lo rampart n i d an , ground,—as the confused mounds inside point to its having been a fortified village rather than a cattle shelter.

7. Water Supply.—A great objectio e defensibilitth o nt fortse th f ,yo is the absence of water within them. This is a recognised defect, 8 34 PROCEEDING SOCIETYE TH F O S , FEBEUARY 10, 1890.

not only of Scottish but of Welsh and Irish forts. In the Upper Ward, the large " Eoinan Camps " in the low ground are, indeed, well supplied. Clydesburn " Camp " has two streams flowing through it, and Cleghorn is in marshy ground where springs must exist. The large curvilinear fort of Knowhead, also in a marsh, contains the head-springs of a little burn; and within Bodsberry Mr Vere Irving discovered the singular basin-lik e UpperTh ef o wel Ward4 2 lf o . figurep t a d Lanarkshire, t witvolBu h . i these exceptions o tracen , f wello s s have been found, and in many cases the nearest water is at a most inconvenient distance for daily requirements, and quite beyond the controe garrison th s possiblei f o lt I . , indeed, that careful exploration might detect wells withi y whic e havy fortsan th ns ma hea , existed must have bee n lateni r time s o prevent fille p u d t accident sheeo t s p or f thecattlesame o t observ th my no et an hollowa ; d time n ei di I , s or surface-indication a hidde f o s n well. Perhaps I hav s a e, sug- gested in former papers, the natural distribution of water may have been different formerly from what it is now; and Mr A. P. Lewis has reminde e tham d t General Pitt Biver ss excavationfounhi n i d t a s Bushmore that the water level was higher there in Boman times than now. I have also suggested that water might have been stored in the fort skinn si potsr so .

8. Names of the Forts.—If we may trust the O.M., very few of the forts, or even of their sites, have local names. The term " Kings," so commonly use Peeblesshiren di , seems hardly know Uppee th n i r Wardonle Th y . notice of it I have met with is in the 0. S.A. of Culter, where it is stated that "four encampments are called Bings by the common people.1' Cow Castle d twan o Chesters probable ar e onlth y y instance whicn i s h forts have special names e firs s i obviouslTh t . y e modernth d an , appellation Chesters, although it may be ancient, is perhaps as likely havo t e been bestowe learnee th y modern b di n timesf fort-sitesO . with names, three Castlehills need not detain us, hut greater interest attache Bodsberry,o st Arlory, d Bizziberry,an y surmism f i correc e b e t that berry lory d corruptione an ar f ~burhso thid ;san seems likely enough in a district so Saxon as to contain motes, and where the principal FORTS, CAMPS, AND MOTES OF LAHAEKSHIEE. 349

villages bear such name s Abingtonsa , Koberton, Lamington, Covington, Thankerton, Walston, Liberton, Dolphmton f CeltiO . c name&c , s con- nected with fort r theio s r sites, Kier, name hila th f l eo wit hfora n o t it at Dolphinton, appears to be the sole representative of Caer, and Dun doe t occu questionabls no i t all a rt I . occurn Du sf sense i eveth en ni of "a hill." It occurs indeed in the names of the following hills —Du (221w La n 7 feet)(1740) w n Grai(2186)w La Du , La nn , Du , Dun Law (1669), Dungavel (1674), Dungavel (1502), Dun Moss (1500) Dunsidd an , (1308)g t seemei Ei t bu s; probable tha nearln i t y l thesal e instance e prefith s s realli x e Saxoth y n colour-namet i s a , s associatei d with Saxon roots witd an ,h hills almos l mucal t h above the elevation at which forts are found in Lanarkshire.

9. Antiquity.—Of the numerous and important class of circular and oval works commonly called Britis e saib hdy thafortsma tt i , they must in all probability have preceded the motes—at least, if we allow that the latter represent the castles of the Saxons in the district,—because it is in the highest degree improbable that works of so primitive a character should have intervened betwee e e Saxonth e moteth th nd f o an ss feuda e n datneithel th ca castlee x e fi w rf ston o st d lime an Bu e . of their beginning nor of their end, although we can say this much, systey thaan f f mfortificatioi to n came betwee e motesnth the d m,an it must have been one which has not left a single trace of its existence.

. Political10 Condition d Builders.—Modee th an f Life f o o A study of the characteristics of these oval and circular works appears to throw some light, dim and uncertain though it be, on the mode of life of their inhabitants. We learn, for example, from the situation of the forts, that they were no mere hill-refuges of a timid people, in constant fear of attack from a race more powerful than themselves. A large proportion hilln hillso e t t allst eithear, th quitfooa e f ebu ,or ,no th o t t era removed from them, in the fields of the dales; and the proportion must originally have been still greater forts a , sucn i s h situations were much more liable destructioo t n from agricultur othed ean r causes than when e placeth n do 0 35 PROCEEDING SOCIETYE TH F O S , FEBRUARY 10, 1890.

hills. We know, indeed, of no less than five of these dale-forts in the Upper Ward that have entirely disappeared, some in our own day, and l withial presene nth t century. Again, those hillsn whico e , ar hwit h very few exceptions, closely overlook the dale, and are not often at an inconvenient height abov e preseneth t cultivate wholee d th lands n ,O . therefore, it would seem that the fort-builders occupied the low ground as effectively as their successors in the present day; and supposing that the forts were their habitual dwelling-places, there is no reason, as far as convenienc f sit eo concerneds i einhabitante th y wh , s shoul t havno d e practised agricultural as well as pastoral pursuits. Indeed, there is some reason to believe that they did so, as the remarkable terraces on the hills of Peeblesshire, if not natural—and it seems impossible to believe that those at Eomanno at least are so—appear to bear witness to a system of terrace-cultivation in remote times. In Lanarkshire I know of only a single example. Looking northwards one day, when the light happened to be favourable, from Snaip fort towards Culter Crags, half a mile off, I was surprised to observe that the steep grassy face was marked with parallel line r terraceso s f whico , hI counte leso dn s than twenty-eight, one above the other at tolerably regular intervals, occupying a space several hundred feet high, and about 300 yards in width, interrupted in the middle and lower part by a steep bluff (Plat t tim no walo et . theme to t d kXIV.)fro e ha bu , th m I . steepness of the slope the most of them must be extremely narrow; perhaps, indeed, although visible from a distance they are scarcely trace- e spotth abl n , o elik e thos t a Torwooe d near Peebles (Proc., 1886, p. 65). A difficulty in accepting those narrow terraces on steep slopes as evidences of cultivation is the rarity of their occurrence, although it must be remembered that in such situations they would be easily obliter- woule atedOn d. also expec meeo t t t with terraces more frequentln o y slopes less steeply inclined, if a system of terrace cultivation were in same t awarvogueth no t e A timm ea .tha eI t these remarkable objects have been claime naturas da geologistsy b l . bottoe descene th th t f mA o t from Nisbe bure t th for nI noticeo t t d x terracessi fiv r o e , which, instea f beino d g horizontal, slantee th p du steep side of the ravine, resembling a series of parallel steep roads. They Proc. Soc. Antiq. Soot. YolJCXIV Hate'XIV

TERRACES ON OULTER CRAGS FROM TOP OF SNA1P HILL FORTS, CAMPS, AND MOTES OF LANARKSHIRE. 351

hav a strangele y artificial I presum lookt bu e ,mor ar e e probably natural. Again, some notion may be formed from the grouping of the forts of the relation of their inhabitants to each other. Here we encounter two difficulties. We cannot tell whether the forts were all occupied simul- taneously r whetheno , t havrno eforty existema s arean i d s whicw no h show no trace of them—both circumstances which seriously affect the significance of the present grouping. But it may at least be affirmed that therreasoo n s ei supposo nt e from their structure tha forte th t s belonged to different periods, and that, on hill sites at all events, it is not likely that forts would entirely disappear without leaving any trace of their existence. If we may accept the present as representative, on the whole, of the original grouping sufficientls i t i , y remarkable fortx si s e neasees th a , n ri Crawford and Abington, the six on Culter Water, and the four on Petti- nain Hill; and we may conjecture that the inhabitants of each group must have bee friendln o n y terms with each t otheralsno o f i ,groupe d for some special purpose t furtheBu . r than this e facth , t thae th t forts of the Upper Ward, which lie to the east of the Clyde, are con- tinuous with those of Tweeddale, while there is almost an entire absence of them in the tributary valleys to the west, seems to point to the pro- bability that the forts of Peeblesshire and of the Upper Ward belonged to one kingdom, of which Clydesdale proper formed the western boundary.

In concluding this review of the Forts of the Upper Ward of Lanark- shire, it is to be regretted that the results should be in general so vague. thiIn s respec count worsthe tno tha resis yoff e n the Scotlandof t . The only means of extending our information, and of building a better foundatio r deductionsou r t nfo beloy excavation ge b e s o i wt ,th s a o s , mere surface of things. Efforts in this direction have been so few and desultor uselesse b t misleading o reaf t no o s f y,a i e l b valu o T Scienceo e.t , they shoul made b dexperty b ea well-organise n o s d syste thed an ym ; would be most hopeful if undertaken by a Society such as ours, or, at least, wit advichits guidanceeand . Here enter questiosthe expensenof , 2 35 PKOCEEDING SOCIETYE TH F O S , FEBRUAR , 1890Y10 .

and scientific Societies, proverbiall t littlbu eo y d withoupoor n ca , t greater endowments tha trifline nth g ones the t presenya t possese th r sfo special purpose of investigation of the kind. There can be little doubt that prehistoric archeology has more to hope for from excavation than any other source; but the ages pass, and with them pass away at our very door materiale sth r investigationsfo , without their true naturd ean meaning being understood o departmenn n I . this ha ts neglect beeo ns marke prehistorie thath n i f s o t da firse th tc s ste a forts reford o pt an , m is to know our wants, it is possible that this paper, and my former ones a osimilaf r nature y eventuallma , y prov f utilityo e t onlno y , b y rousing attention to a deficiency, but by indicating the precise remains which may be expected to yield the best results to systematic ex- cavation.