Proc Soc Antiq Scot, 126 (1996), 871-879

tower-housedatine e th Th f go t Comlongosa d nan Elphinstone Alastair Maxwell-Irving*

ABSTRACT significanceThe greatthe of tower-house castles Comlongonof Elphinstoneand historythe in of Scottish architecture suchis thatdatethe when they were builtmattera is of prime importance. In the past certain assumptions have been made, both in the dating of these two castles and, by analogy, in the dating of similar works; and these misunderstandings have now persisted for so long that they longerno are questioned. purposeThe thisof short evidence paperlookthe to at is to afresh and try to put the record straight.

INTRODUCTION The castle of Comlongon in Dumfries & (illus 1), with its myriad of mural chambers - man lesd o thern an s y e tha- eothear 2 n1 r architectural refinements justls i , y recognize 'perhaps da s the ablest thing of its kind in Scotland, and in it the plain rectangular tower-house . .. may be thought havo t e reache climaxs dit ' (Simpson 1942 distinctioa , s 20)i t I . t shareni s wit formee hth r tower- castl f Elphinstoneo e Eas(illun i ) t2 s sadlw no y, reduce lowess it o dt t storey. Theso etw massive castleth d an se stronghol f Borthwicdo Midlothian ki n have collectively been describey b d Crude s 'pre-eminena n t amon l towergal s ... nevesurpassee b o t r scaln di r strengtho e ' (Cruden 1960, 144). Certainly questione on o n , s their status. Strength , characteristie howeveron e th s i , c that was, in fact, lacking at both Comlongon and Elphinstone. Comlongon's north wall was so weakened s murait y b l chamber thruse th f theid e stabilizeo tb an s o rt vaultd dha ss internall thaha t i t y with massive timber supports, while Elphinstone, similarly handicapped and further weakened by subsid- ence demolishee b , o finallt d ha yc d1963 . Neither would have survive attacn da k with artillery, unlike Borthwick, which was besieged in 1650 and still bears the scars of Cromwell's guns. And ther e otheear r significant differences from Borthwick, notabl thein yi r great halls, bot f whicho h incorporat kitchee e oppositth e th t frod n a e maien emth n fireplace e knoW . w thae licencth t o t e build Borthwic grantes wa k 143n i d 0 (RMS 157)o , n t whevo , bu ,II l n wer othee builto eth rtw ? s 1788a s lono A , gag Cardonne l suggested tha presene th t t castl f Comlongoeo alreads nwa y existencn i time th Davif eo n ei I (CardonnedI l 1788, 'Comlongon')! While nobody would take such a date seriously today, the apparent affinity of the sophistication of Comlongon and Elphinstone with Borthwick has led to the increasing conviction that they must have been built soon after Borthwick, and probably within the period 1440-75. The Royal Commission's Report on Dumfriesshire, pub- lished in 1920, describes Comlongon only as 'an unusually complete and well-preserved example of a 15th-century tower' (RCAHMS 1920, no 537), while Elphinstone was described in 1924 as 'prob-

* Telford House, Blairlogie, Stirling FK9 5PX

874 I SOCIETY OF ANTIQUARIES OF SCOTLAND, 1996

ably erecte e 15tth hn di century ' (RCAHMS 1924 192)o n , . Three years later, Mackay Mackenzie refrained from giving any date to Elphinstone, and only went so far as to ascribe Comlongon 'to this [15th] century' (Mackenzie 1927 ,194y b 187) t 2opinioe Bu Simpsoth . f o ns thanwa t 'to judgy eb the style, this [the building of Comlongon] must have happened about 1450, or soon thereafter' (Simpson 1942 196y ,b 28)d 0 Crudean , confidens nwa t that both Comlongo Elphinstond nan e were onl yyear w 'afe s later [than Borthwick]' (Cruden 1960, 128).' This belie bees ha fn sustainey db other scholars since. Tranter wrote that 'Comlongon resemble contemporare sth y 15th-century tower of Elphinstone' (Tranter 1965, 79); Stell that 'it [Comlongon buils wa ]t probabl aboun yi thire th t d quarter of the 15th century' (Stell 1986, 102); and recently Fawcett has dated it 'around the central decade fifteente th f o s h century' (Fawcett 1994, 244). Zeun edeclaro t eve s a n r e wen fa tha o s t 'three out of a handful of masons' marks at both places [Borthwick and Comlongon] are absolutely identical' (Zeune 1984, 10). More will be said of this presently.

ELPHINSTONE CASTLE With so much scholarship increasingly confident in a dating around the middle of the 15th century, why is there a need to take a fresh look at the question? First, let us consider Elphinstone. Mac- Gibbo Rossn& , writin 1887n gi , were much more guarde thein di r appraisa buildinge th f o l . They wrote that Elphinstone 'resembles ordinary fifteenth-century work, whil e corbellingth e , cable moulding gargoyled an , parapete th t a s , som f whiceo carvee har resemblo dt e cannons, cannoe b t earlier than the end of the century' (MacGibbon & Ross, vol I, 236). The parapet walk is, in fact, a later reconstruction, dating from the latter half of the 16th century, a point fully appreciated by the Royal Commission (RCAHMS 1924, no 192). argumene On t thabees t forwarha t npu supporn di mid-15th-centura f o t y dat r Elphinstonefo e e incorporatioth s i e arm th f Johnsto o sf no f Annandalno e armoriath n i e l panels abov e halth e l fireplace. According to MacGibbon & Ross, this had often led to the assumption that the tower must have been built by Sir Gilbert Johnston, who acquired Elphinstone by marriage to Agnes, daughter and heiress of Sir Alexander Elphinstone of that Ilk, the last of his line, who was killed at Piperdean in 1435 (MacGibbo , Ross237)I n& Fawcetd l an vo ;, t reiterate s 'dateabls thawa t i t heraldin eo c ground arouno t s d 1440' (Fawcett 1994, 245) bee d r Gilbertha Si .n o 'faithfuwh , l squire Kino t ' g James III in 1472, was still alive as late as 1482,2 when he was witness to a charter (RMS, vol II, no 1618); but the Johnston arms could equally well relate to his son, Sir Adam Johnston of Elphin- stils lwa stone witnessino wh , g charter 150n si 3008)o 6n (RMS, poinII A . l t beetvo no , nthas ha t mentioned elsewher face th t s ei thaheraldie th t c displa incompletes ywa photographie Th . c record of Elphinstone (illu ) clearl3 s y shows that, immediatel lase righe th t th f shieldo to y t wale s th , wa l at some period opened up to admit more light to the hall (and later crudely filled in again), with the consequent loss of four or five armorial panels. The truth is that the heraldic evidence is far from

conclusive in providing a precise date. Indeed, Si3 r William Fraser, in his history of the Elphinstone family, points out some anomalies in the arms themselves (Fraser 1897, vol I, xi). On the other hand, Sir William Fraser provides a very convincing argument as to why Elphin- stone could not have been built by Sir Gilbert Johnston. After the death of Sir Alexander Elphinstone in 1435, there was protracted litigation over the succession to the Elphinstone estates between his daughter, Agnes (who is said to have been born posthumously), and the heir-male of the family, his younger brother Henry; and the matter was not finally resolved until 1476, when the appointed arbiters divided the estate between the two claimants, and awarded Elphinstone itself to Agnes (Fraser , 14)x theny , B 1897.I l , howevervo , litigatioe th , proved nha disastrouslo ds y costly that neither Agnes's inheritance, nor her husband's slender means as a younger son of Adam Johnston of that MAXWELL-IRVING: TOWER-HOUSES AT COMLONGON AND ELPHINSTONE I 875

ILLUS 3 Armorial bearings about the hall fireplace at Elphinstone (Reproduced by permission of the RCAHMS© Crown Copyright)

Ilk (Fraser 1894, vol I, cccxxiv), could possibly have supported the cost of any new tower, let alone one of Elphinstone's stature. The new tower had to wait until the family's coffers had once again been replenished.

COMLONGON CASTLE case th f Comlongon eo n I fine w ,d ourselve firmen o s r ground, for, althoug t widelhno y knowne th , licenc constructios it r efo survived.probabls s t datednha wa no t s i bu ,y t 4I grante d 1500-1. t reads5I :

LetteA f Licencro e mad JOHNNo et MURRAf Eo towr a Cokpulf Y fortalico g aerisd s bi ehi an o d t , ean e apon landis ehi s lian Ananderdalen di mao t kd aponan , sammye eth n machcoling, corbalsalze, irnezettis saie th d ang towrbi d d windoisquhao et an s tra hichimpedimenty o ,t pies on etc.e d t h an ,bu , , etc. (RSS, 692).o n vo , I l 6

John Murray of Cockpool succeeded his father, Cuthbert, in 1493, and the following year had sasine of the lands of Ruthwell, etc, which included the lands of Comlongon (Scots Peerage, vol I, 222; Excheq Rolls, vol X, 770). Although he was subsequently knighted, little seems to be known

s lifeohi f , excep s evidentlt wa tha e h t favoun yi s thit courta rs wa Joht I . n Murray7 , who, having received the Licence, built the present magnificent tower, which was to be one of the finest and most sophisticated tower-castles of its age. It was a surprisingly ambitious venture, for despite the high I SOCIET 6 87 ANTIQUARIEF YO SCOTLANDF SO , 1996

offices hel variouy db s members, it f o s famile traditioa th d ha y f keepinno gfairla profilew lo y .

Nor are they known t8 o have been especially rich or prominent at that time. towee Th r itsel s firsi f t mentione 1507/8n di , Februarywhen5 1 n o , , Jame granteV sI r Si o dt John Murra f Cokpooyo chartea l f 'tho r e land f Cokpuleso , Ruvale-tenement [Ruthwell] towee th , r and fortalice of Cunlungane [Comlongon], etc.', which Sir John had personally resigned, and for special favour incorporated them into the free barony of Cokpule (RMS, vol II, no 3194). By the

same charter, James IV also created the 'Town of Ruvale' and the lands of the same in 9 Ruvale- tenemen frea t e burge calleb f baronyBurge o ho t dth w f Ruvalho no , e (Ruthwell).10 This charter was confirmed by parliament the following year (RMS, vol II, no 3346). Zeune's reference to three masons' marks being found at both Comlongon and Borthwick woul erron da e seerb (Zeunmo t e 1984, 10) thorougA . h survefabrie th f Comlongof o cyo s nha been carried outonld marke an ,yth f threso e different masons have been f whicfoundo o htw , bear resemblancdiffereno o s n r o 4 1 t marke eth whatevef o s knowy an o t Borthwickrt na onle Th .y mark that does bear a resemblance, the intertwined W, which in various forms is common to many masons' marks, is not in fact the same. Masons' marks are fairly precise, and there are important differences t Comlongoa betwees W e d Borthwick.th nan n " Moreover s Fawceta , t acknowledges, similar masons' marks do 'not necessarily mean that the same master mason was responsible' (Fawcett 1994, 246).

COMMON FEATURES characteristie on e Th Comlongof co Elphinstond nan e tha moss i t t immediately recognisabl beins ea g different from most other tower-house large th s esi kitchen fireplac greae opposite th f th t o t ea d een hall to the main fireplace. This arrangement, however, is not unique to these two towers, as it is fount leasa n ti d five others groua , fouf po - rLittl e Cumbrae Skelmorlid , Fairliean w n i La , e- Ayrshire Saddeld an , Kintyren i l . These wer l noteeal MacGibboy db Ross n& givo ewh , detailed plans of each (MacGibbon & Ross, vol III, 173-83, 197-200). They all appear to date from the beginning of the 16th century, though actual dates are only available for two: Skelmorlie, which was buil Williay b t m Cunningham Eare f th Glencairn o ,lf eldeso n so t , sometime around 1502 (Close 1992, 86)Saddelld an ; , whicBishoe buils th r hwa f Argyl tfo po l between 150 151d 8an 2 (RCAHMS 1971, no 313). On the evidence of licences granted in 1534 and 1537, it is likely that the tower on Little Cumbrae dates from that time formee th ;grantes wa r Robero dt t Hunte f Hunterstoo r a r nfo 'mansion' (RMS Eare e latte f l th EglintonIIIth o o 1432l o n vo ,,t d r an ) , with 'powe builo t r da tower' (RMS, vol III, no 1675). Cruden was of the opinion that the four Ayrshire castles dated from the 'end of the fifteenth century or the commencement of the sixteenth century' (Cruden 1960, 139). Another distinctive feature that was found at both Comlongon and Elphinstone12 was the moulding of the fireplace jambs in their great halls. They were both alike, and also very similar to the mouldings to be seen at Cardoness, in the Stewartry; but quite unlike Borthwick. No precise date is known for the construction of Cardoness, but it has generally been 'attributed' to the end of the 15th century (MacGibbo , 247RossI n& l ; vo ,Stel l 1986, 101). Perhap e datinth s f Cardonesgo s should als lookee ob t againda . otheTwo r features that were simila botat r h Comlongo Elphinstonenand quitbut , e different at Borthwick, were the ground-floor main entrances and the mural prisons without garderobe facilit- ies. Bot f thesho e feature more ar s e typica f 16th-centuro l y work.

CONCLUSIONS evidence th l lighe al I nth f e o t relatin Comlongoo gt Elphinstoned nan t wouli , d appear that they both belong to the same general period around the turn of the 16th century, or not long afterwards; MAXWELL-IRVING: TOWER-HOUSE COMLONGOT SA ELPHINSTOND NAN 7 87 I E

s self-evideni t i d an t tha t commoa ther s ewa n facto thein i r r design. Ther certainls ewa familya y

connection between the Johnstons of Elphinstone and the Murrays of Cockpool, and it is quite 3

possible that it was one of them who instigated the new kitchen arrangement, late1 r to be copied elsewhere. There was, however connectioo n , n with Borthwick, other than that each representeda supreme example of a tower-house castle of its respective period. By bringing forwar constructioe dth f Comlongonno probabld an , y Elphinston turee tooth n o t , of the 16th century, we have perhaps helped to explain why the new arrangement did not become more widespread. The subsequent absence of tower-house castles of this sophistication in the south of Scotland may be attributable to Scotland's disastrous defeat at Flodden in 1513, when the great families lost not only the flower of their youth, but also their wealth. For many years thereafter, there littls wa e towe Bordere rth buildinn adjacend i l san al t ga t counties. Indeed, ther soms ei e evidence that building in the Borders was discouraged. In 1528, the Council ordered the demolition of a partly built tower south of Coldingham 'sen it stands so nerr the bordouris of Ingland and may turne this realme to hurt and dampnage' (Acts of Lords of Council in Public Affairs, 277).14 It was not until seven years later that parliament, realizing how vulnerable the kingdom had become, passed its celebrate t 'FodAc r biggin f Strenthigo Bordourise y n so ' (APS, vo, 346)II l . This required that

evry landi duellann Inlanma e t y n dr upoi d o bordourie ny s havand yare ane hundreth pundw lanne f do exten sufficiene tan Saig bi l t barmkyn .. f Stan o .lym d ean e contenand thre scor esquare futiy e f o sean El ni Elny v thic d skan Ressetheich e s tennenti y gudee hi r defensth d m fo ttrubloun d an tsi hi san f o s s thinkie h self m l uthi gi t f expedienhi samse I al tour e r t th rf y an fo tymio n landi t ei d n ewy An t me t smallar Rent and Revenew big pelis and gret strenthis as yai pless for saifing of yare selfis men tennentis gudis.

Only then did tower-houses begin to reappear on the scene in southern Scotland.

ACKNOWLEDGEMENTS 1 would like to thank Mr A Ptolomey of Comlongon, Mr and Mrs A Duncan of Tower Farm, Elphinstone, and Mrs Christine Calloway, General Manager of Borthwick Castle Hotel, for kindly

lettin have gm e free accesrespective th o t s e properties,r Stuar M Scottise Allad th W t f an n o h 5

Record s assistanc Offichi r fo e conductinn ei researchy gm I woul. 1 d also lik thano t e r AthokD l Murray, formerly Keepee Recordth f o rf Scotlando s s invaluablhi r fo , e advic n documentaro e y source d Geoffrean s y Stell, Hea f Architecturo d e Royath t a el Commissio e Ancienth n d o nan t Historical Monuments of Scotland (RCAHMS), for his assistance with my researches at Comlongon, and for kindly reading the draft of this paper and sharing with me the benefit of his vast experience.

NOTES 1 It is perhaps worth noting that Cruden also attributes Balvaird in Perthshire to the same period, whereas it is now thought to date to c 1500 (Fawcett 1994, 257). conservatoa s wa truc e e th H ef ro wit 2 h Englan 148n di 4 (Bai 1505)o n n, 1881IV . l vo , 3 National Monuments Recor f Scotlando d (NMRS) photographic recor f Elphinstondo e Tower. 4 The reason the licence is not well known is that there is no mention of 'Comlongon' in it. It is the same date reasof th buildin eo y nearbe th nwh f go y Hoddom Castl Repentancd ean e Towert no n 1565 i , s wa , know r certainfo n until Joachim Zeune recognize significance dth contemporara f o e y referenc theio et r construction: neither was named (See Maxwell-Irving 1987, 204). licence Th e canno5 t have been recorded earlier tha n 1501y precedin e 1date th 3Ma f th , eo ge entrth n yo same Folio (Fo.50e Registeth f o )691)o n r , (RSSI , l thougvo , t couli h d have been granted earlier, 878 I SOCIETY OF ANTIQUARIES OF SCOTLAND, 1996

possibly even prio o 150t r 0 (though certainl t beforyno e 1494, when John Murra sasind yha e (supra)). Entries in the MS Register are not in chronological order. 6 'A Letter of Licence granted to John Murray of Cockpool and his heirs to build a tower and fortalice upon his lands lying in , and to include in the construction machicolations, corbelling, iron yett windowsd an s raiso t ,build etcd e saiean th dan ,d towe whateveo t r r heigh pleasese h t , without impediment, etc.' 7 The family's earlier stronghold had been at Cockpool, on the edge of the Solway, and it was from this that they took their designation. See Reid 1954, 190-2. 8 His father had sat in the Scottish parliament, and had also been one of the commanders at the battle of the Kirtle in 1484, when Albany and the Douglases were defeated; and his grandfather had been a commissioner for the peace with England in 1457 and later Warden of the West March. chartee costh e £10s f Th o twa r 0 (ALHT 9 . 5) , , voIV l 10 See Pryde 1951, 100. 11 The NMRS has photographs of masons' marks at Elphinstone. Two of the simplest marks have parallels at Borthwick unusua n neithet a f bu ,o s i lr form theid an ,r rudimentary natur e- compare d wit more hth e elaborat mord ean e precise marks also found ther e- woul d sugges wore th apprenticesf t ko . 12 NMRS photographic record of Elphinstone Tower. 13 Both families had lands around Moffat, in Upper Annandale; and although they were in dispute in 1504, they had also intermarried. 14 Armstrong's (1883, 245n) interpretatio e reporth f f thio o nt s cas Balfour'sn i e Practicks generaa s a l inhibition on building in the Borders is, however, not supported by the original record. 15 Elphinstone Towe properte th s i r f Loryo d Elphinstone.

REFERENCES

Acts of the Lords of Council in Public Affairs, Hannay, R K (ed) 1932. Edinburgh. ALHT 1877-1978 Accounts Lorde oth f High Treasurer of Scotland, 1473-1580. Dickson t ale (eds)T , . Edin- burgh. 1844-7S AP 5 Acts Parliamentse oth f of Scotland, 1124-1707. Thomson Innes& C ,(eds) C , . Edinburgh. Armstrong 188 B Historye R ,3Th of , , Ewesdale, Wauchopedale Debateablee th d an Land. Edinburgh. Bain 1881-1970 Calendar of Documents relating to Scotland, 1108-1516. Bain, J et al (eds). Edinburgh. Cardonnel e 178d 8A , Picturesque Antiquities of Scotland. London. Close, R 1992 Ayrshire & Arran: An Illustrated Architectural Guide. Edinburgh. Cruden 196 S ,Scottish e 0Th Castle. Edinburgh. Excheq Rolls 1878-1908 The Exchequer Rolls of Scotland, 1264-1600. Stuart, J, Burnett, G, et al (eds). Edin- burgh. Fawcett 199R , 4 Scottish Architecture: From Accessione th Stewartse th Reformation,e f th o o t 1371-1560. Edinburgh. Eraser, Sir W 1894 The Annandale Family Book of the Johnstones, Earls and Marquises of Annandale. Edin- burgh. Eraser, Sir W 1897 The Elphinstone Family Book of the Lords Elphinstone, Balmerino and Coupar. Edinburgh. MacGibbon Ross& D ,1887-9 T , Castellatede 2Th Domesticd an Architecture of Scotland, vols3 . Edinburgh. Mackenzie, W M 1927 The Mediaeval Castle in Scotland. London. Maxwell-Irving 198T 7M 'HoddoA , m Castle Reappraisaa : Architecturs it f o l Placd an eHistory' n ei , Proc Soc Antiq Soc, 117 (1987), 183-217. Pryde, G S 1951 'The Burghs of Dumfriesshire and Galloway: Their Origin and Status', Trans Dumfriesshire Galloway Natur Hist Antiq Soc, Seriesd 3r , XXIX (1951), 81-131. RCAHMS 192 Royae 0Th l Commissio Ancienn no Historicad an t l Monument Construction& s f Scotlando s . Seventh Report with Inventory of Monuments and Constructions in the County of Dumfries. Edinburgh. MAXWELL-IRVING: TOWER-HOUSES AT COMLONGON AND ELPHINSTONE I 879

RCAHMS 192 Royae 4Th l Commissio Ancienn no Historicad an t l Monument Construction& s f Scotlando s . Eighth Report with Inventory of Monuments Constructionsd an Countye th n i of East Lothian. Edinburgh. RCAHMS 1971 The Royal Commission on the Ancient and Historical Monuments of Scotland. : An Inventory of the Ancient Monuments, vol I: . Edinburgh. Reid, R C 1954 'The Site of Cokpule', Trans Dumfriesshire Galloway Natur Hist Antiq Soc, 3rd Series, XXXII (1954), 190-2. 1882-191S RM 4 Registmm Magni Sigilli Regum Scotorum: Registere Th Greate oth f Seal of Scotland, 1306- 1668. Thomson& B Paul J (eds) r M Si J , . Reprinted 1984. Edinburgh. S 1908-8RS 2 Registmm Secreti Sigilli Regum Scotorum: Registere Privye Th th f o Seal f Scotland,o 1488- 1584. Livingstone t ale (eds)M , . Edinburgh. Scots Peerage 1904 The Scots Peerage, vol I. Paul, Sir J B (ed). Edinburgh. Simpson 194D 2W , 'Comlongon Tower, Dumfriesshire', Trans Dumfriesshire Galloway Natur Hist Antiq Soc, 3rd Series, XXIII (1942), 20-9. Stell 198G , 6 Exploring Scotland's Heritage: Dumfries Galloway.& Edinburgh. Tranter 196 N ,Fortifiede 5Th House Scotland.n i . South-West3 l Vo Scotland. Edinburgh. Zeune J 198, Longe 4Th Pause: ReconsiderationA of Scottish Castle-Building 1480-1560.. c Miinchen.