Proc Antiqc So Scot, (1997)7 12 , 855-8 Aspects of the architectural history of Kirkwall Cathedral Malcolm Thurlby* ABSTRACT This paper considers intendedthe Romanesque formthe of Kirkwallof eastend Cathedraland presents further evidence failurethe Romanesque for ofthe crossing, investigates exactthe natureof its rebuilding and that of select areas of the adjacent transepts, nave and choir. The extension of the eastern arm is examined with particular attention to the lavish main arcades and the form of the great east window. Their place medievalin architecture Britainin exploredis progressiveand and conservative elements building ofthe evaluatedare context building. the ofthe in use ofthe INTRODUCTION sequence Th f constructioeo t Magnus'S f o n s Cathedra t Kirkwalla l , Orkney comples i , d xan unusual. The basic chronology was established by MacGibbon & Ross (1896, 259-92) and the accoune Orkneth n i ty Inventory e Royath f o l Commissio e Ancienth d Historican o an nt l Monuments of Scotland (RCAHMS 1946,113-25)(illus 1 & 2). The Romanesque cathedral was begun by Earl Rognvald in 1137. Construction moved slowly westwards into the nave before the crossing was rebuilt in the Transitional style and at the same time modifications were made to the transepts includin erectioe gpresene th th f no t square eastern chapels. Shortly after thi sstara t wa sextensioe madth eastere n eo th befor f m no n ar e returnin nave e worgo t th t thi n .A k o s stage no reason was given for the remodelling of the crossing and transepts in the late 12th century. Aspect f thio s s chronology have been refine receny b d t writers. Stewart Cruden (1977, 87-8; 1986, 119; 1988, 85) examined the extent of the Romanesque build and argued that the eastern arm was probably built quickly so as to accommodate the high altar and the shrine of St Magnus. He observed that 'in 1151 the founder and his bishop departed from Orkney on a two-year crusad Hole th yo et Lan backd dan adventurn a , e whic beed hha n discussed wite hth kin Norwaf go 1148n yi unlikels i t I . y that they would leave befor choie completes eth rwa d dan relice Magnut S th f so s enshrined within it' (Cruden 1977, 88). Cruden suggested tha originae th t l east end had an apse-echelon plan with an ambulatory returning across the eastern bay between St Magnus's Shrine in the apse and the high altar immediately to the west (Cruden 1977, fig 5; ; 1988 1986crossin 12 w 11) g datee g ne ,fi fi ,H . e dgth about 1170, afte presumee th r d collapsf eo Romanesqus it e predecessor observed an , d that 'tharched eol crossinsw adjacenne e gth wero t t e carefully re-fashioned' (Cruden 1977, 90). 2 Cooper Court, Brantford, Ontario, Canada N3T 6G2 85 | 6SOCIET ANTIQUARIEF YO SCOTLANDF SO , 1997 ILLUS 1 Kirkwall Cathedral: plan (after RCAHMS 1946, fig 225) ILLUS 2 Kirkwall Cathedral: longitudinal section (after RCAHMS 1946, fig 226) Cruden's interpretatio liturgicae th f no l divisio Romanesque th f no challenges ewa easd ten d by Eric Cambridge (1988, 119-21) o founwh , d insufficient evidenc n ambulatorya r fo e . Cambridge provided one of the most detailed accounts of the stylistic similarities and differences between Kirkwall and Durham, albeit without any discussion of the choir aisle rib vaults, a point subsequently addresse Eriy db c Fernie (1994, 272-3). Cambridge also presente da carefull y reasoned interpretation of the iconography of Kirkwall Cathedral. Cambridge's account complements the stimulating essay on the entire fabric by Richard Fawcett (1988). Fawcett clarified aspects of Cruden's reading of the first great Romanesque campaign and suggested that the string-course on the aisle side of the easternmost Romanesque pier of the south choir arcade THURLBY: ARCHITECTURAL HISTORY OF KIRKWALL CATHEDRAL 857 extends too far to the east to allow an apsidal termination to the aisle (Fawcett 1988, 89, pi 4a). He observed that work on the Romanesque nave aisles progressed in a rather faltering manner in contras late th e o 12th-centurt y rebuildin crossinge th f g o remodellin e th , transepte th f go d san extensioe stare th easterth e f o tth whicf m no n ar h indicated infusio 'botw ne ha fundf no a t a s critica lwork e stag th alsd f an eo , perhap interventioe sth mora f no e dynamic driving force' (Fawcett 1988, 97) thee H . n mooted that impetuthiw sne s might correspond wit electioe hth f no Bishop Biarni Kolbeinsson aftedeate th rf Williaho n 1188i e continuedI mI H . : 'His long episcopate [1188-1223] would cover much of an important building period, and he would seem hav o somf t o en ebee energyma na ' (Fawcett 1988, 97)liaisoe e attractivn th Th . a t s ni bu e eon implication is that the rebuilding was driven by liturgical and/or aesthetic considerations. This is almost certainl eastere y th tru r enfo extension whic 'probabls hwa accommodato yt chaptee eth r canonf o planne cathedraw e basie th sno r th s fo d a l constitution' (Can twhic d 1972an ) h7 , finds numerous parallels in contemporary churches in England to provide more space and a more glorious settin r majogfo r shrines (Draper 1979; 1980; 1987) requiremente Th . canone th f so s alsy oma explai rebuildine nth transepe th f go t chapels (Cant 1972 . However7) , difficuls i t i , o t t imagin eliturgicaa l rebuildine reasoth r crossinge nfo th f go . Moreover retentioe th ,e th f no Romanesque work in the three-bay choir and in the transepts and nave suggests that the primary motivation for the major structural undertaking of the replacement of the crossing was probably not aesthetic. THE INTENDED TERMINATION OF THE CHOIR AISLES arrangemene Th maie th nf t o arcade Romanesque th f so e choi unusuas ri tha n easternmosli e tth t freestanding pier is asymmetrical; to the east it is flat to correspond to the east respond, while to the west its segmental form agrees with the columnar pier to the west (illus 1-4). The arch of the eastern bay is narrower than either the first or second bay, counting from the crossing, and while its orders toward choie sth r hav same eth e moulding othee archeso th s rtw sa inne e th , r ordes ri carrie pairen do d head corbels rather than wester restine capitale th th n i n go ns s a bay s (illu. s5) Furthermore ordero , thertw e s ear toward westere aisle th sth centran d ei nan l bay t thresbu n ei eastere inclusioe th (illu y Th thire . nba th s4) df n o orde causes rha d problem vaulsb ri wit e thth of the aisle. The rib cuts off the western springer of the outer order of the arch and even cuts into springee th secone th f ro d order (illus 3-5) suco .N h proble wese bayo mth t tw o occurs t e th n si because the ribs there do not have to compete for space on the capital with a third order of the arch. The vault of the east bay goes with the early Gothic extension to the east as witnessed by the form of the wall arch and the transverse arch between bays 2 and 3; the former does not hug hooe th d moul windo, while 2 th pointe e d bayf n dei o th an s ws1 a d for compled man x mouldings transverse th f o e arch contrasts wit round-headee hth d transverse arch betwee wit2 n d bayh an s1 their simple angle rolls (illus 3, 4, 6 & 7). Furthermore, the crease at the apex of the web of the Gothic vault is also witnessed in the eastern severy of the vault of the second bay in contrast to the work to the west. The different articulation of the pier and the arch in the Romanesque eastern bay may suggest that this space had a different function. This would seem to be confirmed by the clumsy junction betweeoutee th Gothird e Romanesque ordenth an th b f cri o r e arcade along wite hth absence of any trace of an earlier vault in the eastern bay. Why the different treatment of this bay? The answer is twofold. In the first place, the use of paired grotesque head corbels on the flat face respondsamthe the choi eas is the pieofrsof the aislr bay e vaults (illus 3-5). Such continuityin THURLBY: ARCHITECTURAL HISTORY OF KIRKWALL CATHEDRAL | 859 articulation suggest continuitsa function yi n (Cambridge 1988, 119). Therefor plausibls i t ei o et reconstruct an ambulatory across the eastern bay of the choir to provide access to the shrine of St Magnus in the apse, as suggested by Cruden. The Kirkwall master mason thereby gave monumental architectural expression to a liturgical division that was probably created with wooden screens at Durham Cathedral to provide access behind the high altar to the shrine of St Cuthbert (Klukas 1983, 173-5; Fernie 1993b, 150-1). segmentae Th le pier e faceth th n sf i so two western bays articulate the sanctuary, as at Norwich Cathedra1 l and elsewhere (Fernie 1993a, 67; Hoey 53)1989seconn e 5 Th . 25 , d aspecarticulatioe th f to eastere th speakf y no nba morf so e ambitious intentions.