Rochester Cathedral: The north choir aisle and the space between it and 'Gundulf's' Tower

ROCHESTER CATHEDRA L: THE NORTH CHOIR A ISLE AND THE SPACE BETWEEN IT AND 'GUNDU LF'S' TOWER J. PHILIP McALEER The north and south aisles of the solid-walled choir of Rochester Cathedral are two of the most complex parts of the building in terms of the sequence of events which has resulted in their present form. Of the two, the form of the south aisle immediately appears the more unusual and leads the inquisitive and perceptive student of medieval architecture to search out an explanation, in an attempt to solve the puzzle its numerous irregularities and variety of parts present. In contrast, our student or, indeed, any visitor may pass through the relatively dimly-lighted, 'straight-forward' corridor of the north aisle, with little hesitation in the journey from the major transept to the minor one. Nevertheless, the constructional history of this aisle is, if anything, as curious and complicated as that of the south aisle, and, in addition, what little has been written about it presents a misinterpretation of the evidence of the fabric.1 Before presenting an analysis of that fabric, consisting of both the inside of the aisle and the space between it and the so-called Gundulf's Tower,2 a brief description calling attention to some of the features which require explanation is needed. The aisle is basically a narrow corridor (Plate IA), its most obvious feature a flight of steps halfway in its length leading up to the higher 1 F.H. Fairweather, 'Gundulf's Cathedral and Priory Church of St. Andrew, Rochester: some critical remarks upon the hitherto accepted plan', Arch. Journ., lxxxvi (1929), 208-10, Pl. II and Fig. 1 (193). 2 The tower was not associated with Gundulf until the eighteenth century. The first to credit the tower to Gundulf may have been Browne Willis, An History of the Mitred Parliamentary Abbies and Conventual Cathedral Churches, 2 vols., London 1718-19, I, 286. Recently, it has been argued, unconvincingly in my opinion, that the tower is a midtwelfth- century campanile: see T. Tatton-Brown, "'Gundulf's" Tower', Friends of Rochester Cathedral: Report for 1990/91 (1991), 7-12. 127 J. PHILIP McALEER PLATE IA Nonh choir aisle, general view to east (author). 128 ROCHESTER CATHEDRAL level of the minor transept and presbytery due to the crypt below them. The south wall of the aisle - the north wall of the choir - is completely solid, from top to bottom and end to end, its only relieving features a continuous bench at the base and the single 'en delit' wall shafts defining four bays relating to the vault (Plate IB). Initially, the north wall appears equally solid, as there are no windows at the level one normally expects or encounters them (Plate IIA). Rather, one discovers the windows are high up in the wall, like clerestory windows, just under the vault. The lower part of the wall surface is broken by several doors and recesses (reading west to east): in the first bay, there is a small portal with a pointed arch; in the third bay, a smaller portal, also with a pointed arch; and in the fourth, last bay, a narrow, semicircularly-arched portal under a wide recess created by a large pointed arch. The number of doors is rather surprising, and some explanation of them must be sought, as also for the absence of windows at the normal (lower) level. But first, several parts of the aisle and the basic sequence of events determining its form can be outlined and quickly accounted for within the context of the known history of the cathedral. The solid south side wall - the north wall of the choir - is generally explained and identified as a survival from the late eleventh-century building, built by the second Norman bishop, Gundulf (1076/7-1108), with the aid of his friend, colleague, and mentor, Archbishop Lanfranc of Canterbury (1070-89). Such solid side walls to a choir are known from several other post-Conquest structures in Britain, and there are parallels in Normandy.3 Whatever the reason for solid walls, in place of the usual arcade between choir and aisles, the custom appears to have ceased about 1100. As no examples are known in the later twelfth and thirteenth centuries, it is assumed that the early Gothic builders retained the late Romanesque choir walls in their new work, adding a register of blind arcading to the inner faces, above which they constructed a new clerestory - replacing the Romanesque one - and ribbed vaults.4 In contrast, the Gothic builders 'demolished the Romanesque aisles; 3 Other examples in England were: Edward the Confessor's Westminster Abbey, c. 1055; Lincoln Cathedral, begun 1072/3; St. Albans Abbey, begun after 1077; Whitby Priory, late eleventh century; and Blyth Priory, begun after 1087/8 (Old Sarum, begun 1075/8, is no longer considered a likely example). Only at St. Albans do part of the walls remain. See A.W. Clapham, Romanesque Architecture in England after the Conquest, Oxford 1934, 20-24, 27, Figs. 4, 5, 7. In Normandy, solid choir walls remain at La Trinite, Caen, begun c. 1059-1066, and at Cerisy-la-Foret, c. 1090-1100(?). They may have existed at St -Etienne, Caen, begun c. 1063. 4 According to British Library, Cotton MS. Nero D. ii, f. 134r [133r], the choir was put into use in 1227 ('Introitus in novum chorum Roffensem'): H. Wharton, Anglia Sacra, 2 vols., London, 1691, I, 347. 129 J. PHJLIP McALEER PLATE 1B North choir aisle, south wall (author). 130 ROCHESTER CATHEDRAL PLATE IIA 131 J. PHILIP McALEER certainly there is no trace of the southern one, and there is no evidence that the north wall of the north aisle belongs to the late eleventh century, or the later Romanesque, although it is impossible to be absolutely certain because the interior surface of the wall is completely plastered over. Certainly, the wall shafts, with their moulded bases, shaft-rings and capitals, all are clearly early Gothic, and relate to the new work of the presbytery and minor transept to the east (c. 1180/1200-1214[?]),5 However, the vault they currently support is definitely not of the early thirteenth century, nor are the tracery-filled 'clerestory' windows (Plate IIB). The vault, with its longitudinal and transverse ridge ribs, the intersections of the various ribs defined by bosses, is much later, its date most likely indicated b y the window tracery, which can be described as late-Decorated verging on the Perpendicular.6 The window in the east bay consists of only two lights; those in the other three bays have three lights. In the latter, in the two western windows, the central light is double-cusped with a circular lobe at the top; the side lights are also double-cusped but the apex lobe has ogee curves; between the heads of the lights, the mullions continuing to the main arch are straight, the areas between them and the central arch nearly solid, as each is indented only by a small curved triangle. The tracery pattern of the easternmost window is slightly different, for the central light is shorter than the side ones, so that the dividing mullion of two additional trefoiled arches rises from its apex, making a total of five rather than three cusped heads in this window. From these details, it is evident that at some time no earlier than the first quarter of the fourteenth century the aisle was revaulted and new windows were created at a high level.7 If the present windows and vaults are the result of remodelling, what of the original arrangement? 5 The traditional date for the start of the new enlarged east end is c. 1200; there is no documentary evidence. Work could have begun soon after a fire in 1179. The presbytery is thought to have been complete by 1214 when Bishop Gilbert de Glanville is said to have been buried in it: B.L., Cotton MS. Nero D.ii, f. 128v[127v] (W harton, Anglia Sacra, I, 347); St. J. Hope, op. cit. in note 11, 50. 6 The bosses of the vault are all carved but lacking visible colour they are now difficult to read. Their subjects seem to be (reading from east to west); (1) bear? face; (2) foliage (four fleur-de-lis?); (3) large rose; (4) encircled face with open mouth; (5) feline face; (6) foliate face; (7) human? face, bearded?, hooded?; (8) four unidentified objects; and (9) feline face. 7 It is sometimes claimed (e.g., Tatton-Brown, op. cit. in note 2, 9 [in this respect he was following St. J. Hope, op. cit. in note 56, 226, and op. cit. in note 11, 85)), that the height of the aisle was raised, but this is not true - at least not to any significant degree - as the new vaults spring from the same capitals as the old one. 132 ROCHESTER CATHEDRAL PLATE 118 133 J. PHILIP McALEER Traces of the jambs and arches of earlier windows have, as a matter of fact, been revealed in the first and last bays; a single narrow window centred in the west bay, a pair of windows filling most of the width of the east bay (Plates IIB, IIIA). They extended in height from the level of the shaft-rings to the level of the capitals (of the shafts) or the springing of. the (later) vault. There is no evidence of windows in the two middle bays. The location and disparate arrangement of the windows is suggestive. Outside the north wall, at a short distance from it, is the formidable mass of the free-standing tower now known as Gundulf's Tower.8 Its position and proximity vis-a-vis the aisle bays means it would have blocked most of the light reaching any windows in the two middle bays, those directly opposite the south wall of the tower. The single west window, however, was opposite the narrow gap between the tower and the north arm of the major transept, while the double eastern window was opposite the wider space between the tower and the north arm of the minor transept. Thus, it seems possible the centre bays were left blind, as little light would have come through windows in these bays, while the windows were doubled in the east bay in order to take advantage of the greater availability of light there. Originally, the aisle was perhaps better lighted than now, for those reasonably large lancet windows were blocked up and replaced by the rather small windows high up in the wall, as already described. Admittedly, the location of these 'new' windows makes sense, but why were they not made wider, and, most especially, why were their sills not extended lower down? The reasons for the size of the windows may relate to the reasons the aisle was revaulted. The evidence for the answers to these questions is, to some extent, found in the outer face of the wall. The exterior face of the aisle wall now forms the south wall of the space between it and the north tower which has been roofed over, the existing roof of a rather flat pitch (draining to the east) at a level just below the upper aisle window sills (Plate IIIB). This outer face is a rough rubble surface without any architectural features except a massive pier buttress (68 in. deep by 61½ in. wide) built against it, about midway in its length. The space is now mostly filled by a wooden stair, entered from the small portal in the third bay, rising to the roof; there is no floor, merely an accumulation of rubble debris which fills the space to a height somewhat higher than the flight of stairs between s According to Fairweather, op. cit. in note 1, 208, the space is about 8 ft., north to south. 9 That is for some 7 or 8 ft.: see Fairweather, op. cit. in note 1, Fig. l and 210. 134 ROCHESTER CATHEDRAL PLATE UIA North choir aisle, north wall. east bay. doorway and former tomb recess (author). 135 J. PHILIP McALEER PLATE 11IB Nonh choir aisle wall. exterior face. two west bays at clerestory level (aurltor). 136 ROCHESTER CATHEDRAL the two levels of the aisle.9 This debris slopes up higher at the west end of the space, rising against the east wall of the major transept's north arm (Plates IVA, B).10 This side of the wall, and its earlier history, have been briefly written about by W.H. St. J. Hope, and more extensively examined by P.H. Fairweather.11 St. J. Hope wrote: The lower part of the outer wall of the north aisle of the quire, if it be not part of Ernulf's reconstruction, may be one of the works done after the fire of 1137. It has a pair of round-headed windows in each bay, now blocked and only visible externally, and, as I ascertained by excavation, it stands upon the base of Gundulf's wall, from which it differs in plane and thickness. The junction has been opened out, and may be seen in the space between the aisle and the old north tower.' 12 From letters of 20 October and 8 November, 1881, written by St. J. Hope to J.T. Irvine, the location of St. J. Hope's excavation can be precisely identified.13 He was digging outside the western window of the Romanesque crypt, in the area between the massive pier buttress and a thin screen wall east of it. He noted 'that the upper part of the wall [that is, the aisle wall] from the [level of the] crypt roof [vaults] upwards was rebuilt but with the old material and reduced thickness'; and that Gundulf's wall [the wall of the crypt] was wider than the wall above and nearly four feet thick. A third, fragmentary letter shows the upper wall, which he later identified as Ernulf's, as thinner, with its exterior face set back 5-7 in. from the outer face of the crypt wall.14 In a last letter on this subject nearly a year later, 23 September, 1882, St. J. Hope, again writing to Irvine, included a sketch elevation of the north aisle wall on top of Gundulf's crypt wall (they are the eastern two bays of the aisle since they are the ones which rise over the crypt). IS He shows three successive building levels: the lowest, identified as 'Gundulph', with two 'later windows (heads gone)'; the middle, '?Ernulph', with a buttress and, on either side at the top of this section (and axially over the headless windows), two round headed windows; and the uppermost, with two (early) Perpendicular windows, one over 10 According to Fairweather, op. cit. in note 1, 210, 19 ft. 11 St. J. Hope's comments are contained in his The Architectural History of the Cathedral Church and Monastery of St. Andrew at Rochester, London 1900, 35, 59-60, 75-6, 85 (Arch. Cant,. xxiii [1898], 228, 252-3, 268-9, 278); Fairweather's in op. cit. in note 1. 12 St. J. Hope, op. cit. in note 11, 35, and note* (emphasis added). 13 Medway Area Archives Office (hereafter M.A.A.O.), DRc/Emf 77/81 and 77/82 respectively. 14 M.A.A.O., DRc/Emf 77/78. is M.A.A.O., DRc/Emf 77/83. 137 J. PHILIP McALEER PLATE IVA onh choir aisle wall. exterior race. two west bays below clerestory level (au1/ror). 138 ROCHESTER CATHEDRAL PLATE IVB Space between north choir aisle wall and north tower as it now exists (facing west) (cf. Arch. Joum .. lxxxvi [ 1929J. Pl. II) (aurhor). 139 J. PHILIP McALEER each of 'Ernulph's' pairs. Presumably, the two lower windows in 'Gundulph's' walling are those of the crypt, but why St. J. Hope identified them as 'later' is not immediately clear. Both windows are still visible from inside the crypt.16 Their jambs are splayed towards the interior and display one peculiar feature - the straight jambs directly abut the curve of the vault. St. J. Hope observed that the original windows had been reduced in size. 17 Only the exterior of the western window is now accessible. 18 On this side, the head of the window is cut off along an horizontal line, i.e. there is no sign of an arch which was probably eliminated in the course of the later rebuilding of the north aisle wall. 19 Fairweather discussed this area in greater detail, with a plan and sketch view, in his paper disputing St. J. Hope's reconstructed plan of Gundulf's church. He concentrated on the western half of this aisle wall. He mentioned a 'small cut-down portion of the wall of the north choir aisle which appears in the angle between the north transept and the later choir aisle wall, the latter having been built upon it, but upon the later axis, thus leaving the outer portion of the older wall exposed, in a rough condition' .20 It appears as 'M' in his Fig. 1, located in the westernmost bay. Of it he further commented: 'It is broken off with a rough fracture against the buttress at its eastern end and on its north face. Its top is partially capped with small roofing stones ... There is so little exposed, that it is perhaps safer not to theorise on its functions' .21 Fairweather's 'outcrop' of early wall ('M') is clearly not 16 The windows are not easily visible, for the north aisle of the crypt is mostly filled by the bellows mechanism of the organ. , 17 St. J. Hope's sketch plan (DRc/Emf 77/78) indicates wider inner jambs (10¾ in. on west, 9 in. on east) identified as 'Gundulph's'. Three of the four jambs are preserved; both windows retain remains of a rebate for a rectangular (wooden?) shutter. Below the level of the bricked-up sill of the western window, there is the line of a rubble-constructed joint on each side. At the east window, there is the line of an earlier jamb of quoins east of the later (damaged) east jamb. (Only the west jamb of the east window is preserved). Hope attributed the reduction in width to Ernulf's rebuilding; but the absence of diagonal tooling on the inserted jamb quoins suggests that the reduction dates to the early Gothic rebuilding of the east end. 1s The exterior of the west window is only seen from the top of a shaft (the result of St. J. Hope's excavation) in a closet-like area now accessible through a door opening off the 'service area' between the east wall of Gundulf's Tower and the west wall of the north arm of the minor transept. On DRc/Emf 77 /78, St. J. Hope noted: 'The window is later. I cd not trace G's opening at all outside owing to a beautiful coat of plaster'. 19 This may be due to the fact that the ground level had risen to cover some of the crypt window by the time the outer side of the aisle wall was refaced (in the fourteenth century). 20 Fairweather, op. cit. in note l , 208. 21 Fairweather, op. cit. in note l, 208. 140 ROCHES TER CATHEDRAL the same as that which St. J. Hope identified as Gundulf's wall, as the former is located in the westernmost bay, and, as seen, St. J. Hope's excavation was in the second bay from the east. In any case, Fairweather's wall 'M' cannot be identified today as a distinct entity in the mass of rubble still in situ (Plate IVA). St. J. Hope had identified the 'lower half' of the aisle wall above the crypt as Ernulf's work or possibly post-1137. Fairweather identified the wall as work perhaps planned by Ernulf but carried out only after the fire of 1137, 'clothed with Early English work' .22 No evidence for these dates was adduced on the basis of the type of stone used, the mortar, or the constructional method. There is no reason not to think that the north aisle wall is not basically of c. 1200 or of the early thirteenth century, contemporary with the new work at the east end (presbytery and minor transept) and the remodelling of the choir (complete by 1227?). The divergence of axis, which St. J. Hope said he observed between the first (west) bay of the north aisle wall of Gundulf's crypt and the wall built on it need not detain us because, if the Romanesque aisle wall - whether of the late eleventh or early twelfth century - was torn down and rebuilt by the Gothic builders, a slight lack of alignment is not surprising or unexpected.23 It is also to be expected that, regardless of date, a wall above the crypt might well be different in thickness - presumably thinner - than the crypt wall itself. An examination of the entire exterior surface of the wall reveals no trace of the windows whose jambs have been revealed on the interior, nor of any earlier windows such as the two round-headed windows shown by St. J. Hope in his sketch.24 The reason for the invisibility of the former is easier to provide than for the latter. In the fourteenth century, there seem to have been structural problems with the north choir aisle wall. This seems to be the only reason which would explain why its windows were blocked up and the exterior of the wall completely refaced. The facing of the wall with rough rubble is carried out uniformly from top to bottom (so far as is now visible) and end to end. There is no question that it abuts the original buttress - located at the western end of bay two - which was at some later date mostly torn out. In addition, the massive pier buttress (mentioned earlier) was built 22 Fairweather, op. cit. in note 1, 202-3, 205. 23 The difference in alignment between the exterior face of the crypt wall and the exterior of the upper wall (shown in M.A.A.O., DRc/Emf 77/78), the construction of which seems to have eliminated the arched head of the crypt window, can be seen in the shaft (mentioned in n. 18) outside the window. 24 In his letter of 23 September, 1882, to Irvine, M.A.A.O., DRc/Emf 77/83. 141 J. PHILIP McALEER PLATE VA Space between tower and north aisle wall at clerestory level, facing east (author). 142 co > Ul 􀀃 .J Cl. ROCHESTER CATHEDRAL 143 I I ' 􀀄 c:, -t: :i s "' ., " bi) C: 1 ·u 􀀅 􀀆 ·ea ... ·5 .,: u 􀀇 0 C: ... " ;> 0 "' 􀀈 :5 .0 bi) C: ·,;, fi: J. PHILIP McALEER exactly in the middle of the wall, with a flyer rising to the choir clerestory, and a pointed arch placed between it and the south wall of Gundulf's tower (Plate VA, B). (This new buttress and flyer are the counterpart to the earlier pier buttress placed directly against the south choir wall.)25 Again, there is no doubt that the facing is contemporary with this large pier buttress (located to the east of the original), or that this buttress was built with the facing (Plates VIA, B).The facing is, of course, dated by the tracery of the windows in its upper half, and the vault behind it,26 As neither St. J. Hope nor Fairweather recognized the fourteenthcentury refacing of the wall, they mistook it either for the work of 'Ernulph's reconstruction' (St. J. Hope) or work done after the 1137 fire (St. J. Hope, Fairweather), or work due to the early Gothic rebuilding (Fairweather). The fourteenth-century refacing could also explain why the wall differs in plane and thickness from the wall of Gundulf's crypt,27 but of course it c a n not account for the 'disappearance' of the outline of pairs of round-headed windows St. J. Hope reported seeing in the two east bays.28 At the same time the wall was refaced, for it appears to course in with the facing, a thin wall was erected between the aisle wall and a large pier buttress, which had been earlier erected against the south face of the south-east angle buttress of Gundulf's Tower (Plate VIIA),29 thus effectively creating a narrow space between the tower and aisle separate from the larger courtyard-like area between the north side of the tower and the west wall of the minor transept's north arm. To return to the interior of the aisle: as mentioned, there are now 25 See the plan, St. J. Hope, op. cit. in note 11, Pl. II. This massive pier buttress, which partly blocks access to the crypt stairs, because of its decorative shafts and niches, is certainly of the (later) thirteenth-century. St. J. Hope, 54, thought it and the buttress and flyer on the north were original to the construction of the choir vaults. 26 Small oblong lights were constructed at the same time to light or ventilate the roof space over the vaults (visible in Plate IIIB). 27 In any case, Fairweather, op. cit. in note 1, 202-3, 204, dated the crypt to Ernulf or later, and not to Gundulf. At the same time, while citing (202-3) St. J. Hope with respect to the divergence of the line or axis between the crypt wall and the choir wall above it (which he then ignored), he claimed a divergence of axis 'takes place at the second nave pier from the crossing on the south side' and consequently affects the axis of 'all' the work to the east which is, therefore, all post-Gundulf! 2s Of course, if such windows had existed, they would have disappeared under the fourteenth-century refacing as have the external jambs of the thirteenth-century windows, which are now revealed on the interior. There is nothing now visible to suggest what might have stimulated St. J. Hope to interpret as pairs of windows. 29 This buttress does not appear on any plan: most especially, it is completely absent from St. J. Hope, op. cit. in note 11, Pl. II. 144 ROCHESTER CATHEDRAL PLATE VIA Exterior face of north choir aisle wall between added buttress (at left) and ripped out buttress (at right) (a111hor). 145 J. PHILIP McALEER PLATE VIB Exterior face of north choir ai,le wall. added buttress at right and portal to third bay of ai,le (awlwr). 146 ROCHESTER CATHEDRAL three doorways in its north wall. That in the western bay (65½ x 27¾ in.) has a pointed arch, and plain chamfered jambs with a small roll above each base spur (like the angles of the north-east buttress added to 'Gundulf's Tower) (Plate VIIB). Jambs and arch are constructed of large blocks which are not bonded in with the wall. Furthermore, the jambs rise from the upper surface of the bench at the base of the wall rather than breaking through it. The second doorway is just above the steps on the east side of the stone screen which divides the aisle (bay three; Plate VIIIA). It is a tall narrow pointed arch (73¾ x 21½ in.) with chamfered jambs, constructed of much smaller blocks than the western doorway. It now gives access to the space between the tower and aisle. The third doorway is at the east end of the aisle (bay four) under a large, steeply-pointed chamfered arch which nearly spans the width of the bay and was probably constructed as a tomb recess (Plate IIIA). It also is tall and narrow (74 X 22½ in.), with the jambs and its semicircular arch simply chamfered. The higher rere-arch is also semicircular. This doorway gives access to the space between the minor transept's north arm and the tower. It may be asked if any of these portals is original to the aisle and what the purpose of each was. The portal at the top of the stairs and that at the west end of the aisle most likely are fourteenth-century insertions, made at the time the exterior of the wall was refaced. This is certainly true of the eastern one, for the east face of the massive new pier buttress forms its western jamb.30 That the jambs of the west doorway rise from the thirteenthcentury bench and that the jamb and arch stones have a continuous outer joint paralleling the arch opening, rather than being coursed into the wall, suggest it, too, is a later insertion.31 This may be further supported by the form of the 'exterior' jamb which is splayed and does not bond with the inner door frame, implying the inner and outer frames were built separately, as would be appropriate for a portal cut through a wall (Plates VIIB, VIIIB). If both portals - those in the first and third bays - belong to the same period, it might now be asked why? One can only speculate that, because of the difference in level between them, they served two different functions, to provide access to two different areas. The smaller, narrower, eastern one, at the upper level, might have given on to a set of steps which rose up against the south 30 A re-used stone, formerly a base and part of the shaft of one of the west front niches, is embedded in the east face of the buttress. 31 St. J. Hope, op. cit. in note 11, 100; 'The present doorway is modern, but replaces an early one'. 147 J. PHILIP McALEER PLATE VIIA Buttress and screen wall between tower and north aisle wall. viewed from west (author). 148 ROCHESTER CATHEDRAL PLATE VIIB North choir aisle, north wall. west bay: doorway (author). 149 J. PHILIP McALEER PLATE YIIIA North choir aisle. north wall. third bay from west: doorway (author). 150 ROCHESTER CATHEDRAL PLATE VIIIB Portal and corridor from north aisle. viewed from tower (awhor). 151 J. PHILIP McALEER wall of the tower in order to permit access to the north aisle roof ( or entrance into the north tower via a window in its south wall?). The western portal, at the lower level, would necessarily open into the space at ground level between tower and aisle, an area which had become separated from that between tower and transept, partially due to the construction of the new pier buttress against the middle of the aisle wall but, more effectively, as a result of the erection of the thin wall extending north from the aisle wall to the buttress at the south-east corner of the tower. This area (between tower and aisle) may have been floored and/or roofed at this time, making it into a 'usable' space. That this entire space was at some time subdivided by a floor and covered by a roof earlier than the existing one is evidenced by raggles of two different roof pitches which remain between the tower and the aisle roof (Plate VA), and several corbels lower down in the aisle wall.32 They suggest several changes in floor levels and roof slopes, although the sequence is not clear. It was probably during this period that the south window of the lower storey of the tower was blocked up and a new window-like opening was broken through the wall immediately above it, but off-axis towards the west (the arched head of the lower window was perhaps eliminated in the process, if that window had not already been extended upwards to form a doorway at an earlier stage) (Plate IXA).33 The third doorway may be best explained in conjunction with yet another, but more elaborate, portal in the south bay of the west wall of the minor transept arm. The latter is now blocked up and only the north half is preserved. It is composed of two thin-coursed jamb shafts with a moulded bell capital; no base remains. The arch profile consists of a filleted roll, two deep hollows, and an outer roll and quarter hollow. This portal may be regarded as an early insertion ( of the second half of the thirteenth-century?), perhaps necessitated by the flow of pilgrims to the tomb shrine of St. William of Perth (murdered 1201; canonized 1256) located on the opposite side of the arm .34 As the portal opened to 32 The pier buttress and the thin 'screen' wall are both pierced by well-made rectangular tunnels as if to accommodate a gutter draining the roof of steeper pitch. The upper nearly flat raggle corresponds to a roof still in position in the 1890s: see St. J. Hope, op. cit. in note 11, 100. 33 This opening is now glazed and serves as an exit from the (restored) floor level of the upper chamber to the 'modern' roof (probably due to the restoration of the tower in 1925 - see below nn. 60, 61) over the space between tower and aisle. 34 Its insertion may have been motivated by a desire to provide pilgrims to the tomb with a quick exit, one not requiring them to return the way they had entei;ed - via .the north portal of the major transept north arm and along the north aisle - thus maintaining a smooth 'one-way' traffic flow. St. J. Hope, op. cit. in note 11, 127, 131, thought it 'of early-Decorated date' (end of thirteenth century). 152 ROCHESTER CATHEDRAL PLATE IXA (Photo: Rochester, Guildhall) Interior of north tower: lower west half of south wall before restoration. 153 J. PHILIP Mc ALEER the space between the north tower and the transept arm, it initially could have provided a direct exit (or entrance?) from the shrine area to the outside. Eventually, however, the space between the tower and transept was converted into an enclosed area, possibly for use as a 'sacrist's checker'. An 'L' plan wall was built between the turret at the northwest corner of the transept arm and the north-east corner of the tower. This wall has a small two-light window with a transom and trefoil arches suggesting a date in the fourteenth century.35 It effectively meant this area could thereafter only .be approached from inside the church. Whether or not the enclosure of this space was motivated by a need for an area to function as a 'sacrist's checker' in connection with the tomb shrine may be open to doubt, for its identification as such is based upon slight and circumstantial evidence.36 At some later date, a floor was inserted, the north-west newel-stair of the minor transept arm becoming the mode of access from one level to another. The original entrance from the transept into the vice was then blocked. A new entrance was made in the south face of the turret facing into the area, and a seco11d opening was made higher up to exit onto the first floor. These changes were probably made prior to the Dissolution, for both the arches inserted into the vice are pointed. Following the Dissolution, there is documentary evidence that this area functioned as lodgings, as would seem to be implied or confirmed by the record of there having been two chimneys in this area.37 The narrow doorway under the (former) tomb recess in the north aisle, however, could only have appeared after the 35 St. J. Hope, op. cit. in note 11, 131, thought this 'domestic-looking' window 'contemporary with the canopy over the tomb of Bishop John of Bradfield (1278-1283)'. 36 S. Denne, 'Memorials of the Cathedral Church of Rochester', in (Ed.) J. Thorpe, Custumale Roffense, from the original manuscript in the archives of the Dean and Chapter of Rochester, London 1788, 174, was the first to so identify this area. St. J. Hope, op. cit. in note 56, 226, accepted this identification ('chambers ... which pertained to the sacrist and other officers') and dated the portal inserted in the west wall of the transept to the early fourteenth century; cf. St. J. Hope, op. cit. in note 11, 131. Denne made this association because of an 'Elizabethian' lease which suggested to him that 'the sextry garden and well . .. were situated contiguous to them'. I could not locate this otherwise unidentified lease which, presumably, should be contained in M.A. A.O., DRc/Elb 1, 'The Booke off Register made by Martyn Cotes off All the Writings leases and other Munyments which have and shall passe under seale of the Dean and Chapter of Rochester from and after the date of the Patente thereof ...' . This is the fifth volume of the registers begun at the time of the refoundation of the cathedral in 1541 and covers the years 1575-1627 (volumes two, three and four do not survive). Conceivably, the lease seen by Denne was in volume four which presumably covered the early years of Elizabeth's reign (1558-1603). 37 To be discussed below.S ee n.4 1. 154 ROCHESTER CATHEDRAL Dissolution and the destruction of the tomb contained in the recess. It was perhaps inserted at that time in order to give direct access to the apartment in its new domestic function. With its creation, the transept portal was blocked up. It would seem that it was also at some time after the Dissolution that the two western fourteenth-century portals were blocked up, for they appear to have been replaced by a rectangular portal with a wooden lintel opening off the aisle stairs. The lintel of this fourth doorway is still visible from outside, but all trace of it has been removed on the aisle side. The creation of this rude rectangular portal, presumably accompanied by the blocking of the fourteenth-century portals to east and west, suggests that the space between the tower and the aisle wall had lost or changed its function. Certainly by the later eighteenth century, judging from the earliest plan of the cathedral (1788),38 the floors had disappeared and the space between the tower and aisle wall was filled by flights of wooden steps which must have given access to the choir aisle roof, much as those now in this (roofed-over) space do today. Access to the wooden steps was at that time gained through the rectangular doorway with the wooden lintel opening off the stairs in the aisle. Its insertion may be associated with the construction of a loft or gallery over the choir steps from which the bells in the central tower were rung. The date of the construction of this loft or gallery is not known, but it was pulled down in 1730.39 This curious arrangement was necessary because, when the major transept and crossing were rebuilt in the course of the thirteenth century, no direct, convenient means of access to the crossing tower was provided: there are no stair vices in the angles of the major transept arms, or above the crossing piers. Nor did the early Gothic builders make any connection between their work in the choir aisles or the north arm of the major transept and the north tower, or even make any provision for temporary arrangements. This is all the more mysterious because, during the later Middle Ages, the north tower apparently served as a bell tower in conjunction with, and as a supplement to, the central tower.40 Its function as a belfry would also seem to have been 38 The earliest plan appears to be that in Thorpe's Custumale (op. cit. in note 36), Pl. XL. 39 Denne, op. cit. in note 34, 182-3. 40 This is not the place to enter into the controversy about the date and function of the north tower. Suffice it to say that St. J. Hope, op. cit. in note 11, 8-10, because he thought the Romanesque church lacked a crossing, identified it as the 'major turris' in which bells were placed as early as the mid-twelfth century. (Tatton-Brown, op. cit. in note 2, 7, accepted St. J. Hope's thesis that the 'major tower' in which bells were placed by Prior Reginald [c. 1150-60] was the north tower, but he failed to take into account St. J. Hope's 155 J. PHILIP McALEER inhibited not only by the absence of a direct connection with the church but by the fact there certainly were no portals in its west, south or east faces, that is, the sides facing the church. Yet, sixteenth-century sources suggest that access to the crossing tower was in some way achieved via the north tower and the areas adjacent to the north aisle and minor transept arm. From a lease granted by the Dean and Chapter on 7 April, 1545, to one Nicholas Arnold, priest, we learn of 'lodgings which was sometimes called the wax chandler's chambers, together with the little gallery next adjoining' which included 'all usual ways, that is to say, through the three-bell steeple, some time so-called, and so up to the north side of the church, and so on to the stairs that goeth to the six-bell steeple' .41 The route described in the sixteenth century lease is difficult to reconstruct, for its understanding depends in part on the identification of the 'wax chandler's chambers' and the 'little gallery next adjoining'. Although the former has been identified as the space between tower and aisle,42 it seems more faithful to the text to identify it as the space between the tower and minor transept arm, especially as the two 'chimneys' - fireplaces - on two stories, one with an oven next to it - formerly in evidence in this area, make it more appropriate as the :r;najor chambers of the lodgings. The 'little gallery next adjoining' would then be the long, narrow space between the tower and aisle, a space which, as has been suggested, may have been floored and roofed over in the fourteenth century and where, well before the Dissolution, there may have been stairs leading up to the north side of the church and to other stairs which went up to the central tower or 'six-bell steeple'. If this was the case, then the route described, which places passage reason for making that identification.) When Bishop Hamo of Hythe (1317-52) completed the Gothic central tower (1343?), he placed four bells in it (St. J. Hope, 83- 4). Just when bells were first placed in the free-standing north tower is not clear: the earliest reference to a lesser tower (parva turris), according to St. J. Hope, 10, is in the Custumale Roffense (Thorpe, op. cit. in note 36, 31) which he dated, 94, to the early fourteenth century. Bells may first have been placed in the north tower in the mid to late thirteenth century when the Romanesque (twelfth? century) crossing tower was taken down to allow the construction of the major transept. Upon the rebuilding of the belfry chamber by Bishop Hamo (by 1343), it would have been appropriate to refer to a parva turris (the north tower) in comparison with a major turris (the central tower). Cf. Denne, op. cit. in note 36, 172-3 and nn. o, p. 41 As cited by Denne, op. cit. in note 34, 174. The lease in question is found in M.A.A.O., DRc/Elb lA (formerly Rochester Cathedral, Dean and Chapter Library, A.4. 15. ), 'Primum Registrum Ecclesie Cathedralis Roffensis ( 1541-7)', f. 84v. 42 Denne, op. cit. in note 36, 174, 'BB' on plan Pl. XL (facing p. 174); he identified the area, 'AA' as 'apartments' belonging to the sacrist: he did not locate the 'little gallery'. 156 ROCHESTER CATHEDRAL through the 'three-bell steeple' - the north tower - before going up to the north side of the church, is confusing, especially as there is no evidence of ground floor openings into the tower from these areas at this time. In the late eighteenth century, S. Denne assumed this route began with the north portal to the tower - the one according to him 'discovered' by F. Grose.43 However, it lay outside the area between the tower and the minor transept arm, and was unapproachable from that area because of the 'L' plan wall (between the north-west transept turret and the north-east corner of the tower) in which there was no door. Since the north tower was constructed without a stone newel-stair and lacked permanent internal subdivision by stone vaults, its floors must have been of wood supported on corbels (the imprint of a few remain), reached by wooden stairs or ladders. So it was no more convenient than wooden stairs placed in the space between tower a n d aisle. Furthermore, as its windows - narrow slits (four on the ground floor, only two on the first floor) - had their sills well above floor level, they could have served as entrances only if widened and lengthened. The only window which provides evidence of such treatment is that on the south - which would have, therefore, communicated only between the tower and the space between it and the north aisle wall. A route through the tower and through the south window/doorway would only make sense if it began to the north, outside the tower and the 'wax chandler's chambers' (as heretofore identified). Equally unclear is how, once the north side was gained, the central tower was reached. Stone steps of a quasi-spiral staircase are still in situ against the sliver of vertical wall between the south wall of the shallow chapel of the north arm of the major transept and the north aisle wall (Plate IXB). They obviously would have provided access to the roof of the north aisle from a set of wooden stairs rising to about the same height as those in existence today. But from this point there is no access to the belfry of the central tower. Today, the lower stage of the tower, which on the interior is still thirteenth-century masonry, is reached by one of four small pointed door-ways - one on the axis of each wall - from a path over the crown of the choir vaults, in turn reached by ascending the newel stair in the north-west angle of the north arm of the minor transept, exiting onto the roof (west) gutter and following that to a doorway in the roof. The problem is how in the sixteenth century they got from the north aisle roof up to the choir roof: 43 Denne, op. cit. in note 36, 173-4. The original appearance of this portal is uncertain. Of its present form St. J. Hope, M.A.A.O., DRc/Emf 77/83, remarked 'They have cobbled up the entrance, to G's tower from outside. It is plain work and unpretentious but fortunately undeniably 19th Cent: - It was a necessity', 157 J. PHILIP McALEER PLATE IXB Space between north aisle wall and tower at clerestory level. facing west (author). 158 ROCHESTER CATHEDRAL was there another stair in the angle of the north choir wall and the east wall of the major transept's north arm, which breached the parapet and led to a now-vanished door in the choir or transept roof?44 The loft or gallery placed over the steps of ascent into the choir, located in the existing crossing, from which the bells were rung until 1730, must have been reached by much the same route, except that, once on· the north aisle roof, access would have been gained through a clerestory window and from the wall passage which ends at the north-east crossing pier.45 Just exactly when the north tower lost its function as a campanile is not known. It was certainly by 1772, for the description of the tower found in The History and Antiquities of Rochester makes it clear there were no longer floors inside the tower or between the tower and the aisle wall.46 The uselessness of the tower is earlier evident in a survey of the cathedral by Henry Keene dated 10 October, 1760: he recommended that the 'five-bell tower' (that is, the 'three-bell' or north tower) be taken down gradually and the material re-used.47 He also suggested 'care should be taken to convey away water that falls between the tower and Cathedral which by being so confined between the buildings must greatly damage those parts where it is suffered to lay'. This, too, implies the area between tower and aisle was unroofed. The most likely date for the despoiling of the tower would be the mid seventeenth century. Daniel Alexander in his survey of November 1799 also noted the 'old tower' could supply material for rebuilding the parapets of the 44 The choir roof, lowered in pitch in- the early nineteenth century (c. 1811), due to dry-rot, was extensively repaired by L.N. Cottingham in 1825. See Rochester Cathedral, D. & C.L., 733.3 COT, 'Repairs to Rochester Cathedral 1825-26: Mr. Cottingham Architect' (a typed transcript of the personal notebook of Dean Robert Stevens [1820-70]), pp. 1, 3, 8, 9. Sir George Gilbert Scott, during his restoration, wished to restore the roofs of presbytery, minor transept and choir to their original pitch, and for this purpose rebuilt some of the gables, but the roofs have never been raised. TattonBrown, op. cit. in note 2, 11, made reference to a '(now blocked) stair above the east wall of the transept': by this did he mean the quasi-spiral stair? He went on to say: 'This would have run up on to the north choir aisle roof, and then on to a door in the crossing tower (Note: Destroyed when the central tower was rebuilt in 1826.)': thus, he begged the question of how one would get from the level of the north choir aisle roof up to the level of the chamber over the crossing. Actually the lower stage of the tower, abutted by the roofs, was only recased on the exterior by L.N . Cottingham in 1826. 45 Denne, op. cit. in note 36, 183: '[the] entrance into the church was by the window next to the aisle leading into St. William's chapel'. 46 [Wm. Shrubsole and S. Denne,] The History and Antiquities of Rochester and its Environs, Rochester 1772, 73. 47 M.A.A.O., DRc/Emf 34, 'A Survey of the State and Condition of the Buildings, of the Cathedral Church of Rochester'. 159 J. PHILIP McALEER 'north cross aisle' (north major arm).48 What was then demolished was not so much any of the eleventh-century fabric of the tower but primarily the corbelled arches supporting a parapet, shown in the engraving of F. Grose,49 which had been added, it appears, in the thirteenth century. The demolition of the upper level with the corbelled arches, therefore, could have begun any time after 1760; perhaps, however, most of the destruction was accomplished shortly after 1799. It is most likely that it was at that time, as a result of the process of demolishing the uppermost remaining part of the tower, that the area between the tower and the north aisle was filled with blocks of Kentish rag rubble which could not easily be removed because of the wall thrown across the space at the east end. As the rubble accumulated to a height of 15 ft., sloping up towards the east wall of the chapel of the north arm of the major transept, it blocked the western-most aisle door. It also completely blocked the rectangular door ( of which the lintel is now visible just above the upper surface of the debris) (Plate VIA). At some date, belatedly following the advice of Henry Keene, the upper surface of the rubble was sloped off (in the manner described by Fairweather) in order to drain the water away. The north tower is now entered from the north aisle by a tunnel of double curvature between the western doorway of the aisle and a portal cut through the west end of the tower's south wall. It is the debris, filling the space between tower and aisle, which is tunnelled through to form this passageway. There is little clue as to the date of the portal to which it leads, except for the use of (y ellow?) brick for the two arches, which define the rubble barrel vault in the thickness of the tower wall: one arch - the outer (towards the aisle) - is semicircular, the other (inner) pointed (Plates X A, B).SO It is decidedly puzzling that neither portal nor tunnel appear on the earliest plans of the cathedral. The plan of 1788 from the Custumale shows only one doorway in the north aisle, that over the stairs: it gave 48 M.A.A.O., DRc/Emf 38, 'Report on a General Survey of Rochester Cathedral', section 5. It is interesting that neither Keene nor Alexander referred to it as 'Gundulf's' tower. 49 F. Grose, The Antiquities of England and Wales, 6 vols., London 1733-87, V (Supplement, I, 1777). so Neither portal nor barrel vault appear in two sketch plans of the tower made by Irvine, presumably when he was working on the restoration of the cathedral under the direction of Sir G.G. Scott in the 1870s. Instead, he shows a kind of recess in this location. See M.A.A.O., DRc/Emf 77/35, and three sheets in DRc/Emf 77/134. St. J. Hope, op. cit. in note 11, 8, referred to this 'recess' without any additional specifics: 'The tower is now entered by a large opening broken through the north wall, and by a door in the south-west corner made by knocking out the back of an original recess there'. 160 ROCHESTER CATHEDRAL PLATE XA South-west entrance to tower. facing north (awhor). 161 J. PHILIP McALEER PLATE XB South-west entrance to tower. facing south (a111hor). 162 ROCHESTER CATHEDRAL on to flights of stairs in the space between tower and aisle (the steps led up to a landing placed along the tower wall followed by a left turn onto a landing against the aisle wall and a further flight against it).51 Rather surprisingly, a number of engraved plans of the cathedral, which date between 1788 and 1838, suggest that the rectangular portal continued in use, giving on to a flight of steps leading up to the north aisle roof and to the belfry, after this date, although one would expect the stairs to have been damaged from falling debris when the corbelled arches were being pulled down and the portal to have been blocked. Yet, J. Coney's plan of 1816 indicates only one aisle door, the same one, leading to flights of stairs in the same general configuration.52 Perhaps, Coney simply copied these details from the Custumale plan, and his plan does not therefore depict the 'reality' of 1816. The same plan shows the portal in the north arm of the minor transept open to the courtyard, as does the Custumale plan. The plan from Winkles '.s Cathedral Churches of 1838 indicates a somewhat different situation:53 the same door (at the foot of the aisle steps) gave access to a single flight rising to a landing against the tower wall from which additional flights departed east and west. By this date, one could hypothesize that the portal had been unblocked and a new wooden stair constructed. There is some slight evidence that this was the case for, in the debris now rising against the wall at this point, there are some rough blocks positioned as if they once formed retaining walls to hold back the rubble from a passage rising from the wooden-lintelled doorway. In Winkles's plan, the portal in the east aisle bay giving access to the court (L = yard) first appears; the transept portal is now shown as blocked. The tunnel between aisle and tower, and the portals at either end of it, are first shown on the plan made by St. J. Hope c. 1881/2. If the debris through which it cuts only accumulated c. 1800, one might wonder why the tunnel was made, considering that the inside of the tower w a s useless, it being unfloored a n d open to t h e air?54 Nevertheless, by the end of the last century, the tunnel was in existence, with the space between the tower and aisle wall roofed over at the level 51 Thorpe, op. cit. in note 36, Pl. XL (facing p. 174). 52 J. Caley, H. Ellis, and B. Bandinel, Dugdale's Monasticon Anglicanum, 6 vols., London 1817-30, I, plan following p. 152 (dated 1825). Another plan, drawn by W. Espin, Junr., published 1 March, 1816, by Sherwood, Neely and Jones, from J. Storer, History and Antiquities of the Cathedral Churches of Great Britain, 4 vols., London 1814-1'9, IV (1819), following Rochester Pl. ·8, is quite similar .• 53 (H. and B. Winkles) W inkles's Architectural and Picturesque Illustrations of the Cathedral Churches of England and Wales, 3 vols., I, London 1838, plan between 106/7. 54 The tower was not returned to use until 1925. 163 J. PHILIP McALEER of the aisle parapet, and the straight-headed doorway leading to a flight of steps going up to the aisle roof.55 The interior of the north aisle assumed its present appearance only at the end of the first decade of this century. The first plan which includes all the existing doorways is that of St. J. Hope, first published in 1885;56 on the plan, they are identified as 'Post-Suppression and Modern'. Some indication of how the present situation was arrived at is contained in documents relating to the 'restoration' of the aisle in 1910 by C. Hodgson Fowler.57 From them we learn that the plaster was to be removed from the interior of the aisle wall, and the wall carefully examined for traces of old painting before being replastered. It was in the course of this operation that the jambs of the lancet windows in the western and eastern bays were discovered and left visible after the replastering. It was also at this time that the doorway with the wooden lintel was filled in (and plastered over) and the built up doorway just east of the cross wall was opened out.58 New wooden stairs were built - 'the outside of the south wall of Gundulf's Tower [ was] to be carefully brushed down and any defective pointing made good; and all the holes made by the present stairs [were] to be carefully filled up with stones' - and the space roofed over. The western and easternmost doorways are not mentioned, but they appear in the several plans and elevations, including one showing designs for three wooden doors with Gothic tracery panels.59 The north tower was not returned to use until the mid 192Os.60 At 55 St. J. Hope, op. cit. in note 11, 100. Cf. G.H. Palmer, The Cathedral Church of Rochester: A description of its fabric and a brief history of the episcopal see, London 2nd edn, 1899, 81: 'The westernmost door in the north wall gives access to Gundulf's tower, the easternmost now leads to the belfry, as the middle one used to do'. 56 W.H. St. J. Hope, 'Notes on the Architecturnl I li,tory or Rochester Cathedral Church', Transactions of St. Paul's Ecclesiological Society, l O 88 l-85 ), 216/217, published again in 'Rochester Cathedral Church', The Builder, lxi, no. 2539 (3 October 1891), 259-61, plan after p. 268 (at the end of the issue), and once more in op. cit. in note 11, as Pl. II. 57 M.A.A.O., DRc/Emf 148, 'Specification of Works required to be done in Alterations to the North Aisle, etc. of the Choir of Rochester Cathedral' (22 pp. plus plans, elevations and sections), July 1910, 8-10, and a letter of 20 October, 1910 (from Fowler to R.A. Arnold). For plans and elevations also see DRc/Emf 71/8. 58 St. J. Hope, op. cit. in note 11, 132: 'A little further west is a small doorway now blocked, that gave access to a small open court between the chambers just described and the wax-house on the other side of the buttress'. 59 M.A.A.O., DRc/Emf 77/8. St. J. Hope, op. cit. in note 11, 132, referred to it as a 'modern' door, 'built within the arch of a much larger one'. 60 A plaque at the base of the tower (just inside the entrance from the north aisle) gives the year as 1925 and provides the additional information that most of the funds for the repairs were provided by the Freemasons of Kent. Work was probably completed in that year, as one of the few and scant references to the work in the Cathedral records reveals that scaffolding was put up as early as 1919/20: M.A.A.O., DRc/Ftb 246 (Treasurer's 164 ROCHESTER CATHEDRAL that time, the decaying masonry was thoroughly restored, the interior, for the most part, being relined in brick, as well as a floor inserted and a roof put on.61 The tunnel from the west doorway in the north aisle became the usual means of access; the doorway at the east end of the tower's north wall formed a back door or service entrance, while the doorway at the top of the stairs in the north aisle continued to provide access, via a set of wooden stairs, to the roofs of the north aisle and the tower itself. In short, the existing arrangements came into being and have been essentially unaltered since then. Ledger, 1919-1925), 8 February (payment to A. Osborne for erecting and hire of scaffolding, £19 4s. Od.). 61 An account of the condition of the tower at the time, and of the repairs, is found in an unsigned, undated copy of a typed report: M.A.A.O., DRc/Emf 77/144. 165