The Church of St Peter Canterbury

THE CHURCH OF ST. PETER, CANTERBURY By TERENCE PAUL SMITH, M.A. With an Appendix by S. E. Rigold, M.A., F.S.A., F.R.Hist.S. THE church of St. Peter at Canterbury does not appear in Baldwin Brown's hst of Anglo-Saxon churches,1 nor is it mentioned by Clapham,2 and does not appear in the catalogue which forms the bulk of H. M. and the late Joan Taylor's magnificent work.3 E. A. Fisher does not mention it.4 The possibility that the tower of the church might be of Anglo-Saxon date must have been considered by Canon R. U. Potts when he wrote that it is 'possibly post conquest'.5 In the present paper, I shall argue that the earhest work in the church belongs to the 'Saxo- Norman overlap', that is to the period between c. 1075 and c. 1115, when Anglo-Saxon building traditions were stiU strong in the smaller churches and were relatively uninfluenced by the more cosmopohtan forms of Romanesque. Mr. S. E. Rigold has added an Appendix on the later work in the church. DESCRIPTION OE THE SAXO-NORMAN WORK The church stands at the corner of St. Peter's Lane and St. Peter's Street, set back a httle from the latter (N.G.R. TR 147580). The Saxo- Norman work comprises: the lower part of the tower, at the southwest corner of the church; one bay—including both piers—at the western end of the north arcade; and, possibly, a short stretch of waning in the south wall, immediately east of the tower. The tower is shghtly wider than the south aisle (Fig. 1). The nave west waU, which is later, runs obhquely to the tower, thus necessitating some alteration to the latter at the time when this west waU was built. The tower rises sheer with no off-sets or string-courses to mark externally the stages into which it is divided internaUy; only the lower two stages 1 G. Baldwin Brown, The Arts in Early England, 2, Anglo-Saxon Architecture, London, 2nded., 1926. 2 A. W. Clapham, English Romanesque Architecture Before the Conquest, Oxford, 1930. 3 H. M. and Joan Taylor, Anglo-Saxon Architecture, 2 vols., Cambridge, 1965, pagination oonseoutive. 4 E. A. Eisher, Anglo-Saxon Towers, Newton Abbot, 1969. 0 R. U. Potts, 'St. Mildred's Church, Canterbury: Further notes on the site', Aroh. Oant., lvi (1943), 22. 99 T. P. SMITH '/ V? TpWER V" B *"' rrrr uzrms :•> «r' » i i n IT 16 HAVE U // ,// IS a t 0 r /T ?/> a i Stones SOUTH AISLE l ST. PETER'S, CANTERBURY I SOUTW-IVOT CORNER W& SURVIVING SAXO-NORMAN 1 EiSi CONJECTURAL SAXO-NORMAU I I i^rax Eia. 1. Plan of South-West Comer of Churoh. are of Anglo-Saxon character.6 This is displayed not only in the general form of the tower, but also in the large-stone quoins at its south-east and south-west angles and at its north-west angle towards the top of the second stage, the lower part of this angle having been destroyed (at the very bottom) or altered (further up). Although not so large as the megalithic quoin-stones in the south-west angle of St. Mildred s in Canterbury7 these stones are still somewhat larger than one would expect in a building erected both after the conquest and by a Normantrained master. The average size of the stones forming these quoins is about 1 ft. by 1 ft. by almost 2 ft., and they are arranged in sidealternate manner, which, though hardly proof, is a strong indication of Anglo-Saxon work. These quoins contrast with the rest of the tower fabric, which is of close-set cut flints; there are also a few re-used Roman tiles in the fabric of the tower, but far fewer than in the west wall. 0 In this paper the ground-floor stage will be referred to as the first stage, the first-floor stage as the second stage, etc. It may be that the upper stage, beneath the parapet, is also Saxo-Norman but with renewed quoins and windows. Of this, however, it is impossible to be certain. 7 R. Hussey, 'St. Mildred's, Canterbury', Arch. Oant., i (1858), 143-6; Taylor and Taylor, op. cit., 146-6; photograph at fig. 419. 100 THE CHURCH OE ST. PETER, CANTERBURY The thickness of the tower waUs varies: north wall 1 ft. 9 in.; east wall 2 ft. 2 in.; south wall 3 ft.; west wall 3 ft. 5 in. At the second stage, the walls are a httle thinner, this reduction being achieved by setting back the walls internaUy, thus providing a narrow ledge which could be used for supporting the floor beams. In plan the tower is sub-rectangular, its east-west axis being slightly greater than its north-south axis; the measurements are 12 ft. 2 in. (along the south waU) by 11 ft. 6 in. (along the west waU). The walls themselves are somewhat irregular, the west wall in particular tapering markedly internally from bottom to top. The two lower stages of the tower have no fenestration in the north and south walls, and there is no fenestration at all in the second stage. In the first stage, in the east wall, is a tower-arch opening from the south aisle of the church. This arch has j ambs of a single square order cut straight through the wall; there are no bases, and the imposts are of simple square type with a plain chamfer on the lower arris; the imposts are returned along the east and west walls of the tower for a distance of about 1 ft. The southern impost is a single through-stone, but its feUow to the north is constructed from two stones. The arch now supported by these imposts is pointed, but—if an Anglo-Saxon character for the jambs and imposts be correct—this arch must be a later alteration. The archway is 3 ft. 7 in. wide by some 7 ft. tall to the level of the imposts, and these latter are 5 in. deep. The stones of the jambs are largish; the plaster has been stripped to reveal these, but the vertical hmits of this stripping do not correspond to those of the stones themselves and it may be that some of them are through-stones. From the foot of the north jamb a large flat, but irregular, stone projects eastwards for 1 ft. (PI. IIA). In the west wall of the tower, also at first-stage level, is a small, single-splayed window. The window surround itself is manifestly not primary, but that it may be a later alteration is suggested by the fact that the stone from which it is constructed is different from any in the rest of the lower two storeys of the tower. The reveals and siU, however, are of wholly different character, being much more crudely constructed. The head is flat and timber-hnteUed. The reveals and sill are deeply splayed to the interior (Fig. 2), and the reveals are also curved in plan. The north reveal, while curving only slightly in plan, is also slightly curved in the vertical plane. At the exterior face the window is 1 ft. 2 in. wide, splayed to 3 ft. 9 in. at the interior wall-face. The siU on the interior is some 5 ft. above the present floor. The north waU of the tower continues eastwards for 1 ft. 1 in. in the form of the respond of the later medieval arcade, but that the original arcade projected in this way also is shown by the large projecting stone at the base of the respond; this stone is rather irregular in shape. Its width of 2 ft. 1 in. corresponds closely to that of the north 101 i i T. P. SMITH FEET METRE EIG. 2. Plan of Window in West Wall of Tower. arcade (2 ft.), and the east face of this stone ahgns with the east face of the western pier of the north arcade, suggesting that the stone was the bottommost member of a similar pier in the south arcade. As such, it would have been a through-stone. The present south wall of the church abuts onto the east face of the tower close to the south-east quoin (Fig. 1). This waU is now pierced by a thirteenth-century doorway, its west jamb against the tower wall. At this point the wall is only 9 in. thick, but is thicker above and beyond the doorway. The waU leaves the east face of the tower at right-angles to this latter, but at a point 9 ft. 6 in. east of the tower it re-orientates shghtly to the south. The stretch of walling immediately east of the tower is, like the tower itself, of close-set cut flints with a few Roman tiles. The rest of the Saxo-Norman work at St. Peter's is at the western end of the north arcade. Here are two piers of markedly different character and spacing from the rest of the arcade (Fig. 3). An original arch links the two piers. The opening to the west of the western pier has been blocked and the blocking plastered over so that only the south faces of jamb and impost are now visible. The arch springing from the eastern side of the eastern pier forms part of the later arcade. The piers are simple, unchamfered blocks of masonry, the jambs 102 THE CHURCH OF ST. PETER, CANTERBURY 1 1= V 7 y xr.f-r: IT EIQ. 3. Elevation of North Arcade. being of a simple square order. The jambs of each pier rise straight from the floor with no bases, and are capped by simple square imposts with their lower arrises plain-chamfered (Fig. 3 and PI. IIB). These imposts are composed of a number of stones, with no through-stones, and are returned along the whole width of the piers on both the north and south faces. The corners are constructed of fairly large stones, and the plaster has been partly stripped to show these, and (as with the tower-arch) some of them may be through-stones. The ahgnment of the top and bottom faces of these stones across the piers certainly makes this a possibihty. The single remaining arch-head is semi-circular, and, hke the piers, of a single square order, unohamfered. It is composed of well-cut voussoirs radially set, though there is some variation in the size of the voussoirs. The western springing of the arch is not exactly above the corresponding jamb, but is set-back some 2 in., thus making the arch some 2 in. wider than the distance between the jambs. The eastern springing is exactly above the corresponding jamb. Immediately above the arch in the north face of the arcade, and reaching for about 2 ft. to the level of the low aisle roof, is a piece of walling which may weU be original. It is not plastered over, so that its individual stones remain visible. Clinging to the western half only of the extrados is a further 'arch' composed of fiattish stones laid in 103 T. P. SMITH \ X c St.PETER I I I XIII c CANTERBURY XIV c . i & ii VZA xiv C.. I ll XV c + mm Hi mil l j | m ill! t i l l IIII IIII + III WH 0 & » £. 7 z S E 7 ^ 10 40 Ft 10 M Era. 4. Tredington fashion.8 The western half of this piece of walhng is of small rectangular stones fairly closely laid; the eastern half contains some similar stones, but has a preponderance of cut flints in its fabric (PI. IB). There is a very shght depression in the plaster surface almost centrally above the arch in the south face of the arcade, seen best from the south arcade of the nave, in the obhque hght entering through the west window. It may well be that the cause of this depression is a blocked window of an original clerestory. The width of the nave, measured between the south face of the western pier and the north face of the large stone at the base of the respond at the western end of the south arcade, is 12 ft. The distance from the western face of the western pier to the west wall of the nave is 7 ft. 8 in. 8 That is, with the flattish stones forming the arch not used as true voussoirs, but so used that the lowest stones are at a sharp angle to the impost, and the others are arranged more or less parallel to these so that a V-shaped gap results at the top of the arch. 104 i A. Tower at South-West Angle of Church. B. Arch of North Arcade with original Work above, from the North-East. PLATE II i A. Large Stones at the Foot of the Tower and of the South Arcade, from the North-East. *•*• % > X m B. Impost of Pier in North Arcade, from the South-East. THE CHURCH OF ST. PETER, CANTERBURY DISCUSSION On the basis of the above description of the relevant parts of the church, I wish now to discuss the date and nature of this work. The large-stone quoins of the tower suggest an Anglo-Saxon or Saxo-Norman date, as does the thin hard walhng of cut flints which forms the rest of the tower fabric. The tower walls, except the west wall, and the piers of the north arcade are very thin, such a thinness being a characteristic of Saxon walhng. The window in the west wall of the tower also has an Anglo-Saxon or Saxo-Norman look; single-splayed windows occur in both Saxon and Norman contexts,9 so that in itself the single splay is no criterion of date. Nonetheless, the roughly made reveals exhibit that lack of concern for the finer points of architecture which Anglo-Saxon and Saxo-Norman builders sometimes show. This can be seen, for example, in the two windows in the chancel north wall at Thursley (Surrey), dateable to the Saxo-Norman overlap.10 These, although double-, not single-splayed, are rather crudely executed and are deeply splayed. Anglo-Saxon windows often show this sharp splay, for example at St. Stephen's, St. Albans, where there is a very sharply splayed singlesplayed window probably of late tenth-century date.11 The single square order of the tower-arch jambs is also indicative of an Anglo-Saxon or Saxo-Norman date; such jambs often, but certainly not invariably, contain through-stones, and it may be that some of the stones in the jambs at St. Peter's are through-stones; but in an area where good buUding stone was not particularly easily avaUable, the absence of through-stones does not militate against an Anglo-Saxon or Saxo-Norman date. The imposts of the archway, which are simUar to those of the arcade piers, are also a common Anglo-Saxon and Saxo-Norman type, a Kentish example being in the north doorway at Aldington.12 Often such an impost is constructed from a single through-stone, as is that above the south jamb at St. Peter's. To these detaUs of the structure may be added the general form of the tower; its smaU size and slim proportions suggest that it should be regarded as an example of the type of beU-tower best represented in the Lindsey area of Lincolnshire. Towers of this type are of rubble or flint walling rising with no external batter or setting-back, though occasionaUy a string-course may mark externally the internal division into stages, as at Bosham (Sussex).13 Some have walls which thin 0 Taylor and Taylor, op. cit., 9-10. 10 P. M. Johnston, 'Witley and Thursley Churches', Surrey Arch. Coll., xxxix (1931), 104-11; Taylor and Taylor, op. cit., 616-7. 11 T. P. Smith, The Anglo-Saxon Churches of Hertfordshire, Hitchin, forthcoming; Taylor and Taylor, op. cit., ii, 630-1. 12 Ibid., i, 21-2. Other examples are: Clapham (Beds.), Reed (Herts.), and Newton-by-Castleacre (Norfolk). 13 Taylor and Taylor, op. cit., i, 81-4. 105 T. P. SMITH internally towards the top, as here; for example, Clapham (Beds.),14 and St. Michael's, Oxford.15 These towers are not normally less than 11 ft. square internally, and all show the shm proportions exhibited by St. Peter's. A late date is suggested by this interpretation, since belltowers do not occur in England prior to the tenth century.16 A tower of the type to which St. Peter's belongs need not be expected before the end of the tenth century, so that this gives the earhest possible date for the tower. However, a date later than this, in the second half of the eleventh century or even early in the twelfth century, would be consistent with the evidence. It will shortly be argued that the early arcade is ofthis Saxo-Norman date, and in view of the similarity between the tower-arch imposts and those of the arcade piers, a similar date may be accepted for the tower. A number of features when considered together date the piers and arch-head at the western end of the north arcade to the Saxo-Norman overlap. The jambs of a single square order rising straight from the ground with no base are typically Anglo-Saxon or Saxo-Norman, and the single through-stone which would appear to have been preserved from the corresponding south arcade provides further evidence for such a date. The imposts are also of Anglo-Saxon or Saxo-Norman type; since these are simUar to those of the tower-arch aheady discussed they need not be given separate treatment here. The arch-head shows Norman influence in its neatly-cut voussoirs with their radial setting. Inequahty in size of individual voussoirs is found after 1100 although there is an improvement in the late twelfth century; and the single square order (unohamfered) of the arch-head suggests an Anglo-Saxon or Saxo-Norman date. One may pretty safely conclude that this surviving fragment of arcade was buUt by Anglo-Saxon workers who were famihar with Norman building methods, perhaps because they were working under a Norman-trained master. The arcade should therefore be dated to the late eleventh century or possibly to the early twelfth century. The whole archway is simUar to, though on a smaUer scale than, the tower- and chancel-arches at Clapham (Beds.). At Clapham the arch-springings are set-back from the line of the jambs beneath, as is the western springing at St. Peter's.17 The walling above the arch, in the north face of the arcade (supra, p. 103), includes an arrangement of stones in Tredington fashion and smaU close-set stones; these features are consistent with a Saxo- Norman dating. 11 T. P. Smith, Anglo-Saxon Architecture in Bedfordshire Churches, forthcoming. 16 Fisher, op. cit., 116. 10 Clapham, op. cit., 118. 17 Taylor and Taylor, op. cit., i, 158; T. P. Smith, 'The Anglo-Saxon Churches of Bedfordshire', Beds. Arch. Journ., iii (1966), 11. 106 THE CHURCH OF ST. PETER, CANTERBURY The Saxo-Norman church would have been of basihcan type, with a nave 12 ft. wide and with simple arcades and narrow aisles; the north aisle was presumably of similar dimensions to its counterpart to the south. There was probably a clerestory above the arcade, one window of which may remain behind the depression above the Saxo-Norman arch. The tower must be seen as belonging to the south-eastern group of early bell-towers, which includes examples in Kent, Sussex and probably Essex.18 UsuaUy these towers are found in an axial position, although the early tower at West Kingsdown in Kent is at the southeast corner of the nave. The tower at St. Peter's is smaller in area covered than most of the towers of the group. The particular interest of St. Peter's lies in its vernacular, Saxo-Norman treatment of a basihcan plan. Of the nature of the east end of this early church nothing is known, aU work in this part of the building being of later date. All that can be said is that the Saxo-Norman church would not have extended as far eastwards as the present buUding, for a survey of c. 1200 shows that the ground now occupied by the most easterly part of the church, and described then as iuxta ecclesiam sancti Petri, was at that time in private ownership.19 ACKNOWLEDGEMENTS The writer would hke to thank the Rev. Derek Ingram Hill, M.A., Master of the Eastbridge Hospital and Rector of St. Peter's, for permission to make measured drawings of the church, and for his kind encouragement. My brother, Mr. D. F. Smith, helped with the measuring. I am grateful to Mr. A. P. Detsicas, M.A., F.S.A., for his editorial comments, and to Mr. S. E. Rigold, M.A., F.S.A., F.R.Hist.S., for his painstaking criticisms and comments and for providing the Appendix. APPENDIX (Fig. 4) THE DEVELOPMENT OE ST. PETER'S AND ITS RELEVANCE TO THE PRIMARY WORK By S. E. RIGOLD, M.A., F.S.A., F.R.Hist.S. The early nave was of four bays and some 40 ft., or 12 m., long: this is almost certain, as the division between nave and chancel, the most persistent demarcation in any church, is shown not only by an extra tie-beam (rood-beam?), in the roof but by a sharp change of orients Examples from Kent and Sussex are: Swanscombe (K); and Bosham, oevington, Singleton, and Sompting (S); some of these may be post-conquest, like St. Peter's. 10 'Rental D' in W. Urry, Canterbury under the Angevin Kings, London, 1967, 209, 304 ff. 107 T. P. SMITH tation. That this begins a trifle to the east in the north arcade may show that the original, and presumably short, chancel did not share this 'weeping', and that a fragment of it may survive. The chancel was extended to its present length, on the deflected hne, quite early in the thirteenth century, and a lancet survives just south of the high altar. Of about the same date is the short length of wall containing the south door, with Purbeck dressings, once 1-5 m. wide and subsequently reduced. There is no evidence that the long chancel was originaUy aisled: the arcades are simply cut through its walls, and despite their simphcity of treatment, they need be no earher than the fourteenthcentury chancel aisles. That on the north generaUy has bar-stops to the chamfers, which could bear an earher date, but that on the south has broach, or pyramid, stops, which point strongly to the fourteenth century. The alterations at this period produced three different sorts of window in the east gables, but the intervals between the phases need not be long. The first phase involved the re-ahgnment of the whole west wall, together with the matching window in the high chancel. The large north aisle (Lady Chapel), which originaUy only covered the length of the chancel, must be a httle later, but has been given the same hatching on the plan. Its north-west quoin, of re-used Caen stone, has a vaguely 'long-and-short' appearance, which has occasionally deceived people. West of this, the waUs of the north aisle are thinner and worse built, generally of field flints, and are similar to those of the whole south wall, right up to St. John's chapel, which disguise the basic change of orientation in the church. The windows of this phase are fairly elementary, but suggest the third quarter of the fourteenth century. The westward extension of the north chapel, to produce a short but wide nave aisle, almost certainly went with the demolition of the eastern piers of the early arcade and the spanning of their positions by single plain arches, as weU as the alteration of the remaining early pier on the south, whether or not its core is primary. The effect of this work was to produce an uninterrupted transverse space in the nave—in fact a preaching-space, likely enough in an urban parish in the later fourteenth century. The roofs are all of crown-post type. That over the nave and chancel s consistent throughout, with caps and bases to the crownposts, quite advanced in character if they are contemporary with the first fourteenthcentury alterations. That over the north has similar crownposts, and deals effectively, if clumsUy, with the change in orientation. I t may be that both these roofs date from late in the century, but if so, they do not match the roof of the south aisle, where the crown-posts are plain and inexpensive. The fittings and monuments can be found in the standard reference books, but the square shelly marble font is worth mentioning as a quasi-architectural feature of c. 1200. 108