Crayford Church

Plate C R _ \ Y F © R P • C I U R C M • K E M T • SCALE-OF • FEET- . ,'P s o l l l l l 1 1 I I I I IO 2 0 30 4 0 50TT.ZT Original Quoin Pl_?_H I Coal Shed Original Quoin Bs^tuCTura. w^rer TW1« Vertical Joint 32'• 10"— 1- fJCWIWRY-iOMITl. 0>W<&._. C K U . 1 . B E FL__n-n (CKAMCEL GROVrW-PL51__3 PL_W* I JG xijl-Sg dottrel? nxvr. FL?«ini • 117- bn=»tCe.*z JitzeO •M»«3 «r Pel. J^oS ( 51 ) CRAYFOKD OHUROH. BY THE REV. G. M. LIVETT, F.S.A. THE Church of St. Paulinus at Crayford is well known to ecclesiologists as the church with the twin naves. A glance at the Plan shews what is meant by the expression " twin naves." The navespace, which measures 61 feet in length and about 44 feet in breadth, is covered by a pair of span-roofs, which rise from a wall carried by an arcade of four and a half arches that runs up the centre of the nave, the half-arch having the nature of a flying arch and abutting upon the east wall of the nave above the crown of the chancel-arch. It is this central arcade, dividing the nave into two nearly equal parts (the northern twin being only 6 inches wider than the southern), that constitutes the peculiar feature of the Church. For a long time it has been a puzzle to archaeologists. The object of this Paper is to record a solution suggested by discoveries made in the course of a long and careful study of the Church by Mr. W. Braxton Sinclair, to whom the Society is indebted for the plans and sketches which have been specially drawn to illustrate the Paper. These discoveries have settled once for all the question of the position and form of the original Church: that Church occupied the area now covered by the northern twin nave. Eemains of the west and north walls of its nave and of the gable-wall of its chancel can still be traced in the west, north, and east walls of the northern twin, while the south wall of its nave must have run upon the lines of the three westerly arches of the central arcade. A very fair idea of the peculiar arrangement of the present Church may be obtained from an " ink-photo " plate that appears in Archceologia Cantiana, Vol. XVIII. , facing page 322. That plate illustrates a Paper from the pen of the late Major Alfred Heales, F.S.A., upon Crayford Church. The Paper is interesting from a historical point of view, but does not afford us much help otherwise. There is a plan of the Church, and some wood-cuts of sections of mouldings, but they are strangely incorrect, and there- • E 2 52 CRAYFORD CHTJRCH. fore of little value as a basis for studying the problem which the Church presents. A Paper of greater value for our present purpose was written by the late Archdeacon of Maidstone, who (as Canon Smith) became Eector of Crayford twelve years after the Church was restored (in 1862) under the oversight of Mr. Joseph Clarke, F.S.A. The present Eector, the Eev. J. P. Alcock, has kindly lent me a copy of that Paper. It bears no date or printer's name. I t appears to be a verbatim report of a lecture delivered by Canon Smith, and internal evidence proves that it was illustrated by tracings made from the architect's drawings of any remains of old masonry which came under his notice.* In the absence of those drawings Canon Smith's description of the features which they represented is not always very clear. His evidence, however, is very important, and in some particulars it affords remarkable confirmation of the interpretation of the recent discoveries advanced in this Paper. We shall discuss each point as it occurs. Of course the problem of the twin naves cannot be considered apart from the rest of the building. Attached to the west end of the southern twin there is a tower with a tall narrow arch of communication. The chancel is about half as long as the nave and not quite as broad as one of the twins. On each side of the chancel there is a chapel extending from the west a little more than half its length. To the east of the north chapel there is a vestry filling the angle which the chapel makes with the eastern part of the chancel. The south chapel projects beyond the line of the sidewall of the nave, and there is a south porch to the nave. Mr. Sinclair's plan and elevations will make all clear to the armchair student. I would direct special attention to the transverse section of the twin naves, shewing the chancel-arch cut by the central arcade, the two side-arches looking into the chancel-chapels, and the arrangement of the twin roofs.f This should be studied in connection with the ground plan of the Church and with the longitudinal sections and elevations. The Plan shews that the chancel, measuring about 30| feet by 18i feet, is perfectly rectangular, and that it lines exactly with the central arcade of the nave, that is to say, if a line drawn on the plan through the centres of the * Enquiries have been made in many directions with a view to discovering these drawings, said to have been preserved in the architect's office, but unfortunately they cannot be found. f The transverse and longitudinal sections of the Churoh, with sections of mouldings and other details, are all shewn on PLATE II. PLATE I. contains a series of ground-plans whioh illustrate the growth of the Churoh. °F§» \ CRAYFORD CHURCH. 53 columns of the arcade be produced eastwards, it coincides exactly with the axis of the chancel. The nave is narrower at the east end than at the west: its side-walls converge slightly towards the central arcade as they run from west to east; the convergence amounts to 3 inches in the case of the northern twin and to 4 inches in the southern. The two side-arches shewn in the transverse section, one on either side of the chancel-arch, formerly tallied with the chancelarch in height and appearance. They were raised to their present height in 1862. The chancel and the twin naves were in existence long before the side-chapels were added, so that these former side-arches (as well as the arches looking from the chapels into the chancel) must have been inserted in outside walls which contained windows. Above the low arch which formerly existed on the south side Mr. Clarke found the remains of such a window. Canon Smith described it as " a narrow lancet," that would have come exactly in the centre of the east end of a " side-aisle." Canon Smith's theory of the original form of the Church may be given in his own words: " A Norman cell with a central nave and side aisles; a chancel in line with a central nave; a tower attached to the south aisle, the tower-arch opening into the aisle at its west end." The Canon accounted for the existing central arcade by thinking that three parallel span-roofs formerly rested on the existing side-walls and on the two supposititious aisle-arcades, and that they were abandoned in favour of two span-roofs resting as the present roofs rest on the two side-walls and the single central arcade, the aisle-arcades being then destroyed. Major Heales' theory was very similar, but he thought that the alteration was made in 1630, the date of the present roofs, shortly after a fire, and that the central arcade was made up of the remains of the two side-arcades. In view of the recent discoveries the " side-aisle " theory must now be definitely abandoned. Indeed, leaving those discoveries out of the question, the theory cannot stand, for it is impossible to squeeze a south arcade into the plan in such a way as to avoid making it abut on to the northern respond of the tower-arch, which therefore cannot have been designed to look into an aisle as supposed. Moreover the tower-arch is too tall to fit into the cross-section of such an arrangement. And again, the east and west walls of the northern twin not-only contain no sign of there having been a north arcade, but they afford positive proof that such an arcade never existed. • 54 CRAIFORD CHURCH. There is one point in Canon Smith's lecture which must be noticed in this connection. He said: " During the last day or two, however, I regard the conjecture of an original tripartite division of the Church as confirmed by the discovery I have made in the tracings of the old walls by Mr. Clarke. For there, in the central hue of the Church where the respond of the arcade now stands, I find that a doorway 5 feet wide was built up in the western wall . . . . If these are the remains of the ordinary western door into a central nave, the theory of a nave and aisles antecedent to the present arcade and twin nave arrangement is confirmed.' Now, as against the existence of such a doorway, we must first note that the language used was not very positive: the Canon speaks of " a discovery he has made in Mr. Clarke's tracings of the old walls," and adds, " If these are the remains," etc. He did not see the doorway, he saw only Mr. Clarke's tracings of the old walls: there is room for a mistake. I think that the theory of a doorway can be easily accounted for, and certainly it can be disposed of. I t is disposed of by the fact that some of the original exterior facing of the early-Norman wall remains under the present groundlevel just where the doorway ought to be. The present nave-floor is about one foot above the original level and about one foot below the present ground-level outside the building. "We have lately examined by digging the face of the wall below the present ground-level and can confidently say that there was no doorway in the position described. Moreover, we have found the sill and the lower part of the jambs of the actual west doorway of the original Church remaining in the wall on the central axis of the northern twin nave. It is shewn in the sketch.* The inner face of the wall is now covered with plaster. Mr. Clarke, however, saw the face stript, and apparently noticed some marks which he represented in his drawing. Probably what he saw was the rough joint of the junction of the original west wall with the inserted west respond of the central arcade. The joint must have run up beside the respond at a distance from it of about a foot or a trifle less. Either Canon Smith " discovered " in this joint, shewn in the drawing, the sign of the jamb of a doorway, or Mr. Clarke himself mistook the rough joint for such a sign and slightly indicated a doorway in his drawing. Be that as it may, it is pretty certain that no doorway ever existed in the position described. Moreover, careful measurements prove * See PLATE II. Pla.te II A SMUT • PAIJUNUS • CRMTORP• C«ount> U7iE |l * o r i w »i laa INHMUIU MMM • 9 * n « _ TOH£ . - LETT rt_Air • RBMSIWS • ©r-KMM___ -womnsf- ^-w?__M RI© • or • • UlAK.l-]!H;W_J_ • •WSn__-«»r-r1I''llV»Di-KL__ • • M©ra _n__ j_w.m» WIHKni_ra BBCTMOT pRi2>MT-E_nrea_ B®RI - _ 9£%HiE fflPFBE _ [OK.TCI-W I Sffi_TI_-V8__I& lilil I I 0 eo y mHHKH^& y K3 (E2 rt SCALE or FEET U& M JWMJ £ I I _E53Z •L®HOPITOIPIf_?_lLi • gUCTWOH-OH-LDIB-A-B • CRAYEORD CHURCH. 55 that the wall to the south of the respond of the central arcade was not laid out at the same time as the wall to the north of it. The west wall of the northern twin makes with the central arcade an angle that is perhaps a few degrees less than a right angle, and the west wall of the southern twin makes with it an angle that is distinctly less than a right angle, so that not only are the two parts of the wall, one on either side of the respond, of different thickness, but they do not run in a straight line with one another. This fact would of itself suggest a fallacy in the theory that the respond was made to abut upon the centre of the west wall of the original Norman Church. Before we consider the site and dimensions of the original Church and endeavour to trace the various stages of the growth of the Church in chronological order, we shall do well to subject the various parts of the present building to further analysis and assign to them approximate dates. The two chancel-chapels are the latest parts of the mediaeval building. They are called respectively the Newbury Chapel and the Howbury Chapel. The former, that on the north side, is also called May Place Chapel. There is a tradition that the name of " May Place" (which belongs to the Manor of Newbury) is a contraction of " Mary Place." The will of Henry Harman, " Clerk of the Crown of the lord King," mentions a Chapel of St. Mary on the north side of the Church which the testator intended to build de novo, and in which he willed that he himself and his wife should be buried. In the same will, which is dated 1502, he bequeathed five marks to " the construction or reparation of the Chapel of St. Mary the New in the parish of Yard" (Crayford). In the will of Wm. Ladd, dated 1504, there is a bequest to " Or Lady " in the same chapel.* Canon Smith, who was followed by Major Heales, identified the north chapel with that of which it is recorded (in the report of a Eoyal Commission issued in 1548) that it was built by one John Marshall, whose father's will was dated 1488.f The question is thus involved in some obscurity, but there can be little doubt that the north chapel is to be identified with " Our Lady chapel" in the Ladd will and with " the chapel of the Blessed Mary" in the Harman will; and judging from the architecture we cannot be in serious error in assigning its construction to a date shortly before * See a Paper by Mr. Leland Duncan in Archceologia Cantiana, Vol. XXIII., p. 138. f Archceologia Cantiana, Vol. XVIII., pp. 324, 325. 5 6 CRAYEORD CHURCH. or shortly after the year 1500. The chief architectural features may be studied in the illustrations. The south chapel seems to be a little earlier in date; its arch shews perhaps a little more skill in construction, and the mouldings are more refined in form. The arch is four-centred, but only slightly removed from a two-centred arch. The form of the base, the height of the column, and the contour of the capital more nearly approach the details of the central arcade than do the corresponding features of the arch of the north chapel. _____ IV« 7—7 ^ S \\ n t i . « \ 6 \> \ \ * H aa _£ > y QUOIN STONES OF NOETH-WEST ANGLE or VESTET. That the north chapel was built after the vestry is proved by a plinth which runs along the west face of the vestry wall (which now divides the two buildings), as well as by its north-west quoin, which remains intact, the north wall of the chapel being built up flush against it. This is shewn in the sketch. A buttress, coeval with the chapel, masks its junction with the earlier wall to the west of it. The chapel is lighted by a large square-headed window of four lights; it is covered by a span-roof running parallel with the roof of the chancel, while inside it has a flat panelled wooden CRAYFORD CHURCH. 57 ceiling of late date. Under the eaves on the north side there runs a cornice, consisting of a broad overhanging casement moulding, which is continued along the east face under the gable and is partly covered by the ridge-roof of the vestry. This, by the way, shews that the vestry was originally covered by a lean-to roof, the slope of which was continuous with that of the chancel-roof. The south chapel, which by a modern enlargement has assumed a transeptal form, has and always had a flat roof with a parapet. Now if these two chapels and the south porch (which seems to be a post-Reformation erection) be eliminated from the plan of the existing Church (PLATE I., Plan I.), there will remain a plan of the Church as it existed towards the end of the fifteenth century. It can be imagined so easily that no plan of it has been included in PL_.T_ I., but to shew its outline the tower and vestry have been •added in dotted line to Plan IV. To reach an earlier stage in the history of the Church we have to eliminate the vestry and the tower and to alter the central arcade. The vestry, which stands on the north side with its end flush with the end of the chancel, seems to have been a Perpendicular or fifteenth-century addition to an earlier chancel. When this addition was made the chancel-wall was made thicker by about six inches in order to carry the plate of the rafters of a lean-to roof of the vestry. In the centre of this wall, some four feet below the line of roof, is a corbel, which no doubt carried a vertical wall-piece connected with a principal rafter. Mediseval vestries in this position are not uncommon.* The tower presents generally a Perpendicular appearance, though it contains a west window which some people have considered to be earlier in date. The west doorway shews two continuous orders with hollow chamfers, which are stopped, each by means of a highly inclined flat slope, immediately above a slight hollow-chamfered plinth. This is exactly the treatment of the vestry doorway of the Stanepit Chantry-chapel in Dartford Church, which may be dated early in the reign of Henry Vl.f In the east wall of the same * Solihull in "Warwickshire and Eising in Norfolk have Decorated examples; Marston in Bedfordshire supplies a Perpendicular example; Hingham in Norfolk is an instanoe of a Perpendicular addition to a Deoorated chancel; while in Southfleet in Kent, and in Floore and Aldwinkle in Northants, Decorated chancels have ooeval vestries standing a little to the west of the east end. All these examples are taken from the plans in Brandon's Parish Churches. t See Archceologia Cantiana, Vol, XVIII., p. 385. The date given in the referenoo is " the reign of Edward IV. probably," whioh seems to me to be too late. Tho (Ire-place may have been inserted late in the fifteenth century. 5 8 CRAYFORD CHURCH. chantry-chapel there is a niche which formerly contained the image of Our Lady; it has a trif oliated' head in which the points of the hollow-chamfered cusps approaching quite close to one another almost form a semi-circle under the upper foil. This gives the niche a distinctive architectural character, and as it is reproduced both in the single and in the double lights of the upper stages of the tower at Crayford, it follows that the two works were being built about the same time. That the tower is all of one date, excepting only the battlement, is evident from the similarity of the different parts, e.g., the way in which the rear arches stop on to the jambs. The same feature is seen in the nave-windows and in the chancel-arch, which will come under notice presently. I t appears, then, that in order to reach an earlier stage of the growth of the Church, and to imagine it as it existed at the beginning of the fifteenth century, we must eliminate from the Plan (I.) not only the late-Perpendicular side-chapels, but also the vestry and the tower, which were probably built respectively a few years after and a few years before tbe middle of the fifteenth century. If, as seems probable, the three-light windows of the twin naves and the central arcade and chancel-arch were built about the same time as the tower, they too must be eliminated. This will leave us with merely the chancel and the shell of the twin naves ;* and these parts will represent all that we now have left of the Church previous to the building of the tower and the present central arcade and the insertion of the three-light windows, all which we have assigned to the first half of the fifteenth century. Fortunately, the story of the growth of the Church, as it will be unfolded in the latter part of this Paper, will require very little modification should experts decide that these dates are not quite correct. It is perilous, perhaps, for an amateur to dogmatize on points upon which even experts are found sometimes to err. We now come to consider the central arcade. It affords a good illustration of the difficulty of fixing dates. A section of the base of the columns of the arcade shews a hollow-chamfered plinth, and above it a bell-shaped base overhanging a shallow surface of " clene hewen asshlar," which intervenes between it and the hollow-chamfered top of the plinth. (In mediseval contracts the plinth is called * The elimination also of the central arcade would leave us bereft also of the roofs of the twin naves, if wo could not find some substitute whereby the roofs were supported previous to its ereotion. This, however, will offior no real difficulty. CRAYEORD CHURCH. 59 the ground-table, and the flat surface and the overhanging bellbase, which in this case are worked on one stone, are together called the ligemeut or ledgemeut-table.) Now this base was never common, but its use extended over a period of not less than one hundred and fifty years. We have it in its earliest form in the bases of the nave-arcades at Dartford, proved by the scrollmouldings of the caps and by the other work with which they are associated to be not later than the middle of the fourteenth century, perhaps earlier. It occurs iu the aisles of the nave and presbytery of Winchester, which are dated circa 1400. After that date it was usually modified by the addition of a " cushion" immediately under the bell; but it is occasionally found without the cushion at a much later date, as in the Crayford chancelchapels. The bases, then, of the Crayford arcade afford but slight evidence of its date.* The bold and unbroken curves of the attached shafts and intervening hollows of the columns are not common in the Perpendicular period, while the design of the arches, shewing a wavemoulding on the chamfer of the lower order and a double ogee on that of the outer order, separated by a plain re-entrant angle, would be consistent with a late-Decorated date. On the other hand, there are distinct Perpendicular features: the bell-bases are much bolder than those of Dartford, and the debased form of the abacus and bell-moulding of the capitals as compared with the scroll and roll of Dartford, seems to be decisive in favour of a Perpendicular date.f Moreover, the designer seems to have been feeliDg his way towards the form that became common in late-Perpendicular work. He was inclined to give each shaft with its base and cap the appearance of having an individual existence. In the west respond he succeeded in doing this: the lower order of the archivolt-mouldings is carried by a semi-cylindrical shaft whose base and cap are quite independent of the rest of the design. The respond to which the shaft is attached is flat, except that its edges are moulded with a double ogee, which alone runs up to form the * The use of the simple hell-base (without ground-table and sunk surface) was very common in late-Deoorated aud early-Perpendicular times. The ground-table associated with other forms of base made its appearance as early as oiroa 1320 (see Willis's Winchester), and its use in late-Perpendioular times associated with some form of bell and oushion base became almost universal. (See a plate of oymagraphed sections in a Paper on Chart Ohuroh in this Volume!) f Major Heales, Archceologia Cantiana, Vol. XVIII., p. 324, favoured a late-Deoorated date. 60 CRAYFORD CHURCH. outer order of the arch. Certainly the spirit of the design is Perpendicular rather than Decorated. The architect of the late-Perpendicular chapels seems to have been influenced by the design of the arcade. His continuous outer orders were cut from a mould which was identical in form with that of the double ogee of the arcade, and his inferior order repeats the wave-moulding. His shaft is a three-quarter shaft, and it is separated from the outer moulding on each side by a hollow casement, which runs up round the arch, separating the two orders. The two designs are shewn side by side in the accompanying sketches. \ I i > I i I i I m Ai ^ «X ( H1 i i ;ti 4 1 - i RESPOND, SOUTH CHAPEL. RESPOND TO CENTEAL AECADE. I take it, then, that considering all these points we cannot put the date of the central arcade later than 1440, while some people might like to put it twenty or thirty years earlier. The chancelarch seems to belong to the same date as the tower and the arcade, despite the bare appearance of its flat-faced responds, which may be accounted for by assuming that the arch was intended to be filled with a screen. The segmental shape was necessary in order to permit the abutment of the central arcade upon the wall above its crown. It is interesting to note that a similar segmental CRAIFORD CHURCH. 61 chancel-arch existed in Dartford Parish Church previous to the recent restoration; it is seen in a picture that hangs iu the vestry. I t is not likely that sUch an arch was tliere when Bishop Hamo de Hethe in 1333 put in his fine five-light window,* so we may assume that it was inserted when the Stanepit Chapel was remodelled in the fifteenth century. The introduction probably had some connection with the erection of the first and lower of the two successive roodlofts, of which many indications remain. The arrangement would be a very uncommon one, and it is not improbable that it was suggested by the low segmental chancel-arch inserted about the same time at Crayford. The Perpendicular windows of the nave have been mentioned more than once. They are of a common type, seven in number: four on the south side, two on the north, and one at the west end of the northern twin. They are square-headed externally; the lights are cinquefoiled; the wall-arches are slightly segmental. I take it for granted that all these windows, which are plainly insertions into au older wall on the south side as well as on the north, are of the same date as the arcade.f They point to a considerable remodelling of the nave in the fifteenth century. This, then, is the nett result of our enquiries so far: the tower, the nave-windows and central arcade, and the chancel-arch were all built iu the second quarter of the fifteenth century; the vestry was built perhaps a little later ; and the two chancel-chapels were built circa 1500. All these must be eliminated from the plan of the existing Church (Plan I.) so that we may realize the form of the Church as it stood in the fourteenth century. But this process leaves the Church with only one window, the south window of the chancel,% and none in the nave, and it also leaves it without any central arcade to support the twin roofs of the nave. Therefore we have now to examine any evidence that may supply the * See Archceologia Cantiana, Vol. XVIIL, p. 384, and Plate xxxix. of Thorpe's Custumale Roff. t There is another Perpendicular three-light window of larger dimensions towards the east end of the north wall of the nave. Its head is segmental externally as well as internally, and its external label is stopped with a pair of carved figures of angels. The whole of the stonework was renewed by Mr. Clarke, but one of the original figures may be seen in the Rector's garden, and there seems no reason to doubt that Mr. Clarke's work is a fairly faithful copy of the old. This window may be somewhat later than the others: perhaps it was in some way connected with the erection of a staircase to the rood, of whioh the remains exist in the wall immediately east of it. X It is said that the east window cannot be relied upon as a copy of the original. 62 CRAYFORD CHURCH. fourteenth-century Church with windows, central arcade, and chancel-arch. The side-windows of the chancel were probably four in number, disposed as in Plan IV., and like the one remaining on the south side (restored by Mr. Clarke). Traces of red paint may be found on the existing arcade and on the responds of the chancel-arch, which bear marks also of fire, and, like the arches looking from the chancel into the chapels, SECTION OF MOULDINGS. T •FRAGMENT" — *NOYM«ir*' •^ •KECTORT?fc. _£ J "^rf30_!_L CRAYFORD CHURCH. 67 The north-east quoin marks the position of the east face of the original chancel-arch wall. Remains of the foundation of such a wall in exactly the required positioii and of the required thickness were discovered by Mr. Clarke. We shall have occasion later on to quote and consider the passage from Canon Smith's lecture which records that discovery. For our present purpose it is sufficient to mention it: it affords trustworthy confirmation of the length of the early-Norman nave as deduced from the foregoing data. On the receipt of Mr. Sinclair's report of his discoveries I immediately marked them on a copy of his plan of the existing Church, and tried to plot a chancel of the common early-Norman type to see how it would fit into the plan. If the reader will glance again at the plans of Padlesworth and Dode (Vol. XXII.) and test them with a pair of compasses he will see that the chancel in each case, measured externally, is approximately a square attached to the east face of the east wall of the nave, i.e., to the chancel-arch wall. A comparison of several Kentish early-Norman chancels suggests that the builders often (though not always) laid them out in this way. They took the internal width of the nave, or a little more, and made it the length of one side of their chancel externally. The result of this method was to make the length of the chancel internally greater than the breadth by the thickness of their wall. Applying this method to the Crayford plan I found that the east wall of the early-Norman chancel would coincide as nearly as possible with the hne of the east wall of the existing twin naves, making it probable that in the gable of the northern twin nave some remains of the east and of the original might be found still existing. Before this theory could be tested I received reports of a further discovery which actually confirmed these anticipations. To this our study must now lead up. The remains of the early-Pointed doorway and round-headed lights in the north wall of the original Church point to an extensive remodelling of the Church in the early-Pointed period. The alterations cannot have been confined to the north wall, and we should naturally expect signs of. early-Pointed work in any remains of the old chancel that might come to light. The discovery referred to in the last paragraph indicates that a triplet of windows, very similar to the early-Pointed lights of the north wall, was inserted at that time in the east wall of the old chancel. The remains discovered consist of nine or ten stones just above the monument affixed to the wall near the north jamb of . 2 68 CRAYFORD CHURCH. the chancel-arch. The material is fire-stone, and the masonry, the facing, and jointing are exactly like the early-Pointed work in the side-walls of the nave. The stones form the upper part of the common jamb and the separate springers of the splayed rear-arches of two adjoining lights. The northern arch springs from a level about 16 feet from the ground, 6_ inches below the tie-beam, and 4f inches above the level of the springing of the southern arch. The common jamb from which these two arches spring at these different heights is 7 inches in width, and its edge is 1 foot 5 inches from the vertical line of the side of the chancel-arch. The turn of the taller of these two arches, which run up behind the tie-beam, may now be seen above it by any one who stands at the west end of the Church. The outline of the lower stones is just visible through the text which has been painted on the surface of the wall. I t was first noticed by the Hector's colleague, the Hev. F . V. Baker, who called Mr. Sinclair's attention to it. Its importance as bearing upon the history of the Church cannot be overestimated. It had escaped the notice successively of Mr. Clarke, Canon Smith, and Major Heales; but the Church of late has been so thoroughly examined that probably not a square foot of it that is visible has escaped attention. On examining the exterior face of the gable of the northern twin, seen above the valley of the chancel and north-chapel roofs, we have found evident remains of the early-Norman rubble walling of the gable-wall of the original Church—the wall, that is, in which the early-Pointed triplet was inserted,* The result of all this is that we have been enabled to plot the early-Norman nave exactly, and the chancel within limits of error amounting at the most to two or three inches, the exact position of the side-walls being alone undetermined.f In this connection the drawings of Mr. Clarke's discoveries, if we had them, would probably help us; and a certain importance attaches to Canon Smith's description, to which we have occasion to refer in considering THE EABLY-POINTED ADDITIONS. So far we have been working on sound principles with plenty of evidence to guide us. We now enter on questions of some * Slightly indicated in Mr. Sinclair's bird's-eye view of the roofs. f The Churoh stands in an angle formed by two roads running respectively north and north-north-west. The old entrance to the ohurohyard is from the latter road, affording a direct approaoh to the west door, The old graveyard evidently lies to the west and south of the Church, CRAYFORD CHURCH. 69 difficulty and a field of speculation. It seems to me impossible to discover and work out any theory of the post-Norman growth of this Church (consistent with the features it presents and with the analogy of other Churches) on the assumption that the present south wall of the nave (with its early-Pointed windows) was built in the early-Pointed period. To begin with, I know of no analogous case. The peculiar feature of a central arcade is not without parallel ;* but a parallel of a side-addition of the same breadth as the nave, and extending the whole length of a Church of the common early-Norman type, would I think be hard to find. Of course nothing is impossible, and exceptions to the ordinary laws of plan X-\ HGsGsetanr/ V-, \ / / WW77%%MA * I H-Ho-rt i ?pnVS>u;s.lic ._*« E CRAYFORD CHURCH. 73 THE DECORATED ENLARGEMENT. If we are correct in our dates the alterations made in the fourteenth century were very considerable, and constitute by far the most important stage in the history of the evolution of Crayford Church. They were quite revolutionary in character. They transgressed all the common laws of growth as illustrated by Kentish Churches. We find extensions that have been made to the east and to the west of the chancel-arch, but as a rule the axis of the Church has remained unchanged, and the line of the division between the nave with its aisles and the chancel with its chapels has remained unchanged. At the Church of St. Lawrence, Bamsgate, a tower seems to have been built over the original small square chancel and a new chancel erected to the east of it, but the dividing line between the people's nave and the eastern parts of the Church remains unaltered.* Possibly at Throwley the dividing line has been pushed to the east of its original position, but this has not been worked out. I know so far of only one Kentish Church in which there is no doubt that such a change has been made,t that of Northfleet, where, in the fourteenth century, a large was huilt in 1839), and the chancel was practically rebuilt, its south wall being made to line with the nave-arcade, and its length being increased by fully ten or twelve feet, while the north wall only of the original Norman chancel was retained. This retention was due to the fact that additions to the north side had been already made, consisting of a narrow aisle (7_ feet wide) and a chancelchapel, whose side-wall was an eastward continuation of that of the aisle. (See a Paper with plan by Major Heales in Archceologia Cantiana, Vol. XVIII.) The north aisle has been destroyed, but the arch opening from it to the ohapel remains, blocked.up, in the present west wall of the ohapel. The significance of this arch has been missed by the writer of the Paper mentioned above. It is perfeotly round-headed, and the diagonal marks of the Norman axe are still clearly to be seen on the surface of its stones. This disposes of the theory that the Church was originally an Early English Church with aisles. It was a Church of the common Norman type like that at Crayford. The Early English east end retains signs of three lancet windows, of which, if the arrangement were not a common one, we should regard the triplet at Crayford as the prototype. Let me here ask the reader not to attach undue importance to the exact form of early-Pointed additions shewn in the Plan No. III. of Crayford Church. The aisle may very well have been two or three feet wider and its wall have been continuous with the side-wall of the chapel to the east of it, like the arrangement that existed at St. Paul's Cray on the north side. There is no evidence to guide us but that of analogy, and that is only general. The general prinoiple on which the growth of the Church is worked out in this Paper is not affected by uncertainty on these points of detail. PS.—Compare Doddington Church. * A similar extension with transeptal chapels was carried out at Godalming Churoh, Surrey. t These nbtes refer to Kent, and might be extended to include the southeastern counties generally. But it is probable that several examples of similar extensions might be found in the large churohes of some other counties, such as 74 CRAYFORD CHURCH. new Church was built up round the old one.* At Crayford, in the same century apparently, not only was Ihe line pushed eastwards, but the axis of the chancel was shifted to the south, and the old chancel literally absorbed and its lines obliterated by the extensive alterations then carried out, including the adoption of the curious device of a central arcade. I t is probable that as a preliminary step to these alterations the arcades on the south side of the old Church were boarded up so that the services might be continued without interruption. The aisle and chapel were then demolished, and at the same time the new south wall was erected, the old doorway and windows being rebuilt with it. The early-Pointed arcade was probably preserved and made to serve as part of a central arcade. Then followed the removal of the south wall of the old chancel with its arcade (the services having been transferred to the nave of the old Church) and the eastward extension of the nave-arcade by the addition of an arch and a half-arch abutting upon the crown of a chancel-arch, whereby a central arcade (the predecessor of the present arcade) was completed, the materials of the old chapel-arcade being used for the purpose. The work would now be far enough advanced for the erection of a span-roof over the new southern twin nave; and an eastward extension of the north wall of the old nave in place of the north chancel-wall now demolished would enable the builders to make the necessary addition to the old nave-roof so as to cover the northern twin nave. The new chancel, which may meanwhile have been in course of erection (by a different set of masons), is so exactly in line with the Norfolk and Lincolnshire, with which I am not so well acquainted. Perhaps the fine Church of All Saints, Maidstone, ought to be noticed in this connection. There is but slight evidence of the position of the early Churoh whioh certainly existed on the site, but what there is suggests that it ocoupied the space between the nave-arcades from the second piers from the west to the line of the ohancel-arch. In spite of Mr. Cave-Browne's arguments in favour of an earlier date, I am persuaded that the existing Church is all of one date, and that it was founded by Arohbishop Courtney, so that one may date it circa 1400. This is a case in which the earlier Church must have been entirely demolished. Its plan and dimensions in no way influenced the design of Courtney's Church. * The tower of the old Norman Church remains at the west end of the present Church; and the three westerly bays of the south arcade are thirteenthcentury work incorporated into the later work. These three bays, which belonged to an added aisle, shew the length of the original nave, whioh ocoupied the site of the western portion of the present nave. All signs of the original chancel have disappeared. PS.—Goudhurst Churoh is another example, seen the day before these proof-sheets oame to hand, Perhaps, also, Minster in Thanet. SAMT-PAULIilK •CR_\TTORPS CAME OF HEE T 2 3 * a io T d j (&J-'i~V P^A^P. W WMJ5B+h ^.& •WCT"IMTta©£_- CRAYFORD CHURCH. 75 central'arcade that it is probable that in laying it out the builders stretched a line along the south face of the early-Pointed arcade and, extending it eastward, used it as a guide. The chancel-roof (the predecessor of the present roof, which is of the same date as the present nave-roofs) was the last to be built, for its west gable, which rises up and overlooks the valley of the twin roofs, is a mere lath and plaster construction that rests on a tie-beam at the extreme west end of the chancel and is built up against the face of the nave-gables. In reviewiug the plan of the Decorated Church as it has issued from the foregoing data, there is an apparent difference between the work of the chancel and that of the nave. The chancel has thick walls, an external plinth or ground-table, diagonal buttresses at its angles, and all new work in its windows, which are of the style of its, period. The uave has none of these: all the windows were old ones rebuilt, and the work of the cheapest character. Is it possible that there was an agreement between the rector aud the parishioners as to the plan to be followed in the enlargement of the Church, and that each party carried out its contract according to their means ? A course something like this seems to have been followed at Battle Church in Sussex in the thirteenth century (see a Paper by the present writer in the Sussex Archceological Collections, 1903) ; probably other instances might be found. FIFTEENTH-CENTURY AUDITIONS. The fifteenth-century additions and • alterations fall into two periods. The earlier works, as we have seen, comprise the building of a new chancel-arch and a new central arcade in the place of the former one, and the remodelling of the lighting of the nave by the iusertion of the three-light windows. The original west doorway seems to have been preserved at this time, for the sill of the window above it is higher than that of its companion windows. The rebuilding of the central arcade must have been prompted by purely aesthetic considerations. It is likely that the three westerly arches were supported by somewhat rude square piers, that the third free pier from the west was composite in construction, consisting in part of the early-Pointed respond and in part of a respond added in the fourteenth century, and that the half-arch at the east end and the arch next to it being fourteenth-century 76 CRAYEORD CHURCH. work were somewhat different from the other three (early-Pointed) arches, so that the appearance of the whole arcade was patchy.* The existing arcade has been fully discussed. It may be noticed, however, that the spandrel of the half-arch is built of larger stones than the other spandrels. Possibly all the old material had been used up in the other spandrels. Mr. Sinclair, however, suggests that the difference in construction is intentional, and that it is based on the system of corbeling, whereby the thrust on the chancel-arch is reduced to a minimum. The foundations of three of the columns of the arcade appear at different levels above the present floor. They are not seen in the case of the west respond or of the east column. These differences are probably due to difference in floor-level, a legacy from the early-Pointed arcade. The floors of our churches remained for a long time unpaved, and the varying level of the ground must account for any difference in the level of bases. The material of the arcade is Kentish rag, a material which builders seldom used for cut-stone before the fourteenth century, while its use does not seem to have become common before the middle of that century. * Doddington, originally au early-Noyman Churoh of the common type, has a south chancel-chapel of early-Pointed date and considerable width, and a south nave-aisle which when it was built was muoh narrower, but now is of the same ROOD-LOFT DOORWAY IN NORTH WALL. CRAYEORD CHURCH. 77 At the east end of the northern twin nave, on the north side, there are remains of an entrance to stairs that at one time led up to a rood-loft. The material is Kentish rag, and the jamb shews a hollow moulding treated much in the same way as in the doorway of the newel staircase of the tower.* There can be little doubt that a rood-loft was erected across the east end of the northern twin nave not very long after the rebuilding of the central arcade was completed, possibly at the same rime. It is not improbable that the fourteenth-century chancel-arch contained a screen surmounted by a small rood. The prevailing taste of the fifteenth century would demand a more gorgeous and conspicuous rood. In any case it may safely be assumed that it was about the middle of that century, or a little later—when, according to the wills (see a Paper by Leland L. Duncan, F.S.A., in the Transactions of the St. Paul's Eccles. Society, vol. iii., part 5), rood-lofts were finding their way into churches all over the country—that a new rood and a rood-loft were erected at the east end of the northern twin nave,f which from old association would retain a character of greater sanctity than the newer part of the nave on the south of the arcade. Whether the window near to the blocked doorway was put in at the same time or at a later date one cannot say for certain. In conclusion, I must emphasize the indebtedness of all who are interested in Kentish architecture to Mr. W. Braxton Sinclair for width as the chapel. (If I remember rightly there are signs of a destroyed north nave-aisle.) It appears that when the chapel was added, the south wall of the chancel was taken down and rebuilt with a pair of arches a little further south. The tie-beams of the chancel seem to be the original beams lengthened by scarfing, so as to cover the increased width. The chancel-arcade therefore very nearly lines with the nave-arcade, but between them stands a mass of masonry (pierced with a curious forked squint looking into both chancel and chapel), on to which the chancel-arch and the west chapel-arch abut. Supposing now that these two arches were removed, the two arcades would appear to be continuous, with a bit of blank walling at their junction; and, supposing a new chancel were added centrically to the east end of the Church, this continuous arcade would run up to the centre of the new chancel-aroh, and we should have the history of Crayford Churoh reproduced—one can imagine how the "patchy" central arcade would suggest its being rebuilt. In fact, Doddington Churoh affords an excellent object-lesson in illustration of the story of Crayford Church, as unfolded in this Paper. * The sill of the doorway is 2 feet 6 inches from the present floor, or about 3£ or 4 feet from the original level, not "6 feet," and must have been approached by wooden steps. f See Canon Scott Robertson's Paper on Kentish Rood-Soreens in Archceologia Cantiana, Vol. XIV., pp. 371, 372, and footnotes. " "We know that there were rood-lofts in side aisles of some churches. In Ashford Church two such lofts were erected in A.D. 1472, one in the north aisle and another in the south aisle." 7 8 CRAYEORD CHURCH. •the prolonged study he has made of Crayford Church and for the careful drawings which have contributed in no small degree to the solution of its history; and to the Bector thanks are due for his kind hospitality and for the ready assistance he has afforded to Mr. Sinclair and myself in our combined labours. APPEOXIMATE DATES. EARLY-NORMAN Church, of which north and west walls of nave and portion of east gable to chancel remain: (?) the Church mentioned in Domesday under Erhede (i.e. Yarde), the old name of Crayford. EARLY-POINTED insertion of windows, and (?) addition of south aisle and chapel (destroyed in fourteenth century) : circa 1200. DECORATED enlargement, of .Church to existing plan (minus tower, vestry, chapels, and porch), with twin roofs over nave: fourteenth century. PERPENDICULAR remodelling of lighting and ritual arrangements; rebuilding of central arcade; addition of tower and vestry: early fifteenth century ; chapels, circa 1500. New roofs, 1630, after fire. Modern restoration, 1862.