Romanesque Fonts in Kent: the French Connections

ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS C. S. DRAKE STONE FONTS A major feature of the early Middle Ages in western Europe and Scandinavia was the growth of trade, encouraged by the development of vessels designed for the transport of heavy loads. By the early years of the twelfth century this trade included a lively traffic in baptismal fonts. There is plenty of evidence for this all round the Baltic, up and down the coastline of northern Germany and Jutland, and along the major river systems of southern Sweden, the Low Countries and the Rhineland. 1 The quarrying of stones of high quality on the island of Gotland, in the valleys of the Maas and Scheidt and in Ostfriesland along the course of the Ems, attracted skilled masons who established their workshops within the quarries, close to the point of extraction.2 The fashioning of products in stone produced large quantities of waste - a piece of given finished size would have originated in a rough block up to double its mass and weight. Transport costs at the time were relatively even higher than in the twenty-first century and it made far more commercial sense, as well as being more practical, to import finished products rather than blocks of stone in its raw state and then throw away a large proportion of what had cost so much in money and effort to deliver in the. first place.3 Work in the quarries also allowed the establishment of significant groups of craftsmen capable of turning out sophisticated work of a high standard. This concentration of skill would not have been possible if every parish had been reliant on its own local neighbourhood stone-cutter to work the rough block when it had been delivered. Despite its long coastline and short sea connections with the Continent, England has very few imported Romanesque fonts. The seven fonts of black limestone (and one fragment) produced in quarries along the banks of the Scheidt near Tournai, have long been 333 C. S. DRAKE thought to be the only imported fonts in the country.4 The reason for their being so few is clear: England is well provided with a variety of stones of high quality in almost every corner of the country and a study of English fonts reveals that there are very few major groups of the same type (Aylesbury, Bucks.; Bodmin and Altarnun, both straddling the Cornwall/Devon boundary; and Silk Willoughby, Lines.) but many small clusters and even pairs of similar design, such as are found in Norfolk, Shropshire, Yorkshire and other counties.5 This suggests the existence of many small quarries serving quite small areas. Support for this view comes from the relatively poor quality of many pieces and the large number of English fonts which are either plain or have very little decoration. The highest quality of execution is almost always to be seen in the larger groups, particularly Aylesbury and Bodmin. The most numerous English group is from the Isle of Purbeck, the source of a dark limestone, and they are found dispersed over a wide area of southern England from Essex to Devon, and especially where there is no naturally-occurring stone suitable for the manufacture of fonts. 6 It is an interesting fact, and almost certainly relevant to their widespread existence, that they were made in seven pieces, bowl, base, central shaft and four colonnettes, with each element weighing very much less than an equivalent single-piece font, making transport much easier. Why then import Tournai fonts? The pieces brought to England represent about an eighth of the total known to have been produced by this School and are among the finest of a group which is of almost uniformly high quality. In the corpus of Romanesque fonts they stand out as true luxury items and it is no surprise that the four in Hampshire - at Winchester cathedral (Plate I), St Mary Bourne, East Meon and Southampton, St Michael) - can all be linked with Henri de Blois, Bishop of Winchester. Similarly the two in Lincolnshire, in the cathedral and at Thornton Curtis (formerly at Thornton Abbey) were acquired by Alexander, Bishop of Lincoln, the close friend of Henri. The presence of the remaining two, at Ipswich, is much more difficult to explain and they may not even have been imported until the fifteenth century, perhaps by Thomas Wolsey.7 Now, however, another foreign import has been identified, in Kent. It is not a Tournai product and the stone did not come from the Scheidt, although it is distantly related. Just as in England a seam of golden limestone runs from the Cotswolds to the Wolds of Lincolnshire, so limestone occurs along the axis from Boulogne to Aachen, though with much greater variation of colour. In the Boulonnais, the hinterland of Boulogne, are the huge Marquise quarries, still an active 334 ROMANESQUE FONTS lN KENT: THE FRENCH CONNECTIONS PLATE I The Tournai font in Winchester Cathedral 335 C. S. DRAKE PLATE II The Newenden font showing the three decorated sides (from top) - lion and dragon; geometrical pattern; mythical beasts and human figures in roundels 336 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS centre of a vast stone extraction industry. During the twelfth and into the thirteenth centuries this stone was used for the making of baptismal fonts and a considerable number are to be found in towns and villages between Boulogne and Amiens. 8 It was while researching for his book on Romanesque fonts9 that the author came across the piece now in the church at Newenden (Plate II). Although it is thought to have been imported originally for that church, it has not always been there and was for a long time in the parish church at Rye. When the Newenden parish demanded its return, at some time during the nineteenth century, the vicar of Rye was so sad to see it go that he commissioned a faithful copy which has stood ever since at St Joseph' s, Rye. On first sight, the author had no reason to believe that the Newenden font was any more than another of the many single-specimen English fonts, made in a pale limestone of no special interest. It was not the first English font encountered during his research which seemed to have no siblings, though its basic 1 +4 support configuration (a massive central shaft and four colonnettes beneath the corners of the bowl) suggested a link in concept, if not in execution, with the Purbeck fonts, of which there are many in East Sussex and in Kent. There are considerable differences in the proportions between this and the Purbeck type, with Newenden being generally more massive, with a thick base and square bowl with prominent rounded 'belly' below the vertical sides. There are also differences of decoration, but it could still be the work of someone who had been commissioned to make a five-support font following a description of a Purbeck 1+4 model, but without his ever having seen one. Three years ago the writer visited the Boulonnais. He did not find the expected Tournai fonts but did discover a whole new family of fonts made of a limestone which varied in colour, from the pale Caen stone so familiar from many of the Gothic cathedrals of England, to the rich golden colour found in the Cotswolds and parts of Northamptonshire. Although the family resemblance is unmistakeable, the considerable variations in size, shape, proportions and decoration would never suggest a link between Newenden and most fonts of this group. At Hesdres (Plate Ill) and Tubersent (Plate IV), however, the link is plain and unequivocal, strongest in the proportions and in the matching decoration of these French fonts with two of the faces of the Newenden bowl. The lion and dragon decoration of the third side at Newenden does not occur elsewhere in the Boulonnais group. The fourth side at Newenden is plain. Where it now stands, the plain face is against a pillar of the nave, begging the question (if it could be shown that this has always been its position in the church) whether it had been commissioned, not only for the church of 337 C. S. DRAKE PLATE Ill The two decorated sides of the Hesdres font (top): detail of left side (below) 338 ROMANESQUE FONTS lN KENT: THE FRENCH CONNECTIONS PLATE IV The two decorated sides of the Tubersent font 339 C. S. DRAKE PLATE V The Marquise font at Halloy Newenden but also for the very spot within the church. Why pay for all four sides to be decorated if one of them was never to be seen? Hesdres and Tubersent each have two sides plain and two decorated, the like sides adjacent, and stand in corners of quite small churches so that any decoration on the wholly plain sides would have been difficult, if not impossible, to see. There is, thus, even a possible link with the existence of undecorated sides. The chief structural characteristic of the Marquise fonts is the sense of weight which they project, an impression given even by the smaller examples (e.g. Halloy, Plate V). The square bowl is thick and is often as deep below the vertical face as the face itself, with a swelling 'belly' between the capitals beneath the corners; these are carved out of the underside of the bowl. The central shaft is short and thick and the corner colonnettes are also so thick that it is impossible to be sure 340 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS whether they are actually separate but touching the central shaft, or whether they have been carved out of the same block but not so deeply undercut as to detach them. The slab which forms the base or plinth is also thick but less so than the bowl; its top surface is carved to provide sockets for the location of the supporting shafts. No Purbeck multi-support piece even approaches the massive form of these Marquise fonts. Only the Tournai products are as big but they are taller, with more slender colonnettes which are clearly separate, and the whole impression is one of greater lightness (Plate I). At Newenden the ornament is different on each of the three decorated sides of the bowl (Plate II). On one are a lion and a dragon, on the second a geometrical pattern and on the third are mythical beasts and human figures in roundels. The side with the lion and dragon has nothing special about it to suggest the Marquise workshops. Both creatures are common in baptismal iconography; the dragon ( an alternative to the serpent) represents the Devil and is a reminder of the forces which lie in wait for the unbaptised and which continue to fight for dominion over his soul right up to the moment he is received into the Christian Church through baptism. The symbolism of the lion, as the Bestiaries make clear, is more complex and it can either be good or evil. Christ Himself is the Lion of Judah, yet monks before they retired prayed to be saved from the Devil 'who, as a roaring lion, walketh about seeking whom he may devour'. As the lion in this case is portrayed in threatening pose, jaws wide agape, he must be construed as evil, a doubling of the threat of his dragon companion. Neither dragon nor lion are known to the author on any other Marquise fonts,· at this scale, though they do appear in the roundels, as will be seen. Both creatures face right. Although many fonts display geometric designs, billet, chevron, sunken star, rings and rosettes, the saltire is very rare as a major motif. At Newenden there are a pair and they enclose bunches of grapes, withfleur-de-lis filling the top and bottom spaces (Plate II, middle). Geometrical patterns based on diagonal lines can be found on fonts but are not common and the author has only met the saltire as a geometrical motif at Killucan and at Kilcooly Abbey, both in southern Ireland, but they are very different and without the grapebunches. The saltire is also found in some of the varied cross motifs such as at Stoke Canon, Devon (Plate VI), even more distinct. The third side is taken up with a row of four 'inhabited' roundels (Plate II, bottom). Reading from left to right they contain: a mythical beast with a human head; a lion regardant (i.e. looking backwards over its shoulder); a six-petalled flower; a dragon looking between its legs. The first two face right, the last faces left. 341 C. S. DRAKE PLATE VI I I The saltire motif at Stoke Canon, Devon Roundels containing human and animal figures are not unique to Marquise fonts but no other fonts use the design in quite the same way. 10 Of English fonts only St Marychurch, Torquay, has occupied roundels the full height of the face but here the bowl is a cylinder, not square. There are a few, also round, which have bands of roundels occupied by men, mythical beasts and birds as a second tier of ornament, always at the top, for example at Alphington, Devon (Plate VII). Of the Tournai fonts, Winchester cathedral has two sides with roundels; on one they enclose pairs of doves, and on the other single doves face outwards in the outer rings flanking a lion passant regardant in the centre (Plate I). Dendermonde (Belgium) has roundels on two sides, doves flanking the Agnus Dei on one, and on the other 342 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS PLATE VII The font roundels at Alphington, Devon (top) and St Michael's, Southampton (below) 343 C.S. DRAKE three winged beasts regardant. St Michael's, Southampton (Plate VII), has roundels on all four sides, three with lions and one with the symbols of three of the Evangelists (St Luke is absent). In the Museum at Bryn Athyn, USA, is a fragment of a bowl which certainly had mythical beasts in roundels on three sides and may well have had all four decorated in the same way, but nothing remains of the fourth side. The difference between these four examples from the Tournai School and the Newenden font is that Newenden has a shallower vertical face to the bowl, with four roundels filling the full width of the face, while the others have only three, much larger, roundels on each side. Of the Boulonnais fonts seen, only Hesdres and Tubersent have roundels. At Hesdres they are on both decorated faces. On Side 1 (Plate III, top) they contain the following images, all facing right: an eagle; a lion regardant; a beast with a human head; a dragon looking between its legs. On Side 2 the images are: a six-petal flower; a human being, head and shoulders; a lion gardant; a dragon looking through its legs (Plate III, below). The last three face right. The Tubersent piece, sadly, has lost the top few centimetres of its height, making it difficult to read the images accurately (Plate IV). On the single side with roundels they appear to frame: a lion gardant; a mythical beast with a human head; what appears to be a human figure, head and shoulders; a dragon looking between its legs; all except the third figure, which seems to be en face, face right. The second face at Tubersent has two lines of chevron top and bottom, so placed that they form pairs of rough saltires. These enclose grape-bunches.and in the spandrels are 'piles' flanked by beads. There are many square, five-support fonts with figures in roundels but at Newenden the combination of style, proportions and decorative programme offers such an excellent match with Hesdres and Tubersent that the affinity cannot be ignored. Now it has been confirmed by an experienced geologist that the material, an ooidal limestone, almost certainly did originate in the Marquise quarries, so the eight fonts of the Tournai School must share their rare distinction as exports to England with a product of the Boulonnais. 11 LEADEN FONTS Lead, as a material for the manufacture of early medieval fonts, seems to have been a speciality of England and France, though in neither country are they numerous.12 (The author knows of only three others elsewhere, two in Germany, at Bremen cathedral and Hellefeld, 344 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS and one in Austria at Salzburg.) However, the ease with which the metal can be melted down, accidentally or by design, makes it hard to rule out the likelihood that there may have been far more. 13 Other parts of Europe, mainly Germany and the Low Countries, also used metal for fonts but there the tradition was that of the bell-founders, casting in bronze and in a bell-shape. Both traditions lasted for many years but that of casting fonts in bronze was more continuous and began 75-100 years later than in lead, though there was a considerable overlap. Both metals were used as late as the sixteenth century. Perhaps the finest metal font of the Middle Ages is that now in the church of St Barthelemy, Liege, but formerly in Notre Dame des Fonts in the same city. This font is in fact of brass. All metal fonts were cast but bronze fonts were made in one piece, like a bell, while the lead, in almost every case, was cast horizontally, in one or more flat sections, then joined together and welded into a cylinder, with a flat disc being finally welded into position to form the bottom. Even if the font were to be cast in one long continuous piece, the craftsmen frequently used the same matrix pressed repeatedly into the sand to form a mould with recurring motifs. A close examination of these lead fonts reveals quite easily how many welds there are and thus the method by which each individual piece was made. The same manufacturing process was used on both sides of the Channel and there has been much debate on the subject of whether lead cylindrical fonts were an English idea transferred to France, or the opposite. Professor Zarnecki thinks that the former is the more likely as generally the English fonts seem, using stylistic criteria, to be earlier than their French parallels. 14 The design was not a new Romanesque concept, for in the Museum at Ipswich is a RomanoBritish cistern of almost identical proportions to those of the twelfthcentury fonts. It bears the Chi-Rho sign and may have been used in some way for ritual cleansing, if not for baptism. Although lead fonts occur in many places in both England and France, there is a concentration, especially of the earlier examples, near the coast on both sides of the Channel. However the idea was conceived and developed, there can be little doubt that there was cross-Channel collusion. The similarities between the products of the two countries occur in shape, size and decoration but are rarely so close as to point to the same hand being at work, though a group of six in Gloucestershire has been shown all to have been made from the same mould. The most common form of decoration consists of one or two horizontal bands of foliate ornament around the top and bottom of the bowl, the space between being occupied by series of panels with seated or standing 345 C. S. DRAKE figures, in many cases alternating with a formalised foliate motif. All these lead fonts stand upon later stone pillars to bring them up to a convenient height. They are smaller than their stone equivalents. In Kent and Normandy, however, are two fonts with decoration which is quite different from all the others; these are in the churches of Brookland and Saint Evroult-de-Montfort respectively. What they have in common and marks them out from the rest, is the decorative programme which uses a combination of the Signs of the Zodiac and representations of the Labours of the Months. This combination in itself is unique on baptismal fonts, though the Labours are known on two fonts in Norfolk (Warnham All Saints and Burnham Deepdale) and four of the common scenes representing the months portray the Seasons at Thorpe Salvin, in Yorkshire. Sagittarius is depicted, and named as such, on the font at Hook Norton, Oxfordshire, where there may also be Aquarius, who is believed to appear at Thorpe Salvin too, but in very schematic form. There are possibly Zodiacal signs on other fonts but they cannot be confirmed with certainty. Is the centaur, so common on southern Swedish fonts, intended as Sagittarius or merely as a representation of evil? The Brookland font is taller relative to its diameter than the norm for English lead fonts (Plate VIII). The main part of the bowl is cylindrical but the bottom band flares slightly, as does the top which turns over into a curled rim. There is a single join, visible inside and out, indicating that the font was cast in one flat sheet, rolled into a cylinder and welded. Below the top flare is a thin cable moulding and a band of saw-tooth. Attached to the rim are three small cast panels depicting Christ's Ascension; He rises from the tomb between two angels as the soldiers sleep below . 15 The main surface area is divided into two tiers of arcading with depressed arches, the columns of the upper row standing directly on the supports of the arches below. Each tier has twenty arches and they match, except for the lower tier being taller. The arches are cast in panels of four, with no attempt to achieve seamless joins between the panels. Clearly the matrix for the single mould was made in sections. The upper tier contains the Signs of the Zodiac and the lower the Labours of the Months, all of which are named in the arches, the Signs in Latin and the Months in Norman French (Plate VIII). Twenty arches were needed to provide a bowl of sufficient diameter - with two tiers of arcading each arch would have been grossly wide if only twelve had been used to provide the desired diameter. The duplication begins with March and ends with October. March is captioned in error in both places CAPRICORNVS instead of ARIES but has the 346 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS PLATE VIII The Brookland font, with detail (below) 347 C.S.DRAKE correct ram Sign. 16 The Month below is MARS with a peasant in a hooded cloak pruning a vine. April has TA VR VS with a bull and AVRIL below, with a woman holding sprays of foliage in both hands. May has GEMINI with two small figures in broad- brimmed hats embracing, above MAI with a rider hawking. June has CANCER with a strange crab above JVIN with a man scything. July has LEO with a leaping lion and JVILLET with a man raking. August has VIRGO, a female with outstretched arms and AOVT with a man cutting corn with a sickle. September bas LIBRA with a woman in a long robe holding scales above SETEMBRB, a man with bare torso, flailing. October has SCORPIO, a toad-like creature with a pointed tail over VITOVVRE with a man treading grapes in a cask. 17 Novem- ber has SAGITARIVS, a centaur armed with bow and arrow and below is NOVEMBRE, a man in hooded cloak knocking acorns to the ground for his pig. December has CAPRICORNVS, a mythical beast with goat's head and DECEMBRE with a man killing a pig with an axe (Plate VIII, detail). January has AQUARIVS, a ma1:1 pouring water from a pot and JANVIER has Janus, the two-headed god, at table feasting, a drinking horn in his left hand. February has PISCES, a pair of fish head to tail above FEVRIER, a man in hooded cloak who sits warming himself at a fire. 18 Saint Evroult-de-Montfort, in Normandy, has a font of remarkable similarity to Brookland, though taller and tapering slightly to the bottom (Plate lX). 19 Slips in the registration of the separate blocks of the pattern compare unfavourably with Brookland in the execution of the preparation of the mould. There is a plain narrow band at the bottom and a single slender cable below the thickened salient rim. The same arrangement of Signs and Labours, one above the other, recurs in panels of four pairs of arches but here in two tiers. A major difference between the two fonts is that at Saint Evroult the names of the months are in Latin and that there are four standing figures in pointed arches the full height of the bowl which separate the Signs and Labours into four blocks. These tall arches, without base or capital, are framed in saw-tooth lines; the shafts have spiral ornament but this is not carried over the arch. The figures are nimbed and bearded, stand en face and hold books, undoubtedly representing the Evangelists. Another variation at Saint Evroult is the use of narrow ornamental bands of double or triple saw-tooth below the half-panels. The double-deck arrangement of the panels gives a total of thirtytwo pairs, providing the artist with a problem of duplication even more complicated than at Brookland.20 The upper row begins with March, running from left to right, and repeats the months November 348 ROMANESQUE FONTS lN KENT: THE FRENCH CONNECTIONS PLATE IX The font at Saint Evroult-de-Montfort, with detail (below) 349 C.S. DRAKE to February, neatly bringing the sequence back to the starting point. In the lower tier the order is reversed and begins with September, October, followed by July, August, May, June, March, April. It then starts again: September, October, July, August, May, June and then July, August. This seems a very strange ordering and may derive from the patterns for the months being made in pairs, so that a complete reversal of the order was not possible. March here has ARIES correctly and the representations of the Signs are smaller than at Brookland, relative to the Labours in the panels below. The Labours depicted are mostly, but not entirely, the same: July depicts a man using a cultivating tool with long handle and curving blade, October has a man cutting grapes from the vine, November a shepherd or swineherd and January a man feasting at a table spread with food. March to June and August, September, December and February have the same scenes as Brookland. Another pointer to the undoubted connection between the two fonts is the insertion of strange demi-lunes at the bottom of the arcading. At Brookland there are three, all empty, and they sit beneath columnbases between J\'farch/April, November/ December and July/August. At Saint Evroult they occur twice in each panel, plain in the upper tier and in the lower filled with five spokes; they are not all placed beneath a column, some filling the full width of the arch. At the top they occur in June, straddle September/October, occur again in February and straddle January/February. At the bottom (in the same panels as the foregoing) they straddle September/October, occur in June, in June again and straddle September/October again. This, with the slantwise displacement of one of the upper half panels suggests the re-use of patterns to create the various parts of the mould. The similarities between the two fonts are such that, if not created in the same workshop, there must have been a common factor in their manufacture; perhaps they were both made by a travelling craftsman or there was a sponsor who took notes of one font while abroad and commissioned a similar vessel in his home village. But which way did the influence migrate? The use of Norman French at Brookland is not conclusive evidence for a French origin for both, especially as this appears to be the earlier of the two fonts and the language was still in use by educated people in England during the twelfth century. The pointed tall arches (Plate IX, detail) and the crocket capitals at the top of the stone support at Saint Evroult are both signs of its Transitional date and, while this cannot rule out the possibility of the Brookland font having been created in Normandy, Saint Evroult-deMontfort must have been inspired by Brookland and not the reverse. 350 ROMANESQUE FONTS IN KENT: THE FRENCH CONNECTIONS CONCLUSION Thus Kent has two baptismal fonts of the twelfth/thirteenth centuries, both with a French connection. In the case of Newenden the direction of influence is clearly from France to England; Newenden is an import from the Boulonnais. With the lead font of Brookland, however, it appears that the migration was one of ideas rather than of hardware and that the direction in this case was from England to France, though one cannot be certain. There is no record of the Labours of the Months appearing on any other French font, though the subject was quite commonly used there in churches of the twelfth and thirteenth centuries as sculptural decoration; they occur frequently with the Signs of the Zodiac, especially on the portals of western France.21 This seems to be a unique baptismal juxtaposition on these two very similar pieces. BIBLIOGRAPHY C.S. Drake, The Romanesque Fonts of Northern Europe and Scandinavia, Woodbridge, 2002. G.C. Druce, 'Lead Fonts in England, with Some References to French Examples', Journal of the British Archaeological Association, n.s. XXXIX, 1934. J. Trouvelot, 'Remarques sur la technique des sculpteurs du Moyen-Age', Bulletin Monumental, LXXXXV, 1936. G. Zarnecki, English Romanesque Lead Sculpture, London, 1957. ENDNOTES 1 Fonts produced on the island of Gotland were exported to mainland Sweden and to Denmark as well as all round the Baltic States. German fonts from the Bentheim quarries, along the present frontier with Holland, are found in that country and as far north as Jutland, where are also found fonts from the Meuse valley which are most common along the Rhine and in the Ardennes. 2 For a discussion on whether raw stone was worked at the point of extraction or on site, see Trouvelot 1936, pp. 103-8. 3 A.C. Leighton, Transport and Communication in Early Medieval Europe, AD 500-1 JOO, pp. 162-70. 4 The seven fonts in use (not all complete with the original supports) are at Ipswich (St Peter), East Meon, St Mary Bourne, Southampton (St Michael), Thornton Curtis, and in Lincoln and Winchester cathedrals. There is a fragment of an eighth font (about a quarter of the bowl only) in the collections of the Ipswich Museums. 5 See Drake 2002, Chapter 1. 6 Some of the fonts generally identified as Purbeck are actually made of a similar stone found in Sussex; they are indistinguishable to the casual eye. 351 C.S. DRAKE 7 The city of Tournai was given to Wolsey by the king. The son of a butcher of Ipswich, he built a college in the town of his birth and the fifteenth-century St Peter's is adjacent to the site of the college of which only a gateway survives. The bowl stands on a fifteenth-century base which was quite clearly made for it, suggesting that only the bowl was imported. · 8 The variety of designs, though all falling within a clear family outline, shows that these fonts must have been produced over a considerable period, probably between about 1150 and 1225. The group includes some fonts which are not square but, as the French would say, bar-long. 9 See note 5. to Not to be confused with these are the fonts in which human and animal figures are enclosed in the undulations of a vine tendril, e.g. St Mary Steps, Exeter. 11 The author is indebted to Mr Robin Sanderson, BSc, CGeol, FGS, who inspected the Newenden font for him. He declared that 'the lithology and structure are correct [for the Marquise ooidal limestone] and the early date and the position of Newenden, at the time being closer to the sea, ... make [the Marquise quarries] the most likely source.' He goes on to suggest the closest matches in England being Stamford or Ancaster limestones, but they lack the same ooidal content. It is well known that Marquise stone is to be found in Kent as a building material but the only other possible font made of this material is at Fawkham, near Dartford; this example is round and undecorated and so is not linked with the Marquise group to which Newenden belongs. 12 This section on lead fonts has been taken largely from Chapter 8 of The Romanesque Fonts of Northern Europe and Scandinavia. 13 It has, however, been suggested that this is not likely to have been the case in churches where Romanesque stone fonts survive today, as few parishes would have been able to afford to acquire two fonts within a short time of each other. 14 Zarnecki 1957, p. 2. 15 G. Zarnecki, p. 37, says these panels were added in the thirteenth century. 16 In this article the captions are given in capitals and spelt as on the font. 17 The spelling of the month of October is clearly wrong. Druce (p. 11) refers to the Liv re des Creatures by Philippe de Thaun, first half of the twelfth century, in which the spelling occurs. He suggests this is a phonetic spelling of HVITOVRE or the eighth month, correct if we accept that the year began with the spring equinox in March. 18 Druce also draws attention to the Classical flavour, with the female figure for April being Venus goddess of fecundity, Gemini being Castor and Pollux, and with the Virgo figure, holding cornucopiae, as Ceres. 19 On a panel in the church at Brookland are large photographs of Saint Evroult for comparison. 20 G. Bouet, 'Note sur les fonts baptismaux de Saint Evroult-de-Montfort', Bulletin Monumental, XVIII, 1852, pp. 423-6, points out that close examination reveals that the patterns for the mould were made with pairs of months together. 21 The only other baptismal context which comes to mind is the great octagonal baptistery at Parma (Italy) where there is a set of free-standing figures representing the Labours but the Signs of the Zodiac do not appear with them. 352