1 For a summary of the history and method of the use of mouldings to identify the styles of individual masons, see Roberts, Eileen, ‘Moulding analysis and architectural research: the late Middle Ages’, Architectural History, 20 (1977), 5–13 CrossRefGoogle Scholar. Mouldings are also an invaluable dating tool for the medieval archaeologist, if supported by proper national and regional surveys (such as that of Dr Roberts for the Hertfordshire area).

2 For a recent exception, see S. Rigold, ‘Romanesque bases in and south-east of the limestone belt’, in Ancient Monuments and Their Interpretation (1977), edited by M. Apted, et al., Chapter 7.

3 Paley, F. A., A Manual of Gothic Mouldings (third edition, 1865)Google Scholar.

4 Bond, F., Gothic Architecture in England (1905), pp. 658–707 Google Scholar. It is not, however, entirely free of inaccuracies, e.g. the Winchelsea arcade arch (p. 669, 4) and base (p. 697, 5) are not correct in every detail; note, Bond uses moulding drawings from a variety of previous works (see his Index of Illus trations, pp. 70 sqq.). The finest of the published nineteenth-century collections are those of Edmund Sharpe, Architectural Parallels (1848), and idem, The Mouldings of the Six Periods of British Architecture from the Conquest to the Reformation (1871-74).

5 H. Forrester, Medieval Gothic Mouldings (1972).

6 Forrester states in his Preface : ‘A profile of mouldings in a position militating against a close-up investigation is represented by the general impression afforded to the eye, attention being directed to the use in the profile of any basic moulds.’ No one who works with mouldings can avoid using at least some drawings done by eye, where it is impossible to get close enough to use a template former or lead wire, or make a detailed measured drawing. In this article, such drawings are indicated by an asterisk, and inevitably they contain some risk of inaccuracy, though precautions have been taken wherever possible to check them. All other drawings were made with template formers, which is a faster method than the measured drawing procedure of the professional architect, and produces accurate copies, if the template formers are used carefully and the salient points of complicated moulding formations checked periodically by measurement.

7 The archive is of full-size moulding drawings, and at the time of writing (September 1977) a card file indexed both according to place and to moulding type is being compiled. At the same time, a computer programme is in preparation for storing and comparing all the data (to take a simple example, it will locate and list all the locations of wave mouldings known, or of waves in conjunction with sunk chamfers). Currently, the author’s collection is of about 5,000 examples (half of them full-size), concentrated mainly on the period f. 1250-1400, with an English emphasis, but including Belgian, French, German, and Spanish examples. Once established, it is planned to enlarge the range of the collection to the whole medieval period, and probably to the post-medieval periods as well : the survey applies to materials such as timber and metal, besides stone and brick. In addition, an analytical survey of tracery patterns is planned on a similar basis. The main full-size drawings are already available for reference at Warwick, and further progress will be reported in the Society’s Newsletter. Enquiries and offers of moulding drawings should be addressed to: Dr R. K. Morris, Department of History of Art, University of Warwick, Coventry.

8 Brakspear, H., ‘A West Country School of Masons’, Archaeologia, LXXXI (1931), pls xvi and xviiGoogle Scholar (Wells Cathedral north porch).

9 Fawcett, R., ‘Later Gothic Architecture in Norfolk’ (unpublished doctoral thesis, University of East Anglia, 1975), pp. 96 sqqGoogle Scholar.

10 Morris, R. K., Decorated Architecture in Herefordshire (unpublished doctoral thesis, University of London, 1972)Google Scholar.

11 See further, R. K. Morris, ‘Worcester Nave: from Decorated to Perpendicular’, in Medieval Art and Architecture at Worcester Cathedral, British Archaeological Association Conference Transactions, 1 (to be published in 1978).

12 In the lists a name refers to the obvious medieval monument in that place with work from c. 1250-1400 (e.g. ‘Wells’ is Wells Cathedral); where two or more important monuments survive, the name refers to all those works unless stated otherwise (e.g. ‘Coventry’ refers to all the Coventry churches, but ‘Bristol Cathedral’ means that no other churches in Bristol have been surveyed in detail).

13 The percentage figure has generally been adjusted slightly upwards for region VI, because so many smaller churches without elaborate and varied masonry detail have been surveyed in it: of the seventy-three listed, only about twenty-five would be regarded as important or well known. By implication, therefore, the percentages given for other regions will probably prove to be rather too high once all the small churches in them have been surveyed.

14 Paley, op. cit., pp. 49-51, where there will also be found a list of other terms that have been applied to the wave moulding.

15 Another early example may be the exterior surround of the sanctuary east window at Winchelsea parish church, after 1288 (Fig. 5 B, i), but this window presents several problems. Though its tracery is a modern replacement of a Perpendicular window, it is probable that the mouldings of the surround are to the original design, re-used in the Perpendicular work. It may be, however, that this window was not amongst the earliest Decorated work to be executed there, for many of its forms find parallels not in the other windows of the east end, but in the transept, the date of which is less easy to gauge. For example, the only other real wave mouldings and sunk chamfer, mouldings there occur in the surrounds of the transept windows, which may not have been executed until after the bulk of work on the chancel (i.e. after 1312, when the establishment of the first Alard chantry suggests the east end was more or less complete). On the other hand, the window need not be as late as the chantry, for other instances of the wave in the south-east appear from their stylistic context to belong to c. 1300 (e.g. Dover, Maison Dieu). So the most likely date is c. 1300-10.

16 All area references with block capitals are to the regions designated in the introduction.

17 Another fruitful line of research might be to investigate whether masons in certain areas tended to carve paired wave mouldings (or any other moulding worked on the chamfer plane) on the same plane (e.g. Fig. 2 E), whereas masons in others canted the mouldings at different angles to each other (e.g. Fig. 2 D). For further details on the latter usage in areas V and VI, see Morris, R. K., ‘The Local Influence of Hereford Cathedral in the Decorated Period’, Transactions of the Woolhope Naturalists’ Field Club, XLI, Part 1 (1973), 59 Google Scholar.

18 For the purposes of this paper, ‘Decorated’ means c. 1290-c. 1350; ‘late Decorated’, c. 1330-c.1350; and ‘early Perpendicular’, c. 1350-1400.

19 I am grateful to John Maddison for noting this example.

20 Coldstream, N., ‘York Chapter House’, Journal of the British Archaeological Association, 3rd ser., xxxv (1972), 18 sqq.Google Scholar. In Aylmer, G. and Cant, R. (eds), York Minster (1977), p. 141 Google Scholar, Eric Gee proposes 1287-93 as the dating for the Southwell carved capitals.

21 See especially the moulding profiles of the tomb’s main arcade, in Archaeologia, CIII (1971), 117, fig. 5.

22 See Addington, H., Some Account of the Abbey Church of SS. Peter andPaul, Dorchester (1845), e.g. p. 26 Google Scholar, south aisle piscina.

23 For the mouldings of the Dorchester screen, see Rickman, T., An Attempt to Discriminate the Styles of Architecture in England (fifth edition, 1848), p. 171 Google Scholar, bottom right illustration.

24 I am most grateful to John Maddison for allowing the material in this paragraph to be published : it will appear at much greater length in his forthcoming doctoral thesis, ‘Major Building in the North-West Midlands and North Wales, c. 1270-1400’ (University of Manchester).

25 Illustrated in Royal Commission on Ancient and Historical Monuments in Wales and Monmouthshire, Caernarvonshire, 11 (1960), 126, fig. 103, A, B.

26 See Thompson, F. H., ‘Excavations at the Cistercian Abbey of Vale Royal, Cheshire, 1958’, Antiquaries Journal, XLII (1962), 206 Google Scholar, fig. 8, 7/8.

27 A. J. Taylor, ‘Castle-building in Wales in the later 13th century’, in Studies in Building History, edited by E. M. Jope (1961), Chapter 7.

28 Valléry-Radot, J., Congrès Archéologique de France, 116, Auxerre (1958), 26–39 and 45-50Google Scholar.

29 The chantry of St Nicholas in the south aisle was founded by Robert Alard in 1322, Stephen Alard having founded the north aisle chantry in 1312. I am grateful to Loveday Gee of the University of Warwick for information on the Winchelsea tombs and chantries : see further her ‘English Monumental Tomb Design in the Province of Canterbury, c.1280-1350’ (University of East Anglia, unpublished M.Phil, thesis, 1976), pp. 66-67.

30 See section 5 for further links between Patrington and the Continent, and section 6 in the forthcoming second part of the article for more detail on mullions using undulating forms.

31 Adenauer, H., Die Kathedrale von Eaon (1934), p. 30 and pl. 1Google Scholar.

32 Quite often in the South East, a bead moulding is substituted for the additional fillet.

33 A small wave moulding is used repeatedly on the miniature piers of the east face of the reredos of Beverley Minster, but as these are carved in an imported stone (Purbeck marble), and seem to be cut from the same template as the marble piers of the pulpitum at Exeter (their marble bases are also identical), they have not been counted as a northern example. They were probably pre-fabricated at Corfe; the closest parallel for their unusually elaborate design of base is to be found in the Purbeck bases of the Wells retrochoir. However, given the element of doubt in their origin, they have been omitted from the list for region VII too.

34 The St Albans example, in the nave south aisle windows, is derived from Neale, J., The Abbey Church of St. Albans, Hertfordshire (1877), pl. 22 Google Scholar, and has apparently been modified since.

35 e.g. eastern Alard tomb and piscina/sedilia in south aisle, Winchelsea, after 1322, Fig. 5 G; Aymer de Valence tomb, Westminster, c. 1325; ‘Y’ panelling over the main windows of St Stephen’s chapel, early 1330s in execution; dado of St Anselm’s chapel, Canterbury Cathedral, 1334-36, Fig. 6 E; Edward II’s tomb, Gloucester, c. 1335. The St Stephen’s mullion is illustrated in Mackenzie, F. A., The Architectural Antiquities of the Collegiate Chapel of St. Stephen, Westminster (1844), pl. xv Google Scholar, 4: both this and the Alard tomb are actually examples of the second variety. The mouldings of the canopy of Edward II’s tomb are so close to those of St Anselm’s chapel and Archbishop Meopham’s tomb in Canterbury Cathedral that there can be little doubt that it is the work of a south-eastern mason : see Sections 8 and 9 in the forthcoming second half of this article for further parallels.

36 Small wave mouldings of about this size, usually arranged in pairs, also occur in Herefordshire and S. Shropshire at this period (e.g. Fownhope, Ludlow church, Pembridge).

37 Hope, V. and Lloyd, J., Exeter Cathedral (1973), p. 15 Google Scholar. The fact that there was no previous tradition at Exeter or in the South West for the use of the basic sunk chamfer suggests that it was already starting to be accepted in the south of England as a stock moulding by this date, probably spreading from the South East. In fact, John Harvey suggests that its occurrence at Exeter is linked with the arrival there of Master Thomas of Witney, who is recorded working on St Stephen’s chapel, Westminster, in the 1290s: for Witney’s career, see Harvey, J. H., The Mediaeval Architect (1972), pp. 133 sqqGoogle Scholar.

38 e.g. the ‘Somerset circle’ works at Lacock and at Dudley Castle, both c. 1540-50, illustrated in Clark-Maxwell, W. G., ‘Sir William Sharington’s work at Lacock, Sudeley, and Dudley’, Archaeological Journal, LXX (1913)Google Scholar, fig. 1. ‘Not frequently met with’ in this period, according to Gotch, J. A., Early Renaissance Architecture in England (1901), p. 114 Google Scholar.

39 Illustrated in R.C.H.M., Caernarvonshire, 11, 126, fig. 103 G.

40 See Morris, R. K., ‘The Remodelling of the Hereford Aisles’, J.B.A.A., 3rd ser., XXXVIII (1974), 32 sqq.Google Scholar. On the Continent, an isolated instance of this demi-version is in the mullion design of the nave south aisle windows of Strasbourg Cathedral, £.1240 sqq. (Fig. 5 E).

41 An isolated instance occurs in the nave of St Katharine’s at Oppenheim (after 1317), where the arches between the aisles and the side chapels incorporate a sunk chamfer that is a cross between the first and third types. This work postdates the earlier examples of the sunk chamfer in England, and may possibly be derived from them, though it is a development with no future on the Continent. See further section 5 for links between England and Oppenheim.

42 e.g. in area C, Evron; in D, Amiens (chevet), Laon (S. transept), Noyon (nave N. chapels), Soissons (N. transept); in E, Auxerre St Germain (chancel), St Thibault, Troyes St Urbain (chancel); in F, Notre-Dame (N. transept), Ste Chapelle; in G, Vendôme. It is not found regularly in the Rhineland (B), except in areas of obvious French influence such as Cologne Cathedral chevet and Strasbourg west façade (dado), but it is taken up quite frequently in late Gothic piers, responds, and ribs in the Low Countries (e.g. Antwerp Cathedral, east end; Hal, S. porch; Malines Cathedral, east end).

43 Of course, the closeness of northern French and English royal works in this period (e.g. the presence of Henry of Reyns at Windsor and Westminster) means that the first example of this moulding need not have occurred on French soil.

44 e.g. Bond, op. cit., pp. 666 (3), 667 (1, 3, 4, 5), 673 (10), 675 (1).

45 Thirteenth-century examples could be cited from all regions of northern France, but it seems to remain a particular feature of Normandy into the fourteenth-century, an area of obvious geographical significance for south-east England (e.g. Evreux Cathedral, east end, c. 1260 sqq.; Rouen Cathedral, north transept, c.1280 sqq., and Lady chapel, 1302 sqq.; St Ouen, 1318 sqq.).

46 Note, for example, its absence in the arcade arches of Westminster Abbey (Royal Commission on Historical Monuments, London, 1 (1924), 95) or of Lincoln nave and St Mary’s, York (Bond, op. cit., p. 668, 2 and 4 respectively), all cases where Continental influence might be expected.

47 An early Perpendicular instance is the chancel of St Mary’s, Warwick (c. 1370 sqq.), in the exterior arches of the window surrounds and elsewhere, thus complementing the use of a prototype sunk chamfer (version iii) for the main mullions of the east window.

48 Other recorded examples — Paris Ste Chapelle, dado of upper chapel (1240s); St Germer Lady chapel, piscina (1259 sqq.); Strasbourg Cathedral, openwork tracery of west front (early fourteenth century).

49 Neale, op. cit., pl. 22.

50 Ibid.

51 Mackenzie, op. cit., pl. xv, 4, mullions of exterior ‘Y’ panelling over main windows.

52 Harvey, Mediaeval Architect, 185, for moulding drawings.

53 Brakspear, ‘West Country School’, ut supra, pl. XVI.

54 In some instances, the profile of the constituent ogees becomes less dependent on the roll and fillet, with the concave part made larger relative to the convex part, and this makes its identification easier as a moulding in its own right (see Paley, op. cit., pp. 59-60, though the distinction between Decorated and Perpendicular examples is not as clear cut as he implies).

55 The St Albans example is illustrated in Neale, op. cit., pl. 53. The remodelling of the east end began around the crossing area as early as 1257, but stylistically the east bays of the choir aisles do not appear earlier than the last fifteen years of the century. The setting up of the shrine of St Alban in the new retrochoir did not take place until the time of Abbot John de Maryns, 1302-08; see Coldstream, N., ‘English Decorated Shrine Bases’, J.B.A.A., CXXIX (1976), 18 Google Scholar. The church at Tintern was finished in 1301 ( Harvey, J. H., William Worcestre, Itineraries, 1969, p. 61, note 2Google Scholar), and the bookroom door appears stylistically to be of about that date, judging from its tracery. The Exeter and Wells examples belong to the later 1280s or the early 1290s.

56 Morris, ‘Worcester Nave’. The Bristol inspiration at Ottery came via the later work on the east end at Wells, where the double ogee makes a belated appearance in the window frames of the two west bays of the north choir aisle (probably 1330s).

57 The curious treatment of the fillets is determined by the triangular plan of the cross base below. Fillets on the same plane as the double ogee occur occasionally in Perpendicular work (e.g. Paley, op. cit., pl. VIII, 19), but there seem to be no other recorded examples in the Decorated period.

58 Fernie, E. C. and Whittingham, A. B., ‘The Early Communar and Pitancer Rolls of Norwich Cathedral Priory ...‘, Norfolk Record Soc, XLI (1972), 30 sqqGoogle Scholar. The Norwich example must surely be related to the Bristol example, which looks so much like it; exactly how is unclear, but the court is the most likely intermediary, and would also provide a common ground with the Geddington cross. For further links between the west and East Anglia, see Morris, ‘Worcester Nave’, and idem, ‘Pembridge and Mature Decorated Architecture in Herefordshire’, Transactions of the Woolhope Naturalists’ Field Club, XLII, part 11 (1977).

59 See Harvey, J. H., English Mediaeval Architects: a Biographical Dictionary down to 1550 (1954), pp. 213 sqq.Google Scholar, and Fernie and Whittingham, op. cit., pp. 30 sqq.

60 For illustrations of all the relevant south-eastern examples from this period, see Harvey, Medieval Architect, pp. 184-85; the example in the vestry arch at Windsor is, admittedly, tenuous, for the double ogees continue into bead mouldings rather then terminating in fillets.

61 e.g. the keep at Warkworth, c. 1420 sqq.; Tattershall great tower, c.1435 sqq.; the Hastings tower at Ashby de la Zouch, licence to crenellate, 1474.

62 For example, in the three major medieval churches surviving in Coventry, where almost all the work dates from the later fourteenth century on, the double ogee appears only in the early sixteenth-century Marier Chapel in Holy Trinity.

63 I am indebted to John Harvey for this suggestion: and see further, Harvey, J. H., The Gothic World (1950), p. 77 Google Scholar.

64 Illustrated in Durand, G., Monographie de l’Eglise ÌSlotre-Dame Cathédrale d’Amiens (1901), p. 295 Google Scholar.

65 The demi-roll and fillet, though not directly pertinent to the ogee, is an important moulding to pursue in its own right, for it becomes a characteristic method in the Flamboyant style for terminating a jamb or respond at the junction with a wall surface (e.g. the fifteenth-century work at Alençon, Caudebec, and Vendôme). Early small-scale examples occur in eastern France (e.g. Troyes St Urbain, porches, 1262 sqq.; St Thibault, presbytery, c. 1290 sqq.), whence it develops into a bolder and more permanent feature in regions B (e.g. Soest Wiesenkirche, after 1331; Strasbourg, St Katharine’s chapel, c. 1340; Aachen presbytery, 1355 sqq.) and then A (e.g. Antwerp Cathedral east end, 1352 sqq.; Malines Cathedral, east end).

66 That is, not continuing directly into an adjacent curve.

67 Though in fact the ogee moulding does not seem to have been used in the Ste Chapelle.

68 In region C, Evron (east end), Rouen St Ouen (east end), St Germer (Lady chapel); in D, Laon Cathedral (east end chapels and south transept rose); in E, Auxerre Cathedral (nave piers, c. 1320 sqq.) and St Germain (east end), Troyes St Urbain (e.g. Fig. 7 K); in H, Avignon papal palace (hall, after 1342). It also occurs in French Rayonnant inspired works in B, e.g. Strasbourg nave (dado at west end, c.1275); Worms Cathedral, St Joseph Chapel (immediately east of south porch, stylistically mid fourteenth century).

69 In region A, Antwerp Cathedral (east end), Hal (east end), Tongres (nave chapels, including the earlier, eastern ones); in B, Aachen (east end), Soest St Thomas, Worms Cathedral (St Joseph Chapel). It seems to be absent, however, from northern French Flamboyant.

70 Harvey, Mediaeval Architect, p. 185, illustrates the Wynford examples.

71 I am grateful to John Harvey for additions to this list.

72 See Morris, Thesis, op. cit., Chapter VII, especially figs 58, 59, 60.

73 Morris, ‘Worcester Nave’.

74 Note that in many Perpendicular instances, the hollow is a casement. A very early example of its use in a pier is in the east end of Bristol Cathedral, but it is not continued in the same form into the arcade arch.

75 See further Section 5. This is one of the parallels that underlies Harvey’s suggestion that the Hull transept is related to work of the Canterbury masons: Harvey, J. H., ‘The Origin of the Perpendicular Style’, in Studies in Building History, edited by Jope, E. M. (1961), p. 162 Google Scholar.

76 Brakspear, ‘West Country School’, pls XVI and XVII.

77 A unique exception occurs in the porch of St Christophe attached to the nave south chapel of that name at Amiens Cathedral, and dated with the chapel after 1358; Durand, op. cit., pp. 45-46 and fig. on 478.

78 Illustrated in Bond, op. cit., p. 668 (3).

79 Illustrated in R.C.H.M., London, 1, op. cit., 95 : drawings 32 and 39 in Forrester, op. cit., p. 33, are not accurate for these details.

80 Also the vestry arch off the Dean’s Cloister at Windsor, 1350, illustrated in Harvey, Mediaeval Architect, p. 180; and the arches of the shrine base of St Alban at St Albans, a miniature example in Purbeck marble, probably 1302-08.

81 A rare instance occurs in the rib design used throughout the nave aisles of Auxerre Cathedral (a keel with two lateral fillets, though not combined with three-quarter hollows), and the wall-rib type used throughout the east end of St Urbain at Troyes also appears to be of this design.

82 In a different context, the formation appears in the lesser ribs of St Stephen’s chapel undercroft (1292 sqq.), another link between Westminster and region III to add to those already observed in section 1 (wave mouldings, fourth variety).

83 Apart from these six tombs, other examples in decorative work among the twenty or so instances recorded are to be found in the dados of Canterbury Cathedral chapter house (1304 sqq.), Ely Lady chapel (1321 sqq.), Hereford chapter house (c. 1340s); in the sedilia at Tewkesbury (c. 1330s); in the pulpitum at Exeter (1317-26) and the reredos at Beverley Minster (1330s). It is also used for ribs in the east of England, e.g. the gatehouses at Battle (licence to crenellate, 1332) and Bury St Edmunds (c. 1340-50), and the miniature vault in the Beverley reredos.

84 If there is a link between Tintern and East Anglia, it may be through the patronage of Roger Bigod, Earl of Norfolk (1270-1306), who held lands in the lower Wye valley and was patron of the rebuilding at Tintern. For example, it is recorded that in 1291 Ralf, the master mason at his castle of Chepstow, returned to Bigod in Suffolk for consultation: Perks, J. C., Chepstow Castle (D.o.E., 1955), 8–9 Google Scholar.

85 Illustrated in Harvey, ‘The Origin of the Perpendicular Style’, 143, fig. 6 (D.2), after Mackenzie.

86 Neale, op. cit., pls 19 and 41.

87 The Winchester stalls are East Anglian work, not only because they are documented as being by a Norfolk carpenter, William Lyngwode, but also because their style of moulding links closely with that of the east cloister walk at Norwich (compare Fig. 8 F and G). A particular trait in the arches of both is the use of a form that is half-way between a fillet and a bead (rather like a beaked half roll) to terminate some of the three-quarter hollows. Work on the stalls is documented in 1308 and 1309 (Harvey, Dictionary, p. 175).

88 Sharpe, E., A Treatise on the Rise and Progress of Decorated Window Tracery in England (1849), pl. E, no. 47 Google Scholar.

89 Ibid., pl. E, no. 43.

90 Illustrated in Harvey, ‘The Origin of the Perpendicular Style’, p. 151, fig. 8. 8, detail C.4.

91 The east window of the south aisle and another in the north aisle, illustrated in Neale, op. cit., pl. 55.

92 The other recorded examples are: Westminster Abbey cloister, S. part of E. walk (head of tracery only, Fig. 10 J); St Stephen’s chapel (upper); Norwich, bishop’s palace chapel (re-used Decorated E. window); Carlisle, triforium (openwork tracery); Beverley Minster, Percy tomb (rib); Lichfield, Lady chapel (if the Scott restoration is reliable on this detail); Tewkesbury, east end clerestory (head of tracery only); Wells, east end clerestory; Ottery St Mary, Lady chapel (Fig. 9 B). For St Stephen’s, see Mackenzie, op. cit., pl. xv, 9; in the same plate (nos. 5 and 10) are illustrated examples of a similar design, except that the undercut part of the hollow chamfer terminates in a sharp ridge rather than a fillet. This idea is found earlier in the retrochoir of St Albans (Neale, op. cit., pl. 56); and, using a chamfer instead of a hollow chamfer, in some of the mullions of the choir screen in Canterbury Cathedral, 1304 sqq. (Fig. 9 D). Also, in the west, the chamfered form of mullion, terminating in a fillet, appears in the retrochoir windows at Wells (c. 1320 sqq.); and, terminating in a sharp ridge, in the Lady chapel windows there (c.1310 sqq.), followed up in the ribs of the Lady chapel screen at Ottery St Mary (1337 sqq.). Another variation occurs in the nave north arcade arches of Herne church in Kent, with a sunk chamfer instead of a hollow chamfer, almost certainly related to the usage of the sunk chamfer in parts of the remodelling of St Anselm’s chapel, Canterbury, 1334-36 (e.g. Fig. 6 E).

93 In the north bays of the east walk, it is used for the interior profile of the cloister tracery, the door to the former slype, and the tierceron ribs; in the other three walks, the tracery and rib usage continues, and in addition it appears in the exterior profile of the tracery. See Fernie and Whittingham, op. cit., 32-33, where the introduction of the undercut hollow chamfer is associated with the arrival of John Ramsey as master mason in c. 1307-08, and where the final execution of these bays (or parts of them, especially the vaults) is put as late as c. 1325-27, by which time both William and John Ramsey are documented as involved. I am very grateful to Eric Fernie for clarifying the distribution of this moulding in the cloister. It is interesting to note that the form does not seem to occur in any other works in the precinct (e.g. the Carnary chapel), except at the bishop’s palace.

94 See further Pernie and Whittingham, op. cit. The only other recorded example in the area in the Decorated period is at Ely, in the rere-arches of the lateral windows of the Lady chapel.

95 As stated in Harvey, ‘The Origin of the Perpendicular Style’, p. 152: see ibid., p. 139, fig. 8.3 (detail C.4), for the Old St Paul’s example.

96 See further Morris, ‘Worcester Nave’, op. cit.

97 This area of the work at Tintern dates between 1288 and 1301 (see J. H. Harvey, Wm. Worcestre, p. 61). If, as seems likely, this form of undercut hollow chamfer is a development of the first type, which appeared in the east in the 1280s, it is tempting to assign this isolated instance in the west to the Bigod connexion with Tintern (see further note 84).

98 The implications of this connexion were noted in the previous section, note 75. A closely related form was the design of the main mullions of the exterior ‘Y’ panelling over the main windows of St Stephen’s chapel, Westminster, probably executed in the 1330s. In these, the only difference is the use of small sunk chamfers instead of hollow chamfers (Mackenzie, op. cit., pl. xv, 4).

99 The Angel choir rib is derived from the designs of the various rib types employed in the nave, and thus the idea is present at Lincoln before the mid-century. Related designs also occur in the nave at Grantham, a derivative of Lincoln (e.g. north aisle, exterior window surrounds). The hypothesis that Lincoln in particular favoured these forms at a relatively early date finds some confirmation in the observation in section 4, that it also seemed to be the main centre for the type of three-quarter hollow formation illustrated in Fig. 8 D).

100 Mackenzie, op. cit., pl. XIV, B.

101 See Morris, R. K., ‘Tewkesbury Abbey: the Despenser Mausoleum’, Transactions of the Bristol and Gloucestershire Archaeological Society, XCIII (1974), 150 sqq.Google Scholar; and Taylor, A. J., ‘Building at Caerphilly in 1326’, Board of Celtic Studies Bulletin, XIV, 4 (1952), 299–300 Google Scholar.

102 The Amiens mullion type is used for the clerestory windows at Tournai, and also a little later in the chapels of the south choir aisle at Laon Cathedral; see further section 6, in the second half of this article.

103 The observation of the transverse arch was made from the ground, and needs to be confirmed by a closer inspection. The tracery of the Frankfurt south transept consists of simple Rayonnant designs employing quatrefoils in roundels and trefoils (as in the nave north aisle at Essen); not until the north transept do forms more typical of the fourteenth century appear, such as quatrefoils framed in curved-sided lozenges.

104 The window in the Franciscan church is a simple flowing design, probably of the early fifteenth century, but the Soest examples belong to the first half of the fourteenth.

105 The relevant tracery at SS Andrew (south transept) and Ursula (over south door of nave) is flowing, but in St Gereon (south transept) the simple Rayonnant forms look to be of c. 1300.

106 e.g. Antwerp Cathedral, 1352-c.1500, but relevant examples all seem to be fifteenth-century (nave north aisle, and clerestories of east end, transept, and nave); Hal, 1341/2-1410 (nave aisles and clerestory, and north-east chapel); also east end arcades at Antwerp and Malines (Fig. 7 L).

107 A similar example occurs in the east crossing arch of St Peter’s at Soest (1272 sqq.).

108 See section 6 in the second half of the article for further detail on this connexion, with regard to mullions employing a semicircular hollow flanked by fillets.

109 Yeovil, c.1380; Coventry, Holy Trinity (nave south aisle), c.1400?; Malvern Priory, c.1420 sqq.; to take a few random examples. The use of the moulding especially in the Low Countries for arcade arches and piers (Fig. 7 L), but also in the Rhineland (e.g. Schwäbish Gmünd, nave and east end; Cologne, the Antonines’ church, nave, both first half of the fourteenth century), appears in Perpendicular too (e.g. Coventry, Holy Trinity, chancel, and St Michael’s, Dyers’ chapel, both probably later fifteenth century). See further section 6, concerning mullions with polygonal terminations.

110 See further sections 6 and 7 in the second half of the article.

111 Mouldings employing a ridge or spike often occur in the same works as the freestanding fillet, as at Patrington, Selby, St Albans, St Stephen’s chapel, Canterbury Cathedral choir screen, and Ottery St Mary: see note 92.

112 Also note the spiked hollows of the mullion mouldings in the gallery chapel of Gloucester south transept, c. 1320s, are filled with ballflower, whereas this idea of the hollow containing ornament does not occur in any of the recorded eastern or Continental examples.

113 e.g. region B — Oppenheim, Schwäbisch Gmünd (nave), Worms (St Joseph chapel); C — Coutances (nave south chapels), Evron (ambulatory, east chapel); E — Auxerre Cathedral (nave, work of c.1320 sqq.), Chaumont (south porch), St Thibault (St Gilles chapel).

114 Aubert, M., ‘St. Thibault’, Congrès Archéologique deFrance, Dijon (1929), 252-66Google Scholar, dates it in the third quarter of the thirteenth century.

115 See further section 6 in the second half of this article.

116 Mackenzie, op. cit., pl. XIV, B. The other instances in the east are at Ely (Bishop Hotham’s west bays of the presbytery, and Lady chapel), and Beverley St Mary (north choir aisle), but the latter does not have a sharp arris. I am grateful to Nicola Coldstream for drawing my attention to the Beverley example.

117 Badgeworth (north chapel), Caerphilly Castle (hall), Worcester (refectory, continued in the early Perpendicular period in the north and south walks of the cloister).

118 See note 101.