m 
 
 D6DEDD 
 
 HRHDCIS • BODD 
 
Digitized by the Internet Archive 
 in 2013 
 
 http://archive.org/details/gotharchitectuOObond 
 
GOTHIC ARCHITECTURE 
 IN ENGLAND 
 
LICHFIELD CATHEDRAL FROM S.W. 
 
GOTHIC ARCHITECTURE 
 IN ENGLAND 
 
 AN ANALYSIS OF THE ORIGIN & DEVELOP- 
 MENT OF ENGLISH CHURCH ARCHITECTURE 
 FROM THE NORMAN CONQUEST TO THE 
 DISSOLUTION OF THE MONASTERIES 
 
 FRANCIS BOND, 
 
 M.A., LINCOLN COLLEGE, OXFORD; 
 
 Fellow of the Geological Society, I. 'union; 
 Honorary Associate of the Koval Institute of British Architects 
 
 WITH L254 ILLUSTRATIONS, COMPRISING 785 PHOTO- 
 GRAPHS, SKETCHES, AND MEASURED DRAWINGS, AND 
 469 PLANS, SECTIONS. DIAGRAMS, AND MOLDINGS 
 
 LONDON 
 
 B. T. BATSFORD, 94 HIGH HOLBORN 
 
 NEW YORK: CHARLES SCRIBNER'S SONS 
 
 1906 
 
Printed at The Darien Press, Edinburgh. 
 
PREFACE. 
 
 In the preparation of this work full use has been made of the materials 
 which have accumulated, both English and foreign. In all important 
 cases an attempt has been made to render due acknowledgment. A list 
 of the sources which have been drawn upon most freely will be found 
 on page viii. ; reference is made to many others in footnotes in the 
 course of the book. But, in addition, the writer has to acknowledge 
 the ready assistance of many friends who have obtained information 
 or verified data for him on the spot. As to the illustrations he is under 
 special obligation to members of the architectural profession for the 
 ready and generous assistance they have given. The difficulty has 
 been to select from the valuable material placed at his disposal. 
 
 For plans, sketches, moldings, or measured drawings his acknow- 
 ledgments are due to Mr Maurice B. Adams, f.r.i.b.a. ; the Com- 
 mittee of the Architectural Association Sketch Book ; Mr H. J. 
 Austin; Messrs G. Bell & Son ; Mr W. H. Bidlake, m.a.; Mr J. Bilson, 
 f.s.a. ; the Council of the Cambridge Antiquarian Society ; the Dele- 
 gates of the Clarendon Press, Oxford ; the Rev. Canon Church, m.a. ; 
 the Rev. R. Corrie Castle; Mr J. J. Creswell, A.R.I.B.A. ; Mr Reginald 
 Fowler; Mr G. Frisch, a.r.i.b.a. ; Mr S. K. Greenslade, a.r.-i.b.a. ; 
 Lord Grimthorpe ; Mr T. G. Jackson, m.a.; Mr Montague Rhodes 
 James, litt.d. ; Mr C. Henman, a.r.i.b.a.; Mr Gerald C. Horsley ; 
 Mr A. H. Kersey, f.r.i.b.a.; Mr J. Langham ; Mr |ohn Murray; 
 Mr J. T. Micklethwaite, f.s.a.; Mr J. Norton; Mr A. Y. Nutt; Mr 
 H. A. Paley, a.r.i.b.a. ; Mr Roland W. Paul, f.s.a.: Mr H. Phibbs ; 
 Professor Beresford Pite, f.r.i.b.a.; Mr E. S. Prior, m.a.: Mr H. A. 
 Prothero, m.a. ; Mr Harbottle Reed; the Council of the Royal Institute 
 of British Architects; Mr J. Oldrid Scott, f.s.a.; Mr C. Wontner 
 Smith, a.r.i.b.a. ; Professor Flsey Smith, m.a. : the Committee of the 
 Society of Antiquaries ; Mr Charles Spooner; Mr Russell Sturgis, m.a., 
 ph.d.; Mr Sydney Vacher, a.r.i.b.a.; Mr H.D.Walker ; Mr F. S.Waller, 
 f.r.i.b.a. ; Mr W. G. Watkins, a.r.i.b.a. ; Mr W. S. Weatherley, f.s.a. ; 
 
vi PREFACE. 
 
 Mr A. Needham Wilson, a.r.i.b.a. ; as well as to others with whom it 
 has been found impossible to communicate. 
 
 A large number of photographs has been placed at his disposal ; 
 and though they necessarily lose in reproduction by mechanical pro- 
 cess, the results show how excellent in many cases were the originals. 
 He is indebted for the use of photographs to Dr F. J. Allen; Rev. 
 W. Tuzo Alston; Mr W. G. Bannister; Mr R. H. Barker; Mr F. 
 Bligh Bond, f.r.i.b.a. ; Mr R. P. Brereton, m.a. ; Dr Oscar Clark ; 
 Mr I. S. Collings ; Mr W. Davidson ; Messrs Dawkes & Partridge ; 
 Mr J. P. Freeman ; Mr S. Gardner ; Mr J. Pattison Gibson ; Mr 
 Donald Gooding; Rev. T. Gough ; Mr E. Gunn, a.r.i.b.a.; Mr 
 
 C. C. Hodges; Mr G. H. Lovegrove; Rev. T. Perkins; Dr H. W. 
 Pigeon; Rev. H. Bedford Pirn; Mr Alan Potter; Rev. G. K. 
 Saunders; Mr F. R. Taylor; Mr G. H. Tyndall ; Mr E. H. Walker; 
 Mr E. W. M. Wonnacott, F.S.I. 
 
 As the preparation of the work advanced, the importance of 
 liberality of illustration became increasingly apparent. It is only right 
 to acknowledge the readiness with which Mr Batsford seconded the 
 author in his wish to widen the scope of the book and to bring it 
 out in worthy form. Special acknowledgment is due to Mr Harry 
 Batsford ; his interest in the subject and acquaintance with archi- 
 tectural literature made his assistance of great value. The whole of 
 the moldings, diagrams, plans, and sections have been drawn by Mr 
 L. R. Stains. Sections are drawn to a uniform scale ; the plans of 
 the parish churches, and that of St Gall, are drawn to half the scale 
 of those of the greater churches. The text has had the advantage of 
 the revision and criticism of Mr John Bilson, from whose sound and 
 accurate scholarship the writer has benefited at all stages of its pre- 
 paration. Various portions of the proofs have been revised by Mr T. 
 
 D. Atkinson, m.a. ; Mr S. B. Beale, a.r.i.b.a.; Mr Harold Brakspear, 
 f.s.a. ; Mr R. P. Brereton, m.a.; Mr J. N. Comper ; Mr J. J. Cress- 
 well, a.r.i.b.a.; Rev. R. A. Davis; Mr C. H. Grinling, m.a.; Mr E. 
 M. Hick; Mr W. H. St John Hope, m.a. ; Mr G H. Lovegrove ; Mr R. 
 Phene Spiers, f.s.a. ; and Mr E. W. M. Wonnacott, f.s.i., to all of 
 whom the writer is indebted for suggestions and criticisms of much 
 value. Not seldom, however, he has ventured to disregard their advice, 
 and has remained of the same opinion still ; for all the shortcomings of 
 the text, therefore, he alone is responsible. Valuable assistance has 
 been rendered by Rev. R. A. Davis in the preparation of the index. 
 
 The student and archaeologist will find in Chapter XLII. a dated 
 list of English buildings arranged in alphabetical order. Such a list 
 
PREFACE. vii 
 
 should be of great service to all who are interested in the history of 
 English architecture. The preparation of this list has involved much 
 labour ; but it is obvious that the first draft of such a chronology 
 cannot be free from imperfections and inaccuracies. The writer will 
 welcome any corrections or additions to it. 
 
 To the architectural student it is hoped that the twenty-eight 
 sheets of moldings will be found specially valuable. In the largest 
 collection hitherto published, that by Mr Paley, the moldings are very 
 minute and crowded together, nor are they to the same scale ; yet 
 it makes all the difference whether, for instance, a capital and arch 
 come from a piscina or a pier arcade ; several species ol moldings are 
 omitted altogether, e.g. those of vaulting ribs, basement courses, door- 
 ways and windows ; and of those which are illustrated the locality from 
 which they come is in many cases not indicated. 
 
 Of the other collections, that in Sharpe's Architectural Parallels 
 is of great value, but it is contained in an expensive book long out of 
 print ; nor does it illustrate any moldings later than the fourteenth 
 century ; that in Sharpe's Mouldings of the Six Periods of British 
 Architecture extends up to the Reformation, but was never finished. 
 The present collection gives a conspectus of English moldings from 
 the middle of the twelfth to the middle of the sixteenth century ; they 
 are drawn boldly and clearly ; they are to the same scale ; the locality, 
 as far as possible, of each is given ; molds of ribs, basement courses, 
 doorways and windows have been included ; and three sheets have 
 been added of the plans of piers, as well of the greater churches as of 
 the parish churches. It has been attempted, moreover, to show the 
 correlation of cognate members. A sheet has been prepared to show 
 the relation of the pier on the one hand to the base and plinth, on the 
 other to the abacus or capital and arch. In the same way illustra- 
 tions have been inserted to show the relation of the arch to the jambs 
 of doorways and windows. These architectural members are not 
 complete in themselves ; each is part of a group, and should not only 
 be beautiful in itself, but should fit the position it occupies as a 
 member of that group or whole. The co-ordination of the various 
 members of the pier and arch has hardly ever been systematically 
 illustrated except in Messrs Johnson and Kersey's valuable Churches 
 of the Nene Valley, to the authors of which special acknowledgment 
 is due. The various members have not always been illustrated on the 
 same sheet. To facilitate reference, however, all the illustrations, 
 including the moldings, have been indexed alphabetically (709-738). 
 In many cases also it will be found that a photograph has purposely 
 
viii PREFACE. 
 
 been given as well as a drawing, e.g. of the foliated capital of West 
 Walton. The index to the illustrations, therefore, should constantly 
 be consulted. In the same way a vault is often shown both in per- 
 spective and in plan, e.g. that of the choir of Oxford Cathedral (331); 
 as also the piers, e.g. that of the nave of Norwich Cathedral (238 and 
 659). It may be added that the photographic representation of vaults 
 on plan has not hitherto been attempted in an architectural treatise (see 
 327-334), and it is believed that this will greatly clear up the intricacies 
 of rib construction. In many later vaults, indeed, e.g. that of the nave 
 of St George's, Windsor (330, 332), the construction is utterly unin- 
 telligible as the vault is usually seen, i.e. in perspective. In conclusion, 
 the writer begs the student to believe that no collection of moldings 
 will absolve him from the task, as delightful as it is indispensable, 
 of drawing- them for himself in situ. 
 
 The following are the Titles of Authorities quoted summarily 
 in the course of the Text. 
 
 Anderson, VV. J., and R. Phene Spiers. The Architecture of Greece and Rome. 
 
 London, 1902. 
 Architectural Publication Society. Dictionary of Architecture. 7 vols. London, 
 
 1849 t0 1892. 
 Architecture and Building, Dictionary of. Edited by Russell Sturgis. 3 vols. New 
 
 York, 1 90 1. 
 
 Barry, E. Lectures on Architecture. 1881. 
 Bell. Series of English Cathedrals. London, 1896- 1904. 
 Beckett, Sir E. Book on Building. 2nd edition. London, 1880. 
 Billings, R. W. Carlisle Cathedral. London, 1840. 
 
 Durham Cathedral. London, 1843. 
 
 Kette?-ing Church. London, 1843. 
 
 Temple Church. London, 1838. 
 
 Bilson, John. The Beginnings of Gothic Architecture. Journal of the Royal Institute 
 of British Architects. March n and 25, April 15, 1899, and May 10, 1902. 
 
 ■ Chapter House of Beverley Minster. Archa;ologia, liv. 425. 
 
 - On the Recent Discoveries at the East End of the Cathedral Church of Durham. 
 Archaeological Journal, liii. 1-18. 
 — Beverley Minster. Architectural Review, hi., 197-204 and 250-259. 
 Bloxam, M. H. Gothic Ecclesiastical Architecture, nth edition. 3 vols. London, 
 
 1882. 
 Bond, Francis. English Cathedrals Illustrated. 3rd edition. London, 1903. 
 
PREFACE. ix 
 
 Bond, Francis. On the Comparative Value of Documentary and Architectural Evidence 
 
 in establishing the Chronology of the English Cathedrals. Journal of the Royal 
 
 Institute of British Architects. November 21, 1898. 
 - Classification of Romanesque Architecture. Journal of Royal Institute of British 
 
 Architects. April 22, igoi. 
 Bowman, H., and Crowther, J. S. Churches of the Middle Ages. 2 vols. London, 1850. 
 Boyle, J. R. Holy Trinity Church, Hull. Hull, 1S90. 
 
 Brakspear, Harold. Hayles Abbey Church. Archfeological Journal, lviii. 350-357. 
 — On the First Church at Furness. Transactions of the Lancashire and Cheshire 
 
 Antiquarian Society, xviii. 
 Lacock Abbev Church. Archaeological Journal, lvii. 1-9. 
 
 Lacock Abbey. Archaeologia, lvii. 125-158. 
 
 Burnham Abbey. Archaeological Journal, lx. 294-317. 
 
 Waverley Abbey. Surrey Archaeological Society, 1905. 
 
 Beaulieu Abbey. Archaeological Journal. 1905. 
 
 Brandon, R. and J. A. Analysis of Got/tick Architecture. 2 vols. London, 1847. 
 
 Open Timber Roofs of the Middle Ages. London, 1849. 
 
 Parish Churches. London, 184S. 
 
 Briti'ON, John. Architectural Antiquiti's of Great Britain. 5 vols. Lond., 1807-1S35. 
 
 Cathedral Antiquities of Great Britain. 6 vols. London, 1814-1835. 
 
 Brown, G. Baldwin. From Schola to Cathedral. London, 1886. 
 
 The Arts in Early England. 2 vols. London, 1 903. 
 
 Browne, Willis. Survey of the Cathedrals of York, Durham, 6>r. 1727. 
 Brutails, J. A. Darcheologie du m<>ven age ct ses mcthodes. Paris, 1900. 
 Buckler, George. Twenty-two Churches of Essex. London, 1856. 
 " Builder," The. Cathedrals of England and Wales. London, 1894. 
 Butler, W. Measured Drawings of Christ Church, Dublin. 1874. 
 
 Christ Church, Dublin. London, 1901. 
 
 Butterfield, W. Shottcsbrooke Church. London, 1S44. 
 
 Carpenter, R. H. Sherborne Abbey Church. Journal of the Royal Institute of British 
 
 Architects. March 19, 1877. 
 Carter, J. Ancient Architecture of England. London, 1795. 
 
 Plans and Drawings published by the Society of Antiquaries, 1807. 
 
 Cattaneo, R. /.'architecture en Italic du I'D au XD Steele. Traduction par M. le 
 
 Monnier. Venise, 1890. 
 Caumont, A. de. Abccedaire, ou Rudiments d'archeologie. 3 vols. 1858-1862. 
 Caveler, W. Specimens of Gothic Architecture. 2nd edition. London, 1839. 
 
 Warmington Church. London, 1850. 
 
 Choisy, A. Histoire de P Architecture. 2 vols. Paris, 1899. 
 
 Dart de batir chez les Remains. Paris, 1873. 
 
 Dart de batir chez les Byzantins. Paris, 1893. 
 
 Christian, E. Skelton Church, Yorkshire. London, 1846. 
 Churches of the Archdeaconry of Northampton. Oxford, 1849. 
 Colling, J. K. Details of Gothic Architecture. 2 vols. London. 1856. 
 ■ Gothic Ornaments. 2 vols. London, 1850. 
 
 English Mediieval Foliage. London, 1874. 
 
 Colson, J. B. Reparations of the Roof of Winchester Nave in 1896. Winchester, 1899. 
 
 Conder, E. L. I-ong Me/ford Church. London, 1887. 
 
 Cox, Rev. J. C-, LL.D., Churches of Derbyshire. 4 vols. London, 1875-1879. 
 
x PREFACE. 
 
 Cox, Rev. J. C, and Sergeantson, Rev. R. M. Church of the Holy Sepulchre, North- 
 ampton. Northampton, 1897. 
 Craddock, Thomas. Peterborough Cathedral. Peterborough, 1874. 
 Cresy, E. Stone Church, Kent. London, n.d. 
 
 Dartein, F. de. D architecture lombarde. 1865-1882. 
 
 Dehio and von Bezold. Die Kirchliche Baukunst des Abendlandes. 2 vols., text ; 601 
 
 plates. Stuttgart, 1 884-1 901. 
 Dollman, F. T. Church of St Mary Overie, Southwark. London, 1881. 
 
 Analysis of Ancient Domestic Architecture. 2 vols. London, 1861. 
 
 Enlart, Camille. Origines francaises de I' architecture gothique en Italie. Paris, 1894. 
 
 Manuel d'arche'o/ogie francaise. Vol. I. Architecture religiei/se. Paris, 1902. Vol. 
 
 II. Architecture civile et militaire. 1904. (Unless otherwise specified, the refer- 
 ences in the text are to Vol. I.) 
 
 Fen and Marshland Churches. Wisbech, n.d. 
 
 Fergusson, J. History of Architecture in all Countries. 2 vols. 3rd edition. Edited 
 
 by R. Phene Spiers. London, 1893. 
 Ferrev, B. Christ Church, Hants. London, 1834. 
 Freeman, Archdeacon. Architectural History of Exeter Catl/ed/al. 2nd edition. 
 
 Exeter, 188S. 
 Freeman, E. A. Window Traceiy. Oxford, 185 1. 
 
 Garbett, E. L. Principles of Design in Architecture. 7th edition. London, 1891. 
 Gardner, J. Starkie. Ironworh. London, 1893. 
 Godwin, E. W. Bristol Cathedral. Archaeological Journal. Vol. 20. 
 Greenwell, Canon W. Durham Cathedral. 4th edition. Durham, 1892. 
 Grimthorpe, Edmund, Lord. St Albans Cathedral and its Restoration. 2nd edition. 
 St Albans, 1893. 
 
 Hadfield, J. Ecclesiastical, Castellated, and Domestic Architecture i7i Essex. London, 
 
 1848. 
 Hodges, C. C. Hexham Abbey. London, 1888. 
 
 Blyth Priory Church. 188 r. 
 
 Hope, W. H. St John. Alnwick Abbey (White Canons). Yorkshire Archaeological 
 
 Journal. 1887. 
 ■ Canterbury, St Pancras. Archaeologia Cantiana. Vol. 25. 
 
 Canterbury, Inventories of Christ Church (with J. W. Legg). London, 1902. 
 
 Castle Acre Priory (Cluniac). Norfolk and Norwich Archaeological Society. 1904. 
 
 - Dale Abbey (White Canons). Derbyshire Archaeological Society, i. 100, 
 
 and ii. 128. 
 
 - Fountains Abbey (Cistercian). Yorkshire Archaeological Journal, xv. 269-402. 1900. 
 
 Furness Abbey (Cistercian). Transactions of Cumberland and Westmorland 
 
 Archaeological Society. Vol. xvi. 
 
 Notes on the Abbey Church of Glastonbury. Archaeological Journal, lxi. 185-196. 
 
 1904. 
 
 Gloucester Abbey (Benedictine). Records of Gloucester Cathedral, iii. 1. 
 
 Hulne (White Friars). Archaeological Journal. 1890. 
 
 Lewes Priory (Cluniac). Archaeological Journal, xl. 
 
 Architectural History of the Cathedral, Church, and Monastery of Rochester. London, 
 
 1900. 
 
PREFACE. xi 
 
 Hope, W. H. St John. St Agatha's Abbey, Richmond (White Canons). Yorkshire 
 
 Archaeological Journal, x. 117-158. 1887. 
 St Radigund's Priory (White Canons). Archaeologia Cantiana, xiv. 140. 
 
 — Watton Abbey (Gilbertine). Archaeological Journal, lviii. 1. 
 
 ■ West Langdon (White Canons). Archasologia Cantiana, xv. 59. 
 
 Hubsch, H. Monuments de F architecture chretienne. Paris, 1S66. 
 
 Johnson, J. Reliques of Ancient English Architecture. London, n.i>. 
 
 Johnson, R. J. Specimens of Early French Architecture. London, 1864. 
 
 King, T. H. Study Book of Mediaeval Architecture and Art. 4 vols. London, 1858. 
 
 Lasteyrie, Comte Robert de. Discours sur les origines de F architecture gothique. Caen, 
 1901. 
 
 - Crypte de St Martin, Tours. Memoires de l'academie des inscriptions et belles 
 lettres. Tome xxxiv., Part I. Paris, 1891. 
 Lethap.y, W. R. Medieval Art. London, 1904. 
 Livett, Rev. G. M. Southwell Minster. Southwell, 1883. 
 Longman, W. St Pauls Cathedral. London, 1S73. 
 
 Micklethwaite, J. T. Westminster Abbey. Arch. Journal. Vol.51. 
 Murray. Cathedrals of England and Wales. 8 vols. London, 1861-1873. 
 
 Neale, J. St Albans Abbey. London, 1877. 
 
 Nene Valley, Churches of. Edited by E. Sharpe, J. Johnson, and A. H. Kersey. London, 
 1880. 
 
 Pai.ey, F. A. Manual of Gothic Moldings. 4th edition. London, 1877. 
 
 Manual of Gothic Architecture. London, 1846. 
 
 Parker, J. H. Glossary of Gothic Architect///-,:. 5th edition. 3 vols. Oxford, 1850. 
 
 — Guide to Architectural Antiquities in the Neighbourhood of Oxford. Oxford, 1846. 
 
 Great Haseley Church, Oxon. Oxford, 1S40. 
 
 Dorchester Church, Oxoi/. Oxford, 1845. 
 
 Petit, Rev. J. L. Remarks on Church Architecture. London. 2 vols. 1S41. 
 
 Boxgrove Priory. Chichester, 1861. 
 
 Potter, Joseph. Tintern, Buildwas, and Wenlock Abbeys. London. 1849. 
 
 Prior, E. S. History of Gothic Art in England. London, 1900. 
 
 Pugin, A. Specimens of Gothic Architecture. 2 vols. London, 1821. 
 
 Pugin, A. and A. W. Examples of Gothic Architecture. 3 vols. London, 1838- 1840. 
 
 Pugin, A. W. True Principles of Pointed or Christian Architecture. London, 1841. 
 
 Chancel Screens and Rood Lofts. London, 185 1. 
 
 Reeve, J. A. Fountains Abbey. London, 1892. 
 
 RlCKMAN, T. Styles of Architecture in England. 7th edition. London, 18S1. 
 
 Ruprich-Robert. L' architect ii re normande aux XI' et XI P slides en Norma ndie et en 
 
 Angleterre. 2 vols. Paris, n.d. 
 Ruskin, J. Seven Lamps of Architecture. 3rd edition. Orpington, 1891. 
 
 Saint- Paul, Anthyme. Viollet le-Duc, ses travaux d'art et sou systeme archeologique. 
 2nd edition. Paris, 1881. 
 
 ■ Histoire monumentale de la France. 4th edition. Paris, 1895. 
 
 Scott, Sir G. G. Gleanings from Westminster Abbey. 2nd edition. Oxford, 1863. ^ 
 
 Lectures on the Rise and Development of Medimval Architecture. 2 vols. London, 
 
 1879. 
 Scott, G. G. (Jun.). Essay on the History of English Church Architecture. London, 1SS1. 
 
xii PREFACE. 
 
 Sharpe, Edmund. Architectural Parallels in the Twelfth and Thirteenth Centuries, selected 
 from Abbey Churches. London, 1848. 
 
 Supplement to ditto, containing full-sized moldings. London, T848. 
 
 - Mouldings of the Six Periods of British Architecture. 3 parts. London, 187 1. 
 
 Decorated Window T?-acery in England. 2 vols. London, 1849. 
 
 Lincoln Excursion of Architectural Association. London, 1871. 
 
 Architecture of the Cistercians. Journal of the Royal Institute of British Architects. 
 
 June 19, 1 87 1. 
 
 Ornamentation of the Transitional Period. London, 1876. 
 
 New Shorehani Church, Sussex. Chichester, 1861. 
 
 Seven Periods of English Architecture. 3rd edition. London ; 1888. 
 
 Sketch-Books. Architectural Association. 32 vols. 1867-1904. 
 
 ■ Abbev Square. 3 vols. 1872. 
 
 John o' Gaunt. 3 vols. 1874-1879. 
 
 Spring Gardens. 8 vols. 1 866-1 890. 
 
 Statham, H. H. Architecture for General Readers. London, 1895. 
 Stewart, Rev. D. J. Architectural History of Ely Cathedral. London, 1868. 
 
 Architectural Histo7-y of Norwich Cathedral and Cloister. Archaeological Journal. 
 
 Vol. 32. 
 Stokes, Margaret. Early Christian Architecture in Ireland. London, 1878. 
 
 Early Christian Art in Ireland. 2 vols. London, 1887. 
 
 Viollet-le-Duc. Dictionnaire raisonn'e de I 'architecture francaise du XI e au XVI siecle. 
 10 vols. Paris, 1858-1868. 
 
 Walcott, Rev. M. E. C. Church and Conventual Arrangement. London, 1861. 
 
 Waller, F. S. Gloucester Cathedral. 1856. 
 
 Weale. Quarterly Papers on Architecture. 4 vols. 1 844-1 845. 
 
 Whewell, Rev. W. Notes on German Churches and Normandy. Cambridge, 1835. 
 
 Wickes, C. Spires and Towers of England. 3 vols. London, 1853. 
 
 Willis, Rev. R. Architecture of the Middle Ages, especially of Italy. Cambridge, 1835. 
 
 Canterbury Cathedral. 1845. 
 
 Chichester Cathedral. Chichester. 1861. 
 
 York Minster. Archaeological Institute. York volume, 1846. 
 
 Worcester Cathedral. Archaeological Journal. Vol. 20. 
 
 Worcester Monastery. Archaeological Journal. Vol. 20. 
 
 ■ Winchester Cathedral. Archaeological Institute. Winchester volume, 1846. 
 
 Lichfield Cathedral. Archaeological Journal. Vol. 18. 
 
 Facsimile of the Sketch Book of Wilars de Honecort. 1859. 
 
 On the Construction of the Vaults of the Middle Ages. Transactions of the Royal 
 
 Institute of British Architects. Vol. I., Part II. 1842. 
 
 Architectural History of Glastonbury Abbey. Cambridge, 1866. 
 
 ■ Architectural Nomenclature of the Middle Ages. Cambridge, 1844. 
 
 — and J. W. Clarke, M.A. Architectural History of the University of Cambridge. 
 Cambridge, 1886. 
 
 Yorkshire Churches. Leeds, 1844. 
 
 Winters, W. Waltham Abbey. Waltham, 1888. 
 
TABLE OF CONTENTS. 
 
 PART I.— THE ORIGIN AND DEVELOPMENT OF THE 
 MEDI/EVAL CHURCH ARCHITECTURE OF 
 ENGLAND. 
 
 PAGE 
 
 INTRODUCTION - - xvii 
 
 CHAPTER 
 
 I. Definitions of basilican, byzantine, Romanesque, and 
 
 GOTHIC ARCHITECTURE - - I 
 
 II. Characteristics of English Romanesque architecture, 
 
 1050 — c 1200 - 14 
 
 III. Characteristics of English gothic architecture, 
 
 c. 1 1 70— c. 1538 44 
 
 IV. Ditto, c 1170 — c. 1315 - 65 
 V. Ditto, c. 1300 — c. 1350 - 80 
 
 VI. Ditto, c. 1330 — c. 1538 - 88 
 
 VII. Chronological history of the greater English 
 
 churches - 97 
 
 PART II.— AN ANALYSIS OF THE MEDI/EVAL CHURCH 
 ARCHITECTURE OF ENGLAND. 
 
 VIII. The basilican plan 145 
 
 IN. The planning of the eastern limb of the cathedral, 
 
 monastic and collegiate churches - 159 
 
 N. The choir, the saint's chapel, the eastern transept, 
 
 THE CRYPT OF DITTO 183 
 
 XI. The central transept of ditto - 195 
 
 NIL The nave, Galilee, western transept, porch and 
 
 chantry chapels of ditto - 201 
 
iv CONTENTS. 
 
 CHAPTER PAGE 
 
 XIII. The planning of the parish churches - 212 
 
 XIV. The supports; the basilican column; the Roman- 
 
 esque pier - - - 230 
 
 XV. The supports; the gothic pier - 244 
 
 XVI. Trabeated and arcuated construction ; defini- 
 tions of arches- - - 257 
 
 XVII. The compound arch; its construction and orna- 
 mentation - 272 
 
 XVIII. Vaulting; the dome; the semidome ; the barrel 
 
 VAULT ; THE HALF BARREL ; THE STONE CEILING ; 
 
 THE SPAN ROOF OF STONE - 28 1 
 
 XIX. Ditto. The groined vault - 289 
 
 XX. Ditto. The construction of ribbed vaults 296 
 
 XXI. Ditto. High vaults with diagonal ribs - 309 
 
 XXII. Ditto. Gothic vaulting; ridge-ribs; tiercerons ; 
 
 LIERNES ; FAN VAULTING - 323 
 
 XXIII. The abutment of the vaulted church ; the 
 
 BUTTRESS, THE PINNACLE 3SO 
 
 XXIV. The transmission of thrusts by the flying 
 
 BUTTRESS ; THE OPPOSITION OF THRUSTS 368 
 
 XXV. The drainage of the roofs. Corbel-tables, 
 
 PARAPETS, BATTLEMENTS, GARGOYLES - - 384 
 
 XXVI. The protection of the walls. Ground-courses, 
 
 dripstones, hood-molds, strings 402 
 
 XXVII. Cubical, scalloped, interlacing, and figure 
 
 capitals - - - - 409 
 
 xxviii. corinthianesque capitals; gothic foliated capitals 420 
 
 XXIX. Molded capitals - 439 
 
 XXX. The base - 447 
 
 XXXI. Romanesque windows, lancets, and plate tracery 456 
 
 XXXII. Geometrical window-tracery - 47 2 
 
 XXXIII. Curvilinear window-tracerv ■ 479 
 
CONTENTS. xv 
 
 CHAPTER PAGE 
 
 XXXIV. Rectilinear window-tracery - 491 
 
 XXXV. Window construction - 505 
 
 XXXVI. The triforium - 519 
 
 XXXVII. The clerestory - - - 543 
 
 XXXVIII. Ol'EN TIMBER ROOFS 550 
 
 XXXIX. Doorways and porches - 573 
 
 XL. Towers - 586 
 
 XLI. Spires - 611 
 
 XLII. Alphabetical list of dated buildings 638 
 
 XLI 1 1. Comparative sheets of moldings of the archi- 
 tectural MEMBERS OF ROMANESQUE AND GOTHIC 
 CHURCHES 658 
 
 INDEX TO ILLUSTRATIONS 709 
 
 INDEX OF PLACES - 739 
 
 INDEX OF SUBJECT-MATTER AND GLOSSARY - 771 
 
INTRODUCTION. 
 
 Of all the artistic achievements of the English race two make unchallenged 
 claim to pre-eminence : our imaginative literature and our mediaeval archi- 
 tecture. Of the former nothing need here be said ; its triumphs are still 
 being won, its end is not yet. With the latter it is not so. Painting, music, 
 novels, play-acting, count their votaries by thousands. The new symphony 
 by Pole or Russian or Bohemian obtains respectful audience and admiration ; 
 columns of appreciation are daily discharged on every second-rate painting or 
 third-rate play. Not so with architecture. There never was a time of such 
 blackness of indifference as to the master-art of architecture. It was not always 
 so. In the old England there was little literature, little painting, little play- 
 acting ; but there was the most beautiful architecture. Everybody loved it, or 
 they would not have paid for it. In the fifteenth century every village mason 
 could build a church, and the village carpenter could crown it with a hammer- 
 beam roof. In Elizabeth's spacious days. Lord Bacon, Lord Burghley, the 
 Secretary of State, the Ambassador to France, were students of architecture ; 
 largely competent to criticise and control the planning and design of hall and 
 mansion. In the Augustan age of English literature and English architecture, 
 no cultured man but had visited and studied the palaces of Palladio and Michael 
 Angelo, and was competent to discuss the proportions of the orders. A 
 knowledge of architecture was a necessary equipment of the gentleman. Lord 
 Burlington was proud to father designs, the paternity of which belonged to 
 others. Those were glorious days for architects, before the English aristocracy 
 had concentrated its intellectual force on the destruction of the pheasant and 
 the fox. 
 
 Nowadays architecture is outside the precincts of culture. Educated 
 people know little and care less about architecture. Classic and Renaissance, 
 Romanesque and Gothic, are naught to them ; their ignorance is naked and 
 unashamed. In this general neglect mediaeval architecture beyond all is immersed. 
 For a brief period indeed interest in this supreme artistic achievement of our race- 
 was revived by Britton, Pugin, Petit, and Willis, greatest of all. That interest 
 was not to endure. 
 
 Nowadays the students of our national architecture are few. It is surprising 
 that there are any in the face of the discouragements which their study meets. 
 At the older universities tens of thousands of pounds are expended every year 
 
 b 
 
xviii INTRODUCTION. 
 
 to encourage the study of classical literature, mathematics, history, or science ; 
 not a penny on architecture. Neither at Oxford nor at Cambridge is there a 
 single professorship, lectureship, scholarship, or fellowship in English mediaeval 
 architecture. France and Germany have several able periodicals devoted ex- 
 clusively to the subject of mediaeval architecture ; we have not one. Government 
 subventions support a great museum of mediaeval art in the Trocadero at Paris ; 
 we have at South Kensington a few casts, and those chiefly of foreign Renais- 
 sance work, mixed up with pitchers and jugs and fiddles and furniture. At the 
 annual exhibition of the Royal Academy one small room is deemed enough for 
 the drawings of the architects. Year by year we have exhibitions of the 
 potsherds of Rome and Greece and Egypt ; not of our own mediaeval art. 
 Immense sums are spent in excavating civilisations in far-away countries with 
 which we have little concern ; our own Byland, Rievaulx, Glastonbury remain 
 lost beneath the soil. 
 
 For this apathy and neglect there must be a reason ; probably there is more 
 than one. In the first place architecture, if it is to be studied to the best advan- 
 tage, must be studied, like botany and geology, in situ. But such study is open 
 to few. Hexham and Dore, Norwich and St David's are far sundered. Yet 
 these and countless others must be visited in any thoroughgoing survey of 
 English mediaeval architecture. Next to actual inspection of the buildings, the 
 best thing is to study them in illustrations. Hitherto, however, it has not been 
 possible, except to the few, to study them even in this form. There are indeed 
 comparatively few mediaeval buildings of the first rank which have not been 
 illustrated in measured drawings. But what private person could afford to 
 become the possessor of the tomes, many of them rare, costly, and bulky, in 
 which they are to be found : Bowman and Crowther's Churches of the 
 Middle Ages, Brandon's Analysis and Open Timber Roofs, Britton's Architectural 
 Antiquities and English Cathedrals, Caveler's Specimens, Colling's Details, Gothic 
 Ornaments, and English Mediaeval Foliage, Hadfield's Essex, Johnson and 
 Kersey's Netie Valley Churches, Pugin's Examples and Specimens, Sharpe's 
 Architectural Parallels, Professor Willis' invaluable papers, scattered about in 
 the Transactions of various provincial societies, the Architectural Association 
 Sketch Book (32 vols.), the Spring Gardens Sketch Book (8 vols.), the fohn d 
 Gaunt Sketch Book (3 vols.), the Abbey Square Sketch Book (3 vols), Neale's St 
 Albans, Hodges' Hexham, Reeve's Fountains, and a host of other monographs. 
 These, in default of personal visits to each church, are the sources to which 
 the architectural student must resort. Such a collection, however, is entirely out 
 of reach except to residents in London. This difficulty of access to adequate 
 illustrations may well explain, to some extent at any rate, the unpopularity of 
 the study of mediaeval architecture. It has been unpopular because the apparatus 
 for its proper study has not been available. To the writer, therefore, the first 
 thing to be done, to advance the study of mediaeval architecture, seemed to be 
 to provide copious illustrations. Fortunately two circumstances combine to 
 make this possible, even in the compass of a single volume to do this in a fairly 
 adequate manner. One is that the copyrights of many large and costly works 
 have run out, and it has become possible to reproduce from them illustra- 
 tions long out of print. The second is the facility of illustration given by 
 
INTRODUCTION. xix 
 
 modern photographic processes. It has been the writer's pleasant task to visit 
 nearly every important church in England, camera in hand, and he has had 
 abundant aid from his brother photographers. But for photography an illus- 
 trated volume so copious in examples would have been out of the question. 
 With its aid, it has been possible to include 20 whole-page collotypes, 785 repro- 
 ductions of photographs, sketches, and measured drawings, in addition to 469 
 further illustrations which are arranged on 12 pages of plans, 2 pages of sections, 
 8 pages of diagrams, and 28 pages of moldings. A great book is a great 
 evil ; but not, it is to be hoped the reader will think, a great picture-book. 
 
 A yet graver reason it may be for the failure of mediaeval architecture 
 to arrest and retain the attention of the modern student is the frag- 
 mentary and disconnected presentation of the subject which has been usual. 
 Open any of the text-books from Rickman downwards and try to obtain a 
 consecutive and complete treatment of any one of the chief features of the 
 medi;eval church — its plan, its vault, the abutments of the vault, the drainage <>f 
 the roofs, the fenestration — what do we find? Perhaps we would like to know 
 about the principles of construction of the vault. On this we get a few isolated 
 scraps of information under " Norman," followed by details about doorways 
 and buttresses and windows and capitals and things in general. The few- 
 scraps of information about Norman vaults are lost in this olla podrida. 
 When we have forgotten all about them, we get perchance some information 
 about "Early English" vaulting. This in turn is overlaid by layer upon 
 layer of other miscellaneous matter. And so on to the end. No subject can 
 be understood nor can any subject interest, when treated in such desultory 
 fashion. There seemed to the writer, therefore, to be room for a connected 
 analysis of mediaeval architecture. In this, first of all, should come the subject 
 of planning — a subject of primary importance, which however has usually 
 been omitted altogether. Secondly should come the important matter of the 
 vault and its supports. Of great importance also is the question of abutment ; 
 it is one thing to put up a vault, it is another to induce it to stay up. This 
 includes the whole machinery of buttresses, pinnacles, and flying buttresses. 
 Then there is the drainage question. How is the rain to be kept from damaging 
 roof and wall? This includes the corbel-table and dripping eaves, and the later 
 contrivances of gutter, gargoyle, parapet, and battlement ; also the protection of 
 wall, window, and doorway by basement course, string, dripstone, and hood-mold. 
 Then there is the whole question of lighting, and the development of window- 
 tracery as controlled by the exigencies of stained glass ; and many other subjects, 
 each needing separate treatment, such as the capital and the base, the triforium 
 and the clerestory, the doorway and the porch, the roof, the tower and the 
 spire. On every one of these a separate treatise seems to be demanded ; not 
 necessarily lengthy, but consecutive in treatment, and as far as space allows, 
 complete. It is precisely to such a collection of short treatises on mediaeval 
 planning and building construction that Tart II., the bulk of the work, is devoted. 
 i'See Table of Contents, xiii, xiv, xv.j 
 
 The fragmentary treatment of mediaeval architecture which has prevailed 
 so long is probably due mainly to the influence of Rickman's work. Just as 
 Linnaeus taught the botanical student to arrange his plants in orders, genera, 
 
xx INTRODUCTION. 
 
 and species, so Rickman taught his followers to classify their churches in archi- 
 tectural periods. Linnaeus' methods long prevailed ; and while they prevailed, 
 botany was a dull science. Later on, botanists arose who taught how plants 
 grew, and botany at once became a fascinating study. Architecture had not the 
 good fortune of botany ; it has remained a classificatory science. No wonder, 
 then, that it has been found void of life and interest. 
 
 Nor is that the only objection to a mere classificatory treatment. It is bad 
 enough that it devitalises the subject of interest ; it is worse still if the classifi- 
 cation is itself unsound. And that is so. We have been told for nearly a century 
 that there are four periods of English mediaeval architecture : Norman, Early 
 English, Decorated, and Perpendicular. But there is no such thing ; the famous 
 four periods are mere figments of the imagination. Take a subject of primary 
 and fundamental importance : that of the planning of the greater churches ; 
 there are not four, but only two periods of planning ; of which the first, the 
 period of the three parallel eastern apses and of the periapsidal plan, ends with 
 the twelfth century, while all the later plans were in use by that time. Or take 
 vaulting as the criterion. Then the periods become five : that of the groined 
 vault, the ribbed vault whether quadripartite or sexpartite, the vault with 
 tiercerons and ridge ribs, the lierne vault, and the fan vault ; the periods are 
 not four but five, and do not coincide with the traditional Norman, Early 
 English, Decorated, and Perpendicular. If the very important matter of 
 abutment be taken as a criterion, we are equally in difficulty. All the main 
 methods of abutment had come into use by A.D. 1200; in the Early English, 
 Decorated, and Perpendicular periods no important novelty as to methods of 
 abutment is introduced. Only to one, and that quite a subordinate member of 
 the building, does the antiquated terminology fairly apply, viz., to the fenestra- 
 tion ; and even here it is badly chosen and inaccurate, and was very properly 
 revised and corrected by Mr Edmund Sharpe.* 
 
 The whole classification, moreover, is mischievous as well as baseless. The 
 novice is led to believe that architecture stopped at the end of each of the four 
 periods, turned over a new leaf, and began again de novo. Nay further, that 
 there is in each of the four periods some inward and spiritual significance, which, 
 could it be discerned, would give us the keynote or character of the whole archi- 
 tecture of the time. But it is just as easy to argue about the deep moral 
 and spiritual significance of the two planning or abutment periods as about 
 that of the traditional four; and just as futile. The greatest objection of all, 
 however, to this cutting up of architectural history into periods is that it obscures 
 the essential unity of the development of the building art. Professor Freeman 
 ever protested against the demarcation of ancient and modern History. Equally 
 important is it to emphasise the unity of architectural art, and to protest against 
 its being cut up into arbitrary sections. Architecture is one, not many. Every 
 so-called style was a transition from that which preceded it, and a transition 
 to that which was its successor. " From Roman to Renaissance the history of 
 architecture is an uninterrupted series of transitions ; it is quite time that we 
 studied the art of the Middle Ages in the fashion in which we study the 
 
 * For Rickman's Norman, Early English, Decorated, and Perpendicular, Mr Sharpe 
 substituted Norman, Transitional, Lancet, Geometrical, Curvilinear, Rectilinear. 
 
INTRODUCTION. xxi 
 
 development of a living being, which from infancy passes to age by a series of 
 insensible transformations, without its being possible from one day to another to 
 say where infancy or youth ceases or where age begins."* 
 
 In the present volume, therefore, the traditional classification into periods 
 has been abandoned,^ except that in Chapters II., IV., V., and VI., the charac- 
 teristics of the so-called Norman, Early English, Decorated, and Perpendicular 
 periods are enumerated. 
 
 It follows from what has been said above that it is here attempted to 
 introduce into the subject of English mediaeval architecture that evolutionary 
 method of treatment which has been so fertile of results in every branch of 
 knowledge to which it has been applied. The book is an attempt not to classify, 
 but to work out processes of development. Evolution, whether in architec- 
 ture or in anything else, was not a flux of blind and unmotived change. For 
 every change there was a reason. What that reason was it may perhaps now in 
 many cases be impossible to discover. We cannot look through the eyes of the 
 old builders. We may think we see what they were about ; but we merely think, 
 we do not know ; we are in the region of conjecture, and conjecture is hazardous. 
 But are we therefore to discard conjecture? It is not discarded from modern 
 science. The naturalist does not know that the colours of insect or of bird are 
 due to protective or sexual reasons ; this is but a hypothesis, i.e. a conjecture 
 of his. So too in architecture hypothesis is not to be discarded, provided that 
 it explains the phenomena, and that the cause it assigns is a vera causa and is 
 adequate to produce the effect. The writer, therefore, has not shrunk from the 
 suggestion of causative relations. Nothing is more interesting than the search 
 for the hidden cause ; nor should the investigator be deterred even if at times 
 his discovery prove but a mare's nest. 
 
 From the adoption of an evolutionary method of treatment yet one more 
 consequence flows. It is that the evolution should be traced back, not half-way, 
 but if possible to the fountainhead ; in other words the question of origins should 
 be dealt with. English mediaeval architecture has been presented too often as 
 a sort of architectural Melchizedek, or as if it sprang forth full-grown like some 
 Pallas from the teeming head of Zeus, in the last half of the eleventh century, 
 in Caen or Canterbury. But the Norman offshoot of the great Romanesque 
 stock had its roots in a distant past. Its history goes back to the earliest days 
 of church building in newly Christianised Rome, to the first years of the fourth 
 century. That history indeed, from the fourth to the eleventh century, is dark 
 and dubious. But that the Romanesque and Gothic minsters are the offspring 
 of the early Christian basilicas there can be no doubt, however difficult it may 
 be at present to establish each step of the pedigree. Throughout the book, 
 therefore, reference has been made, where reasonable evidence exists, to the 
 origin and history of mediaeval architecture not only in our own country but 
 throughout Gaul, German)-, and Italy in the Dark Ages. The statements made 
 are in many cases far from pretending to certainty ; but by the references 
 which have been given to authorities the reader is put in a position to test for 
 
 * Viollet-le-Duc. 
 
 t The French archaeologists have long discarded the arbitrary divisions of Ue Caumont 
 and others. 
 
xxii INTRODUCTION. 
 
 himself the validity of the conclusions presented. English architectural history 
 will lose nothing if it ceases to be so insular. To the Romanesque architecture 
 of Normandy in particular much attention has been given ; in the great abbeys 
 of that country we have the incunabula of the English abbey church and 
 cathedral. 
 
 Nor has the writer hesitated to describe developments which are to be found 
 in the Gothic of France, but which were not reached here. French writers do 
 not fail to include in their architectural treatises an account of those features, 
 such as the open timber roof, the lierne and fan vault, which were developed 
 here only, or reached here the highest stage of development. Similarly it 
 seemed desirable not to conclude the discussion, for example, of the treatment 
 of the triforium without some account of the "transparent" triforium of the He 
 de France. Wherever possible, the comparative method of investigation has 
 been adopted, at any rate as regards the most important of the schools of 
 mediaeval architecture ; those of the He de France and England. 
 
 Many shortcomings there are, and must necessarily be, in this or in any 
 attempt to deal with the vast subject of English mediaeval architecture. It is 
 true that measured drawings of most of the greater churches are to be found 
 scattered here and there in the various Sketch-Books ; in the Builder, B?iilding 
 News, Architect, British Architect, Builders Journal ; and in such collections as 
 those of Bowman and Crowther, Brandon, Colling, the Churches of the Nene 
 Valley, and various monographs. But very few scientific descriptions of 
 churches, with complete apparatus of measurements, plans, sections, elevations, 
 details, moldings, and critical text have hitherto been published. 
 
 Again, a writer on the mediaeval architecture of France or Germany has 
 a vast corpus of facts ready to his hand in the archaeological literature of that 
 country ; in England the Transactions of the provincial societies, though they 
 were founded mainly for the study of mediaeval architecture, are largely devoted 
 to non-architectural subjects.* Their proper task — that of analysing, describ- 
 ing, and classifying the churches of each district — has with a few noteworthy 
 exceptions, made exceedingly little progress. The want of accurate classified 
 information and the lack of an index to measured drawings have made and 
 must make the preparation of any work on English architecture difficult 
 and incomplete ; errors must needs occur in battalions. The author will be 
 grateful for any corrections, suggestions, or criticisms addressed to him through 
 the publisher. 
 
 * Among recent papers may be mentioned one, " On the Ceremonial of the Toda Dairy ;" 
 an interesting" topic, but qu'allait-il ' faire dans cette galire? 
 
GOTHIC ARCHITECTURE 
 
 PART I. 
 
 THE ORIGIN AND DEVELOPMENT OF THE MEDIAEVAL 
 CHURCH ARCHITECTURE OF ENGLAND. 
 
 Chapter I. 
 
 Architecture Defined — Basilican and Byzantine Architecture — Romanesque Architecture 
 — Schools of Romanesque — Gothic Architecture Defined — Relation of Gothic to 
 Romanesque. 
 
 Definition OF Architecture. — The art of Architecture has been defined 
 very variously. It was defined by Mr Garbett * as " the art of well building ; in 
 other words, of giving to a building all the perfection of which it is capable." 
 Mr Ruskin f defined it as "the art which so disposes and adorns the edifices 
 raised by man, for whatever uses, that the sight of them may contribute to his 
 mental health, power, and pleasure." In the American Dictionary of Architec- 
 ture and Building (1901) it is defined as "the art of building with some 
 elaboration and skilled labour " ; and, in a more limited sense, as " the modifica- 
 tion of the structure, form, and colour of houses, churches, and civic buildings, by 
 means of which they become interesting as works of fine art." But it can hardly 
 be held that there is one art of making things well, and another of making them 
 badly. There is not one art of making clothes that fit and another art of making 
 misfits. One and the same art makes flower-pots for the gardener and Worcester 
 ware for the connoisseur. So it is with .Architecture. It is simply "the art of 
 building." ^ Good architecture is indeed the art of building beautifully and 
 expressively ; and bad architecture is the reverse. But architecture is the art of 
 building in general. 
 
 This seems clear enough. But as a matter of fact the definition contains 
 an ambiguity in the use of the term " building." In the erection of every edifice 
 the work necessarily falls into two parts. There is the actual putting together 
 of the materials by manual labour and machinery so as to form roofs, supports, 
 
 * Principles of Design, 1. t Seven Lamps, 13. 
 
 \ So Viollet-le-Duc {Architecture, i. 116), who defines architecture as 'Tart <te bdlir." So 
 also Mr Barr Ferree, for whose discussion of the current definitions see the Architectural 
 Record, i. iqq. 
 
2 DEFINITION OF ARCHITECTURE. 
 
 and abutments. There is also the preliminary process of planning and designing 
 the buildings, and, it may be, of making drawings, whether rough sketches, or 
 drawings to scale or full size, as well as that of superintendence. Now these two 
 operations, the preliminary and the subsequent one, may be carried on by the 
 same individual, or they may not. If a modern builder is erecting a terrace of 
 small tenements, he may conduct both operations himself; he may plan and 
 design the terrace, superintend the actual building of it, and take part in the 
 work with his own hands. In such a case he is both architect and builder. And 
 what is sometimes done nowadays by a modern builder was no doubt at all 
 periods occasionally done by builders. We ma}' be sure that in the fifteenth 
 century many a village builder was capable of planning and designing a new- 
 aisle or chancel as well as of putting it up. In such a case, as in that of the 
 modern builder quoted above, he was, in the modern and restricted sense of the 
 terms, both architect and builder. But when a large and important building is 
 erected nowadays, one and the same man does not undertake both divisions of 
 the work ; one part of the work is handed over to one man, the other part 
 to another ; in modern parlance the first is the architect, the second the builder. 
 And we may be sure that at all periods when any great building was erected, 
 there was a similar division of functions. When the Parthenon was built, or 
 Santa Sophia, or Amiens Cathedral or Salisbury, even if the architect had gone 
 through the "shops," as the British engineer still does, he would have too much 
 to do with planning, design, drawings, and superintendence, to work at the 
 buildings to any considerable extent with his own hands. The more he used his 
 hands, the less time he would have to use his brains. To be accurate, therefore, 
 we must not, except in comparatively small and unimportant work, define 
 "architecture" as "the art of building," but as "the art of planning, designing, 
 and drawing buildings, and of directing the execution thereof." * 
 
 Another difficulty has been raised as to whether Architecture should be 
 classed with the Fine or the Industrial Arts; i.e. whether it belongs to the 
 category in which are found Painting, Sculpture, Music and Imaginative, 
 Literature; or whether it ranks with the Industrial Arts. The difficulty arises 
 from the fact that there is really a third category intermediate between the 
 Fine and the Industrial Arts. No one would contend, except by way of 
 paradox, that farming and cookery are anything but Industrial Arts. But 
 
 * How far the mediaeval magister operis was builder as well as architect has long been a 
 quastio vexata. See Notes on the Superintendents of English Buildings in the Middle Ages, 
 by Wyatt Papworth, Journal of R.I.B.A., xxxviii. ; On the Hope of English Architecture, by 
 W. H. White, Journal of R.I. B. A., December 1874; Architects and Master Workmen, by 
 J. J. Stevenson, Journal of R.I.B.A., January 1S75 \ Die Romanische und Golische Baukunst, 
 by Max Hasak, Stuttgart ; Mr T. G. Jackson, in Builder, 10th April 1897 ; Enlarts Manuel, 
 1. 62; Choisy's History, ii. 518 and 256; Anthyme St Paul's Hist. Monumenlale, 293; 
 Viollet-le-Duc, Dictionnaire, iv. 19S ; The Basis of Gothic Architecture, bv Mr E. S. Prior, in 
 Builder, 23rd February 1901 ; Education in Building, bv Professor Lethaby, in Journal of 
 R.I.B.A., 17th June 1901 ; and his Medieval Art, 255. On Mediaeval Working Drawings, 
 see article by Mr Burges in Journal of R.I.B.A., xxxviii.; article on "Drawing" in The 
 Dictionary of Architecture, issued by the Architectural Publication Society ; list of mediaeval 
 drawings m Journal of R.I.B.A., 25th November 185S ; and in Lethab/s Medieval Art, 
 260; West, in Journal of R.I.B.A., 1S74, 3§ ; Viollet-le-Duc, Dictionnaire, ix. 197 : Enlart's 
 Manuel, 1. 65. 
 
i :■",'■ 
 
 i .:■ 
 
 — 
 < 
 
 IP!':':- 
 
EARLY CHRISTIAN ARCHITECTURE. 5 
 
 it is different when we turn to what are called the Applied Arts or the 
 Decorative Arts. These arts are, in the main, utilitarian ; nevertheless their 
 products may to some extent be beautiful and expressive; in the case of the 
 jeweller, sometimes to a very considerable extent. It is true that if the jeweller 
 make a clock or watch, his main object is utilitarian ; but if he make some purely 
 useless article, such as a ring or necklace, his occupation becomes a Fine Art. 
 So it is with the architect or builder. When he is providing shelter, which is 
 a utilitarian occupation, and the primary function of Architecture, his art to that 
 extent is an Industrial Art. But if he provide shelter in the fashion in which it 
 is provided in Westminster Hall or Westminster Abbey Church, his work ranks 
 among the Fine Arts; without ceasing, however, to be an Industrial Art. 
 When, however, he is building a Triumphal Arch, a Nelson's Column, an 
 Eleanor Cross, his Architecture becomes a Fine Art, pure and simple. For a 
 Fine Art, pure and simple, is one which has no connection with material utility 
 and use. 
 
 Early Christian Architecture. — What we are concerned with here is 
 the Church Architecture which was done in England between the Norman Con- 
 quest and the Dissolution of the Monasteries in the sixteenth century ; the earlier 
 part of which goes by the name of Romanesque or Anglo-Norman or Norman, 
 while the latter part is called Gothic. Church Architecture had a very long 
 history before the Norman style reached these islands in the eleventh century. 
 What Mr Fugin designated Christian Architecture began on a large scale at the 
 commencement of the fourth century of our era, in the reign of Constantine. 
 Almost at once it diverged in two opposite directions. One half of Christendom 
 used the Greek, the other the Latin liturgy. The Greek Christians developed 
 their churches on the lines of such buildings in Rome as the Pantheon, S. 
 Stefano Rotondo, and the like ; producing that great style of domed 
 churches, which, because centred at Constantinople (whose Greek name was 
 Byzantium), goes by the name of Byzantine Architecture. Its origin and 
 history would be far clearer if it were called East Roman. It is an architec- 
 ture in which Roman methods of construction were worked out by Hellenistic 
 craftsmen. 
 
 But those who used the Latin liturgy, i.e. Western Christendom, erected 
 churches of vast dimensions — indeed OLD ST PETER'S (147 J and St Paul's extra 
 muros, Rome, had areas of about 100,000 feet — but cjuite simple in structure.* 
 These Early Christian churches are called Basilicas, and the style is the Basilican; 
 what the French archaeologists call the Latin style.f It is neither Byzantine nor 
 Romanesque, but a style with quite distinct characteristics of its own. Up to 
 the ninth century it may be said to have had the field entirely to itself in the 
 greater churches of Western Christendom. It persisted, in its own country at 
 any rate, here and there throughout the whole Romanesque* period and well into 
 the Gothic days. 
 
 * In section they are practically the same as Ely nave (34.1 ). 
 
 t See article on Latin style by W. P. P. Longfellow in the American Dictionary of 
 Arcliitcctit}-c and Building. 
 
 \ Pisa Cathedral, S. Ambrogio, Milan, and St Mark's, Venice, may well have been building 
 together ; the first is Basilican, the second Romanesque, the third llyzantine. 
 
6 ROMANESQUE ARCHITECTURE. 
 
 Romanesque Architecture* — The last part of the sixth, the seventh, 
 
 and the eighth centuries were the worst times probably ever known in Western 
 Europe ; they were emphatically the Dark Ages. The Roman Empire of the 
 West had sunk beneath the barbarian hordes ; it was not till the redistribution 
 of Europe into nationalities, till Charlemagne arose, late in the ninth century, 
 that civilisation lifted its head again, and a new architecture became possible. 
 To the ninth century may be attributed the elaborate planning of the monastery 
 of ST GALL (194); the eastern parts of S. Ambrogio, Milan, and the 
 ambulatory of ST MARTIN, TOURS (192.3). This new style is called 
 Romanesque.^ It is a term by no means easy to define. Ouicherat's well- 
 known definition is that Romanesque architecture is that which has ceased to 
 be Roman, though it still retains much that is Roman ; and which is not yet 
 Gothic, though it has already something Gothic about it. The definition is an 
 attractive one, but is really but little helpful; it is to explain " obscumm per 
 obscurius" To understand it we must first know what Roman and what Gothic 
 architecture is. The same objection applies to M. Anthyme St Paul's de- 
 finition ; that it is " Roman architecture purified and developed to suit the needs 
 of the Catholic liturgy and the genius of each of the peoples who employed it 
 from the ninth to the thirteenth century"; we want to know what he means 
 by Roman architecture. Turn to Viollet-le-Duc [Dictionnaire, iv. 60) and all 
 becomes clear. The Roman work of which Romanesque is the offspring is just 
 one particular sort of Roman work ; that of the Basilica. " Le probleme que 
 les architectes de l'epoque romane ( = Romanesque) s'etaient donne a. resoudre 
 dtait celui-ci : elever des voutes sur la basilique antique ; " i.e. put shortly, the 
 Romanesque problem was that of vaulting a basilica. Romanesque architecture is 
 the art of building vaulted basilicas. And by a basilica we mean what is basilican 
 both in plan and elevation ; in plan, as having nave and aisles ; in elevation, as 
 having aisle wall, lean-to roof, and clerestory wall containing windows. This 
 then was the problem of problems of the Western builders from the ninth century 
 onward ; to vault an aisled church without destroying its clerestory lighting. 
 
 ROMANESQUE SCHOOLS.— This problem could be solved, and was solved 
 perfectly in more than one way ; nowhere probably till late in the eleventh 
 century. One solution was to ceil the nave with barrel vaults resting on 
 clerestory walls pierced with windows. This was the solution of the two 
 schools of Burgundy and Provence. A second was to ceil the nave with a row 
 of domes ; a method peculiar to Le Puy Cathedral and St Hilaire, Poitiers. A 
 third was to ceil the nave with intersecting vaults ; groined vaults, as at Speyer ; 
 ribbed vaults, as at Durham. + It was the last solution which resulted in Gothic 
 
 * The term "Romanesque" was first proposed by Mr Gurm ; and was adopted by Dr 
 Whewell in his Notes on German Churches, 1835. 
 
 t The application of the term varies greatly. Many apply it, with Mr Fergusson (i. 411), 
 to all Christian Architecture, except Byzantine, done in Western Europe, before Gothic ; i.e. all 
 the work between the fourth and the latter part of the twelfth century. Such an application, 
 which would designate the fourth century basilicas of Rome and the sixth century ones of 
 Ravenna as Romanesque, is confusing in the extreme, and should be discarded. 
 
 % In the above no account has been taken of the schools which evaded one of the conditions 
 of the problem, by omitting either the aisles or the clerestory lighting. On the whole subject 
 see Classification of Romanesque, and table on page 13. 
 
GOTHIC ARCHITECTURE. 7 
 
 architecture. The Burgundian and Provencal solutions with barrel vaults, and 
 that of Le Puy with domes, proved unfruitful ; nothing came of them. The third 
 solution, however, is of the utmost importance. This solution was worked out 
 in several countries, more or less independently ; in particular, in Lombardy, 
 Germany, and Normandy. From Normandy it was borrowed by England. In 
 Germany, at Speyer, the high vault was groined. In Lombard}-, e.g. at S. 
 Ambrogio, Milan ; in Germany, e.g. at Worms ; in Normandy, e.g. at LESSAY 
 (319) ; in England, e.g. at DURHAM (8) ; the high vault had diagonal ribs. 
 
 But at this point a difficulty arises. Such churches as EXETER (9) and 
 Amiens Cathedrals also come within the definition. Both are aisled churches 
 with clerestory lighting ; both are vaulted with diagonal ribs. Yet Exeter and 
 Amiens are as undoubtedly Gothic as S. Ambrogio, Milan ; Worms ; Speyer ; 
 Lessay and Durham are undoubtedly Romanesque. What then is it which 
 constitutes the one set of churches Romanesque, but the other set Gothic ? 
 
 GOTHIC ARCHITECTURE. — The answers given to this question are extra- 
 ordinarily diverse. The term "Gothic" occurs much before the seventeenth 
 century.* Those who invented it were quite clear as to what they meant. 
 They meant that it was something barbarous, because non-classical. Some 
 believed that it was actually invented by the Goths and Vandals who overthrew 
 the Roman Empire. " Then," says Vasari, " arose new architects who after the 
 manner of their barbarous nations erected buildings in that style which we call 
 Gothic." So also Evelyn says that " the ancient Greek and Roman architecture 
 answered all the perfections required in a faultless and accomplished building" ; 
 and that the Goths and Vandals demolished these, and " introduced in their 
 stead a certain fantastical and licentious manner of building ; congestions of 
 heavy, dark, melancholy, monkish piles, without an}- just proportion, use, or 
 beauty"; utterly devoid of all "true and just symmetry, regular proportion, 
 union, and disposition." f 
 
 We may now turn to definitions expressed, as they ought to be, in con- 
 structional terms. The first is extracted from various statements, not very 
 definite, of Viollet-le-Duc ; the substance of which is, that Gothic architecture is 
 the art of erecting buildings in which the outward thrusts of the vaults are 
 neutralised by the inward thrusts of the flying buttresses. The objection to it 
 is that the thrusts of the vaults are far more powerful than the thrusts of any 
 of the flying buttresses, and cannot be neutralised by them ( see 378). 
 
 The next is an attractive one ; it is that of M. Anthyme St Paul ; viz. that 
 Gothic construction is the result of the fusion, in one and the same vault, of 
 diagonal ribs and pointed arches, abutted, when necessary, by flying buttresses. 
 To this it may be objected, first that no mention is made of buttresses, which 
 are of primary, whereas flying buttresses are only of secondary importance ; 
 also that the pointed arch is not of primary importance. Exeter Cathedral 
 
 * See references in Paley's Gothic Architecture, 16, and Lethaby's Med. Art, 135. 
 
 t Mr Ruskin's definition may be inserted as a curiosity. " Our final definition of Gothic," 
 he says, " is Foliated architecture, which uses the pointed arch for the roof proper, and the 
 gable for the roof-mask" {Stones of Venice, ii. zzz). By "foliated architecture" he explains 
 that he means that in which the arches (other than bearing arches and pointed arcading) are 
 cuspcd ; and the apertures foliated. 
 
GOTHIC ARCHITECTURE. 
 
 might be rebuilt with every arch semicircular, yet its construction might remain 
 Gothic. 
 
 To M. Enlart * also the diagonal ribs and flying buttresses appear to be 
 
 Durham Nave from S.E. 
 
 essential elements, though the latter are but organs of transmission ; the real 
 work of stopping the thrusts of the vault being passed on to the buttresses ; 
 
 * Manuel, 435-442. 
 
GOTHIC ARCHITECTURE. 9 
 
 moreover many buildings, undoubtedly Gothic, have no flying buttresses at all ; 
 e.g. Poitiers and BRISTOL (35.4) Cathedrals. He recognises, however, that 
 the pointed arch is non-essential. On the ground that the pointed arch is freely 
 employed in Romanesque construction half a century before the Gothic period, 
 he says, "Cet element doit itre elimine de la definition dn style gotliique." To the 
 vault with diagonal ribs and the flying buttress he adds, " line ornamentation tonte 
 nouvelle, pnisee, /ion plus dans les traditions, mats dans I 'etude directe de la nature ; " 
 a statement hardly true of the foliated capitals and scrolls of the early Gothic 
 
 Exeter Nave from West. 
 
 of England, if they be derived, as suggested below (420), mainly from classical 
 sources. 
 
 On one point at any rate we may agree ; viz. that the one thing of primary 
 importance is the vault. Flying buttresses, buttresses, pinnacles, pointed 
 arches, would none of them be there, were it not for the vault. The pointing of 
 the arches facilitates its construction (322) ; the buttresses, flying buttresses, 
 and pinnacles are the machinery by which its thrusts are neutralised. Now in 
 vaults with diagonal ribs one peculiarity is universally found. They are not 
 constructed like barrel vaults ; they have not a continuous thrust along the 
 whole length of the walls on which they rest. This very fact brings an immensely 
 
IO 
 
 GOTHIC ARCHITECTURE. 
 
 powerful, because concentrated, thrust against certain points of the wall. At 
 these points it is necessary to strengthen the wall. 
 
 It is just at this point that 
 we part company with such 
 Romanesque as that of Durham 
 and Lessay. Compare them 
 with Exeter or Amiens. The 
 aisles of all four have vaults 
 with diagonal ribs, giving inter- 
 mittent and concentrated out- 
 ward pressures. How are these 
 pressures met ? In the first two, 
 by thickness of wall ; in the 
 second two, by buttresses. In 
 Durham and Lessay intermittent 
 pressures are stopped, unscien- 
 tifically, to the great waste of 
 material, by continuous abut- 
 ment ; at Exeter and Amiens, 
 scientifically, to the great 
 economy of material, by inter- 
 mittent abutment. This then 
 is the difference between our 
 Romanesque and our Gothic, 
 between Durham and Exeter ; 
 a difference of abutment. So 
 we may frame a final definition 
 of our Romanesque and Gothic architecture. Anglo-Norman Romanesque is 
 the art of erecting aisled and clerestoried buildings whose vaults have groins or 
 intersecting ribs and the thrusts of whose vaults are stopped by walls. English 
 Gothic architecture is the art of erecting aisled and clerestoried buildings with 
 vaults whose ribs intersect* and whose thrusts are wholly or mainly stopped, 
 directly or indirectly, by buttresses. The second definition excludes Durham 
 nave, though it has a vault with intersecting ribs, flying buttresses, and pointed 
 transverse arches in the vault ; on the ground that the abutment is by thickness 
 of wall, not by projection of buttress. 
 
 One difficulty remains. It is that in many churches which no one would 
 think of calling anything but Gothic, e.g. Salisbury, the thrusts of the vault are 
 not wholly stopped by buttresses, but partly by the wall. If we insist that the 
 buttress shall do all the work, and that the wall shall be reduced to a mere pier, we 
 shall have to exclude nearly all the Gothic work of England — Salisbury Chapter 
 House and GLOUCESTER CHOIR (35.5) would be exceptions— and much of 
 that of the Continent ; and confine Gothic architecture to a few examples in the 
 style of the He de France. A definition so restricted carries with it its own 
 condemnation. 
 
 * It is best not to introduce " diagonal " ribs into the definition ; for some of our later vaults 
 have no diagonals. 
 
 Nave from S.W. 
 
GOTHIC ARCHITECTURE. 
 
 i i 
 
 It remains to see whether the definition given above is of sufficiently general 
 application. First, it applies to such buildings as the TEMPLE CHOIR (35.1), 
 ceiled with vaults which pro- 
 duce opposing thrusts ; the dia- 
 gonal ribs and buttresses are 
 there, though not the flying 
 buttresses. Secondly, there are 
 diagonal ribs and buttresses in 
 buildings without aisles or clere- 
 stories, such as Ely Lady Chapel 
 and the Sainte Chapelle, Paris ; 
 though no flying buttresses. To 
 include this second class we may 
 curtail our definition ; making 
 it read, " Gothic architecture is 
 the art of erecting buildings 
 with vaults whose ribs intersect 
 and whose thrusts are stopped 
 by buttresses." 
 
 But there is a still larger 
 set of buildings to which we 
 cannot deny the term Gothic ; 
 but which have wooden roofs, 
 not vaults ; viz. the vast 
 majority of the parish churches ; 
 and here and there cathedral 
 work; e.g. YORK MINSTER (10) 
 and Carlisle Choir. These have no stone vaults, and therefore no thrusts. 
 The only thing left of our definition is the buttress. They all have buttresses. 
 Revise the definition once more, and we may include Carlisle Choir and the rest. 
 It now reads, " Gothic architecture is the art of erecting buttressed buildings." 
 So it turns out after all that the universal element in Gothic is not the vault 
 with intersecting ribs, but the buttress. 
 
 One case remains to be put. What is to be said of the Eleanor Crosses or 
 of such a monument as the canopied tomb of Archbishop Grey in York transept? 
 The Eleanor Crosses present no difficulties, if we alter the definition to "Gothic 
 architecture is the art of building with the aid of buttresses"; for all the Eleanor 
 Crosses have buttresses. There are no buttresses, however, in Archbishop Grey's 
 Monument.* There are, however, trcfoiled arches richly molded, triangular 
 pediments above them, capitals with stalked conventional foliage, conventional 
 leaf scrolls and crockets, water-holding bases, moldings generally of peculiar 
 design ; all characteristic of Gothic and not of Romanesque architecture ; the 
 monument is unquestionably Gothic. In its widest sense, therefore, Gothic 
 architecture is the art of erecting buildings whose vaults possess intersecting ribs 
 and the thrusts of whose vaults are wholly or largely, directly or indirectly, 
 stopped by buttresses ; and also of doing work which possesses the chief charac- 
 teristics of buildings so constructed. 
 
 York, N. transept. 
 
 * Nor round Clymping Church. 
 
12 RELATION OF GOTHIC TO ROMANESQUE. 
 
 As we have enlarged the definition of Gothic, so we must enlarge that of 
 Romanesque. The majority of our greater Romanesque churches, e.g. Peter- 
 borough nave, have no high vaults ; a few, e.g. Carlisle nave, have not even aisle 
 vaults. Nevertheless, there is such a large amount of Romanesque about them 
 that they must be called Romanesque, not Basilican. E.g. in Carlisle nave there 
 is a clerestory passage and inner arcade ; the pier arches are in recessed orders ; 
 the cylindrical piers are of unclassical proportions ; the bases and strings are 
 unclassical ; there are scalloped capitals ; from the abaci rise roofing shafts ; there 
 are shafts in the jambs of windows ; there is a corbel table. A nave which has 
 so many of the characteristics which we find in such a church as Durham must 
 be classified with Durham as Romanesque. Therefore the definition may be 
 enlarged as follows : Anglo-Norman Romanesque architecture is the art ol 
 erecting buildings whose vaults are groined or have diagonal ribs, and the thrusts 
 of whose vaults are stopped by walls, not by buttresses ; and it is also the art ot 
 doing work which possesses the chief characteristics of buildings so constructed. 
 
 From what has been said above it will be clear that the difficulty of estab- 
 lishing a line of demarcation between Anglo-Norman Romanesque and Gothic is 
 very great. The connection between the two is of the most intimate nature. It 
 is difficult to exclude the nave of Durham from Gothic, without excluding at the 
 same time those of Wells and Salisbury. It follows that the idea that Gothic is 
 an individual and independent style is fallacious. Our Romanesque and our 
 Gothic are not two styles but one style. Gothic is perfected Romanesque ; 
 Romanesque is Gothic not fully developed, nor carried structurally to its logical 
 conclusion. This was recognised long ago by Mr Petit:* " The Romanesque 
 of Normandy, and still more of England, is essentially Gothic ; not indeed 
 fully developed, but quite sufficiently so to mark its direct and inevitable 
 tendency." So also M. Enlart: " L'architecture gothique n'est que la perfec- 
 tionnement de celle qu'on appelle romane." M. Anthyme St Paul takes the 
 same view :f " If, from an artistic point of view, Romanesque and Gothic seem 
 to be, and indeed are, two distinct arts, historically they are one and the same 
 art ; two phases of the same existence. Gothic is not superposed on Roman- 
 esque; has not supplanted or stifled it; on the contrary, it is its supreme result; 
 the last stage in its development ; its apogee, consummation and accom- 
 plishment." So also Comte de Lasteyrie+ says: "Gothic architecture did not 
 
 * Church Architecture, i. 93. 
 
 t On Viollet-le-Duc, 123. 
 
 X On the following page the main systems of high vaults are shown tabularly. The 
 perfected systems are those which include aisles, clerestory lighting, high vaults, and aisle 
 vaults. Peterborough has aisles, clerestory lighting, and aisle vaults, but no high vault. The 
 Perigueux churches have high vaults and clerestory lighting, but no aisles. *" Notre Dame, 
 Poitiers, and Issoire have aisles, aisle vaults, and high vaults, but no clerestory lighting. All 
 the rest, i.e. the four perfected types of Romanesque, have aisles, clerestory lighting, aisle 
 vaults, and high vaults. To the four perfected types may be added the abbey church of 
 I ouinus, which is sui generis. In this the barrel vaults of the nave were set transversely as in 
 the aisles of FOUNTAINS (ioi) nave and London Bridge. To the imperfect types may be 
 added certain Syrian churches with flat stone ceilings and aisles, but no clerestory lighting, e.g. 
 Kalb Louzeh, see 285. The curious church of Loches near Tours may be mentioned : two 
 bays of its unaisled nave are roofed with spires. 
 
CLASSIFICATION OF ROMANESQUE. 
 
 13 
 
 arise from a reaction against the principles of Romanesque ; on the contrary it 
 is the natural development of those principles ; the logical consequence of the 
 germ-idea of the Romanesque builders, of protecting the naves of their churches 
 by vaults of stone" {Discours, 17). 
 
 Vaulting 
 of Romanesque naves 
 
 Perfected 
 
 Imperfect 
 
 I 
 Domes 
 
 Barrel vaults 
 
 With groined With half 
 aisles barrels 
 
 in the 
 aisles 
 
 Intersecting 
 
 cylinders 
 
 1 
 1 
 
 Groined 
 
 Domes Barrel vaults Ceilings 
 
 On pier On 
 arcade triforium 
 arcade 
 
 o 
 
 - 
 
 T3 
 
 fc/J 
 
 — 
 
 U 
 
 — 
 
 o 
 
 - 
 
 — 
 
 z 
 
 
 A. 
 
 *$ 
 
 c 
 
 -« 
 
 < 
 
 • rx 
 
 u 
 
 u 
 
 Pi 
 
 Ribbed 
 (Romanesque) 
 
 --- 
 
 P 
 Ribbed 
 (Gothic) 
 
Chapter II. 
 
 CHARACTERISTICS OF THE ROMANESQUE ARCHITECTURE 
 
 OF ENGLAND 
 
 FROM IO50 TO C. 1200. 
 
 Anglo-Saxon Architecture — Number of Norman Churches— Size of the Norman Churches 
 Planning — Vaulting — Masonry — Foundations — Internal Elevation — External Ele- 
 vation — Clerestory — Abutment — Buttresses — Arches — Piers — Abacus — Capital — 
 Base — Roof Drainage — Ground Courses— Strings — Windows — Doorways — Towers — 
 Ornament. 
 
 NUMBER OF CHURCHES. — The history of Primitive Romanesque or Anglo- 
 Saxon or Pre-Conquest architecture in England is referred to only inci- 
 dentally in this volume ; a full account of it has been given recently in The 
 Arts in Early England by Professor G. Baldwin Brown. It was a backward 
 member of the great Romanesque family ; and was cut off untimely by the 
 advent from Normandy of another branch of the same family, which had there 
 reached a far higher stage of development. For a whole century the history of 
 English architecture is mainly the history of the development of the Romanesque 
 of Normandy. The history commences with the building of Westminster Abbey 
 by Edward the Confessor, which was commenced in 1050, sufficient of the eastern 
 part of the work being complete in 1065 to allow a consecration to take place. 
 The Romanesque of Normandy, therefore, had already found its way into this 
 country before the Norman Conquest. But after the Conquest the progress it 
 made far surpassed anything that had been done in its mother country. Within a 
 century the land was covered with churches, great and small. There was hardly 
 one of the greater Anglo-Saxon churches which was not rebuilt,* and a great 
 number of churches, entirely new, were erected. The resources of the Norman 
 bishops and abbots were of course vast ; conquered England had been divided 
 up in largess ; some of the grantees, ecclesiastics as well as laymen, counted their 
 manors by hundreds. Nevertheless when one remembers that the whole popula- 
 tion of the country was less than half of that of the present metropolis (4^ 
 millions), the bulk of building done seems incredibly great. Very many of the 
 churches then built have perished from the face of the earth; but even if a list be 
 confined to those which remain wholly or in part, or which have been rebuilt in 
 •Gothic, it is an astonishing record of the labour and the piety of the scanty popu- 
 
 * Hexham nave seems to be a solitary exception. 
 
ST. ALBANS, NAVE FROM S.W. 
 
SIZE OF ROMANESQUE CHURCHES. 15 
 
 lation of England in the eleventh and twelfth centuries.* Imagine all those 
 churches enumerated below, and many other great churches like Cirencester and 
 Coventry and Leicester, crowded into one-half of the present metropolis, together 
 with the vast number of parish churches rebuilt throughout Norman England, 
 and some idea may be formed of the enormous bulk of church building which 
 followed the Conquest. 
 
 Size of Churches. — Moreover the churches built from 1050 to 1200 were 
 not only exceedingly numerous, but very many were also amazing in scale, far 
 surpassing the very largest churches of their mother country, Normandy ; f so 
 large indeed that even in Gothic days nothing was set out on so vast a scale. To 
 the very last some of the proudest Gothic minsters remained content with 
 the dimensions that had been laid down in the eleventh century ; with the naves 
 of LINCOLN (151. 1) and Winchester, the transept and nave of CANTERBURY J 
 (149.3). Nowhere in Western Europe was there building in the eleventh century 
 on the gigantic scale of the Romanesque of England. 
 
 Planning of the Greater Churches. — The width of the great Norman 
 churches was conditioned only by the length of the tie-beams by which they were 
 spanned (572). Even when vaulting came into general use, the Romanesque 
 widths were not exceeded. The Norman naves of BURY (150.3), Peterborough, 
 and Gloucester, have a span of 35 feet, a width rarely exceeded in Gothic days. 
 
 In length the greater churches were still more remarkable; except in the 
 West of England ; the greater part of the length being given to the nave, 
 which at Ely and ST ALBANS (153.2) reached a length of thirteen bays, and 
 
 * Cathedrals of Benedictine Monks — Canterbury, Durham, lily, Norwich, Rochester, 
 J I 'inches ter, 1 1 orcester. 
 
 Churches of Benedictine Monks or Nuns — Battle, Bath, Binham, Blyth, Bury St Edmunds, 
 Chepstow, Chester St Werburgh, Colchester, Croyland, Glastonbury, Gloucester, Leominster, 
 Lindisfarne, Mailing, Malvern, Pershore, Peterborough, Ramsey, Reading, Romsey, St Albans, 
 Selby, Sherborne, Shrewsbury, Tewkesbury, Thoruey, Tutbury, Tynemouth, Waltham, West- 
 minster, Wyntondham. 
 
 Churches of Cluniac Monks — Lewes, Castle Acre, Wenlock. 
 
 Churches of Cistercian Monks — Waverley, Buildwas, Fountains, Fumess, Kirks/all, Louth, 
 Rievaulx. 
 
 Churches of Carthusian Monks — Witham, London Charterhouse, Mount Grace. 
 
 Churches of Premonstratensian Canons— Pradsole, Easby. 
 
 Churches of Gilbertine Canons — Old Ma/ton, Scmpringham, Wat/on. 
 
 Cathedral of Augustinian Canons — Carlisle. 
 
 Churches of Augustinian Canons — Bourn, Bridlington, Brinkburn, Bristol, Bolton, 
 Colchester St Botolph, Christ Church, Twynham, Dorchester, Dover, /'unstable, h'irkham, 
 Kenilworth, Lanercost, Lilleshall, London St Bartholomew's, Llanthony, St Frideswide's, 
 Oxford, St Saviour's, Southward, Thornton, Waltham, Walsingham, Worksop. 
 
 Cathedrals of Secular Canons — Chichester, Exeter, Hereford, Lichfield, Lincoln, London, 
 Old Sarum, Wells, York. 
 
 Churches of Secular Canons — Beverley, Chester St John's, Ripon, Southwell, Wiinborne. 
 
 t Mr Prior points out that the Abbaye-aux-hommes at Caen probably had originally an area 
 of less than 30,000 feet. But Norman Winchester and Old St Paul's occupied about 65,000 
 square feet ; while Bury St Edmunds had an area of 6S,ooo feet. Cluny, the largest mediaeval 
 church of the West, had an area of but 54,000 square feet c. 11 3 r. Gothic Art in Ping/and, 34- 
 
 :[ Bath Abbey Church, rebuilt in the sixteenth century, occupies the site of the nave only 
 of the Norman church. 
 
16 PLANNING OF ROMANESQUE CHURCHES. 
 
 at NORWICH (148.4) of fourteen. The choirs varied in length from the two bays 
 of Lincoln to the four of DURHAM (149.1); but by 1096 CANTERBURY ( 149.2) 
 had set out a new choir of nine bays. More accommodation being needed, instead 
 of further prolongation of the church east or west, which would have given it a 
 most unmanageable length, and would have been forbidden by the dimensions of 
 most sites, cross arms (transepts) were set out ; at first at the junction of nave 
 and choir; afterwards, in the Canterbury of 1096, projecting also from the choir 
 to north and south. This second or eastern transept was, however, rare till 
 Gothic days. But every great church, without exception, had a central transept ; 
 that of Bury St Edmunds was 234 feet long ; every great church was cruciform. 
 
 From each arm of the central transept there usually projected eastward one 
 apsidal chapel, as at NORWICH (148.4) ; more rarely two, as at ST ALBANS 
 (153.2). At DURHAM (149.1) instead of an apse there is an eastern aisle to 
 each arm of the transept. Winchester and ELY (153.4) have a western as 
 well as an eastern aisle. Both have, or had, north and south galleries as well. 
 
 In nearly all the larger churches, so far as we can judge from surviving 
 examples, it was usual for the nave to have a single aisle on each side. Ripon 
 built an unaisled nave c. 1170. 
 
 There was occasionally a highly developed western transept; as at BURY 
 (150.3) and Ely. 
 
 Up to the middle of the twelfth century all the choirs of the greater 
 churches ended in a semicircular apse, with the exceptions of Ely, Dover, 
 Southwell, Sherborne, and Romsey. All important choirs possessed aisles.* 
 But there were two entirely different ways of planning the choir aisles. One 
 was to terminate each aisle in a small apse parallel to the central apse of 
 the choir; e.g. DURHAM (149.1). The other was to continue the choir aisle 
 round the apse forming what is called an ambulatory ; and to construct, leading 
 out of it, apsidal chapels, usually three in number, pointing north-east, east, and 
 south-east ; e.g. NORWICH (160). 
 
 A solitary exception to these two plans occurs at ROMSEY (151. 3). Here 
 the ambulatory is rectangular instead of semicircular, and there were no chapels 
 leading out of it, except one to the east ; cf. Hereford and LLANDAFF (164). 
 
 Cistercian Planning.— But about the middle of the twelfth century 
 another influence of Continental architecture has to be taken into account. It is 
 no longer that of the Romanesque of Normandy, but that of Burgundy ; the 
 Romanesque amid which the monks of Citeaux, Clairvaux, Pontigny had been 
 bred. Of all the churchmen of the twelfth century the Cistercians were the most 
 influential ; the greatest of them, St Bernard, practically ruled Western Europe. 
 Vast numbers of Cistercian abbeys were erected c. 11 50. So popular was the 
 Order that in 11 52 the Chapter-General at Citeaux forbade the foundation of 
 more abbeys; a rule broken several times subsequently. In these abbeys, though 
 compelled perforce to adopt mainly the indigenous methods of construction of each 
 country, the planning was largely that of the mother abbeys in Burgundy : so 
 much so that identity of plan prevails in churches as far apart as Kirkstall in 
 Yorkshire, Maulbronn in Wurtemberg, Casamari in Italy, Fontenay in France. 
 
 * Lindisfarne, Melbourne, and others with unaisled choirs, are on a comparatively 
 small scale. 
 
17 
 
PLANNING OF ROMANESQUE CHURCHES. 
 
 19 
 
 In the earl_\- Cistercian churches in England the characteristic plan, e.g. at 
 KIRKSTALL ( 1 52.4), has a short unaisled choir, a transept with a row of eastern 
 chapels separated by solid walls; and an aisled nave, terminating sometimes 
 in a narthex; and, originally, without a tower or with only a low tower at the 
 crossing. Before the end of the century, however, more complex plans came 
 into use among the Cistercians both of Burgundy and England. 
 
 Planning of Parish Churches. — In the planning of the parish 
 churches much more variety prevailed. In fact almost the only common feature 
 the)" exhibit is that they always have a distinct architectural chancel. Unlike 
 the greater churches, this chancel was more often rectangular than apsidal. 
 
 The simplest plan was that of the type of ADEL (220) ; composed of a nave 
 and chancel ; without aisles ; without clerestory ; without tower. Sometimes, 
 as at Kirkburn, Yorkshire, there was 
 
 - 
 
 a western tower. 
 
 A second is of the type of 
 
 IIADLEIGH (214.4.) or NEWHAVF.N 
 
 (17) ; a tripartite church, composed 
 of nave, choir and sanctuary, with- 
 out aisles or clerestory ; usually with 
 a central tower over the choir. 
 
 A third is cruciform ; without 
 aisles or clerestory ; with a central 
 tower. Each arm of the transept 
 may have an eastern apsidal chapel, 
 as at North Newbald, Yorkshire. 
 
 A fourth has an aisled nave ; 
 an unaisled chancel; the nave usually 
 has a clerestory, and a western tower; 
 e.g. Steyning, Sutton St Mary, St 
 Margaret at Cliffe. 
 
 A fifth has an aisled nave and 
 clerestory, an aisled choir, a short 
 unaisled presbytery, and a western 
 tower ; e.g. Northampton St Peter's. 
 
 A sixth is similar to the fourth; but with the addition of unaisled transepts, 
 and with the tower central ; e.g. HEMEL HEMPSTEAD, Hertford. This last, which 
 is a town church, exhibits the highest development of parochial church planning 
 reached in the twelfth century. 
 
 Romanesque Vaulting.— When once the plan of a church had been 
 settled — settled mainly by considerations of ritual — the rest of the task lay with 
 the builder. With him the first thing to consider was how to roof over the area 
 at his disposal : for the primary object of architecture is the provision of shelter. 
 And a secondary object, insisted upon with astonishing persistence in the middle 
 ages, was that the church, if large, should be fireproofed by building beneath 
 the roof a stone ceiling or vault. In the smaller churches, except here and there 
 in an unaisled chancel or beneath a tower, few attempts were made to build 
 vaults. In the greater churches there are a few examples, e.g. the naves of 
 
 Hemel Hempstead. 
 
20 
 
 ROMANESQUE VAULTING. 
 
 Carlisle and Rochester, where no vault was ever built even over the aisles. And 
 in the vast majority of the greater churches no vaults were built in the eleventh 
 and twelfth centuries on the clerestory walls ; i.e. no " high " vaults. They were 
 simply ceiled, as some remain to this day ; e.g. Peterborough and ELY (57). 
 The only examples of Romanesque high vaults remaining, wholly or in part, in 
 aisled churches, are those of Durham, Lindisfarne, and St Cross, Winchester. 
 
 Various forms of vault were in use. The barrel vault was employed in ST 
 JOHN'S CHAPEL in the Tower of London (283); and by the Cistercians at 
 Kirkstall and FOUNTAINS (101). The half-barrel is in use at GLOUCESTER 
 (282). Semidomes were employed in many an apse ; e.g. CHECKENDON 
 (21). But the favourite vault in the eleventh century was the quadripartite 
 groined vault. All the Norman crypts were roofed with it ; and some of the 
 aisles; e.g. NORWICH (238). It does not exist in any high vaults that remain in 
 England. Square, oblong, triangular, and trapezoidal spaces were all roofed by 
 means of the groined vault. 
 
 For the quadripartite vault with groins there was soon substituted the 
 quadripartite vault with ribs: first in DURHAM CHOIR (315), commenced in 
 1093. In these ribbed vaults the transverse ribs were usually rectangular in 
 outline and massive, while the diagonals were molded and were lighter. All 
 the ribs were much more massive than those in Gothic work. The diagonal 
 arches were rarely elliptical ; more often segmental or semicircular ; the trans- 
 verse arches were usually stilted. The pointed arch in a vault first appears in 
 that of DURHAM NAVE (8). The voussoirs were small, and consequently 
 numerous. Being filled in with rubble, both groined and ribbed vaults were 
 very heavy, and required very massive supports. Bosses were not employed till 
 well on in the twelfth century, and they were of small projection. 
 
 Masonry. — The vault and roof of the greater churches rest on aisle walls 
 and clerestory walls ; and the clerestory walls rest on arches and piers. All 
 three — piers, arches and walls — were exceedingly massive. All three were faced 
 with ashlar ; this, however, was but skin deep ; the core was a mass of uncoursed 
 rubble. Of the various methods of construction that had been in use in ancient 
 Rome one had employed faces of ashlar with a core of rubble laid in horizontal 
 layers on a bed of mortar.* This apparently was copied by the Romanesque 
 builders both here and on the Continent ; with the exception that they employed 
 much smaller blocks than those in Roman work. Owing to the badness of the 
 roads and the unbridged rivers, land transport was exceedingly difficult. Water 
 transport was employed, wherever possible. Norwich and Peterborough Cathe- 
 drals are built of stone from the Barnack quarries in Northamptonshire ; that for 
 Peterborough carried down the Nene ; that for Norwich down the Welland, then 
 by sea, then up the Yare ; so also the abbeys of Ramsey, Croyland, Thorney, 
 Ely, Bury St Edmunds ; and the churches at Stamford, Ketton, and Kettering.f 
 Caen stone was sent across sea to Chichester and other cathedrals. Even at 
 the end of the twelfth century, Christ Church, Dublin,} was built of Somerset 
 Oolite. Most of the stone used at Christ Church is of uniform scantlings of 
 2 ft. x 1 ft. x 1 ft. 
 
 * See Choisy's Roman Building, 17. t Assoc. Soc. Reports, xxiii. 143. 
 
 X Builder, May 5, 1894, 350. 
 
2\ 
 
 Checkendon Apse. 
 
ROMANESQUE MASONRY. 23 
 
 To lessen the cost of transport, the blocks were probably roughed out at the 
 quarry. They were then dressed on the bench ; not worked in position, because 
 to do so would have shaken the mortar beds. The capitals, however, seem 
 frequently to have been set up in the rough, and carved afterwards. 
 
 Usually there is one stone for one member of a design ; e.g. there was one 
 stone for the abacus, and one for the capital, till as late as the fourteenth century. 
 It is exceptional to find a capital, like some of the crocket capitals in the Saint's 
 Chapel at Canterbury, composed of two blocks superposed. Stone, owing to the 
 cost of transport, was dear ; handicraft was cheap. Every bit of ashlar was 
 utilised, whether to size or not. If a pattern, e.g. of zigzags, was to be carved 
 round an arch, and one voussoir of the arch was broad, and another narrow, on 
 each was carved a single zigzag ; though in the latter case it was too much 
 cramped, and in the former spaced out too much. 
 
 The stones were usually of moderate size ; again owing to difficulties of 
 transport ; it might be necessary to convey them for some distance by pack horse 
 or cart. They average from about 1 foot square upwards: in early work 
 they are usually rather square than oblong. They are smaller in early than 
 in late work. 
 
 A toothed hammer or axe (bretture) was used in dressing the blocks ; 
 and except in dressing shafts, was used with a diagonal stroke. These marks 
 or hatching are often preserved where the block is a good freestone. In 
 Normandy the hatching is less close in eleventh than in twelfth century work ; 
 the teeth of the tool then used being further apart than later.* Both in Roman 
 and in Anglo-Saxon work hatching occurs occasionally: e.g. in the Roman wall, 
 Northumberland; and in the Anglo-Saxon doorways of Kirkdale and Sherborne.f 
 
 The joints were usually thick, especially in early work. But in late work 
 also, if a porous \ stone was employed, thick joints were necessary. As a rule, 
 however, eleventh-century may be distinguished from twelfth-century masonry 
 by the thickness of the joints. The difference is well seen in the north tran- 
 sept of Winchester ; where the portions with thick joints belong to the work 
 
 * Ruprich- Robert, i. 171. 
 
 t Mr Xeale found that at St Albans the Xorman work is axed ; the Transitional work is 
 chiselled ; the Early English work is bolster-tooled ; the Decorated ashlar is claw-tooled ; the 
 moldings scraped; the Perpendicular finely scraped. Journal of R.I. I',. A.. 1S77. 80. Recent 
 investigations of Mr E. S. Prior have shown that twelve different styles of masoncraft may be 
 recognised in Chichester Cathedral between 1190 and c. 1450. (1.) Dressing with the pick, 
 and with the axe diagonally, c. 1105. (2.) Dressing with the axe diagonally and coarsely, 
 but obliterating the pick marks, c. 11 20. (3.) Caen stone dressed with the axe diagonally and 
 finely, c. 11S0. (4.) Dressing with the axe diagonally; shafts dressed vertically,;". 1 195. (5.) 
 Dressing with the axe diagonally (shafts vertically) and finely : mixed with which is dressing 
 with a claw chisel, the notches six to the inch, 0. 1205. The same mixture of dressing is seen 
 in St Hugh's work at Lincoln, c. 1200. 6.) Dressing with axe vertically to the bed. Diagonal 
 axing also occurs in the vestry, c. 1210. (7.) Dressing with claw chisel with 9, 12, and 14 
 notches to the inch ; bosted always vertical to the bed, <. 1235. 8.) Dressing with claw chisel, 
 of from S to 12 notches to the inch, not always vertical to the beds, but often irregular and 
 crossing diagonally, c. 1260. (9.) Dressing finely with claw chisel, with notches 16 or 20 to the 
 inch, often used as a drag, often crossing one another,!". 1290. (10.) Dressing finely dragged 
 or diagonally clawed, c. 1335. (11.) Dressing diagonally chopped, c. 1420. (12.) Dressing 
 dragged smooth, c. 1450. See Proceedings of Harrow . Irchitectural Club for 1904. 
 
 % Cf. Brutails, 222. 
 
24 ROMANESQUE MASONRY. 
 
 commenced in 1079 ; while the portions with thin joints were built after the fall 
 of the central tower in 1 107. 
 
 Both walls and piers were exceedingly massive ; e.g. the walls of Durham 
 choir aisles have an average thickness of 7 feet. The tradition long survived in 
 English Gothic of trusting rather to thickness of wall than to buttress, flying 
 buttress, and pinnacle ; the thirteenth-century clerestory wall of Salisbury is 
 nearly 7 feet thick at the top.* Equally massive were the Romanesque piers. 
 It has been calculated! that the major piers of the choir of DURHAM (659.1) 
 occupy seventeen times as much space as the cylinders of the choir of CANTER- 
 BURY (106), erected some eighty years later. But the tradition long sur- 
 vived of the massive Romanesque pier as well as of the thick Romanesque wall. 
 Such piers as those of the choirs of WELLS (424.1), LICHFIELD (244), ST 
 SAVIOUR'S, SOUTHWARK (521), have more of Romanesque stability than 
 Gothic grace. 
 
 As we have seen, the substance of wall and pier was rubble and mortar. 
 Everything depended on the quality of this mortar. It seems sometimes to 
 have been excellent ; e.g. in the Bishop's Palace at Winchester ; j at Worcester ; 
 and at Gloucester. When one recollects how Gloucester choir has been pulled 
 about ; how the Norman walls have been made to carry a tall Gothic clerestory 
 and a heavy vault ; and how an enormous Gothic central tower has been poised 
 on the Norman piers of the crossing, it is plain that here at any rate the Norman 
 masonry must have been good. At Binham, too, and elsewhere one may see 
 great masses of Norman masonry hanging on still by the cohesion of the mortar, 
 though their supports have collapsed or have been removed. But this was by 
 no means always so ; § e.g. at Hereford Mr Cottingham found in 1843 that the 
 core of the piers of the central tower was composed of " broken stones, loam, and 
 lime grouting" ; so that the fourteenth-century tower superposed on them really 
 had for support nothing but the thin shells of ashlar which enclosed the core. 
 But this ashlar, not being well bonded and deeply headed into the rubble cores, 
 had split and bulged ; and the core itself was crushed to pieces for want of a proper 
 proportion of lime in the mortar. In Old St Paul's, Sir Christopher Wren found 
 that the piers of the nave were " only cased without, and that with small stones, 
 not one greater than a Man's Burden ; but within is nothing but a Core of small 
 Rubbishstone and much Mortar, which easily crushes and yields to the weight." 
 At St David's the cores of the walls of the central tower had disintegrated into 
 dust; and when a hole was made, the core "began to pour out like an avalanche." 
 Sir Gilbert Scott saw ten buckets of liquid cement poured into one hole. || Nor 
 was bad building unknown in Gothic days. In the west front of PETERBOROUGH 
 (112) the mortar in the joints of the ashlar had crumbled into dust, and the 
 blocks could be lifted from their positions by hand. In the south transept of 
 
 * On the other hand the aisle walls of Patrington, c. 1340, are only 2 feet 3 inches thick, 
 though intended to carry a vault ; those of the Temple Church, Bristol, are 1 foot ioi inches. 
 
 t Prior, 34. 
 
 I See Willis' Winchester, 72 ; and E. Christian in Journal of R.I. B. A., 1877, 151. 
 
 § For the condition of the interior of the piers of Sherborne central tower see R. H. 
 Carpenter in Journal of R J. B. A., 1877, 149. 
 
 |] Report to Dean ami Chapter, 1869. 
 
ROMANESQUE FOUNDATIONS. 25 
 
 YORK MINSTER (523) it was found in 1 87 1 that the cure of the clerestory 
 walls had been made up of stone chippings without mortar. As late as 1323, 
 after the fall of two Norman bays at the east end of the north side of the nave 
 of St Albans, the cores of the new piers were built with such bad mortar that it 
 was found recently that they had disintegrated into dust, and the whole weight 
 of the superincumbent walls was carried by the casing of ashlar. 
 
 FOUNDATIONS. — Equally varied was the practice of the Romanesque builders 
 with regard to foundations. They knew perfectly well what was the right thing 
 to c!g ; sometimes they deliberately did the wrong. Frequently their foundations 
 were both deep and broad. The foundations of the three eastern apses of 
 Norman DURHAM (149.1) were carried clown more than 14 feet, till the solid 
 rock was reached. Those of the wall 1 if the north choir aisle are so broad as to 
 provide a footing both for the buttresses outside and the bases of the vaulting 
 shafts within.* Lord Grimthorpe found that " the foundations of the piers of 
 St Albanst are singularly large and strong." At Ely} the foundations of the 
 thirteenth-century presbytery are about 6 feet deep and rest on the rock. But 
 those of the Norman choir were only 4 feet 6 inches deep and did not go down 
 to the rock. In the Lady Chapel of Glastonbury§ the foundations consist of a 
 rubble wall 12 feet or more deep; so that when a crypt was wanted in the 
 fifteenth century, all that was necessary was to clear out the soil between 
 the foundation walls. At York the first stone of the foundations of St Mary's 
 Abbey || was laid in 1271 at a depth of 9 feet. In places, however, the founda- 
 tions were 24 or 26 feet deep. 
 
 Moreover the builders sometimes took the trouble to provide continuous 
 foundations from pier to pier. Professor Willis' says that he "saw that 
 at Lichfield, El}', Hereford, and elsewhere, the ranges of piers were set on con- 
 tinuous foundations,** walls of rubble constructed with the greatest care." In 
 Gothic work one of the best examples of good building construction is Lincoln 
 nave; in this there are transverse walls underground from pier to wall, as well as 
 longitudinal ones from pier to pier. 
 
 But the temptation to economise on the foundations was not always resisted. 
 At Gloucester the north-west tower fell in 1170; "because of bad foundations," 
 says Giraldus Cambrensis. At Croyland there is a bed of gravel underlying peat. 
 The gravel is about 11 feet from the surface; but the peat was excavated for 
 6 feet only; and the foundations consist largely of layers of quarry dust. This 
 culpable carelessness about foundations is not without parallels in Gothic work. 
 The thirteenth-century Lady Chapel of Chester Cathedral has been found to 
 have been built without foundations of any sort. Peterborough is especially 
 noteworthy among our greater churches for insufficiency of foundations ; the 
 
 * Mr Bilson in Archaeological Journal, liii. 3. 
 
 + The great care with which the foundations of St Albans were prepared is described 
 in Buckler's St Albans, 35. 
 
 % See Journal of R.I.B. A. , [an. 3, 1S76, 70, 71, 79, 80 ; and Stewart's Ely, 20. 
 
 S Willis' Glastonbury, 63. 
 
 || Rickman, 175. 
 
 H Glastonbury, 63. 
 
 ** Continuous foundations are exposed to view at St Mary's Abbey, York. 
 
26 ROMANESQUE INTERNAL ELEVATION. 
 
 Norman portions, the thirteenth-century west front, the eastern chapels of the 
 fifteenth century were all built without proper foundation.* 
 
 Internal Elevation. — Not only were the supports of the Romanesque 
 churches exceedingly massive, but the walls were pierced with but few and small 
 openings. Where in such a Gothic clerestory as that of SALISBURY (i 70) there 
 would be three windows, or as at Exeter one broad window stretching from 
 buttress to buttress, in a Romanesque church there was but one window, and 
 that a small one. So with the aisles In a Gothic church the voids pre- 
 ponderated over the solids; in a Romanesque church it was the very reverse; the 
 building was almost wholly solid. This solidity, this monumental stability, is 
 the special excellence and merit of Romanesque design. The lightness and 
 grace that were already attained in large degree in ELY (57) and Peter- 
 borough are not half so impressive as the massive grandeur and gloom of the 
 earlier work of WINCHESTER (261) and DURHAM (8). 
 
 Internally, all the great churches were three stories high. At the top was 
 the clerestory wall ; at the back of which was a single window in each bay, and in 
 front of the window usually a triple arcade ; e.g. ELY (273). At the bottom 
 was the range of piers and arches — the pier arcade — separating the nave from 
 the aisles. Between the pier arcade and the clerestory was the front wall of 
 the triforium chamber, usually pierced with an arcade. The proportions of these 
 three stories vary very considerably. They were largely controlled by the 
 dispositions adopted for securing adequate light for the central aisle or nave. 
 If the pier arcade was low, the light from the aisle windows was obstructed. 
 Especially was this so, where there was a cloister roof outside one of the 
 aisle walls; unless the aisle windows were set high in the wall, they would 
 not clear the cloister roof. But if the aisle windows were set high, the piers 
 and their arches must be lofty also. Where they were set high, an eleva- 
 tion resulted in which the pier arcade was lofty, and the triforium arcade 
 comparatively small; e.g. GLOUCESTER NAVE (26). In such a design no 
 windows were inserted at the back of the triforium, or at any rate, only 
 small ones. 
 
 But an alternative method was much in favour. This was to raise the aisle 
 wall, and to insert an entirely new row of windows in the upper part of it, which 
 became a back wall to the triforium chamber. And as the light from these 
 windows was wanted for the nave, there could be no solid wall in front of the 
 triforium. Either the triforium chamber opened into the nave by one great arch, 
 as at ST albans (14); or if there were two arches, they were constructed 
 lightly so as to obstruct the light as little as possible ; e.g. ELY (57). In this 
 design so much of the height of the interior was absorbed by the' triforium 
 arcade, that the pier arcade was usually low. 
 
 Besides the two above methods of designing a Romanesque interior, both 
 more or less logical, there were illogical variants and compromises. Thus 
 Durham has windows in the upper part of the aisle wall, but the triforium is low 
 and is blocked by massive arches in front. On the other hand, Romsey has a 
 tall triforium with a light open arcade, but no windows at the back of it. St 
 
 * This is the more remarkable as there is solid limestone rock a feu- feet below the 
 foundations. 
 
ROMANESQUE INTERNAL ELEVATION. 
 
 Bartholomew's, Smithfield, has 
 magnified its triforium at the 
 expense of its pier arcade ; 
 though here also the triforium 
 is a blind-story. 
 
 Another eccentric design is 
 that which is seen in OXFORD 
 CATHEDRAL (525), in which 
 the containing arch of each bay 
 of the triforium is continued 
 down to the ground ; with the 
 result that what is a low three- 
 storied is made to appear a tall 
 two-storied interior. 
 
 Another curious design is 
 that of St Botolph's Priory, Col- 
 chester, where again a three- 
 storied is made to look like a 
 two-storied elevation ; but in this 
 case by absorbing the triforium 
 into the clerestory. Towards 
 the end of the twelfth century 
 this design reappears at ST 
 
 DAVID'S f525). 
 
 In the Cistercian churches 
 usually no windows were inserted 
 in the triforium ; and being con- 
 sequently a blind-story, a solid 
 wall was built in front of it ; e.g. 
 FOUNTAINS NAVE (iOl). 
 
 Walls which do not rest on 
 pier arcades; e.g. the east wall of 
 the south transept of Norwich 
 (168); are also usually divided 
 into three stories, which may or 
 may not be similarly propor- 
 tioned to those of the aisled 
 portions of the rest of the 
 church. 
 
 External Elevation.— 
 
 The usual elevation is one of 
 two stories ; aisle wall and 
 clerestory. But if the triforium 
 has windows, there are three 
 stories, which, subdivided by 
 strings, produce four stories at 
 Ely, six at NORWICH (31). 
 
 ( Kford Cathedral Choir. 
 
28 
 
 ROMANESQUE EXTERNAL ELEVATION 
 
 Eastern Facade.— Of the churches ending eastward in three parallel apses 
 we have no example complete. Peterborough retains the central apse, but has 
 lost the side apses. Romsey retains the eastern side apses, but never had a 
 central apse. The original form of our east ends may, however, be seen in 
 Normandy: e.g. in ST GEORGE'S DE BOSCHERVILLE (160) and CERISY-LA- 
 FORET (160). Of churches with the ambulatory plan and radiating chapels 
 
 Durham Cathedral. 
 
 we have good examples at NORWICH (148.4) and GLOUCESTER (135) ; though 
 in both the eastern Norman Chapel was pulled down in the thirteenth century, 
 to make room for a rectangular Lady Chapel. Good examples of rectangular 
 east fronts survive at Darenth, Barfreston, and PATRIXBOURNE (218). 
 
 Of West Fronts the most important left are those of SOUTHWELL (520), 
 Rochester, and Durham Cathedrals, and of Tewkesbury and Castle Acre ; * but 
 
 * See Plate in Button's Arch. Ant., iii. 
 
ROMANESQUE CLERESTORY. 29 
 
 they have been altered by the insertion of a big window in the centre of each. 
 In Normandy, however, the west front of the Abbaye-aux-hommes remains 
 almost as it was built. Fine west fronts remain in the parish churches of Castle 
 Rising and Iffley. 
 
 Of the Transept Fronts, one, in a church of Monks or of Regular Canons, 
 adjoined the dormitory and other buildings east of the cloister. Where these 
 have been torn away, e.g. from the south transept of Ely, the result is necessarily 
 an unsightly space of blank wall. But the elevation of the other transept is 
 frequently one of much grandeur; e.g. the north* transepts of Winchester, Ely, 
 Peterborough ; and of NORWICH (31); finest of all. 
 
 CLERESTORY. — There was never but one window in each bay. At Southwell, 
 by exception, the clerestory windows are circular. In the Cistercian churches, 
 e.g. FOUNTAINS (101), and a few others, e.g. Leominster, there is no passage. 
 Nearly always there is a passage in the thickness of the wall. In front of this 
 usually there are three arches, of which the central one is the highest, e.g. ELY 
 (57). Comparatively few parish churches had clerestories ; and then usually not 
 till late in the twelfth century. 
 
 Abutment System. — We now come to the most difficult problem of the 
 mediaeval builders ; which was not how to erect a building, but how to keep it 
 up. If a transverse section of one of the great Romanesque naves be examined, 
 e.g. that of ELY NAVE (34. 1); it will be seen that the navet contains two 
 high walls (" clerestory " or " nave walls ") nearly 73 feet high, and two low walls 
 ("aisle walls "j nearly 50 feet high. What keeps them from falling over to north 
 or south ? The nave walls have nothing whatever to keep them from inclining 
 inwards except their vast weight, and the fact that the) - rest on adequate founda- 
 tions. The aisle walls have nothing to prevent them from inclining outwards 
 except buttresses of such slight projection as to be utterly inadequate for the 
 purpose. Like the nave walls, they remain vertical simply because of their great 
 weight and their good foundations. On the other hand, the lower half of the 
 aisle walls is prevented from inclining inward by the loaded arches of the vault ; 
 which thrust outward like a compressed spring. And in the same way the lower 
 part of the high wall — in Ely about 28 feet out of Ji feet — is kept from inclining 
 outwards by the thrust of the same vault, that of the aisles. And as the load on 
 the arches of a Romanesque vault was very heavy — for they were filled in with 
 a thick mass of rubble — the pressure brought to bear by the aisle vaults against 
 the walls on either side was very considerable. So considerable was it that 
 sometimes the aisle wall has been thrust outwards. And though the nave wall 
 is loaded with an enormous weight of masonry, extending up to the top of the 
 clerestory wall, it has been noted — e.g. by Sir Christopher Wren — that the pier 
 arcade on which the nave wall rests has not infrequently received a considerable 
 inclination inwards, owing to the thrust of the aisle vault. So valuable was the 
 aisle vault in the construction of their larger churches that it is very rare to find 
 the Romanesque builders omitting it. Indeed in churches of the first rank in 
 England only two cases seem to occur ; viz. the naves of Rochester and 
 
 * The normal position for the cloister was to the south of the nave ; hence the principal 
 transeptal facade was normally that of the north transept. 
 
 t The roof of the nave is omitted : those of the aisles are shown in dotted lines. 
 
3 o ROMANESQUE ABUTMENT. 
 
 Carlisle cathedrals : in the former of which, however, pilasters were built against 
 the eastern part of the aisle wall, as supports for a vault which at first apparently 
 was contemplated. 
 
 The weakest point of such a construction as that of Ely is that abut- 
 ment is applied to the nave wall so low down. The first improvement on 
 such a section is seen in CHICHESTER CHOIR (34-4). and DURHAM CHOIR 
 (370). It consisted in building semicircular arches in the triforium chamber 
 between the nave wall and the aisle wall. If this had been done at El)-, 
 about 37 feet of the Ji feet of the nave wall would have got abutment, instead 
 of only 28 feet. 
 
 Again, just as the aisle vault prevents the nave wall and the aisle wall from 
 bulging towards one another, so a high vault over the nave, ever trying to 
 expand between the two nave walls, prevents them from bulging inwardly. 
 Such vaults were actually constructed at Durham ; where, in all probability, the 
 high vaults of the nave were built between 11 28 and 1133, and those of the 
 transepts earlier still. 
 
 There can be no question as to the potency of the new ally. With a heavy 
 vault between them, it was utterly impossible for the nave walls to incline 
 inwards. The danger lay in the other direction. So far from being insufficient, 
 the force exerted by the new agency was only too great. The high vault was 
 always tending to thrust apart the clerestory walls. So much so that the high 
 vault built over Durham choir, probably before 1104, collapsed early in the 
 thirteenth century. Evidently the next thing to do was to provide a remedy 
 to prevent the clerestory walls from bulging out. The remedy applied — once 
 more at DURHAM (370) — in its day one of the most advanced churches of 
 Western Europe in science of construction — was to build in the triforium 
 chamber, not arches as in that of the choir, but flying buttresses. This made 
 safe some two-thirds of the height of the nave wall. For the high vault, as may 
 be seen on examining the photograph of the nave on page 8, does not spring 
 from the top of the clerestory wall, but from a level considerably below it. So 
 that it was unnecessary to provide abutment for the clerestory wall much above 
 the springing level. This solution — that of Durham nave — is the one adopted 
 even in much of our Gothic architecture ; e.g. at Wells, at Salisbury, at Tintern, 
 even in Winchester nave in the remodelling commenced c. 1360. And where the 
 high vault springs low down in the clerestory wall, it is entirely scientific and 
 satisfactory. 
 
 If, however, it was desired that the high vault should spring at a higher 
 point, then it was necessary to take the flying buttresses out of the triforium 
 chamber, and to build them above, instead of beneath, the triforium roof. This it 
 was left for the Gothic builders to do ; viz. in the choir of CANTERBURY (34.3), 
 II7S ; CHICHESTER (34.4), 1 1 84; LINCOLN (34.5), II92. 
 
 In the parish churches these difficulties of abutment very seldom presented 
 themselves ; it was most exceptional for a parochial aisle to be vaulted.* 
 
 Buttresses.— These were for the most part little more than decorative 
 pilasters; e.g. at STEYNING (359); and so late as 1175 in WELLS CHOIR (373). 
 
 * A fragment of the groined vault of the aisle remains at St Peter's, Canterbury. 
 
3' 
 
 Norwich North Transept. 
 
ROMANESQUE ARCHES. 33 
 
 But the Cistercians by the middle of the twelfth century were building effective 
 buttresses of considerable projection at Fountains and KIRKSTALL (152.4). 
 
 ARCHES.— One of the special marks of mediaeval architecture is that it is 
 above all things an arcuated, not a trabeated style (257). It may be said that 
 Greek architecture is a trabeated style pure and simple ; that Gneco-Roman 
 work was a mixed style, partly trabeated, partly arcuated ; but that Romanesque 
 and Gothic are wholly* arcuated styles. 
 
 Up to about the middle of the twelfth century the semicircular arch was 
 employed almost exclusively. Here and there a segmental or an elliptical arch 
 occurred in the vaulting or in the heads of doorways. The pointed arch 
 had indeed been introduced in Gloucester choir apses and DURHAM NAVE (8), 
 and at Rochester ; but chiefly in vaulting. In pier arcades the pointed arch was 
 first employed by the Cistercians at FOUNTAINS (101) and Kirkstall, about 
 the middle of the twelfth century. Some time elapsed, however, before it came 
 into general use in doorways, windows and ornamental arcading. To the very 
 end of the century conservative builders were still building their pier arches 
 semicircular ; not only in village churches such as SUTTON ST MARY (42), but a 
 great cathedral as ST DAVID'S (525). In the last half of the century trefoiled 
 arches also occur in doorways and wall arcading. 
 
 All the larger arches were built in recessed orders ; not built like the 
 BRIXWORTH ARCH (274), but as in page 272. At first the edge of each 
 order was left square, as in the transept of WINCHESTER (261) (choir com- 
 menced 1079) and Blyth, founded in 1088. f But very soon, e.g. in Chichester 
 choir, commenced c. 1088, either the edges were rounded off into roll moldings, 
 or the faces were covered with carved ornament. But in the West country, 
 which seems to have had its own school of Romanesque as well as of Gothic, 
 it is very common to retain the plain square-edged arch, without molding or 
 ornament, far into the twelfth century. Examples of unmolded and uncarved 
 Romanesque arches are the pier arcades of Holy Cross, Shrewsbury, Malvern 
 and Leominster naves, and St John's, Chester. The pier arches of the nave of 
 TEWKESBURY (297), consecrated 1123, are but slightly molded. At Romsey, 
 of the four orders of the pier arches of the nave, three are square-edged ; only 
 the outer one is carved with the chevron. At Hereford the arches are much 
 carved; but little molded. In GLOUCESTER NAVE (313) and at Christ Church, 
 Hants, the pier arches are molded ; but the moldings are few and heavy. On 
 the whole, the Romanesque of the West of England is characterised by the small 
 progress made in molding the arch. 
 
 Piers. — In the eleventh century cylindrical piers seldom occur, except in the 
 West of England, as in the choirs of <;i.< >r< kstkk 1294, and TEWKESBURY 
 (165), where they are short and massive. As a rule either all the piers are 
 compound, as at NORWICH (238) ; or compound piers alternate with cylinders, 
 as at DURHAM (239). In the twelfth century cylinders alternate with octagons 
 in PETERBOROUGH CHOIR 318) ; while in the West of England and Southwell 
 nearly all the naves have cylinders; which, at GLOUCESTER (99) and TEWKES- 
 
 * Lintelled doorways arc an exception. 
 
 + It is not necessary to take into account the square-edged arches of St Albans and of St 
 Botolph's, Chichester. They were square edged simply bei ause they were built of bricks. 
 
 C 
 
34 
 
 1. Ely Nave. 3. Canterbury Choir. 
 
 2. Durham Nave. 4. Chichester Choir on left, Nave on right. 
 
 5. Lincoln Choir. 
 
35 
 
 SCALE or FEET 
 
 1. Temple Choir. 3. Exeter Choir. 
 
 2. Westminster Choir. 4. Bristol Choir. 
 
 v Gloucester Choir. 
 
36 
 
 St Mary's, Guildford, North Apse. 
 
ROMANESQUE PIERS. 37 
 
 BURY (297), are of enormous height and bulk. Where the compound pier is 
 employed, it contains, in the best examples, e.g. DURHAM (659. 1 ), a separate 
 shaft or column for each order of the arch and for each rib of the vault. 
 
 In the aisled parish churches, few of which, if any, are earlier than the twelfth 
 century, the pier is almost always a cylinder. NORTHAMPTON ST PETER'S 
 (663.1) is an exception; in this compound piers and banded columns are 
 employed. 
 
 ABACUS — The Norman abacus is always square-edged. Its under surface 
 is usually a straight chamfer, as at YOULGREAVE (421.4) ; or a hollow chamfer, 
 as at CANTERBURY (417.7). In plan it is usually square ; but the cylinders of 
 GLOUCESTER CHOIR (99) have circular abaci and capitals ; another peculiarity 
 of West of England Romanesque. At DURHAM C239) and Buildwas cylin- 
 drical piers have octagonal abaci. Abaci logically subdivided appear as early as 
 the eleventh century in ELY TRANSEPT (506). 
 
 Capitals. — There is a great variety of Romanesque capitals. Imitations 
 of debased Roman versions of the Corinthian and Composite capital are frequent, 
 especially in the eleventh century. At first the band of acanthus is usually 
 omitted ; in the twelfth century it is attempted ; e.g. at CANTERBURY (417.7). 
 These Corinthianesque capitals survive to the very end of the twelfth century. 
 The most common of the Norman capitals is the cubical or cushion cap ; e.g. 
 CANTERBURY (43OJ. At Peterborough hardly anything else occurs. Usually 
 it is a little scalloped. When much scalloped or coniferous, it is usually late ; 
 e.g. in the apse of ST MARY, GUILDFORD (36). In the last quarter of the 
 century, the incurved cone is frequent in the West of England work ; e.g. 
 ST DAVID'S '412.5;. Another capital which persists to the end of the twelfth 
 century is that with interlacings ; e.g. ELY (412. 1). In the last quarter of the 
 century attempts are made here and there to render naturalistic foliage. The 
 water-leaf cap is very characteristic of the period c. 1165 to c. 1190; e.g. 
 WALSOKEN '417.2). 
 
 BASE. — The Norman base is at first quite insignificant ; altogether dispro- 
 portionate to the great spread of the capital. Its moldings are usually of the 
 simplest and rudest. Little attention was paid to the base till well on in the 
 twelfth century ; when a variant of the Attic base was adopted, with flattened 
 lower roll. The plinth was either square ; or if the pier was compound, separate 
 rectangular plinths were provided for the shafts and columns of the pier ; e.g. 
 DURHAM (659.1). The "spur" ornament may occur, where the plinth is square; 
 e.g. NORTHAMPTON (663.1). 
 
 ROOF DRAINAGE. — The roofs had a fairly steep pitch ; as is shown by the 
 weatherings on TEWKESBURY TOWER f 390 ■. The upper courses of the walls, 
 except at Ely, projected on corbels or corbel arches, and the roof coverings 
 again projected beyond these; e.g. SOUTHWELL TRANSEPT (390). For this 
 system of "dripping eaves" the Cistercians substituted gutter, gargoyle, and 
 parapet at FOUNTAINS (385.6), Kirkstall, Roche, and Byland. 
 
 GROUND COURSES. — At first the importance of protecting the foot of the 
 wall from drip and splash was little recognised. Round the base of the 
 twelfth-century work at Hereford, however, a basement course, semicircular 
 in section, exists. About the middle of the century the Cistercians built base- 
 
3 8 ROMANESQUE STRING COURSES. 
 
 ment courses at FOUNTAINS (679.1) and Kirkstall of considerable height and 
 projection. 
 
 STRING COURSES.— On the other hand, strings were employed from the 
 first in great numbers ; not only to shelter the walls from drip, but merely 
 ornamentally. Owing to the great amount of wall space in the Romanesque 
 churches, strings were of great decorative importance. In the strings, carving 
 was employed as well as molding. 
 
 WINDOWS.— The balustered window, being unsuited for glazing, was con- 
 fined to towers ; the baluster was generally set near the outer face of the wall. 
 The usual window was oblong and round-arched ; set near the outer face of the 
 wall, and much splayed internally. In the jambs were frequently set decora- 
 tive shafts. The clerestory window of the greater churches was usually orna- 
 mented with an inner arcade ; e.g. ELY (57). With the exception of a solitary 
 example at ROMSEY (457.2), there is no grouping of aisle or clerestory windows 
 till GLASTONBURY LADY CHAPEL, 1 1 86 (465). On the other hand, circular 
 windows were highly developed ; e.g. PATRIXBOURNE (218). 
 
 DOORWAYS. — The oldest type of doorway is that at ELY (39), with lintel 
 and tympanum. More often these are omitted, as at SEMPRINGHAM (40). 
 The arch of the doorway is almost always semicircular till late in the twelfth 
 century. There are no double doorways.* The arch of the doorway is con- 
 structed in recessed orders ; of which at Malmesbury there are eight. More 
 room for orders was got sometimes by thickening the wall in the neighbourhood 
 of the doorway. Norman porches survive, some of two stories ; e.g. at South- 
 well and SHERBORNE (576). Nor are Norman doors lacking, with the original 
 iron work ; e.g. SEMPRINGHAM (40). 
 
 TOWERS. — All the greater churches seem to have had a central tower, 
 except EXETER (377), whose towers were placed at the ends of the transepts. 
 The normal group was one central and two western towers. Sometimes, as at 
 ELY (587), there was but one western tower ; sometimes, as at Tewkesbury, 
 there was none. None of the greater Norman towers seem to have been 
 octagonal ; they were square. The central towers were meant to be lanterns. 
 Not only have they windows, but they have elaborate arcades round the inner 
 wall, intended to be seen from the floor of the church. Sometimes a central 
 tower barely rises above the roofs ; e.g. at Winchester ; more often it rises to a 
 considerable altitude, as at TEWKESBURY (390), St Albans, Norwich, Castor, 
 Sandwich, St Lawrence. Internally, as well as externally, the towers are 
 usually much ornamented with arcading. Probably they were roofed with low 
 square spires. In flint districts the towers of the parish churches were often 
 circular. 
 
 Norman Ornament. 
 
 Of the Romanesque schools of sculpture the most skilful seem to have been 
 those of Toulouse, Provence, Northern Spain, Poitou and Burgundy. The 
 Normans were among the most backward ; and through lack of skill had to 
 
 * Abroad these are very common ; e.g. magnificent double doorways lead from the cloister 
 into the transept of Tarragona Cathedral, and from the narthex into the nave of Vezelay. 
 
ROMANESQUE ORNAMENT. 
 
 39 
 
 confine themselves largely to geometrical work, simple and easy of execution. 
 The decorative stock-in-trade of the Normans in the eleventh century, with 
 which they started us in England after the Conquest, was composed of billet, 
 
 Ely, Western Processional Doorway of Nave. 
 
 square or round; damiers, patterns like a chessboard; stars; imbrications, or 
 shingle ; interfacings ; chevron, or zigzag ; torsades, or cable ; palmettes, honey- 
 suckle, or anthemion ; and rinceaux, or scrolls of foliage.* All the above occur 
 
 * Ruprich- Robert, 124. 
 
4 o 
 
 ROMANESQUE ORNAMENT. 
 
 also in the twelfth century both in Normandy and England, and in much 
 greater profusion. 
 
 The billet is more common in the eleventh, the chevron in the twelfth 
 century ; e.g. the earliest parts of Ely have the billet ; but it also occurs in 
 Canterbury choir in 1 175, in Lincoln south-east transept in 1192. The billet 
 may be square, as at St Augustine's, Canterbury ; or round, as in Binham 
 
 Priory. 
 
 The chevron is used with great profusion in the twelfth century ; e.g. 
 
 in the western doorway and windows 
 of iffley (574); in the window of 
 ST JAMES', BRISTOL (5 16) ; in the 
 pier-arches of WALTFIAM (521) and 
 STEYNING (273) ; in the ribs of 
 DURHAM VAULT (8). In later work 
 of this century it is often studded 
 with " pearls," or otherwise enriched ; 
 it may be inverted ; and in late ex- 
 amples it may be much undercut. 
 The chevron is an almost exact 
 reproduction of devices found on 
 ancient Roman stones ; e.g. on the 
 fine altar recently discovered at Lan- 
 chester * in Durham. Late examples 
 are seen at ST DAVID'S (412.5) ; in the 
 north porch of Wells ; in Glaston- 
 bury Lady Chapel ; highly undercut, 
 with a roll beneath it, in the north 
 transept chapel of Tewkesbury, c. 
 1230; and in the doorway of Stone 
 Church. f It survives in archivolts in 
 Cyprus and Sicily till the fourteenth 
 and fifteenth centuries. For examples 
 see above. 
 
 The saw-tooth ornament is com- 
 mon in early work ; with teeth first of 
 an obtuse, later often of acute angle. 
 The star ornament is found in 
 altar ; it occurs in Ernulph's work at 
 Norfolk ; Herringfleet, Suffolk ; Upton, 
 
 Sempringham. 
 
 Roman work, e.g. on the Lanchester 
 Canterbury ; at Romsey ; Stringham, 
 Gloucester, and elsewhere. 
 
 The nail-head, being easy of execution, was a great favourite ; e.g. Ely. A 
 band of nail-head was often employed in the first half of the thirteenth century 
 in capitals ; e.g. at KETTON (440.2) ; compare the buttress of ST PATRICK, 
 
 DUBLIN (354). 
 
 * Builder, Dec. 28, 1895, 474. 
 
 t Cresy illustrates it in page 6, and gives reasons for believing that it belonged to an earlier 
 doorway. 
 
ROMANESQUE ORNAMENT. 41 
 
 The pellet or " stud "- might be circular; either flat, as at Stoneleigh, or 
 forming a boss, as at Iffley and Crickfont. Or it might be a lozenge, as at 
 Essendine. 
 
 The patera or medallion is a large flat, circular disc, often containing foliage 
 or figures ; e.g. at LLANDAFF (580) and HALES (575). 
 
 The fret or key or embattled ornament is a most ancient decorative form ; 
 common in Arabia, China, South America ; Greek, Roman and Byzantine 
 work.* Good examples occur in the doorways of Middle Rasen and Kirkstall. 
 
 Imbrications or shingle or scale work is also a very ancient motive ; more 
 common in Normandy than England ; it occurs in Westminster Hall and on 
 Castor tower. 
 
 Interlacings are common in England before the Conquest, and after c. 1090 ; 
 but are somewhat rare between 1066 and iocjo.t Good examples occur at Castor, 
 C. II24; IFFLEY ^256; ; NORTHAMPTON ST PETER'S (41 5.6 ) ; SHOBDON 415.3). 
 Canon Taylor held that the Irish scribes imported them into the Continent ; 
 Professor Boyd Dawkins that they originated with the Franks ; being found 
 in great numbers on Germanic sword-hilts, brooches, buckles, &c, as early 
 as the fifth century. But they are common in Byzantine work, especially in the 
 eighth century ; and very common indeed on Roman sarcophagi and especially 
 in the Roman mosaic pavements which were executed all over the Empire. 
 They occur in Armenia, Hungary, Styria, Wallachia, Mycenae, Chalda;a, 
 Assyria, the Canarese or west coast of India ; in fact, wherever the traditions of 
 plaiting basket work decoratively have survived. 
 
 Interlacing snakes occur on an eighth century bas-relief on the wall of the 
 old Cathedral of Athens;! on the jamb slabs of the Anglo-Saxon doorway of 
 Monkwearmouth ; in Norman doorways at Kilpeck ; on a fourteenth-century 
 capital at Oakham ; and elsewhere. 
 
 The bead and roll occurs in the slype of the south transept of St Albans, in 
 the doorway of HALES (575 .:, and in St Leonard's Priory, Stamford, where it 
 produces a curious molding (705.3.;. It is common in Greek and Roman work ; 
 and is probably motived by a child's necklace. § 
 
 A double cone occurs at Stoneleigh, Warwickshire. 
 
 The chain occurs in St William's Chapel, York, and in the vaulting of St 
 Peter in the East, Oxford ; the dedication of which may have been to St Peter 
 in vineitlis. 
 
 The cable is frequent and effective, especially at Southwell over the arches 
 of the crossing. Sometimes it occurs in bases. 
 
 The nebule is used instead of corbels beneath the eaves of SOUTHWELL 
 (390; and Binham. 
 
 Beak-heads and cat-heads are common in the twelfth century ; e.g. in the 
 west doorways of IFFLEY (574) and BARTON-LE-STREET 42J). Wolf-heads 
 occur over the pier arches of Bayeux nave. 
 
 Pearls are very common in Norman leafage and ornament; e.g. in 
 NORTHAMPTON ST PETER (41 5.6). They have been supposed to be reminiscent 
 
 * Barry's Lectures, 101. t J. T. Irvine in Journal of Arch. Assoc, 48, 26. 
 
 I Cattaneo, 77, Fig. 19. Sj See Statham's Architecture for General Readers, 152. 
 
42 ROMANESQUE ORNAMENT. 
 
 of the ornament (dots of ink) in the Irish missals. But they are particularly 
 common in Poitou, Berri, and Burgundy ; and in Scandinavian wood-carving. 
 
 The palmette, honeysuckle, or anthemion, is common m Greek, Roman and 
 Byzantine work ; especially in Corinthian and Composite capitals ; so also in 
 Norman work ; e.g. at TILNEY (4234)- 
 
 Rinceaux or leaf-scrolls were very common in Greek, Roman and 
 Byzantine work. They are much used in Norman work, especially in capitals ; 
 and had much to do with the origin of the conventional stalky leaved capital 
 of early Gothic 
 SHOREHAM (430). 
 
 (429). See BARTON -LE -STREET (427); ELY (430) ) NEW 
 
 Roses were common in 
 the " lacunaria " or Roman 
 ceilings, and in Corinthian 
 capitals. They occur in the 
 south doorway of IFFLEY 
 
 (577)- 
 
 Reminiscences of 
 
 Classical Mythology occur ; 
 e.g. Centaurs ; the Sagit- 
 tarius ; Sirens ; Mermaids. 
 The stock illustrations of 
 animals are taken from the 
 Bestiaires.* The signs of 
 the zodiac, the works of the 
 months, the virtues and 
 vices all find representation, 
 first in Romanesque, and 
 afterwards in Gothic sculp- 
 ture. 
 
 Duration of the 
 Romanesque Style. — We 
 saw that the first building 
 in the Anglo-Norman 
 variety of Romanesque was 
 the abbey church of West- 
 minster, commenced in 1050. 
 It does not follow that all the 
 world set to work immediately to build to Anglo-Norman design. There 
 have always been Radicals and Conservatives in architecture, as in politics. 
 For another generation or two, well into the twelfth century, we may be 
 sure that many people went on building in Anglo-Saxon fashion. Similarly, 
 when all the world had adopted the Anglo-Norman style, they would not 
 give it up simultaneously. The greater churches would be the first to 
 abandon it for Gothic : but even among these the progress was far from 
 being at a uniform rate. The naves of St David's and Wells were building 
 
 * On Ecclesiastical Zoology see Evans' Animal Symbolism in Ecclesiastical Architecture : 
 and the article on " Physiologus" in the Encyclopaedia Britannica. 
 
 Sutton St Mary. 
 
DURATION OF ENGLISH ROMANESQUE. 43 
 
 simultaneously c. 1190; the nave of St David's is almost as Romanesque 
 as St Botolph's, Colchester, founded in 1102; while Wells nave is in many 
 respects as Gothic as the choir of Lincoln Minster. At no time and in no 
 style was the progress uniform in different parts of the country; e.g. the choir 
 of St Bartholomew's, Smithfield, commenced in 1123, is not so advanced as the 
 Norwich choir of 1096 or the Durham choir of 1093. Still slower would the 
 rate of progress be in the villages ; a fact which has always to be borne in 
 mind in estimating the date of a village church. In village churches rude 
 and archaic work is not necessarily a proof of an early date. If we judged by 
 the rude and archaic exterior and interior of TOWYN CHURCH (458; we should 
 unhesitatingly assign it to the eleventh century ; but it might well be that the 
 new current of Romanesque did not strike the remote coast of Merionethshire 
 till well into the twelfth century. SUTTON ST MARY (42) was not begun 
 till after 1180. On the whole, we may conclude that Romanesque work was 
 still being done in the smaller churches, here and there, till the end of the 
 twelfth century. In the greater churches we may take it that Gothic architec- 
 ture came into being not later than c. 1175, with the commencement of the 
 choir of Wells Cathedral by Bishop Reginald de Bohun ; that of Canterbury 
 under the direction of William of Sens ; and those of Roche, Byland, RIPON 
 (102), and York. In France the choir of St Denis was commenced in 1140; 
 that of Notre Dame, Paris, in 1 163. 
 
 By Mr Sharpe the work done c. 1 145 to c. 1190 has been designated 
 TRANSITIONAL ; by Mr Brandon SEMI-NORMAN. But in the first half of it the 
 presence of pointed arches, e.g. at Fountains and Kirkstall, is not a sufficient 
 ground for admitting them to the fellowship of Gothic ; they are churches in 
 which much more reliance is placed on thickness of wall than on projection of 
 buttress. Nor on the other hand, because of the retention here and there of the 
 semicircular arch, are well-buttressed buildings, such as Canterbury choir, to 
 be denied the name of Gothic. 
 
Chapter III. 
 
 CHARACTERISTICS OF ENGLISH GOTHIC ARCHITECTURE. 
 
 Monastic v. Secular Gothic — Admixtures of Romanesque — Procedure in Rebuilding — 
 Length, Span, Height, Area of English Churches— Proportions — Abutment — 
 Skeleton Construction — Economy of Material— Lightness of Construction — Im- 
 portance of Stained Glass — Reasons for Height of Gothic Churches — The Vertical 
 and Horizontal Line. 
 
 In the Anglo-Norman architecture of the eleventh and twelfth centuries the first 
 landmark is Edward the Confessor's Church at Westminster. The second is the 
 commencement of the building of Cistercian churches c. 1 140 ; in what has been 
 called the Transitional, Semi-Norman, or Pointed Norman style. More than one 
 hundred Cistercian abbeys were founded between 1 1 25 and the end of the century. 
 Of the Cistercian churches remaining the oldest appear to be Fountains, Kirkstall, 
 and Furness. Only a quarter of a century separated these from the Gothic 
 architecture of Canterbury, Wells, Roche, Byland, and Ripon. 
 
 Up to c. 1175 the lion's share of the work had been done by the Monks and 
 the Canons Regular. In the eleventh and twelfth centuries the monastic orders 
 were the progressives and the reformers in the Church. Much energy had 
 been shown even before the Norman Conquest, e.g. by Dunstan, in expelling 
 Secular Canons from their churches, and in replacing them by monks. But 
 gradually the Secular Canons reformed themselves, and regained their influence ; 
 the proof of which is to be seen in the great amount of Gothic architecture to be 
 put to their credit. If we take as a test the cases where whole Romanesque 
 churches were pulled down and rebuilt, not under stress of fire or storm or 
 because of collapse of masonry, we shall find that the Secular Canons were much 
 the more thoroughgoing in Gothic building. To them is to be credited the rebuild- 
 ing of the cathedrals at Wells ; Lincoln;* Salisbury; Lichfield; Exeter; York; 
 Ripon and BEVERLEY (176) Minsters; and Howden. On the other hand, the 
 Benedictine monks rebuilt Whitby, Westminster, St Mary's, York, and Bath ; 
 the latter not till the sixteenth century. The Cistercian abbeys were but 
 partially rebuilt, some not at all. The Augustinian Canons f rebuilt St Saviour's, 
 Southwark. 
 
 I 'onions of the Norman west front were incorporated at Lincoln ; at Exeter the transeptal 
 towers were retained ; York retains a Romanesque crypt ; Ripon allowed some Romanesque 
 work to remain south of the choir 
 
 t The Canons Regular, of whom the Augustinians were the most important order in 
 England, lived a common life in a cloister under a Rule (regula) ; and differed little from the 
 monks, except that none of them were laymen, and that they were attached to a cathedral or 
 other collegiate church. 
 
ADMIXTURE OF ROMANESQUE. 
 
 45 
 
 The comparison is largely in favour of the Secular Canons, If it is true that 
 we owe the majority of our Romanesque churches to the Regular orders, it is 
 equally true that the Secular Canons took the leading part in the development 
 of English Gothic. 
 
 The number of English churches of the first rank built or rebuilt wholly 
 in Gothic is not great. What is rare here is quite common in France. The 
 number of cathedral, abbey 
 and collegiate churches with- 
 out admixture of Romanesque 
 in the Domaine Royale and 
 Champagne is very large.* 
 
 In some of our greater 
 churches the melange of styles 
 is something extraordinary. 
 At Hereford, Chichester, St 
 Albans, Wimborne, every 
 variety and subvariety of our 
 mediaeval architecture may 
 be seen in juxtaposition in 
 one building. As a rule, an 
 English cathedral is not a 
 study in harmonies, but in 
 contrasts. Most often it is a 
 contrast of a Romanesque 
 nave and a Gothic choir ; as in 
 Ely and Hereford cathedrals; 
 sometimes one transept is 
 Romanesque, theotherGothic ; 
 as in Hereford and Chester 
 Cathedrals ; or a Romanesque 
 transept contrasts with a 
 Gothic choir and nave, as in 
 Winchester Cathedral ; or a 
 Romanesque choir has a 
 Gothic retrochoir, as at Peter- 
 borough, Durham and Chichester ; or the eastern bays of the 
 esque; the western Gothic, as at Romsey; Gloucester; Shrewsbury; or the reverse 
 
 * The following list, necessarily imperfect, will give some idea of the extent to which 
 Romanesque was retained in our more important churches : — 
 
 A. Binham, Myth, Bolton, Boxgrove, Bury, Canterbury C. Canterbury St Augustine, 
 Carlisle C, Castle Acre, Chepstow, Chester C, Chichester C, Christ Church, Hants, Colchester 
 St Botolph, Dorchester, Durham C, Ely C, Gloucester C, Hereford C, Leominster, Lindis- 
 farne, London Old St Paul's and St Bartholomew's, Mailing, Malvern. New Shoreham, Norwich 
 C, Pershore, Peterborough, Ramsey, Rochester, Romsey, St Albans, Selby, Shrewsbury, South- 
 well, Tewkesbury, Thorney, Tutbury, Tynemouth, Waltham, Wimborne, Winchester C, 
 Worcester C, Wymondham ; the above contain work earlier than [ 150 
 
 //. Buildwas, Cartmel, Chester St John's, Dore, Dunstable, Fountains, [•'unless, Kirkstall, 
 Temple Church, London, Malmesbury, Oxford St Krideswide's. Wimborne, Winchester St Cross; 
 the above contain work c. 1 150 to c. 1200. 
 
 Elv Lantern. 
 
 ,e are R< 
 
4 6 ADMIXTURE OF ROMANESQUE. 
 
 is the case, as at Rochester. Sometimes early is in juxtaposition to late Gothic; as 
 at WELLS (127), Lincoln, Lichfield, Canterbury, York. Sometimes the substruc- 
 ture is Romanesque, the superstructure Gothic ; as in Selby nave ; St John's, Chester ; 
 and in the naves of Rochester and Malmesbury ; OXFORD CHOIR (27). Some- 
 times the church was poor ; and do all it could, the work went on very slowly ; in 
 the naves of Selby and Binham there is a difference of date and a difference of 
 style almost in every bay. More heterogeneous churches and more picturesque 
 churches cannot be imagined ; as delightful to the artist as to the archaeologist. 
 
 What has been said of the greater is largely true of the smaller churches 
 also. As a rule, an English parish church was not pulled down and rebuilt de 
 novo ; the old church frequently remains inside* forming the nucleus round which 
 all the later additions have crystallised ; e.g. at St Mary's, Guildford ; where all 
 that is left of the original building is the central tower. The chief exception is 
 that in districts where the farmers were making large profits from their wool, 
 and the weavers and merchants from their woollen cloth, e.g. Norfolk, Suffolk, 
 and Somerset, frequently the churches were wholly rebuilt ; the chancel often in 
 the fourteenth, and the nave in the fifteenth century ; leaving no trace of the 
 original church. 
 
 Romanesque largely survived in England, while in the Domaine Royale 
 and Champagne most of it disappeared. The output of Norman building here 
 in the eleventh and twelfth centuries had been enormous ; and at the end of the 
 latter century it must still have been in good repair. The very number and 
 grandeur of our Romanesque churches may have saved them from being 
 promptly rebuilt in Gothic. 
 
 The substitution of Gothic for Romanesque was a long and slow process in 
 most of the greater English churches. Some, like Selby, Chester St John's, 
 Binham, Romsey, at the end of the twelfth century, had Romanesque naves still 
 incomplete ; and finished them in Gothic. More often, however, the nave was 
 complete ; and the new Gothic was first employed at the east end of the church. 
 At Norwich nothing was done but to substitute a rectangular Lady Chapel for the 
 eastern apsidal chapel. At Chichester, Ely, Durham, St David's, and Here- 
 ford, the eastern limb was prolonged or extended. In very many cases a clean 
 sweep was made of all work east of the crossing ; so that Romanesque choirs 
 are now rare with us ; e.g. Winchester, Worcester, Southwell, Boxgrove, Foun- 
 tains, Pershore, Carlisle, in the thirteenth century ; Selby, mainly in the four- 
 teenth ; Malvern and Christ Church, Hants, in the fifteenth and sixteenth 
 centuries. Not only a new choir, but a new central transept also was built at 
 Hexham, Rochester, Rievaulx, in the thirteenth, and at Bristol in the fourteenth 
 century. At St Albans in the thirteenth ; at Shrewsbury and Waltham Abbeys 
 in the fourteenth ; at Gloucester in the fifteenth century, a beginning was made 
 of rebuilding the Norman naves from the west end. At Rochester a com- 
 mencement was made at the east end. 
 
 A less drastic method was adopted at Gloucester and Tewkesbury ; an 
 
 example copied at Winchester and Norwich in the fourteenth, at Sherborne and 
 
 Malvern in the fifteenth century. It was not to pull down the old Norman work, 
 
 but merely to put a new face on it ; to give it a Gothic veneer. At TEWKESBURY 
 
 * Last to disappear generally were the responds of the chancel arch. 
 
47 
 
 C.lourestcr Choir, N.E. Angle. 
 
DIMENSIONS. 49 
 
 (165) the recasting was of a more drastic character than at GLOUCESTER 
 (135). At Tewkesbury the ambulatory plan is still there, and the Norman cylin- 
 drical piers, somewhat heightened ; but the Norman triforium and clerestory 
 were removed. At Gloucester the clerestory was removed ; but the vaulted 
 upper aisle was left, to give abutment to the new licrne vault of the choir. 
 At WINCHESTER (90, 342) and Sherborne Abbey, piers, arches and thick clere- 
 story wall were all left, but transformed into Gothic guise. The piers in Win- 
 chester nave are the original Norman ones, with the moldings* modified ; while 
 the vaulting shafts are the Norman roof shafts unaltered. 
 
 Sometimes the rebuilding was continued till the whole church became 
 Gothic. In some few cases the works were carried on with considerable rapidity ; 
 and in these the result is a uniformity and regularity of style in which a French- 
 man sees nothing remarkable, but which at once strikes one of ourselves as 
 something exceptional. Lincoln (as it was c. 1250), SALISBURY (170), St 
 Saviour's Southwark, and Exeter, were each built in about half a century. Other 
 churches, built wholly in Gothic, but Gothic extending over long periods of 
 time, are Canterbury Cathedral, Lichfield, Beverley Minster, and Westminster. 
 In the last two the later is assimilated to the early work, so that in these two 
 there is remarkable unity and uniformity of design. 
 
 In several cases, when the original Anglo-Norman cathedral had been 
 wholly rebuilt in Gothic, another period of Gothic building set in later, by way 
 of extension of the eastern limb ; in the last half of the thirteenth century at 
 LINCOLN (151.1) and Old St Paul's ; in the first, half of the fourteenth century 
 at Wells, Lichfield, Glastonbury, Carlisle. 
 
 Still more extensive was the Gothic work done at York. Here there was 
 first, a Norman cathedral. Then the choir was rebuilt, 1154-1181 ; the transepts 
 1247-1260; the nave 1291-1345; making the whole cathedral Gothic. Then 
 once more the works recommenced, the choir of 1 154-1 181 was pulled down and 
 the present presbytery and choir were built, 1367-c. 1400 ; and the three towers 
 c. 1 400- 1 474. 
 
 As we have seen above, almost always the short Norman choirs were either 
 rebuilt or lengthened ; and the Gothic choirs themselves were sometimes 
 lengthened a second time. The result of this was that the greater Gothic 
 churches are remarkable for the great length of their eastern limb ; differing in 
 this respect completely from their Norman predecessors, where the excess of 
 length is to be found in the nave; e.g. at ST ALBANS (153.2), Winchester,! ELY 
 (153.4), Peterborough, NORWICH (148.4). 
 
 In total length we can show churches, with their long Romanesque naves 
 and long Gothic choirs, surpassing the largest mediaeval churches of Europe. 
 
 Eeet. Ecet. 
 
 Old St Paul's - 586 Milan 475 
 
 Winchester 530 Florence - 475 
 
 St Albans 520 Amiens 435 
 
 Ely - 517 Rouen C. 435 
 
 Canterbury 514 Reims - 430 
 
 Westminster - 505 Cologne - 427 
 
 The diagram on page 659 shows the Norman pier as remodelled, 
 t Winchester nave was longer still before its remodelling by Wykeham and his successors. 
 
 D 
 
So 
 
 DIMENSIONS. 
 
 n the spans of their 
 
 naves they 
 Feet. 
 
 are surpassed by many. 
 
 King's C. C. - 
 
 York 
 
 Ripon 
 
 • 45* 
 
 - 45 
 
 - 40 
 
 Gerona 
 
 Toulouse - 
 Perpignan - 
 
 Boston 
 Ely- 
 
 - 40 
 
 - 39 
 
 Albi - 
 Milan 
 
 Lincoln - 
 Canterbury choir 
 Glastonbury 
 Old St Paul's - 
 
 - 39 
 • 39 
 
 - 38 
 36 
 
 Seville 
 Florence - 
 Reims 
 Amiens 
 
 Feet. 
 " 73 
 63 
 
 - 60 
 
 58 
 
 - 56 
 
 - 56 
 
 55 
 
 - 43 
 46 
 
 In internal height* they fall far short of their Continental brethren: some 
 being exceptionally low : ^.Lichfield, 57 feet; Chichester, 61 feet; Beverley, 
 Wells, Gloucester, 67 feet ; Worcester, 68 feet ; Exeter, 69 feet. 
 
 Feet. 
 
 Old St Paul's nave 
 Westminster 
 York choir 
 Gloucester choir 
 Salisbury - 
 Lincoln nave - 
 Peterborough - 
 Canterbury nave 
 Winchester 
 
 Feet. 
 
 i°3 
 
 i°3 
 
 102 
 
 S6 
 
 84 
 
 82 
 
 81 
 
 80 
 
 7S 
 
 Cologne 
 Beauvais 
 Bologna ■ 
 Amiens ■ 
 Bourges ■ 
 Chartres • 
 Strasburg ■ 
 Toledo 
 Leon 
 
 In area also they have many superiors on the Continent. 
 
 Old St Paul's 
 York - 
 Lincoln 
 Bury - 
 Winchester - 
 Glastonburv 
 Ely 
 
 Westminster 
 Durham t - 
 Salisbury 
 Canterbury - 
 Peterborough 
 
 Sq. feet. 
 72,460 
 63,800 
 57,200 
 56,270 
 S3.43o 
 48,000 
 46,000 
 46,000 
 44,400 
 430*5 
 43, 2I 5 
 41,090 
 
 Seville 
 
 Milan 
 
 Saragossa 
 
 Amiens 
 
 Cluny 
 
 Toledo 
 
 Cologne 
 
 Florence 
 
 Bologna 
 
 Chartres 
 
 Reims 
 
 Bourges 
 
 !5° 
 
 !5° 
 
 144 
 "7 
 106 
 ior 
 100 
 100 
 
 Sq. feet. 
 
 150,000 
 92,600 
 80,000 
 70,000 
 66,000 
 66,000 
 65,800 
 
 65,7 00 
 65,000 
 65,000 
 65,000 
 59,000 
 
 Much has been written on the subject of the proportions of the Gothic 
 churches here and abroad ; e.g. the assumption being that the interiors were 
 proportioned according to the ratio of the sides of equilateral or of isosceles 
 triangles, as the case might be. No two of these theories agree ; nor are they 
 based on uniform systems of measurement.^ In this, as in all matters, practical 
 considerations may fairly be assumed to have come first with the builders. The 
 
 * These measurements give the height up to the apex of the vault. The lengths, breadths, 
 and heights given in the above tables are internal measurements. 
 
 + Durham occupies nearly an acre. The boundary line of Salisbury, following the angles 
 made by the buttresses and other projecting parts, is nearly half a mile. 
 
 Moreover, the new work was frequently erected on the old foundations, e.g. the nave and 
 central transept of Canterbury. " It is vain to look, as many have done, for any general doctrines- 
 of proportion in work so conducted" (Lethaby, Med. Art, 169). 
 
Si 
 
 Beverley Choir from S.E. 
 
PROPORTIONS. 
 
 53 
 
 span of the nave could not be expanded at will ; it was confined within certain 
 limits by the difficulties and cost attending roofs of exceptional span. Again, in 
 determining the height of the nave, the first thing to take into account was the 
 amount of light desired ; this regulated the height of the aisle window and of 
 the clerestory window ; consequently, of the pier arcade and the clerestory wall. 
 Again, there was the question of borrowing light from the triforium. chamber; if 
 that was desired, the height assigned to the triforium had to be considerable. 
 On the other hand, if no light was desired from this source, the height of the 
 triforium could be greatly diminished. As for the length of each limb, that again 
 could not be determined by geometrical ratios. Its length depended mainly on 
 considerations of ritual ; on the number of monks or canons attached to the 
 church ; and on the number of altared chapels desired. It often happened that 
 the length of a church was curtailed by some obstacle ; by a highroad or a foot- 
 path, or the city wall. Thus the east end of the presbytery of OXFORD 
 CATHEDRAL ( I 52.3) extended up to the city wall ; and there was no room to the 
 east for a Lady Chapel ; it was therefore placed to the south. The Lady Chapel 
 of Gloucester and the chancel of Walpole St Peter's were built oxer rights-of- 
 way ; in these two instances curtailment was avoided by building a vaulted 
 subway. This may have been the case at Hythe also. But, of course, the most 
 weighty factor was the amount of money at the disposal of the monks or canons. 
 Given funds and spaciousness of site, the number of bays in a nave, choir or 
 transept could be multiplied till, as at BURY (150.3), there was a nave of 296 
 feet, or, as at Old St Paul's, a transept of 293 feet and a choir of 224 feet. In 
 England, at any rate, the ratio of height to span varies so greatly, that certainly 
 it cannot be predicated of the builders that they had any abstract scheme of 
 ratios in their heads. The following table shows the height and span of some 
 of the more important vaulted churches*— 
 
 
 Span. 
 
 Height. 
 
 Kali. 
 
 Tewkesbury nave 
 
 33 
 
 5* 
 
 1.8 
 
 Gloucester nave 
 
 34 
 
 68 
 
 2 
 
 Exeter nave - 
 
 34 
 
 69 
 
 2 
 
 Lichfield nave 
 
 - 28 
 
 57 
 
 2 
 
 Wells nave 
 
 3? 
 
 67 
 
 2. 1 
 
 Lincoln nave - 
 
 39 
 
 82 
 
 2. 1 
 
 Winchester nave 
 
 3 2 
 
 78 
 
 2.4 
 
 Gloucester choir 
 
 - 33 
 
 86 
 
 2.6 
 
 Beverley choir 
 
 - 26 
 
 67 
 
 2.6 
 
 Salisbury nave 
 
 3 2 
 
 84 
 
 2-7 
 
 Norwich choir 
 
 - 28 
 
 §3 
 
 2.9 
 
 Westminster nave 
 
 35 
 
 i°3 
 
 2.9 
 
 Noyon 
 
 I .aon 
 
 Chartres 46 
 
 Bouryrs 4'i 
 
 St Sernin, Toulouse 
 
 (Romanesque) 
 Toledo 38 
 
 Amiens 46 
 
 Leon 31 
 
 Beauvais - 45 
 
 Conques( Romanesque) — 
 Cologne - 41 
 
 St Trophime, Aries 
 
 (Romanesque) 
 
 Span. Height. Ratio 
 
 106 
 I 17 
 
 IOO 
 
 144 
 IOO 
 
 '5° 
 '55 
 
 2 07 
 
 2-5 
 
 2 -59 
 2.6 
 
 3 45 
 3-8 
 
 M 
 
 On nothing does the effectiveness of an interior depend so much as on the 
 ratio of height to span. In the naves of GLOUCESTER (26), TEWKESBURY 
 '297 , EXETER (9), and Lincoln the vault is crushingly low. There can be 
 no question that the most successful vaulted interiors we possess are those 
 
 * The dimensions given in this and the preceding tables must be accepted as only approxi- 
 mate : of many churches the measurements are not trustworthy. 
 
54 PROPORTIONS. 
 
 of the naves of Westminster, Salisbury, Beverley, and Winchester. Where 
 height and span are also properly correlated with length, as in the naves of 
 WINCHESTER (342) and WESTMINSTER (63), there an English interior is seen 
 at its very best. 
 
 But there is yet another factor which has very great weight. What it is, 
 may be seen by examining the naves of LICHFIELD (523) and Wells. They 
 are quite sufficiently long ; but their height is only about twice their span. Yet 
 they do not look low ; as do the naves of EXETER (9) and Lincoln, which 
 also are only about half as broad as they are high. The reason for this is that 
 we have taken into account only the breadth of the nave. But the breadth of 
 each of the bays of which the nave is composed is also an important factor. 
 The following table shows the ratio of the breadth to the height of the bay in a 
 few examples. It will be seen how great is the difference of bay proportion in 
 such interiors as those of Westminster * and Exeter. 
 
 
 Breadth 
 
 Height 
 
 Rati( 
 
 
 of Bay. 
 
 of Bay. 
 
 Exeter nave - 
 
 20 
 
 63 
 
 3 1 
 
 Lichfield nave 
 
 - 16 J 
 
 57 
 
 3-5 
 
 Lincoln presbytery 
 
 21 
 
 74 
 
 3-5 
 
 Wells nave - 
 
 16 
 
 68 
 
 4 
 
 Westminster choir - 
 
 18 
 
 100 
 
 5-5 
 
 Yet another factor is treatment of the vaulting shaft. Where it rises from 
 the pavement, as at Lichfield, the apparent height of the bay is enhanced ; but 
 the church looks lower where as at Exeter and Lincoln presbytery it starts 
 from a corbel at some intermediate point. 
 
 Of all our interiors, perhaps that of WINCHESTER NAVE (342) is most 
 successful. If we take 2 as the unit, then if the breadth of each of its twelve 
 bays is 2 ; the span of the nave is 2§ ; the height of the vault and of each bay 
 is 6^ ; the length of the nave is 225. It is to be noted that it retains massive 
 Norman vaulting shafts descending to the pavement. 
 
 So much for the dimensions and proportions of the greater churches. 
 Another factor of enormous importance is the character of the methods of 
 abutment employed. 
 
 Abutment. — Of systems of abutment to Gothic clerestories we may dis- 
 tinguish four. The first is that which was first employed in DURHAM NAVE 
 (34.2), and which was contemplated at NORWICH (371). In this the clere- 
 story walls are abutted low down by flying buttresses concealed beneath the 
 aisle roof. The second is seen in CANTERBURY CHOIR (34.3), commenced 
 1175. Here there is retained the arch spanning the triforium chamber, which 
 was employed in DURHAM CHOIR (370) ; except that it is segmental instead 
 of semicircular. But in addition a flying buttress emerges, for the first time, into 
 the open air. It is constructed in very timid fashion, just crawling along above 
 the triforium roof; unornamented ; regarded, plainly, as nothing more than a 
 builder's expedient. In LINCOLN CHOIR (34.5), commenced 1192, precisely 
 the same system is adopted as in Canterbury choir ; except that the arch in the 
 
 * Westminster choir and Lincoln presbytery, illustrated on pages 55 and 56, 
 may be taken as average specimens of a French and an English internal elevation. 
 
GOTHIC ABUTMENT. 
 
 55 
 
 triforium chamber is pointed. The third 
 system is seen in CHICHESTER NAVE 
 (34.4), which was vaulted in the last years 
 of the twelfth century. Here also, as at 
 Canterbury, flying buttresses are displayed 
 in the open air ; but they are heavy and 
 clumsy. Plainly they are no copies of 
 Canterbury work, but just the flying but- 
 tresses of Durham nave built out of doors. 
 Similar flying buttresses, equally massive 
 and plain, are seen at NEW SHOREHAM 
 and BOXGROVE (373). But down below, 
 in the section on the right, page 34.4, 
 may be seen a second flying buttress, help- 
 ing to support the aisle roof. Here then 
 we have a double set of flying buttresses, 
 one above, the other beneath the triforium 
 roof. The fourth system is that in which 
 all abutment inside the triforium chamber 
 is discarded, and in which, as at EXETER 
 (3S-3)j tne fly' n g buttress is displayed in 
 the open air. The fifth appears at WEST- 
 MINSTER (35.2) c. 1245; and earlier still 
 in Ely presbytery c. 1234. In both these 
 churches the thrusts of the high vaults are 
 stopped by two flying buttresses in super- 
 position, both of them above the aisle 
 roofs. In the Gothic architecture of 
 England two of the five systems remained 
 in employment, viz., the first and the 
 fourth ; with an ever-increasing tendency 
 to employ the fourth. In France, in the 
 Gothic of the Domaine Royale, the fourth 
 and fifth systems chief!)' were employed. 
 Owing to the vastly greater height of their 
 clerestories, the first three systems would 
 have been ineffectual. 
 
 Skeleton Construction. — From 
 the character of the Gothic vault and from 
 the employment of the buttress there flowed 
 consequences which entirely transformed 
 the face of Gothic architecture. Owing to 
 the fact that in a Gothic vault the ribs only 
 descend to the wall opposite the piers, it 
 follows that, while the parts of the wall to 
 which they do descend are exposed to an 
 enormous bursting pressure, the whole of 
 
 Westminster Ch( 
 
56 
 
 SKELETON CONSTRUCTION. 
 
 Lincoln Presbytery. 
 
 the space between the springs 
 of the ribs — i.e. nearly the 
 whole bay — is free from any 
 such pressure. It follows that 
 if the builder chooses to omit 
 the wall space between each 
 pair of buttresses, he can do so, 
 provided that he builds a re- 
 lieving arch across from buttress 
 to buttress to carry the parapet 
 and roof. And where the wall 
 was, he can have glass. To a 
 large extent, therefore, Gothic 
 architecture meant the substi- 
 tution of voids for solids and 
 window for wall. The differ- 
 ence between the Romanesque 
 and the Gothic construction 
 may be seen by comparing ELY 
 NAVE (57), LINCOLN PRESBY- 
 TERY (56), and WESTMINSTER 
 NAVE (55). 
 
 At Ely the distance from 
 window to window in the clere- 
 story is about 1 3 feet ; and the 
 whole breadth of each bay is 
 solid wall, except a window 4 
 feet across. In Lincoln presby- 
 tery the clerestory window 
 occupies 1 2 feet out of a total 
 breadth of 23^ feet ; leaving 
 1 1 1 feet of solid wall ; the voids 
 and solids nearly balancing. 
 But at Westminster the clere- 
 story window occupies as much 
 as 10 feet in a bay of a total 
 breadth of 18 feet; leaving 8 
 feet of solid wall ; so that the 
 voids outbalance the solids. 
 Ely may be taken as an average 
 specimen of late Romanesque 
 construction ; Lincoln presby- 
 tery of English Gothic ; West- 
 minster approaches the con- 
 struction of the He de France. 
 The French churches go far 
 beyond Westminster in the 
 
SKELETON CONSTRUCTION 
 
 attenuation of the clerestory wall. In 
 .Amiens nave the windows of the clerestory 
 are more than three times as broad as the 
 strips of clerestory wall ; which are also 
 narrower than the piers down below between 
 nave and aisles. In the nave of St Denis* 
 (1231 to 1280) the piers below are still 
 broader than the strips of clerestory wall 
 between the windows. While in Metz 
 Cathedral * the piers between the nave and 
 aisles are nearly twice as broad as those be- 
 tween the clerestory windows. Vast is the 
 difference between such construction and 
 that of Lincoln presbytery. 
 
 In such churches as Amiens, St Denis, 
 EVREUX (539), the clerestory wail ceases 
 to exist qua wall. + Really it has become the 
 upper part of a pier : of one of the piers 
 below between nave and aisles. In such 
 examples the piers of the ground story do 
 not stop, as they appear to do, at their 
 capitals : each continues up, between the pier 
 arches, between the bays of the triforium 
 arcade, and between the bays of the clere- 
 story, till it stops about one-third of the 
 distance up the clerestory windows, as at 
 Amiens and Metz; or half-way up, as at 
 Beauvais and St Denis. Such a pier, which 
 may be called the Vault pier, is at Beauvais 
 nearly 140 feet high. How is it kept in 
 position ? The lower part of it, if it be a 
 pier between the nave and aisle, is kept from 
 moving to east or west by the arches which 
 it supports. It cannot incline backward , be- 
 cause of the inward thrust of the vault of 
 the aisle. Nor again can it incline forward, 
 for it is weighted with its own upper portion, 
 which is loaded with its share of vault and 
 outer roof. In the triforium stage the arches 
 of the triforium arcade act as straining 
 arches. To oppose any movement forward or 
 backward there is opposed the weight of its 
 upper portion carrying its share of vault and 
 outer roof. In the clerestory stage, it is pre- 
 
 * Elevations in Dehio, Plates 387, 388. 
 + See especially the longitudinal section of Glou- - 
 cester choir on page 59. 
 
 Ely Nave. 
 
5 8 SKELETON CONSTRUCTION. 
 
 vented from moving to east or west by the arches of the windows, which act as 
 straining arches. It cannot move forward because of the thrust of the high vault; it 
 cannot move backward because of the flying buttress * which acts as a stay,propped 
 up on the top of the aisle buttress; which buttress is loaded with a pinnacle. 
 All this complex mechanism is needed to keep such tall vault piers upright. 
 
 In England, as we have seen, even in the semi-French church of West- 
 minster, usually we did not carry Gothic construction to such logical extremes : 
 eliminating masonry till there remained nothing but a vault pier. It was not 
 that we could not, but that we would not. Even in the thirteenth century the 
 principle of the vault pier was thoroughly understood and properly applied in 
 England. The construction of the Chapter House of Salisbury is precisely the 
 same as that of the clerestories of Amiens, Beauvais, St Denis, Metz. In all five 
 the wall between the windows is reduced to a pier; and the wall ribs of the 
 vault serve also as the arches of the window. In GLOUCESTER CHOIR (59), 
 finished c. 1350, a magnificent pier ascends uninterruptedly from the pavement 
 to the spring of the arches of the clerestory window ; a construction which was 
 repeated, but with more timidity, in Henry the Seventh's Chapel, Westminster, 
 Malvern, and Bath. But what was optional with us was a constructional 
 necessity with the French builders. Even if they had wished, they could not 
 have constructed their lofty churches in our English fashion, with retention of 
 great breadths of clerestory wall. Look at a typical English Gothic elevation, 
 such as that of LINCOLN PRESBYTERY (56). On a pier which is about 5^ 
 feet broad is balanced a mass of clerestory wall, which is no less than \i\ feet 
 broad. Such a pier is top heavy ; the upper part is twice as broad as the lower. 
 In the lofty French churches, to have poised such an enormous weight on the 
 slender piers of the ground story, would have crushed them. Consequently the 
 upper part of the vault pier, as we have seen, had to be made narrower, not 
 broader, than the lower. 
 
 Other considerations no doubt had weight. The generative principle of 
 Gothic architecture has been described, with considerable truth, as the economy 
 of stone.f Labour was cheap, stone was dear. Stone was something precious ; 
 more like ivory than wood. Every care must be used to lessen the cube of stone. 
 Any amount of labour might be expended on ornament ; as little as possible on 
 ashlar. The masons had grown up under this tradition. There was a premium 
 on economy of ashlar. Nowhere is the result plainer than in the construction 
 of the Gothic vault pier. It was an enormous saving in stone. 
 
 Such construction, of course, revolutionised Romanesque practice ; which 
 had been to rely wholly on walls for the stability of the vault. Now reliance was 
 almost wholly on the pier with its paraphernalia of buttresses, flying buttresses, 
 pinnacles. In the nave of a Gothic church in its final development all the 
 windows might be taken away; also the end walls, the walls beneath the 
 
 * In Gloucester choir instead of flying buttresses there is a half barrel. 
 
 t "The most lavish expenditure of labour seems to have been considered no waste, if 
 effecting the slightest saving of material" (Garbett's Principles of Design, 219). "II fallait 
 se suffire avec peu de materiaux ; il fallait traiter la pierre com me une chose precieuse ; tous 
 les efforts devaient tendre a en limiter l'emploi ; on devait batir avec le moins de matiere." 
 (Choisy's Histoire, ii. 526.) 
 
59 
 
 SCALE OF FEET 
 
 G'oucester Choir. 
 
60 SKELETON CONSTRUCTION. 
 
 windows of the aisles and the clerestory, and the spandrils of the pier arcade : 
 it might be reduced to a mere skeleton, consisting of four rows of stone posts— 
 the inner two being the vault piers, the outer two the buttressed piers between the 
 aisle windows, connected by arches— and on these posts, with the winds of heaven 
 blowing through them, the vaults both of nave and aisles would still stand secure. 
 Like the half-timbered house, the Crystal Palace, or the American "sky-scraper," 
 the constructional members are totally independent of the filling in. 
 
 With skeleton construction, moreover, another advance was made to the 
 more complete lighting of the mediaeval churches. Every window, as in the clere- 
 stories of Amiens nave and Gloucester choir, could be widened till it occupied all 
 the whole space from one vault pier to the next. This was no small gain. 
 
 A church so constructed, with the voids so much in excess of the solids, 
 was very light in appearance. Its lightness of construction was still further 
 increased by the superiority of the masonry as compared with that of Roman- 
 esque. The walls could be, and were, made thin.* The piers themselves became 
 surprisingly slender in comparison with their Romanesque predecessors (659, 
 661). All this attenuation of the supports was again facilitated by the lighten- 
 ing of the later vaults ; for the web of these vaults was much thinner ; a shell 
 of ashlar being employed instead of heavy rubble ; nor was it covered with 
 a layer of concrete (304). The result was a wonderful church. A church 
 built logically with vault-pier construction presented an interior such as the 
 world had never seen or dreamt of. It was an " aerial immateriality " ; some- 
 thing spiritual, incorporeal. In such an interior it all but seems that the load 
 might float away from the unsubstantial air or rather from the belt of coloured 
 light on which it rests. In a Romanesque minster like DURHAM (8) one 
 is impressed by the vast downward pressures that exist. Not so in the ethereal- 
 ised later Gothic. " Who, while viewing a stately tree in the pride of its growth, 
 ever thinks of its weight, or of the pressure of its boughs upon the stem ? It is 
 with its upward soaring that the mind is impressed ; and just so it is with the 
 interior of the Gothic cathedral. The perfection with which all the physical 
 forces are met has to the mind the effect, not merely that they are annihilated, 
 but that they are actually reversed!' f 
 
 Nevertheless such construction may be deemed perhaps somewhat non- 
 architectural : a little out of consonance with the material employed ; masonry 
 being made almost as pliant and ductile in design as if it were metal. The great 
 Gothic churches are of stable construction — have they not stood for hundreds of 
 years ? — but however much the intellect appreciates the unseen balance of forces 
 by which their stability is assured, the eye desiderates something more ; solidity 
 as well as stability : and this in its later phases the Gothic preponderance of 
 voids fails to give. " In works of a monumental character which are designed 
 to last for centuries, the strict economy of material, which is sometimes deemed 
 necessary in engineering works, is not advisable ; because mass, solidity and 
 durability are of the very essence of their architectural character."* 
 
 * In late Gothic, e.g. in the Coventry churches and in the choir of ST MARY REDCLIFFE 
 (525), the clerestory wall was made thinner than the pier arches which supported it. 
 + Scott's Lectures, ii. 189 ; cf. Ruskin's Seven Lamps, 64. 
 X Fergusson's History, i. 15. 
 
ALTITUDE. 6 1 
 
 This unsubstantiality of skeleton construction was, however, largely counter- 
 acted by opacity of glass. How essential to Gothic design is stained glass may 
 be seen by visiting any church which has now but white glass. Such a church 
 seems but a collection of stone scaffolding. With stained glass, even if it be 
 one great lantern, like KING'S COLLEGE CHAPEL, CAMBRIDGE (62), an apparent 
 solidity is produced that reassures. " None would have made walls which 
 are literally windows, unless strength of colour had come forward to simu- 
 late strength of substance." * Nothing in the whole history of architecture is so 
 unsatisfactory as an Amiens glazed in white glass; nothing so delightful as that 
 same church filled with stained glass, provided that the glass be good. 
 
 ALTITUDE and VerticALITY of Gothic. — In a Gothic as compared with 
 a Romanesque church or part of a church there is usually a considerable increase 
 of height ; e.g. at Norwich the nave, which retains its Norman clerestory, is 6g\ 
 feet high ; the choir, which has a Gothic clerestory, is 83 -1 feet high. A similar 
 difference between the height of nave and choir obtains at Gloucester. The 
 parts that rise are the pier arcade and the clerestory ; the triforium tends to 
 diminish in height, as its roof is flattened more and more. The primary reason 
 for the greater height of Gothic pier arcade and clerestory is a practical one ; it 
 is due to the desire to have taller windows and more light. It would be useless 
 to make the aisle windows taller if the pier arcade remained low. Tallness of 
 pier arcade is as necessary as tallness of clerestory, if more abundant light is 
 to be had. 
 
 Of the two chief factors in the dimensions of an interior, breadth and height, 
 the former is the master-factor ; the breadth governs the height ; e.g. if an 
 English church is to have a nave of 32 feet span, as at Salisbury and Winchester, 
 each aisle may have a span of about 16 feet. And if the aisle windows are 
 to be sufficiently tall, the aisle should be about 40 feet high ; which should be 
 the height of the pier arcade also. Now a satisfactory elevation is one that 
 allots one-half of the total height of the interior to the pier arcade, one-sixth 
 to the triforium, and one-third to the clerestory ; therefore if the pier arcade 
 has a height of 40 feet, the triforium arcade will occupy about 13 feet, the 
 clerestory about 27 feet, and the total height to the top of the clerestory will 
 be 80 feet; externally, the ridge of the roof will be about 108 feet high. 
 This corresponds pretty closely with the distribution of the three vertical stories 
 of Salisbury, which is 84 feet high, and of WINCHESTER NAVE (90), which 
 is 78 feet high. In such an elevation, the height both of the nave and of the 
 aisles is about two and a half times their span.f 
 
 But in the He de France the builders, in fixing the height of the churches, 
 by no means allowed themselves to be curtailed by the fenestration. Amiens, 
 with a nave of 46 feet span, would, if built with the average English proportions, 
 have an internal height of 1 14 feet ; as a matter of fact, the height is 144 feet. 
 In Amiens the height of the nave and aisles is respectively nearly three times 
 their span. Light enough could have been gained without running up the aisles 
 and nave to such great heights. Partly from ambitions of masoncraft, partly 
 
 * Rensselaer's English Cathedrals, 431. 
 
 t The above dimensions are of course merely an imaginary example ; there are many 
 deviations from such a standard as this. 
 
62 VERTICALITY. 
 
 from exalted ideas of design, the boundaries of the material were far outpassed. 
 The result was a series of buildings surpassing all the other works of man ; in 
 which the builders reached forward to and attained not merely the beautiful, but 
 the sublime. Nowhere does one feel so much the greatness and the insignifi- 
 cance of man. Man who built these towering vaults is crushed and overwhelmed 
 by his own work. 
 
 To a large extent verticality is the dominant note of Gothic architecture ; 
 horizontality of Romanesque. All the vertical lines that were present in the 
 Romanesque building are present in the Gothic ; but they are all elongated 
 owing to the greater height of the building. The piers of Durham give im- 
 portant vertical lines ; but there is a great difference between these and the 
 vault piers of GLOUCESTER CHOIR (59) rising into the clerestory 66 feet 
 from the pavement. So with the vaulting shafts ; they shared in the general 
 uplifting of the interior. The pointing, too, of every semicircular arch carried 
 the eye upwards. The articulation of the piers into shafts and columns and the 
 disuse of the Romanesque cylinder immensely multiplied the number of vertical 
 lines. So also did the multiplication of window mullions. On the other hand, 
 the space from buttress to buttress being occupied with windows, there was less 
 room for the horizontal line either inside or outside the buildings. Bands, 
 too, which checked the upward flow of the shafting, were for the most part 
 abandoned. From the summit of the vaulting shafts, as at EXETER (9), 
 whole sheaves of ribs ran upwards to the ridge of the vault. Externally, the 
 vertical line was still more pronounced ; in the great projection of the buttress ; 
 in the substitution of the pinnacle for the gablet ; above all, in the upper growth 
 of the spire. 
 
 Nevertheless, it is possible to overemphasise the verticality of Gothic 
 architecture. What the builders took away with one hand, they put back with 
 the other. If they added tiercerons to diagonal and transverse ribs, they also 
 added horizontal ridge ribs. If they articulated the vaulting shaft, they usually 
 cut it short at a corbel. If more and more they disused the string, they more 
 and more filled their windows with transoms. If they added the pinnacle, they 
 substituted for the corbel table the far more emphatic horizontality of the pierced 
 or embattled parapet. Whole districts gave themselves up to tower design, and 
 eschewed the spire. So then we may say, with more justice, that Gothic is not 
 the embodiment of verticality alone, but rather the just balance of the two 
 conflicting principles of the vertical and the horizontal line. 
 
INTERIOK OF KING'S COLLEGE CHAPEL, LOOKING WEST. 
 

63 
 
 Westminster. 
 
Chapter IV. 
 
 CHARACTERISTICS OF ENGLISH GOTHIC 
 
 FRc IM C. I I/O To C. 131 5- 
 
 Planning — Internal Elevation — East Front — Transept Front — West Front — Vaulting — 
 
 Piers — Ornament. 
 
 PLANNING. — By the end of the twelfth century* the planning of the greater 
 churches had been revolutionised. Three new systems of church planning had 
 come into use ; differing from one another ; but all agreeing in breaking away 
 completely from Romanesque tradition. No more churches were built with 
 parallel side apses, like those of ST MARY'S, GUILDFORD (36) ; a belated 
 example of this class. Equally the Norwich plan, with semicircular apse, ambu- 
 latory and tangential chapels, went out of use ; except at Westminster in the 
 thirteenth and Tewkesbury in the fourteenth century, where it was revived with 
 polygonal apses. No more semicircular apses were built after those on the 
 east sides of the choir transepts of Canterbury and LINCOLN (66). All the 
 great churches, however, remained cruciform, and most had aisled naves. The 
 Norman western transept was repeated at LINCOLN (151. 1) and Peterborough. 
 The eastern transept of Canterbury was much copied in this period ; e.g. at 
 LINCOLN (66), Rochester, Worcester, SALISBURY (170), BEVERLEY (176). 
 Of the transepts some were without aisles ; some had an eastern aisle ; few 
 had western as well as eastern aisles ; none had return galleries, except the 
 north-eastern transept of Lincoln. 4 " Some of the eastern transepts were as loftv 
 as the choir ; e.g. at Beverley, Worcester, and Salisbury ; others were as low as 
 the aisles ; e.g. at Southwell and Exeter. So also if there was an eastern 
 chapel, it might be low, as at Chichester and SALISBURY I 170J ; or of the 
 full height of the choir, as at Rochester, Worcester, BEVERLEY (176)4 At 
 FOUNTAINS (150.2) and Durham the choir transept was built at the eastern 
 extremity of the church. Of eastern limbs three types came into use about 
 the same time; that of hxi-okh cathkhkal (i 52.3 , 1154-1180, with aisled 
 choir and unaisled sanctuary; that of ST CROSS, WINCHESTER (104) (probably 
 not earlier than I 160;, New Shoreham 'probably c. 1175); and JERVAULX 
 
 * The period c. 1170 to c. 1315 corresponds roughly with the Early English and Early 
 Decorated of Rickman, Bloxam, and Parker ; and with the Late Transitional, Lancet, and 
 Geometrical periods of Sharpe. 
 
 t The south-eastern transept of Lincoln seems to have been remodelled in the thirteenth 
 century. 
 
 \ When tall, it sometimes formed the presbytery. 
 
 E 
 
66 
 
 Lincoln Minster, S.E Transept, 
 
INTERNAL ELEVATION. 67 
 
 (153.3), nl which both choir and presbytery are completely aisled; and that 
 of Chichester, c. 1170, with retrochoir and rectangular eastern chapel* Many 
 Norman choirs were found too small and were pulled down and rebuilt in 
 Gothic. + In several cases, as at Lincoln, this was the prelude to the rebuilding 
 in Gothic of the whole church. 
 
 Little change occurs in the planning of the parish churches till the second 
 half of the thirteenth century. All the Norman forms of plan remain in use. 
 The simple forms of plan, however, tend to be replaced by the more complex 
 forms, as transepts and aisles come more into use. Aisles are still narrow and 
 low ; and clerestories rare. It was not till the second half of the thirteenth 
 century that the aisles became broad, as at St Martin's, Leicester, and War- 
 mington ; or lofty, separated from the nave by tall, slender, graceful piers, as in 
 HOWDEN NAVE (546), HEDON (544) and Stone. 
 
 Internal Elevation. — As in the Romanesque churches, so in our early 
 Gothic work all the greater churches internally were three stories high ; ground 
 story, triforium arcade, and clerestory. And all the various Romanesque dis- 
 positions still survived. In ELV PRESBYTERY ('526 and in WESTMINSTER 
 (379) the triforium still retains windows in its back wall ; giving an exterior 
 three stories high. This arrangement is, however, rare in Gothic. The curious 
 design of the Augustinians of OXFORD (27; and Dunstable is repeated by 
 the Benedictines of GLASTONBURY (536), but with pointed arches. Then this 
 design also disappears. The tall triforium arcade of Romsey, St Bartholomew's, 
 Smithfield — illogical in design because the triforium has no windows at the 
 back — is repeated in the early Gothic of Hexham and WHITBY (1 14;, and later 
 in YORK TRANSEPT (523), and the north side of the nave of BRIDLINGTON 
 (125). More often, however, the height of the triforium is reduced by flat- 
 tening its roof more or less. The space thus gained was sometimes given to 
 the clerestory; as in the south side of BRIDLINGTON NAVE (125;, and in 
 Guisborough choir ; \ and Exeter ; or the height of the piers was increased, as 
 the choirs of CANTERBURY fio6j, Salisbury, and BEVERLEY (51). 
 
 In the Cistercian churches, however, the design of Fountains and Kirkstall 
 naves survives, here and there, as late as TINTERN (524), 1269- 1287. But 
 a much more common and a more important elevation is that in which the 
 jambs of the clerestory window are carried down to the string of the triforium ; 
 e.g. ST DAVID'S (5^5); Dore ; Southwell and PERSHORE choirs (75); and 
 the south side of BRIDLINGTON NAVE (125 j. The most advanced specimen 
 of this treatment is the nave of YORK (10), the foundation stone of which 
 was laid in 1291. Here not only the jambs, but all the four mullions of the 
 clerestory windows, descend to the triforium string. 
 
 The parish churches for the most part are still without a clerestory, and the 
 
 * The Chichester plan occurs also at Dore and Glastonbury, but without the eastern 
 Lady Chapel. 
 
 + E.g. Ripon ; York ; Wells ; Lincoln ; Lichfield ; Salisbury ; St Paul's ; Beverley ; South- 
 well ; Hexham; Southwark ; Rochester; Worcester ; Whitby; Boxgrove ; Chester Cathedral; 
 Pershore ; Rievaulx ; Fountains ; Carlisle ; and after c. 1250 Lincoln, Tintern, Thornton, 
 Exeter, Guisborough. 
 
 % Illustrated in Sharpe's Arch. Parallels, Plate 70. 
 
68 EXTERNAL ELEVATION. 
 
 internal elevation is of one story, even in the great church of Yarmouth. 
 Where a clerestory occurs, in the first half of the century, its windows are 
 often set in an arcade of pointed arches ; e.g. at Darlington, Great Grimsby * 
 Elm, and West Walton. In the last half of the century low clerestories 
 become more common ; their windows are often small circles. The naves 
 of HOWDEN (546) and HEDON (S44) show the clerestory window rising to 
 a considerable height. Where the parish church has a clerestory, the internal 
 elevation is one of two stories. 
 
 EAST FRONT. — Of the east fronts of the thirteenth century several distinct 
 types survive. 1. At DORE (182) there is a rectangular ambulatory, but not a 
 projecting eastern chapel. St Saviour's, Southwark, is similar ; and originally 
 perhaps Winchester, before the Perpendicular Lady Chapel was added. 2. At 
 SALISBURY (170), Chester Cathedral, Hereford in the first half of the century, 
 and at Chichester, Exeter, St David's and St Albans in the second half, a 
 low Lady Chapel forms the eastern termination. 3. At Tynemouth, BEVERLEY 
 (176), and SOUTHWELL (359), the choir ends at full height in a short unaisled 
 presbytery or Lady Chapel. 4. At Whitby, Rievaulx, BOXGROVE (373), ELYf 
 (464), in the first half of the century, both choir and aisles are carried at full 
 height to the east: as they were in the second half at LINCOLN (177), 
 Tintern, Ripon, and Guisborough. The east fronts of the chapels of Ely 
 Palace, London, and MERTON COLLEGE, OXFORD (473), also belong to the 
 last years of the thirteenth century. 5. At FOUNTAINS (150) and Durham the 
 churches terminate to the east in an eastern transept, with nine altars. 6. 
 WESTMINSTER (6f) adopts the polygonal apse of French Gothic. 
 
 TRANSEPT FRONTS. — To the twelfth century belong the transept fronts 
 of Ripon and Canterbury ; as well as those of the eastern transepts of LINCOLN 
 (66). To the first half of the thirteenth century belong the north tran- 
 septs of LINCOLN (69) and Hedon ; four transepts of SALISBURY (170) and 
 four of BEVERLEY (176); two of Whitby and Rievaulx; two of York — 
 differing entirely in design — the north transept of Rochester ; and later in the 
 century those of Tintern. As these fronts were often seen in conjunction with the 
 sides of the transepts, they often followed the dispositions of the latter ; J e.g. at 
 Hedon § and in the east transepts of Worcester the sides of the transept contain 
 two rows of windows ; and beneath the bottom row is blank wall. In the north 
 elevation, therefore, at Hedon there is a doorway, at Worcester a blank wall ; 
 two triplets of lancets superposed, corresponding to the rows of lateral windows ; 
 and a third graduated triplet of lancets in the gable. This is the logical eleva- 
 tion for an unaisled transept ; viz. one of four stories. On the other hand, if 
 the transept have aisles, then on its flanks there may be (1) wall beneath aisle 
 windows ; (2) aisle windows ; (3) aisle roof, which gives a half gable at the end 
 of the transept ; (4) clerestory windows. The normal elevation for such a 
 transept is one of five stories. This logical disposition obtains in all the 
 
 * Illustrated in Building News, March 21, 1875. 
 
 + The eastern terminations of the aisles have been ruined by conversion into chantry 
 chapels by Bishops Alcock and West. 
 
 % So also in Norman transepts ; e.g. Winchester and Norwich. 
 § Illustrated in Builder, Dec. 17, 1S87. 
 
6 9 
 
 Lincoln North Transept. 
 
WEST FRONT. y\ 
 
 transept fronts of SALISBURY (i/O); in each there is (i) wall, with or with- 
 out doorway ; (2) a triplet or quintet of lancets ; (3; a band of arcading or 
 of low windows; (4) another triplet or quintet of lancets; (5) the gable con- 
 taining graduated lancets or a rose window. To this type belongs the noble 
 north facade of Westminster. It is five stories high ; the great rose is placed in 
 the fourth story instead of the gable ; and as the chief entrance to the church 
 is from the north, there are three lofty doorways. But when it was thought fit, 
 such logical dispositions were disregarded ; e.g. in the central transepts of 
 BEVERLEY f 176J and the south transept of York the logical arrangement was 
 disturbed in order to get more headway for doors ; while in all the Beverley 
 transepts, quite illogically, the gables were cut up into two stories by a string. 
 The Whitby elevation also is illogical. The greatest revolution, however, was 
 in the north transept of YORK (11). Here the three central stories were con- 
 solidated into one ; and this one great central story was filled with five enor- 
 mous lancets, all of the same height, the famous Five Sisters. A little later 
 this elevation of three stories was adopted at Tintern Abbey ; except that for a 
 quintet of lancets there was substituted a tall traceried window of six lights ; 
 and in the north transept of Hereford. 
 
 In the east transept of Canterbury ; and the central transepts of LINCOLN 
 (69), Whitby, and Beverley ; and in the south transept of York and the north 
 transept of Westminster circular windows are employed. 
 
 West Front. — Of the artistic problems which came before the mediaeval 
 builder for solution none seem to have presented such great difficulties as the 
 composition of the grand facade of the greater churches. When a civic building 
 was designed, e.g. the Cloth Hall of Ypres,* which is 440 feet long, no one 
 dreamt of making one end of it the grand facade. But this is exactly what 
 the church architect, for ritualistic reasons, everywhere was compelled to do. 
 Otherwise he might have made what is now a side of the church the principal 
 facade ; a facade which in many cases would have exceeded 500 feet in length. 
 In the centre of this might have been placed the main entrance ; emphasised, 
 perhaps, as at Ypres, by a great central tower. Two minor towers, to the far 
 east and west, might have brought together the wings. But to restrict to a 
 breadth of some 80 feet the grand facade of a church 500 feet long, and with 
 transepts spreading out perhaps 200 feet, was to make an adequate solution 
 almost impossible. 
 
 Nevertheless an adequate solution was found. This was to give to the 
 facade in height what could not be Ln'ven in breadth. Such a facade was familiar 
 to the builders in Normandy in the eleventh century ; and was reproduced 
 at SOUTHWELL (S20), DURHAM (28), LINCOLN ''562 , and elsewhere. Early in 
 the thirteenth century it is seen at Ripon ; and at the very end of the century 
 at LICHFIELD (frontispiece). Still greater is the adequacy of the facade if 
 the towers have spires ; as at Lichfield, and formerly at Lincoln and Ripon. 
 And if, behind and between these, there is a central spire, so lofty that this also 
 enters into the grouping of the west front, as at Lichfield, and formerly at 
 Lincoln and Ripon, then, narrow as is the facade, it is adequate even for a 
 church so vastly long and broad as Lincoln. 
 
 * Illustrated in Fergusson, ii. 201. 
 
72 
 
 WEST FRONT. 
 
 This fine type of design was still further strengthened by setting'the western 
 rers clear of the aisles instead of in a line with them. At Lichfield the 
 
 towers clear o 
 
 Howden West Front. 
 
 towers project but slightly to north and south; but at WELLS (154.3) they are 
 quite clear. 
 
 For the success of the twin tower facade, however, it is indispensable that 
 the towers shall be towers all the way to the ground. The towers must be 
 
WEST FRONT. 73 
 
 wholly independent of the central facade: as they are in the Abbaye-aux-Hommes, 
 Caen ; Castle Acre, and Southwell. The distinction between the central and the 
 lateral facades is strongly emphasised at Bayeux and Beverley Minster, and with 
 magnificent effect. In this respect the western towers of Durham and Ripon 
 show some timidity , at Wells and Lichfield the towers are lost in the facade ; 
 at Lincoln and Peterborough they rise in inexplicable fashion in the rear of the 
 facade. 
 
 The towered facade, however, was perhaps an architectural extravagance ; 
 one of the few instances in Gothic architecture of work done mainly for effect.* 
 For this reason, perhaps, and because of its cost, it was adopted in comparatively 
 few churches. Another design was borrowed from the Norman churches which 
 could be turned to religious account. At Ely and Castle Acre, and originally 
 at Hereford, the facade had included a screen wall ornamented with band 
 upon band of arcadings of semicircles, intersecting semicircles, pointed and 
 trefoiled arches. These arcades were built more deeply recessed ; and in each 
 recess was placed a statue. Such a statued screen, an open-air reredos or 
 iconostasis, was defensible on religious grounds. It taught Scripture History 
 and the Legends of the Saints. t Such a great rectangular wall was not designed 
 merely as a facade, and is not to be criticised as a facade. The criticism which 
 it does provoke is that it was ill advised to put sculpture at heights where its 
 meaning was indistinguishable, and where it is exposed to the inclemency of 
 our English climate. Lincoln, Wells, Salisbury adopted this reredos type of 
 facade in the first half of the thirteenth century. At the very end of the century 
 it reappears, for the last time, at Lichfield ; but with a couple of steeples perched 
 on the top of it. After this it disappears from English architecture. 
 
 The simplest method of disposing of the difficulty with the grand facade 
 was to recognise frankly that the west front was not the grand facade ; and to 
 cease to try to make it one. This was the sensible method adopted in far the 
 most churches. The west front was designed in them in the same simple fashion 
 as the north and south fronts of the transepts. Possibly Cistercian precedent 
 had considerable weight; for no Cistercian church had either western towers or 
 the screen-wall facade. So the simple type of west front greatly preponderated. 
 It occurred in the first half of the thirteenth century at Wenlock ; Whitby; 
 Bolton; St Saviour's, Southwark ; Romsey ; BINHAM (471); in the second half 
 at HOWDEN (72); and frequently in later work. It is the same as the west 
 front of a parish church which has no western tower; for distinctness we 
 may call this third type of western facade the parochial. The history of the 
 design of the parochial facade is the same as that of the transept facade. At 
 first it is cut up into four or five stories ; as at Bolton and Byland. Then, in 
 the west front of Romsey, the central tiers of windows are consolidated into one 
 gigantic triplet of graduated lancets ; and the number of stories is reduced 
 to three. But little, if at all later, is the west front of BINHAM (471); here 
 
 * One western tower might be useful as a campanile. But bells were often placet! 111 the 
 central tower ; e.g. LINCOLN (328). The western towers, however, have constructional value ; 
 see pages 381 and 598. 
 
 t The French preferred to teach them in the statued aichivolts of their doorways ; and 
 they were taught both by French and English in the stained glass windows. 
 
74 EARLY GOTHIC VAULTING. 
 
 it seems to have been intended to have superposed rows of lancets, as at 
 Ripon ; but a single great window of bar tracery was preferred. Other three 
 story facades are those of Valle Crucis and Tintern. 
 
 At Peterborough, as at Lincoln, there are two facades. The inner fac;ade was 
 built at the end of the twelfth century ; and was to have flanking towers (as at 
 Wells) of which one only has been completed. The outer fagade was built a 
 generation later and is still broader. 
 
 Vaulting. — None of the Gothic vaults are groined ; * all are ribbed. But 
 there are considerable differences between the ribbed vaults, e.g. of DURHAM 
 AISLE (315) and NAVE (8); and those, e.g. of NEW SHOREHAM (313) and 
 CHICHESTER (3 1 3). In the first place, the filling-in of the latter is of ashlar, 
 and is much less heavy. Rubble "filling-in," however, was frequently retained, 
 e.g. in lichfield nave (3 1 3). The ribs became much less massive ; and were 
 composed of longer blocks. It ceased to be customary to make the trans- 
 verse thicker than the diagonal ribs. At ROCHE (675) they differ much ; while 
 at BYLAND (67 5 J, which can be but little if any later, they are of the same 
 profile. The rectangular is gradually replaced by a triangular profile ; the 
 Gothic moldings being executed more and more on the chamfer plane : e.g. 
 contrast the ribs of WHITBY CHOIR (675.12) with that of LINCOLN GALILEE 
 (677-3)- The lower portion of the ribs ceases to be built independently ; being 
 constructed in solid springers. Sexpartite vaulting received encouragement from 
 Canterbury choir ; but quadripartite vaulting was always the more common, and 
 finally superseded sexpartite. Additional ribs were added in LINCOLN CHOIR 
 
 Tooth Ornament. 
 
 (327), commenced 1192; and to give abutment to these a new rib, the longi- 
 tudinal ridge rib, was invented. Other intermediate ribs, or tiercerons, were 
 added in LINCOLN NAVE (327), c. 1230 ; and to abut these, transverse as well as 
 longitudinal ridge ribs, were employed. At Ripon, Hexham, WHITBY (114), 
 Carlisle, the tradition of the Norman ceiling survived, and no high vaults were 
 built. 
 
 PIERS. — In the greater churches three types of pier were in use in the earlier 
 part of the period. The first is the western pier ; usually short and massive ; 
 not employing marble ; but encircled with slender shafts of freestone, arranged 
 in triplets ; e.g. in WELLS (209), LICHFIELD (244) ; a late example is PER- 
 SHORE CHOIR (75). The second is the southern pier ; usually tall and 
 graceful ; encircled by slender detached shafts of marble ; banded with annulets 
 of marble or bronze; e.g. CHICHESTER RETROCHOIR (245), ELY PRESBYTERY 
 (247);^ late examples are Winchester chancel and Wells retrochoir. The 
 third is the northern pier, which discards slender shafts, and is made up of a 
 cluster of stout columns, which are generally of freestone. Some or all of 
 these columns are usually pointed in section ; e.g. ROCHE (661.2) and BYLAND 
 * Throughout the volume the term " groined " is confined to vaults which do not possess ribs. 
 
75 
 
 Pershore Choir from S.W. 
 
EARLY GOTHIC ORNAMENT. 77 
 
 (661.3). The clustered column is sometimes found where the southern or 
 western type of pier might be expected ; e.g. ST SAVIOUR'S, SOUTHWARK 
 (521) ; ST ALBANS NAVE (14) ; EXETER (241 ). 
 
 Ornament. — The tooth ornament had enormous* vogue in the thirteenth 
 century: e.g. at SKELTON and WARMINGTON (78 and 5/8); partly because 
 of its effectiveness, partly because it was easy to execute ; as is shown 
 in the diagram on page 74. It had its origin in the Norman nail-head.^; 
 It is one of the few ornaments without a classical pedigree. An early 
 example of it occurs in the labels of the aisle windows of the west front 
 of Rochester, which is probably 1125 to 11 37.} It occurs, fully developed, 
 1131-1133, at Terouanne in the North of France.§ It occurs in the west 
 doorways of LESSAY ABBEY (3 1 5), and of Davington Priory; the latter was 
 founded in 11 53; also in the so-called Baptistery at Canterbury, c. 1 1 60 ; ' in 
 doorways at Stillingfleet and Brinkburn; and in the sanctuary arch of Compton ; 
 and among Norman moldings in the north doorway of St Margaret at Cliffe. It 
 occurs as a string at the back of the pier arches of Steyning. It is used profusely 
 at Canterbury in the work of 1175-1184. It is very rare after the thirteenth 
 century ; but an example occurs in the moldings of a Tudor arch at Lichfield," 
 and an imitation of it in another at Congresbury Vicarage. It was common in 
 Continental Gothic also; eg, in Italy at Perugia, Term', and Verona;** and is 
 very common in Spanish Romanesque ; e.g. in Tarragona cloister. It is still a 
 favourite in Cyprus, ff Usually this ornament is designed as a pyramid of four 
 leaves ; but at Salisbury it consists of only two leaves ; i.e. a half pyramid ; 
 the treatment at Binham £ and West Walton J| is similar. In late and rich 
 work scrolls of foliage are carved on each face of the pyramid ; e.g. the north 
 doorway of Lichfield and west front of Dunstable. 
 
 Crockets are said to be derived from the volutes of the Corinthian or 
 Composite capital (425). But our earliest examples are mere incurved hooks, 
 resembling the pastoral staff of a bishop; and corresponding to the earliest knobby 
 type of stalk foliage, e.g. in St Hugh's work at LINCOLN (249; and all round 
 the orders and down the jambs of the west doorway of Strata Florida. These 
 hooks were soon foliated or otherwise ornamented ; e.g. at Wells and Salisbury ;§§ 
 the south porch of the west front of St Albans and Lincoln presbytery. 
 
 Some of the earliest crockets occur at the back of shafts ; e.g. in St Hugh's 
 work at Lincoln and the west porch of St Albans ; soon they are placed between 
 the shafts. From the middle of the thirteenth century their chief use is to run 
 up the straight gables of canopies ; but they are found in many other positions ; 
 e.g. on the flying buttresses of ST MARY REDCLIEFE, BRISTOL tf6 : on the 
 hood-molds of doorways, e.g. CLEY (85); and of windows, e.g. Louth spire 
 and WREXHAM tower (6og) ; on gables, as at LOUTH (397) ; on canopies, as 
 in HOWDEN CHAPTER HoUSE (137); on spires, as at LOUTH (6l 1;. 
 
 Nowhere more than in the Lincoln galilee, which "bristles with tooth ornament, like a 
 cavern of crystals" ; 5355 examples occur in this porch. 
 
 t Sharpe's Nene Valley, 4. % Hope's Rochester, 33. 
 
 § Illustrated in Enlart's Manuel, i. 354. , Willis' Canterbury, S2, note. 
 
 IT Petit's Church Architecture, i. 215. ** Willis' Middle Ages, 196. 
 
 ft Enlart's Manuel, 354, 1. \\ See Colling's Details, i., E.E., Plates 22 and 24. 
 
 §i$ Illustrated in liloxam, 179. ||[ Colling's Cathie Ornaments, i., Plates 56 and 21. 
 
/S 
 
 Skelton Porch. 
 
EARLY GOTHIC DETAIL. 79 
 
 For the remaining members of churches of this period, Chapters XV. to 
 XLI. may be consulted. For arches see especially page 279 ; for buttresses and 
 pinnacles, pages 358 and 363 : tor flying buttresses, page 371: for corbel tables, 
 parapets, pages 392, 393 : for strings, hood-molds, dripstones, and basement courses, 
 page 406 : for foliated capitals, page 429 : for molded capitals, page 442 : for 
 bases, plinth, griffe, page 451: for 'windows and tracery, page 460 : for roofs, 
 P a & e 559 : f° r doorways, page 579: for towers, page 597: for spires, page 617. 
 See also pages 105-126. 
 
Chapter V. 
 CHARACTERISTICS OF ENGLISH GOTHIC 
 
 FROM C. I3OO TO C. I35O. 
 
 Planning — Internal Elevation — East Front — West Front — Vaulting — Piers— Ornament. 
 
 PLANNING. — No new plans were adopted in the greater churches.* The Salis- 
 bury plan was repeated at Milton Abbas, WELLS (1 S4-o). and Ottery St Mary. 
 The aisled choir, with unaisled presbytery, reappears at Bristol ; and at Lichfield 
 an aisled presbytery with a tall unaisled Lady Chapel is built. At Howden, 
 SELBY (86), and CARLISLE (128) aisles are built to the full length eastward of 
 the eastern limb. Eastern transepts are again built ; at Bayham and Wells. 
 The rebuilding of choirs had been carried on with such vigour in the thirteenth 
 century that not much remained to do. However the choirs of Lichfield, Wells, 
 and Carlisle were lengthened ; and those of Howden, Selby, and Bristol were 
 rebuilt. 
 
 In the parish churches all the plans in use in the twelfth were retained in 
 the fourteenth century. Penton Mewsey, Hampshire, has unaisled nave and 
 chancel. Leckhampton, Gloucester, had unaisled nave, towered choir and 
 sanctuary. Shottesbrooke is cruciform, without aisles. BOSTON ( 222), Hol- 
 beach, Hingham have aisled nave and unaisled chancel. The last is by far 
 the most common plan of the parish church to the end of the Gothic period. In 
 large churches, however, the cruciform plan was still in vogue ; e.g. at Tideswell, 
 Nantwich, and Snettisham, where the nave is aisled ; at Patrington, where both 
 nave and transept have aisles on each side ; at HULL (96), where there are 
 full-length aisles to the chancel as well as to the nave. Many chancels are 
 rebuilt ; and aisles are rebuilt broader and loftier. 
 
 Internal Elevation. — In the fourteenth century the internal elevation, 
 as before, in the greater church is one of three stories. One belated example 
 occurs of a tall triforium with windows at the back ; viz. in ELY CHOIR (526). 
 This, however, was so designed in order to assimilate it to the presbytery, with 
 which it is in juxtaposition to the east. In the naves of Beverley, Worcester, and 
 Westminster Abbey, triforium arcades occur of moderate elevation ; in all cases 
 to be in harmony with earlier work with which they are in juxtaposition. But, 
 more commonly, the precedent of Pershore and Southwell is followed ; and the 
 jambs of the clerestory window are brought down to the triforium string ; as in 
 Chester nave and TEWKESBURY CHOIR (165). Sometimes the design of York 
 
 * The first half of the fourteenth century corresponds roughly to the Late Decorated period 
 of Rickman, Bloxam and Parker ; and the years 1315-1360 to the Curvilinear period of Sharpe. 
 
Si 
 
 Hull Chancel 
 
82 EXTERNAL ELEVATION. 
 
 nave or the south side of Bridlington nave is adopted; and the wall passage 
 is protected by a parapet ; e.g. in the choirs of Lichfield and SELBY (390). 
 At WELLS (127) the front of the triforium of the choir of 1175 and that of 
 the fourteenth-century presbytery were alike masked by rows of niches. To the 
 eye all these latter interiors, viz. at Chester, Tewkesbury, Lichfield, Selby, Wells, 
 have the appearance of being but two stories high. 
 
 In the larger aisled parish churches the precedent of Howden and Hedon is 
 adopted generally ; most of them have clerestories, and the elevation is one of 
 two stories. PATRINGTON (133) is an exception. Towards the end of the 
 period, however; above all in the chancel of HULL (81, 474); the clerestory 
 window grows vastly both in height and breadth. And before the century is 
 over, two windows may be found in each bay of the clerestory ; e.g. at BOSTON 
 (222) and Holbeach ; as previously at HOWDEN (546). 
 
 East Front. — 1. In the fourteenth century the Salisbury type of east front 
 is revived at WELLS (602) and at Ottery St Mary ; by the latter in imitation 
 of Exeter. At Lichfield the choir is lengthened and a lofty Lady Chapel is 
 added. 2. At Tewkesbury the semicircular apse and chapels of the choir are 
 made polygonal. 3. But the characteristic east front now is rectangular ; with 
 aisles as long as the choir, and the latter carried up in three stories. Of this there 
 are magnificent examples at Selby ; Hull ; CARLISLE (128) ; and Howden.* 
 
 WEST FRONTS. — The chief west fronts of the fourteenth century are 
 Howden, Exeter, and York. HOWDEN (72) and Exeter f are both of the 
 parochial type. At YORK (82) the lateral facades are blended with the central 
 one, to the great detriment of the towers : as at Wells and Lichfield, the west 
 front is really a single complete facade with a pair of towers perched on the top 
 unrelated to it. 
 
 Beautiful facades of this period are seen in many parish churches ; especially 
 in Mid-Lincolnshire. 
 
 VAULTING. — The simpler forms of quadripartite vaulting were still retained; 
 especially in the North of England ; e.g. Beverley nave ; Howden choir ; Guis- 
 borough ; and also in the choir of Milton Abbas, Dorset. But in the South and 
 West a new rib, the lierne, was highly developed, and led to combinations of the 
 utmost complexity ; e.g. in TEWKESBURY NAVE and CHOIR (332, 330)4 
 
 In BRISTOL CATHEDRAL (329) skeleton vaulting is much employed. 
 
 Owing to the multiplicity of ribs in some of these vaults the filling-in con- 
 sisted of " panels," instead of coursed ashlar. 
 
 In Selby choir a wooden vault was substituted for the stone vault origi- 
 nally intended. In BRIDLINGTON (125) and Howden naves no high vaults 
 were built ; nor in the south transept of St Werburgh, Chester ; nor in the 
 retrochoir of St Albans Cathedral. On the other hand, the churches of St Mary 
 Redcliffe, Bristol ; Ottery St Mary, and Patrington were vaulted, wholly or in part. 
 
 PlERS. — The fourteenth century is marked by the disappearance both of the 
 
 * A restoration of the east front of Howden is given in Sharpe's Arch. Parallels, Plate 86. 
 
 t Exeter facade has been greatly altered by subsequent additions. 
 
 t Lierne vaults occur in Tewkesbury nave and choir ; Bristol Cathedral choir and the 
 south transept of St Mary Redcliffe ; wells choir (332) and lady chapel (325) ; Malmesbury 
 nave ; Ottery St Mary ; ELY choir (329) and Lady Chapel, Nantwich chancel and transept. 
 
YORK MINSTER, WEST FRONT. 
 
FOURTEENTH-CENTURY ORNAMENT. 83 
 
 Western triple * shafts of freestone and the Southern detached and banded shafts 
 of marble. Instead of these the Northern type of pier prevails ; viz. a cluster of 
 engaged columns; e.g. in the choirs of Milton Abbas, SELBY (390), and Howden ; 
 the naves of York, St Albans, and Worcester; ELY CHOIR (251); and Chester 
 south transept. But at BRISTOL (661.11)+ a completely new form of pier is 
 devised. 
 
 ORNAMENT. — The ball-flower is just as characteristic of the first half of the 
 fourteenth century as is the tooth ornament of the thirteenth. It has been 
 supposed by some to be the trollius or globe-flower ; by others to be derived 
 from a hawk's bell \ ; by others to be a horse-bell, in that the thong as well as 
 the bells is sometimes represented. § It is found, however, in late Norman work, 
 side by side with the pellet ; and so may be taken to be but a survival of this 
 Norman ornament. 
 
 In France also it first occurs solid, then pierced with lobes, in the twelfth 
 and thirteenth centuries ; e.g. in the balustrade of the towers of Notre Dame, 
 Paris." During the course of the thirteenth century it was abandoned in France. 
 In England it has been said to be confined almost wholly to the reign of Edward 
 II. 1307-1327). But it occurs in the hollow architrave moldings of the arches of 
 the thirteenth-century clerestory of Beverley Minster** ; and in the west front of 
 Salisbury. Late examples are seen, c. 1380, in the west doorway of St Mary's, 
 Beverley ; and in the late Gothic porch of Worlingworth, Suffolk. It is used 
 with the greatest profusion in the Western counties ; e.g. St Catharine's Chapel, 
 Ledbury ; Hereford central tower; GLOUCESTER (360), south aisle of nave; 
 in every window and doorway of Badgeworth, Gloucestershire. At Gloucester ++ 
 a horizontal line drawn across the head of an aisle window, just above the 
 spring of the arch, cuts no fewer than thirty-two ranks of the ball-flower, sixteen 
 within and sixteen without.^ 
 
 The four-leaved flower, composed of four leaves arranged so as to form a 
 square, is particular!}' common in cornices, e.g. at GRANTHAM and ENSHAM 
 (385). It occurs at all periods, but has specialised forms in each ; e.g. on 
 a Norman arch of Northampton St Peter's; c. 1291 in the Eleanor Crosses ; 
 in the fourteenth century at St Stephen's Chapel, Westminster ; and is very 
 common in all the later Gothic ; both in stone and wood work. 
 
 By the end of the thirteenth century crockets cease to be incurved, and the 
 foliage becomes naturalistic; e.g. in Southwell chapterhouse and Exeter reredos;§§ 
 or the leaves are more conventionalised as at Bridlington and Guisborough and 
 Selby; in either case they are given an undulating ogee curve, which in the 
 work of 1315-1350 is strongly emphasised ; e.g. in Selby choir, the Percy tomb 
 at BEVERLEY f269), and ELY LADY CHAPEL 269). 
 
 * Except in Wells presbytery. + See pages 242 and 255. 
 
 % Glossary, 53. : Si ott's Essay. 
 
 || A solid ball-flower and a fluted pellet occur together at Lincoln ; illustrated in Parker's 
 Manual of Gothic Mouldings, page 14. 
 
 IT Illustrated in Viollet-le-Duc, Architecture, ii. 243, 6. * : Bloxam, 178. 
 
 tt Murray's Cathedrals Glottccstcr, rS. 
 
 XX For other examples of the ball-flower see illustrations on pages 474.4 and 587. 
 SS Colling's Gothic Ornaments, i., Plate 14, and Mediaval Foliage, Plate 56. 
 
 For Bridlington, Guisborough, and Selby see Sharpe's Arch. Parallels, Plate 115. 
 
FOURTEENTH-CENTURY ORNAMENT. 
 
 The word diaper (" d'Ypres or dyaper ") was originally applied to cloth 
 worked in square patterns, which was produced at Ypres. It was common on 
 great festivals to hang the walls of the interiors with tapestry ; and this may have 
 led to diaper work in stone. But rude diaper work or trellis occurs in Ernulph's 
 work at Canterbury, 1096, and Rochester, 11 14; and in Grosstete's work at 
 Lincoln (1235-1253). The spandrils of the Norman triforium of Rochester 
 nave were covered with rude foliated patterns, about the middle of the 
 
 twelfth century. In 
 Gothic it is used in the 
 greatest profusion in the 
 triforium of WESTMIN- 
 STER (119) and c. 1290 
 in the Eleanor Cross at 
 Geddington. Diaper 
 work was in special 
 favour in the fourteenth 
 century ; e.g. in SOUTH- 
 WELL SCREEN (179); 
 and in that of the south- 
 east transept of Lincoln, 
 where it takes the form 
 of expanded lilies. 
 
 Niches occur late in 
 the eleventh century in 
 Remigius' west front at 
 Lincoln ; late in the 
 twelfth century all round 
 Barfreston Church ; * and 
 in the thirteenth century 
 on a vast scale in the 
 west fronts of Lincoln, 
 Wells, Salisbury, and 
 Lichfield. In the second 
 half of thirteenth century 
 they are generally sur- 
 mounted by a straight - 
 sided hood-mold ; f as 
 in the west front of 
 Wells, the interior of the 
 nave and chapter house of York, the buttresses of GUISBOROUGH (354) 
 and the west window of HOWDEN (72). For this triangular hood-mold the 
 fourteenth century frequently substituted an ogee hood-mold ; or used them in 
 alternation. The ogee hood-mold, moreover, may bend forward and retreat ; 
 as in the arcading of ELY LADY CHAPEL (269). The niche with ogee 
 canopy may be considered the characteristic feature of fourteenth-century design ; 
 
 * Illustrated in Britton's Arch. Ant., iv. 
 
 + The monument of Aymer de Valence (c. 1325) in Westminster is a late example of this. 
 
 Leverington Church Porch. 
 
8.s 
 
 12 9 6 J o 1 
 
 Cley, Norfolk. 
 
86 
 
 ■ 
 
FOURTEEXTH-CENTURY ORNAMENT. 87 
 
 it is used in vast profusion in the west front of Lichfield Cathedral, the ruined 
 east front of Howden, and the interiors of the presbyteries of WELLS (127) 
 and Lichfield. So complex and beautiful was the elaboration of the niche- 
 that it usurped the interest which should have been retained for the statue it 
 was designed to enshrine. It is as if some school of artists had spent their 
 main effort not on their pictures but on their picture frames. It appears in 
 arcading ; as in the aisle walls of Beverley nave, and under the west towers 
 of LINCOLN (269); in the screen of wood or stone, as at SOUTHWELL 
 (179) ; in the reredos, as at CHRIST CHURCH, HAMPSHIRE ( 180) ; in the canopy 
 of a monument, a piscina, a stoup, or sedilia ; or in the wall recess of a 
 tomb ; on the font, the chest, the memorial brass, the window (484) ; even 
 in the pinnacle, as in HOWDEN NAVE {J 2), Lincoln nave, and Boston.* To 
 some extent there was a geographical difference in the design of the canopies 
 of niches. To the north and cast the}' were more often solid ; e.g. the 
 Percy tomb at BEVERLEY (269); the arcading of the Ely Lady Chapel ; the 
 sedilia and Easter sepulchres of Hawton, Navenby, and Heckington. In the 
 south and west light open spire-work was preferred. It was appropriate for 
 wood, and had been used all over England in the wooden canopies above stalls. 
 It was equally unsuitable for stone ; nevertheless it was greatly in favour ; e.g. 
 the sedilia of Exeter and Ottery St Mary ; the Exeter throne ; the tomb of 
 Edward II. at GLOUCESTER (294); that of Sir Hugh Despenser (1349) and 
 Sir Guy Bryan (1380) at Tewkesbury ; and the Durham reredos, which is south 
 country work ; made of Dorsetshire clunch, and shipped from London to 
 Durham + via Newcastle, in 1372-1380.* 
 
 For other characteristics of a fourteenth-century church, see Chapters XV. 
 to XLI. For arches, see 279 ; for buttresses and pinnacles, 358, 363 ; for flying 
 buttresses, i" ; for parapets and battlements, 396 ; for strings, hood-molds, drip- 
 stones, and basement courses, 406 ; for foliated capitals, 436 ; for molded capitals, 
 443; for base and plinth, 452; for window tracery, 479 ; for roots, 558: for 
 doorways, 579 ; for towers, 608 ; for spires, 617. Also see 126-134. 
 
 * Illustrated in Prior, 404. 
 
 t Greenwell's Durham, 71. The Selby sedilia are also probably of the same London 
 make. 
 
 t For a full account of the treatment of the niche see Prior, 381-404. 
 
Chapter VI. 
 CHARACTERISTICS OF ENGLISH GOTHIC 
 
 FROM C. I33O TO 1538. 
 
 Planning — Internal Elevation — East Front— West Front— Vaulting — Piers — Ornament. 
 
 PLANNING. — Only three important choirs of the greater churches were rebuilt ; 
 viz. York, commenced 1361 ; BATH (373), commenced e. 1500; both with 
 aisles of the full length of the choir ; and Christ Church, Hampshire ; where an 
 aisled choir with unaisled Lady Chapel was commenced c. 1400. None of the 
 three exhibit any novelty in planning.* 
 
 In the parish churches the normal type is that with aisled nave and unaisled 
 chancel. Some few churches, however, continued their aisles to the full length 
 of the chancel; e.g. Louth, GRESFORD (214). Others, e.g. ST NICHOLAS, 
 LYNN (214) ; North Walsham ; ST STEPHEN'S, NORWICH (228), identical in plan 
 with Louth, differed from it in omitting the chancel arch. But the cruciform 
 plan is never abandoned ; e.g. St Mary Redcliffe, Bristol ; TERRINGTON ST 
 CLEMENT (92). 
 
 INTERNAL ELEVATION. — In this period all the varieties of triforium treat- 
 ment are reduced to one. The triforium arcade, whether tall or short, disappears 
 altogether. At Malvern the triforium chamber is masked with a blank wall, as 
 in the early work of Fountains and Kirkstall. At Bath is the same arrangement ; 
 but the blank wall is less conspicuous ; for the triforium roof is so much flattened 
 that little height is left for the wall in front. Elsewhere the precedent of York 
 nave is followed. The triforium is closed from the nave by a blank wall, to the 
 bottom of which descend the mullions of the clerestory window, which are allowed 
 sometimes, as in GLOUCESTER CHOIR (59), to descend to the hood-molds of 
 the pier arcade. This mullioned wall appears in front of the triforium in 
 Gloucester choir (1337 to c. 1350); and in the last half of the same century 
 in the naves of WINCHESTER (342) and CANTERBURY (90) ; in the south tran- 
 sept of St Mary Redcliffe, Bristol ; and in York choir. In the fifteenth and 
 sixteenth centuries it is seen in the choir and nave of ST MARY REDCLIFFE 
 (525); in the choir of Christ Church, Hants ; and in Sherborne ; in ST GEORGE'S, 
 WINDSOR (330) ; and in Henry the Seventh's Chapel at Westminster. 
 
 Probably the example set in Gloucester choir had most weight in spreading 
 
 * The period c. 1330 to 1538 corresponds roughly with the Perpendicular or Rectilinear 
 period of Rickman, Bloxam, Parker, and Sharpe, except that it also includes the work at 
 ■Gloucester, between 1330 and 1360, which their chronology excludes. 
 
Sy 
 
90 
 
 Winchester Nave. 
 
 Canterbury Nave. 
 
LATE GOTHIC EXTERNAL ELEVATION. 91 
 
 this design. In the Gloucester choir such a design was almost compulsory ; it 
 was necessary to hide away the great semicircular arches of the lower and upper 
 aisles by panelling them over with the mullions of the clerestory windows f 1 3 5 ,). 
 
 When designed, as in Gloucester choir, in conjunction with vault piers, this 
 design gives one the impression, and no doubt was intended to give the impres- 
 sion, that the interior is one of a single story. Unity was the ideal of late Gothic 
 design, and nowhere was that ideal realised so completely as in the choir of 
 Gloucester. Similarly, at CHIPPING NORTON (548 j, the interior is of one story. 
 
 East FRONT. — In the fifteenth century a high Lad)' Chapel and aisled choir 
 are built at Christ Church, Hants ; and less lofty Lady Chapels at GLOUCESTER 
 (132); St Mary Redcliffe, Bristol; and Malvern; the last has been destroyed. 
 To the latter part of the fourteenth century belongs the east front of York ; 
 to the fifteenth century that of LOUTH (89); to the sixteenth century that 
 of BATH (373); in all three the aisles are as long as the choir, and the latter 
 is carried up full height. 
 
 West FRONT. — Of the towered west front there are three examples ; Brid- 
 lington, which is a patchwork of various dates ; Canterbury, of which the south- 
 western tower was Norman till the "restoration" of 1834; and BEVERLEY 
 MINSTER (599), which, with the exception of Peterborough, which is sui 
 generis, has the most successful western facade in England ; the towers are not 
 absorbed by the facade, but are towers all the way to the ground. 
 
 The parochial west front becomes more and more common in the greater 
 churches. It appears at Winchester, Malvern, Gloucester, WINDSOR (49-), 
 Bath. At Winchester and Gloucester it was even substituted for a towered 
 facade. In the parish churches, in this as in all periods, the west front is mainly 
 occupied by a western tower. Fine facades occur at Maidstone ; HULL (96) ; 
 BEVERLEY ST maky's (366) ; Yatton ; Crewkerne ; TERRINGTON ST CLEMENT'S 
 (92). 
 
 In the late Gothic facades the normal elevation is one of three stories ; 
 e.g. at Winchester, Canterbury, Beverley Minster; the third story being that 
 of the gable. But the roofs were flattened more and more ; in addition, the west 
 window might have a four-centred arch. In such a facade there would practically 
 be no gable, and the elevation would be one of two stories only ; the doorway 
 story and the window story ; e.g. Gloucester, Bath, BEVERLEY ST maky's (366), 
 HULL (96), WINDSOR (492). Even with roofs of steep pitch, the elevation 
 is sometimes of two stories only; e.g. at TERRINGTON ST CLEMENT'S (92). 
 
 In all the western facades, from first to last, there was a rivalry between the 
 central doorway and the central window. In France, by moderating the size of 
 the central west window, which was often a rose, a loftier doorway could be had 
 beneath. Still further to increase the importance of this doorway, it was often 
 surmounted with a triangular gable, which in Auxerre Cathedral is filled with 
 open tracery and allowed to rise high up in front of the window. Thus the door- 
 way becomes, as it should be, an imposing and influential member of the facade. 
 In England nothing was too precious to sacrifice to bigness of window, to floods 
 of light and acreage of stained glass.* 
 
 In Beverley Minster the west window is so tall that its head is cut off by the vaulting 
 of the nave. 
 
92 
 
LATE GOTHIC VAULTING. 
 
 93 
 
 Vai'LTINC. — It was in this period that the most magnificent of all our vaults 
 were built.* In the first place, Fan vaulting came into use ; probably its earliest 
 application being in GLOUCESTER CLOISTER ('344J ; afterwards it was em- 
 ployed in high vaults; e.g. SHERBORNE (346), KING'S COLLEGE CHAPEL 
 (62); and HENRY THE SEVENTH'S CHAPEL, WESTMINSTER (348). Lierne 
 
 vaults, however, were in even greater favour; e.g. Bristol, St Marv Redcliffe ; 
 
 Nave. 
 
 Canterbury, Black Prince's Chantry, nave, and St Michael's Chapel ; Christ 
 Church, Hampshire, choir and Lady Chapel ; ELY, Bishop West's Chapel (334) ; 
 GLOUCESTER, south transept (306), choir (334), north transept, west bays of 
 nave, and Lady Chapel; HEREFORD i'}}^), smith transept; NORWICH (330), 
 all the high vaults; OXFORD, the Divinity School (331) and the Cathedral 
 
 * One must not forget, however, the TEWKESBURY vaults (330), which are exceedingly 
 
 beautiful. 
 
94 
 
 LATE GOTHIC PIERS. 
 
 NAVE 
 
 '-/' 
 
 choir (331) ; WINCHESTER nave (342) ; and all the high vaults of ST GEORGE'S, 
 
 WINDSOR (332). 
 
 High vaults were projected at Malvern, but not carried out. Ihose oi 
 
 York are of wood. 
 
 PIERS.— Three varieties of Perpendicular piers may be distinguished. 
 1. Occasionally the cluster of columns survives ; e.g. in York choir, where the 
 design is but a fourteenth-century version of that of the nave. 2. More often 
 the columns become less prominent and the central mass more so, and some of 
 the shafts are reduced to "beads"; e.g. at CIRENCESTER (448); the nave of 
 St Mary, Oxford ; St Mary Redcliffe, Bristol ; Bath ; ST GEORGE'S, WINDSOR 
 (255); Christ Church, Hants; Malvern choir; Gloucester west nave. 3. In 
 all these cases the pier is symmetrical ; and two, four, eight or more shafts are 
 retained. But in Sherborne choir and in HENRY THE SEVENTH'S CHAPEL, 
 
 WESTMINSTER ABBEY, 
 the piers are entirely 
 unsymmetrical masses, 
 their form being wholly 
 
 ^ jm ^f ' ^ fefe regulated by their func- 
 
 ^^-~??yV; v _________._w| jF jf tions. The first step in 
 
 this direction had been 
 taken at BRISTOL 
 (661. 1 1) in 1298. 
 
 In the smaller 
 parish churches there 
 was no scope for com- 
 plexity of plan in the 
 piers. At all periods 
 they may be found 
 circular or octagonal. 
 A cluster of four 
 columns was also very 
 common ; it appearseven 
 in the sixteenth-century 
 nave of Ripon Minster. 
 Ornament. — In late Gothic design the window was all important, and 
 its tracery overspread the church ; e.g. in GLOUCESTER CHOIR (47) ; thus 
 reducing very largely the amount of foliated ornament. What foliage was 
 employed was usually of bulbous or undulatory character, and highly conven- 
 tionalised. Hard square forms or lozenges are characteristic. Square leaves 
 and four leaves arranged in a square are most common in cornices. Stone 
 diaper was abandoned ; but painted diaper occurs ; e.g. in Bishop Beckington's 
 tomb at Wells (1464). The vine and strawberry leaf were favourite forms of 
 leafage. The rose is common in late work ; e.g. KING'S COLLEGE CHAPEL 
 (473), with the portcullis of Henry VII. Shields, heraldic emblems, and 
 grotesque animals are all common. Foliated bosses are frequent in the richer 
 roofs ; e.g. Sail, Tenterden, NEW WALSINGHAM (570). A cornice of vine leaves 
 and tendrils is exceedingly common in the cornices of screens ; it is usually 
 
 Aisle of Henry the Seventh's Chapel, Westminster 
 
LATE GOTHIC ORXAMKNT. 
 
 95 
 
 crested with the Tudor flower. Angels are used in capitals and roofs; e.g. 
 in the pier arcade of St Alary Magdalene, Taunton. The symbols of the 
 
 
 
 
 
 
 Kettering Western Doorway. 
 
 Passion are frequently represented on fonts; also on the ceiling of Winchester 
 presbytery: a capital with the passion flower occurs at TIVERTON '437.'':. 
 
 After c. 1350 CROCKETS lose much of the undulating outline of Decorated 
 
96 LATE GOTHIC DETAIL. 
 
 foliage ; they are usually conventionalised, and become stiff and square ; e.g. St 
 Mary, Bury.* 
 
 For other characteristics of a late Gothic church, see Chapters VIII. to XLI. 
 For arches, see 2S0 ; for buttresses and pinnacles, 361, 364 ; for flying buttresses, 
 ITJ ; for parapets and battlements, 396, 398 ; for strings, hood- molds, dripstones, 
 and basement courses, 406; for foliated capitals, 438; for molded capitals, 4.44.; 
 for base and plinth, 453 ; for window tracery, \g\ ; for roofs, 562 ; for doorways, 
 579 ; for towers, 608 ; for spires, 622. See also 133-142. 
 
 * Illustrated in Colling's Mediaval Foliage, 56. 
 
HOLY TRINITY, HULL, WEST FRONT. 
 
Chapter VII. 
 
 A CHRONOLOGICAL DESCRIPTION OF THE CHIEF 
 ENGLISH CHURCHES. 
 
 [Note. — Except where documentary and architectural evidence coincide, the dates in tlii. 
 chapter arc to be regarded as merely conjectural approximations ; see note on page 638. For 
 references to the documentary evidence see pages 63S to 657.] 
 
 1050—1150. 
 
 XI. CENTURY: THIRD QUARTER {Edward the Confessor, Harold, William I.).— 
 
 WESTMINSTER ABBEY, begun IO5O. Lanfranc's CANTERBURY, begun I070. 
 
 XI. CENTURY: FOURTH QUARTER (William /., icth year, to William //..last 
 year). — blyth, founded 1088. bury; part finished in 1095. Ernulph's Canter- 
 bury, begun 1095. canterbury, st augustine ; castle acre, founded before 
 1089 or in 1090. chichester, begun 1091. Chester, st John's, begun 1067 to 
 1095. Chester cathedral (St Werburgh), refounded in 1093. Christ church, 
 Hampshire, begun c. 1099. Durham, begun 1093. ely, c. 1090. Gloucester, 
 
 begun I0S9. HEREFORD, begun I079-IO95. LASTINGHAM, 1078-1088. LEWES, 
 
 founded 1077. i.iNcoi \, consecrated 1092. LONDON, ST John's CHAPEL in tow i r, 
 
 C. I0S0. LONDON, OLD ST PAULS, 10S7. MAI.LING NUNNERY, 1077-II0S. MALVERN, 
 
 begun c. 1084. Norwich, begun 1096. Rochester, begun 1077-1108. st Allans, 
 begun 1077. selby, begun 101)7. Shrewsbury abbey, begun 10S3. tewkesbury, 
 choir entered in 1 102. thorney, 1085-1 10S. tutbury, founded 1080. Winchester 
 cathedral, begun 1079. Worcester, begun 1084. 
 
 XII. CENTURY: FIRST QUARTER (Henry I., 1st year to 26th year).— binham, 
 re-endowed 1101-1106. bury, gateway, 1121-1130. Carlisle, alter 1101. Col- 
 chester, st botolph, founded 1102. exeter cathedral, towers, 1 1 12-1 136. 
 leominster, consecrated 11 30. lindisfarne, partly finished before 11 28. London, 
 st Bartholomew's, begun 1123. Peterborough, begun 111701-1118. reading, 
 founded 1121. romsey, c. 1120. sherborne, begun 1107. Southwell, begun 
 1108-11 14. waltham abbey, nave, c. 1120. wymondham, founded before 1 107. 
 
 XII. CENTURY: SECOND QUARTER (Henry I., 26th year, to Stephen, 1 6th year). 
 
 — CHEPSTOW. DEVIZES, si JOHN and SI MARY, before II39. DOVER, ST MARTIN'S 
 
 priory, begun 1131-1139. Dunfermline, probably soon after 1124. new shoreham, 
 nave, c. 1 1 30. 
 
 The history of the Norman branch of Romanesque architecture in England 
 commences with the building of Westminster Abbey in 1050 by Edward the 
 Confessor. His church was of great importance to Anglo-Norman design ; for it 
 was the first example in this country of the periapsidal plan (164), derived 
 probably from St Martin de Tours, and anticipating Cluny by thirty-nine years ; 
 
 G 
 
9 8 
 
 CHURCHES OF 1050-1150. 
 
 a plan which was reproduced at Gloucester in 1089 and at Norwich in 1096. Of 
 the earliest churches after the Conquest, Lanfranc's Canterbury was but of 
 moderate dimensions, being closely modelled on the Abbaye-aux-Hommes at 
 Caen and CERISY-LA-FORET (148.3) both in plan and elevation. The choir of 
 the former was rebuilt in Gothic ; and the western bays of the nave of Cerisy 
 have been destroyed ; but from one or the other we can form a fair idea of what 
 Canterbury Cathedral was like, as rebuilt by Lanfranc* 
 
 But the Anglo-Norman was far from being a mere servile imitation of the 
 Norman Romanesque, either in plan or structure. Many of our churches were 
 on a far grander scale than the Romanesque churches of Normandy ; even such 
 early examples as BURY ST EDMUNDS (1070) (150.3), ST ALBANS (1077) (153.2), 
 Winchester (1079), Ely (1083), Old St Paul's (1087); especially remarkable was 
 the vast length of the naves of the above. Some, moreover, e.g. Winchester, Ely, 
 Old St Paul's, had western as well as eastern aisles to their transept ; a great 
 
 St Albans in the Twelfth Century. 
 
 advance on the eastern apse or apses of the transepts of Normandy. As early as 
 1096, CANTERBURY (149.2) set the example of a vast prolongation of the choir 
 also, and in addition built an eastern transept. And in due course BURY ST 
 EDMUNDS (150.3), Ely, and Peterborough provided themselves with vast and 
 complex western transepts. A still greater revolution in planning is seen at 
 Dover, Sherborne, Southwell, Ely, in which the eastern termination of the 
 choir was square; and at ROMSEY (151.3), begun before 1120, where not only 
 was the choir rectangular, but it was encircled by a rectangular ambulatory 
 projecting from which was an eastern chapel. These were the greatest inno- 
 vations in planning. In construction the primacy rests easily with Durham. 
 Durham was designed for vaults with diagonal ribs as early as 1093 ; and high 
 vaults with diagonal ribs seem to have been constructed over the whole 
 cathedral before n 33 (8). To receive the springs of these ribbed vaults piers 
 and abaci were built of logical design (659.1); and to abut the high vaults 
 
 * See interior of the abbaye-aux-HOMMES (319); plan of CERISY (148) ; exterior (160) ; 
 interior of choir (161) ; of transept (199) ; and of nave (521). 
 
CHURCHES OF 1050-1150. 
 
 99 
 
 flying buttresses were built in the triforium chamber of the nave. Tc 
 facilitate the vaulting, 
 every transverse arch of 
 the nave was pointed. 
 It would be difficult to 
 find another church in 
 Western Europe, at the 
 end of the eleventh cen- 
 tury, which had advanced 
 so far as Durham on the 
 way to Gothic* 
 
 Nevertheless it is 
 not to be supposed that 
 ever)- Anglo - Norman 
 Church advanced as far 
 as the Durham of 1093- 
 1 1 33. Even to the 
 middle of the twelfth 
 century or later Durham 
 seems to have remained 
 unsurpassed. For the 
 progress of architectural 
 art is not uniform ; it is 
 not like the steady pro- 
 gress of the steamship. 
 Rather it is as in a yacht 
 race, where first one boat 
 and then another catches 
 abreezeand forges ahead, 
 while others it may be 
 are becalmed and sta- 
 tionary. Peterborough 
 Cathedral was com- 
 menced late ; not before 
 1 117 or 11 18 ; but the 
 improvements of 
 Ernulph's Canterbury 
 and Durham are largely 
 ignored. It had the old- 
 fashioned plan with three 
 parallel eastern apses ; 
 it had neither the 
 ambulatory nor the 
 elongated choir nor the 
 
 eastern transept of Canterbury ; nor the high vaults and pointed transverse 
 of Durham. Still more retrograde is St Bartholomew's, Smithfield ; begun 
 * For durha.m see 149.1, 34, 306, 315, 8, 308, 239, 370, 28, 659.1. 
 
 Gloucester, North Aisle of Nave. 
 
 archc 
 1123 
 
ioo CHURCHES OF 1150-1175. 
 
 where there are no preparations for high vaults and where the aisle vaults are 
 without ribs. Still slower to innovate was the Anglo-Norman builder in the 
 villages ; e.g. the church of SUTTON ST MARY, Lincolnshire (42), a thoroughly 
 Romanesque design, was not commenced till after 11 So.* 
 
 Not only did English Romanesque advance at different rates ; but in distant 
 districts, dissevered by trackless forests and unbridged rivers, it tended to form 
 divergent local schools. Thus the West built its churches less vast in scale, with 
 naves considerably shorter, with less amplification of central transept, and 
 without western annexes, and exhibited a preference for the ambulatory rather 
 than the three parallel eastern apses. So also instead of the compound pier, 
 or of alternation of compound pier and cylinder, or cylinder and octagon, it 
 preferred rows of simple cylinders, short and stout, as in GLOUCESTER CHOIR 
 (294), or immensely tall, as in GLOUCESTER NAVE (26). Of these piers 
 the capitals were often no more than imposts, and the bases were of the 
 most archaic character. The recessed orders of the arches often remained 
 square-edged, with little molding or carving, if any (276). Durham, again, 
 forms a school of its own, with its connections, Lindisfarne, Warkworth, Dun- 
 fermline, Selby, and WALTHAM (521). The school, however, that claimed 
 most adherents was the South-Eastern, with its elongated naves, at NORWICH 
 (148.4), ELY (153.4), BURY (150.3), Peterborough, ST ALBANS (153.2), Old St 
 Paul's, Chichester. It may be that this elongation of the nave is clue to the 
 precedent set by CERISY (148.3). 
 
 1150-1175. 
 
 XII. CENTURY: THIRD QUARTER {Stephen, 16th year, to Henry II, 22nd 
 year). — bolton priory, begun c. 1151. brinkburn, c. 1170. buildwas, c. 1148. 
 byland, the monks entered, 1177. dunstable, nave, c. 1160. Durham, galilee, 
 c. 1 175. ELY", upper parts of west transept and infirmary, and st mary's church, 
 c. 1170. fountains, begun c. 1135. furness, after 1148. kirkstall, e. 1152. 
 lanercost, consecrated 1 169. malmesbury, probably c. 11 50. oxford cathedral, 
 1154-1180. roche, c. 1 165. Stamford, st Leonard's priory, strata Florida, 
 1 166-1203. wimborne, central tower and part of nave. Winchester, st cross, 
 c. 1160 seq. Worcester, west bays of nave, c. n 70. YORK, part of crypt, 1154- 
 
 THIS forms the early part of the period to which Mr Sharpe gave the name 
 Transitional Norman or Transitional.! It - is th e period of transition from 
 Romanesque to Gothic. By Mr Brandon it was called Semi-Norman ; by- 
 others Pointed Norman. Mr Sharpe regarded it as having lasted from c. 1145 
 
 * In this church all the walls have been raised ; and what were originally clerestory windows 
 are now openings looking into the aisle. 
 
 + Owing to lack of documentary evidence as to the date of many of the churches it has 
 been found impossible to arrange and discuss them in strict chronological sequence. They 
 have been arranged, therefore, in this chapter in periods of twenty-five years. 
 
CHURCHES OF 1150-1175. 101 
 
 to c. 1 190. It is characterised, he says, by the simultaneous use in the same 
 building of semicircular and pointed arches. But here again there were retro- 
 gressive builders, who admitted no pointed arches at all into their churches ; 
 e.g. Dunstable nave and OXFORD CATHEDRAL CHOIR (27J ; the latter is 
 1 1 54-1 180. Even so late as 1180 the Cathedral of St David's was designed 
 with all its pier arches semicircular. More often, however, to facilitate the 
 
 Fountains Xave from S.E 
 
 vaulting of the aisles (322), the arches of the pier arcade are pointed. These 
 pointed pier arches are at first very obtuse ; e.g. in fountains nave and in 
 Furness, Kirkstall, Buildwas, all Cistercian; MALMESBURY (522), Benedictine; 
 and the Hospital Church of St Cross, Winchester. More acutely pointed, but 
 covered with Romanesque ornament, are the west transepts of Ely and Peter- 
 borough, the latter probably 1 177-1 193. Still more advanced towards Gothic are 
 
102 
 
 Ripon Choir. 
 
CHURCHES OF 1150-1175. 103 
 
 Brinkburn, Lanercost, RIPON * (102), Roche, and Byland ; though they are all 
 without high vaults. In Durham galileef the arches are semicircular and 
 covered with chevron ; and there is no vault ; but the design is so light and 
 graceful that it has more of the Gothic in it than the Romanesque. The most 
 advanced of all are St Cross, Winchester, probably not begun before 1160, and 
 the Cistercian abbey of Roche. Both had high vaults, which at St Cross still 
 remain. In other respects St Cross is thoroughly Romanesque, relying for 
 stability entirely on immense thickness of wall and pier ; it has neither flying 
 buttresses nor transverse arches in the triforium chamber.^ Indeed St Cross is 
 less advanced than the nave of Durham ; the chief difference being that at 
 St Cross the pointed arch is employed in the arches of the crossing ami 
 the pier arcade, and in the wall ribs as well as in the transverse arches of the 
 vault. 
 
 In the Cistercian churches more progress is made. A distinct tendency 
 is seen to buttress rather than to thicken the walls. § But, in accordance 
 with Burgundian tradition, there was a distrust in these abbeys of the 
 flying buttress, which therefore remained undeveloped. The drainage of 
 the walls was improved by heightening the corbel-table, so as to form a 
 parapet masking a gutter behind (385); and by amplifying the basement 
 course, as at Kirkstall and FOUNTAINS (679.1). Owing to the injunctions of 
 the founders of the Cistercian Order and especially of St Bernard, sculptured 
 ornament was discouraged ; one result of which was to increase the employ- 
 ment of moldings. For the compound pier, cylinder, or octagon a cluster of 
 columns was often substituted, as at ROCHE (661.2). Scalloped, coniferous, 
 and water-leaf capitals and corbels were especially common in the Cistercian 
 churches. Masonry improved most of all, the Cistercians laying great stress 
 on sound construction, and often working at the masonry with their own 
 hands. The triforium was almost always walled in, and the clerestory passage 
 was infrequent. Stone towers and bells were forbidden by the statutes of 
 the Chapter-General. The walls were left plain ; not covered with arcading. 
 Corbels were used wherever possible instead of vaulting shafts or roofing shafts. 
 There was an almost total absence of colour, whether in pictures, wall-paintings, 
 mosaic pavements, or glass. Cistercian architecture may be fairly described 
 as a combination of ascetic ardour, temperate good sense, straightforward 
 procedure, and practical utility. || 
 
 None of the Cistercian churches were of the vast scale of Bury, Lewes, or 
 Old St Paul's. Instead of the western transept they had occasionally a small 
 lean-to western porch ; they had no long choir or eastern transept ; nor had any 
 central transept a western aisle. On the eastern side the transept, as at KIRK- 
 STALL (152.4), had an aisle divided into chapels. The presbytery was usually 
 
 * The greater part of the work of Lanercost, Roche, Byland, Ripon, and the Transitional 
 choir of York was probably done after 1 170, and belongs rather to the period 1 175-1200. 
 
 + Originally the piers of the Durham galilee consisted of but two marble shafts. 
 
 + Section in Dehio, Plate 148. 
 
 S See plan of Kirkstall (152) ; and buttress of Kirkstall chapter house (359). 
 
 || On Cistercian architecture see Dehio, i., book ii., c. xiii. ; and Anthyme St Paul in 
 Enlart's Gothic in Italy, 224-228. 
 
io4 
 
 CHURCHES OF 1150-1175. 
 
 short and without aisles, and it was usually rectangular. Byland had also a 
 rectangular ambulatory, as, later on, had Dore. 
 
 Such an unaisled rectangular presbytery as that of Cistercian Kirkstall was 
 of course a complete breaking away* from the traditions of Anglo-Norman 
 planning, whether with three parallel eastern apses or with semicircular ambula- 
 tory. But others beside the Cistercians were innovating in planning. At 
 OXFORD (152.3) the Augustinian Canons built an aisled choir and unaisled 
 rectangular presbytery. At ST CROSS, WINCHESTER (215.8), a further step 
 was taken ; the rectangular presbytery being aisled as well as the choir. 
 
 In one point all the three new types of plan, those of Kirkstall, Oxford, and 
 
 St Cross, Winchester, from S.E. 
 
 St Cross, agreed ; their presbyteries were all rectangular. Through the influ- 
 ence of these plans, especially of those of the numerous Cistercian churches 
 built at this time, the apsidal presbytery of the Continent, with rare exceptions, 
 disappeared from English architecture. The English became differentiated from 
 the Continental presbytery by being square-ended. 
 
 One more innovation of the utmost importance was made at ST CROSS. 
 This was that the roof of the presbytery was continued to its eastern termina- 
 tion in undiminished height. At St Cross was reached the plan and eastern 
 
 * It was of course a reproduction of the simplest type of Burgundian plan ; probably that 
 of the Clairvaux Church of St Bernard. 
 
CHURCHES OF i 175-1200. 105 
 
 termination which remained in fashion till the very end of English Gothic archi- 
 tecture, till York Minster and Bath Abbey. 
 
 On the whole, the third quarter of the twelfth century was an epoch fertile 
 in change and improvement, except as regards the important matter of vaulting; 
 and for much of the improvement the builders of the new Cistercian abbevs 
 may fairly claim the credit. Their influence was greatest where their abbeys 
 were most numerous, viz., in the North of England. 
 
 1175—1200. 
 
 XII. CENTURY: FOURTH (QUARTER (Henry II, 22nd year; Richard I. . to 
 John, 2nd year). — bishop AUCKLAND, hall, c. 1190. canterbury, choir, 11 75-1 178. 
 saint's chapel and corona, 1179-1184. cartmel, founded iiSS. chichester, 
 retrochoir, &c, 11S6-1199. darlington church, c. 1192. deeping, st james, 
 c. 1180. dore, choir, c. 1190. Dublin, chkist church, after 1171. Glaston- 
 bury, lady chapel, dedicated 11S6: choir of abbey church, commenced 11 84. 
 
 HARTLEPOOL, ('. Il88. HEREFORD, east tranSt.pt, I 186-I 199. JEDBURGH, C. I I 75 — 
 
 c. 1 190. jervaulx, c. i 170 — c. 1 190. Lincoln, choir and eastern transept, begun 
 1 192. llandaff, c. i 1 90. Llanidloes, work of c. 1 1 80 from Cwm Hir. LLAN- 
 thonv. London, nave of temple church, consecrated 1 185. NEW shoreham, 
 choir, c. 1175 — c. 1210. oakham, hall of castle, 1165-1191. old malton, 
 c. 1 180. Peterborough, clerestory of nave, west bays of nave, and west transept, 
 1177-1193. st david's, begun 1180. selby, parts of west nave, west front, and 
 porch, st radegund's, 1191. Shrewsbury, st mary's, nave, c. 1180. wells, 
 
 H74-II9r. WENLOCK, C. IHJO. WIT1IAM, II76-I186. 
 
 The last quarter of the twelfth, like the last quarter of the eleventh century, was 
 a momentous period in English mediaeval architecture ; the latter completed the 
 structural development of English Romanesque, the former that of English 
 Gothic. The former is usually assumed to commence with the building 
 of St Hugh's choir at Lincoln in 1192. Really, however, the first complete 
 Gothic of England commences with the choir not of Lincoln, but of Wells, as 
 begun by Reginald Fitz Bohun, who was Bishop from 1174 to 1191. 
 
 As in our Romanesque, so in our early Gothic, three distinct schools may 
 be recognised : the Western, the Northern, and the Southern. The claims of 
 the Western school have only recently been recognised. In reality not only was 
 it the first to start, but its geographical extension was the most considerable of 
 all, and its output the greatest. In England, Whitchurch Canonicorum, William, 
 Glastonbury, WELLS (209;, DORE CHOIR 1S2), the eastern transept of Here- 
 ford, the western bays of Worcester nave, Wenlock, the nave of ST MARY'S, 
 SHREWSBURY '424, Lilleshall, and the original Gothic choirs of LICHFIELD 
 f244;and Chester; in Wales, Llandaff and Cwm Hir; in Ireland, Christ Church, 
 Dublin, all belonged to this school. It was in the West of England that the art of 
 Gothic vaulting was first mastered ; first, so far as we know, at Worcester ; and it 
 was in the West, first apparently at Wells, that every arch was pointed ami the 
 semicircular arch was exterminated. At the neighbouring abbey church of 
 
io6 
 
 CHURCHES OF 1175-1200. 
 
 Canterbury CI 
 
 Glastonbury, begun in 1 184, the 
 vaults of the Lady Chapel were 
 thoroughly Gothic in character. At 
 Glastonbury and Dore choirs were 
 planned with rectangular ambula- 
 tories, but without the eastern 
 Lady Chapel of Romsey. The 
 Western sculptors were far ahead 
 of the rest of England ; at Wells 
 the craftsman's hand can be seen 
 gaining in cunning, capital by 
 capital, till foliated capitals and 
 scrolls of conventional foliage were 
 produced that remained unsur- 
 passed to the last days of English 
 Gothic* Of this work the earliest 
 is probably that at Worcester. It 
 is earlier in character than the 
 dated work either at Wells or 
 Glastonbury ; and can hardly be 
 placed later, therefore, than c. 
 1 I70.f 
 
 In the desolate regions of 
 Northern England the output was 
 smaller. Byland was completed, 
 or nearly so ; it was complex in 
 plan, but had no high vault. The 
 greatest progress is to be seen in 
 the Cistercian abbey church of 
 Roche, which may have been 
 commenced c. 1 165. It seems to 
 have been vaulted throughout ; 
 and alone of the northern churches 
 compares with the advanced archi- 
 tecture of St Cross, Winchester, 
 New Shoreham, j Wells, and 
 Glastonbury. The works in Selby 
 nave slowly advanced. To this 
 period probably belongs the com- 
 pletion of the choir of York 
 
 * See 412.6, 424.1.2.3. 
 
 + For the Western pier, see 245 ; 
 for the arch-molds, 279 ; for the capitals, 
 422, 412, 424. 
 
 1 Certain resemblances between 
 New Shoreham and Hartlepool are 
 pointed out by Rev. J. F. Hodgson in 
 Arch. Aeliana, xvii. 201. 
 
10S CHURCHES OF 1175-1200. 
 
 Minster (rebuilt in the fourteenth century), and of the choir and transept of 
 Ripon. Jervaulx built a new church planned like that of St Cross. Important 
 churches were erected at Hartlepool and Darlington ; to the same school belong 
 the churches of Holy Trinity, Micklegate, York ; and Nun Monkton. Possibly 
 Hexham choir was commenced. 
 
 In Southern England little was done that can be called Gothic ; but this 
 is of great historic importance. NEW SHOREHAM CHOIR (373), c. 1175, is 
 the first in the South of England to exhibit the St Cross plan on a large scale ; 
 internally, however, in spite of a pointed pier arcade and well-molded arches, its 
 ground story is of massive and Romanesque character. The great work in the 
 South of England was the rebuilding of CANTERBURY CHOIR (149.3) after tne 
 fire of 1 1 74. The architect selected by the monks was a Frenchman, William 
 of Sens ; and he gave them a French design : one modelled to a large extent 
 on that of his own cathedral at SENS* (107). Here then we have a disturbing 
 factor of the first magnitude in the steady development of Anglo-Norman 
 architecture, and it becomes important to consider what was precisely the extent 
 of the Continental innovations introduced by William of Sens at Canterbury. 
 
 As regards the plan, the circular chapel of the Holy Trinity f is directly 
 copied from Sens Cathedral. The internal elevation of the choir, with its tall 
 pier arcade and low triforium, is reminiscent of Sens. The vaulting is sexpartite, 
 as at Sens. Norman sexpartite vaulting exists in the chancel of Tickencote, 
 Rutland, but that of Canterbury is probably copied from Sens. The vaulting 
 shafts, both at CANTERBURY and SENS (106, 107), are alternately massive 
 and slender. Most of the vaulting shafts rest on the abaci at Canterbury, 
 as do the more slender shafts at Sens. In both churches the transverse 
 arches of the aisle vault are semicircular, and are much broader than the 
 diagonals ; whereas at Worcester and Wells they are pointed, and differ little 
 in dimensions. The side cells of the high vault at Canterbury are round arched, 
 as originally were those at Sens. % Many of the pointed pier arches, e.g. in the 
 Chapel of St Thomas and in the crypt, are much stilted ; also they have plain 
 rectangular soffits, as in the twelfth century Gothic of France. The absence or 
 insignificance of the hood-mold over these arches is also a French characteristic. 
 Piers composed simply of a couple of columns put side by side are very rare 
 elsewhere, but are found in Sens choir and the Chapel of St Thomas, Canter- 
 bury. The magnificent Corinthian and Composite capitals (428) are French ; 
 so also are the crocket capitals of the Chapel of St Thomas. The lancet 
 windows are much less slender than the normal lancets of England. § The 
 great circular windows of the eastern transept, undivided except by iron 
 bars, resemble those of Notre Dame de Dijon. || The buttresses have immense 
 
 * Sens Cathedral is commonly said to have been commenced in 1140; but little of the 
 existing church appears to belong to this period. The main structure of the choir is probably 
 that which was consecrated in 1 167, and can be but little anterior to Notre Dame, Paris, 
 commenced 1 163. 
 
 t It was built to enshrine the crown {corona) of the skull of St Thomas. 
 
 % Scott's Lectures, i. 94, 96. 
 
 § Broad lancets occur at Wells and Glastonbury ; but not earlier than those of Canterbury. 
 
 || Illustrated in Viollet-le-Duc, Architecture, iv. 132. 
 
CHURCHES OF 1175-1200. 109 
 
 projection in comparison with their English predecessors or contemporaries. 
 For the first time the flying buttresses emerge from beneath the aisle roofs 
 into the open air, and are of light French construction ; very unlike those 
 built soon after at Chichester and New Shoreham. The French had a long 
 start in Gothic architecture ; St Denis, Noyon, Notre Dame, Paris, St Martin 
 des Champs, St Germer, as well as Sens, were all well advanced before 
 Canterbury choir was commenced. No wonder that it is so reminiscent of 
 the advanced architectural art of Northern France. 
 
 The next important work in the South of England was the building of the 
 retrochoir of CHICHESTER (34.4) and the vaulting of the whole church after 
 the fire of 11S7. The plan of Chichester, with rectangular ambulator}- and 
 projecting rectangular Lady Chapel, and its vaulting of the highly advanced 
 character of that of Worcester and Wells, clearly connect it with the Gothic of 
 the West of England, and not with Canterbury. Of all the French features in 
 Canterbury choir enumerated above, hardly one reappears at Chichester, unless 
 it be the crocket capitals (245) proportioned in depth to the diameter of the 
 shafts or columns. 
 
 The last and greatest work of the period is that of St Hugh at LINCOLN 
 (151. 1); viz. the apse, which has been removed; the north-eastern and south- 
 eastern transepts, with their western adjuncts ; the choir ; and a single bay of 
 the eastern aisle of the great transept adjoining either side of the choir.* The 
 design of St Hugh's architect is full of originality and even of eccentricity. 
 Hut it is impossible to deny that it is largely indebted to the new work at 
 Canterbury, finished in 11S4. Both plans include an eastern transept 
 (149.3, 66_) ; both these transepts have to the east two pairs of semicircular 
 apses (a survival of Romanesque). In both the vault springs at the mid height 
 of the triforium. In both distrust of the flying buttress is shown by the 
 construction of pointed arches spanning the triforium chamber (34). Both at 
 Lincoln and in the Chapel of St Thomas at Canterbury intermediate buttresses 
 are inserted in the centre of each bay between each pair of lancet windows. In 
 both the circular molded abacus is found ; at Canterbury in the crypt ; in 
 Lincoln almost universally. Romanesque billet occurs in the ribs of the vault of 
 the Chapel of St Thomas, Canterbury, and in the south-eastern transept of 
 Lincoln. Marble shafting is used profusely in both churches. The corner 
 piers of St Hugh's transepts closely resemble those in the same position at 
 CANTERBURY (523;. The design of the choir piers of Lincoln, each faced with 
 a single vaulting shaft descending to the pavement,+ occurs sporadically both in 
 the choir and on the east side of the eastern transepts of Canterbury. J The 
 proportioning of the depth of the capitals of the pier arcade to their supports 
 appears at Lincoln and Chichester as well as at Canterbury. The light flying 
 buttresses, displayed in the open air, are reminiscent of Canterbury (112). 
 The buttresses have much projection. It is plain that the obligation of the 
 Lincoln to the Canterbury design is great. It is equally plain that the 
 
 The apses of the eastern transepts and the remainder of the great transept were probably 
 taken in hand c. 1205. 
 
 + The lower parts of the vaulting shafts were cut away when the stalls were inserted. 
 % See in ; and Britton's Canterbury, Plates 11 and 19. 
 
no 
 
 Scale: of Feet 
 
 Lincoln Choir. 
 
1 1 1 
 
 m,M<Ji 
 
 I5H8S 
 
 »* 
 
 ftvWWWW^ 
 
 5c«.e of Feet 
 Lincoln Choir. 
 
I 12 
 
 CHURCHES OF 1 175-1200. 
 
 obligation is almost wholly to the English and not to the French part of 
 that design. For of all the features enumerated above, in which the designs 
 of the Lincoln and Canterbury choirs agree, only the last two, the designs of 
 the buttresses and flying buttresses,* are characteristically French ; they form 
 much too slender a basis for theorising as to the French origin of the design 
 of St Hugh's work at Lincoln. Even were such an origin granted for Lincoln, 
 it would not hold for the early Gothic of the North of England ; still less for the 
 vast amount of work done in the western counties, some of it done before 
 William of Sens reached our shores. Viollet-le-Duc, who studied St Hugh's 
 
 Lincoln from S.E. 
 
 work on the spot, says : " After the most careful examination I could not find 
 in any part of the cathedral of Lincoln, neither in the general design, nor in 
 
 * It is a curious fact that Lincoln choir seems to have been designed with less reference to 
 Canterbury than we now see it. In the opinion of Sir G. G. Scott neither the intermediate 
 buttresses, nor the flying buttresses, nor the transverse arches in the triforium formed part of 
 the original design. But they must have been added very soon afterwards ; when the high 
 vault of the choir was built. The rib-molds of the vault of the central span of the choir 
 are similar to those of the undoubtedly original ones of the sexpartite vault of the minor 
 transepts, except that the tooth ornament is omitted ; so that there can be little difference of 
 date between the high vault of the choir and that of the minor transepts. See Assoc. Soc. 
 Reports, xii. 190 note and 191. 
 
CHURCHES OF 1200-1225. 113 
 
 any part of the system of architecture adopted, nor in the details of ornament, 
 any trace of the French school of the twelfth century, so plainly characteristic 
 of the cathedrals of Paris, Noyon, Senlis, Chartres, Sens, or even Rouen. On 
 the exterior the choir is thoroughly English. . . . The vaults have not at all 
 the same construction as the French vaults of the end of the twelfth century. 
 Arch-moldings slender and deeply undercut, abacus round, the tooth ornament 
 — do not at all resemble the ornaments which we find at Paris, Sens, St Denis, 
 &c. . . . Nowhere in France do we find between 1 190 and 1200 pillars similar 
 to the corner pillars of the eastern transepts of LINCOLN (249J, with the 
 crockets placed between the shafts ; nowhere in France do we find crockets 
 carved like these ; nowhere shafts with hexagonal concave section ; nowhere 
 capitals or abacus similar to those of these pillars. . . . The construction is 
 English ; the profiles of the moldings are English ; the ornaments are English ; 
 the execution of the work belongs to the English school of workmen of the 
 beginning of the thirteenth century." 
 
 1200 1225. 
 
 XIII. CENTURY: FIRST QUARTER {John, 2nd year, to Henry III., 10th year).— 
 bolton priory, aisle and west front. ELY, galilee, 1198-1215. fountains, choir, 
 1208 — c. mo. hexham, choir, c. 1180 — c. 1210; transepts, c. 1215 — c, 1230. 
 Lichfield, choir, finished in 121 1 ; south transept, c. 1220. Lincoln minster, 
 great transept and nave. Peterborough, west front, rievaulx, transept rebuilt, 
 c. 1 2 10. Rochester, presbytery and east transept, begun c. 1200; choir finished, 
 1227. romsey, west nave and west front, c. 1220. ST albans, John de Cella's 
 porches, 1195-1205; Trumpington's work, 1214-1235. st saviour's, southwark, 
 1213-1238. Salisbury, begun 1220. selby, upper nave, c. 1190 — c. 1220. 
 tynemouth, choir, c. 1200. valle crucis, founded c. 1200. wells, west front, 
 1218-1239. whitby, choir and transept. Winchester cathedral, retrochoir 
 begun c. 1202. Worcester, retrochoir, 1202-121S. 
 
 The building of Lincoln choir, 1 192-1200, was followed by a vast output of the 
 new Gothic. In some cases, however, the work was of retrograde character; 
 the ceilings of Anglo-Norman work being yet retained ; as in Hexham, 
 Tynemouth, and WHITBY choirs (114), and the western bays of the nave of 
 ST albans (14). In the last the piers were of massive and Romanesque 
 character. At ST SAVIOUR'S, SOUTHWARK (521), also, there is little of Gothic 
 economy of material ; the piers are short and heavy, the arches low and the 
 walls thick, as in the school of Wells. Nor was the flying buttress, employed at 
 Canterbury, Chichester, and Lincoln, received with much favour. Hexham, 
 Tynemouth, Whitby, dispensing with high vaults, did not require it. Rochester 
 choir was designed throughout so as to dispense with it. At Worcester and 
 Salisbury* it was hidden beneath the aisle roofs, as it had been at Wells. 
 Only in the transept and nave of LINCOLN fii2, 115), NEW SHOREHAM, 
 
 * Those now seen at Salisbury are later additions. 
 11 
 
114 
 
 Whitby Choir. 
 
■15 
 
 Lincoln Nave. 
 
u6 CHURCHES OF 1225-1250. 
 
 Sussex (373), and perhaps in Rievaulx transept and Fountains choir, is it 
 employed. The most striking feature is the sudden decline of the vigorous 
 school of the West, which in the preceding period had had by far the largest 
 output The works at Wells, Glastonbury, Lichfield, and Chester choirs, and 
 Christ Church, Dublin, were completed or continued ; but no new work was 
 commenced wholly in the Western style. The dominant influence now was 
 probably that of Lincoln choir. The most important advances were those made 
 in vaulting in LINCOLN NAVE (327). In the North, the most important 
 work was the rebuilding of the transept and choir of Fountains and Rievaulx. 
 In the South, the rectangular ambulatory was expanded into a retrochoir at 
 Winchester; St Saviour's, Southwark ; and SALISBURY (154.2) ; and in addition 
 was carried in undiminished height eastwards at Rochester and Worcester. 
 The St Cross and New Shoreham type of choir was adopted in the North at 
 Hexham and Whitby. In this period also were designed the facades of St 
 Albans, PETERBOROUGH (112), Wells,* Bolton, Romsey ; and the galilee of 
 Ely Cathedral. f The most important work of all was SALISBURY CATHEDRAL 
 (170), commenced in the same year as Amiens, 1220. 
 
 1225—1250. 
 
 XIII. CENTURY: SECOND QUARTER {Henry III., 10th to 3.5th year).— beverley 
 minster, choir and transepts, c. 1225 — c. 1245. binham, west front, 1226-1244. 
 Bristol, Elder Lady Chapel, boxgrove, c. 1235. Carlisle, choir. Dublin, 
 christ church, nave finished in 1235. Dublin, st Patrick's. Dunstable, west 
 front. Durham, east transept, 1242 — c. 1280. ely, presbytery, 1235-1252. exeter, 
 chapterhouse, 1224-1244. fountains, eastern transept, c. 1 220-1247. Gloucester, 
 vault of east nave, 1243. Hereford, Lady Chapel, lichfield, north transept, 
 chapter house and vestibule. Lincoln, nave and chapter house. London, Temple 
 choir finished 1240. netley, begun 1239. oxford, chapter house, rievaulx, 
 choir, ripon, west front, c. 1233. Rochester, north transept, c. 1240-1255. 
 Southwell, choir begun before 1233. Worcester, choir begun 1224. york, 
 south transept, 1230 — c. 1241. 
 
 The disastrous reign of John (1199-1216), during part of which the kingdom 
 had been under an Interdict (1207-1211), had doubtless greatly checked the 
 advance of the new Gothic. On his death architectural activity revived, and 
 the reign of his successor, Henry III., was productive of an enormous amount 
 of work ; but on the whole marking but little advance on that of the preceding 
 period, or even on that of Lincoln choir. By this time the vault entered into 
 
 * Mr W. H. St John Hope points out that the commencement of the west front of Wells 
 may be fixed by the grant of sixty great oaks in 1220, and of thirty more in 1225 ; in each case 
 ad rogum faciendum, i.e., to make a lime-kiln ; the first thing done when great works were to 
 be commenced. 
 
 t The Ely galilee is assigned by Sir G. G. Scott {Lectures, i. 127) to Bishop Eustace, 
 1195-1214. Professor Willis considered this date too early. 
 
U7 
 
 
 
 
 Ely Presbyter)-. 
 
i iS CHURCHES OF 1250-1275. 
 
 the design of nearly all the greater churches. The chief exceptions were Carlisle 
 choir (burnt in 1 292) ; Lichfield transept (not vaulted till the fifteenth century) ; 
 and YORK TRANSEPT (523) ; where there are the springers of a stone vault, 
 which, however, was constructed in wood. The external flying buttress, how- 
 ever, of Lincoln choir made few converts ; only in BOXGROVE (373), LINCOLN 
 NAVE (115) and BEVERLEY MINSTER (176) is it external. These are the three 
 most advanced works of the period ; LINCOLN NAVE (327) far ahead of 
 all in its rib-system; BEVERLEY CHOIR (51) remarkable for the high spring 
 of its vaults; BOXGROVE quite unique (318). During this period the works at 
 Lincoln probably included the completion of the nave, chapter house, galilee, 
 and part of the west front ; Rochester built its north transept ; Worcester and 
 Rievaulx their choirs; ELY its presbytery (117); Fountains its eastern tran- 
 sept; Christ Church, Dublin, its nave ; Salisbury was in a fair way to comple- 
 tion ; and inspired the design for St Patrick, Dublin ; Durham commenced its 
 eastern transept ; the choir of the TEMPLE CHURCH, LONDON (462), was finished 
 in 1240; the Cistercian church of NETLEY (471) was commenced in 1239; 
 and the western facades of LINCOLN (562), Ripon, BINHAM (471), Dunstable 
 were erected wholly or in part. 
 
 The plans of the choirs of SALISBURY (154.2) and ELY (153.4) represent the 
 two types which remained in use at all subsequent periods. A variant of the 
 former is that of SOUTHWELL (175). The only innovation, which was not 
 repeated, was at FOUNTAINS (150.2) and Durham, where the eastern transept 
 was built at the east end, and not athwart the choir. The most striking change 
 to the eye is that traceried windows commence ; at Binham, Netley, Old St 
 Paul's, and WESTMINSTER (65). 
 
 What is specially remarkable about the architecture of the first half of the 
 thirteenth century is the perfection of the work even in the smallest village 
 churches in some districts, e.g. Northants and South Lincolnshire ; the same 
 artistic fitness of design, the same faultless execution and delicacy of treatment 
 is to be seen in them as in the largest cathedral and the noblest conventual 
 church.* 
 
 1250-1275. 
 
 XIII. CENTURY: THIRD QUARTER (Henry III., 35th year, to Edward I., 4 th 
 year). — chichester, St Edmund's Chapel, 1245-1253. exeter, eastern chapels 
 and retrochoir commenced, c. 1270. Hereford, choir clerestory, c. 1250; north 
 transept, c. 1260. Lincoln, presbytery, 1 256-1 280. lichfield, nave, st albans, 
 choir clerestory begun 1257 ; retrochoir, 1260-1326. Salisbury, upper portions, 
 finished 1266; chapter house and cloister, 1263-1284. tintern, parts, 1269-1288. 
 Westminster abbey, choir, transept, and four eastern bays of nave, 1 245-1 269; 
 chapter house finished, c. 1253. 
 
 Partly because of the disturbed state of the kingdom, partly, perhaps, because 
 the last quarter of the century had been so productive, there remains a much 
 smaller amount of work done between 1250 and 1275 : but what there is, is of the 
 Sharpens Nene Valley, 13 ; and Lincoln Excursion, 50. 
 
U9 
 
 ..... _,i„im«i.-m'! -i ' u^immmtmmmt 
 
CHURCHES OF 1250-1275. 121 
 
 very highest importance. SALISBURY (170) nave and transepts were com- 
 pleted. The choir, transept, and four bays of the nave of Westminster (63) 
 and the nave of LICHFIELD (523) were commenced (Westminster in 1245) ; the 
 latter as characteristically English, as the former is French. St Albans set to 
 work to transform slowly its choir into Gothic and to build a retrochoir. A 
 new Cistercian church was commenced in 1269 at TINTERN (524). HERE- 
 FORD (523) built a north transept, not without reminiscences of the new tran- 
 sept of Durham ; LINCOLN (56) pulled down its apse, and built a magnificent 
 presbytery to enshrine St Hugh. The rebuilding of the whole cathedral of 
 Exeter was commenced at the east end c. 1270. An advance in construction 
 is seen in the fact that ever)' one of these is vaulted, except the retrochoir 
 of St Albans ; where, however, there are preparations for a vault and for 
 flying buttresses. Moreover, except in Cistercian churches, there were flying 
 buttresses, and these were no longer concealed. At this period, perhaps, 
 the art of molding reached its highest level ; the moldings of Durham eastern 
 transept, completed c. 12S0, and of LINCOLN PRESBYTERY (669.2), consecrated 
 in 12S0, are unsurpassed in beauty and refinement. But by far the greatest 
 advance is the recognition to a very large degree of the principle of skeleton 
 construction (55); that the vaults may be supported with safety upon piers, 
 and all the wall between the buttresses suppressed and replaced by glass. 
 
 WESTMINSTER ABBEY CHURCH C 1 5 1.2), like Canterbury choir, is note- 
 worthy as an importation into England of French Gothic, that of the He de 
 France, Picardy, and Champagne.* Once more the apse, ambulatory, and radiating 
 chapels of Gloucester and Norwich appear on English soil, but with polygonal 
 instead of semicircular forms. The eastern limb is short ; whereas at Ely, 
 Lincoln, Old St Paul's, it grew to enormous length. The vault soars to a height 
 of 103 feet ; far above that of our vaulted cathedrals ; the interior is three times 
 as high as it is broad, a ratio common in France, exceptional in England ; the 
 internal elevation is divided in normal French fashion into six parts, of which 
 three are assigned to the ground story, one to the triforium, two to the clerestory; 
 the clerestory is without wall-passage ; the great height of the clerestory is 
 quite non-English at this period ; the geometrical tracery of the windows is 
 of French type. Externally, on the south side (379), there are intermediate 
 piers, and flying buttresses of two flights and superposed in tiers ; the pin- 
 nacle appears, to weight the buttresses ; the magnificent facade of the north 
 transept is wholly French. But, as at Canterbury, the Englishman also had his 
 say ; the transept is of far projection ; the vault has English ribbing, and is 
 not filled in after the normal French fashion ; the back wall of the triforium 
 has windows, and its arcade is open to the rafters ; the capitals are molded ; 
 the abacus is not square on plan, and its lip is molded ; the foliated scrolls 
 are almost wholly of English workmanship; the polygonal chapter house 
 is unknown in France. In Westminster are blended all that is best in French 
 and all that is best in English Gothic. It is noteworthy, that of all the French 
 features that have been enumerated above, not one, except the pinnacle,! 
 
 * " Reims Cathedral is undoubtedly the prototype of Westminster Abbey, which shows 
 evidence of close study of the French Coronation-Church" (Lethaby's Med. Art. 208). 
 t Which, however, we had long employed to weight the angles of walls and towers. 
 
122 CHURCHES OF 1275-1300. 
 
 survived. Otherwise English art went on unmoved, uninfluenced ; almost as if 
 Westminster did not exist. We English are afflicted with a natural incapacity 
 for learning from other people. 
 
 12 75-1300. 
 
 XIII. CENTURY : FOURTH QUARTER {Edward I., 4th to 29th year).— Bridling- 
 ton, nave. Bristol, tower of St Mary Redcliffe, 1292. Carlisle, choir remodelled 
 after fire of 1292. Chichester, Lady Chapel of Cathedral and Chapel of St Mary's 
 Hospital, 128S-1304. Eleanor Crosses, 1291-1294. exeter, Quivil's work, 1280- 
 129 1. guisborough, choir. howden, transepts and nave, kirkham, gateway, 
 1296. Lincoln, cloister, south side, 1296. lichfield, west front. London, 
 St Etheldreda's Chapel, 1290-1300. Norwich, St Ethelbert's gateway, after 1273. 
 oxford, Merton College Chapel. pershore, repair of choir after fire of 1288. 
 ripon, east front, c. 1 288-1300. Rochester, south transept, from c. 1280. st 
 albans, Lady Chapel, 1291-1326. selby, choir aisles. Southwell, chapter house. 
 Thornton chapter house begun 1282. wells, hall of Bishop's Palace, 1280-1292. 
 york, st mary's, 1276-1295. york minster, nave, 1291-1345. 
 
 The work left of this date is not very considerable ; much has been lost. 
 Somewhat retrograde examples are the Rochester south transept, and PERSHORE 
 CHOIR (75), as repaired after the fire of 1288. The works at St Albans 
 went on slowly in the retrochoir and Lady Chapel. The new choir commenced 
 at SELBY (86), advanced slowly. HOWDEN (546) becoming collegiate, built 
 a short unaisled choir, which was rebuilt in the following century ; also a new 
 transept and nave. Guisborough Priory built a new choir late in the period. 
 What was perhaps the masterpiece of the time, St Mary's, York, was built between 
 1276 and 1295. Southwell built its chapter house, with a highly complex vault. 
 YORK NAVE (10) was commenced in 1291 ; BRIDLINGTON (124) completed 
 its nave, on the south side lowering the aisle roof to obtain a loftier clere- 
 story (124). The rebuilding of EXETER (9) was advanced by Bishop Ouivil, 
 1280-1291. 
 
 In the interiors, the pier with detached shafts is almost extinct ; the foliage 
 of capitals, corbels, crockets, scrolls, becomes realistic (434) ; in the tracery 
 combinations of circles make room for other geometrical forms, as in MERTON 
 COLLEGE CHAPEL (473); the triforium dwindles; at BRIDLINGTON (125), 
 York, and Guisborough the clerestory increases greatly in height. Pinnacles 
 appear on the buttresses at EXETER (377) and York. The rib system of LIN- 
 COLN NAVE (324.4) reaches a still higher degree of complexity at EXETER 
 (324.7). A great change commences in the parish churches ; their aisles 
 had been low and the naves dark ; at HEDON (544) a tall pier arcade is 
 built, and is well lighted by the new traceried windows of the aisles ; in its 
 clerestory also are small traceried windows. 
 
12' 
 
 Wells Chapter House. 
 
124 
 
125 
 
i2 6 CHURCHES OF 1300-1350. 
 
 1300—1325. 
 
 XIV. CENTURY: FIRST QUARTER {Edward I., 29th year, to Edward II., 19th 
 year).— boston, tower, begun 1309. Bristol, cathedral choir, 1298-1332. canter- 
 bury, St Augustine's gateway, 1309. exeter, choir, 1291-1307. Gloucester, 
 south' aisle of nave, 1318-1329. lichfield, Lady Chapel, begun 1310. milton 
 abbas, choir and south transept. Norwich, chapel of charnel house, 13 10-1325. 
 oxford, St Mary's spire. ST albans, shrine, 1302-130S. wells, chapter house, c. 
 1310. winchelsea church, c. 1310. Worcester, east bays of north nave, 1317- 
 1324. york, chapter house, c. 1300. 
 
 In this period Worcester rebuilt much of the Norman part of its nave; 
 GLOUCESTER (360) the south aisle of its nave ; Milton Abbas its choir and 
 south transept. Lichfield Lady Chapel and WELLS CHAPTER HOUSE (123) 
 were nearly completed. York also built a chapter house.* York nave and 
 Exeter choir were finished. By far the most progressive work is to be found 
 in the choir and transept of 15RISTOL (329), where signs of the coming revolu- 
 tion in English architecture may already be detected. 
 
 The greatest achievement of the period is its Heme vaulting, which seems 
 to have originated in the polygonal chapter houses ; e.g. of Lincoln, York, 
 Wells. 
 
 The noble churches of Orford and Winchelsea appear to belong to this 
 period, and, like Bristol choir, are of highly advanced type. 
 
 1325—1350. 
 
 XIV. CENTURY: SECOND QUARTER (Edward II, 19th year, to Edward III, 
 24th year). — battle, gateway, 1339. beverley minster, nave, c. 1320-1349. 
 bolton priory, choir. Bristol, St Mary Redcliffe, south transept, bury, gateway, 
 1327. Carlisle, east front, cartmel, east aisle of presbytery. Chester cathe- 
 dral, south transept, ely, Lady Chapel, choir, and Prior Crauden's Chapel, begun 
 1321; octagon, begun 1323. exeter, nave, 1308-1350. Glastonbury, choir 
 prolonged, 1 341-1374. Gloucester, south transept, 1331-1337; choir, 1337 — c. 1350. 
 howden, choir. hull, chancel. lichfield, presbytery and south-west spire. 
 malmesbury, remodelling of nave, nantwich, chancel, 1327-1333. ottery st 
 mary, begun 1337. oxford cathedral, Latin Chapel, before 1355. patrington. 
 Peterborough, south-west spire. Salisbury, spire, sf.lby, choir, tewkesbury, 
 choir remodelled, wells cathedral, Lady Chapel, presbytery and retrochoir ; 
 central tower heightened. Worcester, nave and tower completed. 
 
 ENGLAND'S greatness and prosperity under Edward III. are reflected in her 
 architecture. The first twenty-four years of his reign were marked by one of the 
 most momentous outbursts of building activity in the records of English history. 
 
 * It may be that York chapter house belongs to the preceding period. 
 
CHURCHES OF 1325-1350. 127 
 
 The fourth quarters of the eleventh and twelfth centuries ; the second quarters of 
 the thirteenth and fourteenth centuries are the great building periods of our 
 mediaeval architecture. In the last, 1 325-] 350, were laid, broad and deep, the 
 
 Wells Presbytery from West. 
 
 foundations of all our later Gothic. At this period too the art of architectural 
 composition reached a level that never afterwards was surpassed. The lateral 
 elevation of SELBY CHOIR (86), the poetic facade of CARLISLE (128), the 
 
128 CHURCHES OF 1325-1350. 
 
 solid dignity of the east front of Selby, the grouping of the masses of 
 Ottery St Mary and PATRINGTON (612), the tower composition of ST MICHAEL, 
 COVENTRY (635), show how consummate was fourteenth-century design. Above 
 all, there was a delight in erecting great spires ; the south-west spires of 
 PETERBOROUGH (112) and LICHFIELD {Frontispiece) ; those of SALISBURY (170), 
 Grantham, St Mary, Stamford, KETTON (621), Shottesbrooke, Snettisham, 
 Heckington, and many another, belong to this prolific twenty-five years. 
 
 Of the greater work, the naves of EXETER (9) and Beverley Minster 
 are assimilated to early work ; and the choir of SELBY (86) is the com- 
 pletion of work laid out in the thirteenth century. At MALMESBURY (375) 
 the nave, at TEWKESBURY (165) the choir are remodelled in the style 
 of the period. The great southern transept of Chester Cathedral was com- 
 menced in this period, but not completed till later. Carlisle and Glastonbury 
 both slightly lengthened their eastern limbs. Lichfield built a new presbytery 
 and largely remodelled its choir ; but the work was greatly damaged in the 
 sieges of 1643. The two great examples of the rich, decorated work of the 
 period left to us are the retrochoir and presbytery of WELLS (127) and 
 the Lady Chapel, octagon, and choir of ELY (522). But here again, both 
 in Ely choir and Wells presbytery, the design was largely fettered by the fact 
 that it had to be in harmony with earlier work with which it came into juxta- 
 position to the east or west ; we have not a single interior of the first rank in 
 which the designer had a free hand. This it may be that has so long 
 disguised the essential unity of the Late Decorated style of Edward the 
 Third's reign in England and the great later style of France, the so-called 
 Flamboyant, which is nothing but our English Decorated carried to its logical 
 issue. Yet just as surely as the Romanesque architecture of the eleventh and 
 twelfth centuries of England is the imported and naturalised Romanesque of 
 Normandy, so surely the Flamboyant architecture of the fifteenth and sixteenth 
 centuries in France is the imported and naturalised Late Decorated architecture 
 of England.* Our architectural debt to William of Normandy we repaid during 
 our long occupation of France in the fourteenth century by the loan of our 
 Late Decorated architecture. That century to much of France was a period of 
 the utmost misery. Just when our own Gothic was winning its most brilliant 
 triumphs at Hull, BOSTON (222), Howden, Selby, Beverley, Carlisle, Lich- 
 field, Wells, Ely, France had to pass through the ordeal of the English Wars, 
 La Jacquerie, the Black Death, the alienation of the Duchy of Burgundy, and 
 the troubles in Flanders.j- For a whole hundred years the English Wars went 
 on. French Gothic architecture was practically annihilated. Her thirteenth- 
 century style ceased to be ; and died without an heir. In many districts of 
 France hardly a single important church was erected in the fourteenth century. % 
 
 * It is not intended to deny the considerable extent of Flemish influence in French Flam- 
 boyant ; e.°: in the church of Brou-en-Bresse ; and it must be admitted that, whatever the extent 
 of the foreign elements admitted by France into her later Gothic, she gave them a development 
 all her own, characterised by native harmony and elegance. 
 
 t Gonse, 261. 
 
 " II faut dire que nous n'avons pas en France un seul grand edifice complet d'architecture 
 religieuse du quatorzieme siecle " (Viollet-le-Duc, Architecture, iv. 207, note). 
 
CARLISLE CATHEDRAL. EAST FRONT. 
 
CHURCHES OF 1325-1350. 
 
 129 
 
 The history of the period is written plain in many a French cathedral ; Sens, 
 Keauvais, Limoges, Auxerre, Amiens, Troyes, Senlis, Seez ; where the choir and 
 nave are of the twelfth or thirteenth century, the transepts of the fifteenth. 
 
 Boston, South Doorway of Nave. 
 
 There is a great interval of time between these two building periods ; an interval 
 so long that there was time for the great traditions of French craftsmanship 
 largely to die out ; so that when the foreigner at last was expelled from her soil, 
 
 I 
 
130 
 
 Prior '(rauden's (Ixipel CJy- 1*»^» 
 
 of "IKe ^ocil! ? -^u le - 
 
CHURCHES OF 1325-1350. 
 
 131 
 
 
 ill ■ 
 
 
 and France again was free, she 
 had to make copious drafts on 
 the art of England and of 
 Flanders.* Then once more, 
 but not till the fifteenth cen- 
 tury ,f with the aid of England 
 and Flanders, she suddenly 
 commenced to build in Flam- 
 boyant fashion. The com- 
 plicated vaults of GLOUCESTER 
 (334) w 'ith ridge ribs, tier- 
 cerons and hemes ; the absorp- 
 tion of the triforium at Lich- 
 field and WELLS (127); the 
 molded piers of Bristol choir ; 
 the logical responds of GLOU- 
 CESTER SOUTH TRANSEPT 
 (495) ; the bulbous, undula- 
 tory foliage of the capitals, 
 corbels, crockets, finials, span- 
 drels of ELY LADY CHAPEL 
 (269) ; the flowing tracery 
 of MERTON COLLEGE VES- 
 TRY (480) ; Prior Crauden's 
 Chapel at ELY (130); the 
 
 BEVERLEY REREDOS (486) 
 and the BOSTON DOORS 
 (129); the ogee hood-molds 
 of doorway and window, the 
 presence of the ogee arch and 
 the ogee curve in leafage and 
 molding — everywhere curve 
 echoed by countercurve* — all 
 this was welcomed in France, 
 and received a yet more grace- 
 ful French dress. Strongest of 
 all, as was to be expected from 
 its origin, was the new Flam- 
 boyant in the North of France, e.g. PONT AUDEMER, and Flanders, where 
 it flourished amazingly, and in turn sent forth missionaries to Antwerp, 
 
 \f 
 
 M A T 
 
 * On the obligations of France to Flemish art, see Enlart's Manuel, 587. 
 
 t With the solitary exception of a chapel of 1373 in Amiens Cathedral. Enlart's 
 Manuel, i. 588. 
 
 * " There is no part of the ornamental portions of Heckington Church, the contours or pro- 
 files of which do not exhibit the ogee or wave form, both in outline and section" (Sharpe's 
 Lincoln Excursion, 81). So Enlart, Manuel, 464, says that Flamboyant is a style "011 les 
 ondulations sont multipliers systematiquement." 
 
i 3 2 CHURCHES OF 1325-1350. 
 
 Dordrecht, Milan, Cintra, and Luz* A highly specialised late school of French 
 Gothic is that of Brittany, which, architecturally, would seem during the fifteenth 
 century to have been influenced by Western England.f Up to the present the 
 influence of England on French Flamboyant has been but sparingly admitted. 
 M. Anthyme St Paul indeed recognises that part of St Severin, Paris, is in the 
 English style.} But it has been reserved for M. Camille Enlart § to state 
 definitely the dependence of the fifteenth-century architecture of France on that 
 of the fourteenth century of England. 
 
 " II a ete dit, tome premier, pages 586 et suivantes, que le style flamboyant s'est 
 introduit en France pendant la guerre de Cent ans. On peut dire plus. Ce style est un 
 produit de l'occupation anglaise ; en effet, les traces en accolade, les remplages a soufflets 
 et mouchettes, les crochets de feuillages extremement frises, et les chapiteaux compris comme 
 des frises, existent en Angleterre des le commencement du quatorzieme siecle ; la grande 
 fenetre de facade de la cathedrale d'York, la chapelle de la Vierge d'Ely, a. Beverley le 
 tombeau (of Lady Eleanor Fitz-Alan) presentent ces particularites ; le vestiaire de Merton 
 College a Oxford a des fenetres a remplages flamboyants ; et, des le treizieme siecle les 
 cathedrales de Durham (transept oriental) et de Lincoln (nef) montrent des traces 
 de vofites, qui chez nous sont propres au quinzieme siecle ; la voCite a Hemes et a tier- 
 cerons et la voute a tiercerons sans les ogives. Au quinzieme siecle l'arcbitecture 
 anglaise s'ecarta du style flamboyant ; mais elle en avait fourni depuis cent ans et plus 
 tous les elements a la France ; qui de ces elements a compose un style un peu different, 
 mais dont l'origine n'est pas douteuse pour qui considere les dates, le nombre et 
 l'importance des emprunts, et l'epoque de creation du style, qui est precisement 
 celle de l'occupation anglaise. Avant done qu'une guerre heureuse nous donnat le 
 style de la Renaissance, une guerre malheureuse nous avait valu le style qu'il devait 
 remplacer." 
 
 In the period 1325-1350 a second leading feature was the glorification of the 
 parish church. At all periods indeed noble parish churches had been built ; e.g. 
 in the twelfth century the churches of Northampton St Peter's, Whaplode, St 
 Margaret at Cliffe, Hartlepool, Warmington ; in the thirteenth century West 
 Walton, Skelton, Stone; but it was the exception to build a church de novo ; 
 seldom did the resources of the parish allow more than to build annexes to the 
 original building. But under Edward III. people were prosperous ; the farmers 
 got a good price for their wool, the weavers and merchants for their cloth. Now 
 it became more common than it had ever been before to rebuild the whole 
 church, leaving not a trace of the old. In whole districts, especially in Lincoln- 
 shire, the churches were rebuilt at this time. And very noble churches they are ; 
 lofty, spacious, and spired ; yielding no whit in composition or in beauty of detail 
 to abbey church, collegiate or cathedral. Such are Heckington, Pembridge, 
 Boston, HOLBEACH, patrington (opposite), Cley, Snettisham, Hingham, Tides- 
 well, Shottesbrooke, Nantwich ; the south transept of St Mary Redcliffe, Bristol, 
 
 * Enlart's Manuel, 587, note. 
 
 t Prior, Gothic Art, 332 ; who instances St Pol, Quinrper, Treguier, Folgoet, Lamballe, 
 and the Kreizker. So also Choisy, Histoire, ii. 508, admits that at St Pol de Leon "l'imitation 
 est evidente." 
 
 % Histoire Mo?iumentale, 20S. 
 
 § In page 12 of the Introduction to the second volume of his Manuel dArcheolone, 
 1904. 
 
CHURCHES OF 1325-1350. 
 
 133 
 
 and the chancel of Holy Trinity, Hull* Two of the most advanced in type are the 
 great town churches of Boston and Hull. In both the pier arcades t are tall and 
 stately, as if they were cathedrals ; and owing to excellence of masonry are built 
 with the minimum of material. BOSTON (222) leads the way in the insertion 
 of double windows in each bay of the clerestory. HULL (81) is the first 
 realisation of the stone-lantern type of church which was to be the ideal of 
 English Gothic for two whole centuries.^ Of village churches, in spaciousness and 
 height, and in beaut)- of proportion, the noble church of SNETTISHAM (481) 
 is almost unrivalled. Xor is the minor work in the parish churches any less 
 beautiful. El)-, Beverley, St David's, have nothing more consummate in 
 design to show than the Sedilia and Easter Sepulchres of the village churches 
 
 Patrington. 
 
 Holbeach. 
 
 of Heckington, Navenby, and Hawton. A curious feature about many of the 
 finest churches, e.g. Heckington, is that the interior is bare and barn-like, except 
 for the fittings ; whereas on the exterior the utmost richness of detail is lavished.§ 
 But the history of this great period does not end here. As has been pointed 
 out by M. Enlart above, though between 1325 and 1349 we had worked out a 
 
 * The churches of this period are admirably illustrated in Bowman and Crowther's 
 Churches of tlie Middle Ages. 
 
 t On the rise of the pier arcades see Sharpe's Lincoln Excursion, 131. 
 
 % Just as Hull was the most advanced, so Patrington, with low massive piers and with 
 clerestory omitted, was one of the most retrogressive churches of this period. A comparison of 
 these two neighbouring churches shows the fallacy of supposing that dissimilarity of design 
 necessarily argues difference of date. 
 
 $ Cf. Sharpe's Lincoln Excursion, 80. 
 
i 3 4 CHURCHES OF 1325-1350. 
 
 style so beautiful that it was adopted by France, and remained in vogue there 
 till the very last days of French Gothic, nevertheless in England we had 
 hardly developed it when we threw it aside for something else. This was the 
 architecture of GLOUCESTER ; there elaborated in south transept (495) 
 and choir, between c. 1331 and c. 1350, when it was adopted at Winchester, 
 Canterbury, and York, speedily overran England, superseded Late Decorated 
 design, and became our one and only style till the very last days of English 
 Gothic architecture. This is the famous Perpendicular or Rectilinear style. 
 
 The supreme importance of Gloucester in the history of the later Gothic has 
 never been adequately recognised. She turned the current of English architec- 
 ture in a wholly new direction. But for Gloucester English Decorated work might 
 well have developed into a Flamboyant as rich and fanciful as that of France. 
 But to the remotest corners of the land, to cathedral, abbey church, collegiate and 
 parish church, there was brought the influence of Gloucester by the countless 
 pilgrims to the shrine of Edward the Second in her choir. In the first place, she 
 set the greater churches the fashion, which had long prevailed in the parish 
 churches, of remodelling rather than rebuilding. At GLOUCESTER (135) Norman 
 choir and transept were not pulled down to the ground to be rebuilt ; but only 
 the clerestory. A similar treatment, more or less conservative, was adopted in 
 Malmesbury nave, Tewkesbury choir, Winchester nave, Malvern, Norwich 
 choir, Sherborne, OXFORD CATHEDRAL (27). Throughout England the new 
 stained glass of GLOUCESTER (47) was adopted. Throughout England big 
 aisle and end windows were inserted, and tall clerestories were erected to hold it. 
 Everywhere the tracery of the windows became, as at Gloucester, rectilinear ; 
 and the whole church was brought into harmony by spreading rectilinear panel- 
 ling over wall, buttress, battlement, tower, even over the fan vaulting. Gloucester 
 taught SHERBORNE (576) how to dispense with flying buttresses in her 
 vaulted nave. GLOUCESTER exhibited the uttermost exuberance of heme 
 vaulting (496). Gloucester transept brought every rib of the vaulting into 
 the organism of the pier. GLOUCESTER CHOIR (59) converted clerestory 
 wall into clerestory piers; and on this skeleton construction poised her lierne 
 vaults midway amid the painted glass — an audacity of construction which to 
 the very end of English Gothic remained unrivalled. All these great things 
 were wrought in Gloucester by the middle of the fourteenth century.* 
 
 1350-1375. 
 
 XIV. CENTURY : THIRD QUARTER {Edward III., 24th to 49th year).— canter- 
 bury, Black Prince's Chantry, 1370-1379. edington church, 1352-1361. Glou- 
 cester, north transept, 1368-1373. Norwich, clerestory of choir, after 1361. 
 Westminster, west nave, 1350-1422. Winchester, presbytery, Edington's work, 
 I 34S-i3 ( 56. york, presbytery, 1361-1370. 
 
 IN 1349-50 the prosperity of England and the brilliant art of Ely and Gloucester 
 received a sudden check on the advent of the Black Death or Asiatic plague. 
 * On the superiority of the Gloucester mason, see Willis 1 Vaulting, 57. 
 

 
 
 
136 CHURCHES OF 1350-1400. 
 
 Unfinished work was everywhere stopped ; and for a whole generation little 
 new work was begun. And when the country had somewhat recovered from 
 the decimation of the population, then there began the long miserable Wars 
 of the Roses. Not till Tudor days were there assured peace and prosperity again 
 in England. There was no such halcyon period for architecture as the first half 
 of Edward the Third's reign till Henry the Seventh assumed the crown in 1485. 
 Yet in this long period — 1349 to 1485 — much was done ; but far more in the 
 parish and town churches than in cathedral or abbey. Cathedral and abbey 
 were grand enough already, and there were quite enough cathedrals and 
 abbeys. To the monks' church especially the hearts of men had grown cold ; 
 it was to found collegiate churches, as at Higham Ferrers and Maidstone ; 
 colleges, as at Winchester and New College, Oxford; parish churches and chantry 
 chapels, that religious enthusiasm turned in the later days. Just as in the last, 
 so in the present and subsequent period, the wealth that accrued in any particular 
 district, especially in the wool trade, took monumental expression in new, lofty, 
 and spacious parish churches ; above all, in Norfolk, Suffolk, and Somerset ; of 
 which however, as a rule, not more than the chancel was completed till the 
 fifteenth century. 
 
 At WESTMINSTER (65) the work of building the nave was resumed. 
 Gloucester finished remodelling her north transept. Her next great triumph 
 was the invention of fan vaulting, employed in the eastern walk of her cloister 
 (344). At Edington Church in Wiltshire the new Gothic is seen perhaps 
 for the first time outside Gloucester (1352-1361). It is employed in the 
 presbytery of YORK (199); the design of which is based on that of the nave. 
 At Canterbury is built the Black Prince's Chantry Chapel, 1370- 1379. 
 
 1375-1400. 
 
 XIV. CENTURY : FOURTH QUARTER {Edward III., 49th year, to Richard II, 
 23rd year). — beverley, west front and towers, c. 1380— c. 1430 ; canterbury, nave, 
 c - T 379 — c - 14°° ; cloister, 1397-1412. Coventry, st Michael's, tower finished 1394. 
 ely porta, 1397. Gloucester, cloister, 1351-1412. howden, chapter house, 1380- 
 1407. north walsham, after 13S1. oxford, new college, 1380-1386. st 
 albans, gatehouse, 1349-1396. Thornton, gatehouse, 1382-1388. wells, south- 
 western tower, after 1386. Westminster hall, 1397-1399. Winchester college, 
 r 387-i393- Winchester, cathedral nave, 1371— c. 1460. wymondham, central 
 tower, 1390-1409. york, choir, c. 1380 — c. 1400. 
 
 THE chief works of this period were the remodelling of Winchester nave and 
 the rebuilding of York choir. The great churches of North Walsham, Maid- 
 stone All Saints', Warwick, and TERRINGTON ST CLEMENT'S (92) were 
 built. In two churches, Etchingham and Wymington, there are brasses stating 
 that the founder of the former church died in 1387, of the latter in 1391. 
 
137 
 
 Howden Chapter House. 
 
i 3 8 CHURCHES OF 1400-1450. 
 
 Yet both churches are almost wholly of the style of 1325-1350; proving that 
 not every one was willing to accept at once the new style of Gloucester. 
 At North Walsham both the east end, and the porch, which was built before 
 1405* are largely of the earlier style. 
 
 1400-1425. 
 
 XV. CENTURY : FIRST QUARTER [Henry IV, 1st year; Henry V. : to Henry VI, 
 4th year). — carusle, north transept, 1401. Christ church, Hampshire, Lady 
 Chapel, c. 1400. Gloucester, south porch, west nave and west front, 1421-1437. 
 
 LYNN, ST NICHOLAS, 1413-1418. MANCHESTER CATHEDRAL, after 1422. OXFORD, 
 
 merton, transept, finished 1424. Winchester, chapel in cloister, 1420. york, 
 central tower, c. 1400 — c. 1423. 
 
 THE work at Gloucester still went on. She finished the fan vaulting of the 
 cloister, built a new west front and south porch, and remodelled the two 
 western bays of the nave. Christ Church, Hampshire, starts to rebuild the 
 whole of the eastern limb, commencing with the Lady Chapel. The noble 
 parish church of Bury St Mary and the Chapel of ST NICHOLAS, LYNN (214), 
 are fifteenth-century versions of those of Boston and Holbeach. Fotheringhay is 
 a thoroughgoing specimen of the lantern type of church ; roofed in wood, but 
 with clerestory piers strengthened by flying buttresses. 
 
 1425-1450. 
 
 XV. CENTURY : SECOND QUARTER {Henry VI, 4 th to 29th year).— Bristol, st 
 mary redcliffe. Cambridge, King's College Chapel is commenced in 1446. 
 canterbury, St Michael's Chapel, finished 1439. crowland, north-western tower. 
 
 ETON COLLEGE CHAPEL, begun I44I. OXFORD, ALL SOULS', I438-I442. SHER- 
 BORNE, choir, 1436-1459. wells, north-western tower, after 1424. wimborne, 
 western tower, 1448-1464. wymondham, north aisle, 1432-1445. york, south- 
 western tower, 1433-1447. 
 
 At Canterbury St Michael's or Warrior's Chapel is built with a lierne vault 
 as tangled as that of Gloucester. SHERBORNE remodelled her choir ; cover- 
 ing it with fan vaults (376). The two Royal Chapels of Eton College 
 and of KING'S COLLEGE, CAMBRIDGE (62), were commenced ; but neither 
 made much headway in these troubled times. Blythburgh and Tatters- 
 hall are characteristic parish churches. St Mary Redcliffe, Bristol, carries on 
 throughout the transformation commenced in the south transept ; with its 
 cruciform plan, and nave, transept, choir, and aisles vaulted throughout, it has 
 a cathedral-like air very unusual in an English parish church. 
 
 * Norfolk A.S., v. 341. 
 
139 
 
 jth from N.E. 
 
CHURCHES OF 1450-1525. 141 
 
 1450-1475. 
 
 XV CENTURY: THIRD QUARTER (Henry VI, 29th year, to Edward IV, 15th 
 year). — Bristol cathedral, central tower, c. 1450 — c. 1470. canterbury, Lady- 
 Chapel, 1448-1455. Durham, central tower, c. 1470. GLOUCESTER, central tower, 
 1450-1460 ; Lady Chapel, 1457-1498. malvern, choir begun c. 1450. oxford, 
 Divinity School, 1445-14S0. YORK, north-western tower, 1470-1474. 
 
 GLOUCESTER (132) erects its central tower and commences its Lady Chapel ; 
 a veritable glass-house ; Malvern choir is remodelled ; at Canterbury is built 
 the Lady Chapel or Dean's Chapel with fan vaulting; OXFORD elaborates 
 lierne vaulting yet further in the Divinity School (331). 
 
 1475-1500. 
 
 XV CENTURY: FOURTH QUARTER (Edward IV, 15th year; Edward V. ; 
 Richard III. to Henry VII., 16th year). — ei.y, Alcock's Chapel, 148S. fountains, 
 
 tower, 1494-1526. OXFORD CATHEDRAL, choir vault. 1478-1503. PETERBOROUGH, 
 
 eastern chapels, 1438-1471, and 1496-152S. shereorne, nave, 1475-1504. Win- 
 chester, Lady Chapel, 1487-1500. Windsor, St George's, 1481-1537. 
 
 SHERBORNE NAVE (346) and OXFORD CATHEDRAL CHOIR (27) are re- 
 modelled ; the former covered with fan vaulting, the latter with lierne vaults. 
 The eastern chapels of Peterborough and that of Bishop Alcock at Ely are 
 completed ; both with fan vaults. Another Royal Chapel, ST GEORGE'S, 
 WINDSOR (330), is begun. Rotherham, LONG MELFORD (547), and Fairford 
 are characteristic parish churches. Long Melford has the lantern type 
 developed to the uttermost limit. Fairford retains practically the whole of 
 its original painted glass. Painted glass has by this time passed wholly away 
 from the Gloucester type ; is heavily painted, enamelled and opaque, and betrays 
 Flemish influence. 
 
 1500—1525. 
 
 XVI. CENTURY: FIRST QUARTER (Henry VII, 16th year, to Henry VIII, 17th 
 year). — bath, 1500-1616. bolton priory, west front, begun 1520. Cambridge, 
 King's College Chapel, 150S-1515, begun 1446. Cambridge, Trinity College 
 gateway, 15 18-1535. Canterbury, Christ Church gateway, 1 5 1 7. Hereford, 
 north porch, c. 1520. i.ouni, spire, 1501-1515. oxford, Magdalen tower, 
 finished 1505. oxford, Corpus Christi College, 1516-1520. ripon, aisles of 
 nave, begun 1502 or 1503. Rochester, Lady Chapel, c. 15 12. Westminster, 
 Henry the Seventh's Chapel, 1500-1512. Winchester, presbytery, 1500-1528. 
 
 WITH the Tudors came peace, wealth, and prosperity, and a richness of 
 architectural detail that vies even with the Flamboyant of France. BATH 
 
i 4 2 CHURCHES OF 1525-1650. 
 
 (373) showed its confidence in the stability of the new regime by com- 
 mencing to rebuild its abbey church ; Bolton Priory began a new western facade. 
 The Ripon Canons added aisles to their nave. Rochester built a Lady Chapel. 
 The Royal Chapel of KING'S COLLEGE, CAMBRIDGE (62) was now taken in 
 hand in earnest, and was brought to completion in 1508-15 15. The most 
 gorgeous of all Royal Chapels, except perhaps that of Batalha in Portugal, was 
 built at Westminster, as the Chantry Chapel of Henry VII. Magnificent naves 
 were erected in the churches of Cirencester and St Mary's, Oxford ; LOUTH 
 (139) built its spire; to this period belong most of the towers of Somerset,* 
 e.g. St Mary Magdalen, TAUNTON (595). 
 
 1525—1550. 
 
 XVI. CENTURY: SECOND QUARTER {Henry VIII., 17th year, to Edward VI., 
 5th year). — bangor, nave, 1532. ely, West's Chapel, 1534. wymondham, south 
 aisle, 1534. 
 
 But the days of Gothic were numbered ; the storm was about to burst ; not 
 only the old religion but the old art of England were to succumb to its fury ; 
 Catholicism was to yield to Protestantism ; Gothic to Renaissance art. So little 
 more Gothic was done. Lavenham built a magnificent Gothic porch c. 1529; 
 at ELY (143) was built Bishop West's Chantry Chapel, precursor of the 
 Renaissance.! The greater monasteries were dissolved in 1538. 
 
 Seventeenth Century. 
 
 XVII. CENTURY. — Leeds, St John, consecrated 1634. low ham, before 1624. 
 oxford, wadham college, 1610-1613; fan vault of staircase of christ church 
 hall, 1640. stanton Harold church, Leicestershire, 1653. 
 
 Here and there Gothic lingered on ; especially in " that home of lost causes, 
 and forsaken beliefs, and unpopular names, and impossible loyalties," the 
 University of Oxford. It is astonishing to find the exquisite Gothic design of 
 the fan-vaulted staircase of the hall of CHRIST CHURCH, OXFORD (348), so late 
 as 1640. 
 
 * On the distinctive characteristics and artistic qualities of Tudor work see Scott's 
 Essay, 176, 186: "I regard the Tudor style as the most original and able thing that the 
 English have achieved in art. It was really the discovery of new and quite unlooked-for capa- 
 bility in pointed architecture." See also Freeman's History of Architecture, 394, 395 ; and 
 Statham's Architecture for General Readers, 317. 
 
 + The mixture of Gothic and Renaissance, which is well seen in Bishop West's Chapel, is 
 illustrated by Mr Gotch in chapter ii. of Early Renaissance in Architecture. For similar 
 work in France, where it is much more abundant, see Enlart's Manuel, vol. i., chapter vi. 
 
'43 
 
 Ely, Bishop West's Chapel. 
 


 PART II. 
 
 AN ANALYSIS OF THE MEDIAEVAL CHURCH ARCHITECTURE 
 
 OF ENGLAND. 
 
 Chapter VIII. 
 
 Early Christian Basilicas — Apsidal v. Rectangular Choirs. 
 
 EARLY CHRISTIAN BASILICAS. — Both in Eastern and Western Christendom, 
 the history of architecture begins in Rome ; in architecture, as in law, all roads 
 lead to Rome. In quite early days, before the recognition of Christianity by 
 the State, Christian churches were built in Rome, at any rate in the suburbs. 
 Eusebius tells us* that in the year 260 a.i>. the Emperor Gallienus ordered the 
 restitution to the Christians of churches at Rome, already forty in number. 
 None of these survive. Two venerable monuments, however, of the third century 
 survive, at least in part. It was a custom at Rome, both with Pagans and 
 Christians, to erect a small building, which goes by the name of Schola, over 
 the graves of the members of a burial club or of persons of wealth. In this little 
 lodge-room commemorative banquets were held. The practice grew common ; 
 indeed, too common. In A.D. 384 St Augustine complains that it had become a 
 practice to go drinking from schola to schola, " honouring martyrs." Of these 
 schoke two remain ; one above the catacombs of Soter, the other above those of 
 Callixtus. It was in the latter that the Bishop of Rome, Sixtus II., was murdered 
 by a mob in 25S ; when the building was razed nearly to the ground. In the 
 year 320 it was restored by Constantine, who added a vaulted roof and facade. 
 Then it became a church, and was dedicated jointly to Bishop Sixtus and the 
 martyred maiden, Cacilia, who lies buried in a chamber of the catacomb beneath. 
 It was long a place of pilgrimage ; but ultimately shared the neglect into which 
 the catacombs fell generally, and fifty years ago was a wine-cellar. Now once 
 more, since 18S2, it is a church ; the Church of Sixtus and Ca^cilia.f 
 
 When, however, the Emperor himself, Constantine, became a Christian, 
 Christianity had no need any longer to lurk in the back streets and suburbs, and 
 to build exteriors as unobtrusive as possible, such as are to this day those of the 
 
 * Ecclesiastical History, vii. 13. A church at Nicomedia was destroyed in the persecution of 
 Diocletian ; it must therefore have been built before the end of the third century. Lethaby's 
 Medieval Art, id. 
 
 t See Lanciani's Pagan and Christian Rome : and Baldwin Brown's From Schola to 
 Cathedral. 
 
 K 
 
i 4 6 EARLY CHRISTIAN BASILICAS. 
 
 Coptic churches in the Mohammedan environment of Cairo ; and accommoda- 
 tion was provided on a vast scale for the new cult. At first Pagan buildings 
 were largely utilised. Every Pagan building which was capable of giving 
 shelter to a congregation was transformed, at some time or other, into a church. 
 Smaller edifices, like temples and mausoleums, were adapted bodily to their new 
 office ; while the larger ones, such as thermae, theatres, circuses, and barracks, 
 were occupied in part only.* S. Adriano was the Senate House of Diocletian. 
 S. Andrea was the Basilica of Junius Bassus. And several of the smaller 
 temples became churches : e.g. — 
 
 SS. Cosmo e Damiano Temple of Sacra (Jrbs. 
 
 SS. Sergius e Bacchus „ Concord. 
 
 S. Maria in Cosmedin - „ Ceres. 
 
 S. Nicola in Carcere „ Piety. 
 
 S. Stefano Rotondo „ Mater Matuta.f 
 
 S. Lorenzo e Damaso, built in 370 in the stable-yard of the Factio Prasina. j 
 
 Such buildings, however, must from the first have been inconvenient for the 
 ritual of the new religion ; and churches, many of them of great dimensions, were 
 built in the reign of the first Christian Emperor, Constantine, A.D. 312 to 337; 
 and during the course of the fourth century. Several survive, at any rate in 
 part ; repaired indeed and remodelled again and again ; but still sufficiently 
 intact to show what was the plan of the first great Christian churches. To this 
 period belong basilicas at Bethlehem (part of the nave) ; and at Orleansville, in 
 North Africa. To the latter, which is in ruins, is attributed the date A.D. 325. § 
 At Rome were built churches vast in scale. Of these the five most important 
 were S. John Lateran, Omnium urbis et orbis Eccksiarum Mater et Caput ; S. 
 Peter ; S. MARIA MAGGIORE (148.2) ; S. Paul extra muros ; and S. Lorenzo extra 
 muros. The above form the five patriarchal basilicas. || The old basilica of S. 
 Peter was removed at the end of the fifteenth century to make room for the 
 present Renaissance cathedral. The basilica of S. Paul was enlarged and its 
 orientation was reversed in 388. It was burnt down in 1823, but has been 
 restored mainly on the original lines. S. Maria Maggiore was built 352-356. 
 Constantine helped with his own hands in digging the foundations of S. John 
 Lateran. He is said to have founded S. Lorenzo c. 330 ; it may be a century 
 later. 
 
 These and man)' other great basilicas^ in Rome were familiar to all Christian 
 
 * Lanciani's Pagan and Christian Rome, 160. 
 
 + Or it may have been a Macellum, i.e. a market hall. 
 
 % Lanciani's Ancient Rome. See also Eusebius, Hist. Ecdes., 4, 24 ; and Sozomen, Hist. 
 Eccles., 7, 15. 
 
 § Plan in Fergusson, 1, 510. Cattaneo, 82, holds that the sculpture of the altar belongs to 
 the second half of the seventh century. 
 
 || "Paulus, Virgo, Petrus, Laurentius atque Johannes : 
 Hi patriarchatus nomen in Urbe tenent." 
 
 11 Sta. Sabina, built about 430, is probably the most complete and least altered of the early 
 basilicas. 
 
'47 
 
 Old St Peter's, Rome. 
 
148 
 
 SCALE OF FCETT 
 
 1. Torcello. 
 
 2. S. Maria Maggiore, Rome. 
 
 3. Cerisy. 
 
 4. Norwich. 
 
149 
 
 
 sc^le: of fcet 
 
 i. Durham, f. 1180. 2. Canterbury in 1130. 
 
 3. Canterbury in 1 184. 
 
150 
 
 * *♦♦ 
 
 1 1 1 I I I 1 1 I 14 
 
 *»»**•»♦*« ♦♦♦ 
 
 # • a • «^J ♦♦♦♦♦♦•♦♦♦♦ ♦♦♦ ♦♦ 
 
 ^ 
 
 _3l 
 
 Q oooo®®@>o 
 
 B 
 
 4=iJ 
 
 00000000O00 
 
 
 SSALE OF FEET 
 
 I. Cluny. 
 
 :. Fountains. 
 
 3. Bury St Edmunds. 
 
151 
 
 
 *-T*<^ 
 
 © 
 
 • ••♦♦• 4fc ♦<>*■* <* 
 
 SCALE cif FE.ET 
 
 I. Lincoln, <r. 1250. 
 
 2. Westminster. 
 
 3. Romsey. 
 
 4. Dore. 
 
152 
 
 'i • • * * * q 
 
 . . • ♦ ^ • • • *3f 
 
 <3> 
 
 ♦ ♦ • 
 
 SCALE oF FEET 
 
 1, Beverley Minster, c. 1400. 
 
 2. Southwell. 
 
 3. Oxford Cathedral. 
 
 4. Kirkstall. 
 
153 
 
 w 
 
 ■1 • • « • « ♦ ♦ <^ 
 
 
 1. York Minster. 
 
 2. St Albans. 
 
 3. Jervaulx. 
 
 4. Ely. 
 
154 
 
 » * ♦ 
 
 jj» I I I » M I II I I • *■ 
 
 gs* ♦ ♦<►♦♦♦♦•♦♦♦• til 
 
 I§» #♦•♦♦♦♦ ♦ ♦ ♦ # tp ^ 
 
 Q 
 
 I I I I I 
 
 4+ n 
 
 © 
 
 SCALE OF FEET 
 
 i. Winchester Cathedral. 
 
 2. Salisbury. 
 
 3. Wells. 
 
 4. Exeter. 
 
EARLY CHRISTIAN BASILICAS. 
 
 '55 
 
 people* Thousands and hundreds of thousands of pilgrims visited Rome while 
 all the five patriarchal basilicas were still standing. Their fame went out to the 
 ends of the earth. Bede tells us that when Northern England was barely 
 emerging from heathendom, Benedic Biscop journeyed to Gaul and to Rome for 
 masons and for fittings for the church which he built at Monkwearmouth ; this 
 was in 675. Bishop Wilfrid also, about the same date, when he built important 
 minsters at Ripon and Hexham, 4- sought his inspiration at Rome. Hardly any 
 other type of great church but the basilican was built in Western Christendom 
 till S. Ambrogio was commenced at Milan between 824 and 859. To this type 
 
 S. Aenese, Rome. 
 
 belongs the magnificent group of basilicas at Ravenna, which are of the sixth 
 century : to which period also belongs the remarkable facade of St Saviour, 
 
 * The heavenly temple described in the Hook of Revelation, chaps, iv. to xxi., is but an 
 idealised Christian basilica ; and the ritual an idealisation of early Christian rites. In the apse 
 is the throne of the Bishop ; in the hemicycle the seats of his presbyters, as at Torcello and 
 Norwich (iv. 2, 4). In front of him is the altar (viii. 3) ; and beneath the altar is the confessionary, 
 containing the relics of martyrs (vi. 9). The "sea of glass " is tesselated pavement (iv. 6) ; in 
 front of the church is the atrium or narthex. (See Scott's Essay, 29.) 
 
 + He would hardly have constructed crypts both at Hexham and Ripon but for Italian 
 influence. It is probable that the Ripon crypt was orientated to the west, as originally many 
 of the basilicas at Rome. See Mr J. T. Micklethwaite, quoted in Walbran's Guide to Ripon, 
 1 8th edition, page 39. 
 
i 5 6 
 
 THE APSE. 
 
 Spoleto.* Even when the Romanesque style was in full development, and 
 indeed in Gothic days too, as late as the fourteenth century, the basilican type of 
 church still found adherents, especially in Italy, its mother-country. 
 
 Of the basilican plan the characteristic features are (i) that the church 
 consists of nave and aisles, and that the roof of the nave is raised high above 
 those of the side aisles so as to admit of clerestory windows above the aisle 
 roofs ; (2) that the nave terminates in a semicircular sanctuary or apse ; (3) 
 that no choir is interposed between nave and apse, e.g. ROME S. ACxNESE (r 55)- 
 Some of these churches, constructed in the early days, had double aisles on 
 
 either side of the nave. Such 
 were the basilicas of ST PETER 
 (147) and St Paul at Rome. 
 And this peculiarity was 
 copied in the Romanesque 
 churches of the end of the 
 eleventh century at Cluny ; 
 St Martin de Tours ; La 
 Charite ; S. Sernin de Tou- 
 louse ; S. Remi de Reims ; 
 Ripoli in Catalonia ; S. Ab- 
 bondio, Como ; S. Hilaire, 
 Poitiers ; Souvigny ; Gannat, 
 north of Auvergne ; though 
 in the three last the inner 
 and outer aisles are of different 
 date. This plan is repeated 
 in Early French Gothic in 
 the cathedrals of Paris, 
 Bourges, and Meaux ; at 
 Beaumont-sur-Oise (late 
 twelfth century) and atLagny, 
 c. 1250; and, later, at Troyes, 
 Cologne, Milan, and elsewhere. 
 In England one cathedral, 
 Chichester, and several 
 churches have two or more 
 aisles on each side ; but this has usually come about by the accretion of chantry 
 or other chapels. In Scotland the ruined cathedral of Elgin seems to have 
 had a nave set out in the thirteenth century with double aisles. 
 
 To the east the basilica ended in a semicircular apse. This apse may 
 have existed in larger Anglo-Saxon buildings ; but all their larger churches 
 have perished. In their smaller churches, e.g. St Pancras, Canterbury ; 
 Reculvers; Brixworth; Worth; it certainly existed, but was less common than 
 the rectangular east end. In Norman days nearly all the larger Norman churches, 
 both in Normandy and England, from e. 1040 to c. 1 140, seem to have had an 
 apse. 
 
 * Cattaneo, 146. 
 
 Hales, Exterior of Apse. 
 
THE APSE. [57 
 
 But, as time went on, the Normans were Anglicised ; they sank into and 
 were absorbed into the immensely preponderating population of Anglo-Saxon 
 and Celtic origin in the midst of which their lives were cast : they became 
 naturalised Englishmen. The process, indeed, of naturalisation was long- 
 retarded by the vastness of the possessions held by our early kings in what is 
 now France ; but it was always growing in force and was never arrested. The 
 English language, the English institutions, and the English law emerged into 
 strength ; and in church planning the English oblong type of chancel more and 
 more superseded the apse with its long and hallowed pedigree stretching back to 
 the fourth-century basilicas of early Christian Rome. We may indeed speculate 
 that the fact that the apsidal chancel was largely associated with a foreign priest- 
 hood, foreign bishops, and foreign masters was not calculated to endear it to 
 English minds. In any case, from the middle of the twelfth century, the tendency 
 was, more and more, to build new chancels rectangular, and to square those that 
 had been apsidal ; e.g. MELBOURNE (213 . 
 
 It is to be noted, moreover, that though the greater presbyteries, both in 
 Normandy and England, were very generally apsidal, yet, at any rate in England, 
 some few seem to have been rectangular ; viz. Dover ; Southwell ; Sherborne, 
 and ELY (153). Moreover, the presbyteries of Hereford, Llandaff, and ROMSEY 
 (151) were rectangular, though east of all of them there was probably an ambula- 
 tor}', with an apse or parallel apses projecting east from it. Here and there, more- 
 oxer, there were rectangular eastern chapels, e.g. at Canterbury and Rochester. 
 Such examples — and there may have been many of them in the numerous 
 Norman choirs remodelled or rebuilt in Gothic days — as well as the numerous 
 rectangular chancels of parish churches, must have tended to familiarise the 
 eye with the rectangular eastern form. 
 
 But, apparently, it was reserved for an alien and Continental influence to 
 administer the coup de grace to the apsidal type of choir. This was the advent 
 in England from Burgundy of the Cistercians and of Cistercian planning. 
 During the twelfth century Cistercian influence was predominant throughout 
 the whole Catholic world, from Scotland to Sicily, from Scandinavia to Spain. 
 By this puritanical order the utmost simplicity of planning, as of ritual, was 
 uniformly enforced when the first great group of Cistercian abbeys was built 
 in England ; e.g. Fountains, c. 1 135 ; Buildwas, c. 1 148 ; Furness, c. 1 148 ; Kirk- 
 stall, c. 1 152; Louth, Lincolnshire, founded 1139; Roche, c. 1165; Jervaulx, 
 c. 1 1 80; Netley, 1239; Tintern, 1269; all of which accordingly have rectangular 
 chancels. And when an ambulatory was built east of the choir at Byland, 
 c. 1170, and Dore, c. 1200, the east end still remained square. In nearly ail 
 our Cistercian churches the original plan of the Burgundian abbey church of Clair- 
 vaux was followed. It was exceptional to copy, as at Croxden, c. 1188, and 
 Beaulieu, c. 1221, both Cistercian, the apse and ambulatory of the second 
 Clairvaux and Pontigny. Reinforced by the example of these great builders, we 
 English threw off the yoke of the foreign plan. The Sens plan indeed arrived and 
 found realisation at Canterbury in 1175 ; but it made proselytes nowhere; and 
 the He de France plan, with polygonal instead of semicircular apse, arrived in 
 1245, and gave us the chevet of WESTMINSTER I 15 1.2;, but with English emenda- 
 
158 THE APSE. 
 
 tions.* And in 1192 St Hugh's architect had given us a curious polygonal apse 
 at LINCOLN (151.1), only to be destroyed in 1256. A similar apse was built 
 at Pershore, c. 1223. In the fourteenth century Tewkesbury built a polygonal 
 apse, probably on Norman foundations. A few minor examples may be seen 
 in parish churches ; e.g. Madley, Herefordshire; fourteenth century. These, 
 however, were the last expiring efforts to retain the apse. What had been the 
 predominant type of east end both before and after the Conquest, in our smaller 
 churches, but which in Norman days had been superseded by the apse in the 
 larger churches, had become, as it was to continue to the end, the characteristic 
 eastern termination of the English church, f 
 
 * For these see Sir G. G. Scott's Gleanings, 23. 
 
 t Enlart's Manuel, 483, 485, gives lists of numerous French churches with square east end ; 
 chiefly small parish churches ; it is particularly common in Normandy, Burgundy, Champagne, 
 and the south-west. 
 
Chapter IX. 
 
 PLANS OF THE EASTERN LIMB OF THE GREATER 
 CHURCHES. 
 
 First Plan: Type, Norman Durham: Three Parallel Eastern 
 Apses; or variants thereof. 
 
 We have now to consider the planning of the aisles of the choir. At first, 
 in basilicas of the simple type of S. Maria Maggiore, or S. Maria in Cosmedin, 
 Rome, the central apse projected clear of the aisles, which were square-ended. 
 But this simple plan seems soon to have been complicated by the need of a 
 sacristy and a library ; in the earliest basilicas room may have been found for 
 these near the entrance, as at Old St Peter's, Rome, where there was a building 
 arranged like a little basilica, to the left of the narthex.* But it was more 
 convenient to have them nearer the high altar. And so Paulinus of Nola, who 
 died in 431, describes on the right of the apse a sacristy where the bread and 
 wine and the church plate and the vestments are kept — 
 
 " Hie locus est veneranda penus qua conditur, et qua 
 Promitur alma sacri pompa ministerii :" 
 
 and on the left the library — 
 
 " Hie poterit residens sacris intendere libris." t 
 
 The same arrangements occur in the sixth century in the churches of S. John 
 the Evangelist and S. Vitale, Ravenna ;* in the fifth and sixth century in Central 
 Syria : e.g. Tourmanin and St Simeon Stylites ; § three parallel eastern apses 
 are found. A recent discovery at Hispalis proves that the central apse contained 
 the episcopal chair ; the one on the right the church plate : the one on the left 
 the library. From the Eastern Church the triple eastern apse passed to Rome : 
 an early example, perhaps, is seen at S. Maria in Cosmedin, which was re- 
 modelled for the Greeks who had been exiled by the Iconoclasts of Con- 
 stantinople. Pope Adrian is described as " tres absides in ea constituens " ; this 
 was in 872. A still earlier example at Rome is that of S. Maria in Domnica, 
 
 * Cattaneo, 60. 
 
 t Scott's Essay, Si. The basilican church of the monastery of St Catherine at Sinai, which 
 is undoubtedly of the time of Justinian, has apse and side chambers. Lethaby's Medieval 
 Art, 60. 
 
 I The basilica at Bethlehem has three parallel eastern apses. These are held by De Vogue, 
 R. de Fleury, and Kraus to be Constantinian. Professor Lethaby (Med. Art, 58) thinks that 
 they are later than Constantine, and may be anterior to Justinian. 
 
 S Illustrated in Scott's Essay, 62. || Lanciani's Ancient Rome, 187. 
 
i6o 
 
 Cerisy from S.E. 
 dloucester, N.E. Apse. 
 
 St George's de Boscherville from E. 
 Norwich, S.E. Apse. 
 
[6i 
 
 Wells from East. 
 
 Peterborough N. Transept. 
 
 Cerisy Choir. 
 
l62 
 
 Castle Rising. 
 
 St Margaret's, Lynn. 
 
 St David's. 
 Sudbury St Peter. 
 
THREE PARALLEL EASTERN APSES. 163 
 
 817-824.* The central apse of TORCELLO CATHEDRAL (148.1) is dated by 
 Cattaneo c. 650 ; the absiodoles c, 864. 
 
 From these basilicas the plan passed to the earliest Romanesque 
 churches. It appears in the plan of ST call (194J, and in the oldest work 
 of S. Ambrogio, Milan, A.D. 824-859; again, c. 850, in the same city, at 
 S. Vincent in Prato ; and again at Alliata, near Monza, A.D. 881.+ In 
 France it appears in Carlovingian work at Germigny-les-Pres, consecrated 
 806 ; and at St Generoux, which also may be of the ninth century. When we 
 come to the great churches of Normandy in the eleventh century, the triple apse 
 or some modification of it is the invariable arrangement* (1.) Three parallel 
 apses occur in the eleventh century at Guibray and the Abbaye-aux-Dames at 
 Caen.§ (2.) The semidomes, however, of such little side apses are much easier 
 to roof with wood, if they are squared externally ; and this is often done. So 
 that while the central apse is semicircular externally and internally, the side 
 apses may be semicircular internally, but externally rectangular. This arrange- 
 ment is characteristic of the Romanesque churches of the Como district ; e.g. 
 S. Abbondio. In Normandy it occurs in the eleventh century at Lessay ; St 
 Gabriel ; St Nicolas, Caen ; St George's de Boscherville ; and in the transept 
 apses of the Abbaye-aux-Hommes. ; (3.) But when the apses of the aisles had 
 been squared externally, it was not a long step to square them internally also. 
 And so we get a central apse flanked by aisles squared within and without. 
 In Normandy this plan occurs at CfiRISY-LA-FORET (148.3), which may have 
 been set out in the eleventh century ; and on the eastern frontier of Normandy 
 at Gournay, where the work is probably early in the twelfth century."" 
 
 This, then, was the normal plan of eastern limb which the Norman 
 builders brought over to England at the Conquest : a central apse flanked 
 (1) by absiodoles, (2) which might be rectangular externally, or (3) by square- 
 ended aisles. In England, unfortunately, the greater part of the eastern 
 limbs of the large Norman churches have disappeared. Either they were 
 pulled down, as at Lincoln, to be rebuilt on a more magnificent scale in 
 Gothic ; or, where the church was served by monks, the parishioners at the 
 Dissolution retained for their use only the nave, as at Binham, Leominster, 
 Wymondham ; and the choir fell into ruin, was pulled down, and dis- 
 appeared. Nevertheless, excavations have made clear the eastern terminations 
 of several of the large Norman choirs. 
 
 Peterborough retains the central apse ; the foundations of semicircular lateral 
 apses exist underground. At DURHAM (149. 1 ) the foundations of three parallel 
 eastern apses have been found. In both cases the lateral apses were semi- 
 circular within, square without; so also probably at Selby.** The same arrange- 
 
 * Plans in Cattaneo, Figs. 80 and 90. t Cattaneo, Figs. 123 and 128. 
 
 \ It must be borne in mind, however, that several of the east ends have been destroyed, or 
 have been replaced in Gothic ; some of those may not have possessed the triple parallel eastern 
 apses. Rouen Cathedral has a periapsidal plan of Gothic date, which may perhaps be a survival 
 of a similar Romanesque plan. 
 
 § Ruprich-Robert, i., Plates 8, 9. || Ruprich-Robert, 60, 1, and Plates 8, 9, 93. 
 
 TT Ruprich-Robert, Plate 54. 
 
 ** iMr J. Bilson in Arcfuzological Journal, liii. 8, on Durham apses ; and Mr C. C. 
 Hodges in the Archaeological Journal, liii. 1 et seq., on Selby apses. 
 
i6 4 THE PERIAPSIDAL PLAN. 
 
 ment occurred at ST ALBANS (98) * St Mary's, York, Castle Acre, f and Exeter. 
 At ROMSEY (151.3) lateral apses remain; these are semicircular within, rect- 
 angular without. ST MARY'S, GUILDFORD (36), retains lateral apses, built in the 
 last years of the twelfth century. 
 
 Both Lindisfarne and Melbourne originally had central apses ; but they 
 had no lateral apses, as their choirs were without aisles. 
 
 Not a single example remains entire. If we wish to see the effect of 
 the parallel apse plan, we must go to Normandy and visit CERISY-LA-FORET 
 and ST GEORGE'S DE BOSCHERVILLE (160, 160) ; in both of which, however, the 
 aisles are square-ended externally, while at Cerisy they are also square inter- 
 nally, except in the triforium, where they are semicircular. 
 
 Two important variants remain to be noticed. One is that in which the 
 aisles end in semicircular apses, but the presbytery is rectangular. Ely j pres- 
 bytery, originally intended to be apsidal, was made rectangular between 1103 
 and 1106. The other three are Sherborne, § begun 1107; Southwell, || between 
 1108 and 1 1 14 ; St Martin's Priory, Dover^T between 1131 and 11 39.** The 
 other variant is seen at Hereford, LLANDAFF (580), and ROMSEY (151.3). In 
 these the presbytery is rectangular, but it opens by a semicircular arch or arches 
 into an aisle or ambulatory running north and south. This ambulatory survives 
 at Romsey. Romsey had also a central chapel or chapels, probably apsidal, pro- 
 jecting eastward from the centre of the ambulator}-. This was rebuilt in Gothic 
 days. Hereford yf and Llandaff may have had a similar plan. This second 
 variant is of the utmost importance, and is peculiar to the West of England 
 school of Romanesque. 
 
 Second Plan : Type, Norwich Cathedral ; Semicircular Ambula- 
 tory with Radiating Chapels. 
 
 Periapsidal Plan. — But another more complicated and improved type 
 of planning was more common in England than that of the three parallel eastern 
 apses. In this the central eastern apse was encircled by a semicircular aisle, 
 which is called the ambulatory. 1 ^ Of this we may distinguish three species — 
 (1) The ambulatory with tangential chapels ; (2) the ambulatory without 
 chapels ; (3) variants of the above. 
 
 By far the most common type is that in which chapels radiate to the north- 
 
 * The late Lord Grimthorpe's Guide to St Albans, 5 ; and Buckler's St Albans. 
 
 t For the plan of Castle Acre see paper by Mr W. H. St John Hope in Norfolk and 
 Norwich Arch. Soc, 1894. 
 
 J Willis, in Stewart's Ely, gives plan of the discoveries in Ely choir. 
 
 S Carpenter in Journal of R. LB. A., March 1877. 
 
 || E. Christian in Journal of British Archaeological Association, January 1853. 
 
 IT Rev. Dr Plumptre in Arch. Cantiana, vol. iv., has plan of Dover Priory. 
 
 ** Bristol also is said, by Mr E. W. Godwin, to have had a square-ended choir, but this is 
 merely a conjecture. Arcliaol. Journal, xx. 47. 
 
 tt Mr A. Moore, Hereford, sends the following extract from the I lercford Journal of 
 June 13, 1S63 : — "At the commencement of the present restoration the foundations of the 
 original apsidal (or semicircular) terminations of the choir and its aisles were discovered." 
 
 \% No ambulatory with tangential chapels is older than c. 900 a.d. Lenoir, Architecture 
 monastique, ii. 35 ; Yiollet-le-Duc, Architecture, ii. 456 ; and Comte Robert de Lasteyrie's 
 Monograph on St Ma? tin de Toms. 
 
i6 5 
 
 Tewkesbury Choir. 
 
THE PERIAPSIDAL FLAX. 167 
 
 east, east, and south-east. The usual number of these radiating chapels is 
 three. Lewes, a Cluniac abbey, as remodelled soon after 1 100, had five of these 
 chapels, probably copying the arrangements of CLUNY (150.1). Not one 
 of these choirs survives entire. GLOUCESTER (135) and NORWICH (160) have 
 retained their north-east and south-east chapels, but in either case the eastern 
 chapel was demolished and rebuilt much larger as a Lady Chapel, which also 
 at NORWICH (148.4) has disappeared. At GLOUCESTER (192.1), however, the 
 original eastern termination is seen perfectly in the crypt. At CANTERBURY 
 CATHEDRAL, in Ernulph's choir, 1 096-1 107 (149.2), there were three chapels, 
 of which the eastern one was square ; tangential chapels, those of St Anselm and 
 St Andrew, still survive. Diggings have shown that the periapsiclal plan existed 
 at St Augustine, Canterbury, which had an apse of seven bays ; * also at the 
 collegiate church of St Martin-le-Grand, Dover. So it did at Tynemouth, and 
 Bury St Edmunds ;f in the present cathedral of Chester ;J at Leominster ; § 
 and at Battle and Pershore. From the present form of the choir of TEWKES- 
 BURY (165), and from its proximity to Gloucester, Pershore, Leominster, and 
 Worcester, it is probable that this church also had the same plan. All but 
 one of the above were Benedictine churches, and Canterbury and Norwich were 
 Benedictine cathedrals ; Lewes 'was Cluniac. The same plan was also adopted 
 by Secular Canons at Chichester ;V and by Cistercian monks at Croxden. The 
 Benedictine abbey of Reading had central apse and ambulatory; it is uncertain 
 whether there were radiating chapels. 
 
 Of the second type of plan, viz. without radiating chapels, there are few 
 undoubted examples. Edward the Confessor's Benedictine church at West- 
 minster had central apse and ambulatory ; so far no chapels have been dis- 
 covered. Lichfield Cathedral, built by Secular Canons, had central apse and 
 ambulatory, but no chapels have yet been found.** The same is the disposition of 
 St Bartholomew's, Smithfield. X4 " But an ambulator}- without radiating chapels is 
 so rare in Romanesque that these three examples should be regarded as doubtful. 
 
 Thirdly, variations are played on the periapsiclal plan. At Worcester 
 Cathedral, as the remains of the CRYPT ( 192.2, show, there was central apse and 
 ambulatory ; but instead of radiating chapels" the choir was flanked by elongated 
 chapels projecting eastward from the transepts. This was the plan of the Abbaye- 
 aux-Dames at Caen, except that in that church there was no ambulatory; so also 
 of the Cistercian church of Yaux-le-Cernay (1128); and probably of Cluny 
 before the rebuilding of 1082. Winchester possessed apse and ambulatory ; but 
 the side chapels were not set tangentially, but due east, parallel to an eastern 
 chapel ; as is well seen in the CRYPT (192.5). This is a sort of blend of the plan 
 with three parallel eastern apses and that with apse and ambulator)-. 
 
 In plan the normal type of the tangential chapel is semicircular. But at 
 
 * See paper by Mr St John I lope in Archico/ogia Cantiana, xxv. 
 
 t Paper by Dr M. R. James in Cant/'. Aniiq. Society's Octavo, xxviii. 
 
 I Sir G. G. Scott in Chester Architect, and Archaol. Soc. Journal, iii. 169. 
 
 % Arch. Journal, 10, [II. Scott's Gleanings, 19. 
 
 IT See Willis' plan and evidence in monograph cm Chichester. 
 
 ** See Mr Hope's plans in Builder, Feb. 7, 1S91. 
 
 + t Plan in Architectural Review, i. 1, 22. 
 
 XX If radiating chapels exist, they are blocked up, and have still to be found. 
 
THE PERIAPSIDAL PLAN. 
 
 SMI 
 
 !, I 
 
 
 Cluny. 
 
 Gloucester (192.1) the chapels are pentagonal ; while at Norwich (148.4), each 
 
 is composed of part of a 
 small circle, serving as 
 choir, and part of a large 
 circle, serving as nave. 
 At CANTERBURY (149.2) 
 the north-east and south- 
 east chapels are oblong, 
 but with eastern apses. 
 
 Whence came these 
 apses with circumambient 
 aisles in Ernulph's Canter- 
 bury; St Augustine's, 
 Canterbury; Norwich; 
 Gloucester; Bury St 
 Edmunds ; Leominster ; 
 Battle; Tewkesbury; Per- 
 shore ; Chichester; Lewes; 
 Lichfield ; Worcester ; St Bar- 
 tholomew's, Smithfield ; Win- 
 chester ; St Werburgh, Chester ; 
 St John's in the Tower of 
 London, and the Norman abbey 
 of Westminster? Hardly from 
 Normandy ; for " till Fecamp, 
 A.D. 1082, there was no ' Rond- 
 Point' in Normandy; and Fe- 
 camp remained without imi- 
 tators for at least a century."* 
 Nor again is the ambulatory 
 characteristic of the Roman- 
 esque schools of Lombardy, 
 German} 7 , Provence, or Peri- 
 gueux. Three schools employ 
 it most ; those of (1) Burgundy, 
 (2) Poitou, (3) Auvergne and 
 Toulouse.f It is difficult to 
 connect historically eleventh- 
 century England with Auvergne 
 and Toulouse. In Burgundy 
 the periapsidal plan is charac- 
 Norwich, East Side of South Transept. teristic ; especially in the great 
 
 * A. St Paul 111 Planat's Encyclopedic, vi. 23. Enlart quotes, however, the substructures of 
 the cathedral of Evreux, where there was a consecration in 1072 {Manuel, 227). Avranches 
 may be added. 
 
 I- See paper by the author on Classification of Romanesque in the R.I.B.A. Journal 3, 
 
 viii. 12, p. 2S2. 
 
THE PERIAPSIDAL PLAN. 169 
 
 churches of the Cluniac Order. CLUNY (168), the largest Romanesque church 
 in Christendom, had this plan. So has its daughter church, Paray-le-Monial, 
 which was dedicated in 1104. La Charite, another Cluniac church, which 
 originally had three parallel eastern apses, received an ambulatory in the twelfth 
 century.* The same transformation of a triapsal into a periapsidal choir occurred 
 at Tournus and Vignory ; and in England at Lewes ; all early in the twelfth 
 century. The choir of St Etienne, Nevers, a great church which architecturally is 
 on the frontier of Burgundy and Auvergne, has an ambulatory ; it was begun 
 in 1063, but was not consecrated till 1099. It were natural to suppose that the 
 grand church of CLUNY (150. 1) would be the leading influence in England in the 
 eleventh century. But, in the first place, the Cluniac Order was never strong in 
 England; and nearly all our periapsidal churches were Benedictine. (2.) Chrono- 
 logy forbids the supposition. The ambulatory of Cluny was not commenced 
 till 1089 ; but that of Winchester was begun in 1079, that of Worcester in 1084, 
 of Gloucester in 1089 ; while the little chapel in the Tower of London is c. 1080 ; 
 and Edward the Confessor's abbey church at Westminster, consecrated in 1065, 
 had an ambulatory ; for a contemporary writer \ says that the " ambitus ip sins 
 aedis duplici lapidum arcu clauditur."\ The bases of three of the piers of this 
 duplex lapidum amis, i.e. apse and ambulator) - , exist below the pavement of the 
 present sanctuary ; and one of them may be seen, by means of a trap-door, to 
 the north of the high altar. § So far as dates go, therefore, it would be more 
 reasonable to derive the ambulatory of Chin}- from that of Westminster than 
 that of Westminster from that of Cluny. 
 
 The truth is, both are derived from one common source ; and that source 
 is to be found in one of the most important abbey churches in mediaeval 
 Europe, a special resort of pilgrims, ST MARTIN, TOURS (192.3). Excava- 
 tions made in i860 have shown that the great double ambulator)' and 
 radiating chapels — a work of the thirteenth century, destroyed at the French 
 Revolution — were an amplification of an earlier eastern limb, built between 
 997 and 10 14, which consisted of a choir of two bays and an apse of five 
 bays, surrounded by a single ambulatory and five radiating chapels. | Immedi- 
 ately following this, in all probability, was the work at Xotre Dame de la 
 Couture, Le Mans, where to a ninth-century aisleless choir and apse, similar 
 to that shown in the plan of ST (.ALL (194), were added an ambulator)- and 
 five radiating chapels. This was built in the time of Abbot Gauzberts, 990- 
 1007 ; who came from Tours, and had been a great builder there. Close on 
 this follow S. Remi at Reims, c. 1035 ; St Savin, between 1020 and 1030 ; 
 St Hilaire de Poitiers, consecrated 1049 ; Notre Dame du Pre, Le Mans, re- 
 modelled c. 1050; St Sernin de Toulouse, whose choir was consecrated in 
 1096 : all great pilgrim-churches. And so we arrive at the unexpected con- 
 
 * Dehio, Plate 120, 2, and 121. 3. 
 
 + Quoted in full in Scott's Essay, 131, note. 
 
 X See plan by Mr Micklethwaite in Builder, Jan. 6, 1894. 
 
 Jj This base is figured in the Builder, lee. cit. 
 
 || Bulletin Monumental, vol. 40, 147. Dehio, text, i. 267; anil Comte de Lasteyrie's mono- 
 graph on the excavations, in the Memoires de l' Academic des Inscriptions et Belles-lettres, 
 Paris, 1 So 1. 
 
i 7 o THE PERIAPSIDAL PLAN. 
 
 elusion that the great majority of our Norman churches are probably not 
 Norman in their planning, but hail from St Martin de Tours. 
 
 The convenience and superiority of the periapsidal as compared with the 
 triapsal plan must have been from the first manifest. The fact that it occurs 
 first in the great pilgrim-churches of Western France may point to its origin. 
 The triapsal plan meant danger to life and limb on days of crowded pilgrim- 
 ages. But, with an ambulatory, the pilgrims could proceed up one choir aisle, 
 pass behind and round the apse, and down the other choir aisle, without having 
 to retrace their steps. It was equally convenient in processions, especially the 
 great Sunday Procession,* when the monks had to circumambulate the church in 
 order to asperge every altar before the supreme Mass of the week.f Moreover, 
 three or five altars which before, probably, had cumbered the nave, could 
 now be placed in the new radiating chapels of the ambulatory. And the 
 ambulatory afforded the necessary access to them. Again, each apsidal 
 chapel could be treated as a sanctuary to be entered only by the officiating 
 priest and his attendants, and the ambulatory provided the necessary nave for 
 the worshippers. At GLOUCESTER (135),+ indeed, the three radiating apses 
 of the ambulatory of the choir, as well as the two eastern apses of the transept, 
 were built three stories high ; one in the crypt, one on the ground-floor, and one 
 in the upper aisle. And, for access to these, three ambulatories were con- 
 structed in a similar position ; the uppermost ambulatory being open to the 
 choir, floored, lighted by windows at the back, and approached by broad stair- 
 cases. So very convenient were these arrangements found at Gloucester, that 
 they were allowed to remain almost intact even when the central apse was 
 made square, c. 1350; the ambulatory of the upper aisle reappearing in the 
 so-called Whispering Gallery. 
 
 In France the periapsidal plan subsisted to the end. The five choir-chapels 
 of St Martin de Tours, St Savin, Cluny, St Sernin de Toulouse, St Jago de 
 Compostella, amplify into the seven apsidal chapels of Amiens, Beauvais, 
 Cologne, and that masterpiece of French Gothic, Le Mans. But because arches 
 curving on plan are difficult to construct, the semicircular apses and absiodoles 
 of Romanesque became polygonal in Gothic. 
 
 These then were the two characteristic plans of the greater Romanesque 
 churches in England ; either (1) three parallel eastern apses, or some variant of 
 these ; or (2) an ambulatory, almost always with radiating chapels. These two 
 plans held the field to the almost total exclusion of all others for about a century 
 after the Norman Conquest. The one important exception is that of ROMSEY 
 (15 1.3), and perhaps Hereford, Llandaff, and Sarum. 
 
 * For the route of the Sunday Procession see Hope's Rochester, 217. 
 
 + In the account, quoted in Willis' Canterbury, 61, of the rebuilding of Canterbury 
 choir in 1 180, the monk Gervase expressly says that "the master preserved as much as he 
 could the breadth of the passage outside the choir on account of the processions which were 
 there frequently passing." 
 
 % In the illustration of Gloucester Cathedral (160), the three tiers of windows in each 
 absiodole correspond to the three superposed chapels. 
 
" m n 
 
PLAN OF SALISBURY CHOIR 
 
 w 
 
 Third Plan: Type, Salisbury Cathedral: Rectangular Ambulatory ; 
 sometimes without, more often with a low rectangular 
 Eastern Chapel. 
 
 Roth the above plans — that with the parallel eastern apses, and that with 
 the semicircular ambulatory — were, as we have seen, direct importations from 
 the Continent. The next plan, which came into use mainly in the last half of 
 the twelfth century, is of twofold origin. In the early Gothic of the North of 
 England it is of Cistercian, i.e. of Burgunclian origin. The plan of Byland 
 Abbey is plainly but a simplified version of such a Cistercian plan as that 
 in the Sketch Book of Villard de Honnecourt ; or those of Ebrach, Lilien- 
 feld, Arnsburg, Riddagshausen.* 
 
 But in the South of England it appears in the Benedictine nunnery church 
 of ROMSEY (i 5 1.3), quite early in the twelfth century ; too early for any Cistercian 
 influence. Romsey was built not later than 11 20. Here the high choir was 
 given a square east end, but the ground-story of this was pierced with two 
 arches. Behind these was built a processional aisle or ambulatory connecting the 
 north and south aisles of the choir, and from the ambulatory projected an eastern 
 chapel or chapels. The new eastern aisle may be regarded as the rectangular 
 equivalent of the semicircular ambulatories which were exceptionally numerous 
 in the abbey churches of the West of England ; e.g. Worcester, Pershore, 
 Leominster, Gloucester, Tewkesbury. The Romsey plan may have existed 
 also at Hereford, Llandaff, and Old Sarum, which William of Malmesbury 
 tells us was built anew by his contemporary, Roger, Bishop of Salisbury ; i.e. 
 between 11 15 and 1139. 4 - It is quite clear that the rectangular ambulator}- 
 was adopted at Wells Cathedral, c. 11S0; and at Glastonbury,^ after the fire of 
 1 184. Another early adoption of the Romsey plan is in the choir of Lichfield 
 as built c. 1 190. From its plan, as well as from the moldings and capitals, 
 this choir clearly belongs to the West of England School of Gothic. Lichfield, 
 like Romsey, had two arches from the choir to the ambulatory. § 
 
 In Cistercian work it occurs twice : at Byland, in Yorkshire, to which 
 the monks removed from Stocking in 1 1 77 ; and at DORE (182), in Hereford, 
 where the eastern termination of the choir seems to have been remodelled c. 
 1200. Dore choir is practically a reduced version of that of Ebrach, consecrated 
 1 178. The Cistercian churches in Europe were so commonly built from plans 
 inspired by the mother-abbey at Citeaux or by its daughter churches, that it 
 is hardly safe to regard the plan of DORE (15 1.4) as derived from that of Romsey. 
 
 Two more doubtful cases occur in Yorkshire. In the choir of York, as 
 rebuilt by Archbishop Roger, 1 1 54 to 1 18 1, it has been conjectured that 
 there was a rectangular ambulator}-. If so, this may have been the case also 
 
 * Oehio, Plate 193 (Byland) ; 191, 195. 
 
 t A plan of it is given in Prior, 67 ; the evidence for it is not strong. 
 
 I There is no evidence as to the existence of a projecting eastern chapel at Wells. Mr 
 James Parker in his paper in the Somerset Arch. Proceedings, vol. xxvi., gives a plan of 
 Glastonbury east end; in 1190 it was exactly the same as that of DORK (151. 4). Willis 
 assumed that Glastonbury had a projecting eastern chapel. Mr W. H. St John Hope has 
 recently proved by excavations at Glastonbury that there was no eastern chapel in 1190. 
 
 J; See Mr St John Hope's plans in Builder Cathedrals. || Willis' York, 11. 
 
i 7 2 PLAN OF SALISBURY CHOIR. 
 
 with his choir at Ripon ; the east end of which, however, had to be rebuilt at 
 the close of the thirteenth century. 
 
 At Romsey, Byland, and Wells there was but one eastern aisle. The 
 next improvement was to increase the number to two, as at Dore and Glaston- 
 bury. At Dore there are remains of five altars against the eastern wall, and 
 of low walls cutting up the easternmost of the two aisles into so many chapels.* 
 The westernmost aisle was employed as ambulatory or processional path. The 
 next improvement was that of Bishop de Lucy at WINCHESTER, 1202 (154.1), 
 viz. to construct three eastern aisles with three chapels east of them ; thus 
 providing plenty of space for worshippers to attend the services at the eastern 
 altars; as well as room for processions. At St Saviour's, Southwark, f 1213- 
 1238, which was connected with Winchester, similarly three eastern aisles were 
 provided, but without the three eastern chapels of Winchester. 
 
 The Romsey eastern chapel as well as the rectangular ambulatory soon 
 inspired imitation. At Chichester the Secular Canons had finished a rectangular 
 eastern chapel before 1 1 75. After the fire of 11S7, for the Norman ambulatory 
 and its radiating chapels two eastern aisles were substituted ; and at the end 
 of the thirteenth century the Lady Chapel was prolonged still further. At 
 Hereford, c. 1190, the Norman ambulatory appears to have been raised and 
 vaulted, eastern aisles added, and the Norman eastern chapel replaced by a 
 square one; this chapel was still further prolonged c. 1230, and forms the 
 present Lady Chapel (464). 
 
 Finally, all these experiments were summed up in the beautifully sym- 
 metrical plan of SALISBURY (154.2); with triple eastern arch, two eastern aisles, 
 and projecting Lady Chapel. St Patrick's Cathedral, Dublin, finished in 1235, is 
 the same in plan as Salisbury ; except that the choir has but one arch in the 
 eastern wall and one eastern aisle. The choir of Milton Abbas, Dorset, 1290 to 
 1300, had three eastern arches, an aisle or aisles, and a Lady Chapel. EXETER, 
 c. 12S0 (154.4), was satisfied with two eastern arches, one aisle, and Lady 
 Chapel. ST ALBANS (153.2) has three eastern arches, three eastern aisles, and 
 Lady Chapel. WELLS, c. 1340(154.3), presents the plan in its most attractive 
 form ; with piers, arches, and vaults set so as to produce fairy-like vistas and 
 perspectives. Ottery St Mary, also of the fourteenth century, is a reduced 
 copy of Exeter. The great parish church of St Mary Redcliffe, Bristol, was 
 remodelled, c. 1442, very much on the lines of Ottery St Mary.* 
 
 The geographical distribution of the Salisbury plan is remarkable. It is 
 
 * Piers and capitals from Dore retrochoir are illustrated on page 422. 
 
 t Later on a projecting Lady Chapel was added ; this was destroyed to make room for 
 the south approach to London Bridge. 
 
 I It is often assumed that every eastern chapel was intended as a Lady Chapel. This is 
 not certain in all cases ; especially in early examples, such as at Romsey and Winchester. For 
 it was not till Pope Innocent III. and St Bernard urged increased veneration of the Blessed 
 Virgin that Lady Chapels rose to importance, and finally received such great augmentation in 
 scale and splendour, as at Peterborough and Ely. Even then it was not always thought 
 desirable to dislodge the shrine of the local saint east of the high altar, and the new Lady 
 Chapel was placed to the north of the choir as at Oxford, Wymondham, Llanthony, Bristol, 
 Peterborough, Canterbury, Ely ; or south of it, as at Rochester ; or west, in the galilee, at 
 Durham. 
 
173 
 
 ^ 
 
 
 ^ ^ ,f "'" 1 
 
 Salisbury Lady Chapel from N.W. 
 
PLAN OF SOUTHWELL CHOIR. 175 
 
 entirely confined to the South and West of England. In fact, it is one of the 
 marks which differentiate the Southern from the Northern Gothic of England.* 
 
 It is to be regarded as a home-grown development of the Romsey plan ; the 
 first complete breaking away from Continental traditions of planning. It is a 
 very beautiful plan. Internally, its shadowy recesses and broken lights add 
 mystery and distance. Externally, Lady Chapel, Retrochoir, Choir, rising up in 
 successive ranges, Alp behind Alp, leading up to, and culminating in the central 
 spire, as at SALISBURY (i/O), group marvellously. It is a most worthy rival 
 of the perapsidal plans of Amiens and Westminster ; internally, it rivals them ; 
 externally, its superiority is beyond question. 
 
 Same Flax : Type, Southwell Minster ; with High Rectangular 
 Unaisled Presbytery or Lady Chapel. 
 
 But neither was this beautiful type destined to endure, any more than that 
 of the three parallel apses or the semicircular ambulatory. It was developed and 
 perfected in the last half of the twelfth and the first halfofthe thirteenth century. 
 After that, with the belated exceptions of Wells and Ottery St Mary in the four- 
 teenth, and St Alary Redcliffe, Bristol, in the fifteenth century, it fell wholly out 
 of use. It conquered for itself much of the South and the West of England ; 
 it failed to win over the Gothic of the North. 
 
 There must have been some inherent fault in the Salisbury plan. What 
 was it? It ma}' have been that the English mediaeval builder had already 
 begun to show his deep dislike of intricacy of planning. One practical con- 
 sideration, however, may be suggested. The great churches of the eleventh, 
 twelfth, and thirteenth centuries were exceedingly dark ; there was often but 
 one small window in each bay of the clerestory or aisle, and the walls were 
 so thick that but little light could penetrate the gloom of the centre of nave 
 or choir. What little light there was, was reduced to a minimum by the thick- 
 ness and opacity of the stained glass. The chief source of light was through 
 the end walls of the nave or transepts, as at PETERBOROUGH (161) ; or of 
 the apse. With the ancient triapsal plan, as is well seen at CEklSY-LA- 
 EORET (161), the central apse of the choir was most effectively lighted by 
 three tiers of windows, north, east, and south ; but when an ambulatory was 
 added, the light was usually reduced to that from the clerestory windows, 
 which were too high up to be of much service. 
 
 The gloom of St Scrnin, Toulouse, or St Etienne, Nevers, and even of 
 Chartres, must be felt to be appreciated. To a practical builder like the 
 Englishman, the bad lighting of Winchester and Salisbury, just at that spot 
 where light was most essential — that all might follow with the eye each 
 movement of the priest officiating at the Mass — may well have seemed an 
 insuperable objection to either plan of choir. 
 
 A much simpler method, which at any rate ensured good lighting — for the 
 light entered on three sides — was to construct a short unaisled oblong sanctuary 
 immediately adjoining the central transept. It was a plan especially suitable to 
 
 * Unless the Ripon and York choirs of the twelfth century had projecting eastern chapels 
 as well as rectangular ambulatories. 
 
i 7 6 PLAN OF CHOIR OF YORK MINSTER. 
 
 a monastic church in which the monks sat in the crossing and the eastern bays 
 of the nave. And as it was a simple and easy solution of the lighting problem, 
 it was adopted in very many of the large Cistercian churches ; e.g. in England 
 at KIRKSTALL (152.4), Furness, Buildwas, and Roche, and no doubt in other 
 Cistercian abbeys where, as at Rievaulx and Fountains, the choirs of the 
 churches were afterwards enlarged and remodelled.* But, ritualistically, it 
 was a retrogression ; it did not admit of processions, or of a continuous flow of 
 pilgrims round the sanctuary. But as the Cistercians did not wish or expect 
 to have a concourse of pilgrims in their sequestered churches, this probably 
 seemed no disadvantage in their eyes. It was, however, a plan eminently 
 suited for churches where the number of clergy was too small to admit of 
 elaborate processional ritual ; and accordingly was, above all others, the normal 
 plan of the parish churches. The vast majority of the village churches of 
 England have unaisled chancels ; even such large churches as Gedney and 
 Walpole St Peter's, and such important town churches as BOSTON (216.4). 
 But such a plan was unsuitable for, and was seldom adopted, in the cathedral 
 and larger collegiate churches. 
 
 Some churches, however, had already progressed beyond the simple plan 
 of the unaisled chancel. In several churches built in the last half of the twelfth 
 century, aisles were constructed to the western bay or bays of the chancel, 
 leaving to the east an unaisled presbytery. This plan was much more con- 
 venient than that of Kirkstall ; for the unaisled presbytery with the high altar 
 was excellently lighted. It was adopted by the Augustinian Canons of St 
 Frideswide, OXFORD (152.3), 1154 to n8o;t and not much later by those of 
 Lanercost Priory and Cartmel ; c. 1191 by the Premonstratensian Canons of St 
 Radigund's Priory at Bradsole ; and c. 1 190 by the Benedictines of Tynemouth. 
 In the thirteenth century it was adopted by the Benedictines of Rochester and 
 Worcester; and by the Secular Canons of SOUTHWELL (152.2), Beverley (152.1), 
 and Wimborne ; in the fourteenth by the Augustinians of Bristol. The same 
 plan was followed later at Lichfield and at Christ Church, Hants.j 
 
 Fourth Plan : Type, York Minster : Aisled Parallelogram. 
 
 But a much simpler plan than either that of Salisbury or that of Oxford 
 came into use simultaneously with those two plans, and in the end superseded 
 both. Like them it originated in the last half of the twelfth century ; and 
 originated, probably independently, in the South and North; in the South at St 
 Cross and New Shoreham ; in the North at JERVAULX (153.3), Whitby, and 
 Hexham. In these choirs, pillars and arches, clerestory windows, vault and roof 
 ran in undiminished height without break from the central tower to the east end ; 
 
 * Cf. the plans of Fontenay, Casamari, Chiaravalle, Maulbronn, in Dehio, Plates 191-194. 
 It was the most common of all the Cistercian plans ; and probably was the plan of the Clairvaux 
 church in which St Bernard, the greatest man of the order, had worshipped. 
 
 t It is possible that this was the plan also of the church built by the Secular Canons of 
 Wimborne early in the twelfth century. It may even have been the plan of Abbot Richard's 
 presbytery at Ely in 1103. 
 
 I It should be noted that the unaisled member was not usually a presbytery. It was a 
 Lady Chapel at Worcester, Lichfield, and elsewhere. 
 
PLAN OF CHOIR OF YORK MINSTE1 
 
 177 
 
 and choir was separated from presbytery, presbytery from Saint's Chapel, Saint's 
 Chapel from ambulator}-, ambulatory from Lady Chapel, merely by a series of 
 screens, whether of stone or wood ; and the east wall was filled with as many 
 windows as possible, as at ELY (464) : or was made all window, as in Lincoln 
 presbytery. This simple arrangement solved both the lighting and the ritualistic 
 problems. It was the final and definitive solution of the English cathedral builder; 
 
 Lincoln, East Front. 
 
 and was the full and final break with Continental tradition ; a plan which made 
 the later English cathedral and abbey church utterly dissimilar both within 
 and without to the interiors and exteriors of the Gothic of the He de France. 
 Externally, it is a plan which has much grandeur ; a roof of the unbroken height 
 of that of the choirs of Ely or Lincoln, some 160 feet long, is imposing in the 
 highest degree ; internally, it is ruined by destruction of the screens. 
 
 M 
 
1 7 8 
 
 St Albans, Rood Screen. 
 St Albans. Reredos. 
 
 Lavenham. Spring Pew. 
 Newark from SAY. 
 
SCREENS. 
 
 179 
 
 Like the Kirkstall and Oxford plans, it originated in the last half of the 
 twelfth century. It is seen in the church attached to the Hospital of ST CROSS, 
 Winchester (104), founded by Bishop Henry de Blois ; * New Shoreham 
 (c. 1175); Portsmouth (c. 1185); Boxgrove (c. 1235). But it was worked out on a 
 far grander scale in the North, and probably independently of Southern Gothic, 
 in the early years of the thirteenth century. Cistercian Byland had built a 
 long choir, c. 1 177, but with a rectangular ambulatory. Cistercian JERVAULX 
 (153.3) ar| d Benedictine Whitby followed with choirs without ambulatories. 
 Early in the thirteenth century the Augustinian Canons of Hexham built a great 
 choir of this plan. These were followed by the magnificent choir of Cistercian 
 Rievaulx. All four choirs have full length aisles. The new plan soon re- 
 appears in the Cistercian abbeys of the South of England, Netley, c. 1239, and 
 Tintern, 1269 ; in the Benedictine Cathedral of ELV, c. 1235 (153.4) ; it produced 
 the enormous choirs of Old St Paul's, London (dedicated in part in 1240), and 
 the Angel choir of Lincoln, 1256, both of Secular Canons. The influence of the 
 three choirs of Ely, Lincoln, and St 
 Paul's must have been immense. We 
 may add Thornton Abbey,T 1264; 
 Ripon, Guisborough, and Selby (1280- 
 1300) ; Howden and Carlisle, c. 1340 ; 
 the magnificent choir of YORK (153. 1), 
 begun in 1 361; and finally, Bath Abbey, 
 c. 1 500. The cruciform church of York, 
 with full-length aisles on either side of 
 nave, transepts and eastern limb, repre- 
 sents the plan of the English cathedral 
 in its complete and final form. 
 
 Screens. 
 
 The number of screens which 
 might occur in a church of the first 
 rank was considerable. Of these the 
 
 most important was the pulpitum. In the Early Christian basilicas there had 
 been on either side an ambo or pulpit. Pairs of ambos are still retained in most 
 of the Spanish cathedrals, e.g. Toledo and Burgos. In our mediaeval churches the 
 ambos were as it were connected by a broad platform, either end of which was 
 used as an ambo for reading the Epistle and Gospel ; sometimes it contained an 
 altar ; sometimes on it stood a pair of organs. This platform was reached by 
 a staircase. The position of the pulpitum varied. If the stalls were in the nave, 
 the solid pulpitum stood west of the stalls, and therefore some distance down the 
 nave. The pulpitum remains in this position at Gloucester. But where the 
 stalls were in the choir as in Lincoln Minster and Southwell, Rochester,~York, 
 Ripon, Canterbury, the pulpitum was at the west end of the choir. The pulpitum 
 had a central " quire door," leaving room for an altar on either side. 
 
 * The date of the Church of St Cross is uncertain ; probably it belongs to the Bishop's 
 later years ; he was Bishop from 1 129 to 1 171. 
 t Plan in A. A. Sketch Book, 1872. 
 
 Southwell Choir from S.E. 
 
i So 
 
 SCREENS. 
 
 One bay west of the pulpitum, but in monastic churches only, was the rood 
 screen. This also was surmounted by a platform or loft, in front of which stood 
 the great Rood or Crucifix, with the attendant images of St Mary and St John 
 Sometimes the Rood rested on a separate beam, fixed a little above the loft * 
 Unlike the pulpitum, the rood screen had two side doors. This was for the two 
 ranks of the Sunday procession. Between these doors was placed the altar of 
 the laity, usually called the altar of the Holy Rood or of St Cross, or the Jesus 
 
 Christ Church, Hants. Reredos. 
 
 altar. The rood screen still exists at ST ALBANS (178). The Jesus altar seems 
 to have been protected by a wooden screen west of it, as at Dunstable. 
 
 East of the choir was the sanctuary or presbytery. At ST DAVID'S (162) 
 is a low open screen of wood between choir and presbytery.f 
 
 At the east end of the presbytery was the high altar, behind which in later 
 
 * The sawn offends of the rood-beam may occasionally be seen. In the parish churches 
 the rood-beam not infrequently remains ; e.g. at the east end of Knapton and Sail naves, 
 t There remain the supports of another at Michaelchurch, Hereford. 
 
SCREENS. 
 
 |S| 
 
 days there was often a reredos, which had two side doors in order to allow the 
 priest to pass completely round the high altar while censing at High Mass. 
 The reredos seems originally to have been low and inconspicuous, but in the 
 fourteenth century at CHRIST CHURCH, Hants (180), and later at ST ALBANS 
 (178), Winchester, and St Saviour's, Southwark, it became lofty and magnifi- 
 cent. The reredos or retablo is the special glory of the Spanish cathedrals.* 
 
 If there was a Saint's Chapel, as of St Alban at ST ALBANS (186), this 
 also was fenced off to the east from the processional path and the eastern 
 chapels by a screen, as it was to the west by the reredos of the high altar. 
 
 Finally, if there was an eastern Lady Chapel, this was separated from the 
 processional path or ambulatory by a screen west of it, as at OTTERY ST MARY, 
 Devon. 
 
 At YORK (153.1) the eastern limb consists of nine bays; which, counting from 
 the east, were distributed as follows. The easternmost bay was the Lady Chapel. 
 This was separated by a screen from the ambulatory, to which two bays were 
 assigned. The ambulatory was 
 separated by a screen from the 
 Saint's Chapel, that of St William 
 of York, which occupied one bay. 
 Then came the existing reredos 
 screen east of the presbyter} - , 
 which occupied two bays. West 
 of the presbytery was the choir, 
 occupying three bays, and screened 
 off from the nave by the still 
 existing pulpitum.t 
 
 It should be added that the 
 sides also of the choir and sanc- 
 tuary were guarded with screens. 
 In the parish churches, e.g. LYNN 
 
 ST MARGARET (162), Collumpton, 
 
 SUDBURY" ST PETER ("162), these side or parclose screens were usually of oak. 
 In the greater churches, as at Canterbury, Selby, Exeter, Winchester, they were 
 often of stone. 
 
 None of the larger screens in England or in France are earlier than the 
 thirteenth century ; and from this VioIlet-le-Duc drew the conclusion that the 
 
 IXMKKKXNXMKKV 
 
 Ottery St Mary, Lady Chapel. 
 
 * On the Reredos see paper by Mr T. Garner in Assoc. Societies' Reports, xvi. 136. 
 
 + See Willis' York, Plan 5. It should be noted that the screen at the entrance of a 
 parochial chancel corresponds both to the pulpitum and the rood screen of the monastic 
 churches ; it often carried the rood, and was also used to put lights on. Many of these 
 chancel screens survive ; especially in East Anglia, Devon, and West Somerset, and the Welsh 
 Border. Usually they are of oak, as at NEWARK (178) ; but occasionally of stone ; e.g. at 
 Compton Bassett, Wilts. (Illustrated in Weale's Quarterly Papers, vol. i.) On either side 
 of the central door, if there were no aisles, there was an altar. Recesses foi these side altars, 
 one Norman, the other Early English, remain at CASTLE rising (102). At Ranworth the 
 painted wooden reredoses of the side altars remain. On the Devonshire screens, see Mr 
 F. Bligh Bond's illustrated paper in Transactions of Devonshire Association, 1902. Some 
 hundred and fifty screens remain in Devon. 
 
182 
 
 SCREENS. 
 
 thirteenth-century cathedral differed from the older monastic type in being 
 laic ; * in having no barrier between priest and layman ; an idea which has 
 
 been carried out in several 
 modern French restorations. 
 From the first, however, low 
 screen walls doubtless sur- 
 rounded choir and sanctuary. 
 At Ely, indeed, the original 
 Norman pulpitum forming the 
 western termination of the choir 
 was still standing in 1757/f It 
 occupied the whole of the eastern- 
 most bay of the present nave ; 
 so that there was one bay east 
 of it before the construction of 
 the present octagonal crossing. 
 Moreover the monk Gervase, 
 describing what he had seen 
 before the fire of 1 174 at Canter- 
 bury, says that " at the base of 
 the pillars was a wall of marble 
 slabs ; which, surrounding 
 Ernulph's choir and presbytery 
 (built 1096-1115), divided the 
 church from its sides, which are 
 called alae." 
 
 Chantry chapels, and later, 
 
 family pews, e.g. that of the 
 
 Spring family at LAVENHAM 
 
 (178), were also screened off. The eastern bay of each aisle was almost always 
 
 screened off as a chapel, with an altar of its own, which, in the unaisled church, 
 
 had stood in each eastern corner of the nave. 
 
 * " Viollet-le-Duc tried to explain the appearance of Gothic by the substitution of laymen 
 for monks in the direction of architectural work, by the triumph of the laic over the monastic 
 spirit. It is a figment" (Comte de Lasteyrie, Discours, 14). 
 
 + Willis' Canterbury, 43 ; and Stewart's Ely, 43. 
 
 Abbey Dore from East. 
 
Chapter X. 
 
 PLANNING OF THE GREATER CHURCHES. 
 
 Ritualistic Divisions of Church — Length and Position of Choir -Saint's Chapel — 
 Eastern Transept — Crypt. 
 
 It has been shown how the various ritualistic divisions of the church were 
 marked off by screens. In the earlier work the architecture itself sometimes kept 
 distinct the ritualistic divisions of the greater churches : as is well seen at 
 Winchester and ST ALBANS (153.2). In the latter, first, from the west, comes 
 the nave, occupying ten bays ; the tenth forming the sanctuary of the Jesus altar. 
 The remaining bays of the nave, together with the crossing, formed the choir * 
 of the monks. Then, east of the central tower, three bays formed the sanctuary 
 of the high altar ; at the back of this is the great stone reredos. Behind this 
 reredos is the Saint's Chapel or Feretory, containing the shrine of the great local 
 saint ; that of St Alban at St Albans ; of St Swithin at Winchester ; of St 
 Thomas the Martyr at Canterbury ; of St Cuthbert at Durham ; of St Werburgh 
 at Chester.-)- Next comes the ambulatory. Beyond that, architecturally distinct, 
 comes the projecting Lady Chapel. 
 
 These then were the divisions marked out either by stones and mortar or 
 by screens in a great mediaeval church. It remains to treat of each in order. 
 First we will turn to the choir. 
 
 The Choir. 
 
 In the earliest Christian churches, e.g. ST MARIA MAGGIORE ( 148.2 , there 
 were but two parts, a nave and sanctuary ; there was no architectural choir. 
 The sanctuary occupied the apse, and the apse was joined immediately to 
 the nave ; or, in the double-aisled basilicas of the fourth century, such as 
 those of St Peter and of St Paul at Rome, to the transept ; there was no 
 interposition of a choir between nave and apse. The choir was simply the 
 east part of the nave, and was fenced off by low walls, usually of marble, 
 carved or perforated with interlacing patterns, peacocks (the symbol of 
 immortality), lions, doves, &c. Many specimens of these screen-walls remain, 
 e.g. in S. Vitale, Ravenna, and Ancona Cathedral.] This arrangement is well 
 seen in S. Clemente, Rome; as rebuilt in 1108 with the material of the under- 
 ground basilica demolished by the Norman, Robert Guiscard, in 1084. 
 
 * " Since the close of the seventeenth century quire has been fictitiously spelt choir : but the 
 spelling quire has never been altered in the English Prayer Book" {New English Dictionary). 
 
 t At St David's, Rochester, and Oxford Cathedrals, the shrines of the local saints were 
 placed north of the choir. 
 
 I These walls were called cancclli ; hence our English word "chancel." 
 
1 84 LENGTH OF CHOIR. 
 
 But in all the larger mediaeval churches a choir is interposed between 
 presbytery and nave.* Such a choir is clearly seen in the ninth-century plans of 
 ST GALL (194), and of Notre Dame de la Couture, Mans.f In the earlier 
 churches of Normandy, such as Bernay and CERISY (148.3), it is but short; 
 only two bays. St Stephen's, Caen ; Lanfranc's Canterbury ; Selby and Lincoln 
 resembled Cerisy in having choirs of two bays. But the greater part of our 
 eleventh-century cathedrals and abbeys set out choirs with more bays ; usually 
 with four: e.g. BURY (150.3); St Augustine's, Canterbury; DURHAM (149.1); 
 Ely; Peterborough; St Albans; WINCHESTER (192.5); NORWICH (148.4). 
 And, as all the above were Benedictine, and the monks usually sat in the crossing 
 or the eastern nave, the apsidal eastern limb of some four bays provided a 
 very dignified presbytery ; occupied only by the high altar, and entered only by 
 the priest and his attendants officiating at the Mass. 
 
 Lanfranc's choir at Canterbury, of two bays, was a very unworthy sanctuary 
 of a church which was at once the chapel of a large Benedictine monaster}- 
 and the church of an Archbishop and Primate. It was completed in 1077 ; 
 but in 1096, only nineteen years after, Prior Ernulph commenced to build 
 a new eastern limb (149.2), vast in scale, and with all the latest improvements in 
 planning, such as were to be seen already at Gloucester and Cluny. It con- 
 tained an apse and a choir of no less than nine bays ; and must have been 
 at the time by far the longest choir in Western Christendom; even surpassing 
 Cluny, whose enlarged and remodelled choir had been consecrated the previous 
 year. The chief reason for this great eastern extension no doubt was that the 
 monks were very much cramped in Lanfranc's short nave of nine bays. How 
 short it was in comparison with other Norman naves in the South and East of 
 England may be seen from the following table : — 
 
 Canterbury St Augustine's - 1 1 bays Ely - - - 1 3 bays 
 
 Peterborough - 11 „ St Albans - - 13 „ 
 
 Bury - - - 12 „ Norwich - - 14 „ 
 
 Winchester - - - 12 „ 
 
 while Bury, Ely, and Peterborough had western transepts in addition. It was 
 impossible to extend Lanfranc's nave to the west ; for the new western towers 
 were in the way; the only extension possible was eastward. Ernulph's vast 
 choir of Canterbury, half choir proper, half sanctuary, turned the current of 
 English planning. After this, with the exception of plans of foreign ex- 
 traction, such as those of the earliest Cistercian churches, e.g. Kirkstall 
 and the French plan of Westminster, the eastern limb of the English church 
 was of vast length. The new position of the clergy east, instead of west, 
 of the eastern arch of the central tower, was recommended not only by con- 
 venience of accommodation, but by the enhanced dignity which it gave to 
 the monks and canons. Nevertheless there had been no alteration in the 
 relative position of the clergy^ and the altar. In the earliest Christian churches 
 the altar had been placed at the west end. The clergy, therefore, as at S. 
 CLEMENTE, Rome (3), in order to face east, had to be seated west of the altar. 
 
 * A choir of one or more bays was not interposed in front of the apse till c. 900 (Quicherat 
 and Comte Robert de Lasteyrie). 
 + Dehio, Plate 119,7 and 7A. 
 
SAINT'S CHAPEL. tSs 
 
 When it became customary to place the altar at the east end, the clergy retained 
 their seats west of the altar. The position of the altar was changed, not that of 
 the clergy. 
 
 Hardly was Canterbury choir completed when Glastonbury built a choir of 
 six bays ( 1 1 86) ; to be enlarged to eight bays in the fourteenth century. Where 
 the naves were exceptionally long, as at NORWICH (148.4), Peterborough, ELY 
 (153.4), ST ALBANS (153.2); or where the choirs were short, as in the early 
 Cistercian churches, the Benedictine abbeys of GLOUCESTER (135) and WEST- 
 MINSTER (15 1.2), and the cathedrals of the Secular Canons of Chichester and 
 Hereford, the stalls were allowed to remain in the east nave ; or in the crossing 
 and east nave ; or else in the crossing and west choir. But in by far the greater 
 number of churches, no matter by whom served, the stalls were placed in new 
 choirs which were built in Gothic. This was especially so with Canons' 
 churches. In the end we find nearly all the Canons, Secular or Regular, housed 
 in new Gothic choirs : Augustinian Canons in the thirteenth century at Hexham, 
 Southwark, Thornton, Carlisle, and Guisborough ; in the fourteenth century 
 at Bristol ; in the fifteenth century at Christchurch ; Secular Canons in the 
 twelfth century at York ; in the thirteenth century at Wells, Lincoln, Lichfield, 
 Salisbury, St Paul's, Exeter ; all seven cathedrals. In collegiate churches, the 
 Secular Canons enlarged or built choirs In the twelfth century at Ripon ; in 
 the thirteenth at Beverley and Southwell ; in the fourteenth at Howden. The 
 example of the canons was largely followed by the monks also. Even the 
 Cistercians built long Gothic choirs at Rievaulx, Fountains, and Tintern ; and 
 the Benedictines at Canterbury; Glastonbury; Rochester; Worcester; Whitby; 
 Boxgrove ; St Werburgh, Chester ; Pershore ; and Selby. Many others might 
 be named. Enough have been enumerated to show how widespread was the 
 tendency to remove the stalls eastward ; a tendency which was to make the 
 elongated eastern limb of the greater English churches so strikingly dissimilar 
 to such plans as those of Amiens and Westminster, and to constitute one of the 
 most marked differences between English and Continental Gothic architecture. 
 
 Saint's Chapel. 
 
 We have seen that it was at Canterbury in 1096 that the great prolongation 
 of the eastern limb * commenced. Strangely enough, it was Canterbury that was 
 to set an example of yet further prolongation eastward. For when the great 
 
 * There is much ambiguity about the terms Choir- and Presbytery. Strictly speaking, the 
 Choir is that part of the church where are the stalls of the clergy. (1) As these stalls may be 
 in the east nave, e.g. at Westminster, it may occur that no part of the choir is in the eastern 
 limb. (2) If the stalls are in the eastern limb, the choir will occupy its western bays ; i.e. the 
 space between the crossing and the sanctuary ; e.g. at Canterbury. But (3) the term Choir is 
 also used loosely of the whole of the eastern limb ; including choir proper, sanctuary, retro- 
 choir, &c. 
 
 The Presbytery (1) is the space between the choir and high altar. It was raised on steps, 
 gradus presbyterii ; and there were doors from it into the north and south aisles, ostia presby- 
 terii. (2) But the term is also used loosely of all the space in an eastern limb between the choir 
 and the eastern wall ; e.g. at Ely, Lincoln, and York, fin non-collegiate churches the eastern 
 limb is called the Chancel, the eastern portion of which is the presbytery or sanctuary.) 
 
1 86 
 
 SAINT'S CHAPEL. 
 
 choir of nine bays built by Priors Ernulph and Conrad between 1096 and 1115 
 was burnt down, instead of its central eastern apse there were constructed four 
 more bays eastward. But this prolongation was not made for the same reason as 
 before. It was not to provide more accommodation for the monks ; they had room 
 enough. It was to provide east of the high altar space for the shrine of the great 
 local saint, St Thomas the archbishop, martyred in 1 170, and now the most famous 
 saint in Western Christendom. After the murder, following an ancient usage 
 which goes back as far as the burial of the bodies of the first Christian martyrs 
 in the catacombs of Rome, the archbishop's remains had been deposited below in 
 the crypt. But they were resorted to, for their miraculous powers, by thousands 
 
 Saint's Chapel, St Albans. 
 
 and tens of thousands of pilgrims. Ernulph's crypt was vast, and had escaped 
 damage by the fire ; but even this great crypt was inadequate to accommodate 
 the multitudes who desired to pass round the shrine, to touch if it might be the 
 sacred relics, and to say one prayer before them. It was determined, therefore, 
 to remove the relics from the crypt, and to build a Saint's Chapel or Feretory 
 for them east of the high altar. Into this Saint's Chapel* which at Canterbury 
 has usurped the name of the older Trinity Chapel, in which St Thomas had been 
 used to officiate, the body of the martyr was removed, and was deposited in a 
 
 * Mr St John Hope has shown that the so-called Trinity Chapel was the Chapel of St 
 Thomas, and that the Corona was the Trinity Chapel, and was designated " Trinity Chapel 
 ad coronam." 
 
1 8; 
 
 \ 1 
 
 Btitf+n 
 
 
 f 
 
 
 i 
 
 f? 
 
 il Till ffi 
 
 
 '^1 
 
 r 
 
 1 
 
 V . i 
 
 1. ...rr. ../■? 
 
 1 1 
 
 Pontigny. 
 
 Hereford. 
 
 Pontigny. 
 St David's. 
 
i88 EASTERN TRANSEPT. 
 
 magnificent shrine, amid a concourse of kings, princes, and ecclesiastical digni- 
 taries from all over Western Europe. This shrine has utterly vanished ; but we 
 have considerable remains of others ; but, except at Westminster, of the pedestal 
 only. The other shrines, which were usually of wood, were broken up in all cases 
 for their jewels, gold, and silver, by the commissioners of Henry VIII. The 
 most important of these pedestals are those of St David at St David's, thirteenth 
 century ; of Edward the Confessor at Westminster, 1269 ; of Thomas Cantilupeat 
 HEREFORD ( 1 87) and St Frideswide at Oxford, both c. 1290 ; of St Alban at ST 
 albans,* 1 302- 1 308 (186); of StEtheldreda at Ely f and St Werburgh at Chester, 
 both c. 1340. We may reconstruct, to some extent, the appearance of these 
 shrines by visiting PONTIGNY (187), between Sens and Auxerre, the largest 
 of the Cistercian churches in Europe. Here there is a shrine to Edmund Rich, 
 Archbishop of Canterbury, who died near PONTIGNY (187) in 1240. His body 
 began to work miracles, and he was canonised in 1246. The shrine, a work of the 
 eighteenth century, stands considerably back from the high altar, and towers much 
 above it. So must our English shrines have looked, elevated on high, behind and 
 above the high altar, conspicuous far down the church. At the back of St Edme's 
 shrine are the staircases by which the pilgrims ascend to and descend. In 
 the shrine of ST DAVID (187) are holes through which the pilgrim could insert 
 a diseased limb for healing. At Westminster a still more ancient Pagan and 
 Christian custom survived ; for in the shrine of Edward the Confessor \ there 
 are niches in which the sick were left for the night in hope of cure. 
 
 From Canterbury the new fashion passed in a few years to Chichester ; 
 where, after a great fire, a retrochoir was added in 11 86. Here in later days 
 stood the shrine of St Richard (died 1253) ; a harper used to play and sing the 
 praises of the saint. § 
 
 Eastern Transept. 
 
 Yet another change was made to increase the convenience of the eastern 
 limb. It was the addition of an eastern or choir transept. For the third time 
 it was Ernulph's CANTERBURY that led the way. But (149.2) in Ernulph's 
 choir the pier-arches ran right across its entrances; so that, internally, it was 
 a masked transept.H This peculiarity seems to connect it with the transeptal 
 basilicas of Rome, e.g. St Paul extra muros and Sta Prassede, rather than with 
 the eastern transepts of Cluny, St Benoit-sur-Loire, or Souvigny. This eastern 
 transept of Ernulph was rebuilt by William of Sens in 11 80, with the omission 
 of the pier and two arches which had previously masked each of its arms. To 
 the east of each arm he built two semicircular apses. His eastern transept was 
 
 * On the right of the photograph of the Saint's Chapel at St Albans is a glimpse of part of 
 the three narrow arches leading eastward to the ambulatory ; in the foreground is the pedestal 
 of St Alban's shrine ; to the left is the Watching-Loft ; and, above it, an arch leading into the 
 north aisle. 
 
 t For this pedestal see Index of Illustrations. 
 
 t See article and illustrations by W. Burges in Scott's Gleanings, page 127. In a woodcut 
 on page 136 a sick person is seen crawling into the pedestal of the Confessor's shrine. 
 
 § Rev. T. Hugo. 
 
 || So also probably at Exeter; see Professor Lethaby's plan of Norman Exeter in 
 Architectural Reviciv, April 1903. 
 
EASTERN TRANSEPT. 189 
 
 copied, apses and all, in St Hugh's choir at LINCOLN (151. 1) in 1192. Simul- 
 taneously with Ernulph's work at Canterbury, there was going on a great 
 remodelling of the eastern limb of Lewes,* 1091-1098. Lewes was a daughter- 
 church of CLUNV (150.1), and copied not only the eastern transept with eastern 
 apse, but also the five tangential chapels of Cluny ; elsewhere English Romanesque 
 was content with three. About 1200, eastern transepts were set out by the 
 Benedictines of Rochester and Worcester, which show much similarity in the 
 plan of the eastern limbs; and by Secular Canons in 1220 at SALISBURY (154.2), 
 and a little later at BEVERLEY MINSTER (152.1). The Rochester plan also 
 appears c. 1300 in the Premonstratensian church of Bayham, Sussex. In all 
 these cases the transepts rose to the full height of the choir, and were invalu- 
 able both for internal and external effect. Internally they added shadowy 
 recesses and half-seen, mysterious distances. Externally they were of the very 
 greatest value in breaking up the exceptionally long, monotonous horizontal lines 
 of the elongated choirs of English Gothic. 
 
 Elsewhere the choir transept took a humbler form ; it was satisfied to be a 
 projection to north and south, not of the choir, but of the choir aisles. Judging 
 from the foundations in the crypt, something of the sort had been built by 
 Archbishop Roger at YORK (199), c. 1160. And even when his choir was 
 demolished and rebuilt, c. 1361, this secondary type of eastern transept was pre- 
 served, and adds immensely both to the external and internal effect of the long 
 Perpendicular choir. So also at SOUTHWELL, c. 1230 (152.2), at EXETER, c. 
 1280(154.4), and at wells, c. 1340 (154.3), l° w eastern transepts were thrown 
 out from the aisles of the choirs. 
 
 Two noteworthy departures were made in the setting out of the new eastern 
 transept. At FOUNTAINSf (150.2) it was placed not across the choir but to the 
 extreme east of it. This disposition was copied a few years later at Durham. In 
 both, the eastern transept goes by the name of the Chapel of the Nine Altars ; a 
 name which sufficiently explains its destination. 
 
 As we have seen, it was at St Benoit-sur-Loire, Cluny, and Souvigny that 
 the eastern transept first appeared. It is strange that, with a few exceptions, 
 such as Verdun, Besancon, and St Ouentin, it found no welcome in the 
 later Romanesque or in the Gothic of France. It appears sporadically else- 
 where; e.g. at Nivelles, Ferrara, and Milan; and in the Liebfrauenkirchen 
 at Treves. But in the thirteenth-century Gothic of England it is one of the 
 noblest and most characteristic features. Even in English Gothic, however, 
 its reign was short. In its grandest form, at full height, it commences at 
 Lincoln in 1192, and, with the exception of Bayham, ceases c. 1240. 
 
 What was the object of this choir transept? Probably, as the apses and 
 aisles on the eastern side show, it was to provide more altared chapels ; each 
 apse or bay of the aisle serving as a choir, and the central space of the 
 transept as a nave. At Rochester, however, the north transept was certainly 
 appropriated from the first for the shrine of the local saint, a Scotch baker, 
 William of Perth, who was murdered near Rochester in 1201, and buried in the 
 
 * See plan in Mr St John Hope's paper read to Royal Archaeological Institute, Aug. 1, 
 1883. 
 
 + At Fountains it was finished in 1247. 
 
i go THE CRYPT. 
 
 cathedral " miraculis choruscando" "amid a coruscation of miracles."* No 
 doubt the offerings at his shrine helped to pay for the great eastern extension at 
 Rochester. 
 
 The Crypt. 
 
 Beneath several of our choirs is a crypt. The crypt has a long and vener- 
 able history, which goes back to the early days of Christianity at Rome. Both the 
 Jew and the Christian were treated, on the whole, with much toleration at Rome ; 
 in spite of brief periods of persecution, such as that of Decius, 249 to 251 A.D., 
 and that of Diocletian, 303 A.D. They were allowed to form Burial Clubs, like 
 the Romans themselves ; the)' were at liberty to have cemeteries of their own ; 
 and to dispose of their dead as they thought fit. The only restriction imposed 
 was that no one, Pagan, Jew, or Christian, might bury within the city. Even 
 before the introduction of Christianity, the Jews had cemeteries in the catacombs 
 of the Roman suburbs. The Christians followed the Jewish mode of burial, not 
 employing cremation, the general practice in Pagan Rome. Now in the suburbs 
 of Rome there is a variety of volcanic deposits ; one of sand, which gives the 
 pozzolana pura, of which the famous Roman mortar was made ; another, the tufa 
 litoide, a very hard building stone ; a third, the tufa granolare, not so hard as 
 the last, nor so soft and crumbling as the first. This stratum was worked just 
 as coal nowadays, or Caen stone. Shafts were sunk, which afterwards served 
 both as staircases and air-holes ; and when the tufa granolare was reached, hori- 
 zontal galleries were driven all round the area of the proposed cemetery ; and, 
 later, cross-galleries as well. In the walls of these galleries, which were usually 
 only 2-i ft. to 3 ft. wide, were cut horizontal niches, much like the tiers of berths 
 in a ship's cabin ; each usually holding a single corpse. Such a niche was called 
 a locus ; nowadays it is designated loculus. In the case of richer people a solid 
 stone coffin was excavated out of the side of the gallery, closed with a horizontal 
 slab ; and an arch was hollowed out above it ; this is an arcisolium. But in the 
 case of still more important personages, chambers, aibiada, were hewn out of the 
 rock at right angles to the gallery, arranged like the bedrooms leading out of an 
 hotel corridor. Opposite the entrance of such a cubiculum was the rectangular 
 raised tomb ; a sarcophagus of solid rock, closed with a stone slab. This slab or 
 mensa was used as an altar. Prudentius, towards the end of the fourth century, 
 says, " The same table gives the Sacrament, and is the faithful guardian of the 
 bones of the martyr (Hippolytus) ; which it keeps laid up there in expectation of 
 the Eternal Judge, while it feeds the dwellers on the Tiber with holy food." The 
 cubiculum or burial-chamber of such a martyr was called a Confessio, where lay 
 one who had confessed and given witness to his faith by his blood.T It is 
 improbable that these cubicula were used for purpose of worship, except for the 
 Eucharistic service on each anniversary of the burial. But from c. 350 the graves 
 of the martyrs in the catacombs received much attention. The Bishops of Rome, 
 especially Damasus, had the more important graves marked by inscriptions, 
 improved the means of access, and constructed shafts to light and ventilate the 
 
 * Hope's Rochester, 40. 
 
 t The term " Martyrdom," applied to the north transept at Canterbury, is an exact 
 equivalent to " Confessio." 
 
THE CRYPT. 191 
 
 cubicula. For one or two centuries they were greatly frequented by pilgrims 
 from all Christendom. St Jerome says that when he was a boy, c. 364 A.D., he 
 used to go to the catacombs every Sunday, "to visit the tombs of the Apostles 
 and martyrs, and to go into the crypts there excavated in the bowels of the 
 earth." 
 
 What was done at Rome set a precedent for Christendom in general. At 
 Rome the graves of the martyrs had not been dug in the surface soil, but deep 
 down in a subterranean gallery. Outside Rome, therefore, where there was no 
 subterranean gallery, churches, which possessed distinguished martyrs, built 
 crypts, that should be reminiscent of the cubiculum or confessio of the Roman 
 catacombs. There is an early example at Ravenna at S. Apollinare in Classe ; 
 A.D. 534. At first these were sometimes as deep sunk as the Roman " cubicula " 
 themselves ; e.g. at St Germain, Auxerre, and at Chartres Cathedral. Or, they 
 they were but partly above ground, and were lighted by small windows placed 
 in their side walls ; e.g. Ernulph's crypt at Canterbury. Occasionally their 
 floor was but little below the surface of the ground, as in the eastern crypt at 
 Canterbury ; or was even on a level with the pavement of the nave, as in 
 S. Miniato, Florence. In these latter cases the crypt practically became a 
 second or lower church ; e.g. St Faith's, under Old St Paul's, London. Such a 
 crypt, however, entailed a raised choir ; hence it is that one ascends high flights 
 of steps to such choirs as those of S. Miniato ; Rochester ; and Canterbury.* 
 
 That the connection of the mediaeval crypt f with the early Christian 
 " confessio " is no mere fancy is clear from the words of Edward the Chanter 
 (quoted by the monk Gervase ; see Willis' Canterbury), who expressly says 
 that the crypt under the Anglo-Saxon Cathedral of Canterbury was copied 
 from the Confessio of St Peter at Rome; "ad instar confessionis Sancti Petri 
 fabricata." 
 
 In our own country we have two examples undoubtedly of the seventh 
 century; the crypts at Ripon and Hexham,* both built by Wilfrid between 671 
 and 678. Another, later than these, occurs at Wing, in Buckinghamshire ; with 
 an archaic barrel-vault, possibly added later. All three consist of a walled 
 cubiculum, with a gallery round ; not on one side only, as in a Roman catacomb. 
 Originally, probably, they had separate entrances and exits § on either side of 
 the chancel arch. Later still, but yet Anglo-Saxon, is the pillared crypt of 
 Repton ; from which it is but a short transition to the eleventh-century Norman 
 crypts of Lastingham, WINCHESTER (192.5), WORCESTER (192.2), Rochester, 
 GLOUCESTER (192.1) ; CANTERBURY (193) ; which latter was again lengthened 
 after the fire of 1174; and to the twelfth-century crypts of St Peter in the 
 East, Oxford ; and York Minster.!) 
 
 With such a long pedigree, Roman, Gaulish, and Anglo-Saxon, it is a little 
 
 * S. Apollinare in Classe had a crypt and raised choir from the sixth century. 
 
 t St Jerome used the terms "crypt" and "catacombs" as synonymous in the passage 
 quoted above. 
 
 \ On the Anglo-Saxon crypts see Baldwin Brown's Arts in Early England, ii. 263 seq. 
 
 •5 Cf. the crypts of St Peter in the East, Oxford ; and Madley. 
 
 || The latter is of various dates ; some of it built with re-used material. See plan in Builder 
 Cathedrals. 
 
192 
 
 Crypts. 
 
 i. Gloucester. 4 . St John's Chapel, Tower of London. 
 
 2. Worcester. 3. St Martin de Tours. 5. Winchester. 
 
THE CRYPT. 
 
 '93 
 
 surprising that the crypt did not come into still more general use. Most of our 
 churches, however, whether great or small, have no crypt and never had. It 
 might be thought that it was where there were no relics buried of a famous 
 local saint that the crypt was omitted. But in many churches where the local 
 saint, e.g. St Etheldreda, St Cuthbert, St Alban, was held in the highest repute, 
 there was never a crypt. Nor does the objection to building deep down in a 
 watery soil explain the omission ; for the cathedrals of Ely, Durham, St Albans 
 all stand high ; whereas the crypt of Winchester Cathedral is below the level 
 
 Canterbury Crypt. 
 
 of the neighbouring stream. It may be that the omission and disuse of the crypt 
 were largely due to Cluniac example.* 
 
 In the end, probably because the number of pilgrims to the more noted 
 shrines, e.g. to that of St Thomas the Martyr at Canterbury, passed beyond what 
 even the very largest crypts could contain, the more important shrines were 
 removed from the crypt and placed on the pavement of the presbyter)-, behind 
 and above the high altar; in such a position as is seen in PONTIGNY (187). 
 Thus at Canterbury the body of St Thomas was translated to the chapel above 
 in 1220, after lying in the crypt for fifty years. So in France Abbot Suger had 
 
 * Dehio remarks that the omission of the crypt was especially characteristic of the Cluniac 
 churches. 
 
194 
 
 THE CRYPT. 
 
 translated the body of St Denis and his two companions in 1 144 from the crypt 
 of St Denis. The crypts, therefore, deserted by the great saints, lost their value, 
 and no more were built ; except in later days the crypt of the present St Paul's, 
 destined to receive the remains of great generals and admirals. 
 
 In addition to the above there were others of quite different purpose. 
 Some were merely constructional ; e.g. in France at Bourges the east end of 
 the cathedral, and at Madley * in Herefordshire the fourteenth-century east end 
 of the church, were built on steep declivities ; and the substructures necessarilv 
 took the form of crypts. Moreover, it was not uncommon to excavate an 
 Ossuary or Golgotha. One remains at Norwich, a little west of the cathedral ; 
 formerly there was one north of Worcester Cathedral ; the crypt at Hereford may 
 be a Golgotha. That at Hythe till lately was filled with bones. Both at Hythe, 
 however, and at Wimborne Minster, the crypt may be but a subway to give 
 communication between the choir aisles without crossing the sanctuary. 
 
 * But here no doubt, as often elsewhere, a crypt, constructional in origin, was utilised for 
 ritualistic purposes. 
 
 I a l D □■ u 
 
 I— ■ r # * #— 1 
 
 
 • FO * 
 
 "F3 ' 
 
 • ra • 
 
 TFJ « 
 
 • © • 
 
 FT 
 
 FT 
 FT 
 FT 
 
 SCALE OF 
 
 St Gall, showing Apse. 
 (Same scale as Pla?is of Parish Churches) 
 
Chapter XI. 
 
 THE CENTRAL TRANSEPT. 
 
 Early Transepts — Position and Object of Central Transept — Enlargement of Transept. 
 
 All the greater churches in Normandy and England, and many of the smaller,* 
 seem to have been cruciform ; i.e. the nave and choir, which ran east and west, 
 were crossed by another arm running north and south ; the intersection of the 
 four arms is what is called the " crossing." f The " crossing " is generally covered 
 with a central tower. This sometimes was low ; e.g. Winchester Cathedral ; but 
 often rose to a great height, and was richly ornamented ; e.g. at St Albans, 
 Tewkesbury, Castor, Norwich. The question naturally arises, " Was the central 
 tower anterior to the transept, or the transept to the central tower?" It has 
 been argued that the four walls of a central tower rest on four arches, which 
 again, in the larger churches, rest not on solid walls, but on piers. Now these 
 arches, loaded with the weight of the tower walls, exercise enormous thrusts. 
 To the west and east these thrusts are met by the walls or by the piers and 
 arches on either side of nave and choir. But to the north and south they 
 would have no abutment if there were no transept. It therefore would seem 
 that the transept is but a structural necessity due to the pre-existence of a 
 central tower. 
 
 Chronologically, the very reverse is the case. Transepts occur many 
 centuries before central towers. They appear in all the double-aisled basilicas 
 of Rome in the fourth century of our era. The basilica at SILCHESTER (21 5.1), \ 
 which is very much like a church in plan, has a transept ; it cannot be later than 
 the end of the Roman occupation of Britain. There still exist in Syria the 
 remains of the immense cruciform church of St Simeon Stylites,§ built between 
 459 ar >d 560. At Como, beneath the present eleventh-century church of St 
 Abbondio, [| there have been found the complete foundations of a church with 
 an apse attached almost directly to the transept ; it is held to be of the fifth 
 century. At St Denis, as at Como, the church was originally a crux cow- 
 
 * E.g. our Pre-Conquest churches of Deerhurst ; Worth ; and St Mary's in Dover Castle ; 
 and such Norman churches as North Newbald, Yorkshire ; St Martha's Chapel, Guild- 
 ford ; Melbourne ; St John, Devizes ; Hemel Hempstead. 
 
 t The term " transept " is used indifferently either of the whole arm or of the part which 
 projects from the crossing, north or south. In French the former is termed " transept," the 
 latter " croisillon." 
 
 { The plan of Jataghan, an early Christian church in Asia Minor, is very similar to that 
 of Silchester. It is shown in Lethaby's Mediaval Art, 25. Compare also the plan of S. Maria 
 Antiqua in the Roman Forum. 
 
 § Plan in Scott's Essay, 60. 
 
 || See De Dartein, 312 ; lieito's monograph, Milan, 1868 ; and plan in Dehio, Plate 66. 
 
196 THE TRANSEPT. 
 
 missa, not a Latin cross, crux immissa ; i.e. as in the Constantinian basilicas, 
 the apse was attached directly to the transept without the interposition 
 of a choir : this Viollet-le-Duc assigns to the sixth century ; * Professor 
 Baldwin Brown gives the date of 628.+ In the sixth century the church which 
 Namatius caused to be built at Clermont had a transept ; so also the prede- 
 cessor of St Germain-des-Pres, and of St Martin de Tours, \ which belong to 
 the same period. The Spanish church of St John de Bonos, Palencia, has two 
 small square transeptal projections opening opposite the east bay of the nave 
 arcade ; it has an inscription set up in 661. In the ninth century the transept 
 begins to be common ; e.g. in the plan of ST GALL (194); at Hersfeld ; § at 
 St Generoux ; where, as in our Pre-Conquest churches, the transepts are 
 lower than the nave. Instances of small Anglo-Saxon transepts are given 
 above ; and it is clear from the descriptions that have come down to us that 
 some at any rate of their greater churches had central towers and transepts ; 
 e.g. the church of Ramsey is described as having two towers, one western, 
 the other " in the centre of the square, standing upon four columns con- 
 nected by arches stretching from a/a to a/a, i.e. from transept to transept." 
 This was written in the last half of the tenth century. || In France and Italy, 
 however, transepts are exceedingly rare till the eleventh century. They seem 
 to be a special mark of the German and Anglo-Saxon schools of Primitive 
 Romanesque. 
 
 When, however, transepts were introduced, there arose at once a difficulty in 
 the roofing of the church, where the span-roof of the transepts met the span- 
 roofs of nave and choir ; a difficulty which was all the greater when the transepts 
 and their roofs were lower than the nave and choir. The simplest remedy was 
 to raise low walls on the four arches of the crossing, and to run all four roofs up 
 to these four walls. This, then, is suggested If as the origin of the central 
 tower ; and at the same time as the reason why, at first, in Norman churches at 
 any rate, it was so low. 
 
 In England transepts were the rule from the first in the greater Norman 
 churches. Their length was out of all proportion to the needs of northern and 
 southern abutment to a central tower. Some of the internal lengths ** are : — 
 
 St Albans 
 
 175 feet. 
 
 Winchester 
 
 209 feet 
 
 Ely - 
 
 185 „ 
 
 York - 
 
 222 ,, 
 
 Peterborough - 
 
 185 „ 
 
 Lincoln 
 
 223 „ 
 
 Reading 
 
 19° » 
 
 Bury - 
 
 234 ,. 
 
 Glastonbury - 
 
 192 ,, 
 
 Old St Paul's - 
 
 3°° .. 
 
 In France, on the other hand, the transepts of some of the largest cathedrals 
 do not exceed 1 50 feet ; e.g. Paris and Reims. And even at Amiens, where 
 the transept is prolonged to nearly 200 feet, it projects but little beyond 
 nave and choir, owing to the fact that the choir has double aisles and a 
 ring of buttress-chapels, while the nave has a single aisle and buttress- 
 
 * Plan in Dehio, 42. t Builder, Nov. 9, 1895. 
 
 % Enlart's Manuel, 122. § Dehio, Plate 42. 
 
 || Scott's Lectures, ii. 33. 1 Enlart's Manuel, 123. 
 ** Beckett, 376. 
 
THE TRANSEPT. 197 
 
 chapels. It is indeed owing to the utilisation * of the buttresses to form the 
 side walls of chapels that the French Gothic cathedrals are able to dispense 
 with the elongation of the transept, thus gaining great compactness in plan. 
 These cathedrals are broader and far loftier than our own ; their vaults are 
 proportionally far heavier ; and the thrusts of the vaults, transmitted over the 
 aisle roofs by flying buttresses, have to be stopped by buttresses of excessive 
 projection. Given such long thin buttresses, it was a short step to roof over each 
 pair, providing thus an altared chapel. To such a scheme the comparatively 
 small English buttresses did not so readily lend themselves. But there were 
 as many saints to worship in England as in France, and as much need there- 
 fore of the multiplication of altared chapels. Not being able to utilise his 
 buttresses,"} - the Englishman turned to his transepts, already much elong- 
 ated, in order to find room for additional chapels. These he prolonged still 
 further to north and south. The eastern sides of the long English tran- 
 septs thus became fringed by a row of altared chapels. As late as St Mary 
 Redcliffe, Bristol, this was the ritualistic use of the transept ; there it was built 
 to hold four altars, two in each arm ; to St Catherine, St George, St Blaise, and 
 All Souls^ 
 
 For each one of the altars a sanctuary was needed ; the main body of the 
 transepts serving as a nave where the worshippers stood or knelt. At first the 
 sanctuary took the form of a little eastward-pointing apse. Usually there was 
 one apse in each arm of the transept ; as at Notre Dame du Pre, Le Mans ; and 
 Cerisy. This was the usual arrangement in England ; e.g. at Melbourne, 
 GLOUCESTER ( 135) ; ROMSEY (151.3) ; Chichester; Chester; NORWICH (148.4) ; 
 Lindisfarne ; Christ Church, Hants; Tewkesbury; Castle Acre; Evesham. At 
 Hereford the Norman south transept has a square sacristy instead of an apse. 
 If the transept was long, there might be two apses in each arm, as at CLUNY 
 (150.1); Canterbury, St Augustine's ; and ST ALBANS (98). But as early as 1079 
 WINCHESTER CATHEDRAL (154. i ) was set out with an eastern aisle instead of 
 eastern apses to its transepts; and in 1083 ELY (153.4); and in 1093 DURHAM 
 (149.1). This was a great improvement ; for in each of these three cathedrals 
 it was now possible to get three eastern chapels instead of one or two apses in 
 each arm of the transept. Peterborough, begun 11 17, copied the Durham plan. 
 This plan is characteristic also of the earlier Cistercian transepts here and 
 
 * In France the utilisation of the buttresses commenced with Notre Dame in 1290 : first, 
 with the building of Saints' Chapels between the buttresses of the choir. Then, the shops which 
 seem to have been allowed to be built and rented between the nave-buttresses were dislodged, 
 and in their place family or chantry chapels were built : the whole range of chapels being 
 finished to the west in 1320. The example of Notre Dame was followed in nearly all the northern 
 Gothic cathedrals, except where the nave walls were exceptionally thick, as at Chartres and 
 Reims ; or where the church was monastic, as St Ouen, Rouen. At Notre Dame, Paris ; 
 Troyes ; and Amiens ; the chapels were added continuously, and according to a fixed and 
 symmetrical plan. More often they were built intermittently, and different in area and design, 
 as at Bourges. 
 
 t They were utilised, by exception, at Chichester, producing an additional outer aisle on 
 either side of the nave. Cf. the chantries of Bishops Russell and Longland at Lincoln ; 
 Bishop Audley's at Hereford ; and the chapels between the buttresses of Eton and KING'S 
 COLLEGE chapels (199). 
 
 + Norris, 37. 
 
ip8 THE TRANSEPT. 
 
 abroad;* e.g. KIRKSTALL (152.4) and Roche; and of Ripon Minster (Secular 
 Canons). In Gothic it was reproduced at LINCOLN (151.1), Whitby, Lichfield, 
 Hereford (north transept), Selby, Howden. Such transepts, with eastern aisles 
 only, are rare in France ; nearly all the larger cathedrals have both eastern and 
 western aisles. Lisieux, however, has eastern aisles only. A further step is 
 seen at Winchester and Ely, where the Norman transepts have both eastern 
 and western aisles. It is remarkable to find such complication of plan so 
 early in Anglo-Norman work ; it shows how very rapidly the builders on 
 this side of the Channel left behind the precedents of the Normandy abbeys. 
 Even in Gothic days but few of our abbeys or minsters indulged in the luxury of 
 a double-aisled transept: Old St Paul's; the Cistercian Abbey of Byland, c. 1170; 
 BEVERLEY (152.1) and YORK, c. 1240 (153.1); WESTMINSTER, 1245 (north arm 
 only; 151. 2); Chester, c. 1330 (south arm only) are the chief. It is found also in 
 a few parochial or collegiate churches; e.g. Faversham; PATRINGTON (215.11); St 
 Mary Redcliffe, Bristol. BURY (150.3) had both an eastern aisle and an apse in 
 each transept arm. Glastonbury had two aisles,! both on the eastern side of the 
 transept. The Gothic Cathedral of Rouen, which is curiously Romanesque in 
 plan, has a double-aisled transept ; and from each eastern aisle projects an 
 elongated eastern apse. Where a western aisle was built to a transept, it 
 could not have been very serviceable for worship. If the altar were placed to 
 the west, there would be room for the congregation, but the altar would 
 point in the wrong direction. If, on the other hand, the altar were placed 
 under the arches between the western aisle and the transept, it would point 
 correctly to the east, but there would be very little space for the congregation. 
 In later clays this western aisle was utilised to provide sacristies or vestries ; 
 as it does to this day in the south transepts of Peterborough and Ely and 
 Winchester ; and in the north transept of Wells. 
 
 Of all our transepts, those of Ely and WINCHESTER (154.1), built by two 
 Norman brothers, were set out on the most magnificent scale. They not only 
 had eastern and western, but return or end aisles as well. These existed also 
 at cerisy (199); at St George's de Boscherville; at St Stephen and St Nicholas, 
 Caen ; and in Lanfranc's Canterbury. Gervase, the monk, lets us see the purpose 
 of these north and south aisles in the transepts of Lanfranc's Canterbury. For 
 he tells us that in the upper part of the eastern apse of the south transept 
 there was the altar of All Saints ; and in that of the north transept that of 
 St Blasius. Thus the return aisle provided a means of access to these altars. 
 Moreover, he says, the vault of the south aisle of the south transept carried an 
 organ.}: Another object may have been to provide continuous communication 
 between the triforiums of the nave and choir. § Winchester retains the original 
 aisles ; at ELY (506) they have been set back nearer to the end walls of the 
 
 * But in the early Cistercian churches, such as Kirkstall and Pontigny, there is no con- 
 tinuous aisle ; the eastern chapels being- completely separated by solid walls. 
 
 t The eastern of these was divided up, in Kirkstall fashion, into separate chapels. 
 
 I Willis' Canterbury, 39. 
 
 S M. Besnard is of opinion that at St George's de Boscherville these transeptal galleries 
 were employed for the exposition of relics, which were placed on a beam extending from the 
 east to the west wall of the transept above the gallery {Monographie de St George's de Boscher- 
 ville, 54). 
 
199 
 
 York Choir. 
 
 Lincoln South Transept. 
 
 King's College Chapel. 
 Cerisy South Transept. 
 
200 THE TRANSEPT. 
 
 transepts. It is curious that a return aisle occurs also in the north arm of 
 the eastern transept of Lincoln ; and perhaps formerly in the south arm also. 
 Even in Gothic days no further development of the transept obtained than 
 that which had been reached at Winchester in 1079. Indeed many of the 
 Gothic transepts were built on Norman foundations ; and, as at Canterbury, 
 remained aisleless, with merely the substitution of rectangular for apsidal 
 eastern chapels. 
 
 Between the chapels of the transeptal aisles there were originally solid 
 walls ; such as still remain in the north transept of Ely, in the Cistercian 
 transept of Kirkstall ; later on, only low walls were employed ; as at Abbey 
 Dore and Lincoln ; later still, they were usually separated merely by screens. 
 Screens also were interposed between each chapel and the transept. 
 
Chapter XII. 
 
 The Nave — Narthex — Galilee — Western Transept — Porch — Chantry Chapel. 
 
 As we have seen, the basilican prototype of the mediaeval minster usually had 
 aisles. Here and there, however, churches, even of great scale, were aisleless ; 
 e.g. the so-called basilica of Treves,* which still stands. Of aisleless naves there 
 were originally many in England ; especially in the churches of the Secular and 
 the Regular Canons. Of the former the nave built by Archbishop Roger at Ripon 
 (1154-1181) was one of the most remarkable ; it was 40 feet across ; f but aisles 
 were added in 1502. On the south or north side of the naves of the churches 
 of the Regular Canons, on the other hand, there was a cloister ; so that an aisle 
 was only added on the opposite side of the nave to the cloister ; e.g. at Brinkburn, 
 Dorchester, Easby, Hexham, Kirkham, Lanercost. Some naves, like those of 
 Bayham and Lilleshall, remained without aisles to the last. The great peculiarity 
 of the Anglo-Norman nave was the enormous length which it reached in such 
 examples as Winchester, Norwich, Ely, Peterborough, and St Albans. 
 
 In front of some of the largest of the early Christian churches, such 
 as OLD ST PETER'S (1 47) at Rome in the fourth century, and S. Sophia, 
 Constantinople, in the sixth century, was a quadrangle with colonnaded 
 walks on its four sides; what is called an atrium. It formed a grateful 
 transition from the noise and glare of the streets and the outer world to the 
 stillness and seclusion of the Christian church. In it were gathered together 
 reminiscences of more than one type of building familiar to the early Christians. 
 There was the great outer court of the Gentiles, which preceded Solomon's 
 temple at Jerusalem. At Rome too, in front of such great secular basilicas as 
 that of Trajan, there was often a square with covered colonnades all round. 
 And in the great mansions of the Roman aristocracy, where ten generations of 
 the Christians at Rome had worshipped, there was a threefold division ; first, 
 on entering, a court called the atrium; then, further in, another colonnaded 
 court called the peristyle ; and then the tablinum, where was probably placed 
 the altar at their services. 
 
 Such a fore-court, however, to a church required an area of land costly and 
 difficult to obtain in a great city. So the atrium survived but sporadically 
 in Eastern and Western Christendom ; e.g. at Novara and Parenzo (seventh 
 century; ; in Salerno, c. 1077 ; in S. Clemente, Rome ; S. Ambrogio, Milan ; and 
 the German monastery of Laach ; all probably twelfth century. In England we 
 have no example of it. 
 
 Where, however, it was impossible to find space for all the four 
 
 * It is not certain that this was originally a church, 
 t Illustrated in Archtrological Journal, 31. 
 
202 THE NARTHEX. 
 
 covered walks of an atrium, it was often possible to build the eastern walk 
 in front of the main entrance to the church. This eastern walk we may call 
 the narthex. The basilicas at Ravenna seem to have had usually a closed 
 narthex : while those at Rome * were open to the west. But a mosaic in S. 
 Apollinare Nuovo, Ravenna, shows an open narthex closed by curtains. 
 
 In the Burgundian churches, but not in England, the narthex was 
 developed to an enormous extent ; in fact it became a great ante-church. It 
 occurs in the archaic church of Gannat, and at the Benedictine abbey of 
 Tournus ; both apparently early in the eleventh century ; and in the Benedictine 
 abbey of Fleury or St Benoit-sur-Loire, commenced 1062. At Tournus and 
 Fleury it is two stories high. Of all the Orders the Cluniac seems most to have 
 favoured it; it occurs at CLUNY (150.1), begun 1089; Vezelay ; Souvigny ; 
 Faray-le-Monial : all Cluniac abbeys. There is an eleventh-century example 
 at Romainmotier ; it occurs also in the Cathedral of Autun ; c. 1150. But at 
 the great abbey-church of La Charite-sur-Loire, instead of an independent 
 fore-church, the number of bays of the nave is simply increased to ten plus 
 a western transept.! This Cluniac church affords a curious parallel to our 
 long Benedictine naves at St Albans, Bury, Ely, Peterborough, Winchester, 
 Norwich ; and it is possible that in them, as in La Charite, the excessive length 
 of the nave is due to the fact that it was intended that its western bays should 
 serve as a narthex. 
 
 Another survival of the narthex in England is to be found in the form 
 of western porches ; some of which from the earliest times have been called 
 Galilees. At DURHAM (149.1), Glastonbury, £ ELY (153.4), and Snettisham, 
 Norfolk, the galilee is to the west of the facade; at LINCOLN (151.1) it is west 
 of the south transept. St Woolos, Newport, has a fore-church of earlier date 
 than the church west of it. The western porch seems to be of the most vener- 
 able antiquity. It occurs at St Pancras, Canterbury ; South Elmham, Suffolk ; 
 Ythanchester, Essex ; Monkwearmouth, Durham : all probably of the seventh 
 century. § And as at Ythanchester and Monkwearmouth the porch was after- 
 wards raised to form a tower, it may be that in these western porches we have 
 at least one of the origins of the western tower which is so common in Anglo- 
 Saxon and English Gothic architecture. 
 
 With the Cistercians the fore-church was in much less favour than with the 
 Cluniacs ; probably because their churches were purposely planted in most 
 remote and sequestered sites, and there was little or no resort of pilgrims to 
 them. So the narthex dwindles down with the early Cistercians to a low 
 building west of the facade, covered by a lean-to roof. Remains of such a 
 narthex exist at Fountains, Rievaulx, and Byland. At Byland it went by the 
 
 * Cf. the narthex of St Nicholas, Caen ; Ruprich-Robert, Plate 8. 
 
 + Nowadays the pier-arcade, triforium, and clerestory of the western part of this nave 
 have been converted into house-fronts ; and the nave into a street. 
 
 X Willis {Glastonbury, 76, 78) quotes John Glaston as calling the east part of the Lady 
 Chapel (St Joseph's Chapel) at Glastonbury the galilee. Willis says that there is no docu- 
 mentary evidence for calling the south porch of Lincoln a galilee. At Durham the galilee was 
 a Lady Chapel. 
 
 § See Mr Peers in Archeeological Journal, December 1901. 
 
THE WESTERN TRANSEPT. 
 
 203 
 
 name of Galilee. Fine examples of the Cistercian narthex survive in France at 
 PONTIGNY (599) ; in Italy at Casamari and elsewhere.* The later Cistercian 
 churches, e.g. Tintern, seem to have dispensed with the narthex altogether. 
 
 There was a curious revival of the narthex or vestibule in Post-Reformation 
 days. In the Fen and Marshland churches of Norfolk and Lincolnshire much 
 woodwork seems to have been introduced in the time of James I. and 
 Charles I. YValpole St Mary, happily unrestored, retains a fine Jacobean screen 
 separating off the western bay of the nave ; as well as Jacobean font cover, open 
 seats, pulpit, and chandelier. Terrington St Clement's retains much of the 
 western screen. t It is noteworthy that Wren also in most of his London 
 churches separated off a vestibule by means of an oak screen. 
 
 Hereford, X. Porch. 
 
 Melbourne from East. 
 
 Western Transkit. 
 
 Another form assumed by the "narthex" was the western transept. It 
 occurs in the Romano-British basilica of Silchester ; and in the seventh century 
 at Brixworth. In our Romanesque and Gothic churches it may be seen in three 
 forms. In the fine Norman church of MELBOURNE the aisles end as usual in 
 a pair of western towers. These towers have a groined vault ; and the space 
 
 * Illustrated in Enlart's Gothic in Italy, 43. 
 
 t Tilney All Saints' retains a Jacobean choir screen ; that at Wig 
 has been brutally destroyed, like many others of this period, in 
 Hereford, retains much of its Jacobean fittings. The best examples of all are Crowcombe 
 Somerset, and St John's, Leeds. 
 
 nhall St Mary Magdalen 
 'restoration." Dore Abbey, 
 
204 
 
 THE WESTERN TRANSEPT. 
 
 between them also has a groined vault, the upper surface of which provides a 
 gallery. The arrangement is curiously like that given by Mabillon, "Dime turres 
 sint in ecclesiae fronte statuta, et subler ipsas atrium ubi laid stare debent, ut 
 non impediant processionem (monachorum)." '* A similar arrangement occurs 
 in the two Caen f abbeys. A second plan was to build the two western towers 
 not in a line with the aisles, but flanking them, and opening into them by lofty 
 arches. This was the transept plan at the Norman priory of St Botolph, 
 Colchester; at Ripon the towers originally flanked an aisleless nave ; and in the 
 Transitional transept of Peterborough behind the thirteenth-century facade. 
 It was planned at St Albans by John de Cella c. 1200, but never carried 
 out. It is the plan of Bishop Jocelyn's transept at WELLS, c. 1220 (154.3). At 
 
 PETERBOROUGH (112), perhaps in 
 rivalry with its neighbour Ely, three 
 western porches of colossal height were 
 built in Early Gothic in front of the 
 Transitional western transept. The 
 third plan is far more complex, and 
 produces an interior of the very highest 
 beauty. It survives in part at ELY 
 (153.4); an d in foundations at BURY 
 (150.3). % That at Ely has an immense 
 central arch to the west ; showing that 
 it was intended in the twelfth century 
 to be preceded by a great porch, per- 
 haps even loftier than the present 
 galilee. The magnificent vestibules of 
 Ely and Bury were reproduced in 
 Gothic at LINCOLN (151. 1); save with 
 the substitution of rectangular for apsi- 
 dal chapels ; and a chapelled western 
 transept is one of the noblest features 
 in the planning of Wren's St Paul's. 
 
 The only other Romanesque province 
 where the western transept is common 
 is Germany ; where indeed it abounds : the flanking octagonal towers of Bury 
 and Ely find analogues at Gernrode and Ouedlinburg § in the tenth century and 
 in many later examples. On the other hand, the simple Colchester plan with 
 its flanking towers and the complexity and great projection of the Bury and Ely 
 plans have little in common with the normal type of German western transept. 
 There seems no reason for doubting that with us, as with Germany, it was 
 an independent indigenous growth. In England the broad western transept, 
 whether of the Wells or the Bury type, was built no more after the middle of the 
 thirteenth century, with the exception of the seventeenth-century St Paul's. 
 After that, the old type of Normandy, with two axial western towers, prevailed ; 
 
 Llanl: 
 
 * Dehio, i. 390. 
 
 \ The western transept of Bury was 260 feet long. 
 
 t Ruprich -Robert, Plate 9. 
 § Plan 47, Dehio. 
 
THE PORCH. 205 
 
 as at Chichester, Lichfield, and York ; or else even the towers were omitted, 
 as at Exeter, Salisbury, Worcester, St David's ; giving a very unimpressive and 
 parochial aspect to the main facade of many of our most important churches. 
 
 Porches. 
 
 In France we may perhaps regard the great cavernous western porches 
 — three, as at Amiens and Reims ; five, as at Bourges and Bayeux — as lineal 
 descendants also of the ancient narthex. Setting aside the projecting galilees 
 of which we have spoken above (202), we have few analogues in England. 
 There is a western porch at Chichester; added c. 1250; there are the superbly 
 designed porches of John de Cella at St Albans, and the enormous triple 
 portal of Peterborough. As a rule we did not care to develop the western 
 doorways. The reason may be that our churches are all comparatively low ; to 
 give the west doorways, therefore, any considerable elevation would be at 
 the expense of the western windows. We needed western light badly in 
 our English naves, especially in the twelfth and thirteenth century, and pre- 
 ferred to develop the western window at the expense of the western doorway ; 
 reaching in the end such a facade as that of ST GEORGE'S, WINDSOR (492). In 
 France, on the other hand, the cathedral was frequently half as high again as 
 in England ; and however much the doorways were developed, there was 
 abundant space left for windows under the soaring vaults. 
 
 But there was ritualistic need of large porches in England as much as in 
 France. Not being able to introduce them into the western facade except in 
 the inartistic fashion of the Durham and Ely galilees, we built big porches in 
 large numbers on the sides of the nave. So that the single big side porch in the 
 end became as characteristic of the English church, whether cathedral, monastic, 
 collegiate, or parochial, as great western porches, three or more, came to be of 
 the French cathedral. 
 
 The porch was often two stories high, even in the twelfth century, as at 
 Southwell, SHERBORNE (576), and Kelso. Sometimes the upper room contains 
 a piscina, and therefore was used as a chapel ; sometimes, as at Southwell Minster, 
 there is a chimney, and it was used as a dwelling-place for the sacristan,* or 
 for a chantry-priest. After the Dissolution it was often used as a church library. 
 
 What was the precise use of these western adjuncts, whether narthex, 
 galilee, western transept, or porch, it is now difficult to determine. No doubt it 
 differed in different branches of the Catholic Church, and in different ages of the 
 Church. But there must always have been a large number of functions which 
 on the one side were of an ecclesiastical, on the other of a civil character ; e.g. 
 the payments of tithes and rents : and for such functions an adjunct, half within, 
 half without the sacred building, would have a natural suitability. Moreover, 
 especially in early Christian days, special arrangements were made for the 
 accommodation of the catechumens, who had as it were been but half initiated 
 in the Christian rites. In the Cluniac churches the great narthex may have 
 formed a waiting-room for pilgrims. Here, too, took place important ceremonies 
 
 * In 1294 Archbishop John Romanus of York ordered a sacrist to lie within the church of 
 Southwell, to ring the bells at the appointed hours. 
 
206 
 
 ®«Nfel mJM 
 
 
 Burford, South Side. 
 
MEDIAEVAL CHURCH PLANNING. 207 
 
 on Palm Sunday ; the Attollite portas of Psalm xxiv. In some cases the abbot 
 administered justice in his porch. Almsgiving took place here, and the rite of 
 exorcism ; and part of the ceremonies of baptism, matrimony, and the churching 
 of women. At Southwell the Pentecostals, or " Whitsun-farthings," were paid 
 in the north porch.* Here too was placed the holy water stoup (204). At 
 Dereham there are two external stoups in addition to one inside the porch. 
 
 Just as the "atrium" had dwindled down to the porch, so the great 
 fountain in the centre of the " atrium " (which still exists at S. Sophia, Con- 
 stantinople) dwindled down to the stoup. So also the early Christian practice 
 of washing hands and face as well in the fountain has survived in the 
 Mohammedan world ; but in Western Christendom dwindled to dipping the 
 tip of the finger in the water of the stoup; e.g. LLANBELIG (204). 
 
 Complexity of the Mediaeval Plan. 
 
 We are now able to see what a complex organism was the mediaeval 
 cathedral ; how manifold and diverse were the requirements prescribed to 
 the architect by the ecclesiastical authorities ; and how able was the planning 
 by which those requirements were satisfied. It was no longer the simple 
 basilica of early Christian days, with its single altar to the one God. It 
 was as if there had to be designed a temple for some great Polytheistic 
 religion, with a crowd of deities all to be housed within one common roof. It 
 was no longer for the Supreme Deity only the churches built ; but for an 
 ever-growing multitude of Saints and Doctors of the Church. In earlier days 
 it was thought enough sometimes to group several separate churches within 
 one precinct ; such as the seven churches of Cashel, and those ten churches in 
 the graveyard of Twyneham t which Bishop Flambard pulled down c. 1099 to 
 make room for his Norman Christ Church. The next stage is seen in the ninth 
 century at ST GALL (194); where all the churches are assembled under one 
 roof, but in such a manner that all the aisles and nearly all the nave are blocked 
 up with chapels ; congregational worship is impossible ; and the main body of the 
 church is impervious to processions. % The next step, perhaps, is illustrated at 
 ST ALBANS (14), where, as is shown by the paintings on the piers of the nave, 
 altars were placed against the western side of each pier instead of in the centre of 
 nave and aisles, as at St Gall. And this may be one reason for the enormous 
 length of the naves of St Albans, Ely, Norwich, Peterborough, Bury ; that a 
 
 * Livett's Southwell, 11. On the whole subject see Lenoir's Architecture Monastiquc, ii. 
 73 ; and Thiers' Sur les Porches. 
 
 + Cart, Twyneham, Tiberius, D. VI. 
 
 \ In the south aisle of St Gall, counting from the west, were the altars of St Agnes, St 
 Sebastian, St Mauritius, and St Lawrence ; and in the north aisle those of St Cecilia, Holy 
 Innocents, St Martin, and St Stephen. In the western apse of the nave was the altar of St 
 Peter ; then the font ; then the altar of St John the Evangelist and St John the Baptist ; then 
 that of the Crucified Saviour ; then in the western part of the eastern limb the high altar ; and 
 in its eastern apse that of St Paul. On the east wall of the north transept was that of St Philip 
 and St James ; and in the south transept that of St Andrew. Thus there were altogether 
 fifteen altars, not including those in the crypt. At Chartres there were in the fourteenth 
 century thirty-nine altars. 
 
208 
 
 MEDI/EVAL CHURCH PLANNING. 
 
 nave so elongated provided accommodation for an exceptionally large number of 
 altared chapels. The next step is to remove these altars from the nave and its 
 aisles to a chevet of chapels round the apsidal choir, as at St Martin de Tours 
 and Norwich ; to a central transept, as at Romsey and St Albans ; to an eastern 
 transept, as at Cluny and Canterbury ; to a western transept, as at Bury and 
 Ely ; to an upper aisle, as in the five upper chapels of Gloucester choir and 
 transept ; leaving the nave free for the assembled laity, and the aisles un- 
 encumbered for processions. Eastward of the nave, as we have seen, it became 
 usual to provide an enclosed choir for the clergy and singers ; east of the choir, 
 a raised* sanctuary for the high altar. And when there was a famous local saint, 
 
 efficacious in answer to 
 , J prayer, another great eastern 
 
 annexe was added ; the great 
 eastern presbyteries of Ely 
 and Lincoln were but the 
 glorified shrines, not in gold 
 and silver, but in stone, of 
 St Etheldreda or St Hugh. 
 Then there was provision 
 to be made more and more 
 for a multitude of chantry 
 chapels. One of the earliest 
 was that at Lincoln Minster 
 of Bishop Hugh of Wells, 
 who died in 1235. In the 
 thirteenth century they were 
 few ; f the greater number 
 were founded in the fifteenth 
 century. They were founded 
 by kings and princes ; e.g. 
 Henry the Seventh's chantry 
 chapel at Westminster ; and 
 that of his eldest son, Prince 
 Arthur, in Worcester choir ; 
 by bishops, abbots, barons, 
 commoners, and guilds ; by 
 any one who had the money to provide for the saying of masses for the 
 repose of the soul of " himself and his parents and his relatives and his 
 benefactors and of all the faithful dead."j Some two thousand chantries had 
 been founded up to the time of Henry VIII. ; when they were suppressed their 
 income was found to be under £10,000 a year, which we may perhaps multiply 
 by twelve to bring it to the present value of money. In a great church it was 
 
 * In England the sanctuary as a rule was but little raised, except of course where there 
 was a crypt. 
 
 t The Taxatio records only two. In Yorkshire less than a dozen are recorded before the 
 fourteenth century. Cutts' Parish Priests, 438. 
 
 % Louth Records, 163 ; " Chantry of Canon Thomas de Luda." 
 
 St Albans, Ramryge Chantry Chapel. 
 
209 
 
 Wells, Bubwith Chantry Chapel in N. Aisle of Nave. 
 
MEDI/EVAL CHURCH PLANNING. 21 1 
 
 not difficult to find room for the combined tomb and chapel, what is often called 
 the " chantry " ;* which, in larger examples, consisted of a screened enclosure, 
 containing the tomb, the slab of which served as altar ; in late examples, as in 
 Fox's chantry at Winchester, there might be also a vestry at the east end. 
 These were placed in great numbers under the arches separating the aisles from 
 the choir or nave ; e.g. at WELLS (209 J and ST ALBANS (208). Owing to the 
 protection given by Fairfax, an old Winchester boy, the Cathedral of Winchester 
 retains an unusually fine set of chantries of its bishops ; those of Edington and 
 Wykeham between the piers of the nave; those of Beaufort, Fox, and Gardiner 
 between the piers of the retrochoir and choir. Frequently the eastern bay of 
 each choir aisle was turned into a chantry chapel ; as by Bishops Alcock and 
 West at ELY (14.3). At Durham the great family of the Nevilles was allowed 
 to screen off two whole bays of the south aisle of the nave as their chantry 
 chapel. Sometimes, as at LIN'COLN (112) and Eton and Windsor, chantries 
 were inserted between the buttresses. Sometimes there was such an agglo- 
 meration of chantries that an additional aisle, or aisles, was produced, as at 
 Chichester Cathedral, Dorchester Priory, Manchester Collegiate Church. All 
 this added greatly to the difficult)' of planning the great mediaeval churches. 
 But the problems and difficulties were triumphantly surmounted. The mediaeval 
 cathedral is a masterpiece of ingenious and scientific planning. Complicated as 
 it is, there is no part but was necessary ; and each part is situated where most 
 convenient for the requirements of ritual : there is no waste, no awkwardness of 
 spacing. Ever)- part had its use ; and was in daily use. The plan was as 
 utilitarian as it was magnificent. 
 
 * Strictly the term applies to the ecclesiastical foundation, not to the building. 
 
Chapter XIII. 
 
 THE PLANNING OF THE PARISH CHURCH. 
 
 So far we have spoken of the planning of the larger churches ; cathedral, 
 monastic, or collegiate. They were originally the offspring of the early Christian 
 basilicas. It has often been assumed, therefore, that the humble village church 
 also is again but the Italian basilica, cut down, curtailed, simplified in plan. 
 The truth is the other way. The parochial church is really not so much a re- 
 duction or simplification of a higher organism as an amplification of a very simple 
 and lowly form of ecclesiastical architecture, gradually throwing out extensions, 
 now here, now there, till in the end there is sometimes but little difference in plan 
 between the parish church and the cathedral ; as may be seen on comparing 
 Patrington Church with York Minster ; St Mary Redcliffe, Bristol, with Salisbury 
 Cathedral ; or Louth with Llandaff. 
 
 Apsidal v. Rectangular Chancel. 
 
 Of the Norman parish churches many were apsidal. The following retain 
 apses at the end of the chancel : * — Berkshire, Padworth, Finchamstead, Remen- 
 ham ; Cambridge, Iselham ; Cumberland, Warthwick ; Derbysliire, Steetley ; 
 Essex, Great Maplestead, Little Maplestead, East Ham, Haversfield, Hadleigh, 
 Little Braxted, Langford, Copford ; Hampshire, Nateley, Easton ; Hereford, 
 Aston, Kilpeck, Moccas, Pencombe, Peterchurch ; Hertfordshire, Bengeo ; Kent, 
 Sutton, near Dover ; Norfolk, Heckingham, South Runcton, Gillingham, Tritton, 
 Hales ; Oxfordshire, Checkendon, Woodcote, Swyncombe ; Suffolk, Fritton, 
 Dunwich ; Sussex, Newhaven, Upper Waltham ; Warwickshire, Bilston ; Wilt- 
 shire, Manningford Bruce ; Yorkshire, Feliskirk, Birkin, Lastingham ; besides 
 others in remote parts of England. 
 
 But, as in the smaller Anglo-Saxon, so in the smaller Norman churches of 
 England, rectangular chancels are more common than apses ; e.g. Avington, 
 Berks; Adel, York; Barfreston, Darent, and PATRIXBOURNE (218), Kent.f The 
 origin of this rectangular chancel is somewhat doubtful. We know that the 
 greater part of England was not converted to Christianity by Augustine and the 
 monks sent from Rome, but by Irish and Scotch missioners whose head- 
 quarters were Lindisfarne, Iona, and Ireland. Now in Ireland the churches 
 were of stone, and rectangular ; none were apsidal. It has therefore been 
 assumed that these missioners brought with them the lithic architecture of 
 Celtic Christianity. But this is to ignore the fact that the Irish stone churches 
 * Parker's Glossary, 29. + See list in Bloxam, 84. 
 
Leicester St Nicholas from SAY. 
 Melbourne S. Choir. 
 
 St Lawrence from S.W. 
 Louth from S.W. 
 
214 
 
 Lynn St Nicholas from S.W. 
 Wrexham from S.W. 
 
 Gresford from N.E. 
 Long- Melford from S.W. 
 
siS 
 
 B 
 
 ® 
 
 © 
 
 ID- 
 
 © 
 
 Qa 
 
 fr-ri rri 
 
 SCALE OF FEET 
 
 1. Silchester. 
 
 2. Temple Balsal. 
 
 3. Tangmere. 
 
 4. Hadleigh. 
 
 5. Cassington. 
 
 6. Bucknell. 
 
 7. Achurch. 
 
 S. St Cross, Winchester. 
 
 9. Howell. 
 10. North Minims. 
 1 r. Patrington. 
 
2l6 
 
 SCALE OF FEET 
 
 1. Westham, Sussex. 
 
 2. Toot Baldon. 
 
 3. Baginton. 
 
 4. Boston. 
 
 5. St Mary, Eastbourne. 
 
 6. Kidlinoton. 
 
217 
 
 lo Ze }o *o 5© 6o 7o 
 SCALE OF FEET 
 
 Burford. 
 
 St Nicholas, Lynn. 
 
218 
 
 THE RECTANGULAR CHANCEL. 
 
 were of the rudest possible character in the seventh century and long after ; little 
 oblong sheds without aisles or chancel ; without an arch ; probably built without 
 mortar. How could a race, architecturally so backward, impose its building 
 methods on a country which, even in the seventh century, as we read in the 
 description of Hexham, by Wilfrid's own chaplain, Eddius, was able to build 
 aisled churches, three stories high, with towers, crypts, galleries, and porticus ? 
 A race, which architecturally is on a higher plane, imposes its building methods 
 on a race which is on a lower plane. This was the case with the Normans when 
 they came into contact with Anglo-Saxon architecture. But when a race meets 
 another in a more advanced stage of architectural development, its own archi- 
 tecture succumbs or makes but little headway. This was the case with the 
 Normans themselves when they came in contact with the advanced art, Byzantine 
 or Saracenic, of Sicily and Southern Italy. It is impossible then to believe 
 
 that the rude lithic architecture of Ireland * 
 can have had any considerable influence on 
 the far superior work of seventh-century 
 England. 
 
 Simple constructional reasons are suffi- 
 cient to explain the preference for the rect- 
 angular over the semicircular chancel. In 
 the first place, though the earliest Celtic 
 churches in Ireland were of stone (perhaps 
 because the Irish builders had not the tools 
 for working wood), yet we read in Bede that 
 when the first Celtic missioners landed in 
 Lindisfarne, their first church was not of 
 stone, but of wood. Saxon records, from 
 Bede downwards, contain abundant references 
 to timber-built churches. And the cases are 
 very numerous in which an earlier structure 
 of wood was rebuilt in stone, e.g. at York, 
 Tynemouth, Chester-le-Street, Wilton, Last- 
 ingham.t It is reasonable to suppose that in 
 such instances the earlier plan might survive in the later structure. Bede, too, 
 speaks of building " more Scottorum non de lapide" which can only mean that 
 the characteristic church, as planned by the first Celtic missioners in England, 
 was of osiers or wood. Again, he says, % that " Naitanus, rex Pictorum, archi- 
 tectos sibi mitti petiit, qui juxta morem Romanorum ecclesiam de lapide ingenti 
 facerent,"^ which shows that stone-construction was exceptional on the north 
 as well as on the south of the English Border ; where, he says, it was a " mos 
 Britonibus insolitus." The first churches of the Celtic missioners in England, no 
 
 Patrixbourne, East End. 
 
 * It is not intended to deny that the Irish missioners did here and there build in stone. 
 Bede describes the erection of a stone church in the south of Scotland, at Whitherne or Stone- 
 house, in Galloway. Such churches would be of the archaic Irish type. 
 
 + Baldwin Brown in Builder, Dec. 21, 1895. \ Ecrfes. Hist., v., c. 22. 
 
 § " Naitanus, king of the Picts, asked for architects to be sent to him (from Gaul), to 
 b uild him a church of large stones after the manner of the Romans." 
 
THE RECTANGULAR CHANCEL. 
 
 219 
 
 doubt, were constructed of osiers* plastered over with mud : a little later, these 
 temporary oratories would be rebuilt in timber: ultimately, in stone. In these 
 osier or timber churches, then, of the first Celtic missioners the origin of the 
 rectangular chancel might be looked for. 
 
 But it is unnecessary to assume a Celtic origin. To build an apse would 
 have been as unnatural to their converts as it was to the Celtic missioners. 
 For the converts in England were Angles, Saxons, Jutes ; and later on, Danes 
 and Norwegians ; and all these came from forest lands ; and were shipwrights and 
 carpenters by trade, not stone-masons; and well able to build substantial log-huts, 
 such as those of the Norske bonder to this day. Of churches of this latter or 
 timber type we have an example surviving in the Pre-Conquest church of Green- 
 stead, Essex. Now it 
 would be difficult to 
 build a log-chancel in 
 the form of a semicircle ; 
 and just as easy to build 
 it rectangular. Secondly, 
 if a log-chancel were 
 actually built apsidal, 
 it would be difficult to 
 cover it with a timber 
 roof; though it is done 
 at Worth and Brixworth. 
 We may conclude that 
 the churches of timber 
 would invariably have 
 rectangular chancels ; 
 and that this form, when 
 hallowed, as in time it 
 would be, by the associa- 
 tions of centuries, would 
 become the natural and 
 characteristic form for 
 the chancel of the 
 smaller, if not of the 
 greater churches also, 
 even when they were constructed in masonry. Thus we may explain the pre- 
 ference of the Anglo-Saxons for the rectangular form of chancel in spite of 
 Continental tradition. 
 
 St Columba, Kel 
 
 Planning of Parish Churches. 
 
 The simplest form in which the church occurs is seen in the early churches 
 of IRELAND. There it is just an oblong shed ; no chancel ; no chancel-arch. 
 In England this form constitutes a chapel ; not a church. It occurs, for instance, 
 in St Cuthbert's Chapel at Lindisfarne ; and in the Chapel of St Mary Magdalen 
 
 * Such a church long survived at Glastonbury ; see IJritton's Arch. Ant., v. 96. 
 
220 
 
 PLANNING OF PARISH CHURCHES. 
 
 at RIpon and at Skirlaugh * It is but rarely seen in the smaller churches ; e.g. 
 at TEMPLE BALSAL, Warwick (215.2); which is rather a chapel than a church; 
 built by the Knights Templars, and therefore anterior to the suppression of the 
 Order in 13 1 2. 
 
 Plan I. : Unaisled Nave and Chancel. 
 
 The normal church, however, in Normandy t and in England has a distinct 
 chancel, at any rate in early days. The chancel may be square, as at TANGMERE 
 (215.3); ^ may be polygonal, as at Brixworth and Wing; semicircular, as at 
 HADLEIGH (2 1 5.4) ; but it is always there in our early churches. The presence 
 of some form of chancel and chancel-arch in England, long before it appears in 
 
 W!ulM~* 
 
 
 
 'V' 
 
 1I1 1'- Ji 
 
 Adel from N.W. 
 
 Ireland, makes a Celtic origin for the English parish church impossible. Of 
 this first type of English church, without aisles, without clerestories, with or 
 without towers, many examples are left us ; such as the Pre-Conquest church 
 of CorhamptonjJ Norman churches such as ADEL and Barfreston, which are 
 square-ended, and Hadleigh, which is apsidal ; § Gothic churches, such as Elsfield 
 and Tangmere (thirteenth century). It should be noticed that though architec- 
 turally a church of this type divides into two parts, nave and chancel, nevertheless, 
 ritualistically there is a triple division into nave, choir, and sanctuary ; for the 
 
 * Plans and illustrations in Churches of Yorkshire. 
 
 + R. Robert, 57. + Illustrated in Rickman, fifth edition, xxxvii. 
 
 § Buckler's Churches of Essex, 171, 4. 
 
PLANNING OF PARISH CHURCHES. 
 
 22 I 
 
 sanctuary is usually differentiated from the choir by having its floor at a 
 hisrhcr level.* 
 
 Second Plan : Unaisled Nave ; Choir ; Presbytery. 
 
 A second type, however, is common, at any rate in earl)' churches. In 
 this a choir is interposed between the nave and the eastern limb ; so that the 
 division is no longer into nave and sanctuary, but into nave, choir, and sanctuary. 
 It occurs in the Anglo-Saxon church of Brixworth, but with the addition of 
 aisles to the nave; and in the unaisled Norman church of HADLEIGH, Essex 
 (215.4). Frequently the central compartment, the choir proper, is covered with a 
 central tower ; e.g. in the Anglo-Saxon churches of Dunham Magna, Norfolk ; 
 St Mary's, Guildford ; and Barton-on-Humber, 
 as recent excavations have shown. t It is com- 
 mon in Norman churches, e.g. Stewkley ; J St 
 Andrew's ; Cassington ; Newhaven ; in the last 
 the sanctuary is apsidal. At BUCKNELL, Ox- 
 fordshire, the Norman towered choir remains ; 
 nave and sanctuary were remodelled later. 
 At Iffley all is Norman, except that the 
 sanctuary was lengthened in the thirteenth 
 century. Leckhampton§ has a towered choir 
 and sanctuary, both vaulted ; built in the 
 fourteenth century, yet Norman in plan. So 
 also Fairford, Kempsford, Tong. 
 
 It is noteworthy that in many of the Nor- 
 man churches with triple division the sanc- 
 tuary is vaulted, whether it be apsidal or 
 rectangular, as at CASSINGTON, Oxfordshire 
 (215.5); Darent, Kent; Tickencote, Rutland; 
 Compton Martin, Somerset ; Stewkley, Buck- 
 ingham. In the latter case there is a space 
 
 above the sanctuary vault and the timber roof ; and this seems to have been 
 often utilised as a priest's room : as may be seen by the windows inserted in 
 the gable to give light to it ; e.g. St Peter in the East, Oxford ; Stewkley ; 
 Darent. 
 
 * It is curious how closely the triple disposition of the Christian church reproduces that of 
 the Jewish temple : (1) The Holy of Holies, entered only by the High Priest. There was the 
 Ark, the resting-place of the Shekinah ; where God was in visible presence. It was screened 
 by a veil, cf. the Lenten Veil, the Iconostasis of the Greek Church, and the veiled Baldachinos 
 above the High Altar in the Sanctuary of the ancient basilicas. (2) The Holy Place ; entered 
 only by the Priests ; this corresponds with the choir of the Christian Church. (3) Next 
 came the Court of the Israelites ; accessible to Jews only, not to Gentiles ; this corresponds 
 to the Christian nave. (4) The outer court, or Court of the Gentiles ; corresponding to the 
 atrium ; e.g. St Clemente, Rome. Moreover, the Jewish temple, like the earliest of the Christian 
 basilicas, was orientated not to the east but to the west. (See Exodus xxvi. 27 ; and Scott's 
 Essay, 64.) 
 
 t Illustrated in Baldwin Brown's Arts in Early England, ii. 210. 
 
 I Britton, ii. 2. § Illustrated in Brandon's Parish Churches, 65. 
 
 Buckne 
 
222 PLANNING OF PARISH CHURCHES. 
 
 Third Type : Cruciform. 
 
 As time went on, both the above types of church proved inadequate to 
 accommodate an increased population, and to provide an adequate number of 
 chapels. It was necessary to enlarge the church. To get this enlargement, 
 two methods suggested themselves. One was to add aisles ; this involved 
 difficulties of construction, and at first was avoided. The other was to add 
 transepts; this method had the additional recommendation that much more 
 light could be obtained from the end walls of transepts than from the 
 side walls of aisles. If the church consisted of nave, towered choir, and 
 presbytery, as at Dunham Magna and Guildford, the obvious course would 
 be to add transepts. The tower over the choir would be all the safer for 
 having abutment to north and south in the form of transepts, as it already had 
 to the east and west in the form of presbytery and nave. And for the cruciform 
 shape thus given to the church there was abundant precedent. In the little 
 church in the Romano-British town of SILCHESTER (21 5.1) there is a rudimen- 
 tary transept of early Christian type. So also foundations of transeptal chapels 
 may be seen in the very ancient church of St Pancras, Canterbury. Other 
 Pre-Conquest examples exist or formerly existed at Deerhurst ; Stanton Lacy, 
 Salop ; Repton ; Stow ; Norton, Durham ; Worth, Sussex ; St Mary in the 
 Castle, Dover. Bradford-on-Avon has a north transept ; and foundations of 
 a southern one. And it is plain from numerous descriptions that transepts and 
 a central tower were common enough in the greater Anglo-Saxon churches, now 
 all gone. It is true that these cross-arms were, in the Anglo-Saxon churches, 
 more of the character of transeptal chapels than of transepts ; being usually 
 lower and narrower than the nave ; nevertheless they gave the church the 
 cruciform outline. In Normandy and in England all the greater Romanesque 
 churches had central tower and transepts. It was therefore but natural that in 
 many of our earlier churches enlargement should take the form of transepts. 
 And when churches were built de novo, the cruciform type, almost universal in 
 the larger churches, and already introduced into some of the smaller ones, would 
 here and there be adopted at any rate in churches of all but the humblest dimen- 
 sions. Hence the plan of Barton Seagrave, North Hants, which has a south 
 transept only, built as a chantry chapel ; and of such cruciform churches as 
 North Newbald and achurch (215.7). 
 
 A different process of extension would lead to the same result, when 
 applied to a church without a central tower. Here it would be easy to enlarge 
 the church eastward by pulling down the sanctuary ; building on its site a central 
 tower ; and projecting from the central tower transepts and a new sanctuary. 
 This is the process through which ST NICHOLAS, LEICESTER (213), and Bakewell 
 seem to have passed. By a similar change the aisled church of Sedgefield ob- 
 tained transepts and a new chancel, c. 1 290, some forty years after its rebuilding.* 
 
 Or thirdly, the easternmost bay of each aisle of the nave might be prolonged 
 to north and south so as to form a transept ; as at Weston. t 
 
 * If we accept the view of C. C. Hodges in Archceologia Aeliana, xvi. 389. 
 + Plan in Fen and Marshland Churches, vol. iii. Or it might be raised and perhaps 
 partially prolonged also, so as to form the quasi-transept, characteristic of Herefordshire. 
 
if ^' 
 
 •-. I ■ ---. • , 
 
 ,- jctSSS 
 
 BOSTON CIICKCH FROM S.E. 
 
PLANNING OF PARISH CHURCHES. 223 
 
 Such transepts, in whatever way they originated, were exceedingly con- 
 venient. They provided room for one or more chapels. And where a church 
 was built de novo by a single benefactor, it seems to have been not unusual 
 or him to reserve one transept as his chantry chapel ; e.g. the fourteenth-century 
 church of Shottesbrooke was built by Sir William Tressell, who with his lady, 
 Maud, is buried in the north transept.* 
 
 This fine church is without aisles. But transepts were added to, or built with, 
 aisled naves as well. At I'ATRINGTON (215.1 1), built, like Shottesbrooke, de novo, 
 c. 1330, the nave has aisles, and the transept has both eastern and western aisles ; 
 so that, but for the absence of aisles to the chancel and of western towers, it is 
 identical in plan with York Minster. A still grander and later example of the 
 parochial use of the transeptal plan is to be seen at St Man- Redcliffe, Bristol. 
 
 The transeptal plan without aisles occurs at all epochs ; and not sporadically, 
 but in quite large groups of churches here and there ; e.g. in North Wilts and the 
 adjoining parts of Oxfordshire, Berks, and Bucks ; in South Somerset and West 
 Dorset ; in Cornwall ; in West Sussex ; in Mid-Essex.f A fine example of the 
 thirteenth century is the large church of Llan-badarn-fawr, just outside of 
 Aberystwith. In later examples, of course, it occurs more often in conjunction 
 with an aisled nave. 
 
 Fourth Type : With Aisled Nave. 
 
 Common, however, as the transeptal plan became, especially in the Norman 
 churches of the twelfth century, it was in the end superseded in the vast majority 
 of churches by a simpler type. This was a church with aisled nave, elongated 
 chancel, and western tower. It may be regarded as simply a development of 
 the second type, that of Anglo-Saxon Corhampton or Norman Adel. This is 
 the type par excellence of the English parish church. There are thousands of 
 churches, both great and small, so planned. Its most magnificent outcome is 
 BOSTON CHURCH (216.4) ; which has an aisled nave 93 feet long. 
 
 In the early churches the aisles were often obtained in the simplest possible 
 manner. Aisle walls were built ; windows were inserted in them ; a lean-to 
 roof was put on ; then the walls of the old nave were pierced with arches, leaving 
 big blocks of masonry between them to serve as piers ; as at ST NICHOLAS, 
 LEICESTER (213) ; St Michael's, St Albans; Ickham and Stum', Kent ; Wing, 
 Bucks. 
 
 For the aisled nave there was abundant precedent. It was present in 
 all the larger early Christian basilicas and in the Romano-British church at 
 Silchester, and in the larger Anglo-Saxon and Norman churches. It existed 
 even in many of the smaller Anglo-Saxon churches ; e.g. Reculvers, Wing, and 
 Lydd ; and in some Norman parish churches ; e.g. TOWVN (458) ; Steyning ; 
 Northampton St Peter ; St Margaret at Cliffe. And so, just as many a Guildford 
 had enriched itself by the accretion of a Romanesque transept ; so many an Adel 
 expanded by the addition of a basilican aisle. And, as we have said, many 
 churches adopted both aisle and transept. 
 
 Sometimes, at first, it was but a single aisle that was added ; and here and 
 
 * Butterfield's Shottesbrooke. + Prior, Gothic Architecture, 54. 
 
22 4 
 
 PLANNING OF PARISH CHURCHES. 
 
 there a church remains still with no more than a single aisle ; e.g. EASTWOOD, 
 Stanwick, and North Creake. More often there was built afterwards, or simul- 
 taneously, another aisle on the opposite side of the nave. The early aisles, e.g. 
 at TOOT EALDON, Oxon. (216.2), those built in the eleventh, twelfth, and thirteenth 
 centuries, were usually narrow and low. They were therefore very frequently 
 pulled clown — sometimes more than once, as at Wakefield — and rebuilt broad 
 and high. Or, beyond an original aisle was added another aisle, as at BAGINTON, 
 Warwick (216.3), or more than one ; giving the church, three, four, five, or even 
 six aisles, as at Wisbech and Abingdon. Frequently, however, the newer outer 
 aisles were really chantry chapels, belonging to private persons or families, and 
 shut off from public use by screens ; e.g. the outer south aisle at Collumpton is 
 
 the chantry chapel of John Lane, 
 Merchant ; as appears from the in- 
 scription on his tomb therein (1510- 
 1528). At Ludlow the whole of the 
 south aisle and of the western part of 
 the nave appears to be an agglomera- 
 tion of various chantry chapels.* At 
 Baginton, the outer aisle seems to be 
 a chantry chapel founded in honour 
 of St Thomas the Martyr, c. 1250, by 
 Richard de Hertbull.f 
 
 In the great majority of the 
 churches the aisles were roofed with 
 lean-to roofs. Sometimes, however, 
 instead of building an aisle, the nave 
 was enlarged laterally by building 
 another nave parallel to it ; or two 
 parallel naves were built contem- 
 poraneously. This plan occurs in 
 early churches, and even so late as 
 St John's, Leeds (1634). Sometimes, 
 as at ST LAWRENCE, Thanet (213), 
 and at Rye, in the last half of the 
 twelfth century, three parallel naves 
 were erected. In this way a very broad church could be got. It is a plan 
 especially common in Kent and Sussex. But whether there were two or three 
 parallel naves, each with its saddle roof, the plan was open to the objection 
 that the gutter between any pair of saddle roofs, in the autumn became 
 choked with leaves, in the winter with snow ; and that at all times it was an 
 unsatisfactory and risky method of carrying off the rainfall. So it remained 
 exceptional. 
 
 Another method was to construct one saddle roof covering nave and aisles 
 alike ; as at Seaford, Sussex ; and TOOT BALDON, near Oxford (225). This 
 method of roofing was common in mediaeval aisled barns ; e.g. that of the Abbaye 
 
 * So also the south aisle of Dorchester Priory Church. 
 + Brandon's Parish Churches, 114. 
 
 Eastwood, Essex. 
 
PLANNING OF PARISH CHURCHES. 
 
 2 *> c 
 
 d'Ardenne, near Caen, and of Bradford-on-Avon. But the aisle walls, to be 
 thus roofed, have to be kept very low ; little room is left for windows ; and 
 the church is dark ; consequently this method had but little vogue. 
 
 Not only the nave received aisles, but the chancel was elongated. We 
 have seen that nearly every cathedral had its Norman eastern limb re- 
 built and lengthened, 
 as at Lichfield, Lincoln, 
 Wells, Exeter, York, 
 Canterbury. The same 
 process of elongation 
 was applied to vast 
 numbers of parish 
 churches ; in Kent usu- 
 ally in the thirteenth 
 century; elsewhere 
 more often later. And 
 so room was obtained 
 in the chancel eastward 
 for a spacious elevated 
 presbytery, westward 
 for the stalls of the 
 clergy and choir. Nor- 
 man chancels so en- 
 larged are IfHey and MELBOURNE (213 
 truncated apse may still be seen outside. 
 
 Toot Baldon. 
 In the latter, the curve of the 
 
 Fifth Type: Aisled Nave with Rectangular Chancel 
 
 partly aisled. 
 
 In one way or other, nearly all churches "f any consequence were either 
 built with aisled naves, or aisles were added to the nave subsequently. The 
 next step is seen when the aisles begin to creep into the chancel. 
 
 At first the chancel aisle appears only in a rudimentary and imperfect form 
 in the parish churches. One side of the chancel is aisled, the other is not. Or 
 part of one side or of both sides have aisles. These additions of bits of aisle 
 supplied chapels and sacristies ; both containing altars. The chancel of Hemel 
 Hempstead consists of two vaulted bays ; to the north of the western of these is 
 a vaulted chapel or sacrist)-; all this work, including the vaults, appears to belong 
 to the third quarter of the twelfth century. At Higham Ferrers the chancel has 
 a north aisle partly separated from it by a solid wall ; it is a Lady Chapel with a 
 sacristy east of it. So also at North Mimms, I lerts, the whole space north of the 
 chancel is occupied by a chantry chapel and a sacristy; at Howell, Lincoln, by a 
 chantry. At Worstead and Hinghain, Norfolk, there is a sacristy north of the 
 presbytery. At Barn wood, Gloucester, the whole space north of the chancel is 
 occupied by a chapel of Norman date. At Marston, Beds, is a chantry south, 
 and a sacristy north of the chancel. At Merstham, Surrey, the Early English 
 choir is flanked by a Decorated chapel on one side, and a Perpendicular 
 
 P 
 
226 PLANNING OF PARISH CHURCHES. 
 
 chapel on the other; leaving a small projecting presbytery: cf. Outwell, Norfolk. 
 At Solihull, Warwick, north of the presbytery is a vaulted sacristy, with a 
 chantry above it ; c. 1 200.* 
 
 But, here and there, much earlier than is generally supposed, symmetrical 
 aisles were built along the western bay or bays of the chancel. Examples occur 
 in the first half of the twelfth century at Northampton St Peter ; in the last half 
 of the century at Walsoken in the Norfolk Marshland ; at St Nicholas at Wade, 
 and originally at St Lawrence, both in the Isle of Thanet. So again in the first 
 half of the thirteenth century at West Walton, near Wisbech ; and, later, at 
 Heme, Kent, and kidlington, Oxon. (216.6). In the great churches of the 
 later fourteenth and fifteenth centuries in East Anglia, a very favourite plan was 
 to provide both nave and choir with aisles, but to build the presbytery of a 
 single unaisled bay. Thus the presbytery could be lighted by a side window on 
 either side ; e.g. at Loddon, Blythburgh, Lowestoft, Southwold. At Covehithe 
 these side windows rose to a vast height, and must have lighted the altar 
 brilliantly. 
 
 Sixth Type : With Nave and Chancel fully aisled. 
 
 It was but a short step to continue these abbreviated aisles eastward to the 
 full length of the chancel ; and the plan of LOUTH (139) is reached. The full- 
 length choir aisles which had been a special possession of the later great 
 monastic churches, such as Rievaulx and Whitby, and of cathedrals such as Ely, 
 Lincoln, York, were appropriated by the parochial builders ; and their church 
 became yet more dignified. This plan also seems to have been reached in the 
 latter half of the twelfth century, e.g. at EASTBOURNE (216.5). It was naturally 
 confined chiefly to large town-churches. LOUTH CHANCEL (213) is c. 1400 to 
 1445-t 
 
 The chancel of Burnham Overy, Norfolk, has a beautiful Early English aisle 
 running along the whole of the south side of the chancel. Wimmington, Beds, 
 which was built before 1391,+ has full-length chancel-aisles ; but the}- are partly 
 separated from the chancel by solid walls. 
 
 Seventh Type : The Aisled Chapel. 
 
 As we have seen, a constant characteristic of the English church was its 
 division into nave and chancel. These were separated by a chancel-arch. But 
 as time went on,§ the ritualistic distinction between the layman's nave and the 
 chancel of the clerk and of the holy altar was emphasised still further by a 
 screen. These screens constantly grew in magnitude and importance ; e.g. that 
 
 * For plans of the above see Brandon's Parish Churches; Parker's Churches of Northants; 
 Ringstead, 73. For Methley, where the Waterton chantry is flush with the chancel, see 
 plan in Churches of Yorkshire. 
 
 + J. J. Creswell, Associated Societies' Reports, July 6, 1897. 
 
 % Brandon's Parish Churches, 94. 
 
 § It should be noted that some screens at any rate existed in Norman times ; e.g. in 
 front of the upper story of the presbytery of Compton, Surrey. 
 
PLANNING OF PARISH CHURCHES. 227 
 
 of NEWARK (178). To get room for them the chancel-arch was built broader 
 and broader, till, as at LOUTH (213), where, unfortunately, the screen has dis- 
 appeared, and still more at NEWARK (178), there is internally little left of the 
 transverse wall in which the chancel-arch is set ; while, externally, the distinction 
 between nave and chancel almost ceases to exist. It was but a short step to 
 remove the chancel-arch and its responds altogether, and the church was thrown 
 into the shape of a great oblong aisled chapel.* In such a church, of course, a 
 screen to delimit nave and chancel was indispensable. 
 
 We have called it " the aisled chapel " plan. And indeed an early and impor- 
 tant example of it is at ST NICHOLAS, LYNN (2 1 7), which, though on a vast scale,f 
 was nothing but a chapel-of-ease to the great Church of St Margaret's, and was 
 originally debarred from the solemnisation of the rites of baptism and marriage. 
 A long account might be given of the attempts of St Nicholas to get rid of its 
 dependency on St Margaret. I At any rate, when it was rebuilt between 141 3 
 and 1419J it would seem that the authorities of St Margaret insisted that even in 
 plan St Nicholas should not be a church. And so it was built chapel-fashion 
 with a chancel-screen instead of a chancel-arch. || But the new chapel was a 
 glorious success, and man}' a church proper, when prejudice against adopting a 
 chapel-plan had had time to evanesce, was built after the fashion of St Nicholas 
 of Lynn. It was not, however, for nearly a century that this fine type of church, 
 nearly always built with loft)' and almost continuous clerestory windows,' came 
 into general use. St Stephen's, Bristol, A.U. 1455, is an early example. It 
 occurs in the village church of GRESFORD (214). Not far away is WREXHAM 
 (214), which was originally of this type. But the Wrexham people seem to have 
 been shocked at the absence of a chancel ; and so converted the east window into 
 a chancel-arch (leaving the stumps of the tracery sticking out), and built a little 
 chancel behind it. In London we have St Andrew Undershaft, Leadenhall Street, 
 1520-1532. At York there is St Michael's, 1 535-1 545. But the special habitat 
 of this grand and final type of the English parish church is East Anglia. 
 There magnificent examples occur in St .Andrew's, St Stephen's, and St Peter 
 Mancroft, Norwich; and, above all, at LONG MELFORD, Suffolk (214); all, 
 however, have lost their screens. It should be noticed how gloriously lighted 
 
 * In churches with unaisled naves the screen naturally had been employed instead of the 
 chancel-arch long before ; e.g. at Temple Balsal, which is before 1312 ; and at Ewerby, which 
 is c. 1340. Almost all the churches of Cornwall omit the chancel-arch. 
 
 t It is internally 193 feet long by 81 feet wide ; and is divided into thirteen bays. 
 
 % See Monograph on St Nicholas by Mr E. M. Beloe. 
 
 § A resolution of the Town Council, dated 14 19, speaks of it as having been recently rebuilt 
 by voluntary subscription. 
 
 || Unfortunately the chancel screen has been destroyed here, as in most churches of the 
 aisled chapel type. From the interior of King's College Chapel, Cambridge, one can form an 
 idea of the original aspect of these interiors. 
 
 1" Still earlier is the plan of the great church at North Walsham, built after the destruction 
 of the old church in the peasants' rising of 1381. The porch seems to lie a later addition to the 
 church, and contains the Royal Arms, in which the first quarter is ancient ; i.e. semie de fleurs- 
 de-lis ; a bearing which was disused in 1405. This church therefore would seem to have been 
 rebuilt, together with the porch, between 1381 and 1405. This early date is borne out by the 
 flowing tracery of the eastern windows and of some of those of the porch. 
 
228 
 
 PLANNING OF PARISH CHURCHES. 
 
 were these " Tudor lanterns " ; with a large window in every bay of the aisles,* 
 and two in every bay of the clerestory ; and a great east window, and a great 
 west window ; especially when, as at St Nicholas, Lynn, ST STEPHEN'S, NORWICH, 
 TERRINGTON ST CLEMENT'S (92), and St John's, and many other churches in 
 the Norfolk Marshland, the tower was built to the south-west or the north-west, 
 so as not to obscure the great west window. The seventh type of plan, as it was 
 the last, so was the grandest of all. 
 
 One further step might have been taken, and probably but for the collapse 
 of church architecture at the Reformation would have been taken ; that was 
 to remove the aisles as well : it would have been to bring in the unaisled 
 chapel type. This indeed had always been employed with frequency in smaller 
 
 Norwich St Stephen. 
 
 Norwich St Stephen. 
 
 churches ; e.g. at Skirlaugh. Moreover, vast unaisled Lady Chapels, such as 
 those of Peterborough and Ely, clearly pointed that way. KING'S COLLEGE 
 CHAPEL, Cambridge (62), and Eton College Chapel are but versions of the 
 Lady Chapels of Peterborough and Ely. Had this type come into general use, 
 we should have had the logical outcome of that desire for simple and straight- 
 forward planning which was ever at the heart of the English builder, whether 
 he was building for monk or canon, for townsman or for villager. 
 
 South of the Loire, this Hall-type of church became exceedingly common'; 
 from Angers, where the cathedral was built (with transepts, however) c. 11 50'; 
 south to the Mediterranean ; leaving us the great Church of the Cordeliers, 
 
 * Long- Melford actually has two windows in each bav of the aisles. 
 
PLANNING OF PARISH CHURCHES. 229 
 
 Toulouse,* and the magnificent Cathedral of Albi ; which has neither aisles 
 nor transepts, but is simply one vast hall, 355 feet long. 
 
 With us the history of church planning ends with the great Chapel of 
 King's College, Cambridge, one of the supreme glories of Christendom. Per- 
 verted as unpatriotic must the imagination be that can see decadence and 
 debasement in this great monument of mediaeval art. The three Royal Chapels, 
 at Cambridge, Windsor and Westminster, with the East Anglian churches, 
 present to us English Gothic planning at its very noblest and its best. 
 
 * Illustrated in Fergusson, ii. 70 ; recently burnt. 
 
Chapter XIV. 
 
 THE BASILICAN COLUMN AND THE ROMANESQUE PIER. 
 
 Basilican Colonnades — The Basilican a Distinct Style — Why the Basilican Style 
 Perished — Pre Conquest and Norman Piers — Origin of Compound Pier — 
 Romanesque Piers — Vaulting Shafts — Cylindrical Piers. 
 
 The history of the Christian Church practically commences in the fourth 
 century with the great Christian basilicas built by Constantine and his successors 
 (see 146). Unfortunately, however, for the early Christian architects, the great 
 ages of Roman building had passed away. Nothing was ever done again 
 greater than the Pantheon, erected by the Emperor Hadrian, greatest of all 
 the Roman builders. One great building, indeed, commenced by Maxentius in 
 the last years of the third century and finished by Constantine, rivalled the 
 triumphs of the second century. This was the so-called Basilica of Maxentius 
 or Constantine.* To what a depth of degradation Roman art had sunk is clear 
 from the history of the triumphal arch of Constantine himself. Master though 
 he was of a whole world, he could only get giallo antico columns of the yellow 
 marble of Tunis for his arch by robbing that of his predecessor Trajan ; so it 
 was also with the sculptures. And it was the methods of Constantine's arch, 
 not the engineering triumphs of the Basilica of Constantine, that the Roman 
 Christian copied ; not the genuine, indigenous arcuated building in brick 
 and concrete, with the scientific disposition of thrust and abutment which the 
 mediaeval builders groped after all through the dark ages till the twelfth 
 century, oblivious of the fact that in the Basilica of Constantine the problems 
 that puzzled them were already solved ; but the Romano-Greek style, the style 
 of the colonnade ; a style which shirked every engineering problem. But a 
 colonnaded style requires a great wealth of columns. Those in the better days 
 of Rome had not been built in drums as in the Parthenon, but were monoliths 
 of rare marbles, of red granite and blue granite, of serpentines and breccias, ot 
 onyx and alabaster, and the yet more precious porphyry ; from Paros and 
 Lesbos and Rhodes and Thasos and Eubcea ; from Sicily and Spain ; from 
 Tunis and the Atlas mountains ; from the islands of the Nile and Nubia ; from 
 Phrygia, Cappadocia, and Persia : t picked for their beauty or their rarity ; 
 quarried and transported at vast expense all over the Roman world ; hewn, it 
 may be, above the precipices east of the upper Nile ; let down the mountain side 
 by cable and capstan ; transported on rollers over the plain ; put on board barge 
 or raft ; floated down the Nile ; transhipped from barge to sailing ship ; brought 
 across the Mediterranean to Ostia ; put on barge again ; towed up the Tiber to 
 
 * It used to be called, wrongly, the Temple of Peace. 
 
 t See Corsi's list of Roman marbles in the Builder, March 9, 1889 ; and \Y. Brindley in 
 Builders' Journal, Nov. 15, 1899. 
 
BASILICAN COLONNADES. 231 
 
 Rome ; and then as Martial tells us in one of his epigrams, the corner houses of 
 the narrow streets of old Rome had to be taken down to let them pass. Such 
 magnificence of procedure was beyond the resources of the Christians ; even 
 beyond the resources of the Christian Emperor. As it was with Constantine's 
 Arch of Triumph, so it was with the early Christian churches ; it was easier to 
 beg or buy and re-use than to cut and transport new columns. It is true that at 
 S. Maria Maggiore, Rome, the magnificent ranges of white Ionic columns were 
 cut for the positions they occupy, and there are other instances ; but the) - are 
 exceptional. The usual practice was to collect as many marble columns, bases, 
 capitals, strings, architraves, friezes, and cornices as could be got, and to piece 
 them together as best might be. Thus at St Paul's extra muros, before the fire, 
 there were thirty-two major columns ; of these sixteen were from some Pagan 
 building, and the other sixteen had been cut to match. In Pelagius' work 
 (c. 578)* in S. Lorenzo there are two columns of Hymettian marble and ten of 
 pavonazzetto ; ten of the capitals are Corinthian ; two are composed of greaves 
 and bucklers and armour, plainly from some temple of Bellona or Mars ; above, 
 hardly any two blocks of the entablature correspond either in design or in height. 
 At S. Maria in Ara Coeli, where the colonnade may be sixth century, the 
 melange is something extraordinary. Of the twenty-two columns eighteen are 
 of Syenitic granite, two of fluted white marble, two of cipollino from Eubcea. 
 Bases and capitals vary still more ; some are perfect, some are truncated 
 Corinthian capitals ; some are Composite, some Tuscan ; plinths and bases are 
 of all sizes and design. Such was the church even in the metropolitan city of 
 Western Christendom : like some tramp's suit of clothes ; coat begged from one 
 man ; shirt stolen from another ; trousers a patchwork, pieced together from the 
 remains of three separate scarecrows. 
 
 The Basilican a Distinct Style. 
 
 But heterogeneous as were the materials of the Roman basilicas, when put 
 together they produced interiors which were convenient for the collective worship 
 of a great multitude, while as yet there was but one altar; and the) - were well 
 lighted, economical, and handsome. Few interiors in Christendom are more 
 impressive than that of St Paul extra muros at Rome, as rebuilt. 
 
 What was done at Rome was imitated elsewhere. Constantine built at 
 Bethlehem t a great double-aisled basilica. At Ravenna we have still surviving 
 a magnificent group of basilicas of the sixth century. In Constantinople the 
 traditions of the indigenous Roman building — that in brick and concrete — sur- 
 vived and were yet further developed ; and a new and grander style was evolved 
 in the days of Justinian (sixth century), and produced S. Sophia and the Church 
 of the Apostles. But even in Constantinople, the home of the Byzantine style 
 and the domical church, basilicas also were built; e.g. St John Studius. j To 
 the same type belong Baquoza in Syria ; § Orleansville in North Africa, and 
 man) - others in the Byzantine dominions. 
 
 In Western Christendom the basilican church long prevailed with hardly a 
 
 * Cattaneo, 41, ascribes these columns to A.D. 325. t Plan in Fergusson, i. 419. 
 
 + Illustrated in Fergusson, i. 421. § Illustrated in De Vogue. 
 
232 BASILICAN COLONNADES. 
 
 rival. In Germany many early naves of this type survive; e.g. Limburg (1030- 
 1042); Hersfeld (1040); Alpirsbach (1 100) ; to which maybe added Constance 
 (1052) and Schaffhausen (105 2- 1064) in German Switzerland* In France a good 
 example is seen in the nave of St Genou (Indre), which may be early in the 
 eleventh century;f and quite as early are the two bays of a basilica which now 
 terminate the Gothic Cathedral of Nevers to the west. In England the nearest 
 approach we have to the columnar basilica is St Peter's, Northampton.^ 
 
 In Italy, its native country, the basilica naturally is very common. We 
 may mention Agliate, 881 (Cattaneo) ; Grado, ninth century (Rohan de 
 Fleury) ; Como, S. Abbondio, consecrated 1095 ; Pisa Cathedral, commenced 
 c. 1063 (Rohan de Fleury); Lucca, S. Frediano, 11 12; and with pointed 
 arches, in the thirteenth century, the cathedrals of Genoa and of Monreale, 
 Sicily ; and in the fourteenth century, Toscanella. Even well on in the 
 Renaissance period the Basilican type was still employed ; e.g. the splendid 
 Church of S. Annunziata, Genoa. § A very high development of the basilica 
 was reached at Pisa and Lucca, where it has been denominated sometimes 
 Round-arched Gothic, sometimes Pisan Romanesque. It is simply Basilican. It 
 is no more Gothic at Monreale than it is Romanesque in Pisa. In Sicily, 
 perhaps, its most notable triumphs were won. When decorated, as at Monreale, 
 with veneer of marble and mosaic, irradiated by a brilliant southern sun, it gave 
 a glory of internal colour that vies even with Chartres and Le Mans. As we 
 have seen, its distribution was wide as Christendom, though it was more 
 especially the heritage of the Western Church. And it went on in use from 
 the fourth to the fourteenth century. 
 
 Why the Basilican Style Perished. 
 
 Wide, however, as was the distribution of churches of Basilican style, that 
 style was doomed to perish ; it carried within it the seeds of its own dissolution. 
 On the one hand it relied on a supply of columns which it had not wrought 
 itself; on the other hand, even if obtainable, columns were unsuitable for the 
 work the mediaeval builder was planning to do. 
 
 Where possible, as at Rome, the columns were monoliths, often brought 
 great distances. || But the supply of monoliths failed, even in Rome itself. No 
 wonder ; for whole families made a respectable living for many centuries by kilns 
 built in the middle of the city, in which they burnt to lime the marbles of 
 Imperial Rome. Further afield, in our own remote country, for instance, mono- 
 liths of any considerable size would be still scarcer. In any case the supply 
 failed ; and the monolithic column had to be built in drums or in small blocks. 
 
 Secondly, the column of classical proportions was too weak. Three types 
 of column had been in use in ancient Greece and Rome ; Doric, Ionic, and 
 
 * Dehio, Plates 52, 55, 56. 
 
 t Enlart's Manuel, 424 ; for illustration of interior see Corroyer's Architecture romane, 177. 
 
 I Illustrated in R. Robert, ii. 67. § Illustrated in Fergusson's Modern Architecture. 
 
 II We hear of columns brought from Ravenna in the ninth century to Aix-la-Chapelle by 
 Charlemagne. Columns were brought to Venice for St Mark's in the eleventh century from all 
 round the Mediterranean. Some of the columns of rare marbles round the so-called Trinity 
 Chapel at Canterbury must have been brought from some Mediterranean countries. 
 
PRE-CONQUEST AND NORMAN PIERS. 233 
 
 Corinthian. Of these the Roman Doric, with its variant, the Tuscan, seems 
 seldom t<> have been employed in the Basilican churches. The columns were 
 usually Corinthian ; much less frequently, as at S. Maria Maggiore, Ionic. Both 
 the Ionic and the Corinthian column are slender and graceful, but by no means 
 strong. They had not been designed to carry heavy weights. When, therefore, 
 the early Christians put on a pair <>f columns an arch, and that arch bearing 
 its share of clerestory wall and roof, they were giving the column work to do 
 which it had not been designed to do. The load might indeed be lightened, 
 as it was in the Byzantine basilicas — e.g. St John's Studius, Constantinople, 
 and S. Lorenzo and S. AGNESE, Rome (155J, in both of which Byzantine 
 influence appears — by inserting an open arcade in the wall above the colonnade, 
 but the columns had to be set very close together ; whether they carried an en- 
 tablature (i.e. horizontal lintels or beams of stone), as at S. Lorenzo ; or arches,* 
 as at St Paul extra muros. Set so closely, the long succession of columns 
 and narrow arches produced a converging perspective, which, as later on in 
 Romanesque and Gothic, increased the apparent length, and greatly added 
 to the internal effect of the churches. On the other hand, such a close-set 
 colonnade so obstructed the aisle that from it little of the service in the 
 apse could be seen. So therefore, in spite of the beauty of the classical column 
 with height nicely adjusted to diameter, partly because no more monoliths were 
 available, and coursed columns were difficult to build, and when built not very 
 handsome ; partly because the colonnades blocked the aisles ; most of all 
 because the colonnades were unable to carry the heavy weights with which it 
 was desired to load them ; some other form of support had to be devised. That 
 form of support was found in the pier.t 
 
 Pre-Conquest and Norman Piers. 
 
 As regards our own country, it is clear that in the time of the Roman 
 occupation colonnaded buildings were erected. The remains of one on a 
 vast scale have recently been discovered in constructing a sewer at Lincoln. % 
 
 * It is usually assumed that the first example of an arch resting direct on columns with- 
 out the interposition of a piece of entablature is that of a gateway in Diocletian's palace at 
 Spalato, late in the third century. But a wide arch so resting is seen at Atil in the Hauran, 
 A.D. 151 ; and still earlier in the Propylasa at Damascus; illustrated in Anderson and Spiers, 
 167, 168. Also columns carrying an arch without intervening entablature occur at Pompeii. 
 
 t All through the book the term pier has been used in the generic sense of a "support" ; 
 as one speaks of the " piers" of a bridge ; i.e. the masses of masonry on which the arches rest. 
 A row of arches with their piers has been called a pier-arcade. The term column has 
 been avoided except of a cylinder classical in its proportions ; i.e. with height and diameter 
 correlated, or of a stout shaft forming part of a compound pier. A very broad, massive column 
 has been designated a cylindrical pier or a cylinder. A very thin, attenuated column is spoken 
 of as a shaft. A somewhat stout shaft partly sunk in the masonry of a compound pier is an 
 engaged column. Shafts and columns are either detached; or engaged, i.e. "attached." 
 
 A compound pier is one which is made up of an assemblage of shafts or columns round a 
 central mass, which may be circular, rectangular, or polygonal : e.g. the Norman piers of 
 Durham and Peterborough ; the Gothic piers of Ely choir and Chichester presbytery. Or, it 
 may be composed of a cluster of engaged columns, without any shafts ; e.g. those of Peverley 
 Minster and Exeter Cathedral. 
 
 I Two of the bases of these immense columns are to be seen in a cellar. 
 
2 34 ORIGIN OF COMPOUND PIER. 
 
 To Anglo-Saxon days, probably,* belong the two columns, which originally 
 formed part of the triple chancel-arch of Reculvers, and which now stand on 
 the lawn to the north of Canterbury Cathedral. Two more, also in drums, have 
 been placed in the porch of the Anglo-Saxon church of Repton. The vault of 
 Repton crypt also is supported by four columns ; these are monoliths. And 
 there are examples in Northumberland,! where the columns may be Anglo- 
 Saxon, or may be Roman columns re-used. No general use, however, of the 
 column between nave and aisles in Anglo-Saxon churches can be substantiated. 
 Had colonnades been in general use, fragments of columns, perhaps whole 
 columns, would frequently be found on the site of Anglo-Saxon churches ; 
 which is not the case. We can hardly doubt that where aisles existed in 
 the Anglo-Saxon churches, they were separated from the nave, not by a 
 colonnade, but by a pier-arcade. At any rate, that is the case in the only 
 aisled Anglo-Saxon churches which survive ; those of Lydd, Brixworth, and 
 Wing. | The history of supports of our English churches begins then with the 
 pier, not with the column. 
 
 The same is the case with the greater Norman churches of the eleventh 
 century, whether in this country or in Normandy. Two forms of support are 
 usually found in the eleventh century ; either compound piers, as at Bernay § 
 and Chichester ; or an alternation of compound pier and cylindrical or octagonal 
 pier, as at Jumieges, Durham, Peterborough ; seldom cylinders || or octagons 
 only. Plainly, here and in Normandy, the pier is as much the normal form of 
 support as the column in Southern Christendom ; perhaps because there were 
 fewer Roman columnar buildings to rob in Northern Europe, or to suggest a 
 church with colonnaded aisles. 
 
 Origin of Compound Pier. 
 
 For the genesis of the compound pier we must go abroad. As has been 
 pointed out, we have no Romano-British church, and but few Anglo-Saxon 
 aisled churches surviving. Every one of the larger Anglo-Saxon churches 
 was pulled down long ago,H and most of the smaller ones, to be rebuilt in 
 Romanesque or Gothic fashion. But, abroad, we can trace back the genesis and 
 development of the pier, at any rate to the ninth century. Sometimes it 
 replaces the column altogether. But sometimes the column makes a struggle. 
 For instance, at S. Prassede, Rome, as built in the ninth century, all the supports 
 were columns. But it was desired to introduce diaphragmatic roofing.** For 
 this purpose it was necessary to build broad arches across the nave carrying 
 a heavy load of masonry.tt Columns were utterly inadequate to support such a 
 
 * See G. E. Fox in Builder, Oct. 27, 1900. t Baldwin Brown, 8. 
 
 % Plan of Wing in Scott's Essay, 44. § Rupvich-Robert, Plate xi. 
 
 || There are short, massive cylinders in the choirs of GLOUCESTER (294) and Tewkes- 
 bury (165) ; both begun in the eleventh century. 
 
 IT The last to survive was the nave of Hexham. See C. C. Hodges' sumptuous monograph 
 on Hexham Abbey. 
 
 ** Every fourth clerestory window was blocked up by an arched stone principal of the roof, 
 but may still be traced in the outer wall. 
 
 ft As shown in Dehio, Plate 45. 
 
DIAPHRAGMATIC ROOFING. 
 
 load; and accordingly even' third column was replaced by a massive pier. 
 Similarly the eleventh-century church of s. mixiato. FLORENCE* and the 
 twelfth-century church of S. Zeno, Verona, and the contemporary cathedra] 
 of Modena.f were built with a similar alternation of column and pier. In 
 Normandy also and in Northern France roofs with arched stone principals 
 were constructed here and there: e.^. at Notre Dame du Pre, Le Mans;} at 
 St George's de Boscherville ; § at Cerisy-la-Foret. The same form of roof may 
 have been in use at the Abbaye-aux-Hommes at Caen." 
 
 These diaphragmatic roofs, with arches carrying a gabled transverse wall, 
 simplified the construction of the timber roof, for the purlins could rest directly 
 on the gables ; and moreover they divided 
 it up in separate compartments ; so that 
 if one compartment of the roof was burnt, 
 the flames, owing to the party-walls, could 
 not reach the adjoining compartments. 
 As was natural, they were employed in 
 the aisles also. The remains of such 
 arches are to be seen over the aisles of St 
 Peter's, Northampton ;** they are to be seen 
 perfect at Chivy, near Laon, in the eleventh 
 century ;ff they were intended at Thaon.** 
 Enlart §§ cites twelve other examples 
 of aisles with diaphragmatic roofs in Nor- 
 mandy, Brittany, and the north of France. 
 
 It is possible that in this system of roofing we see a constructional origin 
 for the compound pier. In its simplest form, as at S. Eustorgio, Milan, to the 
 back of each pier there was added a pilaster facing another pilaster on the aisle 
 wall. On these pairs of pilasters arches were built across each aisle, of which 
 two still remain.* r A similar example occurs in the tall oblong piers oi STE. 
 
 S. Miniato, Florence. 
 
 * Illustrated in Norman Shaw's Architectural Sketches, Plate 47. 
 
 + Willis' Middle Ages, 89 ; and Dehio, Plate 66. 
 
 1 Dehio, i. 286 ; and Enlart's Manuel, 264. 
 
 S See perspective in Choisy, ii. 193. 
 
 || See Dehio, Plate 90: Enlart's Manuel, 264 ; St Paul in Planat, vi. 24 : and R. Robert, 
 Plate 72. Before the present vaults of lath and plaster were put up in 1868, R. Robert was 
 able to see still in situ the three lowest voussoirs of one of the transverse arches of the nave. 
 
 1" As shown in Dehio, Plate 87. It may have been intended, even if not carried out, in the 
 transepts and nave of Durham. This may be suggested as an explanation of the roof-shafts, 
 rendered objectless by the present vault, which run up in the end bays of the east sides oi the 
 Durham transepts. So also in the nave, the bottommost voussoirs of the great transverse 
 arches of the vault are much broader than those above them. This must mean that broader 
 transverse arches than the present ones were intended and were actually commenced; arches 
 strong enough to carry either arched stone principals or else a groined vault. (See section 
 in Hilling's Durham, Plate 15; and Bilson's Beginnings, 315, 26. It should be noted that 
 Modena Cathedral has gone through a similar series of changes to those suggested above for 
 Durham. (See Willis' Middle Ages, 89 ; and Dehio, Plates 66 and [62. 
 
 ** See R. Robert, ii., Plate 67. ++ Illustrated in Enlart's Manuel, 263. 
 
 \% R. Robert, 136. i; Manuel, 204. 
 
 Illl Tenth century, Cattaneo and Kivoira. HU Cattaneo, 245. 
 
236 
 
 ROMANESQUE COMPOUND PIERS. 
 
 SKETCH PLAN 
 Beziers. 
 
 *±l 
 
 APHRODISIE, BEZIERS, which appears to be of the tenth or early eleventh 
 
 century ; and in the church of Vaux-sous-Laon. 
 
 The next step is seen at the dated Church of SS. FELIX 
 AND FORTUNATUS, outside Vicenza (A.D. 985), where there 
 are cased up in the walls columns alternating with com- 
 pound piers. The latter are cruciform ; they have pilasters 
 to east and west, and engaged columns to north and south, 
 as at S. Miniato, Florence, illustrated above. This is a 
 great step in advance.* A further advance is seen in the 
 primitive Romanesque church of S. Stefano, Bologna ; 
 where there is a square pier with four engaged columns. 
 A yet further stage is seen at Jumieges.f and in our church 
 of ST PETER'S, NORTHAMPTON (663. 1), c. 1 160 ; in the latter 
 there is simply a cluster of four engaged columns. Each 
 one of these has its own special work to do. The front 
 column rises to the roof, and supported the tie-beam ; the 
 back column supported the diaphragmatic arch spanning 
 the aisle. The pier-arches are in one order only ; and the 
 springing of each arch rests on an engaged column. The 
 whole pier, therefore, is thoroughly logical and rational ; 
 each member of it has its own function ; each has a 
 different load.^ A 
 less logical pier is 
 
 seen in the Norman church of St James, 
 
 Bristol ; it also consists of a cylinder with 
 
 four columns attached. 
 
 In such examples as the above, it is 
 
 plain that the compound pier is construc- 
 tional in origin. 
 
 But possibly the idea of such construc- 
 tion was also suggested by the sight of 
 
 actually existing Roman examples of it. 
 
 For the compound pier was perfectly well 
 
 known to the Romans. It occurs in the 
 
 peristyle of a house at Pompeii ; in the 
 
 portico of the court of the Temple of the 
 
 Sun at Palmyra ; on the arch at Damas ; 
 
 in the Temple of Vespasian at Brescia ; at 
 
 Palatitza.§ In the Piscina Mirabile at 
 
 BaiaeJI the piers of the cistern are cruciform, 
 
 there being pilasters on each of the four sides. Rectangular compound piers occur 
 
 in the Roman baths at Bath. At Germigny, consecrated 806, antique shafts are 
 
 * Illustrated in Cattaneo, 248. f R. Robert, Plate 13. 
 
 I See section of St Peter's in R. Robert, ii. 67. On the compound pier see Willis' Middle 
 Ages, go, seq. 
 
 § De Dartein, ii. 467. 
 
 || Illustrated in Willis' Middle Ages, Plate v., 5, and text, 79. 
 
ROMANESQUE COMPOUND PIERS. 237 
 
 applied to the piers:* as indeed was the constant practice in the Roman Thermae. 
 Compound piers or responds occur also in Anglo-Saxon, where they arc- 
 probably an indication of late date ; e.g. in the tower arch of Sompting ; and at 
 Bosham, Wareham, and Broughton.f 
 
 When, therefore, the Normans commenced to build their great churches in 
 England, the compound pier had already a long pedigree. Their piers, however, 
 were much more advanced than such as those of S. Miniato, Florence. This was 
 partly because they had further developed the arch, and partly because they had 
 learnt to vault the aisles. Both their arches and their vaults needed additional 
 supports. Arches in the eleventh century were now generally double ; i.e. 
 each arch was compounded of two independent arches, of which the upper — 
 it is called the outer order of the arch — was broader than the lower — or inner 
 order — and overlapped it. Such an arch required three supports at each 
 springing on which to rest. At the junction of two arches, therefore, six columns 
 or shafts were needed. Moreover, usually another shaft ran up the face of the 
 pier — what is called a roofing shaft or a vaulting shaft — up to the wooden roof, 
 or to the vault, if there was one, of the nave. This increased the number of 
 supports to seven. But at the back of the pier there was a low transverse arch 
 which spanned the aisle, and of which one end rested on a wall shaft or a wall 
 pilaster, while the other required for its support a column attached to the pier. 
 This raises the number of supports to eight. And it is just this pier, with four 
 engaged columns and four detached shafts, which we find at Cerisy-la-Foret ; \ 
 St Nicholas, Caen ; in the eastern bays of the nave of Lessay ; in the triforium 
 of the Abbaye-aux-Hommes, in St George's de Boscherville, and in the west aisle 
 of the north transept of ELY (659.2 .§ In all these the outer order of the pier- 
 arch is supported by shafts ; the other four supports are stout columns. Such 
 a compound pier may be described as a pair of intersecting oblongs, with eight 
 columns and shafts. 
 
 In the west aisle of the Norman NORTH TRANSEPT OF WINCHESTER || 
 (659.4;, the number of supports is increased to ten. This is because shafts 
 were added to carry the groins of the vault of the aisle. These are carried by 
 a pilaster in the AISLE OF Norwich NAVE 23S . 
 
 In DURHAM choir* ('659.1; the number of supports rises to twelve. This is 
 because the pier has on its north and south faces not single, but triple vaulting 
 shafts. 
 
 * Enlart's Manuel, T7~. t Illustrated in Baldwin Brown's Arts in Early England. 
 
 1 R. Robert, Plate 73. 
 
 § But the eight supports might be utilised for other purposes ; e.g. in the Ely pier, '>59.- ; 
 where d, d carry the inner orders and e, e the outer order of the pier-arches; b carries the 
 transverse arch and c, c the groins A, a of the aisle vault ; a is the rooting shaft of the central 
 span of the transept. Plans of all these Norman piers in Bilson's Beginnings, 291, 306, 310. 
 
 || In 659.4, A, A are the groins of the aisle vault: B, 1:, B, B are the outer orders and 
 C, c the inner orders of the pier-arches ; a is a roofing shaft of the central span of the transept ; 
 /' supports D the transverse arch of one order of the aisle vault ; c. c support the groins of this 
 vault ; (/, d support the inner orders and e, <\ <•, c the outer orders of the pier-arches. See 261. 
 
 1 In the pier of Durham choir, 659.1, d, d support the inner orders and e, e, e, e the outer 
 orders of the pier-arches ; /; supports the transverse arch and c, c the diagonal ribs ot the 
 aisle vault; while a, /,/ is a triple vaulting shaft rising nearly to the clerestory. See 306, 
 315- 
 
238 
 
 ROMANESQUE COMPOUND PIERS. 
 
 Norwich Nave, North Aisle. 
 
 Peterborough Nave, South Aisle. 
 
 In the central transept of Lincoln and in the western bays of the nave 
 of Worcester Cathedral, as at SENS (107) and Laon, there were also wall-ribs 
 (formerets) round the clerestory windows. In RIPON CHOIR a vault of similar 
 character seems to have been intended (102). For each of these wall-ribs two 
 more vaulting shafts were required, raising the total of vaulting shafts to five,* 
 and still further increasing the complication of the pier. Again, we have spoken 
 of the pier-arches as constructed in two orders. But sometimes, as at St Albans 
 and PETERBOROUGH, they were constructed in three orders ; in this case two 
 more supports on the east and two on the west side of the pier would be 
 necessary to carry the additional outer order ; an addition of four shafts. 
 Again, the transverse arch spanning the aisle was usually built in two orders ; 
 this would necessitate two more shafts at the back of the pier. 
 
 So far we have spoken as if the supports devised for the orders of the pier- 
 arch and the groins or ribs of the vaults were always either shafts or columns. 
 This was not so ; pilasters might be substituted for any or all of the shafts and 
 columns. Thus, at ST ALBANS (659.5), where, owing to scarcity of freestone, 
 the pier and arches are constructed with Roman brick from the ruins of the 
 neighbouring Roman town of Verulamium, pilasters only are employed ; as 
 also at Autun.f and constantly in the Romanesque of Provence and Palestine. 
 
 It has been pointed out above that in the best examples, such as Durham 
 
 * There are five vaulting shafts in the western bays of Worcester nave, 
 t Illustrated in Viollet-le-Duc, Architechtre, vii. 161. 
 
VAULTING SHAFTS. 
 
 239 
 
 choir, the composition of the pier is rigidly determined by its function as a 
 group of supports. But strict logic was often disregarded either for practical 
 or for aesthetic reasons. Thus at NORWICH (238), the pier-arches consist 
 of two recessed orders; and the second pier ("the one with the tablet) has, 
 quite correctly, a single stout and 
 ugly column to carry the broad 
 inner order. But in the alternate 
 piers, for this massive column 
 three * shafts are substituted, 
 making the pier less obstructive 
 and less heavy in appearance. 
 For the same reason each trans- 
 verse arch of the vault of the 
 aisle is supported not by one, 
 but by two shafts, t For artistic 
 reasons, also, there are four vault- 
 ing shafts in front of these piers, 
 instead of three. Thus the alter- 
 nate piers of Norwich} (659.7) 
 come to consist of sixteen shafts 
 and pilasters. In Gothic archi- 
 tecture this logic of the pier — i.e. 
 the correlation of the articulation 
 of the pier to the orders of the 
 arch and to the ribs of the vaults 
 of nave and aisle — is less often 
 insisted upon. Usually, though 
 not always, the builders were 
 willing to correlate the members 
 of the pier with the orders of the arch ; but much less frequently, at any rate 
 in England, with the far-away vault of the nave. It became customary more 
 and more to exclude the vaulting shafts from membership in the pier. 
 
 Durham Nave, North Aisle. 
 
 ; Shafts. 
 
 a vaulted nave, it is supposed that 
 in front of the pier a group of three 
 
 Vaultini 
 
 In theory, indeed, where there i^- 
 there should rise from the pavement 
 shafts to support the transverse rib and the diagonal ribs of the high 
 vault ; and if the vault has wall-ribs, there should be a group of five, But 
 we may observe that as a matter of fact the vaulting shafts do not really 
 support the ribs of the vault at all. ^ This is plain from the fact that in 
 ruined churches vaulting shafts may be seen to have perished, but the vaults 
 
 * The photograph shows two of these shafts ; but there are three. 
 
 + It should be noted that the diagonal groins or arrises of the vault are not supported by 
 shafts as at Durham and Peterborough, but by pilasters. 
 
 \ In the major pier of Norwich, 659.7, and 238, d, d, d, d, d, d carry the inner orders 
 and e, c, e, c the outer orders of the pier-arches ; </, a are roofing shafts of the nave ; /', /' 
 carry the inner order and c, c the outer order of the transverse arch of two orders of the aisle 
 vault. No separate shafts are provided for the groins of this vault. 
 
 3 So also Dehio, ii. 549. 
 
2 4 o VAULTING SHAFTS. 
 
 remain perfectly safe without them. This is because the weight of the vault 
 does not descend vertically as it appears to do ; but owing to the outer thrusts 
 of the arches of the vault, passes out more or less laterally. Therefore, if we 
 are to chop logic, the English Gothic builders were quite right, on scientific 
 grounds, in refusing, as they usually did, to bring the vaulting shafts down to 
 the pavement. It is true that there is great value in the latter treatment. 
 It provides the strongest possible vertical lines ; as well as an emphatic 
 line of demarcation between the respective bays of which an interior is 
 composed. On the other hand, it is open to the practical objection that in a 
 choir, if a group of vaulting shafts is placed in front of the piers, the choir 
 is considerably narrowed thereby. So much was this felt at Lincoln, that 
 when the present stalls were placed in St Hugh's choir, his vaulting shafts were 
 chopped off from the piers. And it is open to the artistic objection that the 
 group of vaulting shafts breaks in most ruthless fashion into the group of 
 capitals which ought, but is unable, to encircle the pier. Examples of the 
 retention of the vaulting shaft in the pier are Worcester nave, Rochester 
 presbytery, ST SAVIOUR'S SOUTHWARK (521), YORK NAVE and CHOIR (10), 
 Chester nave, Bath Abbey, and LICHFIELD NAVE (523), where great unity is 
 given thereby to the triple horizontal division of each bay. So again, when 
 WINCHESTER NAVE (342) was remodelled c. 1360, the Norman vaulting shaft 
 was retained. Some ten years later CANTERBURY NAVE (90) copied 
 Winchester ; but with the unfortunate addition of bands to the vaulting shafts, 
 obstructing the upward flow of vertical line. At ST MARY REDCLIFFE, Bristol 
 (525), c. 1442, the bands are omitted. 
 
 The more common Gothic treatment in England was to stop the vaulting 
 shafts at or above the pier. Of all the methods adopted the worst is that 
 adopted in the transept of PETERBOROUGH (161), at MALMESBURY (522), in 
 Canterbury choir, in the nave of the Temple choir, and in the central transept of 
 Lincoln ; in all of which the vaulting shafts rest on the abacus of the piers below. 
 It is very common in French twelfth-century Gothic ; e.g. Notre Dame, Paris, 
 and LAON (528). To the eye the abacus appears, and is, utterly inadequate 
 to carry the weight of the vault ; and the vaulting shafts, perched on this thin 
 slab, look painfull}- insecure. The most common English treatment is to insert 
 a corbel at a little distance above the abacus ; as in the presbyteries of LINCOLN 
 (56), ELY (117), and EXETER (241); where there are foliated corbels of 
 extraordinary beauty. Sometimes the vaulting shaft is stopped at the sill of 
 the triforium ; e.g. at New Shoreham ; ST DAVID'S (525); Hexham; and 
 Rievaulx choir. This is a less successful treatment ; it leaves a blank spandrel 
 between the abacus and the sill of the triforium. This, however, was filled up 
 at Guisborough and SELBY (390) by a statuette under a canopy. Worse still 
 is Salisbury ; where the vaulting shaft is stopped about half-way up the triforium. 
 Worst of all is WELLS NAVE (524);* where the vaulting shaft is stopped at the 
 sill of the clerestory. In one way or other, the result usually was to exclude 
 the vaulting shafts from membership in the pier. The builders thus obtained a 
 freer hand, and were able to compose their piers with reference only to the 
 orders of the arch and the ribs of the vault of the aisles. 
 
 * But see page 534. 
 
VAULTING SHAFTS. 
 
 241 
 
 Before long even this was felt to be a restraint and a bondage ; and both in 
 France* and in England the strict subordination of the supports to the load was 
 abandoned. Indeed, if we examine such a beautiful pier as that of Exeter, it is 
 plain that so far from subordinating the columns of the pier to the orders of the 
 arch, the very reverse is the case. The builder has designed the pier first, with 
 sixteen columns ; and has then made the design of the arch follow that of the 
 pier ; e.g. he has actually so grouped the moldings of his arch that he gets 
 four sham orders instead of two genuine ones ; in order to make each of the latter 
 correspond with the pair of columns beneath it. In this he has reverted to a 
 very ancient practice (for not all the Romanesque builders were logicians first, 
 artists afterwards) ; even as early as St Remi, Reims, in the tenth or eleventh 
 
 ir wpE_ • r 
 
 W 
 
 NP ro^. 
 
 £ I 
 
 V 
 
 ^jjfi7«j!L f ~V 
 
 >t 
 
 my 
 
 A 
 
 ll\ M ^ 
 
 V 
 
 \ V ' 
 
 :i 
 
 \ 
 
 \ V 1 
 
 \ 
 
 r* 
 
 \ 
 
 \t 
 
 ; 
 
 Exeter Choir. 
 
 Ely Presbytery. 
 
 century, to support arches of two orders, the builder erected a pier of fourteen 
 clustered columns.! The fact is that in Gothic, now that the piers were built 
 in good masonry, they had become so slender that there was no longer room for 
 such an apparatus of shafts and columns as we saw at DURHAM \ (659.1) and 
 Norwich. Moreover, in England, the number of constructional ribs in the vault 
 had been increased greatly by the addition of the tiercerons ; e.g. at Exeter; 
 and if all these, together with the wall-ribs, had to be found separate supports in 
 the pier, there would be no end to its complication. In fact, logic had become 
 impossible. Being sensible men, the builders recognised this, and made the best 
 compromise they could between logic and art.^ 
 
 * E.g. at Amiens. t Illustrated in Viollet-le-Duc, Architecture, vii. 155. 
 
 % This pier occupies 160 square feet. § Illogical piers are seen in hedon nave (544)- 
 
 Q 
 
242 LATE GOTHIC PIERS. 
 
 The Compound Pier in Late Gothic. 
 
 There was, however, here and there, a curious survival of strict logic as to 
 pier-composition in the late Gothic of England, and still more of France. 
 Indeed the logician bethought himself of yet further refinements. In the south 
 transept of Gloucester a slender shaft, almost a " bead," with due capital and 
 base, is provided for every rib of the high vault ; viz. one transverse rib, two 
 diagonals, two tiercerons, two Hemes ; as well as for the wall-ribs. Sometimes 
 there is a desire that even the moldings of the vault-ribs should reappear down 
 below in capital, shaft, and base. With us it produced such piers as those of the 
 nave of ST GEORGE'S, WINDSOR (255). Very interesting it is to trace out the 
 interdependence of load and support in such examples, and still more in such 
 work as that of St Urbain, Troyes ; * S. Severin, Paris ; and S. Juan de los 
 Reyes, Toledo.! Ultra-logical, however, as such piers may be, their appeal is 
 more to the intellect than to the eye : art is often sacrificed to logic ; beauty 
 of curve, alternation of high light and deep shadow may be lost ; they are not to 
 be compared for one moment with the illogical piers of the Ely presbytery or 
 Exeter choir. 
 
 Cylindrical Piers. 
 
 So far we have spoken of the Romanesque compound pier and of its 
 developments. In the eleventh century it was the prevailing form in all the 
 larger Romanesque churches built here or in Normandy. In the twelfth century, 
 on the other hand, there was a curious revival of the much older form, the 
 column ; but a column that no Greek could have looked at without a shudder ; 
 a column whose thickness was not regulated by artistical consideration of the 
 fitting ratio of breadth to height, but by the weight of the load it had to carry ; 
 a column without any tapering upward, without any central entasis ; without 
 any flutes ; with a mere impost for capital, as at GLOUCESTER (26), or 
 with a capital unknown to or strangely changed from that of Ancient Greece. 
 It is better to provide it with a new name at once, and call it a " cylindrical 
 pier" or a " cylinder," not a " column." 
 
 The present geographical distribution of our Norman cylinders is a little 
 curious : so many Norman minsters, however, have been rebuilt in Gothic, 
 that it is impossible to speak with precision. At present Norman cylinders 
 appear in the south at Colchester, St Botolph's ; St Bartholomew's, Smithfield. 
 In the Midlands they are seen at SOUTHWELL (520) ; MELBOURNE (203) ; 
 Oxford ; in the north at Carlisle and Dunfermline. But in the province of what 
 later on was to be a distinct school of Western Gothic, the surviving examples 
 are much more numerous; GLOUCESTER (294), TEWKESBURY (165), Malvern, 
 SHREWSBURY (521), Hereford, Pershore, St David's, CHESTER ST JOHN'S (448.1). 
 It is notable that these examples occur principally in naves, and that in all but 
 the later examples, such as St Cross and St David's, it seems to have been 
 intended that the nave should not have a vault. 
 
 At GLOUCESTER (26) and TEWKESBURY (297) the cylinders are of brob- 
 dingnagian proportions. Both are Benedictine churches ; and because they have 
 * Choisy, ii. 350. t Illustrated in Dehio, Plate 567. 
 
CYLINDRICAL PIERS. 
 
 ^43 
 
 a cloister on one side of the nave, the aisle windows have to be placed high in 
 the wall to clear the cloister roof. By raising the piers and arches of the nave, it 
 was possible to get into the aisle wall a taller window, and thus to improve the 
 lighting both of aisle and nave. This may explain the extraordinary proportions 
 of these piers.* 
 
 In our parish churches the cylinders, or their variant the octagonal pier, 
 remained in constant use till the end of Gothic. This was simply because a 
 more highly organised support was beyond the resources of most villages. But 
 in our larger Gothic churches we abandoned it. The French, on the other hand, 
 in the face of all the logic of construction, frequently employed the cylinder ; 
 not only round the apse, but flanking choir or nave; e.g. Rouen Cathedral. 
 Indeed in their Flamboyant work, e.g. at St Lo and Montargis, it is one of the 
 two most characteristic forms.f 
 
 ALTERNATION of PIERS. — Frequently cylinders alternate with compound 
 piers ; e.g. in Lombardy in S. Stefano, Bologna ; in Germany at Worms ; in 
 Normandy at Jumieges ; in England at DURHAM (8), Castle Acre, Furness, 
 and elsewhere. \ 
 
 * Similar colossal piers occur early in the eleventh century at Tournus, on the Saone, 
 which has a cloister south of the nave. Illustrated in Classification of Romanesque, 275. 
 
 + The other is that in which the moldings of the arch and vault are carried down the 
 pier (242). 
 
 % On alternating supports see page 317. 
 
Chapter XV. 
 GOTHIC PIERS. 
 
 The Western Pier — The Southern Pier — Shafts, Bands, Marble — The Northern Pier — 
 Clustered Columns — Ornamentation and Plan of Shafts. 
 
 In English Gothic both forms of Romanesque pier were abandoned, or at any 
 rate totally transformed. The rectangular pier nearly surrounded by shafts, as 
 at Norwich, was composed altogether of columns and shafts, with its core 
 seldom visible. The cylinder was surrounded by shafts, and produced a second 
 type. In the first type the columns or shafts are engaged ; in the second they 
 are detached. Both were in use simultaneously for about a century. Of the 
 
 Beverley Minster. 
 
 Lichfield Choir. 
 
 piers with engaged columns or shafts there were two sorts ; those which were a 
 cluster of columns, and those which were encircled by triplets of shafts ; but they 
 agreed in eschewing marble, and in being constructed usually in courses of free- 
 stone, and not in monolithic blocks. The former we may call the Northern ; 
 the other the Western pier ; the pier with detached shafts, of the Canterbury 
 type, of which we have spoken above, we may call the Southern pier. 
 
GOTHIC PIERS. 
 
 -45 
 
 The Western Pier. 
 
 This is a pier very short and massive in proportion to its height, but which 
 is prevented from looking heavy and squat by the exceptional amount of slender 
 shafting encircling it. The special feature of the shafts is that they are arranged 
 in groups of three ; and that they are constructed in coursed freestone. In the 
 western bays of Worcester nave, which may be as early as 1170, all are of 
 coursed freestone ; some are pear-shaped. The shafts are numerous and slender. 
 At WELLS (209), c. 1 175, the pier retains the Romanesque disposition of two 
 intersecting oblongs, as in the typical pier shown in 661.4, but round these rect- 
 angular forms there are disposed 
 eight triplets of shafts, twenty-four 
 in all. Piers of the same type occur 
 in St Cuthbert, Wells. In the piers 
 of Cwm Hir, now at LLANIDLOES 
 (552),* all traces of the rectangle 
 have disappeared ; the pier is fully- 
 developed and consists of eight 
 groups of triplets ; so also in PER- 
 SHORE (75J and LICHFIELD choirs 
 (661. 8). t In St Mary's, Shrewsbury, 
 being a parish church, the piers are 
 small and of early type ; a rectangle 
 surrounded by four triplets.* LLAN- 
 DAFF (424) has a simpler pier, con- 
 sisting mainly of plain chamfers ; 
 but still there are triplets of shafts 
 attached instead of single columns. 
 In the western piers of LICHFIELD 
 CHOIR (244) the shafts are arranged 
 in triplets. Of these the nave-piers 
 are a variant (424.8). Similar triplets 
 occur in the piers at the back of the 
 high altar of Dore Abbey, Hereford. 
 They occur again at Christ Church, 
 Dublin. Their distribution is very 
 
 remarkable. They are all to the west of a line drawn from Wells to Lichfield § 
 and Chester, and including Christ Church, Dublin ; which we know was built 
 of Somersetshire oolite by masons from Bristol and Pembroke. In this western 
 area the triple-shafted pier reigned almost alone from c. 1 170 to c. 1200. 
 
 Chichester Retrochoir. 
 
 * See capitals of these piers (422). 
 
 t A curious parallel is seen in the fuurteenth-century piers of S. Lorenz, Nuremberg ; 
 which have eight triplets (I)ehio, Plate 566). 
 
 I See capitals of these piers (424.7). 
 
 § The Transitional piers of St Sepulchre's, Northampton, are an "outlier" {Nene Valley, 
 Plate 16). 
 
246 GOTHIC PIERS. 
 
 The Southern Pier ; with Detached Shafts. 
 
 We will turn next to the pier with a cylinder for core, and round it dis- 
 posed a number of shafts, some or all of which are detached.* The detached 
 shafts are seldom built in drums ; but if the shaft be short it is a single block ; 
 if long, it is composed of several blocks joined together. The use of monolithic 
 shafts was well known to the Romans ; they were turned in the lathe.f Roman 
 monolithic shafts remain with the characteristic tapering and entasis. One 
 occurs in situ at Chester. \ Two are re-used in the west doorway of St Woolos, 
 Newport. Others, with the marks of the turning-tool upon them, may be seen 
 in the museums of Leicester and Rouen. The Anglo-Saxons were expert in 
 turning baluster shafts in the lathe ; and four of their monolithic columns remain 
 in Repton crypt. Monolithic shafts occur in the Norman work of Gundulph in 
 Rochester crypt, and of Ernulph in Canterbury crypt. They are very common 
 in Norman doorways and windows at all periods. Moreover, about the middle of 
 the twelfth century and onward, we have a large number of fonts, e.g. that in 
 Winchester nave, the bowls of which rest on a central cylinder surrounded by four 
 or eight shafts. This is precisely the plan of the Early Gothic piers of CHICHESTER 
 RETROCHOIR (245), Boxgrove, and St Thomas, Portsmouth. In these three 
 the shafts are very widely spaced from the central cylinder ; on the other hand, 
 in the piers of Canterbury choir they are set close to the cylinder ; it was the 
 latter precedent which most often found favour. The piers of ELY RETROCHOIR 
 (247), however, incline to the Chichester type. 
 
 The number of shafts in these piers varies greatly. Often, e.g. in Chichester 
 retrochoir and Salisbury nave, as at Chartres, there are but four ; when these are 
 set close up to the internal cylinder, the effect is somewhat naked and unsatis- 
 factory. More successful is the pier with eight spaced shafts, as at Ely. 
 In the choir of Westminster the piers have four shafts ; in the nave, which was 
 built later, they have eight.§ Sometimes the number of detached shafts rose to 
 sixteen, as in the vestibule to the Chapter House of St Mary's, York ; |[ where the 
 central cylinder is surrounded by a ring of twelve detached shafts, with four 
 columns at the cardinal points of the pier outside the ring. At LINCOLN 
 (249) is a remarkable pier — all the more remarkable because it is so early — 
 it is part of St Hugh's work of 1192 — at the junction of the choir with the 
 eastern transepts ; it has eight shafts ; of these four are circular, four are fluted 
 hexagons. Piers of surprising height and slenderness were erected by means 
 of these monolithic blocks ; daringly beautiful are those of the north-west chapel 
 of Lincoln nave, and of SALISBURY LADY CHAPEL (173): the latter are nearly 
 30 feet high and carry a vault. Britton If points out that the central cylinders 
 of the Salisbury piers are composed of small square stones laid in mortar. 
 These must have settled considerably as they gradually got their load. If, 
 therefore, the shafts had been built at the same time as the cylinders, not 
 
 * This may be styled the Southern pier, because, though it occurs elsewhere, it is far more 
 common in the South of England. 
 
 t Enlart's Manuel, 328. % Baldwin Brown, 8 to 10. 
 
 S In the western bays of Westminster nave, built later still, the shafts are engaged. 
 
 || Plan in Scott's Lectures, ii. 148. IT Salisbury, 80. 
 
247 
 
GOTHIC PIERS. 
 
 H9 
 
 Lincoln, 
 Eastern Transept. 
 
 being able to settle so much, being monoliths, they would inevitably have been 
 fractured. It is probable, however, that the shafts were not added till all settle- 
 ment had ceased. We can still see, inside the western wall of St Alban's, 
 bases, bands, and capitals prepared by John de Cella for the insertion of shafts ; 
 which, however, were never inserted. The same was the 
 practice in France.* If we can trust Leland, the shafts 
 of the choir of Worcester, built 1202- 1240, were not 
 added till the time of Bishop Giffard, 1268-1301.T 
 
 We have spoken of the core of the central pier as a 
 cylinder ; and this was the most common form. But 
 other forms, e.g. the octagon, were common. At Foun- 
 tains Abbey, in STONE CHURCH (665.7), and in All 
 Saints', Stamford, it is noteworthy that the builders have 
 set their shafts round the special form of pier, which for 
 its strength was adopted in some of the best of four- 
 teenth-century Gothic of France ; e.g. at St Urbain, 
 Troyes.* 
 
 Baxds. — The joints between the monoliths of which 
 the longer shafts were composed were masked and 
 
 strengthened by stone or marble bands {annulets). Usually the bands were solid. 
 In the north transept of Tintern, however, they are pierced ; so that the shaft 
 passed through them ; elsewhere at Tintern they are solid. § The method of 
 attaching the annulet to the central cylinder is shown in the diagram. || Some- 
 times, however, as in the Salisbury pier, bands of 
 copper or bronze" were employed; the method 
 of attachment is shown in the diagram. Bands of 
 stone occur occasionally abroad ; e.g. at Noyon, 
 Laon, Dol, Lausanne, Magdeburg. In England 
 they usually disappear about 1280 ; but sometimes 
 reappear in late work, as at Canterbury, Bath, 
 Wrington, and Yatton. One of the earliest ex- 
 amples is the columns of St Peter's, Northampton, 
 each of which consists of two monoliths, banded 
 together. 
 
 Marble Shafts. — Not only is the employ- 
 ment of the pier with detached shafts far more common in England than abroad, 
 but the further refinement, specially English, was added, that most often the shaft 
 is constructed of marble: turned in the lathe. We have special mention of lathes in 
 the account of the rebuilding of Canterbury choir by William of Sens, who, says 
 
 * Choisy, ii. 264. + For examples of piers with detached shafts see 665.6, 7. 
 
 I For moldings of capital, band, and base see west waltox (432). 
 § T. Blashill's Guide to Tintern, 12. 
 
 || In early work the band is constructional; being a bondstone into which the detached 
 shafts were socketed. But in the thirteenth-century work of the nave of All Saints', Stamford, 
 there are decorative bands, running round the coursed work of the solid pier. Nene Valley. 14. 
 
 II On the bands of gilded brass which occur in Worcester retrochoir, but only where 
 exceptional pressure had squeezed out the beds of the monoliths or flushed their edges, see 
 Mr A. B. Pickney \x\ Journal of R.I.B.A., 27th April 1 901, 300. 
 
 Salisbury. 
 
250 GOTHIC PIERS. 
 
 the monk Gervase, " torneumata ad lapides formandos fecit valde ingeniose." * 
 The marbles most commonly employed are formed of comminuted shells ; e.g. 
 those of Bethersden, Kent, used at Canterbury ; those found near Petworth, 
 Sussex ; and those found in the Isle of Purbeck, near Portland. The last, being 
 of fine quality, and on the sea-board, was shipped in vast quantities wherever 
 there was water-carriage up the rivers ; up the Avon to Salisbury ; up the Ouse 
 to Ely ; up the Witham to Lincoln ; up the Humber and Hull to BEVERLEY 
 (51); up the Humber and Ouse to York; up the Wear to Durham for the 
 galilee ; even across the sea and up the Liffey for St Patrick's Cathedral, Dublin. 
 Where there was a good local marble, that of course was usually substituted for 
 Purbeck ; Langport lias at Wells and Bristol ; Frosterley encrinital marble for the 
 eastern transept of Durham ; Yorkshire marbles for Fountains and Jervaulx. 
 The marble shaft is first found in general use in William of Sens' work at 
 Canterbury, begun 1 175. In the pier-arcade he uses the Purbeck shafts in 
 hesitating, tentative fashion. One can see, as one walks from the eastern 
 transept eastward, that the material is a novel one, and that he is trying all sorts 
 of combinations of shafting to his cylinders and octagons ; in the triforium and 
 clerestory, however, it is employed with the greatest propriety and success. But 
 the curious thing is that for this profusion of dark marbles against white free- 
 stone, which was to be the dominant note of most English Gothic for a century, 
 there was little precedent in the country from which he came. A few piers shafted 
 with monoliths occur in the aisles of Notre Dame, and in two bays of the nave 
 of Laon ; j- but they are too rare to have suggested the use of Purbeck on the 
 vast scale in which it is used at Canterbury. Besides, we were employing 
 marble shafts in the galilee of Durham, c. 1160-1175 ; these were of Purbeck; 
 a remarkable example of the great distance to which Purbeck marble was 
 exported. Marbles also occur in piers and shafts in Jervaulx chapter house, 
 Pudsey's Hall at Auckland and his church at Darlington ;} all c: 1190. Plainly 
 this characteristic decoration of English Gothic is of English origin. 
 
 For detached shafts marbles were in employment chiefly from c. 1160 to c. 
 1300. After this they appear in engaged shafts; e.g. c. 1330 in the beautiful 
 piers of ELY CHOIR (251) and the back of BEVERLEY REREDOS (452), c. 1330. 
 
 POLYCHROMY. — The use of Purbeck marble is one of the few attempts on a 
 large scale at polychromy in mediaeval art.§ In Auvergne polychromy was 
 popular ; e.g. at Le Puy ; brilliantly coloured stones being obtainable from the 
 extinct volcanoes in the neighbourhood. But the nearest parallel to the attempts 
 of our thirteenth-century builders to make their church as it were studies in 
 pen and ink i] is to be found in Northern Italy; where may be seen a whole 
 cathedral, Siena, built in alternate stripes of black and white marble. In 
 Verona red and white marbles are used alternately. From this the Italians 
 
 * Willis' Canterbury. 
 
 + c. 1200, ace. to Viollet-le-Duc. See Plate 66 in Johnson's Early French Architectii>-c. 
 
 % Some of the Darlington piers have eight marble shafts. This use, however, remained 
 characteristic of Southern piers. In the Northern pier-arcades a different use prevailed. 
 
 § There is a certain amount of polychromy also in the Norman work of Worcester ; in the 
 walls of Westminster choir, and throughout its high vault. So also Ham Hill stone is alternated 
 with lias in Somerset. Red ironstone is similarly used in the Midlands. 
 
 || " Albo nigroque " (Metrical Life of St Hugh). 
 
25' 
 
252 GOTHIC PIERS. 
 
 went on, as in the cathedral and its campanile at Florence, to panels of coloured 
 marbles separated by mosaic borders ; about as artistic as " an elaborate Tun- 
 bridge work-box." * In early work, indeed, in our own country, such as at 
 Canterbury, we find a good deal of hesitancy in the use of the new material ; and 
 at Rochester, c. 1 200, some of the shafts have rather the appearance of glorified 
 drain pipes. But almost immediately the new marble shafting produced con- 
 summate design : the arcading of St Hugh's work at Lincoln ; of the porches of 
 John de Cella at St Alban's ; of the galilee at Ely ; of Peterborough west front ; 
 of the Lady Chapel of Salisbury ; of the unparalleled choir of Beverley. And, 
 with the aid of the lathe, marble was used, not only for cylinders and shafts, 
 but for bases, capitals, and abaci. And when Ely presbytery was built, the 
 carvers had become so expert that foliated caps also were wrought of this 
 intractable material. It is interesting to note that at Christ Church, Dublin, 
 the hundreds of Purbeck shafts employed are of a uniform length of 16^ 
 inches each. Being brittle, they were probably sent in short lengths from 
 Purbeck to lessen risk of breakage in sea-transit. The uniformity of dimension 
 of these shafts largely conditioned the builder's design. He was not able to 
 make his doors, windows, piers, &c, any height he chose ; each jamb or pier had 
 to be so many multiples of \6\ inches.! 
 
 Disuse of Marble. — We cannot but ask why a material which produced 
 such splendid results was abandoned, as in the end it was. It may have been 
 that the quarries ceased to be able to furnish a supply of sufficient good marble. 
 But this reason hardly seems adequate. If the Purbeck quarries gave out, 
 there were other quarries. At any rate, other reasons may be adduced. 
 When the shafts were put up, they were highly polished, and, to the 
 mediaeval eye, dazzlingly beautiful. Geraldus Cambrensis, j speaking of St 
 Hugh, says that " Lincolniensem beatae Virginis ecclesiam ex Pariis lapidibus 
 marmoreisque columnis, alternatim et congrue dispositis, et tanquam picturis 
 variis, albo nigroque, naturali tamen colorum varietate distinctis, incompara- 
 biliter erigere curavit eximiam." And the Metrical Life of St Hugh, speaking of 
 the dancing lights of the polished shafts, says — 
 
 " Inde columellae quae sic cinxere columnas 
 Ut videantur ibi quandam celebrare choream ; " 
 
 and again he says that the surface of them 
 
 " Clara repercussis opponit visibus astra."§ 
 This polish, however, does not last. In the course of a century, if not 
 before, the surface of the shafts, inside just as well as outside the church, probably 
 began to peel and flake off in unsightly patches. What had been the glory 
 of the church became a spectacle of mouldering decay. The very hardness of the 
 surface was the cause of the mischief. On a damp day the moisture in the air 
 percolated into the porous freestone ; and on the next dry day was exuded again, 
 without injury to the stone. But on a damp day moisture collected in drops, 
 and streamed down the cold impervious Purbeck marble ; till gradually the 
 surface was disintegrated. For one reason or other marble shafts went out of 
 
 * Willis' Middle Ages, 12. t T. Drew, in Builder, May 5, 1894. 
 
 | Vita 5. Remigii, vii. 97. § « Dazzles the eyes like a mirror." 
 
GOTHIC PIERS. 
 
 fashion 
 column. 
 
 and the way was left clear for the rival form of pier ; the cluste 
 
 The Northern Pier. 
 
 The Northern pier resembles the Western pier in one respect only, viz. 
 that both are in courses of freestone. Instead of shafts, it employs columns ; 
 seldom more than eight. The examples are very numerous. They include 
 piers at Kirkstall, Furness, Cartmel, Roche, Jervaulx, Ripon, Byland, Selby nave, 
 Hartlepool, Darlington, Hexham, all c. 1 150- 1200; and but little later Whitby 
 
 and Rievaulx. South of the Humber, New Shoreham 
 columns (423) ; and in the western 
 province Wenlock Abbey; late in 
 the twelfth century. In the follow- 
 ing century they occur out of their 
 own area at St Saviour's, South- 
 wark ; in St Alban's nave (which 
 was intended, however, by John de 
 Cella to have piers encircled by 
 Purbeck shafts; ; and at Netley and 
 Tintern ; which, being Cistercian, 
 may have been influenced by the 
 Yorkshire abbeys. In the Mid- 
 lands they occur in Southwell* choir 
 and at St James, Deeping. A 
 special feature of this pier is that 
 usually some or all of the columns 
 are pointed; e.g. at ROCHE, BYLAND 
 (661), and Beverley. Anything 
 more majestic and at the same 
 time elegant than the pier of BEVER- 
 LEY TRANSEPT (244) it is im- 
 possible to conceive. It has the 
 grace of the Southern pier with- 
 out its fragility and insincerity of 
 construction ; and one would cer- 
 tainly have expected that, of the 
 three rival piers, this would be 
 the one to prevail. In the end it did prevai 
 its Southern rival. 
 
 has some clustered 
 
 Beverlev St Mary, Nave. 
 
 But at first the victorv was with 
 
 Conquests of the Southern Pier. 
 
 The first to disappear was the Western pier. After 1200 it is seldom seen 
 again; but we may recognise it in the 1280 piers of EXETER (661.9), and in 
 the fourteenth-century choir of WELLS (437). In the Western district the 
 detached marble shafts of the Southern pier find their way into the west front 
 of Wells, the very cradle of the Western School of Gothic ; to the Lady 
 
 * Southwell was in the diocese of York, and so susceptible to Northern influence. 
 
^54 
 
 SCALE OF 
 
 Beverley St Mary, Chancel. 
 
GOTHIC PIERS. 
 
 255 
 
 Chapel of St Patrick's, Dublin — the last reminiscent of the Lady Chapel of 
 Salisbury — to the choir and retrochoir of Worcester ; where the contrast be- 
 tween the Western Gothic of the western bays of the nave and the Southern 
 Gothic of the eastern limb is particularly striking. 
 
 Equally sweeping were its conquests in the North. Southern influence 
 has been recognised in the east transepts of Fountains and Durham.* In the 
 Midlands it appears in Lincoln Minster; the lovely presbytery of ELY (117); 
 and the unparalleled west front of PETERBOROUGH (112). 
 
 Clustered Columns. 
 
 But in the end, as we have said, the Purbeck fashion of the Southern 
 Gothic passed away, c. 1300; Western Gothic had been long defunct; and 
 the rest of the story of the English pier is 
 that of the clustered column of the North. + 
 In Beverley transept, in Merton College, Ox- 
 ford, and at Exeter the number of columns 
 had risen to sixteen. For the rest of the 
 story, with the exception of such ultra- 
 logical piers as those of ST GEORGE'S, 
 WINDSOR, the tendency was all in the direc- 
 tion of simplification. The sixteen columns of 
 Exeter were very commonly reduced to four. 
 A beautiful example is seen in the fourteenth- 
 century choir of ELY (251), which, though it 
 has four columns, is given eight bases ; and 
 moreover is worked in marble, that it may be 
 in harmony with the piers of the presbytery 
 to the east. The pier of four columns had, of 
 course, been long with us ; as far back as St 
 Michael's, York, and ST PETER'S, NORTH- 
 AMPTON (663.1). It is a pier of the simplest 
 possible plan ; merely four columns put 
 together. There was, however, a decided 
 difficulty as to how to treat the space between the columns. At SKELTON 
 (663.3) c - II 45. the hollow is deeply undercut to get a black shadow; at 
 HIGHAM FERRERS (663.4; there is a flabby hollow ; at NORTHBOROUGH 
 (663.5 j a wedge; at MARKET DEEPING (663.2) the columns interpenetrate; 
 all these are thirteenth-century work. In the fourteenth century, e.g. at 
 BOTTISHAM (663.6), COGGESHALL (663.8), HEDON (544), and LONG MELFORD 
 (663.9), the hollow is filled up with a small, illogical shaft. From the fourteenth 
 century onward, by far the most common treatment is to leave between the 
 columns a shallow hollow, called the " casement." But where the piers were 
 
 * By Mr C. C. Hodges in Arch. Aeliana, xvi. 382. 
 
 t Examples of clustered columns in the North of England are illustrated from Kirkstall, 
 Roche, Byland, Bridlington, St Mary's York, and Guisborough, on 661. 
 
 St George's, Windsor (Plan of Pier). 
 
256 
 
 GOTHIC PIERS. 
 
 very lofty and massive, a simple casement was not enough ; and a double 
 ogee ("bracket") might be added, as at LAVENHAM (663.7). At Cromer* a 
 wave-molding is employed ; and there are many other treatments ; e.g. LONG 
 MELFORD (665.12).! 
 
 Ornamentation of Shafts. 
 We may now turn to the decorative treatment of the column and shaft. 
 The first thing to be noted is that Romanesque shafts are often carved ; 
 
 Gothic shafts but seldom: the Gothic 
 builders preferred the keel-molding 
 or the fillet. The Durham School 
 of Norman, e.g. at Durham, Selby, 
 Dunfermline, Waltham, often in- 
 cised their big cylinders with deep 
 grooves ; \ sometimes the patterns 
 ran spirally, as in Canterbury crypt, 
 Lincoln west doorway, Pittington.§ 
 Enriched shafts are seen at ST 
 peter's, Northampton, and 
 IFFLEY ; and in the doorway of 
 SHOBDON (415.3). 
 
 Northampton St Petei 
 
 Plan of Shafts. 
 
 In Norman work the shafts and columns are semicircular ; nor was this 
 form ever wholly abandoned; e.g. it is seen c. 1280 at EXETER (241); in 
 fifteenth-century piers it is almost as common as in Norman. But from c. 
 1 1 50 various other forms appear. One of the earliest is the pointed column or 
 shaft ; it is especially characteristic of the twelfth-century Gothic of the North. 
 It appears, e.g. at ROCHE (661.2) and BYLAND (661.3); m the latter it alter- 
 nates with the semicircular form. The next step was to emphasise the sharp 
 edge (" arris ") of the pointed column by hollowing the column a little on either 
 side ; this produces what is called the " pear-shaped molding." Frequently 
 the sharp edge was cut off; producing a narrow "fillet." At GUISBOROUGH 
 (661.6) the smaller columns have the "pear-shaped molding"; the larger have 
 the "fillet." || Towards the end of the thirteenth century the fillets often become 
 broad ; at Tideswell in the fourteenth century to an excessive extent ; in the 
 fifteenth century they are less common. 
 
 * Glossary, ii., Plate 153. 
 
 t For the final development of the compound pier in later Gothic, turn back to page 242. 
 
 \ In the Cluniac Priory of Castle Acre, Norfolk, also the cylinders were grooved with 
 spirals, lozenges, and frets (Mr W. H. St John Hope). 
 
 § Cf. the remarkable columns of the ruined chancel of Orford, Suffolk. With these we 
 may compare the spiral friezes of the column of Trajan at Rome. 
 
 || Filleted shafts are shown in SOUTHWELL choir (448.2) and SHREWSBURY ABBEY (440.4). 
 
Chapter XVI. 
 
 THE FORMS OF THF ARCH. 
 
 On Trabeated and Arcuated Construction — Semicircular, Stilted, Segmental, Horseshoe 
 Elliptical, Pointed, Four-centred, Foiled, and Ogee Arches. 
 
 TRABEATED AND Arcuated Styles. — In Greek architecture the voids between 
 column and column, or between column and wall, were spanned by lintels ; 
 i.e. by horizontal beams of stone (Latin, trades). The Greek, therefore, is 
 a Trabeated style. From a constructional point of view it does not rank high ; 
 first, because long blocks were very costly ; being but rarely found ; and when 
 found, difficult to extract ; and if found and worked, costly to transport ; and 
 costly also to raise to their position ; secondly, because Trabeated construction, 
 except on quite a small scale, is to use stone for a purpose for which the nature 
 of the material unfits it : to use it as if it were a tough material like an oak 
 beam or an iron girder, whereas it is a granular material, and therefore 
 incapable of supporting any serious strain. From 
 stone post and lintel construction the Romans and 
 Byzantines gradually freed themselves, and the 
 mediaeval architects freed themselves altogether : 
 using stone only where its strength is greatest, that 
 is, in resisting compression. With a moderate amount 
 of pressure you may squeeze wood or iron out of 
 shape ; but not so with stone. When stone blocks 
 
 are used in an arch, they are in a state of compression ; i.e. stone is used just 
 in the way where its strength is at the greatest. Such a method of con- 
 struction is called Arcuated. It is as scientific and practical as the Trabeated 
 construction of the Greeks is the reverse. And as the arches are constructed 
 in small blocks, which are to be obtained in abundance, are easy to quarry 
 and work, and cheap to transport and to put in position, it is as cheap as 
 it is scientific. Besides, in many districts, and in some whole countries, 
 large blocks for lintels are not to be had at all ; whereas small blocks, or clay 
 hardened into bricks, are to be found everywhere ; so that the application of 
 Arcuated construction is universal and world-wide; whereas Trabeated construc- 
 tion is confined to a few favoured restricted districts. 
 
 Mem hers of THE ARCH. — Each wedge-shaped block of which an arch 
 consists is called a claveau or voussoir (4 to 13 above). The central voussoir, if 
 there is one, is called the keystone or key (French, elef) (9 above). Where several 
 arches intersect in a vault, the keystone may become a boss or a pendant (302). 
 A pointed arch sometimes has a keystone ; but more often a straight joint, as 
 
 R 
 
 Skewback. 
 
> 5 S 
 
 I. Semicircle. 2. Stilted Semicircle. 
 
 5. Horseshoe. 6. Stilted Horseshoe. 
 
 9. Pointed Obtuse. 10. Pointed Segmental. 
 
 13. Four-Centred. 14. Quasi-Four-Centred. 
 
 17. Trefoiled. 18. Trefoiled. 
 
 3- 
 
 Segmental. 
 
 4- 
 
 Stilted Segmental 
 
 7- 
 
 Pointed Equilateral. 
 
 8. 
 
 Pointed Lancet. 
 
 1. 
 
 Three-Centred. 
 
 12. 
 
 Four- Centred. 
 
 5- 
 
 Ogee. 
 
 16. 
 
 Ogee. 
 
 9- 
 
 Pointed Arch Trifoliated. 
 
 20. 
 
 Shouldered Arch. 
 
FORMS OF ARCH. 
 
 259 
 
 below. The lowest voussoirs are called springers (French, sommz'er; Old English, 
 " summer " or "summer"), 1, 2, 3, 15, 16, 17 in 257. The springers usually 
 have one or both joints horizontal ; and are said to be laid in tas de charge 502 . 
 The upper surface of the top springer, 3 or 15, against which the first voussoir 
 of the real arch, 4 or 14 (in which both joints radiate) starts, is said to be skew- 
 backed. The skewback is shown in 2\~ at X, Y. 
 
 The under surface of the arch, BD, is called its intrados or soffit. Its upper 
 surface, AE, is its extrados. Mediaeval arches are extradossed ; but for decorative- 
 reasons some Roman and Renaissance arches were not ; e.g. the upper surfaces 
 of a Renaissance arch sometimes form a horizontal line: instead of forming a 
 curve, and that curve the same curve as that of the intrados. If an arch is 
 enclosed, or is imagined as being enclosed in a square, e.g. in the doorway 
 of KETTERING (95), or WINCHESTER (261 ), then the spaces between the arch and 
 that square are its spandrels. 
 
 Forms of Arch. — In Rome and Western Europe (but not necessarily so 
 in the East) the oldest and normal type of arch is the Semicircular 'arc en 
 plein cintre), 258.1. In this the centre is in the middle of the diameter. In 25S.2, 
 the arch does not start at the ends of the diameter, but at some distance 
 above it ; a vertical piece of masonry or stilt being interposed. (This differs 
 from BX, DX in 257, in that the intrados is straight, not 
 curved.) The centre is therefore at a point above the dia- 
 meter. This is called a Stilted arch (air sitrhansse). Or, 
 instead of taking a whole semicircle, we may select a 
 curve that is less than a semicircle, and we shall have a 
 much flatter arch, a segment of a circle (258. 3 J, with its 
 centre beloic the diameter. This is a Segmental arch arc 
 surbaissc). Or we may select a curve that is greater than 
 a semicircle, and we shall have an arch with its centre 
 
 above the diameter. This is the Horseshoe arch, 258.5 (arc outrepasse or arc enfer 
 a cheval). 
 
 All the above four arches were struck each from one centre. The 
 second class is struck from two. This arch is the Pointed arc aigu or arc 
 briscj. There are three chief varieties. The first is the Equilateral arc en 
 tiers-point . In this the two centres coincide with the ends of the diameter, 258.7. 
 The second, more acutely pointed, is the Lancet* (lancette), 258.8. In this the 
 centres are on the line of the diameter, but outside it. The third is the obtuse or 
 Diop arch (arc en tiers-point surbaissc ', 258.9. In this the centres are still 
 on the line of the diameter, but inside. Sometimes the arch is nearly straight- 
 sided, as in Boxgrove clerestory and in Hereford north transept. 
 
 The third class consists of arches struck from three centres, as shown in 
 258. 1 1. This is the three-centred or " basket-handle " arch a/c en anse depanier . 
 
 The fourth class consists of arches which are struck from four centres. 
 The first variety is the Four-centred ox Tudor arch from Bath (258.12); and from 
 
 Radiating Joints 
 
 * It is often asserted that the Lancet is characteristic of the first half, the Equilateral arch 
 of the second half of the thirteenth century. But Sharpe points out that in the Nene valley 
 (page 12) both forms occur indiscriminately throughout the whole of that century. In the 
 clerestory of Bath Abbey the equilateral arch occurs at the beginning of the sixteenth century. 
 
26o 
 
 SEMICIRCULAR ARCH. 
 
 St George's, Windsor (258.13). The different way in which the curves are struck 
 in these two examples should be noticed. In some examples, e.g. in the 
 windows of the Divinity School, Oxford, the long curves are replaced by straight 
 lines (258.14). In Jacobean Gothic both curves may be replaced by straight 
 lines with a short curve at their junction. 
 
 Another variety of arch struck from three or four centres is the Ogee arch 
 {arc en accolade), 258.15, 16. In this one or two of the centres are below, 
 but the other two are above the arch. So that the two upper curves of the 
 arch are concave, the two lower convex. 
 
 Foiled arches. These are arches with three or more lobes or leaves 
 (Latin, folia). The simplest are the Round-headed Trefoil (258.17) ; the Pointed 
 Trefoil (258.18); the Square-headed Trefoil (258.20); which also goes by the 
 
 name of the Shouldered arch. 
 
 tss^M 
 
 £ 
 
 Spiff® if 
 
 A trifoliated arch is a trefoiled arch enclosed in a 
 pointed arch, as in 258.19. A trefoiled 
 arch is not enclosed in any other arch. 
 Besides the trefoil there is the cinque- 
 foil arch with five lobes or foils, and 
 the multifoiled arch with several. 
 
 Again, the voussoirs may be all 
 wedge-shaped ; but the extrados and 
 intrados composed not of curved but 
 straight lines ; this is a Flat arch 
 {plate bande) ; it is in common modern 
 use, instead of a lintel, over a window, 
 door, or fireplace ; it occurs over 
 TEWKESBURY porch. Sometimes, 
 to strengthen a flat, or even a curved 
 arch, the voussoirs are notched or 
 joggled ; good instances occur at 
 Kirkstall in the cellarer's lodgings ; 
 Tewkesbury Porch. at Fountains over the big fireplace 
 
 of the calefactory, 16 feet wide, of 
 too great a span for a lintel ; another survives over a fireplace in CONISBOROUGH 
 
 KEEP (261 ).f 
 
 Semicircular Arch. — This arch is specially characteristic of Romanesque 
 architecture. To us, perhaps, the pointed arch, from its association with 
 Gothic architecture, seems the more beautiful. But of old the semicircular 
 was the favourite form ; as was natural ; " for its type is ever before us in 
 that of the apparent vault of heaven, and the horizon of the earth." J Cer- 
 tainly, when the pointed arch first came into use, it was unpleasing to eyes 
 that had ever been accustomed to the older form ; and though it was 
 employed where it facilitated building construction ; yet, wherever an arch 
 was to be built for decorative purposes, it still usually remained semi- 
 
 * This is not an arch at all, but a lintel ; it is very common in Carnarvon Castle. 
 + A coursed lintel occurs at the end of the third century in the Golden Gate of Diocletian's 
 palace at Spalato, illustrated in Anderson and Spiers, 259. 
 % Sez'en Lamps, 189. 
 
STILTED ARCH. 
 
 261 
 
 circular. A curious example of belated fondness for the older form is 
 seen in the aisle windows of Glastonbury nave (a.Ii. 1184) which exter- 
 nally are pointed, but internally semicircular. The pointed arch is the arch 
 par excellence of Gothic architecture ; nevertheless, in early Gothic archi- 
 tecture, sometimes an arch was needed that was lower than a semicircle. 
 Sometimes a very obtusely pointed arch, i.e. a SEGMENTAL roiNTED ARCH, 
 was employed ; e.g. 258.10; or else a low segmental arch (258.3); but some- 
 times a semicircular arch. Gothic semicircular arches occur in the thirteenth 
 century in the doorway of the transept of IJEVERLEY (574) ; in the pier-arcades 
 of St Gilles, Caen ; and Woodford, North Hants: in the fourteenth century in 
 
 ^^B^IBBi 
 
 Winchester, X. Transept. 
 
 Conisborough. 
 
 the pier-arcade of Water Newton, Hants; and the doorway of Badgeworth, 
 Gloucester ; in the latter it is covered with ball-flower.* 
 
 STILTED ARCH. — By stilting, a narrow semicircular arch is enabled to rise 
 to the same level as a broad arch. Where, therefore, broad and narrow arches 
 come into juxtaposition, the eye is saved the pain of seeing the narrow arches 
 rise to a lower level than the broader ones. There are three common cases 
 of this. (1.) The arches between an apse and its ambulator}', e.g. at St 
 Bartholomew's, Smithfield, are about half the span of those on either side of 
 
 * In the north-west of North Hants and in Rutland they are in some places as common as 
 the pointed arch in the thirteenth and occasionally in the fourteenth century ; e.g. in the pier- 
 arcades of Bainton, Castor, Werrington, Garwell, Manton, Great Casterton, Preston, Seaton, 
 Edith Weston, Clipsham, Barrowden ; over two lancet lights, espei tally in belfries ; e.g. Barnack, 
 Etton, Tansor, Wadenhoe ; and over doorways ; e.g. Whitwell, Barrowden. — R. P. I!. 
 
262 
 
 POINTED ARCH. 
 
 the choir ; the former, therefore, are stilted ; if. St Germer and La Charite (2.) 
 The transepts are sometimes narrower than the nave or choir ; and its north and 
 south arches may be stilted, as at Malmesbury and SHERBORNE.* (3.) The pier- 
 arches are narrower than the diagonal arches of the aisle vault ; and to facilitate 
 vaulting (322) the former may be stilted, as at WINCHESTER (261). At MEL- 
 BOURNE (203) the pier-arches are stilted, even though the aisles are not vaulted. 
 Segmental Arches. — These occur occasionally even in Norman work, 
 e.g. at Southwell there was little room under the window of the south transept, 
 and the doorway, therefore, has a segmental head. 
 
 Horseshoe Arches. — Well-marked examples occur in Norman work, 
 but are rare ; e.g. Holywell Church, Oxford ; under the west towers of Southwell 
 Cathedral ; in the chancel-arch of Patrixbourne, Kent ; and a little later in the 
 
 west nave of Kilwinning Abbey. They are not 
 uncommon in Norman ribbed vaults ; e.g. in 
 PETERBOROUGH AISLES (318). They occur in 
 the aisled basilica of Dana,t near the Euphrates, 
 which has the inscription A.D. 540. In Eastern 
 work the horseshoe arch is frequently not round- 
 headed, but acutely pointed. This facilitates con- 
 struction, as the upper and more difficult portion 
 of the arch or dome can then be constructed by 
 corbelling and without centering ; as in many 
 Indian domes. 
 
 Elliptical Arches. — It may be doubted 
 whether any true elliptical arches occur ; at any 
 rate otherwise than accidentally. The anse de panier is sometimes called an 
 elliptical arch, but is really three-centred. It is not likely that the mediaeval 
 builders knew how to draw a true ellipse. Sometimes, however, a vault has 
 given way somewhat ; its crown flattening, and its sides receding ; and thus 
 something like an ellipse is produced. The following, among others, have been 
 cited as examples of elliptical arches ; the transverse arches of Vezelay nave 
 and Laach ; \ the transepts of the Abbaye-aux-Dames and Montivilliers, and 
 Gueron apse ; § and the diagonals of Devizes St Mary and Dunstable. || 
 
 POINTED Arch. — Though, as we shall see (314), in Durham transepts 
 our Romanesque architects, without the pointed arch, had solved triumphantly 
 in the first years of the twelfth century the greatest problem of the Middle 
 Ages, viz. how to vault throughout with stone a clerestoried church Basilican 
 in plan, yet the employment of the pointed arch greatly facilitated building 
 construction. Next to the use of diagonal ribs and flying buttresses, it was 
 the greatest improvement introduced into mediaeval architecture. 
 
 Origin of the Pointed Arch — (1.) In discussing the origin of the 
 pointed arch and of that beautiful style with which it is associated, our authorities 
 almost break into poetry. Stukeley in 1755 says "this pointed architecture and 
 its slender pillars are taken from the groves sacred to religion ; those verdant 
 
 Sherborne. 
 
 Illustrated in R.I.B.A., 1S77, 144. 
 Petit's Church Architecture, i. 101. 
 Bilson's Beginnings, 308. 
 
 + Illustrated in Scott's Essay, Plate 13. 
 § R. Robert, i. 169. 
 
POINTED ARCH. 263 
 
 cathedrals of antiquity." "We were overshadowed," says Washington Irving,* 
 "by lofty trees, with straight smooth trunks like stately columns; and as the 
 glancing rays of the sun shone through the transparent leaves, tinted with 
 the many-coloured hues of autumn, I was reminded of the effect of sunshine 
 among the stained windows and clustering columns of a Gothic cathedral. 
 Indeed there is a grandeur and solemnity in some of our spacious forests of 
 the west, that awakens in me the same feeling that I have experienced in 
 those vast and venerable piles ; and the sound of the breeze sweeping through 
 them supplies the deep breathings of the organ." '2.) For a long time the 
 theory of Dr Milner, 1798, held the ground; viz. that "the pointed arch 
 arose from the intersection of semicircular arches." So it does at Southwell ; 
 where the arcading of the north-west tower has the latter; that of the south- 
 west tower the former. But, unfortunately for this theory, just where in- 
 tersecting semicircular arches were most employed, i.e. in Normandy and 
 England, there the pointed arch is latest in making its appearance ; and where 
 they are little employed, e.g. in Central and Southern France, there it appears 
 very early. (3.) Or Western Europe may have borrowed it from the East. 
 It has been held that it may have travelled by an overland trade-route across 
 Russia to Wisby, the great Hanseatic depot in the island of Gotland. But 
 the Wisby churches, in which the pointed arch is used so largely, are now 
 recognised not to be earlier than the middle of the twelfth century. (4.) On the 
 other hand, it is possible that the pointed arch was borrowed from the East — 
 perhaps through acquaintance with Saracenic work in Egypt and Sicily — by 
 Southern France: for it is there that it seems to make its first appearance; and 
 there not till the second half of the eleventh century. In Egypt the great mosque 
 of Tulun, built in 879, has pointed arcades ; so also the Xilometer in the island 
 of Roda, + which is probably of 861. At Diarbekir in Armenia, two pointed 
 arches crown a Roman colonnade, and may be contemporary with it; also the 
 great gateway of the palace of Ctesiphon (fifth century) is pointed. + "The 
 pointed arch is certainly found in the great aqueduct near Constantinople, and 
 in one of the city cisterns ; both probably are of the age of Justinian. They 
 are correctly illustrated in Miss Pardoe's Bosphorus." § The Roman bridge of 
 Severus, illustrated in Hogarth's Levant, has pointed arches. 
 
 Of the antiquity then of the [jointed arch in the East there can be no 
 question : in many districts it is as much the normal form as is the semicircular 
 in the Romanesque of Europe. But it does not follow that the latter borrowed 
 it. Like the axe, it has probably been invented again and again as necessity 
 arose. In countries where there was no timber or no tools to work it, the 
 natives had to build shelters in stone, usually without mortar. Frequently the 
 only way known of roofing these was to pile flat stones on one another, i.e. with 
 horizontal beds, not with radiating joints, each course projecting a little further 
 inward as the wall went up. Plainly these walls would topple in if a semi- 
 circular roof had been attempted ; but they could be got to stand if the roof was 
 built in the form of a pointed arch, at any rate if the arch was very acutely 
 pointed. If a roof thus formed, i.e. by corbelling in, was constructed on a rect- 
 
 * Tour on the Prairies, 47. t Mr Spiers in Fergusson, ii. 45. 
 
 \ Enlart's Manuel, 293, 3. § Lethaby and Swainson's Santa Sophia, 220. 
 
264 
 
 POINTED ARCH. 
 
 angular base, it would produce a pointed tunnel. If on a circular or oval base 
 a pointed dome. People far apart in geography and in time, such as the Pre- 
 Conquest Irish, and those who built the Treasury at Mycenae and the Jaina 
 domes in India, adopted perforce the pointed form, but as a condition of corbel- 
 construction ; no one borrowing from the other, but each driven to invent by 
 necessity, the mother of invention. So again, several churches in Ireland, e.g. 
 St Columba's, Kells, and Killaloe* the latter probably c. 1007, have above the 
 lower semicircular tunnel vault a pointed tunnel which also forms the external 
 stone roof. This remarkable roof-system is not due to borrowing ; but is the 
 outcome of roof builders in a rainy climate without tools to work timber. And 
 so it may have been with those who first used the pointed arch in the tunnel 
 
 vaults and domes of 
 Southern and Central 
 France. 
 
 Value of the 
 Pointed Arch. — More 
 important than the ques- 
 tion of the origin is the 
 question of the value of 
 the pointed arch. It is 
 valuable in many ways. 
 First, it is obvious that 
 a flattish arch, e.g. the 
 segmental arch in 258.3, 
 cannot carry such a heavy load as an acute arch, e.g. the lancet arch 
 in 258.8. So with other arches. Any form of pointed arch with the 
 same span as a semicircular arch, but rising to a higher level, must be the 
 stronger of the two. Secondly, a pointed arch has a more vertical, a less 
 lateral, thrust than a semicircular one. This is pointed out by Samuel Ware 
 in a paper in the Archceologia in 18 14. Indeed, an arch may be so acutely 
 pointed, that it has practically no thrust ; the inward and outward pressures 
 being exactly balanced.! This important property of the pointed arch was 
 of enormous value in Romanesque districts such as Provence, Auvergne, and 
 Burgundy, where it was customary to roof the naves with BARREL VAULTS 
 (283). Indeed, when, in Provence and in Burgundy, the barrel vault rested on 
 a wall pierced with clerestory windows, it was impossible to provide direct 
 abutment against the continuous thrusts of these high tunnels, which necessarily 
 therefore had to be pointed, e.g. at St Trophime, Aries, and Paray-le-Monial.t 
 Again there were other Romanesque districts, e.g. in and round Perigueux, where 
 the churches were roofed with STONE DOMES (282). Here, as also at St Croix de 
 Montmajour,§ to point the dome reduces the lateral thrust. In both these cases, 
 as in the Irish churches, the introduction of the pointed arch was plainly due to 
 an exigency of building construction. For the same reason, the diagonals as 
 
 * Illustrated in Fergusson, ii. 44S. t See Statham's Architecture, 91. 
 
 t See Classification of Romanesque. 
 
 i, To this the date a.d. 1019 used to be assigned, but it has been proved by M. Brutails, 
 196, 4, to be later. 
 
POINTED ARCH. 
 
 265 
 
 well as the transverse arches of vaults were sometimes pointed ; e.g. at Wells. 
 Thirdly, there is another reason which may have brought the pointed arch into 
 use early in districts where domes were constructed on pendentives. It is that 
 if the arches on which the dome ultimately rests are pointed, the pendentives 
 between them will both be broader and more vertical, and consequently more 
 stable ; as may be seen in the diagrams below. Fourthly, over the barrel- 
 vaulted naves of Southern France there seem originally to have been n<> 
 outer roofs of wood. Masonry was added on the back of the vault till a 
 straight slope on either side was obtained. Now if the diagrams be com- 
 pared, it will be seen that while on the apex of a pointed barrel vault little 
 masonry will be required, over the semicircular one a heavy load must be placed, 
 and that just on the weakest part of it, its crown.* From the above considera- 
 tions, it seems probable that it was in Southern and Central France that the 
 pointed arch was first employed in Western Europe. Fifth, but the most 
 valuable property of the pointed arch remains to be mentioned. It is that 
 while the height of the semicircular arch was rigidly regulated by its span — the 
 height must be just half the span, neither more nor less — the height of 
 the pointed arch was independent of the span. It is, we may say, an elastic 
 arch. Wherever, therefore, we may wish to retain the span unaltered, but to 
 alter the height, we can effect this by employ- 
 ing a pointed arch. And so all the old-fashioned 
 shifts and dodges of stilted and horseshoe 
 arches were superseded at once ; the entrances 
 to narrow transepts, as at the Abbaye-aux- 
 Dames, St Bartholomew's, f Smithfield, Oxford 
 Cathedral, Bolton Prior}-, were given pointed 
 arches ; while those of nave and civ iir remained 
 
 semicircular. So also the narrow arches of the apses ; as at La Charite, St 
 Germer, Westminster, Tewkesbury. Still more valuable was the pointed arch 
 in vaulting ; e.g. in vaulting an aisle the four narrow arches, i.e. the two trans- 
 verse arches, the pier-arch, and the wall-arch were pointed ; while in most cases, 
 for a long time, the two wide arches, the diagonals, remained semicircular. So 
 again if three arches were required, the two side ones to be narrow and solid, 
 the central one broad and pierced for a door, all three could be constructed of 
 the same height by pointing the narrow lateral arches, but leaving the broad 
 central arch semicircular ; as in Ketton faqade and Selby north porch. 
 
 That the pointed arch came into English architecture for constructional 
 and not for aesthetic reasons is plain from the fact that even when it had been 
 adopted wherever it facilitated construction, in other cases, especially if it 
 was merely a form of decoration, the semicircular arch was long used ; e.g. in 
 doorways, which often had to be kept low not to interfere with a window above 
 the semicircular arch long retained its ground. At New Shoreham, in spite of 
 the advanced type of its vaults, piers, capitals, and moldings, the windows are 
 round-headed, and there is semicircular arcading on the aisle walls. In Ripon 
 
 Barrel Vaults. 
 
 * See Petit's Church Architecture, i. 114. 
 
 + These are said to have been reconstructed {Glossary, 39, note). 
 
266 POINTED ARCH. 
 
 choir, though internally the pointed arch forced itself into use for vaulting 
 reasons, there is hardly a sign of it outside ; except that narrow blank arches 
 are pointed on either side of the clerestory windows. So abroad ; no one from 
 the exterior of Souillac would suspect that all the great arches and domes 
 within were pointed. Even in the Durham galilee, all the arches remain semi- 
 circular. 
 
 In the end, however, for the sake of harmony, pointed arches were used all 
 over the church, whether constructionally necessary or not. But a good deal of 
 difference of practice occurred as to what variety of pointed arch should be 
 preferred. In the He de France the preference was rather for the equilateral 
 arch ; in the thirteenth-century work of England perhaps rather for the lancet. 
 As regards this, construction is at war with aesthetics. It is obviously better to 
 have obtuse arches, as in LINCOLN PRESBYTERY (56), in order to have fewer of 
 the obstructive piers between nave and aisles ; but there can be no doubt that a 
 far more graceful elevation is obtained by the employment of a more acutely 
 pointed arch, as at WESTMINSTER (55). Sir Gilbert Scott* remarks, that on 
 comparing a series of pointed arches of different curve, the two which pleased 
 him best, viz. the pier-arches of WESTMINSTER (258.8) and Wells, turned 
 out on examination, the former to have exactly three, the latter two equilateral 
 triangles in the height of each : so that beauty of curve here seems to rest on a 
 geometrical ratio. 
 
 Early Pointed Arches. — Perhaps the two earliest examples we possess 
 of the pointed arch are at Gloucester and Rochester. At Gloucester, com- 
 menced 1089, the front compartment of the south-east apse is a pronounced 
 oblong ; and the wall-arches of its groined vault become pointed.! In 
 Rochester nave, 1115-1130, the passage in the wall above the pier-arcade 
 has a pointed head. But it is not till the middle of the twelfth century 
 that the pointed arch enters into pier-arcades, to facilitate vaulting. The 
 earliest example has been assumed to be Alalmesbury nave, thought to be 
 built by Roger of Salisbury, who died in n 39; but it is more likely 
 work done some twenty years later. The pier-arches of Fountains nave 
 are pointed : to the eastern bays of which Mr Reeve in his monograph 
 assigns the date 1147-1150. In the opinion, however, of Mr Hope, J the nave 
 was built between 1135 and 1147. Kirkstall was moved to its present site in 
 1 1 52 ; and the pointed pier-arches of its nave may therefore be c. 1160. The 
 pier-arches of Furness also are pointed ; they were probably built when Furness 
 became Cistercian in 1 147. Fountains and Kirkstall also were Cistercian. And 
 as the Cistercian mother abbeys in Burgundy exported Burgundian construction 
 all over Europe,§ we shall probably be right in concluding that the introduction 
 of the pointed arch into Yorkshire pier-arcades is due to designs or designers 
 sent over from Citeaux or Clairvaux. But though we were so slow to point 
 our pier-arches, we had pointed the transverse arches of the high vault of 
 Durham nave as early as 1128-1133. 
 
 Four-centred Arches. — As the name implies, these arches are parts of 
 four different circles. The position of the centres varies greatly, and with them 
 
 * Gleanings, 26. t Illustrated in Bilsoir's Beginnings, 294. 
 
 t Paper in Yorkshire Archceol. Journal, vol. xv. § See Enlart's Gothic in Italy. 
 
FOUR-CENTRED ARCH. 267 
 
 the beaut\- of the arch. Perhaps the most usual position is for the upper 
 and lower centres of each side of the arch to be in the same vertical line. 
 It may happen, as in Winchester nave, that the lower curves on either side are 
 parts of the same circle, and therefore are struck from one and the same centre ; 
 in which case the arch would more properly be called a four-curved, and three- 
 centred, arch. As has been mentioned before, the long curves may be, if not 
 straight lines, at any rate very nearly so. The four-centred arch has been 
 considered peculiar to England ; but it is common enough in Flanders at the 
 same time as in England ; a well-marked example occurs in the entrance to 
 the sixteenth-century Hotel de Ville at Saumur. 
 
 As to the value of this arch, it may be regarded as the complement of the 
 pointed arch. The latter answered admirably if an arch taller than a semicircle 
 was wanted, without a change of span. But sometimes an arch lower than a 
 semicircle was wanted, without a change of span. In this case three expedients 
 presented themselves. A segmental arch (258.3) might be employed; as over 
 Southwell doorway. Or a pointed segmental arch (258.10) might be used. Hut 
 both of these give an ugly junction where they spring from the wall. The 
 remedy was to round off the junction. If the springing of a segmental arch be 
 rounded, a three- centred arch is produced ; if the springing of a pointed segmental 
 arch be rounded, we have a four-centred arch. The former was the form of arch 
 favoured in late French Gothic ; but it occurs occasionally with us, e.g. in Bishop 
 Alcock's Chapel at Ely. The latter arch came more and more into favour with 
 us from the fourteenth century. At first, as in Winchester nave, it is often 
 considerably pointed ; in later work it tends to become a very depressed arch, 
 with its longer curves nearly, if not quite straight. At no period, however, did 
 it wholly oust the pointed arch : least of all in pier-arches. Nevertheless pier- 
 arches also, in late work, were sometimes four-centred ; e.g., in Bath Abbey ; 
 St Stephen's, Norwich; KILKHAMPTON (568); sometimes they even occur in 
 oak, as at Wingham. 
 
 There is an abnormally early example of four-centred pier-arches at 
 Stanwick* North Hants, where " the hood-molds are so decidedly Early English 
 that the arches cannot be later" ( G. A. Poolei.t Others of the thirteenth 
 century are said to occur in the Lady Chapel of Oxford Cathedral, and in the 
 doorway to the city schools at Bristol; also in the crypt of the Glastonbury 
 Lady Chapel. 
 
 As to the origin of the four-centred arch, it is usual to say that it was 
 arrived at by abstract mathematical considerations.} But as a matter of fact, 
 the four-centred arches had long been presenting themselves, in more or less 
 accidental fashion ; e.g. when a curved brace was inserted beneath a tie-beam 
 or collar, as at Knapton, Norwich St Andrew's, Leicester St Margaret, a four- 
 centred form was originated. Again.it was not uncommon to " humour " the 
 curves of the diagonal ribs of a vault at their foot, so that they might start 
 
 * Illustrated in Nene Valley, Plate 36. 
 
 t Inspection suggests that the hood-molds at Stanwick have been reset ; and thai the an lies 
 therefore are not thirteenth-century work. 
 
 X See Choisy's History, ii. 2S0 ; Viollet-Ie-Duc, Architecture, ix. 533 ; Willis' Vaulting, 25 
 scq. ; Garbett, 173. 
 
2 68 OGEE ARCH. 
 
 at the same angle as the shorter transverse ribs, and not present the awkward 
 collocation of curves seen in the aisles of Norwich and Peterborough and in 
 the eastern crypt at CANTERBURY (334). But this straightening of the spring- 
 ing of a diagonal arch converts it at once from a pointed into a four-centred 
 arch. Again, the rounding of the springing of a pointed segmental arch converts 
 it into a depressed four centred arch ; and this practice was of frequent occur- 
 rence ; e.g. in the great gateway of Wingfield, Suffolk, c. 1370, the outer archway 
 is pointed segmental ; but the inner one, which is of the same date and character, 
 is four-centred. 
 
 Whatever its origin, it greatly facilitated vaulting. Indeed, in fan vaults it 
 disposes of all the difficulties of adjusting the curvatures of the ribs by allowing 
 them all to be constructed of the same curvature* And having found its way 
 into the vaults, in the end it permeated the whole building. It is very useful in 
 doorways ; for being so much lower than a pointed arch, it was not necessary 
 to raise the windows, and consequently the whole wall of the church, to make 
 room for the doorways beneath them. So also a window with a four-centred 
 instead of a pointed head effected an economy of one or two feet of masonry all 
 round the walls, both of the aisles and clerestory ; and great reduction of expense 
 in constructing a large church. In some cases, even the pier-arches were con- 
 structed four-centred, for similar motives of economy ; e.g. at Bath. It is true 
 that being so much flatter than a pointed arch, it was proportionately less 
 adapted to carry the heavy load of a clerestory wall. But the fifteenth-century 
 mason had become so scientific that he may well have felt that the weakness of 
 the form of the arch was largely compensated for by the perfection of his 
 masonry. Moreover the clerestory walls by this time had come to be nearly 
 all glass, and the weight that fell on the pier-arches was reduced thereby. In 
 any case, the great principle of harmony, which is the dominant note of late 
 Gothic design, was offended by the juxtaposition of arches of different curves, 
 and tended to eliminate the pointed form. 
 
 FOILED ARCH. — The remaining arches, viz. the foiled and the ogee, are not 
 of constructional value, but decorative. The round-headed trefoiled arch is 
 less common than the pointed. Examples of the former occur in Winchester 
 Chapel, c. 1200; the triforium of CHESTER CHOIR (524); and the south door- 
 way of St Michael's, Coventry. The cinquefoil is usually later than the trefoiled 
 arch ; but it occurs in Wells facade ; another charming example is the aisle 
 arcading of CARLISLE CHOIR (498) The early thirteenth century delighted in 
 beautiful combinations of pointed and trefoil arcades ; e.g. the porches of St 
 Albans and El)- ; the upper part of Ely facade and Durham west towers. The 
 trefoiled arch appears first in doorways; e.g. in Ely nave, c. 11 50, and Ripon 
 transept, c. 1170. It was not for nearly a century that the precedent set by 
 cusped doorways led to cusping the arches of the window tracer)'. 
 
 OGEE Arch. — As the upper curves of this arch (258) are reversed, it 
 cannot bear a heavy load, and so does not occur in pier-arches. It occurs, 
 exceptionally, in doorways ; as at TENBY (580), and in the palace at St 
 David's.f Ogee heads occur in windows, as not unfrequently in North 
 Hants, e.g. at W1LBV (270). Fine specimens of ogee arcading occur in the 
 * Willis' Vaulting, 43. t Illustrated in Prior, 39S. 
 
Beverley Minster, Percy Tomb. 
 Ely Lady Chapel. 
 
 Lincoln, S.W. Tower. 
 Gloucester, Western Processional Doorway 
 of Nave. 
 
OGEE ARCH. 
 
 ff frWV 
 
 PERCY TOMB, BEVERLEY 
 (269), where even the cusps 
 are ogees ; and, in a more 
 advanced form, beneath the 
 western towers of LINCOLN 
 (269). Finest of all are the 
 ogee canopies in the LADY 
 CHAPEL (269) and PRIOR 
 CRAUDEN'S CHAPEL, ELY 
 (130). Soon after 1300 
 ogee canopies almost wholly 
 replace their predecessors, the 
 straight-sided canopies of 
 GUISBOROUGH (354) and 
 HOWDEN (72). 
 
 Amongst the earliest 
 specimens of ogee arches are 
 those, both acute and de- 
 pressed, of the Eleanor Cross, 
 near Northampton: * the date 
 of which, as shown by the exe- 
 cutors' accounts, is between 
 1 29 1 and 1294.T Not much 
 later, probably, are the ogee 
 dripstones over the windows 
 of Lichfield Lady Chapel 
 and Wells chapter house ; 
 and the ogee arches in the 
 buttresses of Winchelsea. 
 In France the ogee arch 
 seems not to come into general 
 use till late in the fourteenth 
 century, J e.g. in the north 
 chapel of Amiens Cathedral, 
 c. 1 373-§ In England it comes 
 into general use, c. 1315- 
 And when once introduced, 
 there was a mania for it. Late English Decorated and French Flamboyant are 
 simply a glorification of the ogee arch ; the builders could not have enough of 
 it : above all, in niches and window tracery. 
 
 * This cross has been much rebuilt. See Proceedings of Society of Antiquaries, 1903. 
 
 T Ogee arches occur in Prior Eastry's choir screen at Canterbury, 1304: illustrated in 
 Caveler's Specimens, Plate 27. Other early examples are on the south side of Northfleet Church, 
 Kent ; in the east window of St Mary Stratford, Suffolk ; in the piscina at Fyfield, Berkshire ; 
 and on the heads of the buttresses of Winchelsea choir. 
 
 \ Viollet-le-Duc, Architecture, i. 9, and iv. 279. 
 
 § The ogee arch makes its first appearance in France, but only as a very rare exception, in 
 the last years of the thirteenth or the first years of the fourteenth century. Enlart's Manuel, 588. 
 
 Wilby West Front. 
 
OGEE ARCH. 271 
 
 " In its origin, it is, unquestionably, Oriental," says Mr G. G. Scott.* It is 
 true that it is used in India on a vast scale in those domes which are 
 constructed by corbelling. But in England it is not used constructionally at 
 all, but only decoratively. Moreover, all the links are wanting which would be 
 necessary to justify us in connecting our fourteenth-century ogees with those 
 of India. A much simpler origin may suffice. It is that wherever in early 
 geometrical window tracer}-, c. 12 40- 1290, a circle was set on two lancets, a 
 reversed ogee arch is almost irresistibly suggested. Indeed it is often difficult to 
 say that a reversed ogee was not really meant. If the other half of that figure 
 be added, we have at once the reticulated patterns which appear so early in the 
 fourteenth century; e.g. in the vestry of MERTON COLLEGE, OXFORD (480.2, 
 
 487-5)- 
 
 The ogee arch, like the pointed arch, may vary greatly in form : according 
 to the character of the arch whose curve is reversed to give the upper part of the 
 ogee, and according to the length assigned to the upper curve: e.g. in 25S.16 
 the lower curves are parts of a pointed arch, and the upper curves are very 
 short ; in 258.15 the lower curves are parts of a semicircle, and the upper curves 
 are very long. Of all the various combinations attempted, perhaps the most suc- 
 cessful are the very acutely pointed, and its opposite, the much-depressed ogee 
 arch. The former is constructionally strong, as it approaches the vertical : it 
 is a common and admirable feature of French Flamboyant window tracery.t In 
 our fifteenth-century work the lower curve of the acute ogee arch is often so 
 small as to be almost unnoticeable ; and sometimes it is suppressed altogether. 
 This acute ogee arch with the lower curves omitted is particularly common in 
 late canopies and tabernacle work ; e.g. the doorway leading into the cloister 
 from the western bays of GLOUCESTER NAVE (269) ; the buttresses of ST NEOTS 
 (356J. On the other hand, the depressed ogee is so weak that it is more often 
 executed in wood ; as in the doorway of the screen of ST MARGARET, LYNN 
 (162 j. When it is executed in stone, it is usually placed beneath a pointed arch 
 which acts as a relieving arch ; as in the tracer}" of the east windows of the 
 aisles of Hull chancel. 
 
 * Essays, 126, e. 
 
 t Compare the Hedon window, 480.3. 
 
Chapter XVII. 
 
 THE COMPOUND ARCH. 
 
 The Orders of the Arch — Recessing of the Orders — Sculpture on Arches — Arch-Molds. 
 
 COMPOUND ARCHES. — At first, e.g. in the early Christian basilicas and in much 
 Norman work ; e.g. the aisles of LEICESTER ST NICHOLAS (213), the arches were 
 simple, as in i.A below. But if the arch needs to be unusually strong — say 6 feet 
 thick — it is better to construct two independent arches, one on the top of the 
 
 other, each 3 feet thick. 
 Or we may construct in 
 three separate rings, as 
 shown in the right in IB. 
 Each of these sub-arches 
 or rings of which the whole 
 compound arch is com- 
 posed is called an order* 
 It is a safer form of arch 
 than the simple arch. If 
 anything happens to a part 
 of the latter, the stability 
 of the whole is Involved. 
 But in the compound arch 
 even if one or two of its 
 orders are fractured, the 
 remaining orders or order 
 may be able still to sup- 
 port the load. 
 
 Such a system of con- 
 centric arches or orders was employed by the Romans early in the sixth century 
 B.C. in the Cloaca Maxima at Rome ; three occur where it enters the Tiber. So 
 also in the Anglo-Saxon church of BRIXWORTH (274), probably built in the 
 eighth century, two concentric orders occur ; constructed of bricks from a neigh- 
 bouring Roman station. Another example occurs in the doorway of the Anglo- 
 Saxon church of Clee.t A pointed arch in two non-recessed orders occurs in the 
 west doorway of Etchingham, Sussex. \ 
 
 * To ascertain of how many orders an arch is compounded, it is necessary to note care- 
 fully whether the joints fail to coincide ; this is seen clearly in DURHAM transept (8), and 
 
 GLOUCESTER NORTH AISLE (99). 
 
 t Illustrated in Baldwin Brown, Arts i?i Early England, 164. 
 
 \ At Dendera in Egypt, Professor Flinders Petrie found passages, 6 feet wide, covered 
 with tunnel vaults of three rings of voussoirs, built in crude brick, and dating from 3500 B.C. 
 Anderson and Spiers, 121. 
 
 Arches. 
 
r 
 
 ■ 1 -■% 
 
 ^p?| - 
 
 
 
 1 ys| 
 
 ■ 
 
 # 
 
 
 • ^■jj* SE!S ®Si ■ ,:, 
 
 1 H^" 
 
 I 
 
 91 ' 
 
 J 
 
 y 
 
 I. . 
 
 
 
 Jr^ . a 
 
 v— J 
 
 ^ fl 
 
 > 1 
 
 
 tf*jMm: ^s| 
 
 ^TIPP^ — '^ ,,r Ni- 1 
 
 :f *W* • *yM 
 
 " i ^mmr- 1 -M 
 
 *~~*— -■ -<i^» 
 
 
 Steyning. 
 New Shorebam. 
 
 Ely Transept. 
 Norwich Apse. 
 
274 
 
 RECESSED ORDERS. 
 
 Recessing OF the Orders. — In the above compound arches the faces of 
 the orders are in the same plane. But far more often the orders are successively 
 recessed ; i.e. the innermost sub-arch or order is narrow ; the next above it 
 is broader ; the next is broader still, and so on ; as in Figs. 3 and 4, page 272. 
 
 This principle of construction also was known to the Romans. It 
 occurs in the Roman amphitheatre at Aries ; * in the palace at Treves in 
 windows which have arches of three recessed orders ; in S. Sophia, Constanti- 
 nople, and S. Apollinare in Classe, Ravenna ; both of the sixth century ; in 
 S. Pudenziana, Rome ; probably of the eighth century. In France an early 
 example is St Philbert de Grandlieu, which is assigned, doubtfully, to 815 A.D.f 
 
 Ickleton. 
 
 Brixworth. 
 
 It occurs in Anglo-Saxon work, e.g. in Yorkshire at Kirkdale and Kirk 
 Hammerton. J 
 
 The question now arises, why did the Romanesque and Gothic builders 
 practically universally, construct their compound arches, not in concentric, 
 but in recessed orders? What was the special value of the latter? The 
 following answer is often given. In early Norman work, and where ashlar 
 was expensive, e.g. in the Isle of Thanet till late in the twelfth century, it 
 was customary to build not only the walls and piers but the arches also with 
 
 * Illustrated in Petit's Church Architecture, i. 28. 
 
 t Enlarfs Manuel, 178. 
 
 % Illustrated in Baldwin Brown, 98. 
 
KKCKSSKD ORDKRS. 
 
 2/5 
 
 a core of rubble : * e.g. the pier-arch of LEICESTER ST NICHOLAS (213). When 
 an arch, however, was to be a compound one, then its construction was facilitated 
 if the outer orders only were constructed with rubble cores, but the inner one in 
 ashlar. And to economise ashlar, the inner order was made narrower than the 
 orders above it. Such a method was undoubtedly often adopted : it is seen quite 
 clearly in the tower arch of LEICESTER ST NICHOLAS (213), and in WHITBY 
 choir. On the other hand, in early examples of the compound arch, the inner 
 as well as the outer orders often have a rubble core : so that the explanation 
 given above does not cover the whole ground. Moreover it does not meet the 
 case of Roman recessed arches, constructed wholly in brick. 
 
 The true explanation probably lies in a fact which had the most profound 
 influence in conditioning the methods of building construction not only among 
 the mediaeval builders but among Oriental, Roman, and Byzantine builders 
 also. It is impossible to study in M. Choisy's 
 three classical works the methods of construction 
 in use among the latter without being forced 
 to the conclusion that the key to what seems to 
 a modern builder the roundabout and eccentric 
 practice so often pursued was nothing but the 
 desire, or to speak more accurately, the necessity 
 to economise centering and planks. Scaffolding 
 poles no doubt were cheap and plentiful, as they 
 are at present. But imagine a builder — Roman, 
 Byzantine, Persian, Indian — let loose in a wood 
 with nothing but a handsaw, perhaps with only 
 axe and adze, with instructions to turn trees 
 into planks. The fact is that the invention of 
 the circular saw, and the application of water- 
 power or steam-power to drive it, have revolu- 
 tionised the methods of building construction. 
 The methods of building construction now are 
 such as to encourage the use of planks to the 
 greatest possible extent ; in the ancient and 
 in the mediaeval world it was just the reverse. 
 It is only by constantly bearing in mind the 
 excessive cost of wooden centering till modern 
 
 times that one can apprehend the rationale of what seem to us the ab- 
 normal forms of arch and vault and dome which M. Choisy has explained so 
 lucidly. They are, each and all, the outward and visible sign of some non- 
 modern method of construction conditioned by the necessity of building with 
 the aid of as few planks as possible. So it was probably with the recessing 
 of the orders of the arch. Look at the Norman arch (272. 3;. It consists of 
 three orders.! Let us suppose that the wall to be carried by this compound 
 
 * It is noted by Professor Baldwin Brown that this method does not occur in Anglo- 
 Saxon work. 
 
 t It is represented as constructed throughout in ashlar: generally the two upper orders 
 would have rubble cores. 
 
 Whitby. 
 
27 6 ARCH MOLDINGS. 
 
 arch is 6 feet thick. If the arch had been constructed in three non-recessed 
 orders, the innermost order would be 6 feet broad. Therefore, in order to 
 construct this innermost order, it would be necessary to construct a temporary 
 wooden arch (a " centre ") 6 feet broad. Instead of that, a centre has been 
 put together, say 2 feet broad. On this the first order, a narrow stone arch, 
 is built ; also 2 feet broad. On the back of that is built another stone arch, 
 say 4 feet broad ; on the back of that is built another, 6 feet broad. If 
 that is not enough, order may be built round order, till there are as many 
 as eight or more. The Gothic west doorway of Elgin Cathedral is in eight 
 recessed orders. And all this without having to enlarge at all the original 
 wooden centre of a breadth of 2 feet only. The economy of centering- 
 is enormous.* 
 
 Such, then, may be the origin of the recessing of the orders of the arch. 
 Like almost everything else in mediaeval architecture, though afterwards turned 
 to decorative account, it was in its origin and in principle, nothing but a bit 
 of engineering. The compound arch of WHITBY (275), with three recessed 
 orders and a hood-mold, is nothing but the arches of Bernay and Jumieges 
 beautified. 
 
 The principle thus introduced was of course applicable to all arches ; not 
 merely to the pier-arches, but to the arches of doorway, triforium, clerestory, 
 window, and arcading. The side-walls of the Norman aisles were thick ; and 
 still thicker the end-walls. And it was in the openings in these, especially 
 in the western doorways, that the Norman architects delighted in carrying the 
 principle of the recessing of the orders to its furthest limits. Sometimes they 
 even built the wall round the doorways exceptionally thick in order to get more 
 orders for their doorways ; not because there was any constructional need of 
 them, but simply to increase the field for decoration. And even the small 
 Norman windows were sometimes recessed, quite unnecessarily, in two or three 
 orders. 
 
 Moldings and Ornaments of Arches. — In the earliest Norman 
 arches ; e.g. at Jumieges (c. 1040), the arches were square-edged : and thus they 
 sometimes remained in Normandy till the very end of the century or later, e.g. 
 at LESSAY and the ABBAYE-AUX-DAMES (319). In England also the unmolded 
 arch was frequently employed. It seems to have been a special favourite in 
 the greater Norman churches of the West and South of England ; e.g. Chester 
 Cathedral, Chester St John's, SHREWSBURY ABBEY (521), Leominster, Malvern, 
 GLOUCESTER CHOIR (294), Chichester before it was remodelled after the fire of 
 1 186, Winchester transept, Colchester St Botolph's, and St Albans; in the 
 two latter it may have been conditioned by the employment of brick. It 
 even occurs so late as c. 1170 in the transept of OXFORD CATHEDRAL (423.8). 
 
 As to the ornamentation of these square-edged arches, two methods were 
 employed ; sometimes one of the two exclusively ; often both together, as in 
 Steyning.t One was to carve the faces, and even, as in the central arch of 
 STEYNING (273), the soffit or intrados of the arch ; the other was to mold 
 
 * Cf. Garbett's Principles of Design, 202. 
 
 t It should be added that painted ornament was sometimes added ; e.g. on the brick 
 arches of the Norman tower at St Albans are painted sham voussoirs. 
 
HOOD-MOLDS. 277 
 
 both face and soffit. The former method became immensely popular in the first 
 half of the twelfth century ; the various orders were smothered with a profusion 
 of Romanesque ornament ; which, in Normandy and England, owing to our 
 unskilfulness then as carvers, consisted chiefly of geometrical patterns. In 
 NEW SHOREHAM CHOIR (273), c. 1 1 75, the ornament consists of scrolls of 
 Romanesque leafage. In late twelfth and in thirteenth century work the tooth- 
 ornament is common. But after this carving almost wholly ceases ; and molding 
 is preferred. 
 
 One step to the molded arch was to chamfer off the square edges, as at 
 SHREWSBURY ST MARY'S (424.7), where the inner order is chamfered and the 
 outer molded. The chamfered pier and the chamfered arch are specially 
 characteristic of the late twelfth-century work at LLANDAFF (424.6). And at all 
 periods in the parish churches, simply because of the expense, the arch was as 
 a rule chamfered, e.g. at Yarmouth ; and the pier was usually either a cylinder, 
 or an octagon ; which is a chamfered cylinder. 
 
 But even so early as the ABBAYE-AUX-HOMMES (319) a small change was 
 being made, which was destined to have the most momentous results. It was 
 the insertion of a small roll on the square edge of the arch ; a roll which was 
 a near relative of the small shafts of Romanesque doorways, windows, and 
 buttresses. Just the same treatment is seen in TEWKESBURY NAVE (297). 
 Reiterate these small rolls, and we have the moldings of the arches of the piers 
 and triforium of NORWICH APSE (273): in their multiplicity, delicacy, and 
 refinement, unmistakable harbingers of such lovely moldings of the thirteenth 
 century, as those of Ely and Durham, which are the special distinction of 
 English mediaeval art. 
 
 More often, however, the early moldings were bold and large : offsprings 
 plainly enough of the jamb-shaft, as seen at ICKLETON (274). In Bernay, 
 c 1 140, the same heavy roll, resting on the same heavy shaft, is seen in the 
 soffit of the pier-arches. At Christ Church, Hants, the heavy rolls of the pier- 
 arches "may almost be characterised as a mere continuation of the semi-columns 
 from which they rise." * Before the end of the eleventh century these massive 
 rolls were coming into use in Northern and Eastern England, though in the west 
 and south, as in Normandy, the square-edged arch was still preferred. But in 
 the end the molded arch superseded the unmolded in the south and west also ; 
 e.g. in the pier-arch of the NORTH AISLE OF GLOUCESTER, seen at the foot of 
 the illustration on page 313. 
 
 In the east, ELY (57; and Peterborough carried on the tradition of 
 Norwich, with more vigour and less delicacy, and paved the way for Gothic 
 moldings. 
 
 HOOD-MOLDS. — Above external arches, e.g. of doorways and windows, there 
 is usually a narrow projecting hood to throw off the wet ; it is called a dripstone 
 or weather-mold. Inside the church a dripstone is of course unnecessary, and in 
 the He de France is omitted ; but examples may be found elsewhere ; e.g. in the 
 late Norman nave of Bayeux, at Graville, and at Vezelay. But though unneces- 
 sary internally, its projection gives a beautiful shadow effect ; and serves to 
 demarcate more distinctly the arch from its load. Inside a building, it is called 
 * Ferrey's Christ Church, 42. 
 
278 ARCH-MOLDS. 
 
 a hood-mold, not a dripstone. Both dripstones and hood-molds are often carried 
 along horizontally as strings ; and their moldings are described under the head 
 of " Strings " (404). 
 
 The omission of the hood-mold over the pier-arches is characteristic of late 
 churches in EAST ANGLIA (552) and Somerset. 
 
 Chamfering. — Originally the orders of the arch were square-edged, and 
 in the West of England sometimes remained so till late in the twelfth century, 
 eg. at St John's, Chester. It is probably because these sharp edges were often 
 broken in transit or in working, that the practice of chamfering arose. 
 
 ARCH-MOLDS. — The history of these divides roughly into three periods : the 
 first, that of the eleventh century up to c. 1160, when the molds employed are 
 chiefly semicircular rolls and shallow, i.e. segmental, hollows. The second lasts 
 till the end of the thirteenth century or a little later. In this period the roll may 
 be pointed or have a keel or a fillet, or two or three fillets. The hollows 
 generally form a semicircle or a segment larger than a semicircle, and the 
 rolls are greatly undercut on either side, in order that, to the eye, they may 
 seem not to be part of the arch behind them ; but detached from it just as 
 much as shafts of dark Purbeck marble are detached from the core of a 
 central cylinder. And under certain conditions of light, so black are the 
 shadows of these undercut hollows, e.g. in the arch from CWM HIR (422.6), 
 that the roll-molds really do seem to be detached. It may be said to be an 
 attempt — and a successful attempt — to do in the arch by means of shadows 
 what was done in the pier by means of dark marbles — to get as vivid a con- 
 trast between roll with high lights and black-shadowed hollows as is that of 
 Purbeck marble shafts against a white limestone pier. In the third period, from 
 the fourteenth century onward, the rolls tend to disappear from the arch, just as 
 the shafts from the pier; and the hollows become "casements," broad and shallow; 
 cut to hold, not one uniform blackness, but varying shades, grey white and black, 
 or the reverse, according to the direction from which the light proceeds. The 
 rolls were softened away almost entirely ; represented by a single undulation of 
 surface (an ogee), or by two undulations separated by a quirk (a double ogee 
 or bracket); the latter is well seen in ST MARY'S, BEVERLEY (253). In these 
 arrangements in grey there is subtle and exquisite beauty ; perhaps because it 
 is subtle, and so does not force itself on the observation, that our later Gothic 
 system of molding has been held to be " degraded " and " debased," in com- 
 parison with the vivid contrasts of blackest black and whitest white which were 
 the joy of the thirteenth-century mason. 
 
 Before studying or copying a suite of moldings, it is necessary first to 
 ascertain what was the shape of the arch, with the exact dimensions of each 
 member of it, before the moldings were sunk on it. If the arch is not later than 
 c 1260, the arch was probably built originally in square-edged recessed orders; 
 and the moldings were carved on the faces of these. After 1360, usually the 
 square edges were previously chamfered off, and the moldings were then 
 sunk in these chamfer planes. Between c. 1260 and c. 1360 both practices 
 prevailed. 
 
 Arch Moldings, c. 1050— c. 1 160.— Till about 1160 arch-molds consist, as 
 a rule, of little more than suites of angle-rolls. The CHICHESTER (666.1) 
 
ARCH-MOLDS. 279 
 
 example is from the western bays of the nave, built after the fire of 11 14. 
 The arch at CASTLE HEDINGHAM (666.2) is said by Mr Hadfield to be between 
 1088 and 1 107. 
 
 The examples from WARMINGTON and WALSOKEN (666, 667) belong to 
 the early part ; those from NEW SHOREHAM and BYLAND (667) to the latter part 
 of the period of 1 145 to 1 190. Notice that at New Shoreham, though the hood- 
 molds, A, A, are the same in section, and the arches are nearly if not quite 
 contemporaneous, the arch on the right is quite different in type to the other ; 
 for all its rolls are pointed. Pointed shafts and pointed rolls are especially 
 common in early Yorkshire Gothic ; e.g. Byland. Note that both in New 
 Shoreham and Byland (choir entered 1 177) the rolls are already more numerous, 
 and consequently have to be smaller. The New Shoreham capitals and arches 
 are illustrated on page 667. 
 
 Arch Moldings, c 1190 — c. 1245. — The example from wawne (666) 
 shows a simply chamfered arch, carried by a cylindrical pier ; these, from their 
 economy of labour and skill occur at all periods in the smaller parish churches. 
 The arch-mold from GRIMSBY (667.5), which probably belongs to the first 
 years of the thirteenth century, is interesting as a survival of Romanesque 
 molding. It has resemblances to that of the pier-arches of Peterborough nave, 
 built about a century before. Little rectangular projections and curving 
 horns are not uncommon ; e.g. at West Walton church, Norfolk, where 
 the rolls are either keeled or filleted. In the TEMPLE CHOIR (668.1), on the 
 other hand (finished 1240;, only one roll is filleted, viz. A. Note how greatly 
 the rolls are undercut, showing the excellence at once of the stone and the 
 mason. LINCOLN NAVE (668.2), c. 1 120, is much more advanced; here all the 
 rolls are filleted ; and, what is very important, the)' are neither semicircular 
 nor pointed. Perhaps the most exquisite moldings are those executed at the 
 end of the period in Durham eastern transept. 
 
 Arch Moldings, c. 1245 — c. 13 15. — In such arches as those of the Temple 
 and Ely, the hollows and rolls were so very numerous that they were necessarily 
 very narrow. For a time they remain so. The arch in the TEMPLE CHOIR 
 (668.3) i s a glorification of yet another advance in type, viz. the roll and triple 
 fillet. In this example also the necks of the rolls are perilously undercut ; 
 in fact the mason had been allowed too much play ; he was executing in stone 
 moldings which ought to have been confined to wood ; he had become too clever. 
 ST MARY'S, YORK (668.4), 1 273-1 295, is of early type, but this excessive under- 
 cutting of the rolls has been restrained. The rolls, however, are still numerous 
 and small. But in most of the later work of this century, the small delicate rolls 
 and undercut hollows disappear, never to return. A bolder, more vigorous style 
 comes in ; the rolls are large, and the hollows are large and are comparatively 
 little undercut. Moreover, as in LINCOLN PRESBYTERY (669), they are arranged, 
 more often than before, in groups according with the orders of the arch ; and 
 the form of the filleted roll is varied with great freedom, as in TIXTERN CHOIR 
 (669) at A, A. And sometimes, as at Tintern, to facilitate the undercutting, the 
 joint is placed in the hollow. 
 
 Arch Moldixcs, c. 1315 — c 1360. — In the wtnchelsea arcpies (669), 
 which may be as early as 13 10, the rolls are much broadened, and that at the 
 
28o ARCH-MOLDS. 
 
 expense of the hollows ; and either side of each roll forms an ogee or undulatory 
 curve. A vast change in the formation of moldings generally sets in about this 
 time. The roll and triple fillet is retained at B, B, B ; but with modifications. The 
 LEADENHAM ARCH (670) also retains the three fillets. One of them is broad, 
 usually a mark of late date ; the upper order has the wave-molding at B. But 
 at HELPRINGHAM (670) the side-fillets C, C, C, C, and the hollow of the 
 Leadenham lower order have been softened away altogether ; so also with the 
 hood-mold D ; so that both orders and the hood-mold also have wave-moldings. 
 So also at NORTHBOROUGH (670) the two adjoining arches have the same 
 characteristics of rolls with ogee curves. At BOTTISHAM (671) a new char- 
 acteristic appears ; two ogees meet at E, E, E, E, and produce a double ogee 
 (" bracket "). In various forms this " bracket " remained in constant use to the 
 end of Gothic architecture. In the chancel of BEVERLEY ST MARY (670) we 
 have an example later in date, but less advanced in type. 
 
 Arch Moldings AFTER 1360. — The moldings of the first half pass 
 imperceptibly into those of the second half of the fourteenth century. The nave 
 of ST mary's, BEVERLEY (253), exhibits both the "bracket" and a narrow "case- 
 ment." This " casement " soon became broad and shallow ; and varying in 
 contour according to its height above the line of vision, was to abide to the 
 last ; it is a special characteristic of fifteenth and sixteenth century Gothic. 
 CHELMSFORD (67 1 ) is a more advanced example; the "casement" F is two- 
 centred only; and more complicated forms of the "bracket" occur. Somewhat 
 simpler molds are shown from the south side of ST MARY'S, OXFORD (671). 
 All exhibit " casements " at F. 
 
Chapter XVIII. 
 VAULTS WITHOUT DIAGONAL GROINS OR RIBS. 
 
 The Dome — The Semi-dome — The Barrel Vault — The Half Barrel — The Stone Ceiling — 
 
 The Span Roof of Stone. 
 
 As in the plan and the supports, so in the vaulting of the churches, the story 
 begins with Rome. In ancient Rome, towards the end of the third century 
 A.D., four methods of vaulting were in use — the dome ; the semi-dome ; the 
 tunnel, wagon, barrel, or cradle vault ; and the groined vault. Of these the 
 groined vault is of supreme importance ; it is one of the greatest inventions in 
 the architectural history of mankind. It is the groined vault, reconstructed 
 indeed by different methods, not the pointed arch or the flying buttress, which 
 is the generative principle of Gothic architecture. We defer the detailed treat- 
 ment of it till later. 
 
 The Dome. — The dome is a vault of masonry, which, in its original form, 
 as in the Roman Pantheon, was circular, and rested on all sides of the space to 
 be covered. In the Pantheon the dome rests on a circular wall. A barrel vault 
 rests on two sides of a rectangular space to be covered, and leaves two sides free. 
 A groined vault rests only on the angles of the space to be covered, and leaves 
 all four sides free. The dome and the barrel vault are of immemorial antiquity. 
 In some districts men were compelled to build in stone or brick or mud because 
 there was no wood — e.g. in Babylonia, as Strabo says ; in other districts because 
 they had not the tools to work wood, e.g. Ireland — perhaps not having learnt 
 to use iron or even bronze. In all such cases some form of dome or tunnel 
 had to be devised for shelter. In Eastern Christendom the dome became 
 the dominant factor in church design ; whether a single dome, as at St 
 Sophia, Constantinople ; or a central dome encircled by other domes, as 
 at St Mark's, Venice ; or a row of domes, as at Angouleme.* North of 
 the Loire medkeval domes are exceedingly rare. In the early Norman 
 abbey of Bernay the aisles are roofed with domes, which have usually been 
 attributed to the eleventh century ; but from the precision of the masonry it 
 is clear that they belong to the remodelling of the church in the time of 
 Louis XVI. At Rucqueville t (Calvados) enough remains to show that there 
 was originally a dome on pendentives ; but the capitals in this church are of 
 Burgundian type, and show that the church is an "outlier." At Goring-on- 
 Thames, \ under the tower is a Norman ribbed dome.?; The courses are hori- 
 
 * The new Roman Catholic cathedral in Westminster reproduces the plan and domes of 
 Angouleme. 
 
 + Ruprich-Robert, Plate v. \ Illustrated in Scott's Lectures, ii. 170. 
 
 § This is really a variant on the so-called Cloistered 'vault, or squared dome. See Scott's 
 Lectures, loc. cit. 
 
282 
 
 Gloucester, Upper Aisle of Choir. 
 Scarborough St Mary. Scarborough St Mary. 
 
THE SEMI-DOME. 
 
 283 
 
 zontal, but they stop at each rib, and start again at different levels on the 
 other side ; so that it is a compound of sections of four domes rather than 
 one dome. Of domical ribbed vaults there are many instances, especially in 
 France, e.g. all the Angevin vaults ; but their courses are not horizontal with 
 radiating joints, i.e. they are not of the form produced by the revolution of an 
 arch round its own central vertical axis, and therefore are not domes. 
 
 The Semi-dome. — But though we, with Northern France, lost the old 
 Roman art of constructing domes, there never was a time when semi-domes could 
 not be built. Hemicycles were common in the Roman Thermae, palaces, and 
 other large buildings, and were roofed with semi-domes. And throughout the 
 whole of the Dark Ages, from the sixth century onwards, wherever an apse was 
 erected, it was the rule to ceil it with a semi-dome of masonry. In Normandy 
 in the eleventh century every apse was vaulted with a semi-dome (cul de four).* 
 In one of the earliest of 
 our own Norman build- 
 ings, St John's Chapel in 
 the Tower of London, the 
 apse has a semi-dome. 
 In village churches there 
 remain several examples 
 of semi-domes,^. CHECK- 
 EXDON(2i). Usually they 
 are strengthened by ribs. 
 When, however, the east 
 end was squared, the semi- 
 dome was no more of ser- 
 vice, and went out of use. 
 
 The Barrel Vault. 
 — This simple and ancient 
 type of vault produced, 
 south of the Loire, whole 
 schools of Romanesque ; t 
 that of Poitou, with barrels 
 resting on the pier-arches ; that of Auvergne, with barrels resting on the 
 triforium ; that of Provence, with barrels resting on the clerestory walls, and with 
 half-barrels in the aisles ; and that of Burgundy, with similar barrels but with 
 groined aisles. No such grand developments of the barrel vault were worked 
 out in Normandy and England. Sometimes foreign influence — e.g. that of the 
 Cistercians of Burgundy — imported the barrel vault ; e.g. at Kirkstall Abbey 
 the walled chapels of the transept are ceiled with pointed barrel vaults ; at 
 FOUNTAINS (101), also Cistercian, pointed barrels set transversely and resting 
 on semicircular arches, were employed in the aisles of the nave. * We have one 
 very remarkable example of the use of the barrel vault — ST JOHN'S CHAPEL 
 in the Tower of London (c. 1080J. The chapel has both upper and lower 
 
 * Kuprich-Robert, i. 51. t For all these see Classification of Romanesque. 
 
 \ See plates in Sharpe's Arch. Parallels; and if. Fontenay ; and St Nicolas, Girgenti ; 
 illustrated in Enlart's Gothic in Italy, 75, 76, 247. 
 
 St John's Chapel, Tower of London. 
 
284 
 
 THE BARREL VAULT. 
 
 vaulted aisles ; the nave and the upper aisles have barrel vaults ; the lower 
 aisles are groined. The barrel vault was in use in Anglo-Saxon days, e.g. in 
 the crypts of Hexham and Ripon, which are undoubtedly of the seventh century ; 
 and in the porch of Monkwearmouth, which may be contemporaneous. The 
 slype, or passage between a monastic cloister and the cemetery, often has a barrel 
 vault ; as in Oxford Cathedral. At Norwich, beneath the Bishop's Palace, is 
 a barrel vault, 20 ft. in span. St Cormac's Chapel, Cashel, a specimen of 
 the twelfth-century Romanesque of Ireland, has a barrel vault strengthened 
 with transverse arches.*' The chapter house of Reading Abbey had a barrel 
 vault 42 ft. across : that of Gloucester has a pointed barrel of about 35 
 
 ft. At ABBOTSBURY, 
 Dorset, high on a hill 
 above the sea, is St 
 Catharine's Chapel, the 
 roof of which is stone 
 within and without: 
 the internal roof is a 
 pointed barrel ; t so also 
 is that of Bothwell in 
 Scotland, { with which 
 may be compared the 
 semicircular barrels of 
 Roslyn Chapel. § 
 
 The barrel vault 
 then left very few visible 
 signs of its existence 
 in English mediaeval 
 architecture. 
 
 The Half Barrel. 
 — Of the barrel vault the 
 demi-berceau is an im- 
 portant and valuable modification. If only one side of a barrel vault is built, it is 
 called a half barrel or demi-berceau ; it is simply one side of a barrel vault. In 
 the Romanesque of Auvergne it was employed to vault the upper aisles, a whole 
 barrel being employed to vault the nave. Then, the haunches of the nave 
 barrel being filled up with solid masonry, it was easy to construct one solid roof 
 of stone covering up both barrel and demi-berceau, as at Issoire. || But in 
 the twelfth-century Romanesque of Provence a pointed barrel was set on the 
 top of the clerestory wall, and the aisles were vaulted with half barrels in order 
 to provide continuous abutment for the clerestory walls. Of the use of the 
 half barrel we have one important example in England. In the upper aisle of 
 GLOUCESTER CHOIR (282), on either side of the choir and originally round it to 
 
 * Illustrated in Scott's Lectures, ii. 23. 
 
 t Illustrated by J. D. Wyatt in Building News, 6th August 1880. 
 
 J Illustrated in Fergusson, ii. 435. 
 
 S Illustrated in Fergusson, ii. 434. 
 
 II See Classification of Romanesque, 277, 278. 
 
 Abbotsbury. 
 
THE STONE CEIEING. 
 
 285 
 
 the east also, is a Norman half barrel strengthened with ribs. The same con- 
 struction appears to have existed at Tewkesbury and Pershore.* 
 
 Exactly the same system occurs in the upper aisles of the nave of the 
 Abbaye-aux-Hommes, Caen. This demi-berceau is held by Ruprich-Robert not 
 to have been built till 
 
 the seventeenth cen- 
 tury, when the church 
 was restored by the 
 Benedictines with the 
 greatest care and ac- 
 curacy. But it seems 
 very unlikely that at 
 such a period they 
 would have devised 
 such an antiquated 
 method of abutment 
 as the demi-berceau ; 
 especially as the sex- 
 partite vault of the 
 nave requires inter- 
 mittent, not continu- 
 ous abutment. Prob- 
 ably what they did 
 was to repair or rebuild 
 the old demi-berceau 
 of the eleventh century. 
 If so, the Gloucester 
 demi - berceau would 
 probably be copied 
 from that of Caen. 
 
 The Stone Ceil- 
 ing. — We may here 
 introduce two forms 
 of roofs which might 
 be designated semi- 
 vaults ; for they have 
 in part, though not 
 throughout, the arcu- 
 ated construction of a 
 genuine vault ; more- 
 over they are con- 
 structed of masonry. These are stone ceilings and span roofs of stone. 
 
 The stone ceiling occurs in a rudimentary form in Roman ruins at Aries 
 and Nimes.t In Northern and Central Syria it gave rise to a distinct style, 
 producing the buildings erected between A.D. 105, when Syria became a 
 
 * Scott's Lectures, i. 90. 
 
 + See Dr West in Journal of R.I. B. A., November 1874, 32; and Anderson, Greece and 
 Rome, 179. 
 
 Abbotsbury. 
 
286 
 
 Minchinhampton, South Transept. 
 
THE SPAN ROOF OF STONE. 
 
 i%7 
 
 Roman province, and c. 630, when these districts were deserted for ever on the 
 approach of the Mohammedan invaders from Arabia. Of these the earliest 
 dated buildings are of the time of Constantine — early fourth century. They 
 have been fully described by De Vogue.* 
 
 If we had to ceil an oblong room we might put beams across it, and on 
 these beams put joists longitudinally. Being short of wood, the Syrian Christians 
 erected transverse arches of stone instead of the beams, and on the haunches 
 of these arches built up walls to form a horizontal bearing. The arches were 
 
 Minchinhampton, South Transept. 
 
 built 8 or 9 ft. apart, and then the distance from the top of each cross wall to 
 the next was spanned by long blocks of basalt, which in those districts are 
 plentiful. Of this method, which is only applicable where very long blocks of 
 stone are to be had, we have no example. 
 
 Tme Span Roof OF Stone. — But where large flagstones were obtainable, 
 it was not uncommon to construct porches of a series of arches set very close 
 together ; on these it was easy to construct a substantial roof of flagstones. On 
 
 Syrie Ccnfrale, Paris. See also Scott's Essay, 54-62 
 
288 THE SPAN ROOF OF STONE. 
 
 the south side of ST MARY'S, SCARBOROUGH (282), once a Cistercian priory, 
 three chapels have been built in this fashion ; the roofs consist of overlapping 
 flagstones. In the transept of MINCHINHAMPTON (287) the roof is of stone, 
 and is supported by skeleton arches.* The room over Leverington porch is 
 roofed in the same fashion. 
 
 * Mr Spiers mentions the Church of Montataire in Picardy, where the aisles have a 
 roof of very steep pitch, consisting" of long flagstones resting on arches nearly 10 ft. apart. 
 R.I.B.A., 1874, 3, 46. So also Willingham, Cambridge ; illustrated in Rickman's 7th 
 edition, 218. 
 
Chapter XIX. 
 
 GROINED VAULTS. 
 
 Object of Vaulting — Roman Groined Vaults — Romanesque Groined Vaults — Groined 
 Vaults in Crypts — Groined Vaults in Aisles-High Groined Vaults. 
 
 It was, however, to none of the foregoing that the wonderful development 
 of vaulting is due in England and Northern France. It is clue to the fact 
 that in the darkest ages the mediaeval builders had never lost the tradition 
 how to construct another type of Roman vault — the groined vault. But before 
 we go further it may be well to ask what was the reason why there was such 
 a consuming anxiety from the eleventh century onwards to provide the 
 churches with the stone ceilings which we call vaults. We may also ask 
 why were the other Romanesque provinces of France anxious so earl)' to 
 vault their churches, and why did the Normans and English postpone vaulting 
 so long. The answer generally given to the first question is that in the tenth 
 and eleventh centuries many churches had perished by accidental fires, some by 
 lightning, and that a very large number had had their roofs burnt off by the 
 Northmen all round Western Europe, Great Britain, and Ireland, wherever they 
 could get access by sea in their ships or up the rivers in their boats. The idea 
 is that the tenth century was marked by a sort of Norske feu-de-joie of the 
 churches of Western Christendom. The picture is dramatic ; but it must be 
 borne in mind that the Auvergne province, where vaulting seems to have been 
 developed very early, is wholly remote from Viking inroads ; and, secondly, it is 
 just in the districts which we know certainly were devastated by the Northmen, 
 Normandy itself and Northern France, just where the timber roofs had been 
 burnt, and where fireproof construction would have been most appreciated, that 
 vaulting was most backward. Perhaps humbler constructional reasons had 
 something to do with it. One people is a race of masons ; that may have 
 been so in Auvergne, where there are scores of miles of lava-streams from the 
 breached craters of the old volcanoes, providing in abundance the material, light 
 volcanic cinders (pumice and tufa;, which is best adapted for vault construction. 
 Using these materials, the Auvergne builders were able early to construct churches 
 without any wood in them anywhere, with vault and roof all one, constructed of 
 solid stone. But the Normans and the Franks and the English were a race of 
 carpenters and shipwrights more than of masons. Those who could lay the keel 
 of a ship and bend up its ribs could invert the ribbed keel and make a roof of it. 
 It may therefore have been easier in Auvergne to build a stone roof than a timber 
 one ; in Normandy, England, and Northern France easier to construct a timber 
 roof than a vault. Nor did we in England ever lose our love for carpentering. 
 Nowhere else were such magnificent wooden roofs constructed. Indeed in the 
 North of England vaulting never made its footing sure : there the elaborate 
 
290 
 
 GROINED VAULTS. 
 
 Heme vaults and the fan vaults of the South seldom found their way ; the four- 
 teenth-century vaults of Guisborough, Howden, and Beverley are of simple 
 character ; and after these no high vaults of importance seem to have been con- 
 structed. York indeed and Selby prepared springers and pinnacles and flying- 
 buttresses for vaults of stone in the fourteenth century ; but in the end the 
 carpenter had his way, and Selby choir and York transept, nave, and choir had 
 to be content with vaults of wood. Moreover the Gothic vault is but a thin 
 shell, and requires a second roof of timber to shelter it from the weather. Our 
 old builders, being practical men, may well have objected to a roof which itself 
 required to be roofed. So the story of the English vault begins in Normandy 
 
 with the fact that though the builders vaulted 
 the aisles, they omitted to put up a high* vault, 
 i.e. to vault the central aisle or nave. This is 
 the position of things at Bernay Abbey, begun 
 1013-1017, and finished in part in 1050 ; and 
 at Jumieges Abbey, begun 1040, and conse- 
 . crated in 1067. Nor were either of the great 
 Caen abbeys, or Cerisy, or St George's de 
 Boscherville intended originally to have high 
 vaults. In many instances the builders hesi- 
 tated ; hardly knew what they wanted ; de- 
 signed the church for a vault, but ended by 
 giving it merely a wooden roof, as at Ely and 
 Peterborough. But there is no sign of vault- 
 ing shafts, or of any intention to vault, even 
 in the twelfth century, in such important 
 churches as Dunfermline, St Bartholomew, 
 Smithfield, Malvern, Melbourne, Gloucester, 
 Rochester, and Shrewsbury. Even the Cister- 
 cians, who came to England full of traditions 
 of Burgundian vaulting, could not influence 
 the Yorkshire people to put high vaults on the 
 naves of Fountains, Kirkstall, Furness, Byland ; nor would the Benedictines vault 
 the naves of Whitby and Selby, or the Augustinian canons the choir of Hexham. 
 
 Basilica of Maxentius. 
 
 Groined Vaults. 
 
 But to return to the Roman groined vault. It was constructed in mortar, 
 in which stones were laid in horizontal layers.! It was not a thin shell like the 
 tunnel vaults of the Cluniacj churches or the ribbed vaults of Gothic architec- 
 ture, but very thick and very heavy. Being so thick and heavy, it was con- 
 structed on arches of brick or tiles ; which, when the concrete had been built 
 up on them, remained as permanent centres. The under surface of the arches, 
 as also of the concrete, was plastered over ; and then this under surface received 
 a decorative treatment. The arches were invisible when once the vault was 
 
 * This is a convenient term to distinguish the vault of a central from that of a side aisle. 
 + Choisy's L'Art de batir chess les Romains, 25. J Classificatio7i of ' Rovianesgue, 279. 
 
GROINED VAULTS. 
 
 291 
 
 finished. When once the concrete had set, the vault had little lateral thrust.* 
 But nevertheless the Roman builder provided a very large amount of abut- 
 ment. His practice was the very reverse of that of the Gothic builders. Me 
 liked to have an overplus of stability ; they were ever trying to find what was 
 the minimum to which supports and abutments might be reduced. Thus in the 
 BASILICA of MAXENTlUSt (290) the plan shows enormous buttresses on 
 either side of the nave. J The dimensions of these vaults far transcended 
 anything that was attempted in mediaeval times ; for each rectangular bay of 
 this nave had a span of 83 ft, and its diagonal groins a span of 117 ft. The 
 nave of Lincoln has a span of 39 ft. ; that of Amiens, 46 ft. ; Milan and Seville, 
 56 ft. Another vast groined hall, the tepidarium of the Baths of Diocletian, 
 forms the transept of the Church of S. Maria degli Angeli, Rome. 
 
 It was the massive Roman barrel vault which was copied by the Roman- 
 
 producing churches with 
 
 Roman Groined Vault. 
 
 esque builders of France south of the Loire 
 ponderous supports, and with high vaults of 
 barrel form. In the Gothic cathedral of Albi, 
 however, there is a genuine survival of the 
 Roman groined vault ; 
 it is built of brick, 
 with internal b u t - 
 tresses, and with mas- 
 sive groined vault 
 " holding together by 
 cohesion like an arti- 
 ficial monolith," with 
 ribs which are decorative, not constructional. § 
 
 A Roman groined vault consists of two intersecting semi-cylinders. Or 
 imagine two drain-pipes to interpenetrate. It may have arisen more or 
 less accidentally ; imagine two corridors at right angles, of the same breadth 
 and height, both roofed with semicircular tunnels; where they cross there 
 will be a groined vault. It is a case that must often have occurred in the 
 large mansions and underground reservoirs, and other public buildings of 
 Rome. Often one corridor would be smaller than the other, and its 
 tunnel vault at a lower level than that of the other. This is indeed the 
 case with most actually existing Roman vaults ; and there are a few examples 
 
 * On the homogeneous nature of the Roman vaults see Professor Middleton's Ancient 
 Rome, 2 vols., London, 1S93. Objection has been taken to this theory, which certainly is con- 
 tradicted by Roman practice. 
 
 + In order to make the system of abutment clear, the screen walls from C to D have been 
 omitted ; as well as the arches from A to V, which convert the three compartments on either 
 side of the nave into a continuous aisle. 
 
 X On the whole subject of Roman vaulting see the classic work of M. Choisy, I.'.lrt dc 
 bdtir chcz les Remains. 
 
 § Choisy's Histoire, 467, 510. 
 
 I This is only true in a general way. Some Roman, some Romanesque, and most Byzan- 
 tine groined vaults are domical ; i.e. bulge upwards at the centre ; and are such as would be 
 produced by the intersection of two barrels or casks, not of two agricultural drain-pipes ; they 
 are intersecting barrel vaults, not intersecting tunnel vaults. 
 
29 2 GROINED VAULTS. 
 
 of it in Romanesque, e.g. in St Remi de Reims and St Benoit-sur-Loire.* 
 Usually in Roman work there are no transverse arches between the different 
 compartments of a groined vault ; so also in the aisles of the nave and in the 
 crypts of the ABBAYE-AUX-DAMES, CAEN (293),! and Rochester ; and the under- 
 croft of Durham refectory. But in Norman work far more often each bay of the 
 groined vault rests on a pair of massive square-edged transverse arches. Early 
 examples are S. Vitale, Ravenna, sixth century ; Aix-la-Chapelle, ninth century. 
 This is the normal form in which the vault appears from Roman times to the 
 eleventh century. 
 
 It is in the Romanesque crypts that the mediaeval groined vault appears 
 earliest. Even when the builders were afraid of vaulting anything else, they 
 usually vaulted the crypt. In a crypt vaulting was easy ; the crypt was low ; the 
 pillars in it could be placed so as divide it up into numerous small compartments; 
 and each compartment could be kept square ; and the vault did not need to be 
 very strong, as all that it carried was the pavement of the floor above. \ A rude 
 attempt at a groined crypt may be seen in Wing Church. § In the eighth 
 century it occurs in the crypt of the older cathedral at Brescia ; in the ninth 
 century in those of St Vincent in Prato, Milan ; of Alliata ; and in two crypts 
 at Orleans. || The tradition preserved in the crypts also survived in the vaulting 
 of the little semicircular ambulator)' of S. Stefano, Verona ; and the eastern 
 bay of each aisle of Alliata ; both probably ninth century; and both the nave 
 and the rectangular oratory of Cividale were vaulted even in the eighth century. 
 
 To the end of the eleventh century all the greater churches in Normandy 
 had groined vaults to their aisles, e.g. Bernay choir and Jumieges (in the latter 
 upper and lower aisles both have groined vaults) ; Cerisy, and the Abbaye-aux- 
 Hommes.^1 So also in the twelfth century at LESSAY (293), Blois, Poissy, 
 Pontigny ; and the upper aisles of the choirs of St Germer and Vezelay : in the 
 latter at least as late as 1170 ; in some of the latter the builders were certainly 
 acquainted with the construction of ribbed vaults : yet constructed groined 
 vaults by preference in the aisles. In England also in the eleventh century 
 the normal practice was to give the aisle a groined vault, e.g. at NORWICH (238). 
 And even till the middle of the twelfth century the same form still occurs 
 occasionally, e.g. in St Bartholomew's, Smithfield. 
 
 But, here and there, the builders took courage to construct high groined 
 vaults as well. The photograph shows in the distance the semi-dome, and in the 
 foreground two oblong bays of the high vault over the aisled choir of ST 
 NICHOLAS, CAEN,r. 1080 (293),** with which that of St George's de Boscherville 
 
 * Choisy's Histoire, 151; and Roman Construction. 71. 
 
 + Illustrated in Scott's Lectures, ii. 157 ; and Hope's Rochester, Plate iii. 
 
 % See the western bays of the crypt of WORCESTER (192.2). 
 
 § Seventh century, Micklethwaite ; comparatively late, Baldwin-Brown. 
 
 || Cattaneo, 202, 231, 238, 250; and Enlart's Manuel, 166, 176, 184. 
 
 H The term groined vault is often used in a loose way of any intersecting" vault, whether it 
 has diagonal ribs, or only diagonal groins, i.e. sharp edges or arrises. It is convenient to 
 restrict it to the latter. Throughout, therefore, when speaking of a groined vault, I mean an 
 unribbed intersecting vault. 
 
 ** This church is now divided by floors, and used as a granary ; the photograph is taken 
 from the top floor ; the groined vault is seen in the foreground. 
 
29 3 
 
 
 
 
 
 WJ00wm% 
 
 
 x^^—^^^ 
 
 
 ///^~ ^^x\ X 
 
 . 
 
 \ 
 
 QyttMsiiF^NifeiJft 
 
 
 
 
 Abbaye-aux-Dames, Caen. 
 
 St Nicholas, Caen. 
 
 Lessay Abbey, Normandy. 
 Gloucester Choir Aisle. 
 
294 GROINED VAULTS. 
 
 may be compared. The choir of the Abbaye-aux-Dames is vaulted in the same 
 way, except that the bays are nearly square ; and that the high vault rests not 
 on isolated piers, but on solid walls separating the choir from the aisles, as at 
 Rochester, St Albans, and Cerisy. This vault may be earlier than that of St 
 Nicholas, Caen.* The vast naves of the German Cathedral of Speyer and of 
 the Burgundian Abbey of Vezelay received groined vaults ; the former c. noo, 
 the latter probably after the fire of 1120. Any high groined vaults we may 
 
 Gloucester, North Ambulatory. 
 
 have had in England have disappeared. Sir G. Scott was of opinion that Here- 
 ford choir was designed for one.f 
 
 As the illustrations of Cerisy, Lessay, and Norwich show, the bays or 
 severies of these aisle vaults were separated by massive unmolded semicircular 
 arches. These arches were built first. Then on them a temporary vault 
 of boarding was put together ; what is called " centering " ; the upper 
 surface of the boarding was covered with a thin layer of mortar, and 
 
 * See Ruprich-Robert, i. 71 ; and Plans, i. 8, 9. 
 
 t See his restoration of it in Building News, August 9, 1878. 
 
GROINED VAULTS. 295 
 
 on this was constructed the permanent vault of masonry. This masonry was 
 mere rubble ; packed together almost anyhow, without any attempt at regularity 
 in the courses. When the mortar had set, the centering was removed and the 
 lower surface of the vault was plastered over.* Usually the lower part of the 
 groins is well marked, but towards the top they die away, leaving the summit 
 of the vault a sort of flattish dome. In modern restorations these vaults have 
 usually been plastered over neatly, and the groins have been carried up to the 
 apex of the vault. These rubble vaults were of course very thick and very 
 heavy : but if they were built with good mortar, which was not always the case, 
 each bay would practically form one homogeneous block of concrete, with little 
 if any lateral thrust. They required strong supports more than strong abutment ; 
 and therefore were well suited for Xorman architecture, with its thick walls 
 and piers, but shallow buttresses. 
 
 With these rude vaults the builders became exceedingly expert ; they 
 succeeded in vaulting not merely squares, or oblongs, as in the high vault of St 
 Nicholas, Caen, but trapeziums and triangles, as in WINCHESTER CRYPT, c. 10S0 
 (192.5), and in GLOUCESTER AMBULATORY (293). There are even hexagonal 
 groined vaults in the eastern apses of Gloucester transept. Or sometimes, 
 as in WORCESTER CRYPT (192.2), they covered the apse not with a semi-dome, 
 but a series of triangular groined vaults. Or, in the ambulatories, the)' banished 
 the trapezium by vaulting the bays alternately in squares and triangles. 4 - 
 
 * Sometimes portions of the boarding" stuck to the layer of mortar on their upper surface, and 
 were left there. Fragments of the original centering may still be seen adhering to the mortar 
 in Lastingham crypt. 
 
 + This had been done in the sixth century at S. Vitale, Ravenna ; in the ninth at Aix-la- 
 Chapelle, and in S. Stefano, Yerona. 
 
Chapter XX. 
 CONSTRUCTION OF RIBBED VAULTS. 
 
 Vaults with Diagonal Ribs — Use of Ribs — Early Diagonal Ribs — Breadth of Ribs — 
 Wall Ribs — Rib-molds — Voussoirs — Tas-de- Charge — Materials of Cells — Thickness 
 of Web — Keystone, Boss — Pendant — Height of Spring. 
 
 We have seen how very valuable the groined vault proved to be ; nevertheless 
 it was very heavy, and needed massive and obstructive piers, and the groins were 
 difficult to construct. So there was devised an immense improvement, which 
 superseded the groined vault after a life of twelve centuries, and completely 
 revolutionised the history of mediaeval architecture. It was the substitution of 
 ribs for the sharp edges of the groins. It had many good results ; some of them, 
 no doubt, unforeseen by the inventors. First of all, it got rid of the special 
 defect of the groined vault, the weakness of the diagonal groins.* But what 
 was probably still more at the heart of the builders — it enabled them to con- 
 struct the filling in with much less centering. Instead of building up the whole 
 vault first of all in wood, a difficult, tedious, and expensive matter, the mason now 
 first put up centres for the two diagonal arches and the two transverse arches, 
 and built these in ashlar.f When the masonry of these arches was set, the 
 four compartments ("cells") into which the vaulting bay ("severy") was now 
 divided, could be vaulted one at a time. Instead of having to prepare 
 centering for all four cells at once, he could employ it for one cell, take it down 
 and employ it for another, and so on with the rest. And the centering required 
 for vaulting each cell was of the simplest character, merely a set of planks of 
 short lengths. When with these he had bridged the space from one of the 
 diagonal ribs to one of the four outer arches, he could then fill in with rubble 
 held together by mortar. But a further improvement was possible. It was 
 that two planks should be so fastened together, that one would slide on 
 the other; forming a sort of extensible plank or templet (" cerce"). % Thus 
 the cerce, drawn out as it could be to any length required, might supersede 
 planking altogether. No centering would be required at all for the " filling in " 
 of the cells ; only for the construction of the arches (diagonals, transverse arch, 
 and pier-arch) on which the cells rested. In those days, before steam and before 
 circular saws, it was an important consideration to be able to economise in 
 planks. 
 
 * Cf. Bmtails, 98. t Ashlar is squared and dressed building stone. 
 
 I The " cerce " is illustrated on page 298. 
 
2y7 
 
 Gloucester West Nave. 
 Oxford Cathedral Choir. 
 
 Exeter Choir Aisle. 
 Tewkesbury Nave. 
 
298 
 
 ADVANTAGE OF RIBS. 
 
 Another thing greatly desired by the Gothic builder was to lighten the vault, 
 and to reduce the area of the thick piers with which the Norman naves were 
 encumbered. This led the way to the light piers of Gothic, occupying sometimes 
 only one-nineteenth of the area of such piers as those in DURHAM CHOIR (659.1). 
 To effect this, all the mason had to do was to substitute a thin shell of worked 
 stone for the thick mass of concrete of the groined vault. Indeed, if he wished 
 to use the " cerce," * the vault could not be constructed at all in heavy concrete ; 
 the " cerce " would not bear it. So on the upper surface of the " cerce," which 
 had a segmental curve, a segmental arch was constructed, not in rubble, but in 
 ashlar. Then the " cerce " was closed up, taken down, extended, set up again 
 (one end resting on a diagonal and the other on an outer arch), and a second 
 segmental arch was built on the first, and so on till the ridge of the cell was 
 reached. Then, starting near the bottom t of the other half of the cell, the mason 
 
 Vaulting'. 
 
 worked upward till again he reached the ridge. Such work could only be done 
 with worked blocks, not with rubble. 
 
 Another great advantage of ribs was that in a severy of several cells, each 
 cell was independent of its neighbour. In the groined vault, if fracture occurred 
 in any part of it, the safety of the whole vault was imperilled. But any cell in 
 a ribbed vault might be fractured without necessary injury to the adjoining cells. 
 Indeed, sometimes one cell was taken out, as in the high vault of SENS (107), 
 and a cell of different shape substituted for it, in order to get room for a larger 
 clerestory window. The cells of the vault of the south aisle of Chichester nave 
 are similarly elevated to admit tall windows (see 34.4). 
 
 Still another advantage was that there was a very considerable amount of 
 elasticity in the shell of such a vault. If, owing to pressures and strains, the 
 
 * We have no actual evidence that such a "cerce" was employed in mediaeval construc- 
 tion : in large spans certainly it could not be used. Choisy's Histoire, ii. 274. 
 
 + The bottom courses, e.g. those shown in fig. 4, being vertical or nearly so, could be 
 built without either "cerce" or centering. 
 
BREADTH OF RIBS. 299 
 
 diagonal arches of the vault were thrust apart a little,* the arched courses would 
 sink ; if the diagonals were forced nearer together, the courses would rise ; and 
 all this without necessarily causing the vault to fall. In some French churches, 
 e.g. in the collegiate church of St Ouentin, the twisting and contortion 
 of the vault surfaces is something extraordinary ; the vaults look most un- 
 pleasantly unsafe ; yet they have stood since the thirteenth century. This 
 elasticity was again further increased by the fact that the end vaulting stones 
 (voutains) of the courses were not bonded into the arches on which they 
 rested. At first the courses of voutains passed over and rested on the backs 
 of the arches. In later work the arches were usually sunk (" rebated "J to 
 receive the voutain, or at least the lower part of its thickness.! Whichever 
 method was adopted, the cells of the vault had a good deal of " play." Some- 
 times, however, even in late vaults, the construction of the groined vault was 
 retained, i.e. the cells were still built in rubble,^ e.g. in Lichfield nave. When 
 this was the case, the under surface of the vault was plastered. In modern 
 restorations this plaster has often been removed, as at Lichfield. But a vault 
 so constructed has little elasticity ; and needs to be strengthened by other ribs 
 in addition to the diagonals ; it is so strengthened at LICHFIELD (313). 
 
 Again, in a vault without ribs the eye has nothing to rest on but the 
 webs ; and naturally, looking to them for beaut)- of curve, demands that their 
 curves shall be true ; that the surfaces shall be portions of regular cylinders ; and 
 notes with displeasure any deviation from regularity of surface. But in a ribbed 
 vault it is to the ribs that the gaze is directed, not at all to the surfaces ; unless 
 they are exceptionally distorted, as at St Ouentin. § So long, therefore, as the 
 ribs present agreeable curves to the eye, the builder may " play " with his 
 surfaces ; and in Gothic he does so to a very large extent. Many difficulties 
 of Gothic vault construction are got rid of by a change of the direction of the 
 vault surface, which escapes notice unless one goes out of the way to look for it. 
 Thus, then, not in one but in many ways, the substitution of the rib for the groin 
 was an immense advance. 
 
 BREADTH OF Ribs. — So much for the value of the rib. Later on, its early 
 and simple forms received an astonishing development. But before we go on to 
 speak of that, it may be well to describe more fully the construction of a bay 
 or severy of an earl}- ribbed vault. In its simplest complete form it consists 
 of two diagonal arches, AD, BC, and four outer arches, AB, BD, DC, CA. Of the 
 outer arches, BD is a pier-arch, the most massive of all. AB, CD, are transverse 
 arches spanning the aisle. AC is the wall-arch (French formeref), with a window 
 below it. Compare 30S.7, 298.3, and 313 (GLOUCESTER). 
 
 In earl}- examples, (•.£-. in Canterbury choir, the transverse arches are usually 
 much broader than the diagonals. This is merely a " survival." In the groined 
 vault, as there were no diagonal arches, the whole of the weight of the vault was 
 
 * E.g. Hereford north transept. 
 + Scott's Lectures, ii. 194. 
 
 J Chichester and New Shoreham have early and beautiful examples of vaults accurately 
 filled in with ashlar (313). 
 
 § E.g. no one but an expert ever notices the frequent ploughshare vaulting in a Gothic 
 lerestory (see page 311). 
 
3 oo RIB-MOLDS. 
 
 taken by the outer arches, which consequently had to be broad and strong, as in 
 NORWICH AISLE (238). But in a ribbed vault the diagonals carry most of the 
 weight* This at first was not realised, e.g. in the aisle vault of DURHAM NAVE 
 (315); hence the unnecessary breadth of many early transverse arches. Later 
 on, both sets of arches are made of the same section. An early example of 
 the improved method is the aisle vault of the western bays of Worcester 
 nave t (c. 11 70). 
 
 Wall Ribs. — Wall ribs are frequently absent in the early Gothic, both of 
 France and England. The eye distinctly requires them ; and the absence of 
 them is an artistic loss. Examples of wall ribs occur in the aisles of Aix- 
 la-Chapelle and S. Ambrogio, Milan ; the Norman groined aisles of Glou- 
 cester choir and Norwich ; the Norman ribbed aisle of Gloucester nave ; in 
 Glastonbury Lady Chapel; Wells Cathedral; NEW SHOREHAM and CHICHESTER 
 
 (313)- 
 
 Early Diagonal Ribs. — The ribbed vaults of S. Ambrogio, Milan, are of 
 uncertain date. The damaged vault of the nave was repaired in 1196. Cattaneo 
 and Rivoira suggests. 1050 — c. 1100 for the construction of these vaults ; Comte 
 Robert de Lasteyrie puts them after 1 100. In France early examples of diagonal 
 ribs occur at Quimperle, Poitiers, St Gaudens, Moissac, St Guilhem du Desert, 
 St Gilles, Marseilles St Victor ; chiefly under towers. The chief early examples 
 in the He de France and Picardy are St Denis, 1140 ; Morienval, which may be 
 either before or after 1122 (Brutails, 213, 215); the ambulatory of St Martin- 
 des-Champs, c. n 30 to c. 1150; the aisle of St Etienne, Beauvais, c. 11 15 to 
 c. 1125 ; Bellefontaine, after 1125. Enlart {Manuel, 440) is of opinion that in 
 1 120 diagonal ribs had been in use but a short time in these districts ; this is 
 shown by "the clumsiness with which they were still employed." In England, 
 diagonal ribs appear in the eleventh century, viz. in the aisles of Durham choir, 
 which was built 1093 to I0 99 (Bilson's Beginnings, 261 to 264). Early examples 
 of Gothic diagonal ribs are seen in the two western bays of the south aisle of 
 Worcester nave, c. 11 70, where the wall arches and transverse arches are pointed 
 and are composed of a double roll; in Glastonbury Lady Chapel, 11S4-1186, 
 where also the wall arches and transverse arches are pointed ; as they are in the 
 round church of the Temple, dedicated 1185 ; and in the transept and nave of 
 Wells Cathedral. Earlier than any of these are the pointed arches of the vaults 
 of St Cross, Winchester.^; 
 
 Rib-molds. — As we have seen, e.g. at Lessay, Norwich, and Gloucester 
 choir, the earliest transverse arches were unmolded. But when diagonal ribs 
 were introduced in Durham choir aisles {c. 1093), pier-arches were often molded ; 
 indeed, in some cases, e.g. in the arches of Norwich apse and choir (1096) 
 ( 2 73). tne moldings were already numerous, delicate, and refined. Therefore, 
 except now and then in crypts, e.g. CANTERBURY (193), or under towers, 
 unmolded diagonals do not occur. Nevertheless, for a long time, even 
 in Gothic, the rectangular outline of the ancient unmolded arches was pre- 
 
 * Thus on page 675 the transverse and diagonal arches of ROCHE (1, 2) differ in section ; 
 but those of byland (3) are the same. 
 
 t Prior, 91, adds Hereford east transept, Chester, and Llandaff. 
 X See quotation from Viollet-le-Duc, Architecture, 440 note. 
 
VOUSSOIRS. 301 
 
 served ; all the moldings being obtained by sinking hollows into ribs of 
 rectangular section. Of the early molded ribs three species are most common. 
 The first is a single roll ; e.g. in the transverse ribs of KIRKSTALL NAVE (673. 5 ) 
 and the octopartite vault of CANTERBURY TREASURY (673.2), The second is a 
 pair of rolls; as in the transverse arches of BUILDWAS CHAPTER HOUSE 
 (673.10), GLOUCESTER (3 13) and Southwell nave aisles; and in Gothic, at 
 Wells, Salisbury, Chichester, NEW SHOREHAM (313), and Lincoln. The hollow 
 between these rolls is variously treated. A third arrangement is to set between 
 the pair of rolls a third projecting roll ; as at the ABBAYE-AUX-iiommes (673.4) 
 and BUILDWAS (673. 1 1). 
 
 In all the above, and other such variants, the moldings are all inscribed in 
 a rectangle. But, after a time, the rectangular was abandoned for the triangular 
 form ; as soon as it was seen that, for strength of rib, depth is more valuable 
 than breadth ("just as one lays iron girders on the flange, not on their sidej. In 
 other words, instead of molding the faces and the soffit, the masons molded 
 the chamfer planes ; having cut the voussoirs of the ribs into a triangular shape 
 before beginning to mold them. A good example of this is seen in LINCOLN 
 GALILEE (677.3;. I' 1 l at;e vaulting the principle was sometimes carried further 
 still, and triangular became what we may call " knife-edge " ribs, as in the 
 lower story of the Old Convocation House at Oxford ; Bishop Booth's Porch, 
 Hereford ; and Bishop West's Chantry at ELY (334). 
 
 Another reason for altering the quadrangular to the triangular form was 
 that three, five, or more ribs had to descend to the same abacus ; but if each rib 
 was square, it was difficult to find room for them all. 
 
 Other alterations were mainly to gain various shadow effects. As in 
 the diagonal rib of FURNESS (673.8;, the rib became pointed in section ; or a 
 keel-molding was employed, as at HOWDEN (677.10); or a narrow fillet, as at 
 RIEVAULX (675.14) ; or a broad fillet, as at JERVAULX (675.5). Sometimes the 
 roll (or bowtelE) had a double fillet, as at RIEVAULX (675.14); or even a triple 
 fillet. Finally, all this alternation of roll and hollow was run into meagre 
 undulatory curves and hollow chamfers, as in WELLS CLOSE (677.14). 
 
 It should be added, that as only one side of the wall ribs supports the webs, 
 it is made smaller than the diagonal and transverse ribs ; e.g. in the Norman 
 aisle of GLOUCESTER NAVE (313) ; and at BYLAND, RIEVAULX, WHITBY (<'>J3). 
 
 Occasionally molded ribs are ornamented ; e.g. in St Cross, Winchester, 
 with zigzag; in St Peter in the East, Oxford, with chain ornament; 
 in Lincoln south-east transept, with billet; in Canterbury choir and Lincoln 
 galilee, with the tooth ornament; in the south aisle of Gloucester nave, with 
 ball-flower. 
 
 VoussoikS. — In early examples the ribs are composed of a large number 
 of small blocks (voussoirs) ; in later work much longer blocks are employed. 
 The length of the voussoirs may therefore be of value in determining the 
 date of a vault. Thus the photographs of the vaults of the choir aisle of 
 DURHAM, the choir of LESSAY, and the nave of the ABBAYE-AUX-HOMMES 
 (315) show that Norman ribs are composed of quite short blocks. In the 
 north-east transept of Canterbury {c. 1175 :, which is 30 ft. wide, the transverse 
 rib is made up of about a hundred voussoirs. In Westminster transept (c. 1245), 
 
302 
 
 TAS-DE-CHARGE. 
 
 the transverse rib above the solid springer contains thirteen or fourteen vous- 
 soirs ; in the nave, which is much later, six only.* 
 
 TAS-DE-CHARGE. — Turning to the diagram 308.3, and the photograph of 
 the vault of Chichester nave on page 313, it will be seen that five ribs — two 
 diagonals CJ, GJ, one transverse arch DJ, and two walls ribs IJ and NJ — may 
 all descend on to one capital or corbel J. So also five ma)' descend on to 
 D. But in 324.7 there are no less than thirteen ribs, all trying to find room 
 on the capitals B and D. It was impossible to find room for them all ; and 
 
 so, in Gothic days, 
 the lower parts 
 {springers) of the 
 groups of ribs were 
 constructed in hori- 
 zontal courses bonded 
 to the wall : i.e. the 
 courses of the lower 
 part, the springers of 
 the arches, were not 
 arcuated at all ; had 
 not radiating joints : 
 the arch proper only 
 began where the 
 springers ended. 
 Such an arch is 
 said to be built in 
 TAS-DE-CHARGE: the 
 lower part of its 
 curve is merely cor- 
 belled out ; it con- 
 sists below of courses 
 of horizontal blocks, 
 each upper course 
 projecting further for- 
 ward than the course 
 below it. It may ex- 
 tend upward to as 
 much as one-third of 
 the height of the vault. 
 Many great advantages sprang from the construction of the springers in the 
 solid. The weight from above was brought down on to horizontal blocks which 
 could not slip, instead of being brought down to inclined voussoirs which might 
 be forced out. In diagram 1 there are four horizontal springers (French 
 sommiers) on which the weight descends. In diagram 2 there are three 
 springers, A, B, C ; in c radiating joints are seen from which three ribs are 
 to spring — a broad transverse rib in the centre, a slender diagonal rib on either 
 side. D shows the first voussoir of the transverse rib ; D 1 that of either of 
 
 * Willis' Middle Ages, 6, 9. 
 
 1, 2. Tas-de-Charge. 
 3. Keystone. 4. Pendant. 
 
MATERIALS OF WEB. 
 
 303 
 
 the diagonals. Really tas-de-ckarge means continuing the pier so much upward; 
 e.g. in diagram 2 the pier does not really terminate as it appears to do 
 at the capital beneath the sommier A, but at the top of the skewback C* By 
 a logical corollary, in later French Gothic, eg. at St Lo, the capitals of the main 
 piers are omitted altogether, as telling a false tale. 
 
 Another happy result is that since the haunches of the arches are executed 
 in corbelling, the span of the real arches with radiating joints is much reduced ; 
 and as it costs less to construct a narrow than a broad arch, owing to the 
 reduction in centering, tas-de-charge effects considerable economy. hi the 
 transept of Westminster the real span of the vault arches is thus reduced by 
 about one-sixth. t 
 
 As to the mode of execution, the easiest way would be to erect the springers 
 in block, and then when the arches were built on them and had set, to work 
 moldings downwards to correspond with the forms of the ribs above. % This 
 may well have been a common practice. Professor Willis, § however, proves that 
 the molds of the springers were not worked in situ in the nave aisles of St 
 Saviour's, Southwark, or in a late vault beneath Lanfranc's tower at Canterbury, 
 both of which he saw in the course of demolition. In these the surfaces of the 
 springers were covered with lines and profiles of moldings to enable the blocks 
 to be cut to shape on the bench. Possibly the practice varied. 
 
 The use of tas-de-charge does not occur in England till towards the end of 
 the twelfth century ; e.g. at Glastonbury, and in the central pillar of the guard- 
 room of the Castle at Newcastle. In France it does not appear till the 
 thirteenth century; Soissons choir, finished in 1212, seems to be the first 
 example of it ; it does not appear in Notre Dame, Paris, finished in 1220.T 
 
 Materials. — The cells of the vaults were constructed of rubble or of ashlar ; 
 of a light stone if it could be had. Thus at Rome there were deposits of pumice 
 and tufa from the neighbouring volcanoes ; these were used largely ; e.g. in the 
 vaults of the Colosseum, and the Baths of Titus and Caracalla.**" In the 
 haunches of the vault, e.g. in the Temple of Minerva Medica, and the vaults of 
 tombs, pots were often used. They were light, and they could be had for the 
 fetching, for there was a mountainous dust-heap of old wine jars at Rome — the 
 present Monte Testaccio. In England, when we could get it, e.g. in Kent and 
 Gloucestershire, we used an aqueous tufa, ft a deposit of so-called petrifying 
 
 * Curiously enough, the real termination of the vaulting shaft in the clerestory of York- 
 presbytery is marked by a couple of little capitals set on either side of the top of the solid 
 springers. (See int. elevation in Britton's York, Plate 24.) 
 
 t Willis' Middle Ages, 7. In Mr Pearson's church at West Croydon, the span is thus 
 reduced from 22 ft. to 19 ft. (R. P. Spiers, Architectural Drawing). 
 
 % Scott's Essay, 180. § Vaulting, 10, 11, 12. 
 
 || Illustrated in Willis' Glastonbury, 66. 
 
 1 Enlart's Manuel, 506 ; and Choisy's Histoire, ii. 272, 274. 
 
 ** Choisy's L 'Art de batir chcz les Remains, 96. 
 
 tt At Worcester the Lady Chapel is vaulted with rough lumps of tufa ; so also the two 
 western bays of the nave. This tufa is found and is still used occasionally in the Stockton-on- 
 Teme neighbourhood, some 15 miles to the west of Worcester. The rock is excessively soft 
 when quarried, but hardens in the air. The Normans were fond of it in the West of England ; 
 the church of Moccas, Hereford, is largely built of it (Mr A. B. Pickney). 
 
3 04 THICKNESS OF WEB. 
 
 springs issuing from limestone. It was not only light and tough, but gave a 
 good hold to the mortar. The vault webs of Bredon porch and of Sherborne 
 are or were of tufa ; and the vaults of Canterbury choir are, says Gervase, "ex 
 lapide et tofo kvi." The webs of the vaults of Salisbury and of Canterbury 
 cloisters and St Augustine's gateway* Canterbury, are of chalk ; those of the 
 fourteenth-century nave of Beverley are of brick. t 
 
 Thickness of Web. — Another way to reduce the weight of the vaults 
 was to construct the webs in thin shells of cut stone or brick. Thus while the 
 original rubble vault of the Abbaye-aux-Dames, taken down in the eighteenth 
 century, was 12 in. thick, the vault of hollow tiles and Portland cement put 
 up recently by Ruprich- Robert is only 6 in. thick. | Those of Chartres are 
 nearly 1 1 in. thick : the result is seen in the extraordinary massiveness of the 
 walls, buttresses, and flying buttresses of that great cathedral: a vast and 
 unnecessary addition of expense. Many French Gothic vaults over naves, even 
 in the twelfth century, are not more than 4 to 5 in. thick ; that of Xotre Dame, 
 Paris, an early example, is 6 in. The panels of the vault of Henry the Seventh's 
 Chapel at Westminster were found by Mr Wonnacott to vary from 4-i to 5 inches 
 in thickness. 
 
 As regards the question of thickness, the builders show a certain amount of 
 hesitation. The logic of the matter was, that if the vault webs were constructed 
 in rubble, they ought to be thick. Again, if they were to be covered externally 
 with tiles or with flagstones resting directly on their backs (extrados), they 
 ought to be thick ; so also, if they were to be sheltered with a timber roof, but 
 one resting on their extrados. § But if the timber roof was to be quite inde- 
 pendent of the vault, not touching it at all, then the thinner the vaults could be 
 made, compatibly with stability, the better : a thin vault fireproofed the church 
 just as well as a thick one. In practice, however, the builders were not always 
 logical. They frequently added to the webs, on their extrados, a thick irregular 
 course of rubble work, often covered with concrete ; e.g. in Winchester, Wells, 
 Ely, Bristol Cathedral and St Man- Redcliffe, most of Westminster and 
 Hereford, Sherborne crossing, Oxford Cathedral, and elsewhere. But in the 
 west nave of Westminster and in the south transept and tower of Hereford 
 the upper surfaces have no such covering ; and at Ely it was lately hacked 
 away. In late work not only is this outer covering dispensed with, but a good 
 deal of the surface of the blocks of the cells themselves was chipped away ; 
 e.g. in the high vault of the nave of St George's, Windsor ; || while in Henry the 
 Seventh's Chapel at Westminster the whole of the outer surface is chipped off 
 (347)- Probably it was necessary to lighten the later vaults as much as 
 possible, for they were far flatter than the earlier ones, and had much greater 
 lateral thrust. The more the weight could be diminished, the less would be 
 the thrust. 
 
 KEYSTONE, BOSS. — In an arch even" one of the voussoirs of which it is 
 composed is a keystone ; but the central voussoir, that at the apex of the arch, 
 
 * Carder's Specimens, 36. + Mr John Bilson in Archit. Review, xviii. 252. 
 
 I The high vaults (brick) of the nave of S. Ambrogio average 16 in. in thickness (Mr C. A. 
 Cumming's Architecture in Italy, i. 105). 
 
 § Choisy's Histoire, ii. 162. Illustrated on page 341. See Willis' Vaulting, 9. 
 
THE BOSS. 305 
 
 is generally called the keystone (French clef). In a pointed arch this central 
 keystone is sometimes found, but usually the junction of the two curves of the 
 arch is marked by a vertical joint. When diagonal arches, crossing one another, 
 as in DURHAM AISLE (3 IS), were first introduced, a difficulty arose because each 
 diagonal could not have a keystone of its own. In some early examples, e.g. in 
 the crypt of Gloucester, the eastern apse of the south transept of Tewkesbury, 
 and Upton, Bucks, first one diagonal arch was set up complete ; then the other 
 in two halves, abutting awkwardly and obliquely against the first. Durham 
 has keystones of unscientific form.* But in the repair of Winchester transept 
 vaults after the fall of the central tower in 1107, there is one common key- 
 stone, which is " shouldered to receive the ribs, which abut against them with a 
 joint at right angles to the direction of the rib." t Sculptured Norman key- 
 stones occur beneath Iffley tower and in the semi-dome of Kilpeck. In the 
 eastern transept of Canterbury^ (V. 1180J there are bosses; but their arms 
 are straight instead of being curved like the ribs. 
 
 The builders soon gave the boss increased dimensions. In Winchester 
 nave there are keystones weighing about 2 tons.§ In Heme vaulting (340) 
 bosses are especially necessary ; for it is very difficult for the moldings of the 
 obliquely set ribs of such a vault to form true intersections, or as it is termed 
 "to mitre." Nevertheless the builders occasionally amused themselves by con- 
 structing elaborate " mitred " vaults without bosses ; e.g. in the south transept 
 
 of GLOUCESTER (306), and WEST'S CHANTRY, ELY (334). 
 
 Other uses also were found for bosses. Unusually large bosses with 
 sculptured legends were placed over the spot where was the high altar or the 
 shrine of some great saint ; and from the boss we may be able to ascertain where 
 formerly that shrine or altar stood ; e.g. in Ely a very elaborate boss, repre- 
 senting the Coronation of the Virgin, points to the place where St Etheldreda's 
 shrine stood, behind the high altar. Foliated bosses of the rarest beauty were 
 common ; albeit all their detail invisible except to a modern binocular ; foliage 
 changing from conventional to naturalistic, as at EXETER (297) ; from natura- 
 listic to undulatory leafage, as in the BERKELEY CHAPEL (329), Bristol 
 Cathedral ; and finally to the square leaves of Perpendicular work. Or, 
 in later work, coats of arms are on the bosses, and help, as in Winchester 
 nave, to tell who provided the funds for the construction of the vault. In 
 later work the carved boss and either the whole or the adjoining portions of 
 the ribs were often painted and gilt. 
 
 Through the centre of the boss a hole was often pierced, through which a 
 chain was passed. From the chain was suspended a lamp, or a censer, which 
 on certain days was swung backward and forward, sending forth a " cloud " of 
 incense. 
 
 Pendant. — In Tudor work the boss is often greatly elongated, and is then 
 called a pendant (302.4). In a transverse arch, as in the Divinity School 
 and the high vault of OXFORD CATHEDRAL choir (297, 331), more than one 
 pendant may occur. Or between two adjacent pendants skeleton pointed 
 
 * Bilson's Beginnings, 299. t Bilson's Beginnings, 301. 
 
 I A boss is a keystone bossed out for carving ; a keystone not so decorated is not a boss. 
 
 § Colson's Winchester, 22. 
 
 U 
 
3o6 
 
 THE PENDANT. 
 
 arches may be built up ; with the result that, to the untrained eye, the vault 
 appears to rest on the arches, and the arches on the end of the pendant, and 
 the pendant on nothing at all (302.4). In Henry the Seventh's Chapel at 
 Westminster there are two pendants in each transverse arch of the vault (347). 
 It is the fashion to abuse these as unconstructional ; but as a matter of fact 
 each is an elongated voussoir of a transverse arch ; and each is the founda- 
 tion on and round which is built up with marvellous skill one of the inverted 
 conoids or trumpets of the fan vault. 
 
 Height of Spring. — Much diversity exists as to the height at which the 
 spring of the vault commences. Constructional and artistic considerations pull 
 in opposite directions. The lower the spring, the safer the vault is, and the 
 
 Durham, S. Aisle of Choir. 
 
 Gloucester, S. Transept. 
 
 worse it looks. On the other hand, if the spring is very high up, the clerestory 
 windows are unmasked ; the vault itself is well lighted, its beauty can be seen ; 
 all heaviness disappears ; as in GLOUCESTER CHOIR (59), it seems to float on the 
 illumined air. But a vault set on narrow piers half-way up such vast clerestory 
 windows as those of Gloucester and Norwich requires strong flying buttresses to 
 take its thrusts; unless, as in SHERBORNE NAVE (346), it is a fan vault of pointed 
 section ; or as at GLOUCESTER, it is abutted by a half barrel vault below (35.5). 
 
 In TEWKESBURY NAVE (297) the vault actually springs from the abacus of 
 the nave piers ; the result is that the clerestory windows are almost wholly 
 blocked, and the nave is gloomy in the extreme. In the monks' vault over the 
 eastern bays of GLOUCESTER NAVE, c. 1220 (26), the vault springs from the sill 
 of the triforium. In the Norman vaults of Durham choir, nave, and central 
 
HEIGHT OF SPRINGING OF VAULTS. 307 
 
 transept; in the choir vaults of CANTERBURY (106); in the thirteenth-century 
 vaults of Lincoln choir, nave, central transept, presbyter)' ; and in Hythe 
 chancel ; and in the fourteenth-century vaults of Pershore and Canterbury 
 nave, the vault springs about midway in the triforium. At the end of the 
 twelfth century in New Shoreham and Christ Church, Dublin ; in the thir- 
 teenth century in St Saviour's Southwark, Salisbury, Worcester choir, Hereford 
 choir and north transept, ELY presbytery (117) and choir, Lichfield nave, and 
 Exeter ; and in the fourteenth century in Winchester nave, the vault springs 
 from the sill of the clerestory.* Finally, the vault may spring from somewhere 
 midway in the clerestory ; e.g. late in the twelfth century at Wells and 
 Chichester; in the thirteenth century at Rochester, BEVERLEY (51), West- 
 minster ; in the fourteenth century in Gloucester choir, and in its imitators 
 Sherborne and Norwich. Probably no part of mediaeval design was, from an 
 artistic point of view, of such overwhelming importance as the combination of 
 towering clerestory and high-set vault. 
 
 * So it does at BOXGROVE (318), but is much stilted. 
 
3 o8 
 
 A C 
 
 F H 
 
 M M 
 
 A C 
 
 Q 
 
 Ye 
 
 /x 1 
 
 Yj 
 
 B D ~ G I 
 
 /v 1 
 
 D \/ 
 
 \vi 
 
 \<H 
 
 / V* J 
 
 /x 
 
 © 
 
 1. Peterborough Nave. 
 
 2. Durham Nave. 
 
 3. Lessay Choir. 
 
 4. Boxgrove Choir. 
 
 5. Lincoln Great Transept. 
 
 
 '© 
 
 6. Florence Cathedral Nave. 
 
 7. Angers Cathedral Nave. 
 
 8. Lincoln Aisle of Nave. 
 
 9. Lincoln South-west Chapel. 
 10. Oxford Proscholium. 
 
Chapter XXI. 
 VAULTS WITH DIAGONAL RIBS. 
 
 Part I. 
 
 Domed and Undomed Vaults — Vaults of Square and < (IjIoiil; Scveries. 
 
 DOMED AND UNDOMED VAULTING. — So much for the general principles 
 of vault construction. We will now turn to their application in English and 
 French hands : beginning with the groined vaults constructed by the Romans 
 in the aisles of S. Sophia, Constantinople, and in S. Vitale, Ravenna ; which may 
 be called Byzantine groined vaults. These are not composed, like the normal 
 vault of Rome (see 291) of the intersection of two half cylinders, or two tunnel- 
 vaulted corridors, with level ridges. In their construction the groins as well as 
 the four outer arches are semicircular. The result is that since the semicircular 
 groins are wider than the outer 
 arches, they are also propor- 
 tionately loftier ; and conse- 
 quently, the ridges of the vault 
 are not level ; the vault is 
 more or less domical. In the 
 first diagram, supposing ABDC 
 to be a square space covered 
 with a vault, of which the 
 groins AED and BEC are semi- 
 circles, and the outer arches AGB, 
 BHD, Die, CFA, are also semi- 
 circles, it follows that since the 
 diagonal of any square is longer 
 than its sides, E, which is the 
 
 apex of the two groins, must be higher than G, H, I, F, the apices of the four outer 
 arches. Compare the arch on AH or on any one of the sides of the square with 
 the arch on the diagonal AD or BC. It was this type of groined vault that was 
 usually adopted in Lombard)-, German}-, Burgundy, and Northern France, 
 but seldom in Normandy and England. It was constructed at first with 
 diagonal groins ; later, with diagonal ribs. It is a very good form of vault, 
 for it has the merits of being stable and of being easy to construct. It is 
 easy to construct ; for when all its arches are semicircles, they can be executed 
 without full-sized drawings on the ground ; so even when some of them were 
 
 Forms of Seveiies and Arches. 
 
3 io DOMED VAULTS. 
 
 pointed, the diagonals usually remained semicircular. Secondly, the more 
 domical the cells are, the more support each course of the cell gets from the 
 course below it. The flatter a vault surface is, the greater the difficulty in 
 constructing it. Thirdly, just as a pointed arch has much less lateral thrust, 
 and therefore requires less abutment than a semicircular or a segmental arch, 
 so the more domical a vault is, the less is its thrust, and the less the abutment 
 required. These are great advantages. One disadvantage is that since a 
 domical vault is loftier than an undomed vault, the aisle or nave walls on either 
 side of it have to be raised in proportion, and much time and material are 
 wasted. This objection may have weighed with the Norman and English 
 builders, and have caused their dislike of domical vaulting. We have indeed 
 examples ; e.g. at Xew Shoreham and the east side of Westminster cloisters ; 
 but they are not common. In France, east of Normandy, the early Gothic 
 vaults, e.g. at St Denis, are usually excessively domical ; later on, they are 
 less so, but their ridges seldom become quite horizontal. As early as the 
 twelfth century the diagonal as well as the outer arches were pointed in the 
 aisles of St Cross, Winchester, in St Joseph's Chapel, Glastonbury, and in 
 Wells Cathedral ; the result being to make the vault domical, and at the same 
 time to bring the thrusts down more vertically. Domical vaults of different 
 types are shown in 325.6 and 321. 
 
 SQUARE SEVERIES. — In Normandy and England the type usually adopted 
 was that of the vault with horizontal ridges (see 291). Now if this vault be studied, 
 it will be seen that all the arches — the two diagonals, and the four outer arches — 
 all start from the same level, and rise to the same level ; and therefore both the 
 ridges are horizontal. It follows that the diagonal arches are not semicircles. 
 For we saw in 309.1, that if they were semicircles, the}- would rise much 
 higher than the other arches, and the ridges would not be horizontal. Since the 
 ridges are horizontal, it follows that the arches are of some other form than the 
 semicircular. As a matter of fact they are half ellipses. So that by adopting 
 the undomed rather than the domical type of groined vault, the builders 
 had committed themselves to the adoption of the elliptical form of arch 
 for the diagonals. But an ellipse is difficult to draw and difficult to 
 execute. Probably in the eleventh century the builders could not draw an 
 ellipse at all. Moreover, it is more difficult still to construct diagonal ribs of 
 elliptical curve. So semi-elliptical diagonals are rare; the}' seem not to occur 
 till the middle of the twelfth century. Examples* are Devizes St Mary and 
 Dunstable Priory;! both probably c. 11 50; in Normandy the transepts of the 
 Abbaye-aux-Dames and of Montivilliers ; t and Gueron apse; in the thirteenth 
 century the choir of the Temple Church, London ; § the south and west sides of 
 Westminster cloister and the passage leading into Dean's Yard. Generally, 
 however, the semi-ellipse was avoided. Some easier way had to be found out of 
 the difficult}-. The difficulty was that the diagonals were too tall and the 
 transverse arches too low for horizontal ridges. Evidently the thing to do was 
 to lower the tall diagonals or to raise the low outer arches, or to do both. This 
 
 * It should be noted that some of these occur in high vaults, not in vaulted aisles. 
 
 t Bilson's Begirmings, 308. % Ruprich-Robert, Plates 74 and 95. 
 
 § Illustrated in Scott's Lectwes, i. 119. |] Garbett, 166, 167. 
 
SQUARE AND OBLONG SEVERIES. 
 
 3ii 
 
 is precisely what was done in Norman England. For the semicircle in the 
 diagonals the builders substituted the upper part of a semicircle ; i.e. a seg- 
 mental arch ; the outer arches they raised on stilts. Segmental diagonals occur in 
 the aisles of Durham, Winchester, Peterborough, Gloucester nave, Southwell, 
 Lindisfarne, Selby, Warkworth ; * all probably between 1093 and l ' 5°- While 
 beneath the transverse arches a few vertical courses of masonry were constructed ; 
 i.e. the arches were stilted; in the aisles of Durham, PETERBOROUGH (318;, 
 Romsey, Southwell, Lindisfarne, and others. Sometimes the "stilt" is not vertical, 
 but incurved ; making the arch more of a " horseshoe " than a stilted semicircle. 
 
 But all this was to break two of the principles of vault construction ; one, 
 that all the ribs of a vault should start from the same level, which the outer ribs 
 no longer did ; and secondly, that all the arches must be semicircular, which 
 the diagonals no longer were. The penalty was that the 
 webs would be no longer regularly cylindrical surfaces,! 
 but would present irregular curves of surface. But having 
 secured by these devices what they valued most — horizontal 
 ridges — the builders did not trouble about the surface 
 contour of the webs ; they let it take its chance. So 
 convenient in practice was this disregard of geometry 
 found to be that the best Gothic of France and England 
 constantly twisted its vault surfaces where there was the 
 slightest convenience to be gained thereby. This was 
 especially the case with the cell next to a window of an 
 aisle or clerestory : this cell was often wide and very 
 difficult to construct ; but by disregarding the proper 
 curve, and by constructing the lower part of it nearly 
 vertical (298.4), that part of the cell, perhaps one-fourth, 
 could be built without centering, each course resting securely 
 on the course below it. Moreover, by pinching back the 
 webs on either side of the window instead of letting them 
 project forward at once more or less in front of it, the light 
 was less obstructed. In such a case, therefore, French and English alike dis- 
 regarded regularity of web-surface in favour of facility of construction and 
 improvement in lighting. This twisting of the surfaces is called ploughshare 
 vaulting: examples are the high^: vaults of Wells ; ST SAVIOUR'S, SOUTHWARR ; 
 Salisbury ; Westminster. But in Gothic, instead of stilting the ribs, they were 
 often supplied with small upper shafts, from which they sprang, and not from 
 the vaulting shafts ; e.g. sens ( 107). 
 
 Oblong SEVERIES. — But we are anticipating matters. So far we have 
 been speaking as if all vaults covered square spaces, as in 309.1. But 
 oblong spaces also often had to be vaulted over. Then a still further difficulty 
 
 St Saviour's, South- 
 wark. 
 
 * Bilson's Beginnings, 299-307. 
 
 t Willis' Middle Ages, 74, says "the whole difficulty of vaulting a parallelogram with 
 quadripartite vaulting resided in the assumed necessity of preserving the apex of the vaulting 
 cells horizontal, and making their surfaces cylindrical throughout." 
 
 1 It is in the high vaults that this occurs most ; because it was more important to clear the 
 clerestory windows of obstructions. 
 
3 i2 HIGH VAULTS WITH DIAGONAL RIBS. 
 
 harassed the builders. Instead of semicircular arches rising to two different 
 levels, as they did over a square, over an oblong they rose to three ; e.g. on the 
 figure ABDC the two diagonals rise higher than the arches on AB and CD, and 
 these again rise higher than those on AC and BD (309.4). So that now the 
 builder had to humour three sets of arches instead of two, if he insisted on having 
 horizontal ridges. Sometimes he was successful. The bays of the ribbed vaults 
 of the Norman aisles of Durham, Southwell, and Kirkstall are all markedly oblong. 
 
 Part II. 
 
 High Vaults with Diagonal Ribs. 
 
 1. Durham — 2. Lessay — 3. Boxgrove — 4. Lincoln Great Transept — 5. Abbaye-aux- 
 Dames — 6. Florence Cathedral — 7. Angers Cathedral. 
 
 As we have seen, here and there, as at St Nicholas, Caen, the builders had 
 succeeded in constructing groined vaults, not merely over the aisles, but on the 
 top of the lofty clerestory walls. What had been successful in the case of groined 
 vaults had still to be attempted with vaults possessing diagonal ribs instead of 
 groins. In the last years of the eleventh century this was the problem throughout 
 Western Europe with all Romanesque builders, except those who had adopted 
 the barrel vault or the dome to roof and fireproof their churches. The problem 
 was solved not in one, but in many ways. 
 
 If we examine the plan of such a nave as that of NORWICH (148.4), it will 
 be seen that if we divide each aisle into squares, the corresponding bays of the 
 nave are necessarily oblongs ; the high vault consists of a number of oblong 
 bays, and, as we have seen above, is consequently difficult to vault. In but too 
 many cases the builders set to work not to remove, but to shirk this difficulty. 
 In PETERBOROUGH (308.1), as the diagram shows, they made things extremely 
 easy by vaulting the aisles in squares, but covering the nave with merely a roof 
 of wood. And that apparently * is what was done in all our large Norman 
 naves except Durham and its copy, Lindisfarne. The naves of Peterborough, 
 Ely, Binham, Carlisle, Chepstow, Chester St John's, Dunfermline, Leominster, 
 London St Bartholomew's, Melbourne, Rochester, Romsey, St Albans, Selby, 
 Shrewsbury, Southwell, Waltham, Wymondham, all had wooden ceilings and no 
 high vaults. Not till Gothic days were high vaults constructed in the Norman 
 naves of Chichester, Christchurch, Gloucester, Hereford, Norwich, Sherborne, 
 Tewkesbury, Winchester. In Normandy also, and apparently throughout 
 Northern France, e.g. St Germain des Pres, Paris, this unworthy solution seems 
 to have been adopted generally ; although, as in the case of the Abbaye-aux- 
 Hommes, Caen, the wooden roofs were often superseded during the course of 
 the twelfth century by stone vaults. 
 
 FIRST SOLUTION. — To this survival of the wooden ceiling in so many of 
 the Romanesque buildings of Normandy and England there seems to have been 
 a conspicuous exception at Durham. Contemporary chroniclers tell us that 
 Durham choir was commenced in 1093, ar >d that in 1099 the work had been 
 * Almost all our greater Norman choirs have been rebuilt or remodelled : some of them, 
 e.g. Gloucester and Hereford, may have been vaulted. 
 
313 
 
 Lichfield W. Nave. 
 Chichester W. Nave. 
 
 < .Inn. ester, North Aisle of Nave. 
 New Shoreham Choir. 
 
3 i4 HIGH VAULTS OF DURHAM. 
 
 extended as far as the nave ; that between 1099 and 11 28 the nave had been 
 built up ad sui usque testudinem ; and, lastly, that the church was finished 
 between 1128 and 1133. It is contended that these dates cover both the aisle 
 vaults and the high vaults of Durham. High vaults of early character 
 (149.1, 8) remain over transepts and nave ; and the marks of the wall-cells of an 
 earlier vault than the present one are plainly visible in the clerestory of the choir. 
 Aisle vaults, still earlier in character (306, 315), remain in choir, transepts, 
 and nave. The vaults of the choir aisles are the earliest, and are clearly part of 
 the work of 1093 to 1099; for the masonry and moldings correspond in char- 
 acter with those of the pier-arcades of the choir ; and, for constructional reasons, 
 the aisle vaults must, as usual, have been erected along with the main walls. The 
 aisle vaults of Durham, therefore, are as early as 1093 to 1099. The next thing is 
 to fix the date of the latest vault, the high vault of the nave. That may be fixed 
 from what we know of the vault of the apse of the chapter house. This build- 
 ing, we are told by the chronicler, was finished between 1133 and 1140. Now, 
 both as to the character of the chevron (zigzag) and the moldings, this vault is of 
 more advanced type than the high vault of the nave. It follows that what was 
 constructed in the cathedral between 1128 and 11 33 was the high vault of the 
 nave.* Moreover, as there is a regular sequence of change of profile and orna- 
 ment in their ribs, we may arrange in chronological order all the vaults that 
 remain, viz. (1) all the aisle vaults ; (2) the high vaults of the north transept; (3) 
 those of the south transept ; (4) those of the nave ; (5) those of the chapter 
 house; all built between 1093 and 1140. Finally, as the original high vault of 
 the choir was no doubt built after the aisle vault of the choir, and before the 
 high vaults of the transepts and nave, it becomes probable that the high vault 
 of Durham choir is but little later than 1099. And as we know that the body 
 of St Cuthbert, the patron saint, was translated in 1104 to the east end of the 
 choir, the probability is that 1104 is the year in which the high vault of the 
 choir, at any rate that portion of it above the shrine, was completed. It is 
 of course to be borne in mind that the plan of the piers (659.1) shows that 
 from the very first, i.e. from 1093, a choir-vault was intended. 
 
 After Durham, none of the aisled Norman churches remaining in England 
 have high vaults built in the first half of the twelfth century, with the exception 
 of the neighbouring church of Lindisfarne, c. 1 125 ; where the whole church was 
 vaulted. So many Norman churches, however, have been destroyed, especially 
 the choirs, which were most likely to be vaulted, that this negative evidence has 
 not much weight. 
 
 But though we cannot substantiate the existence of other high ribbed 
 vaults over aisled buildings, except at Lindisfarne, in the first half of the twelfth 
 century, the)' occur in considerable numbers over unaisled buildings, e.g. Kirk- 
 stall choir ; and the chancels of Warkworth, Hemel Hempstead, Stow, the 
 two churches at Devizes, and others. Moreover, a vault, probably ribbed, was 
 built in the nave of Lincoln between 1141 and 1148. 
 
 In Normandy, however, there is an example of a high ribbed vault, of even 
 more advanced character than those of Durham, and not necessarily any 
 later in date than the high vaults of Durham nave, viz. at Lessay. If a date 
 
 * The chronicler states that Bishop Flambard (who died in 1128), " iiavem usque ad sui 
 testudinem erexerat" ; and that in the next five years " navis peracta est'' 
 
3i! 
 
 Lessay Choir 
 
 Abbaye-aux- Homines Nave. 
 Durham, S. Aisle of Choir. 
 
3i6 HIGH VAULTS OF LESSAY. 
 
 early in the twelfth century can be substantiated for the Lessay vaults, it 
 becomes easier to accept the dates above assigned to the vaults of Durham. 
 
 On the whole, the chronology of the Durham vaults, as given above, appears 
 likely to receive increasing support.* If once the date of the choir aisle vaults 
 (1093-1099) be admitted, there is no inherent difficulty about the dates of the 
 high vaults. When once oblong severies had been vaulted with diagonal ribs 
 in the aisles, there was no difference in principle in constructing similar 
 vaults on the clerestory walls ; the oblongs merely had to be placed side by 
 side instead of end on. The height of the high vaults is no difficulty ; for c. 
 1083 the clerestory of the choir of St Nicholas, Caen, was given a groined vault ; 
 a more difficult piece of construction than a ribbed one. The wide span of the 
 high vaults of Durham also present little difficulty ; for before 1 140 the Norman 
 builders had succeeded in vaulting the apse of the chapter house of Durham, 
 which has an internal span of no less than 34 ft. 6 in.f 
 
 Second Solution. — At Durham there was still one defect. Except in 
 the single bays at the end of the transepts and nave, transverse arches were 
 omitted between each pair of bays (see 149. 1), a curious omission which 
 occurs also in Magdeburg Cathedral, \ A.D. 1 209. At Lessay § the high vaults 
 are arranged in single bays (308.3) ; all of which have diagonal ribs ; those 
 in the eastern part of the church are of the character illustrated in 315; in the 
 west bays of the nave they are of less heavy profile. Now Lessay Abbey was 
 begun in 1043 ; || or according to others, in 1050 or 1064 ; the founder died in 
 1098, and was buried in the middle of the choir. The choir therefore is earlier 
 than 1098. With the exception of the four westernmost bays of the nave, the 
 church was not originally designed for a high vault. There appear to be at least 
 five stages of work: (1) all except the west nave; (2) the vaults with heavy 
 profile ; (3) the west nave ; (4) the vaults with lighter profile ; (5) the west front. 
 The west doorway has a band of tooth-ornament, but need not be later than c. 
 1 1 50. The scalloped capital occurs in some of the western piers of the nave; 
 but is not necessarily evidence of very late Norman date, for it appears in the 
 earliest work of Norwich Cathedral ; nor are the griffes. If we assume an interval 
 of ten years between the different sets of work, this would give the date of c. 
 1 140 for the lighter vault ; c. 11 30 for the west nave ; c. 1 1 20 for the heavier 
 vault. Whatever be the exact dates it is undoubtedly a very remarkable and 
 early example of ribbed vaulting. Here at Lessay, we have precisely the type 
 of high vault which was employed in the early Gothic architecture both of the 
 He de France and England ; a complete solution of the great problem of the 
 mediaeval vault-builders. 
 
 * It is accepted by Professor Lethaby, Mediceval Art, 301. 
 
 f For the architectural and documentary evidence on which the above account of the 
 Durham vaults is based, see the important papers by Mr John Bilson in the Journal of the 
 Royal Institute of British Architects, Third Series, vi. 289, 345, and ix. 350. 
 
 % See plan in Dehio, Plate 447. 
 
 § Lessay is a Benedictine Abbey Church at the foot of the Manche in Normandy. It is 
 illustrated by Ruprich-Robert in Plates 89, 90 ; see also text 137, 138. His Plate 89 illustrates 
 the -west nave only. See 319, 315, 30S, 293, 412. 
 
 || Anthyme St Paul in Planat, vi. 23. Ruprich-Robert, i. 282, gives "fin du n e siecle," 
 as the date of the Abbey, and thinks (ii. 10) that it was finished c. 1130. There was a 
 consecration in 1178. 
 
HIGH VAULTS WITH SQUARE SEVERIES. 317 
 
 THIRD Solution. — Other solutions, however, were obtained ; all of them 
 of considerable interest. One of the most important is that adopted in the 
 ribbed vaults of S. Ambrogio, Milan. It is characteristic both of the Lombardic 
 and the German school of Romanesque ; and it was on these lines, i.e. with 
 coupled bays and alternating supports, that the vast cathedrals of Worms and 
 Mayence were vaulted, well on in the twelfth century. 
 
 We may suggest that it originated out of the diaphragmatic roof system ; 
 by which the churches, for roofing purposes, were crossed at every second 
 or third bay by strong broad transverse arches carrying gables. An arch, 
 so heavily weighted, required very massive supports ; and these had to be 
 piers, while the other supports might remain columns. Sometimes, as at 
 S. Prassede, Rome, and S. MINIATO (-35), Florence ("eleventh century, 
 the alternation of supports was pier, column, column, pier ; in the end 
 the customary arrangement was pier, column, pier. By far the greater 
 number of the churches which adopted the alternation of supports had 
 the latter disposition.* One solution, therefore, of the problem how to vault 
 a church with alternating supports was to vault the nave in large squares, and 
 the aisles in little squares (30S.4). Each large square bay (or "severy" = 
 " ciborium ") of the vaulting rested on two major piers on either side of the 
 nave ; each little square sever}- of the vaulting rested on one major and one 
 minor pier. This was a very neat solution. There were no oblongs ; and the 
 
 * Several churches in and near Normandy have the alternating piers ; e.g. Jumieges, 
 1040-1067, the earliest of all the large Norman churches except Bernay ; the Abbaye-aux- 
 Hommes; Graville; Notre Dame du Pre, Le Mans; and in French Transitional Gothic, Noyon 
 and Laon. In England we have alternating supports at Durham, Selby, Waltham, Lindisfame, 
 Norwich, Ely, Boxgrove. 
 
 It has been asserted that Lanfranc, who was a native of Pavia, introduced from Lombardy the 
 alternation of piers and columns into Normandy, lint Jumieges was begun in 1040. Lanfranc 
 did not arrive till c. 1042 ; and some years elapsed before he became powerful and influential. 
 Before Lanfranc's time, William, abbot of Fecamp from 1010 to 103 1, had great influence in 
 Normandy ; in his time forty new churches and monasteries are recorded to have been built. 
 He was a Lombard. So also, a Lombard, born near Novara, " St William, abbot of S. Benigne 
 Dijon, was, at the end of the tenth century, the great restorer of the arts in Burgundy. He- 
 rebuilt his abbey in magnificent fashion. He brought his colony of monks, together with 
 artists, from Lombardy in 996, according to the eleventh-century chronicler, Raoul Glaber. His 
 abbey seems to have been the most ancient model of the style prevailing in the Maconnais in 
 the eleventh and twelfth centuries" (Yirey, author of a valuable treatise on the Romanesque 
 Diocese of Macon, 1892). But the difficulty arises that there seems to be no adequate evidence 
 that any church in Lombardy with alternating supports is so ancient as the period either of 
 William of Dijon or of William of Fecamp. 
 
 Ruprich- Robert was of opinion that in Normandy the alternating supports, the big open 
 triforium, the windows at the back of it, the vaulted upper aisle, must have come from 
 Lombardy ; but not the masonry, nor the capitals (pages 54, 74, 84, 109, 115; so also Dehio, i. 
 2S6). On the other hand, the province of German Romanesque included the Netherlands, and 
 that of Normandy extended up to the Netherlands ; so that there may be a secondary influence 
 of Lombardic work through its very important connection, the German Romanesque. On 
 some minor resemblances between English and Lombardic Romanesque, see Colling, Early 
 Mediaval Foliage, 20; and Petit's Church Architecture, 73, 75, 77. It may be added that 
 while Norman is doubtless indebted to Lombardic or to German Romanesque, or to both, it 
 is a moot question whether Lombardic is the parent of German, or German of Lombardic 
 Romanesque. 
 
;i8 
 
 Peterborough Choir Aisle. 
 
 Angers Cathedral Nave. 
 
 Boxgrove Choir. 
 
319 
 
 Abbaye-aux-Hommes, Caen, Nave. 
 Lessay Abbey Nave. 
 
 Abbaye-aux-Dames, Caen, Nave. 
 Lessay Choir. 
 
32o SEXPARTITE VAULTS. 
 
 church was vaulted throughout. We have only one example of this vault 
 system ; the choir of BOXGROVE PRIORY, c. 1235 (308, 318). Here both nave and 
 aisles are vaulted ; but the nave vault rests on every other pier ; while all the 
 piers help to support the vaulting of the aisles. In such a plan, for every bay 
 or severy in the nave there are two severies in the aisle.* 
 
 Fourth Solution. — A better solution suggested itself to the builders 
 of Normandy (308.5). This, like the third, adopted the alternation of 
 supports ; but utilised them in a different way. Dividing the nave, as before, 
 into squares, it utilised the minor piers also as supports of the nave vault, by 
 constructing an additional intermediate transverse arch across the nave, from 
 each minor pier, D, to the minor pier opposite, J. This was a distinct improve- 
 ment, for each of the two cells CQG, IQN, was now divided into two ; and 
 it was easy to stretch planking across, e.g. from DQ to CQ and GQ, in filling in 
 the cells. But the solution was imperfect ; the cells CIQ, GNQ, were still much 
 too large. However, this plan found great favour in Normandy ; and from 
 Normandy it was exported into France and into England. East of Normandy 
 it is seen in Notre Dame, Paris; Laon ; Senlis ; Mantes; Sens; Beauvais; 
 Bourges ; Noyon nave ; in Burgundy at Notre Dame de Dijon ; Vezelay 
 St Pere ; Pont sur Yonne ; Chaumont ; Rouvres ; in Lausanne Cathedral, 
 Switzerland; in Italy at Casamari ; S. Galgano ; S. Martino; St Francis 
 Bologna ; Certosa of Pavia ; f not disappearing from Italy till the end of the 
 fourteenth century. In Normandy probably it was adopted first at the ABBAYE- 
 AUX-HOMMES, CAEN (3 1 5, 319) ; where the vaults maybe between 1 130 and 1 1604 
 From Sens it was brought by the architect, William of Sens, to Canterbury, 1 175 ; 
 it occurs at Rochester, c. 1200; in Lincoln great transept, c. 1215 ; St Faith's 
 Chapel, Westminster ; Lincoln galilee ; Durham east transept ; Canterbury, 
 St Augustine's gateway. § It appears to have been intended in the Norman 
 nave of Norwich ; || and of St David's.*! Practically it disappears in England 
 about the middle of the thirteenth century. A Norman sexpartite vault occurs 
 in the chancel of Tickencote, Rutland ; otherwise we might assume that our 
 sexpartite vaults were copied from those of the choir of Canterbury. 
 
 If diagrams 6, 7, 4, be compared with diagram 5, 308, it will be seen 
 that the high vault of Florence, Angers, and Boxgrove, and all the vaults in the 
 various aisles, have only four cells in each severy ; while Canterbury choir has 
 six. The former is therefore called a quadripartite, the latter a sexpartite 
 vault. The SKELETON of a sexpartite vault in diagram 321 shows on the left 
 a pointed transverse arch AHB spanning the nave ; on the right two small wall 
 arches BFC, CED (above two clerestory windows) ; between them is the new inter- 
 mediate transverse arch CGJ spanning the nave. As the diagonals are semi- 
 circular, this particular vault is highly domical. 
 
 * Note that the flying buttresses (373) correspond to the bays of the nave, not of the aisles. 
 
 t Enlart's Gothic in Italy, 251. 
 
 % Note that the major and minor piers have respectively three vaulting shafts and one ; 
 and that the diagonal ribs are elliptical. In 315 note the large number of voussoirs in each arch. 
 
 § This is a compromise between a sexpartite vault and a quadripartite one with a trans- 
 verse ridge rib ; illustrated in Caveler's Specimens. 
 
 || Ruprich-Robert, i. 144, and ii., Plate 91. IT Scott's Lectures, i. 117, Plate 76. 
 
HIGH RIBBED VAULTS. 
 
 321 
 
 Fifth Solution. — Of the sexpartite solution there is a variant (the fifth), 
 which we may call the quasi-sexpartite : it looks sexpartite, but is not ; it seems 
 to have six cells, but really only has four. It is well seen in the high vault of 
 the ABBAYE-AUX-DAMES (319), as recently rebuilt. Like the true sexpartite 
 vault, it has an intermediate arch. But this merely carries a wall ; whereas in 
 the sexpartite vault below, it carries half a curved web on one side of it, and 
 half on the other. In diagram 308.5, CDQ and GDQ are separate hollow cells, if 
 the vault be sexpartite; if it is quasi-sexpartite, CGQ is all one hollow cell, but 
 with a partition wall DJ run up the centre. This wall is well seen in 319, on 
 the left. The quasi-sexpartite vault is not found except in Normandy. It is an 
 unscientific form ; the intermediate wall having no constructional value. This 
 vault occurs at Bernieres, St Gabriel, and Ouistreham ; a variant of it is seen in 
 La Trinite, Angers. It does not occur in England. 
 
 Sixth Solution. — This was to divide the nave, not the aisles, into squares 
 (308.6). Unfortunately the result was that the oblongs, which had been 
 expelled from the nave, turned up again in the aisles. The contrast between 
 the two forms of vault is exceedingly unpleasant ; but the Italians, who in Gothic 
 architecture preferred engineering to art, con- 
 tinued to employ this form till late in the 
 Gothic period ; e.g. S. Theodore, Favia (" 1 1 50- 
 11 80; De Dartein) ; Florence and Verona 
 cathedrals; S. Fetronio, Bologna. This 
 method was not adopted in England. 
 
 Seventh Solution. — This was the 
 drastic solution of the iconoclast who built the 
 
 vaults of ANGERS CATHEDRAL, I 1 50 (318). 
 
 He divided his nave into great squares, 
 pulling down the aisles.* There were no 
 oblongs to vex him then. We never carried 
 out this plan on the magnificent scale of Angers Cathedral ; but we adopted 
 it in smaller work ; e.g. in the Cistercian choir of Kirkstall. In France, however, 
 Angers Cathedral produced a school ; and very large and noble Gothic churches 
 in the South were designed on the hall plan, i.e. without aisles ; but with 
 oblong instead of square bays ; e.g. St Vincent, Carcassonne, with a span of 
 69 ft. ; the nave of Gerona in Spain, with a span of ~2 ft. ; the Church of the 
 Cordeliers at Toulouse, now burnt to the ground ; t and the masterpiece of the 
 style, Albi Cathedral. 
 
 POINTED Arch IN VAULTING. — What is particularly remarkable about 
 the high vault of DURHAM nave (8) is that its transverse arches are pointed.* 
 Next to the introduction of ribs instead of groins, it was the greatest improve- 
 
 Sexpartite. 
 
 * Originally Angers Cathedral had aisles. Afterwards the whole church was thrown into a 
 single span. The same change took place in Notre Dame de Couture, Le Mans ; in several 
 of the churches of Zamora ; and probably in the nave of Bordeaux Cathedral. For the original 
 plan of Angers Cathedral see Lethaby's Med. Art, 16S. 
 
 t Illustrated in Fergusson, ii. 69, 70, 488. 
 
 \ Other early examples are the west aisle of Peterborough transept ; Malmesbury ; St 
 Cross, Winchester ; The Lady Chapel, Glastonbury ; and Wells Cathedral. 
 
 X 
 
322 POINTED ARCH IN VAULTING. 
 
 ment that was ever made in vault construction. It removed at one stroke all 
 the difficulties which had been worrying" the builders in their endeavours to 
 construct vaults with horizontal ridges, but without elliptical diagonals. The 
 semicircular arch is non-elastic ; its height is one-half of its span ; cannot be 
 more. But the height of a pointed arch is an elastic dimension. Without 
 changing its span, we can make the arch as high as we please (258.7-10). 
 In other words, by employing pointed arches for the narrower spans of a vault, 
 we can make their ridges rise to just the same level as those of the semi- 
 circular diagonals, in spite of the wider span of the latter ; so both the ridges 
 will be horizontal; e.g. in the sexpartite vault, figured on 321, it would have 
 been quite feasible to make all the six pointed outer arches more acutely 
 pointed still, till their apices were on a level with the central point of the vault ; 
 then the ridges, instead of curving, would be horizontal. It was a simple, but 
 a most scientific and successful way out of the difficulties that so long had 
 beset vault construction. At first, however, and indeed for a long time in 
 Gothic architecture, though the outer arches were pointed, the diagonals re- 
 mained semicircular.* But, curiously enough, they are found pointed in some 
 of the earliest vaults ; e.g. all those of St Cross, Winchester ; and Wells 
 Cathedral ; so also in the French cathedrals of Chartres and Reims. The 
 result was to bring the thrusts down more vertically. All four are churches 
 in which the massiveness of the construction argues an anxiety to secure an 
 overplus of stability more characteristic of Roman or Romanesque than of 
 Gothic architecture. 
 
 * A vault with the diagonals semicircular and the outer arches pointed is figured in 298.3. 
 
Chapter XXII. 
 
 GOTHIC VAULTS. 
 
 Web Construction — Ridge Ribs — Tiercerons — Liernes — Fan Vaults. 
 
 WEB CONSTRUCTION. — We have seen how the builders, after much experi- 
 menting, arrived at the construction of quadripartite ribbed vaults. After this 
 the history of the vault diverged in the schools of the He de France and 
 England. This seems to have been largely due to a difference in mason-craft 
 in the two countries. It was this. In France it was customary to construct 
 each arched course, of which the web was composed, of voutains of different 
 sizes. In England * the mason constructed his web without troubling to cut 
 the voutains to shape. Let FEDB represent an oblong bay of a nave which it 
 is required to vault (326.1). BAD is one of its four cells; in which BA, I>A are 
 diagonal ribs, BC, DC, transverse ribs. The dotted line AC represents the ridge 
 of the cell. We have to vault BAD in two portions ; first one half, then the 
 other. We will take the half BAC. This half cell is shown separately in 
 diagrams 2 and 3. Biiil is a series of horizontal courses, built solid. It 
 remains to vault the space above this BGH. First of all, we will fill up this 
 space with a wooden vault of planks. We have two workmen, a Frenchman 
 and an Englishman. Both are supplied with planks of the same width. First 
 let us look at the Englishman's work (326.2). He cuts off as much of a plank 
 as he wants, and rests one end of it on the diagonal BA and the other on the 
 transverse rib BC ; this is plank 1, 1. Then the same with plank 2, 2; and so 
 on till he reaches the ridge. But at the ridge AC the planks do not make a 
 straight joint. The Frenchman, on the other hand, in 326.3, alters the breadth 
 of his planks, so that they are narrower towards the transverse rib and broader 
 towards the diagonal. Then, when he has put them all on, he gets a straight 
 joint at the ridge AC. Next, let us suppose each to be supplied, not with 
 planks, but with blocks of stone ("voutains"). Where each plank is shown in 
 the diagram, he will have to build an arched f course of stones. The French- 
 man treats each course as before ; i.e. he shapes each voutain on the scaffold by 
 a few strokes of his tool, and makes one end of the course broader than the 
 other. This takes trouble and time ; but it results at the ridge in a clean 
 straight joint. The Englishman's method is cheaper and more expeditious. 
 He builds each course of the same breadth from end to end, and thus has 
 
 * "Le poseur (Anglais) pourra ainsi n'avoir a placer que des moellons egalement epais" 
 (Viollet-le-Duc, Architecture, ix. 522 ; cf. iv. 108). 
 
 + The arched form of the course adds another complication, which need not be taken 
 into account at present ; it is shown in 298.2. 
 
3 2 4 
 
 
 ® 
 
 
 ^^ 
 
 1 //^L 
 
 
 
 
 ^ — 
 
 
 vl|SNr~~^i 
 
 
 L / m/ ry / 
 
 
 © 
 
 F / ^*^z \ G 
 
 \ ^^^F ^^^ / 
 
 A S C 
 
 
 ^ =::: \~~--i 
 
 
 
 ^Sx. 
 
 
 (6) 
 
 ^ — ~~* 
 
 
 1. Westminster Choir. 
 
 2. Lincoln Choir. 
 
 3. Chester Chapter House. 
 
 4. Lincoln Nave. 
 
 5. Lichfield South Transept. 
 
 6. Lincoln Presbytery. 
 
 7. Exeter Nave. 
 
 8. Oxford Schools Tower. 
 
 9. Lichfield Nave. 
 io. Durham Kitchen. 
 
325 
 
 ABA 
 
 ABA 
 
 © 
 
 bx /\\\\ / /y/\ >P 
 
 B\ \ \ / / \ \v\ / Vb 
 
 © 
 
 © 
 
 © 
 
 H J 
 
 © 
 
 © 
 
 WEST 
 
 r. Lincoln Chapter House. 
 
 2. Southwell Chapter House. 
 
 3. Wells, Undercroft of Chapter House. 
 
 4. Wells Chapter House. 
 
 5. Oblong Severy with curved ridges. 
 
 6. Wells Lady Chapel. 
 
 7. Gloucester Choir. 
 
?26 
 
 10 
 
 I0\ 
 
 
 
 
 
 IO 
 
 9 
 
 9\ 
 
 
 
 
 
 9 
 
 a 
 
 £\ 
 
 
 
 
 
 Ft 
 
 7 
 
 t\ 
 
 
 
 
 
 7 
 
 6 
 
 (D ^ 
 
 
 
 
 
 f. 
 
 5 
 
 •\ 
 
 
 
 
 <> 
 
 4- 
 
 -> 
 
 
 
 <t- 
 
 > 
 
 z" 
 
 
 ^ 
 
 2 
 
 1 
 H 
 
 1 \ 
 
 2 
 1 
 
 H 
 
 X I 
 
 4. St Saviour's, Southwark. 
 
 5. Buildwas. 
 
 6. Westminster Cloister. 
 
 7. Square Severy with curved ridges. 
 
327 
 
 Lincoln Choir. 
 Lincoln Nave. 
 
 Lincoln Presbytery. 
 Ely Presbytery. 
 
Lichfield S. Transept. 
 Lincoln S.W. Chapel. 
 
 Lincoln Central Tower 
 St Hilaire, St Florent. 
 
3?9 
 
 Bristol S. Aisle of Choir. 
 Bristol Berkeley Chapel. 
 
 Kly Choir. 
 
 ( Ixford Proscholium. 
 
33° 
 
 
 u ' ! .ml^v> / " : 
 
 /• 
 
 
 Norwich Nave. 
 Norwich Nave. 
 
 St George's, Windsor, Nave. 
 Tewkesbury Choir. 
 
S3' 
 
 Oxford Divinity School. 
 Oxford Cathedral Choir. 
 
St George's, Windsor, Nave. 
 Wells Choir Tewkesbury Nave 
 
333 
 
 Hereford S. Transept. 
 Wells Lady Chapel. 
 
 Sherborne Nave. 
 King's College Chapel. 
 
334 
 
 Ely, West's Chapel. 
 Canterbury Eastern Crypt. 
 
 Gloucester Choir. 
 Gloucester Lavatory. 
 
RIDGE RIBS. 335 
 
 not to shape his blocks, but he gets a ragged joint. As the two diagrams 
 show, the French method divides the two ribs into an equal number of parts; 
 while the English method divides them into an unequal number; here the 
 diagonals have fourteen, the transverse ribs ten parts. Compare ST SAVIOUR'S 
 (326.4), and CARLISLE CHOIR AISLE (498). The French method of filling in 
 is found occasionally in England; e.g. in the Chapter House of BUILDWAS 
 ABBEY, c. 1 1 55 (326.5), where the filling in is of coursed rubble; in the Transi- 
 tional aisle of Ripon transept; at NEW SllnREHAM and CHICHESTER (313;; 
 LINCOLN PRESBYTERY (327), and very frequently in late work. On the other 
 hand the English method occurs now and then in France : e.g. in the cloister 
 and crossing of Fontefroide Abbey ; and in the aisles of Eu, Normandy. 
 
 RlDGE RIBS. — Let us imagine a vault at which a French and an English 
 mason have been at work together, as in diagram 326.6. Plainly three webs 
 were built by the Frenchman ; the remaining web by the Englishman. The 
 Englishman's ragged joint is not only unsightly, but less strong than the other. 
 Therefore, partly to mask it, partly to strengthen it, a new rib was invented 
 which is called the ridge rib, DI, with a number of saw-like teeth cut in its 
 side to receive the ends of the top-courses of the web, where they abut on 
 the ridge.* The first example we have is in the transept aisle of Ripon, c. 1 170, 
 where the ridge ribs are so slender that the} - can be but decorative ; moreover, 
 they are unnecessary, as the cells are filled in after the French fashion. Ridge 
 ribs occur in the high vaults of Angers and Poitiers cathedrals, c. 1165 ; and, 
 perhaps as early as 11 30, at Airaines and Luchueux, Somme.t These slender 
 unconstructional, decorative ribs form indeed one of the most striking features 
 of the Plantagenet Gothic of Anjou (328). Constructional ridge ribs do not occur 
 in England till Lincoln choir, which was commenced in 1192.^ In the He de 
 France and Picardy, and in the Burgundian Gothic exported to Italy, they 
 are employed first to strengthen the vault of the crossing, because of its great 
 span ; e.g. of Fossanova, consecrated 1208 ; Casamari and Arbona in Italy; and 
 of Amiens, c. 1265. There was indeed no structural need for them in the rest of 
 the vaults, if constructed in the French manner. 
 
 Each compartment of a vault has ridges both ways ; both across the church, 
 and in the line of its axis ; the former are called transverse, the latter longitudinal 
 ridge ribs ; the latter run the whole length of nave or choir and transept. 
 
 In a non-domical vault the transverse ridge-rib^ will be horizontal: in a 
 domical vault it will curve downward at each side of each severy ; i.e. if it occurs 
 in a severy of a domical high vault, it will curve downwards both ways from the 
 central boss of the vault to the tops of two opposite wall arches above the 
 clerestory windows at II and F in 325.5. But it is just as feasible to make 
 
 * " Elles concourent au soutenement, ct leur principale utilite est de former des couvrejoints 
 pour masquer certains raccords defectueux d'appareil, au sommet des voutains " (Enlart's 
 Manuel, 39). 
 
 + Illustrated in Enlart's Manuel, 38. 
 
 % In the Chapter House of St George's de Boscherville, probably built between 1175 and 
 1200, is the commencement of a longitudinal ridge rib. See M. Besnard's monograph, 150. 
 
 S The transverse ridge rib as shown on the plan of a vault must not be confused with 
 the transverse ribs, which are parallel to it ; nor the longitudinal ridge rib with the wall ribs 
 which are parallel to it. 
 
336 INTERMEDIATE RIBS. 
 
 it curve up as to make it curve down. And sometimes the transverse ridge, 
 and consequently the transverse ridge rib, is curved upward to allow a higher 
 clerestory window to be inserted ; e.g. in the upper aisle of Notre Dame, Paris ;* 
 at Norwich and DURHAM (306) in the south aisle of the nave ; and in the south 
 aisle of CHICHESTER NAVE (3 I 3). 
 
 Artistically, the addition of the other ridge rib, the longitudinal one, running 
 uninterruptedly from west door to eastern wall, and from one transept end to 
 the other, is a very great improvement. It ties together the disconnected bays 
 of an interior, and gives it a unity which no string-courses can effect. Without it 
 such a church as Amiens or WELLS (524) or BOXGROVE (318) is like some inverte- 
 brate creature ; ribbed, but without a spine. What the strong fourteenth-century 
 parapets do for the exteriors of Wells and Eincoln, that the longitudinal ridge 
 rib does for interiors. Yet it was not till late, till the fifteenth century ; e.g. in 
 Souvigny Abbey;! that the French adopted this great artistic improvement; 
 and then they could only make a success of it by adopting also the English 
 method of vaulting ; i.e. with horizontal ridges. For if the vault be domical, 
 then in every severy the longitudinal ridge rib rises up from the summit of one 
 transverse arch to the central boss and then plunges down to the summit of 
 the next transverse arch ; then climbs to the next boss, descends, and so on. 
 The result is, as may be seen in one or two of our early vaults, e.g. those of 
 Lincoln great transept and its imitation in Southwell choir, and at St Riquier, 
 a rib wobbling up and down distressingly all the way from the central tower to 
 the end-wall window. In late vaults as many as three longitudinal ridge ribs 
 occur ; e.g. in Gloucester west nave and choir ; TEWKESBURY NAVE (332) ; 
 and ST GEORGE'S, WINDSOR (332). J 
 
 Intermediate Ribs or TiERCERONS. — The next, or perhaps a simul- 
 taneous step in advance, was to introduce an additional pair of ribs in each 
 cell where they might be useful as intermediate supports to the ridge rib ; rising 
 from the same capital as the diagonals, but not rising to the central boss 
 where the diagonals intersect, but to some point between that boss and the 
 summits of one of the four outer arches ; i.e. the two transverse and the two 
 wall-arches (see 324). As we have seen, the English vault was much more 
 flat-topped than the domical French vault, and accordingly more difficult to 
 construct. By the addition of these extra pairs of ribs the span of each cell was 
 reduced by one-half: and the centering required for the courses of the webs 
 was greatly diminished. 
 
 The first attempt to reduce the distance from diagonal to diagonal was 
 made in St Hugh's choir at LINCOLN (324.2 and 327). This is usually 
 regarded as an architectural freak, without rhyme or reason ; really it is an 
 ingenious attempt to provide additional permanent centering to make the web 
 construction easier. § It will be seen on reference to 324.2, that instead of 
 dividing the longitudinal ridge rib, ST, into two halves at the centre, and con- 
 
 * Illustrated in Viollet-le-Duc, Architecture, ii. 289. t Illustrated in Fergusson, ii. 170. 
 
 \ Examples of this rib are illustrated in the naves of YORK (10) and exeter (9). In 
 WESTMINSTER CHOIR (324.1 and 63) there is a longitudinal, but not a transverse ridge rib. 
 § So Prior, 95. 
 
IXTFRMLDIATF RIBS. 
 
 337 
 
 structing diagonal arches intersecting at that centre, diagonals have been 
 omitted, and the ridge rib has been divided into three equal parts. Thus several 
 of the cells of the vaults are much narrowed, and therefore their webs are easier 
 to construct. But the solution is an imperfect one, for it still leaves two of the 
 cells, AVii and CXI), too broad ; and for this reason probably this curious vault 
 was never imitated. 
 
 The correct solution is seen in LICHFIELD SOUTH TRANSEPT (324.5 and 
 328). In this the longitudinal ridge rib, ST, in each severy is divided, not into 
 three parts as in Lincoln choir, but into four, at the points J, E, K. At E the 
 diagonals AD, BC intersect. From the two bottom corners of the severy, 1; and D, 
 a new pair of ribs, BK, DK, is built up to the point K on the ridge rib. Similarly 
 other pairs are built from A and C up to J. These new ribs, intermediate between 
 the diagonal and outer ribs, are called tierccrons. In LINCOLN PRESBYTERY 
 (324.6) the tierccrons occur in the east and west cells ; in CHESTER CHAPTER 
 HOUSE f'324.3) in the north and south cells ; in LINCOLN NAVE 324.4) and LICH- 
 FIELD SOUTH TRANSEPT (324.5) in all four cells. 
 
 The practice varies. One reason for this is, that a severy oblong on plan, 
 e.g. over a nave, requires different treatment from a seven - square on plan, 
 e.g. over an aisle. In 324.3, AEB and CED are broader cells than AEC and BED ; 
 therefore sometimes, as in Chester chapter house, it is only in these broad cells 
 that pairs of tiercerons, AF, FB and CG, GD are placed. In LINCOLN PRESBY- 
 TERY (324.6), on the other hand, it is the narrow cells that receive pairs of 
 tiercerons. In LINCOLN NAVE (324.4 , LICHFIELD SOUTH TRANSEPT (324.5 , 
 and ELY PRESBYTERY ( }2j 1, each of the four cells of the vault receives an 
 additional pair of tiercerons ; in Lincoln nave and Chester chapter house trans- 
 verse ridge ribs occur, but are abbreviated. The vault of LICHFIELD SOUTH 
 TRANSEPT (324.5 is the same as that of Lincoln nave, except that the transverse 
 ridge rib is not abbreviated, but is prolonged to the summits of the wall-arches 
 H and I. 
 
 In 324.7 the complication increases ; for two cells contain one pair of 
 tiercerons each, and the other two cells contain two pairs each. At Oxford the 
 large square vault under the Schools Tower, and also the central compart- 
 ment of that of the DIVINITY SCHOOL (324.SJ, have three pairs of tiercerons 
 in each cell (331 ). 
 
 Then comes in another great improvement. For if the vault of the south 
 TRANSEPT OF LICHFIELD I 328; be examined — it is the same in plan as that of 
 Westminster nave — it will be seen that the space from rib to rib is now so small 
 that in the greater part of the vault it can be bridged over by single flat stones ; 
 no arched courses being necessary. The difficulties of web-construction are not 
 merely reduced, they are annihilated. The vault has become, '>r is on the verge 
 of becoming, a rib ami panel vault : a result little foreseen by Geoffry de Noiers 
 when he designed Lincoln choir. When this was once realised, as it 
 could not fail to be, that given plenty of ribs, the webs could be filled in 
 with panels, more and more ribs were provided ; till the vault became nearly 
 all rib. EXETER (324.7) is an example of this latter type; with one pair of 
 tiercerons east and west ; and three pairs north and south ; a most noble 
 vault ; ever}- rib doing definite work ; fastening down and clamping the bays 
 
 Y 
 
333 
 
 INTERMEDIATE RIBS. 
 
 like the lid of some ancient strong-box ; every curve guiding the eye upward 
 to the horizontal ridge rib, and along it to furthest east and furthest west ; our 
 very noblest achievement in ribbed vaulting. These vaults were designed, in 
 EXETER CHOIR (9), as early as 1280. 
 
 Such are some of the main types of the quadripartite* vault with ridge ribs 
 and tiercerons. But all sorts of variations occur. In LICHFIELD NAVE (324.9) 
 the builder amused himself by omitting the transverse arches ; dividing the 
 ridge EU into three equal parts instead of the usual four. It is an improved 
 version of the vault of LINCOLN CHOIR (324.2). 
 
 With the aid of the new tiercerons the builders were able to obtain sur- 
 prising results ; e.g. at Durham it was desired to vault the PRIOR'S KITCHEN, 
 which is an octagon, but in such a way as to leave a hole in the centre of the 
 
 vault for the smoke 
 ^-NA to escape. This was 
 done by transverse 
 arches alone ; without 
 any tiercerons or dia- 
 gonals. A transverse 
 arch was thrown 
 across from each 
 corner to the opposite 
 corner ; and the inter- 
 sections of these eight 
 transverse arches pro- 
 duced an octagonal 
 central space, which 
 was left open.f Or 
 suppose that it is 
 desired to vault a very 
 large square or oblong 
 space, with a central 
 pillar at D. All that is 
 necessary is to group 
 together four simple 
 quadripartite vaults: allowing the capital of the pillar D to rise to the same height, 
 or thereabouts, as the capitals or corbels at A, B, G, H, I, F, E, C. Then the diagonals 
 JD, LD, MD, KD, and the transverse ribs ND, RD, PD, OD, will descend on to the 
 capital D of the central pillar ; and the remaining ribs will descend on to the 
 capitals or corbels B, A, C, E, F, I, H, G. This is the plan of the N.W. CHAPEL 
 OF LINCOLN NAVE (308.9). I The s.W. CHAPEL (328) is precisely the same in plan: 
 but there is no pillar at D. This is managed by not allowing the diagonals AJ, GL, 
 LM, EK, and the transverse arches BN, CO, FP, HR, to begin to descend as before 
 at the small bosses, but making them go on rising till they reach D, which now 
 is no longer on the same level as A, B, G, H, &c, but at a much higher level, 
 
 * Note that it is still quadripartite ; it has only four cell-surfaces or webs, however much 
 they are cut across by ridge ribs and tiercerons. 
 
 t Scott's Lectures, ii. 203. \ Illustrated by Sir G. Scott in Lectures, ii. 19S. 
 
 Durham Cathedral, Vault of the Prior's Kitchen. 
 
RIDGE RIBS AND TIERCERONS. 
 
 339 
 
 making the vault domical. Both these plans, i.e. with and without a central pier, 
 occur in the vault of Glasgow crypt* And it is obvious that, applying similar 
 methods to cover an octagonal or polygonal space, we can produce a series of 
 chapter houses ; either with a central pier, as at LINCOLN (340), Westminster, 
 Salisbury, WELLS (123) ; or without a central pier, and therefore highly domical, 
 as at York, where the vault, however, is but of wood, and at Southwell, where it is 
 of stone. The following is the plan of Southwell chapter house (325.2). 
 AA are diagonals ; BB ridge ribs ; AD, ae tiercerons. The former class of chapter 
 houses may be descended from the twelfth-century chapter house of Worcester, 
 which again may be but a development of the ribbed semi-dome of the apse in 
 the crypt. The second class may perhaps be fathered on the Norman octo- 
 partite vault which covers the Treasury of Canterbury ; both the above are 
 c. 1 160. In some cases, e.g. in vaulting a TOWER, AS AT LINCOLN (328; and 
 York, where a central pillar is obviously impossible, the vault of the S.W. CHAPEL 
 OE LINCOLN NAVE (328), just provides the model required. It will be noticed 
 in the tower that a large hole has been left in the centre, i.e. round the point 
 D, through which the bells could be drawn up (308.9). 
 
 Among other combinations ma}' be mentioned Tripartite and Quinque- 
 partite vaults. Tripartite vaults occur in the undercroft of WELLS CHAPTER 
 HOUSE (325.3). Ouinquepartite vaults, with five cells, are useful sometimes 
 where the seven" of an aisle is polygonal ; or where there are two lancet 
 windows in each bay, as in the aisles of LINCOLN 1 308.8;, Salisbury, and South- 
 well. In the Salisbury aisle the vault in front is quadripartite; that at the 
 back is quinquepartite. But there was practically hardly any limit to the 
 combinations of ribs. Lincoln, in particular, was a vast laboratory of vault 
 experiments in the thirteenth century. 
 
 RlDGE Ribs AND Tiercerons. — It is sometimes asked whether the ridge 
 rib or the tierceron appeared first. In Anjou, e.g. at ST HILAIRE, ST ELORENT 
 (328), near Saumur, and in importations from Anjou, e.g. Airaines and 
 Ripon transept, the ridge rib undoubtedly appeared first ; and in these ex- 
 amples it is of no constructional use whatever ; as the filling in is clone in 
 the French fashion with a straight joint at the ridge. But the solitary Ripon 
 example created no school ; and the ridge rib does not appear again with us 
 till Lincoln choir, 1192. Here plainly it is brought in to form something for 
 the ribs to abut on ; because they do not meet in pairs and form arches 
 (324.2). So again, wherever a pair of tiercerons is employed, they are not 
 in the same plane and therefore do not form an arch : here again a ridge-rib 
 is necessary.f On the other hand, a ridge rib is employed in WESTMINSTER 
 CHOIR (324.1); here there are no tiercerons, but the filling in of the webs is 
 done in the irregular English fashion, and a "cover-joint" at the ridge is 
 desirable.:): Ridge ribs, therefore, while purely decorative in Angevin vaulting, 
 
 * See Mr Watson's monograph on Glasgow Crypt. 
 
 + It will be noticed that the transverse rid^e ribs are not always constructed across the 
 whole span of the vault, as in Lichfield transept 524.5; ; but only just so far as they are 
 needed to provide abutment to the tiercerons ; e.g. in (II KM kr CHAPTER HOUSE and LIN( 01 \ 
 NAVE (324.3, 4). 
 
 \ There is, however, no transverse ridge rib, 
 
34-0 
 
 LIERXE RIBS. 
 
 are constructional in English vaulting ; either because they mask and strengthen 
 irregular jointing, or because they give abutment to other ribs, or because they 
 do both. 
 
 SKELETON VAULTING. — Sometimes a secondary rib system is constructed 
 beneath the ordinary ribbing to give the vault additional support. This is 
 especially common in Bristol Cathedral ; the photographs are from the BERKELEY 
 CHAPEL and the CHOIR AISLE (329) ; cf. Warwick chancel. 
 
 LlERNE RlBS. — At Exeter the builders might well have stopped. But the 
 logic of vault construction urged them on yet further. If such a vault as that 
 shown in 308.10 be examined, it will be seen that the tierceron DE is only 
 obliquely abutted by the tierceron BE and the ridge-rib HI. It seems to call for 
 abutment in the same plane. So a little strut, EF (French Her, to bind) was 
 inserted, which provided the direct abutment required. In the same way, a 
 lierne, EG, was inserted to abut BE ; and so on, all the way round. This form 
 of vault, the Stellar, became exceedingly popular in square compartments,^, in 
 
 the OXFORD PROSCHOLIUM (329) ; under a 
 tower, in the oriel of a window, in a cloister, or 
 in a gateway. But probably what commended 
 it much more than logic, was the prettiness of 
 the patterns which were obtained, e.g. TEWKES- 
 BURY CHOIR (330). And very soon all sorts 
 of hemes were added, many of them not making 
 even a pretence of being of constructional value. 
 And not only hemes, but unconstructional 
 ridge ribs as well ; e.g. in the westernmost bay 
 of Gloucester nave there are three longitudinal 
 and three transverse ridge ribs. Triple ridge 
 ribs occur also in GLOUCESTER CHOIR (334) and 
 Lady Chapel, and in the naves of TEWKES- 
 BURY (297) and ST GEORGE'S, WINDSOR 
 (330). In GLOUCESTER (496), where most of 
 the chief developments of later English Gothic 
 art originated, lierne vaults of the most amazing 
 complexity were constructed ; that of the SOUTH TRANSEPT (306) before 
 1337, that of the CHOIR (334) before 1350. In spite of the rise of a rival 
 type, the fan vault, the lierne vault remained in constant use till the extinction 
 of Gothic architecture. 
 
 As to the name of the new rib, it is a very unfortunate one. For in the 
 French of Philibert de l'Orme (sixteenth century) and in modern French 
 architecture, a lierne is what we have called a ridge rib. However, the present 
 English connotation of the term is too deeply rooted to be got rid of. Liernes 
 may be defined as all those ribs which merely cross from rib to rib. They 
 neither spring from an abacus, nor rise to the central boss. 
 
 The earliest example of the lierne ribs has been supposed to be that of the 
 vault of LINCOLN CHAPTER HOUSE* (325.1), which can hardly be later than 
 
 * Mr W. C. Watkins and Mr John Allan, of Lincoln, have kindly supplied moldings of 
 this vault. 
 
 Lincoln Chapter House. 
 
LIERNE RIBS. 
 
 34' 
 
 c. 1230 Moldings occur in the nave of similar section to those of the ribs of this 
 vault; so that the vault is contemporaneous with the chapter house itself, and 
 is not a later addition. But the ribs which make up the inner decagon arc really 
 ridge ribs. The Heme rib proper seems not to come into general use till the 
 beginning of the fourteenth century ; e.g. St Stephen's crypt, Westminster ; 
 ELY CHOIR (329) and GLOUCESTER S. TRANSEPT (306), c. 1336. Other fine 
 examples are the naves of WINCHESTER 342) and Canterbury; the OXFORD 
 DIVINITY school (331); and the CATHEDRAL CHOIR (331) ; all the high vaults 
 of NORWICH (330); and of ST GEORGE'S, WINDSOR (330, 332. In this last 
 they are often described erroneously as fan vaults ; really they consist of two 
 longitudinal sections of a Heme vault, separated by a segmental barrel vault. 
 
 St George's, Windsor. 
 
 In the choir of Oxford Cathedral this disposition is reversed ; for each severy 
 contains a pair of segmental barrel vaults separated by a lierne vault. WELLS 
 has a vault in the chapter house without Hemes, and a very complex vault 
 in the Lady Chapel (325.6). In Wells chapter house (325.4), AC, BC, EC, FC, 
 are diagonal arches ; tiercerons rise from A and 1: to 1 and 2 ; m; is a ridge rib ; 
 6> 5i 3> 4. 7> are also ridge ribs. 
 
 In the earlier and in the best examples, the main constructional ribs, the 
 diagonals and tiercerons, are retained. Hut in some later examples, e.g. in 
 WELLS choir (332; and in Dreux transept, the whole construction of the vault 
 is revolutionised. It ceases to be a ribbed vault at all ; the ribs are merely 
 surface carvings ; the vault has become a group of purely geometrical forms, 
 
342 LIERXE RIBS. 
 
 which are called conoids ; i.e. a solid form semicircular on plan, terminating in 
 a point. A highly developed type of this kind of vault, the " Reticulated," is 
 specially characteristic of the late Gothic of Germany.* It is a particularly 
 objectionable type ; for its surface decoration has no connection whatever with 
 the divisions of the bays below it. In France, fine examples of lierne vaulting 
 occur in the Flamboyant churches of St Riquier, Montargis, Caen St Pierre, 
 and Chaumont ; and in Rue Chapelle, + Somme. The late vaults of Spain, 
 e.g. at Segovia, Zaragossa, Salamanca, S. Juan de los Reyes, Toledo, are 
 magnificent, and alone may bear comparison with those of England. But, from 
 first to last, even in such a simple type as ELY CHOIR (329), lierne vaults are 
 vicious ; the Hemes interrupt the flow of upward curve ; for the beauty of a 
 vault it is essential that the ribbing should follow the lines of main strength. 
 In such complex vaults as that of GLOUCESTER CHOIR (334), the rib system, 
 to the eye, is simply a maddening chaos. Yet, on paper, it is simple enough. 
 It ought to have remained on paper. Omitting, for simplicity, all non-essentials, 
 its main structure, as seen in 325.7, is, that it has an additional set of diagonal 
 arches spanning two bays ; and all the various sets of diagonals intersect one 
 another, producing an astonishing medley of curves, but nevertheless leaving 
 spaces for the heads of the clerestory windows. This vault is a panel vault 
 rather than a ribbed vault. 
 
 It should be noticed that the symmetrical appearance of a vault plan on 
 paper does not necessarily ensure that it will be symmetrical when seen in 
 perspective on the surface of the vault, which is curved, not flat like the paper. 
 The ribbing of the vault over the monument of Archbishop Stratford at Canter- 
 bury is symmetrical, as seen in perspective ; but drawn on paper is much 
 distorted. The construction of such a vault argues great power of foreseeing 
 the result of a design as seen in the solid. But, sometimes, as in WELLS CHOIR 
 (332), a design is symmetrical on paper, unsymmetrical in execution, j because 
 account has not been taken of the concavity of the vault. 
 
 Fan Vaulting. — The same school of masons, that of Gloucester and 
 Tewkesbury, which had given such an astonishing development to the lierne 
 vault in those two abbey-churches, hit upon a still more advanced type of vault, 
 the fan vault, perhaps to some extent accidentally. It was the custom to suspend 
 over a corpse lying in state a pall supported by four posts. This velvet pall was 
 translated into wood in the canopy which is still suspended over the tomb of the 
 Black Prince at Canterbury ; and very soon it was translated into stone. Such 
 a stone canopy occurs in Tewkesbury choir over the tomb of Sir Hugh Despenser, 
 who died in 1 349-§ Inside these stone canopies it was customary to carve on a 
 diminutive scale an imitation of the elaborate vaults of the day. But in this 
 canopy, to save himself the trouble of carving a multiplicity of tiny ribs 
 branching up from one abacus, the mason simply carved a plain half funnel, and 
 then painted the ribs on it. In the Trinity Chapel, however, which was probably 
 erected after the death of Edward, Lord Despenser, in 1375, the half funnels or 
 fans are decorated with real ribs of stone. Following such examples, it was but 
 
 * See illustration in Dehio, ii. 571. 
 
 t Illustrated in Enlart's Manuel, 598. t Willis' Vaulting, 39. 
 
 § This tomb is illustrated in the Spring Gardens Sketch Book, vol. vi., Plates 29. 30, 31. 
 
WINCHESTER CATHEDRAL, NAVE FROM S.W. 
 
;he i'-' 
 
FAN VAULTING. 343 
 
 a short step for Abbot Horton of Gloucester (1351-1377) to put a fan vault over 
 the portion of the east walk of the cloister which he built — viz. from the nave to 
 the door of the chapter house. 
 
 In the Tewkesbury monument we have but a suggestion of fan vaulting. 
 Apart from any such suggestion, however, the fan vault had become inevitable. 
 For a long time, without noticing it, the builders had been working towards an 
 inverted conoid as the form of the lower courses of their sheaves of ribs. In 
 the early vaults, <?.£-. in Peterborough aisle, and the eastern' CRVI'T ok CANTER- 
 BURY (334), the springing of the vaults has decidedly an ugly appearance, 
 because the long diagonal ribs leave the abacus at a different angle to that of 
 the short transverse ribs. Various expedients were adopted to remedy this ; t\t, r . 
 the diagonal might be made to spring further back on the abacus, or the trans- 
 verse rib spring further forward ; or the true curve of the ribs was tampered 
 with.* Now the more successful these and such-like expedients were, the more 
 the springing of the vault approached the form of an inverted conoid. The 
 spandrel of the vault, which had at first been rectangular as in 344.1, had reached 
 the form shown in 344.2, and was then far on the way to the semicircular 
 spandrel of fan vaulting 344.3). 
 
 A more scientific origin, however, is usually claimed for the fan vault. Each 
 fan is held to be due to the revolution of a four-centred arch half-way round 
 its vertical axis. Hitherto each rib had been half a pointed, semicircular, or 
 segmental arch ; and therefore had a simple curve. Now each rib, it is said, had 
 come to be half of a four-centred arch, and therefore was a two-centred rib ; its 
 curve was a compound of two curves. Each of the new ribs consisted of two 
 curves ; the lower curve, rising from the abacus, very short ; the upper curve 
 very long ^see 258). These two-curve ribs were made of different lengths, but 
 of precisely the same curve. It follows that if the whole of the sheaf or group 
 of ribs, where it rises from the abacus, have the same curve, the sheaf will be 
 semicircular in plan ; in fact it will be like the lower part of a trumpet. In 
 344.4, from ALL SOULS', OXFORD, the two-centred form of the ribs is clearly 
 seen on the left. But there is a serious objection to this theory. It is that the 
 earliest fan vault of importance, that of Gloucester cloister, is not constructed with 
 four-centred, but with pointed arches ; as also is that of Sherborne nave. It 
 would seem, therefore, that though the four-centred arch facilitated the con- 
 struction of fan vaults, it did not suggest them. We may add that to think that 
 theorising about mathematics or geometry led the old men to improvements in 
 building construction is to look at past times through modern spectacles. 
 Building was not taught in those days in polytechnics or science and art 
 classes; the masons learnt to improve by making blunders and having to 
 correct them.+ 
 
 As we have seen, in such examples as the lierne vault of HEREKORD 
 SOUTH TRANSEPT (S33)^ tne builders had approximated so closely to the form 
 of the conoids of the new vault, that fan vaulting resulted inevitably. But the 
 two types of vault are distinct. In the lierne vault the ribs which rise from the 
 
 * See diagrams in Willis' Vaulting, 65, showing the differing plans of the spandrels of the 
 vaults in Norwich cloister, executed at different periods, 
 t Cf. Brutails, 3 ; Prior, 88. 
 
344 
 
 FAN VAULTING. 
 
 abacus are of different lengths, and do not terminate in horizontal lines. But 
 in a fan vault they are all crossed by horizontal lines, and each piece of rib 
 between these lines is of the same length. And the solid spandrel of the fan 
 vault is an inverted conoid ; whereas that of the heme vault only approximates 
 to that form.* 
 
 It is usually said that the fan vaults are confined to England. Examples, 
 however, occur in the Baltic lands. 
 
 Various combinations of the conoid may be arranged ; whole circular 
 conoids or quadrants, or semicircles, e.g. all along a corridor, such as the walks 
 of GLOUCESTER CLOISTER (344) or LAVATORY (334), or Peterborough retrochoir, 
 we may arrange on either side a series of semicircular conoids, meeting one 
 
 4. All Souls', Oxford. 
 
 Springing of Vaults. 
 
 5. Westminster South Transept Window. 
 
 another in the centre ; to the eye the effect will be that of a series of pairs 
 of semicircular corbels, each pair carrying a small section of stone ceiling. Or, 
 as in the passage leading into the gardens of St John's College, Oxford, 
 we may have a small quadrant in each corner, and a very large spandrel. 
 In the Dean's Chapel, Canterbury, there is a quadrant, semicircle, quadrant, on 
 each side. In the Central or Bell Harry Tower of Canterbury four quadrants 
 and four semicircles are arranged alternately. Across Henry the Seventh's 
 Chapel at Westminster are semicircle, circle, circle, semicircle. In the very late 
 STAIRCASE OF CHRIST CHURCH, OXFORD, 1640 (348) the space to be vaulted 
 
 * The high vaults of Oxford Cathedral and St George's, Windsor, and that of the Divinity 
 School, Oxford, are frequently, but incorrectly, described as fan vaults. 
 
GLOUCESTER CATHEDRAL, THE CLOISTERS. 
 
FAX VAULT IXC. 
 
 345 
 
 is treated as if it were a square chapter house ; the centre having a tall column 
 carrying a whole conoid, while round the walls are arranged quadrants and half 
 conoids projecting forward till they meet the central conoid. It is just the 
 vault of Canterbury central tower plus a central pier carrying a conoid. 
 
 The treatment of the spandrel between the fans varies. In Gloucester 
 cloister it is flat. More often it is slightly arched ; it is necessarily so, when, as at 
 SHERBORNE (333), the ribs do not stop at the horizontal rim of the fan, but are 
 continued into the spandrel. It would be feasible to erect a little dome over the 
 space between the fans. Sometimes there is a large boss, as in the Salisbury 
 chantry at Christ Church, Hants. Or, as in Bishop Alcock's chantry at Ely, 
 the boss may be enlarged into a long pendant, which weights and steadies the 
 vault. If the pendants are very long and heavy, they may be constructed as 
 hollow tabernacles, as in OXFORD cathedral choir (297). 
 
 The construction of the fans also varies. Sometimes the space between 
 the ribs is filled by panels, except where the tracer)' requires to be sunk out 
 of larger stones; e.g. at SHERBORNE* (346). Sometimes, even in the earliest 
 fan vaults of Gloucester cloister, the vault is all rib ; what seems to be a panel 
 is really part of one of the ribs on either side. It was found to be less trouble 
 to cut a thickish block to be rib and panel at once than to cut rib and panel 
 separately. When they are cut separately, the joints run alongside the ribs. 
 When they are cut out of the solid, the joints run down the centre of the web; 
 e.g. in CLOUCESTER CLOISTER (344J. Or, thirdly, in Henry the Seventh's 
 Chapel at WESTMINSTER (34/j, the vault, as may be seen by examining the 
 back of it, is all panel, no ribs at all ; except decorative ribs carved on its 
 under surface. Most often, the first and second systems are combined ; the 
 broader spaces being closed by panels ; while, in the narrower, rib and panel are 
 carved out of the solid ; e.g. in the eastern chapels of Peterborough. 
 
 As for the decorative system of such a vault as that of Henry the Seventh's 
 Chapel, the artist had a free hand. What he did, however, was to reproduce on 
 his conoids the familiar tracery of a window of rectilinear type. And so it 
 came about, by a strange reverse of fortune, that the vault, which hitherto had 
 been the dominant member of a Gothic building, marshalling its obedient array 
 of piers, bases, capitals, buttresses, flying buttresses, and pinnacles, in the last 
 days of Gothic fell under the subjugation of the window. In Tudor days it was 
 the window that was the dominant note. The church was, and was meant to be, 
 a Lantern church. Rectilinear window tracer)' had descended from the lofty 
 clerestories of r.LOUCESTER CHOIR (59; and WINCHESTER nave (90), and had 
 panelled triforium and pier-arcade and walls ; had passed outside, as in the 
 Beauchamp Chapel, Warwick, to panel buttress and wall and battlement alike; 
 had panelled great towers at Evesham, Boston, Wrexham. Its last triumph 
 was to panel the fan vault. It was but to acknowledge facts. The craftsman in 
 glass had been a more important person than the mason for a century or more. 
 
 A remarkable feature about the fan vault is that it has little thrust. At 
 SHERBORNE (376) the choir was vaulted first, and flying buttresses were erected. 
 But in vaulting the nave the)- were omitted ; the vault merely being given a more 
 
 * Carpenter m Journal of R.I.B. A., 1S77, 145. 
 
346 
 
 FAN VAULTING. 
 
 acutely pointed section (576). Still more daring is the southern outer aisle or 
 chantry of Collumpton Church (1510-1528); where, on the north side, the fan 
 vault rests on the piers between the two aisles. The inner aisle and nave are 
 unvaulted ; so that there is no counter-thrust. In Cirencester an inner aisle has 
 a fan vault, although both the central aisle of the choir and the outer aisle have 
 wooden roofs. Unfortunately, Gothic architecture came to an end before the 
 builders had realised the full value of the new construction, enabling them to 
 dispense straightway with the cumbersome apparatus of external stone scaffold- 
 ing which had been the curse of mediaeval building. One sees, however, timidity 
 and hesitation in adopting such a revolutionary step as the abolition of the flying 
 buttress. All Saints', Maidstone, which was practically complete in 1395, was 
 planned with broad and lofty aisles to the nave, and narrow and low ones to the 
 
 Sherborne Nave. 
 
 choir. The choir aisles were designed for a fan vault. But the time was not 
 ripe for the adoption in Kent of the West of England improvement ; and the 
 choir aisles were raised and covered with a wooden roof. So, at Rochester, a 
 nave of three bays was added to the Lady Chapel, in just the same position as 
 the Collumpton chantry : but the intention of covering it with a fan vault was 
 abandoned ; probably it was regarded as too risky, as two sides of the chapel 
 were open arches.* Double the Oxford staircase (348) ; provide two central 
 piers and two conoids instead of one of each, and we have the original 
 Rochester design. Another sign of distrust is the construction of massive 
 transverse arches, as at KING'S COLLEGE CHAPEL, CAMBRIDGE, 1512-1515 (62), 
 passing up the centre of each half conoid and spanning the nave. 
 
 * See Hope's Rocheste?; 87. 
 
FAN VAULTING. 
 
 347 
 
 The suggestion of these transverse arches seems to have come from Oxford, 
 where they were employed in two superb Heme vaults of novel character. The 
 first is that of the DIVINITY SCHOOL (331), begun c. 1445 and 
 finished c. 1480. It is a long and very broad room. The obvious 
 way to vault it was to divide it into nave and aisles, as here shown, 
 and then to ceil these with three vaults; e.g. as in the choir 
 of the Temple Church. But the eight piers would have been 
 obstructive ; moreover there would have been no credit to be got 
 for merely repeating what had been clone before in scores of 
 churches. Imagine, however, that it had actually been vaulted in 
 three divisions of nearly equal height ; and that each of the piers 
 carried a capital. Would it be possible to retain the capitals, while 
 withdrawing the piers, without bringing the three vaults down ? 
 What would a modern contractor do if he were asked to leave the vaults of 
 the Temple choir (35.1 standing, while withdrawing the supports? Obviously, 
 from A in one wall to E in the opposite wall, he would insert an iron girder ; 
 and if this girder passed under the capitals of the piers 1, J, he could then 
 take away the piers. So with the other piers. What a modern contractor 
 
 Divinity School. 
 
 Henry the Seventh's Chapel 
 
 would do with iron girders, that the Oxford builder dirl with strong transverse 
 arches of stone. The pendants of each transverse arch, which represent the 
 imaginary capitals, are but two of the voussoirs of the arch, greatly elongated. 
 This beautiful novelty was imitated soon after in the vault of the CATHEDRAL 
 
34§ 
 
 FAN VAULTING. 
 
 CHOIR, c. 1478-1503 (331, 297); where each compartment of the high vault is 
 similarly divided into nave and aisles ; the latter ceiled with segmental tunnel 
 vaults, the former with a Heme (stellar) vault. Compared with the complex 
 loveliness of such a vault, the monotonous uniformity of fan-design is tame 
 indeed. 
 
 In one great fan vault, however, that of Henry the Seventh's Chapel at 
 WESTMINSTER (347), by developing further the Oxford improvement, the builder 
 was enabled to break away entirely from the somewhat monotonous repetitions of 
 the fan vaults of Sherborne, Bath, and King's College, Cambridge. This famous 
 vault, the masterpiece of English masonry, the wonder of foreign lands, is but 
 the Heme vault of the Divinity School with certain modifications. First, it is 
 executed in fans. Secondly, from the bottom of each of the two pendants in 
 
 Staircase of Hall of Christ Church, Oxford. 
 
 each transverse arch is built up a complete circular conoid. Thirdly, each 
 transverse arch passes through the vault, and consequently its upper portion, as 
 may be seen in 347, is invisible from below ; so that from below the astonishing 
 effect is produced of inverted conoids resting on pendants, which themselves rest 
 on nothing but the unsubstantial air. Fourthly, as is seen in 347, the vault is 
 purely a panel one : constructional ribs do not exist in it* And so, in the most 
 amazing way, all in a moment the history of the English vault has been 
 revolutionised. It began in the far-away eleventh century, as a vault all web, 
 and no ribs. In the aisles of Durham, in 1093, diagonal ribs were added ; at 
 Lincoln more ribs still ; ridge ribs and tiercerons. More and more the ribs multi- 
 
 * So also there are no ribs on the extrados of the fan vault of Wells central tower ; nor on 
 the central part of the vault of the nave of ST GEORGE'S, WINDSOR (341). 
 
FAN VAULTING. 349 
 
 plied ; more and more close-set the rib structure became, till a single flat panel was 
 enough to bridge the rib openings. Then, in Gloucester cloister, even the panel 
 disappeared ; the vault became one solid mass of ribs. Finally, in the moment 
 of its triumph, the rib is struck down ; and the vault becomes a thin shell of 
 panels ; the most consummate achievement of the masonry of the Middle Ages ; 
 put together with as unerring science and precision as the parts of a steam- 
 engine or an astronomical instrument.* 
 
 * Truly, if vault construction is the be-all and end of Gothic architecture, it is not with any 
 foreign country, but with England, that the artistic supremacy in mediaeval architecture rests. 
 From first to last — in the vaults of Durham nave, of Lincoln, of Exeter, of the choir of Oxford 
 Cathedral, of Henry the Seventh's Chapel at Westminster — we were the envy of less happier 
 lands. We English are too modest in claiming our artistic dues. 
 
Chapter XXIII. 
 
 Part I. 
 The Buttress. 
 
 Functions, Dimensions, and Origin of the Buttress — The Columnar Buttress — The 
 Pilaster Strip — Gothic Buttress — The Diagonal Buttresses — Stages of Buttress — 
 Omission of Buttress — Internal Buttresses. 
 
 A BUTTRESS is a thickening of a wall at intervals by projecting masonry. It 
 has economical, constructional, and decorative value. A wall of 4 feet in thick- 
 ness, with buttresses projecting 3 feet, is much stronger and far cheaper than a 
 wall of 6 feet thick without buttresses* Moreover, the wall may help to carry 
 a wooden roof; and the tendency of the roof to spread and to force the wall 
 out has to be guarded against by buttresses applied to the walls where the 
 rafters may bring down pressure. And if the wall be pierced with windows, 
 especially if they be large windows, so that the wall perhaps is nearly all glass, 
 as in the Tudor Lantern churches, all the more is there need of strong buttressing. 
 Yet more is it needed if the roof is composed of arches loaded with masonry 
 — i.e. a vault. 
 
 If indeed the vault be an unribbed barrel vault, like that in St John's 
 Chapel in the TOWER OF LONDON (283), the thrust of the vault is exerted 
 against the whole length of the wall, and continuous abutment is required ; 
 intermittent abutment would be inadequate ; it is necessary to thicken the whole 
 wall. But if the vault be composed of a series of arches carrying flagstones, 
 as at MINCHINHAMPTON (287), then only intermittent abutment will be needed. 
 The usual vault, however, is not a barrel, but one with intersecting groins or 
 intersecting diagonal ribs. Therefore at certain points ; e.g. in 308.3, at the 
 point B, half of the pressure of the vault of the aisle is brought down by 
 the diagonal BE, and the outer arches BD, AB ; and similarly half of the pressure 
 of the adjoining vault. In the same way, in 324.7, the pressure is brought down 
 to the point B by the diagonals EB, WB, the transverse rib TB, and the tiercerons 
 LB, MB, NB, OB, PB, QB, and the wall ribs up,, RB. At these points, D or B, 
 therefore, great pressure is brought against the wall, and very strong buttresses 
 are needed to withstand it. On the other hand, between the points from which 
 the ribs of the vault spring ; e.g. between BA and BX in 324.7, there is little 
 pressure, and the intervening wall may be replaced, if desired, by mullioned 
 
 * Pugin's True Principles, 3. 
 
THE BUTTRESS. 351 
 
 windows. This, then, is the main use of the buttress in mediaeval architecture; 
 to provide resistance at intervals, and not continuously.* 
 
 The dimensions or mass required for a buttress will depend on several 
 considerations. A tall wall will need more abutment than a low one ; a thin 
 wall than a thick one ; a badly built wall of rubble more than a well-built wall 
 of ashlar. Again, the greater the amount of glass, the greater must be the 
 mass of the abutment ; and if there be inside the wall a vault, yet more strength 
 must be given to the buttress. Again, a vault which springs high up on the wall 
 requires more abutment than one which, as in GLOUCESTER NAVE (26) or 
 HEREFORD LADY CHAPEL (464), springs low down. A vault which is of flattish 
 curve, like that of WINDSOR (341) or Ely Lady Chapel, has greater lateral 
 pressure than an acutely pointed vault, like that of Wells, which has diagonal 
 as well as transverse ribs pointed. Xo mathematical theory can take into 
 account all these and other differences ; and it was doubtless purely by 
 empirical methods that the old builders estimated the height, breadth, pro- 
 jection, and shape required for their buttresses. + 
 
 Roman buttresses occur in the apses of the Temple of Minerva Medica 
 at Rome ; in S. Vitale, Ravenna, and S. Lorenzo. Milan, both sixth century, 
 according to Rivoira ; in S. Ambrogio, Milan, and the Baptistery of Biella, 
 ninth century. In Syria buttresses are frequent from the fifth century. Africa 
 had churches from the fourth century with vaulted aisles and buttresses. In 
 our own country, the churches of Ythanchester (Essex; ; Canterbury St Martin 
 and St Pancras ; and the apses of Brixworth and Wing, all of which are ascribed 
 to the seventh century, have buttresses. They do not taper upwards ; and at 
 Ythanchester and St Martin's have sloping heads of brick in horizontal courses. 
 At Ythanchester they project 2 feet. J But after that date the Anglo-Saxon 
 buttresses are merely pilaster strips. Plainly the Roman buttress went out 
 of use in England, and had to be reinvented in the twelfth century. We must 
 resist the temptation to derive from the Roman buttress that of English 
 Romanesque and Gothic. 
 
 Frequently in Northern Italy, e.g. S. Abbondio, Como ; and in France 
 south of the Loire ; the column is elongated and made to serve as a buttress, 
 especially to an apse. It is highly decorative, but almost wholly useless. It 
 occurs in Normandy in the apse of St Gabriel. We have a few examples ; the 
 Norman tower of Northampton St Peter's has three columns at each angle ; 
 Norman examples occur in the aisle of Ely and in the clerestory of Peterborough ; 
 Transitional in Leicester St Mary's. It is rare in Gothic ; but occurs in the Ely 
 galilee. 
 
 The Norman buttress is descended not from the Roman buttress, but from 
 the Roman pilaster strip, employed in S. Apollinare in Classe, Ravenna, in the 
 sixth century, in Rome at S. Balbina A.D. 600; and S. Pudentiana, eighth 
 century. This pilaster strip in turn may be the descendant of the classical 
 pilaster, employed instead of a column to cany an entablature. In our Anglo- 
 Saxon work it seems to be a mark of late date ; e.g. at Barnack, Bradford-on- 
 
 * See also the illustrations of the naves of NORWICH and WINDSOR (330, 332). 
 + See Viollet-le-Duc, Architecture, iv. 63, for suggested method. 
 I Peers in Archceological Journal, vol. 58, p. 431. 
 
552 
 
 X 3£ 
 
 £>I 
 
 
35: 
 
 i.33J r 
 
 S1IVX3Q 2JOJ 33VD9 
 
 ~i 1 1 
 
 Si 
 
 :q ,91 
 
 -.C2 
 
 &£a- 
 
554 
 
 SCALE OF 
 
 Guisborousrh. 
 
 St Patrick's, Dublin. 
 
355 
 
 12 1 b } 
 
 Pucklechurcli. 
 
356 
 
 12 fc O 1 
 
 iL^kLi^feiiki^ 
 
 
 Austrey. 
 
 St Neots. 
 
iS7 
 
 Cottingham. 
 
 Frampton. 
 
358 THE BUTTRESS. 
 
 Avon, Earl's Barton, Sompting. It is common in the eleventh-century work 
 of Normandy and England ; e.g. STEYNING (359). 
 
 These pilaster strips, rising only to the eaves or short of the eaves, when 
 attached to the massive walls of Norman architecture, can have little con- 
 structional value. Artistically they give relief and intermittent shadows to 
 the flat surface of an unbroken wall ; they serve also to define externally the 
 limits of the bays into which the interior is divided. But when the walls are 
 thin and are built of rubble, as in most Anglo-Saxon work and in some 
 Norman, e.g. the aisles of Carlisle and Rochester, they have a real con- 
 structional value ; for they act as binding-courses to stiffen the wall. In 
 no case are they of use as buttresses. It is to be noticed that where ashlar 
 was scarce, or had to be imported, e.g. from Caen, the windows are some- 
 times set not in the walls, but in the buttresses, as at Old Shoreham, to economise 
 the imported stone. 
 
 Soon, however, the pilaster strip, instead of being broad and thin, became 
 square, as at Birkin ; and in such Transitional work as Glastonbury Chapel is 
 converted into a buttress. The next step is to give it more depth than breadth ; 
 as in KIRKSTALL CHAPTER HOUSE, c. 1160 (359). A good many village 
 churches retain this simple squat buttress in the thirteenth century, not divided 
 into retreating stages. It occurs at Pembridge in the fourteenth century. 
 
 To decorate the Norman buttress, slender shafts, with cap and base, were 
 sometimes attached at each angle ; e.g. outside NORWICH TRANSEPT, c. 1096 
 (31); in Rochester nave; and outside LINCOLN CHOIR and TRANSEPTS 
 (66, 484). Very effective, too, is the sharp chamfer of LINCOLN NAVE (115), 
 as if cut with a knife ; imitated in SOUTHWELL CHOIR (opposite). This 
 chamfering of the edges of the buttress is common in the thirteenth century. 
 A few examples occur later. Another note that distinguishes Romanesque from 
 Gothic buttresses is that the former never terminate in a gable. But the 
 special distinction of the late Transitional and Gothic buttresses is that they 
 are usually divided into stages, each drawn back more than the stage next 
 below ; so that the buttress as a whole tapers upward. In the first half of the 
 thirteenth century the aisle-buttress often has but two stages,* as at ST PATRICK'S, 
 DUBLIN (354), LITTLE WENHAM (352) ; later, three stages are more common ; 
 but there was no period at which the buttress of two stages was not employed, 
 e.g. AUSTREY (356). 
 
 Of the buttresses of the period 1245 to 131 5, a good example is seen at 
 warmington, c. 1260 (352); and richer ones at MERTON COLLEGE chapel, 
 OXFORD (352), GUISBOROUGH CHOIR (354), and GLOUCESTER NAVE (360), where 
 the niche has the characteristic straight-sided gable. 
 
 Later differ from Early Gothic buttresses in that the moldings may contain 
 undulatory ogee curves ; and that where a niche is introduced it usually has an 
 ogee arch for head, and crockets of bulbous foliage, as at ST NEOTS (356) and ST 
 MARY MAGDALEN CHURCH, OXFORD (353). It must be remembered, however, 
 that plain unmolded set-offs may occur in any period ; e.g. at HINGHAM, 1316- 
 1359 (489); GIMINGHAM (576); PETERBOROUGH (365), depriving us of a valuable 
 criterion of date. Usually the later buttress is marked by the imposing molded 
 
 * Salisbury, however, and others have three. 
 
;S9 
 
 Steyning. 
 Southwell Choir. 
 
 Kirkstall Chapter House. 
 Hedon Nave. 
 
360 
 
 THE BUTTRESS. 
 
 basement courses of the period carried round its foot ; e.g. at HAWTON (355) 
 and FRAMPTON (357). Many of the fourteenth-century buttresses are singularly 
 simple and sober in design ; e.g. HEDON (359) ; this was the golden age of 
 composition ; when the best men anxiously avoided the crying sin of Gothic 
 over-elaboration and exuberance of ornament. Hence, while some buttresses 
 
 Gloucester, S. Aisle of Nave. 
 
 of the period, e.g. those of the nave of Beverley Minster, are rich in the 
 extreme, many, such as those of SELBY CHOIR (86), Madley, and Patrington, 
 are kept studiously simple, that all the ornamentation may be concentrated 
 at one zone, and that the uppermost — in the parapet and pinnacle. It is the 
 same feeling that supplied the pinnacled spires of PATRINGTON and OXFORD 
 
THE BUTTRESS. 
 
 |6i 
 
 ST MARY'S (634, 631; with a plain tower as basement and foil. Viollet-le- 
 Duc notes the same tendency to sobriety and restraint in the fourteenth- 
 century buttress of France. 
 
 The later buttresses differ but little from those of the fourteenth century, 
 except that they tend to become thinner (and therefore to project further , e.g. 
 in the Divinity School, Oxford, and KING'S COLLEGE CHAPEL, CAMBRIDGE 
 (199), in order to leave as much room as possible for the threat windows of the 
 period. In the richer examples the presence of panelling or of a very acutely 
 pointed incurved canopy, as at Aylsham, is significant. Examples are ROX- 
 
 WELL, C. I4OO (353); STRATFORD, SUFFOLK (353); LAVENHAM (^76) \ the 
 
 inserted buttresses at Aylsham, Norfolk, and pftfrki irough RFTROCHOIR 
 
 At the angles of a church, or of its tower 
 or porch, it was usual to place a couple of 
 buttresses at right angles to one another ; 
 and this disposition was never wholly aban- 
 doned. But from the fourteenth centurv it 
 was also common to employ one buttress only 
 at the angle, but to set it diagonally, as in 
 TF.RRINGTOX ST Clement's tower, aisle, and 
 porch (92). The angle buttress, set diagon- 
 ally, may occasionally be found even in the 
 thirteenth century ; e.g. in the tower of Pole- 
 brook Church ; * also at Warmington and 
 Morton Pinkney. In France it is employed 
 almost invariably from the beginning of the fifteenth century, but hardly ever 
 
 till thrll.-'- 
 
 As we saw above (358), the early buttresses usually rose vertically, 
 undivided into stages, as in 367.1. But as the thrust passes diagonally down- 
 wards to the point Ii, the masonry which occupies the triangle BDA seemed, to 
 a large extent, unnecessary. In 367.2, it is supposed to have been removed. 
 In such a buttress, however, it would be necessary to protect the joints from 
 the infiltration of rain all the way from D to 1;. It is better construction, there- 
 fore, to construct as much of the buttress as possible in vertical stages, uniting 
 its different stages by short slopes ("set-offs") nil, FG, which can be protected 
 from the weather. This is an artistic as well as a practical improvement ; 
 and involves the use of but little more masonry. 
 
 the ancle OF SLOPE of these weatherings or set-offs varies according 
 to the distance from the eye. A set-off high up slopes much more than one 
 near to the ground. The higher the buttress, the steeper will be its upper 
 set-offs fsee above). If this is not attended to, parts of the upper stages of 
 the buttress will be hidden by the too flat set-off, and to the eye the buttress 
 will lose in height. Nevertheless, at Netley Abbey, the lower set-offs are 
 steeper than the upper ; this may be scientifically correct, because the line of 
 pressure tends to become more and more vertical as it approaches the ground. % 
 
 Slope of Set-offs. 
 
 * Illustrated in Rickman, 142. 
 } Sharpe's Parallels, Plate 44. 
 
 + Choisy's Histoire, ii. 29S. Cf. page 131. 
 
362 THE PINNACLE. 
 
 In Anglo-Saxon churches, except in some of seventh-century date, 
 buttresses are dispensed with ; so also in several of our Norman churches, 
 e.g. Abinger, Gillingham, Kippax ; and in Normandy, Thaon. They are absent 
 even in Northampton St Peter's, though the aisle is crossed by transverse arches 
 (236) ; and in the unaisled choir of Creully, Normandy, though it has a quadri- 
 partite vault of two bays, the side walls have no buttresses. At Ensham, near 
 Oxford, the fourteenth-century choir has no buttresses ; but its walls are 3 feet 
 thick. Some towers also, e.g. Southfleet and Barnwell, are without buttresses, 
 and many churches in chalk districts, e.g. Tangmere, Sussex, where freestone had 
 to be brought from a distance. 
 
 In Roman construction, e.g. in the BASILICA OF MAXENTIUS (290), the outer 
 walls, instead of being set inside, were set outside the buttresses. Thus the but- 
 tresses became internal instead of external. Much additional space was thus 
 gained, simply and cheaply. And when the buttresses were pierced with arches, 
 as in the Basilica of Maxentius and the Cordeliers' Church at Toulouse, a 
 continuous aisle was obtained. This improvement we seldom adopted in 
 England ; * probably because our buttresses had much less projection than 
 those of France, and hardly lent themselves to such a treatment. 
 
 Part II. 
 
 The Pinnacle. 
 
 Pinnacles may be divided into two classes : those which rest on the flanking 
 buttresses of the side walls of a church, whether aisle-walls or clerestory walls : 
 and those which are set over the angle where two walls meet ; e.g. at the corners 
 of a nave, a tower, a polygonal chapter house. 
 
 Corner Pinnacles. — The latter were employed on a large scale from 
 the first. Norman or Semi-Norman pinnacles occur at the corners of Peter- 
 borough choir ; at Oxford St Peter's, where they are circular ; at BREDON and 
 BISHOP'S CLEEVE (363), where they are square ; in Rochester west front, 
 where they are octagonal ; in GLASTONBURY LADY CHAPEL (465) ; over South- 
 well porch. In the thirteenth century a shafted octagon is a favourite form ; 
 e.g. in LINCOLN EASTERN (66) and CENTRAL (69) TRANSEPTS and PRESBY- 
 TERY (177), and SALISBURY (458). In the east end of GUISBOROUGH (354) is 
 a massive combination of gables and pinnacle ; in the west end of Howden 
 the pinnacles become tabernacles of open work. The eastern pinnacles of 
 SELBY (86) are reminiscent of those of HOWDEN (72) and Guisborough. Fine 
 pinnacles crown the angles of Canterbury central transept, LOUTH STEEPLE 
 (6ll), WREXHAM TOWER (609), ST GEORGE'S, WINDSOR (492), and BEVERLEY 
 ST MARY (366). 
 
 Flanking Pinnacles. — Though, as we have seen, the pinnacle was 
 employed early to weight the angles of a building, it is a curious fact that both 
 
 * Except in KING'S COLLEGE CHAPEL, CAMBRIDGE (199). 
 
THE PINNACLE 
 
 ;'.; 
 
 here and in France it was long before it was recognised that it was of value in 
 weighting the lateral buttresses also ; indeed, of exceptional value, where there 
 are flying buttresses. Great cathedrals, like Bourges, with tier upon tier of 
 flying buttresses, had at first no pinnacles. The early practice was to give 
 a flanking buttress no finial except a saddle-back roof to keep the rain 
 out of the joints. This then was the first step towards the pinnacle ; a mere 
 gable; e.g. in Whitby choir, in York transept, and the chapel of MERTON 
 COLLEGE, OXFORD (473). But in 
 LINCOLN CHOIR and NAVE ( 1 1 5 ), 
 and still more in SOUTHWELL CHOIR 
 (359)i 't begins to be recognised 
 that the gable may be more than a 
 roof-covering ; it may be enlarged and 
 heightened so much as to act as a 
 weight also. The diagram below shows 
 this use of it. C is supposed to be an 
 arched spring (which we may imagine 
 to be loaded above, just as the arches 
 of a vault are loaded by the masonry 
 of the cells), with a thrust equivalent 
 to 12 in each direction. W is a wall 
 and Ii its buttress. WB together have 
 an inert resistance of 10; therefore!' 
 will push over WB. But if on B we 
 put A whose weight is equivalent to 
 3, A and WB together have a strength of 13, and are able to stop the outward 
 thrust of C. The addition of the pinnacle to the flanks of churches was one of 
 the few contributions made by Gothic to the art of mediaeval building con- 
 struction. Almost everything else was invented in Romanesque days ; for the 
 most part Gothic art had but to develop and make beautiful what Romanesque 
 genius had invented. 
 
 Before the fourteenth century, pinnacles on flanking buttresses are rare. 
 Perhaps the earliest were those (now disappeared) which 
 formerly crowned the curious polygonal buttresses of the aisle 
 added to Chichester nave* At WESTMINSTER (379J the pin- 
 nacles are restorations ; but probably they reproduce pinnacles 
 of 1245; which the builders would hardly fail to introduce in 
 imitation of Amiens, which had pinnacles c. 1230. At the end 
 of the century flanking pinnacles appear in EXETKR CHOIR 
 (377), where they are square with crocketed edges. At the beginning of the 
 fourteenth century the)- appear in York NAVE (366;, SELBY CHOIR (86), 
 MALMESBURY NAVE (375), and in the chapter houses of Southwell, Wells, 
 Lincoln, and York. 
 
 In the fourteenth century the pinnacle received considerable development 
 in minor work ; such as the spire-like open-work of the canopied monument of 
 Edward II. at GLOUCESTER (294), of that of Sir Hugh Despenser at Tewkes- 
 * Illustrated in Willis's Chichester, Plate 3. 
 
 Bishop's Cleeve. 
 
364 
 
 THE PINNACLE. 
 
 Scale of 
 
 bury, and of the Exeter sedilia. The 
 characteristic design in minor work 
 of the fourteenth century in the north 
 and east was more solid and massive; 
 e.g. the PERCY TOMB (269), and the 
 Ely Lady Chapel arcade. Having" 
 therefore such a decorative origin, the 
 fourteenth century pinnacles passed 
 without any transition into composi- 
 tions of exuberant beauty — triangular 
 gable, ogee niche, crocketed finial all 
 blending into consummate design 
 even in village churches ; e.g. at 
 HECKINGTON. In SELBY and ELY 
 CHOIRS (86, 365), and in the ruined 
 choir of Howden, the pinnacles are 
 square. In Lichfield choir they are 
 crocketed octagons. In the fourteenth- 
 century clerestory of Lincoln nave, 
 and at Boston, the front of each 
 pinnacle contains a niche ; very 
 beautiful must both have looked when 
 the range of arched pinnacles was 
 peopled with angels, on guard round 
 the sacred walls. Frequently the 
 pinnacle consists of a spirelet rising 
 out of a cluster of gables, as in the 
 CHOIR OF ELY (365) and YORK NAVE 
 and CHOIR (366). 
 
 In late Gothic the pinnacle loses 
 in variety and interest : the tall, thin 
 spike of York choir is only too 
 common. In minor work, and some- 
 times elsewhere ; e.g. in the tower of 
 St Neots and in YORK EASTERN 
 TRANSEPT (199), a slender spike some- 
 times sticks up from the buttress; 
 it may be called the " spear-head " 
 pinnacle. It is usually set diagon- 
 ally ; as in the PORCH OF BEVER- 
 LEY st mary (365). These pinnacles 
 look painfully unsafe. In Tudor days 
 a few interesting variants occur. Ogee 
 cupolas are not infrequent ; as in 
 Winchester choir, Henry the Seventh's 
 Chapel at WESTMINSTER (378), 
 WREXHAM TOWER (609), and King's 
 
 Heckineton. 
 
3^5 
 
 Ely Choir. 
 Peterborough Retrochoir. 
 
 Beverley St Mary. 
 Lavenham Nave. 
 
366 
 
 THE PINNACLE. 
 
 York Minster from S.E. 
 Beverley St Mary from W. 
 
 College Chapel, Cambridge. 
 Sometimes the pinnacles are 
 flat-topped; to carry statuettes 
 of saints ; as in Norwich choir 
 and PETERBOROUGH RETRO- 
 CHOIR (365) ; or of angels, 
 as in Blythburgh nave and 
 Sail chancel ; or of birds and 
 beasts rampant, wrought in 
 iron, as originally in St 
 George's Chapel, Windsor. 
 On some late towers, e.g. 
 TAUNTON ST MARY'S (607) 
 and GLOUCESTER CATHE- 
 DRAL (132), the pinnacle is 
 a square steeple of open-work 
 divided into as many stages 
 as there are stories in the 
 tower. 
 
 So far we have spoken of 
 the constructional value of 
 the pinnacle in weighting a 
 flanking or a corner buttress. 
 But if a parapet or battle- 
 ment is light and open, the 
 pinnacle is of much service 
 in securing it to the wall, 
 as at SELBY (86). It can 
 hardly be maintained, how- 
 ever, that at TIVERTON (390) 
 the pinnacle is anything but 
 a bit of constructed decora- 
 tion. 
 
 Sometimes the presence 
 of a range of pinnacles, such 
 as those of the aisles of 
 SELBY CHOIR (86) or of 
 YORK NAVE, is a sign of an 
 intention, never carried out, 
 to construct a vault of stone 
 with flying buttresses thereto. 
 Both at Selby and York in 
 the end the vaults were put 
 up in wood. Outside the 
 clerestory of Selby choir 
 and York nave there may 
 still be seen the " tushes " ; 
 
THE PINNACLE. 
 
 367 
 
 pinnacle 
 be seen 
 
 °K 
 
 i.e. the projecting courses on which the heads of flying buttresses were to 
 rest. 
 
 From a constructional point of view, pinnacles, being tall detached masses 
 of masonry, are no improvement on the gablets of Lincoln and Southwell choir.* 
 Eor they are exceptionally exposed to the weather ; and indeed in many cases 
 have perished and have had to be renewed again and again. Artistically, too, 
 the row of sharp spikes gives a restless, fussy look to many late exteriors ; far 
 better seems the simple unbroken flow of horizontal line of earl)- thirteenth- 
 century work. 
 
 It was perhaps because the pinnacled buttress was felt to be bad con- 
 struction that the gabled buttress was employed late in the thirteenth century ; 
 e.g. in Tintern ; Lichfield nave ; Hereford north transept ; Howden nave ; 
 Merton Chapel, Oxford ; Bridlington nave ; Lincoln presbytery ; and in the 
 fourteenth century in the choirs of Guisborough, Carlisle, and Beverley St Mary. 
 And in some examples, e.g. at HEDON (359), HOWDEN (72), and HAWTOX 
 (483J, the very construction of the buttress was modified, so that no 
 should be required.! If the following diagram be examined, it will 
 that the triangle of masonry BDA in 1 is omitted 
 in 2, giving the buttress a tapering form. But 
 it is evident that, if retained, it will act as a weight 
 on the lower part of the buttress just in the same 
 way as a pinnacle. There was, as a matter of fact, 
 from about the middle to the end of the thirteenth 
 century, e.g. in the Saint Chapelle at Paris and that 
 at St Germer, a frequent reversion to the upright type 1 a C 2 B c ?" 
 
 of buttress which had prevailed in Romanesque. 
 
 In England it occurs in Hedon nave, and on a large scale in the north transept 
 of HEREFORD (587), and the Lady Chapel of LICHFIELD (369); where the 
 contrast of the choir with, and the Lady Chapel without pinnacles, is very 
 striking. Another motive for the vertical type of buttress would be that the 
 drippings from the gurgoyle above, as at Hedon, would be projected further 
 from the face of the buttress if it were vertical or nearly so, than if it were 
 built in retreating stages. And in a town church like Notre Dame de Dijon, 
 built in a narrow street, the vertical buttress blocked up the street less. Also 
 the superficial area of the church was reduced, and with it the cost of the site. 
 
 The chief justification of the importance given to the late Gothic pinnacles 
 is that the walls were so much weakened by the enormous breadth of the windows 
 that the buttressed piers between the windows needed to be reinforced as much as 
 possible ; hence the array of pinnacles on the naves not only of KING'S COLLEGE 
 CHAPEL (199), where the vault within calls for exceptional abutment, but of 
 HULL (474) and LOUTH '139), where all the roofs are of wood. 
 
 * Loth at SOUTHWELL ^359) and at MALMESBURY (375) the lowering of the pitch of the 
 high roofs has projected the pinnacle against the sky-line ; which was never intended. 
 
 + In the second diagram, BH should be an unbroken line ; BA, AD should be dotted lines. 
 
 ± 
 
Chapter XXIV. 
 
 Part I. 
 The Flying Buttress. 
 
 Definition — Need, Value, and Origin of the Flying Buttress — Internal and External 
 Flying Buttresses — Substitutes — Curve, Inclination, Mass, and Spring of Flying 
 Buttress — Clerestory Buttress — Pinnacle — Design of the Flying Buttress— Flying 
 Buttresses Superposed or in Two Flights — Function of Flying Buttress. 
 
 A FLYING buttress, as usually constructed, is a half arch springing from the 
 buttress of the aisle wall, and abutting the clerestory wall. 
 
 The need of it may be seen on examining LICHFIELD CATHEDRAL (369) 
 at the junction of the unaisled Lady Chapel with the aisled choir. Even with- 
 out being weighted by pinnacles, the projecting buttresses effectually stop the 
 thrusts of the vault of the Lady Chapel. But the aisled choir also has a high vault 
 whose thrusts must be stopped. A carpenter might have kept the clerestory wall 
 from being forced out by putting up a beam sloping from the top of the aisle 
 buttress to the clerestory wall. What a carpenter would do with a beam, the 
 builder has done by a stone bar carried by an arch. Both the arch and the bar 
 act as stays ; but the bar more than the arch ; the main function of the latter is 
 to carry the straight bar. It is very rare to find the bar omitted, as in CANTER- 
 BURY CHOIR (34.3), Hereford choir, and Fotheringhay,* where the flying buttress 
 is all arch. On the other hand, in St Urbain de Troyes,f the central voussoir 
 of the arch is also the central stone of the bar ; and to that extent this flying 
 buttress is more of a bar than an arch. 
 
 The flying buttress, however, is not confined to clerestory walls within which 
 is a vault of masonry. In Winchester ch©ir, remodelled by Bishop Fox, 1500- 
 1528, there are elaborate flying buttresses, though the vault is of wood; 
 perhaps, as at York and Selby, the original intention had been to vault it 
 in stone. So there are in the parish church of Fotheringhay, although it 
 has a roof of wood, and an elevation of but 40 feet. The vast hammerbeam 
 roof of Westminster Hall j also has flying buttresses. On the side of Worstead 
 Church and of TERRINGTON ST CLEMENT (92) is a single flying buttress, 
 set where a pier has settled outwardly. At Rye and Tewkesbury are flying 
 buttresses set against a wall, just as a farmer sets a prop against a haystack, 
 
 * Illustrated in Glossary, i., Plate 43. + Illustrated in Viollet-le-Duc, Architecture, i. 76. 
 I Illustrated in Pugin's Specimens, i. 32. 
 
ORIGIN OF FLYING BUTTRESS. 
 
 369 
 
 where the wall has bulged out. At Glastonbury, instead of building inside the 
 cloister obstructive buttresses outside the wall of the vaulted aisle of the nave, 
 flying buttresses were constructed in the fifteenth century over the roof of the 
 cloister ; flying buttresses are used in a similar position in Exeter and CHESTER 
 Cathedrals (522). 
 
 As to the origin of the flying buttress, it has been held * that it may 
 have been suggested by a half-barrel roofing an aisle. Imagine a ribbed half- 
 barrel such as that in the upper aisle of GLOUCESTER CHOIR (282,1; retain 
 the ribs, but omit the 
 half-barrel, and we have a 
 series of flying buttresses. 
 But where the flying but- 
 tress was employed earl}-, 
 i.e. in England and Nor- 
 mandy, half- barrels are 
 exceedingly rare. And 
 secondly, where they are 
 exceedingly common, e.g. 
 in the upper aisles of the 
 Romanesque churches of 
 Auvergne, the flying but- 
 tress appears exceedingly 
 late. So far as Normandy 
 and England are con- 
 cerned, a much more prob- 
 able origin of the flying 
 buttress is to be found 
 in the transverse arches 
 which crossed the triforium 
 of several of our vaulted 
 churches. Semicircular 
 arches still span the tri- 
 forium of DURHAM CHOIR 
 (370), begun 1093. At 
 CHICHESTER (34.4) they 
 seem to have been removed 
 when the cathedral was 
 re-roofed and vaulted and 
 
 provided with flying buttresses after the great fire of 1 1 $6. The ragged ends of the 
 old arches of the work of 1091 still remain in the triforium chamber. Similarly 
 transverse arches, but not semicircular ones, appear in the triforium of the choir 
 of CANTERBURY, c. 1175 (34.3); and of LINCOLN CHOIR, c. 1192(34.5).! Both 
 at Chichester and Durham the arrangement apparently was felt to be unsatisfac- 
 tory, for the transverse arch abutted the clerestory wall far too low down to 
 
 * Viollet-le-Duc, Architecture, i. 61. 
 
 + So also in Bayeux choir ; at Narbonne and Bari, and in the transepts of St Pierre-sur- 
 Dives and St Quentin. 
 
 2 A 
 
 Lichfield from S.E. 
 
3/0 
 
 INTERNAL FLYING BUTTRESS. 
 
 enable it to resist the thrust of the vault. At CHICHESTER (34.4) the transverse 
 arches were removed, and were rebuilt as shown in the section on the right. 
 At DURHAM (34.2), when the nave was built (it was finished in 1 133), instead of 
 the semicircular arches employed in the choir, the builders built half arches, 
 which are still standing, and are shown below. Still earlier, in the triforium 
 of Norwich choir, begun in 1096, arrangements were made for flying buttresses 
 beneath the roof* Along both sides of the triforium chamber of nave and 
 choir and apse are supporting shafts ; short ones along the wall, tall ones at 
 the back of the arcade of the triforium. f Internal Romanesque flying buttresses 
 occur also at the Abbaye-aux-Dames; and c. 11 50, in the upper aisle of St 
 
 Durham Choir Triforium. 
 
 Durham Nave Triforium. 
 
 Germer near Beauvais ; at St Gabriel % (Calvados) ; and Creil, § near Paris. All 
 these are genuine flying buttresses ; and plainly they were developed by the 
 Anglo-Norman builders at a very early period. For the invention of the flying 
 buttress — which next to that of diagonal ribs was the most important invention 
 in Western mediaeval architecture — the rival claimants are Norwich, Durham 
 nave, and the Abbaye-aux-Dames. || Of the last, unfortunately, the chronology 
 cannot be fixed definitely. Durham choir was commenced in 1093 ! Norwich 
 choir in 1096. At the end of the eleventh century, the builders of England 
 and Normandy had made great advances towards the development of Gothic 
 
 * In the west end of the choir triforium it almost looks as if flying buttresses were 
 actually built : if so, they were probably taken down when the choir received external flying 
 buttresses, i.e. 1472-1499. 
 
 t They are shown as a restoration in Ruprich-Robert, ii., Plate 92. 
 
 % Illustrated in Ruprich-Robert, ii., Plate So. § Illustrated in Enlarfs Manuel, 449- 
 
 il The pier-arcade and upper parts of the Abbaye-aux-Dames do not belong to the eleventh- 
 century work. 
 
EXTERNAL FLYING BUTTRESS. 
 
 371 
 
 architecture; and of the two countries England was ahead, having developed 
 diagonal ribs and flying buttresses as well; whereas the vaults of Normandy 
 did not get either till the twelfth century was well advanced. Only in one point 
 was England behind the times ; viz. in 
 the employment of the pointed arch. 
 
 It must be noted that though the 
 internal flying buttresses, e.g. of Wells 
 choir, were put there for the purpose of 
 abutment just as much as the external 
 flying buttresses of WELLS PRESBY- 
 TERY ($73), yet it is quite probable 
 that another object than that of abut- 
 ment was also present in the builders' 
 minds. This was to provide support 
 for the purlins (563J of the lean-to roof 
 of the aisle. In the case indeed of 
 CHICHESTER NAVE (34.4) it cannot be 
 doubted that one object of the employ- 
 ment of the lower flying buttress was 
 to facilitate the roofing of the aisles. 
 
 The next step was apparently 
 taken earlier in France than in Eng- 
 land and Normandy. It was to con- 
 struct the flying buttress, not beneath 
 the aisle roof, but above it, in the open 
 air. From one point of view, it was an 
 improvement. For it enabled abutment 
 to be given to the clerestory wall at a 
 much higher level ; and therefore put 
 
 it in the power of the builders to construct much taller clerestory windows, 
 and proportionately to improve the lighting of the church. From an artistic 
 point of view, at any rate when used in superposition and in double flights, 
 as at Le Mans and Westminster (379), it almost wholly hides from view 
 the building round which it is placed ; most especially is this disability 
 felt round the east end of an apsidal choir. And from the point of view 
 of a practical builder, the external flying buttress has grave defects. It is 
 difficult to construct, requiring lofty and expensive scaffolding. And when 
 the aisle roof has been put on, a flying buttress high above it is difficult to 
 repair. Worst of all, this stone stay, on which the stability and the very 
 existence of the whole building depends, now that it has been brought out of 
 doors from the shelter of the aisle roof, is exposed to all the mischances of the 
 weather ; to rain, and frost, and storm. To the English builder, who was nothing 
 if not practical, these objections may have seemed grave ; and for nearly a century 
 the English flying buttress remained concealed beneath the aisle roof, as it was 
 built in Durham nave. It was not till Canterbury choir was rebuilt, and then 
 probably only because the architect, William, was a Frenchman (from Sens, 1 175- 
 1 178), that it emerged from inside the triforium. Even at Canterbury it only 
 just crawls along close to the aisle roof, almost unseen from below. And the 
 
 Norwich Nave Triforium. 
 
372 SUBSTITUTES FOR FLYING BUTTRESS. 
 
 builders have shown their distrust of it by inserting a transverse arch in the 
 triforium below it (34.3). Next we come to CHICHESTER (313),* where high 
 vaults were constructed after the fire of 11 86 (34.4). Here the flying 
 buttress is much more advanced in type than at Canterbury ; but, like its imi- 
 tators and neighbours at NEW SHOREHAM and BOXGROVE, is heavy and ungainly. 
 On examining the section of Chichester, on the right, it will be seen that here, 
 as at Canterbury, the builders had not much faith in their new ally ; and have 
 supplemented it by another flying buttress concealed beneath the aisle roof.-f- 
 In France, on the other hand, flying buttresses had been constructed in the 
 open air long ago. Those of the St Denis of 1140 have disappeared ; but that 
 church was so light in construction that it is hardly conceivable that its clere- 
 story walls could have stood without the support of flying buttresses. Those of 
 Poissy may be f. 1 135 ; and Poissy very soon after inspired Sens. Examples of 
 the last half of the twelfth century occur at St Germain les Pr6s, Paris ; St 
 Remi choir, Reims ; Notre Dame, Chalons-sur-Marne ; S. Ouiriace, Provins ; 
 Langres, Paris, Mantes, Senlis, Noyon cathedrals ; and the churches of Domont 
 and Ebrueil. j By the thirteenth century the flying buttress had become the 
 trade-mark of the Gothic of the Domaine Royale ; no church in this style, 
 whether in France, Spain, Italy, Cyprus, was complete without it. 
 
 As we have seen, the English builders were by no means in a hurry to 
 construct flying buttresses out of doors. And even when it had come into 
 common employment in current Gothic, there was frequently evinced an 
 anxiety to avoid it. In the twelfth century external flying buttresses were 
 dispensed with at Durham, Lindisfarne, St Cross, and Wells ; in the thirteenth 
 century at Rochester (except one north of the choir) ; in Worcester choir ; § 
 in Salisbury (except those which were added when the steeple was built) ; in 
 SOUTHWELL CHOIR (400) (the present ones are c. 1355); in Gloucester nave; 
 at Dore Abbey ; at Tintern (except over the aisle of the north transept, which 
 was not finished till the middle of the fourteenth century); while in Rievaulx the 
 flying buttresses which existed were so low and slight that the builder evidently 
 trusted for his abutment to other means. In the fourteenth century the very 
 lofty clerestory of Gloucester has no flying buttresses although its windows are 
 of vast size ; nor has the neighbouring nave of Tewkesbury. In the fifteenth 
 century the fan-vaults of the choir of Sherborne have flying buttresses, but 
 not those of the nave : so also the lierne vaults of Norwich choir have flying 
 buttresses, but not those of the nave. Finally, they are dispensed with in the 
 lierne vault of Oxford Cathedral choir. Rather than employ such a dangerous 
 ally, the builders had resort to all sorts of expedients. The most drastic alter- 
 native was to construct the vault not in stone, but in wood. In the North of 
 England dislike of the external flying buttress seems to have been particularly 
 strong. In Selby choir, and on the north side of York nave, the builders had 
 
 * There were flying buttresses on either side of the choir of St Radegund, Bradsole ; 
 arranged as at Rievaulx. See Plan by Mr W. H. St John Hope in Archceologia Cantiana, xiv. 
 147. The Bradsole work was commenced in or soon after 1191. 
 
 t A similar disposition occurs in the choir of the Abbaye-aux-Hommes. 
 
 I Gonse, 41. 
 
 § It is possible that the present external flying buttresses of Worcester choir were sub- 
 stituted in the fourteenth century for internal ones. 
 
.v 5 
 
 New Shoreham. 
 
 Wells Choir. 
 
 Doxgrove. 
 liath Choir. 
 
374 DEFECTIVE FLYING BUTTRESSES. 
 
 actually constructed pinnacles (86, 366) to weight the feet of flying but- 
 tresses ; nevertheless they altered their minds ; no flying buttresses were built, 
 and .both were vaulted in wood. Sometimes, as in the twelfth century, in 
 Durham nave and Wells ; in the thirteenth, at Salisbury and York transept ; in 
 the fourteenth, in Winchester nave ; the flying buttresses are concealed beneath 
 the aisle roof* Sometimes, the clerestory wall being Norman, as in the naves 
 of St Cross, Norwich, Blyth, Gloucester, Tewkesbury, Oxford, Sherborne, is so 
 exceedingly thick that it takes the thrusts of the vault quite safely. Or, though 
 Gothic, as at Salisbury, it is built nearly seven feet thick at the top, with 
 the same object. Or the pier below is made unusually thick and massive, in 
 order to carry a strong clerestory buttress,-}- as at WELLS (373). Or the vault is 
 made to spring from a very low level (see 306) ; from the capital of the pier at 
 TEWKESBURY (297) ; or but little above it, as at Blyth. Or the thrust is 
 brought down as vertically as possible, by pointing the diagonal ribs of the 
 vault, as at Wells ; or by constructing the vault as a pointed barrel, as in 
 GLOUCESTER CHOIR (35.5) ; or by accepting the abutment of an ancient demi- 
 berceau (282), as in the same choir. It is indeed only the scientific combina- 
 tion of pointed barrel and demi-berceau that allows the existence of that 
 splendid clerestory and vault. Again, in SHERBORNE NAVE (346) account 
 was taken of the smallness of the thrust of a fan-vault of pointed section to 
 dispense with flying buttresses. Finally, in Bristol choir a totally different force 
 was called to aid,; that of opposing thrusts (see 381). All these vaults stand 
 safe ; a monument to the engineering skill of English builders, determined 
 to secure Gothic architecture without Gothic flying buttresses. 
 
 For a time mistakes were made in the construction of the flying buttress. 
 Some of the earliest had arches which were quadrants of circles. But, obviously, 
 the greater the curve of the arch, the more likely is it under the pressure 
 of the vault to buckle up in the middle. What is wanted is as flat an arch as 
 possible. So the later flying buttresses, e.g. in Malmesbury and BATH (373), 
 are much depressed. 
 
 Again, the early buttresses, e.g. at BOXGROVE (373), were not tilted up 
 enough. The result was that it was difficult to construct buttresses and 
 pinnacles substantial enough to resist the strong lateral thrusts transmitted to 
 them by the flying buttresses ; it will be seen how excessively massive are 
 buttress and pinnacle at NEW SHOREHAM (373). Still worse was it in many 
 of the flying buttresses of the twelfth century in France ; e.g. those of St 
 Remi, Reims. + But in later examples, e.g. MALMESBURY (375) and BATH § 
 (373)3 it is remarkable at what an acute angle the flying buttress is set. || 
 
 * In Italy also there are internal flying buttresses ; these, however, are not arches but 
 solid walls ; built upon the transverse arches which span the aisles, and reaching, at S. 
 Ambrogio, Milan, up to the roof boarding ; and at S. Michele, Pavia, one or two feet above it. 
 
 t Such a buttress is in what is called "false bearing" (porte-a-fau.x). See the sections of 
 Canterbury, Lincoln, Westminster, Exeter, and Gloucester (34, 35). 
 
 X Illustrated in Viollet-le-Duc, Architecture, i. 62. 
 
 § So also when Ely presbytery was remodelled, the flying buttresses were given a much 
 sharper slope ; for the original flying buttresses, being too flat, had pushed the buttresses out of 
 the perpendicular. See illustration 365, and Stewart's Ely, p. 80, and section in Plate 4. 
 
 II So also at Bourges, Semur-en-Auxois, Notre Dame de Dijon, and Pontigny choir. 
 
DEFECTIVE FLYING BUTTRESSES. 
 
 375 
 
 Again, too much masonry was employed in the earlier examples. CANTER- 
 BURY CHOIR (34.5) is an exception ; but there the flying buttress is reinforced 
 by a transverse arch beneath. At Chichester, Boxgrove, and New Shoreham 
 the flying buttresses are excessively massive and heavy. But it must be noticed 
 that at Boxgrove, owing to the peculiar system of vaulting (see 320;, each seven- 
 of the high vault is twice as broad as usual, and therefore an exceptionally 
 powerful thrust has to be dealt with. We may surmise from the analogous 
 disposition of the but- 
 tresses at New Shore- 
 ham, that there a simi- 
 lar vault system was 
 planned, though not 
 carried out. The light- 
 ness of such flying 
 buttresses as those of 
 Lincoln choir (34.5) 
 comes on one as a sur- 
 prise, considering that 
 the vaults of New 
 Shoreham and Lin- 
 coln choirs are con- 
 temporaneous work.* 
 In this we may sur- 
 mise the influence of 
 Canterbury choir; 
 followed by Lincoln 
 not only in lightening 
 the flying buttresses, 
 butinreinforcingthem 
 by transverse arches 
 in the triforium cham- 
 ber^ 
 
 By the time the 
 choirs of Canterbury 
 and New Shoreham 
 were built, the French 
 had learnt to con- 
 struct light and ele- 
 gant flying buttresses. 
 The fact that our earliest examples arc so massive and ugly, or so unusual in 
 type, tends to show that those who built them had little acquaintance with what 
 was going on in France, or at any rate did not imitate the French work. 
 
 * But their unusual lightness supports the opinion of Sir Gilbert Scott and Professor 
 Lethaby that the vault and flying buttresses of Lincoln choir were added later ('see 1 12). 
 
 t In France also the earlier flying buttresses often exceeded in massiveness ; e.g. Pontigny ; 
 Notre Dame, Chalons-sur-Marne ; Ourscamps. Later they sometimes exhibit astonishing 
 tenuity ; e.g. at Narbonne, and above all, at St Urbain de Troyes. 
 
 Malmesbun 
 
376 
 
 DEFECTIVE FLYING BUTTRESSES. 
 
 Instead of avoiding the mistakes which the French by 1175 had corrected, we 
 insisted on making the mistakes for ourselves, and the corrections for ourselves. 
 The development of the flying buttress here seems to be mainly independent of 
 its development in France. 
 
 Another point that seems to have troubled the builders was to settle to 
 what height on the clerestory wall the flying buttress should be tilted up. One 
 cannot see a thrust ; and even now mathematics cannot tell us where exactly 
 the thrust is concentrated. The matter had to be settled empirically. Some- 
 times the head of the flying buttress was set exactly opposite to the spring 
 of the ribs of the vault inside ; that was found to be too low. In some French 
 examples, on the other hand, the flying buttress was set too high ; so that, as at 
 Evreux, it has been necessary to pull down vault and flying buttresses, and give 
 the vault abutment lower down.* Really, the thrust is felt most at some point 
 
 Sherborne Choii 
 
 St Mary Redcliffe. 
 
 between the spring of the ribs and half-way up the vault ; but the exact situation 
 of that point varies in every vault. Only practical experience, correcting itself by 
 the study of failures, enabled the old builders to get their flying buttresses right. 
 One expedient, and a very successful one, was very commonly adopted, to 
 meet the difficulty of estimating the height of flying buttress required. It 
 was to build a stout buttress outside the clerestory wall, and to set the head of 
 the flying buttress against this, not against the wall.f It is well seen in CANTER- 
 BURY CHOIR and WELLS PRESBYTERY (34.3, 373). In this way the flying 
 buttress gets a grip of the whole height of the clerestory wall, and the thrust 
 of the vault can hardly escape it. On this point the Lincoln builders differed 
 from the common practice ; first they built the eastern transept with tall 
 clerestory buttress ; then choir and nave with one half size ; then the presbytery 
 with none at all : going from bad to worse. 
 
 * See Ruprich- Robert, i. 140, on the abutments at the Abbaye-aux-Dames. 
 t For examples of clerestory buttresses see the sections on 34, 35. 
 
DEFECTIVE FLYING BUTTRESSES. 
 
 177 
 
 At BOXGROVE (373). as in the apse of Reims St Remi, the upper surface 
 is flat. This is objectionable, because the rain finds its way into the joints. At 
 NEW SHOREHAM ($7$) there is a double slope or coping, which sheds the rain, 
 and at the same time improves the 
 appearance of the flying buttress. In 
 WELLS PRESBYTERY (373), in addition, 
 each stone of the outer surface is cut so 
 as to overlap the one beneath ; protect- 
 ing the joints still better. And if the 
 coping be molded, the flying buttress 
 becomes something more than the piece 
 of brute engineering which it had been 
 at Boxgrove. Occasionally one more- 
 change was made ; its upper surface 
 was channelled that it might serve as 
 an aqueduct, as at SOUTHWELL* (400). 
 
 At BOXGROVE (373) the buttress has 
 to stop, unaided, the pressure trans- 
 mitted to it by the flying buttress. 
 But, in later examples, the buttress 
 is generally weighted by a pinnacle. 
 Sometimes this pinnacle is placed, 
 as at Amiens and EXETER, close to 
 the outer face of the buttress. But 
 the flying buttress is liable to burst up 
 at its haunches ; and therefore, more 
 correctly, the pinnacle is often placed close to the inner face of the buttress, or 
 even resting partially on the flying buttress in porte-a-faux or " false-bearing." 
 Cf. NEW SHOREHAM (373), NORWICH CHOIR (160), and MALMESBURY (375).+ 
 
 Beyond molding the coping, the builders, as in the choir of EXETER, 
 usually left the flying buttress plain. Often even moldings were omitted, 
 that there might be less access given to rain and frost. In Winchester presby- 
 tery, however, and the transept of ST MARY REDCLIFFE, BRISTOL (376), a ruw 
 of crockets runs along the upper surface of the flying buttresses ; and in this 
 and several late examples, e.g. SHERBORNE (376), Windsor, Bath, the upper 
 spandrel of the flying buttress is pierced, and is filled with tracery. 
 
 Frequently also late spires were connected with their corner pinnacles by 
 ornamental flying buttresses. It is remarkable how unsuccessful almost all 
 are. Some are thin and flimsy ; some have a weak compound curve ; e.g. 
 Batrington and St Michael, Coventry. LOUTH (139) has the one success; 
 a flying buttress far surpassing both in strength and grace any attempted else- 
 where. It consists of an arch cusped beneath ; the arch carries vertical balusters; 
 the balusters carry a first bar ; tracery on this bar carries a second bar, which 
 again is crocketed. The only other flying buttresses which will compare with 
 
 * So also the thirteenth-century flying buttresses of Burgos are utilised as aqueducts. 
 + At Ely, when the presbytery was remodelled, the pinnacles were set back three feet 
 nearer to the clerestory wall. 
 
 Exeter S. CI 
 
5/ 
 
 78 
 
 COMPOUND FLYING BUTTRESSES. 
 
 the highly decorative flying buttresses of France, e.g. at Eu and Abbeville, 
 are those of Henry the Seventh's Chapel at WESTMINSTER (378). Here 
 an arch carries a first bar ; the bar carries tracery, the tracer) 7 a second arch ; 
 and this second arch a third ; but this third arch is inverted. The object of 
 this inversion, which occurs often in late French Gothic, as at Alencon,* is to 
 stop the tendency of the flying buttress to rise at its haunches. 
 
 We have seen above that the precise point where the thrust of the vault 
 should be met could not be ascertained with exactness ; and that this difficulty 
 was met by the addition of a clerestory buttress. But in the lofty French 
 churches, and in Westminster, the loftiest vaulted church in England, it was 
 
 met in another and very ingenious 
 way ; viz. by superposing two or three 
 flying buttresses, which held firm the 
 clerestory buttress at two or three 
 points, and effectually steadied it 
 against the outward thrust of the 
 vault. An early example is seen at 
 CHICHESTER (34.4), where one flying 
 buttress is concealed by the aisle roof, 
 while the other is built in the open 
 air. In those bays of Ely presbytery, 
 c. 1240, which have not been re- 
 modelled, two flying buttresses are 
 superposed. WESTMINSTER (35.2), 
 following French precedent, has super- 
 posed flying buttresses ; two over the 
 aisle roof; and three over the roof of 
 the cloister ; which in this example 
 corresponds to the outer aisle of a 
 double-aisled choir such as that of 
 Le Mans or Beauvais. 
 
 In Westminster cloister it will be 
 seen that the flying buttress does not 
 pass at one flight all the distance from 
 the cloister buttress to the clerestory 
 wall, but in two flights ; the piers be- 
 tween the aisle and the cloister being 
 raised to such a height that they sup- 
 port the foot of the upper flying but- 
 tress and the head of the lower one. This intermediate pier is weighted by a 
 pinnacle. Flying buttresses of two flights occur also in Roslyn Chapel.-r In 
 France many choirs and some naves have double aisles, and this disposition is 
 very commonly employed. It existed originally in Notre Dame, Paris ; but 
 afterwards for two flights of flying buttresses there was substituted a single 
 flying buttress of the vast span of 50 feet. 
 
 The function of the flying buttress has been much misunderstood. It 
 
 Henry the Seventh's Chapel, Westminster. 
 
 Choisy, 
 
 + Illustrated in Britton's Architectural Antiquities, vol. 
 
FUNCTION OF FLYING BUTTRESS. 
 
 379 
 
 has been held that the flying buttress exercises a powerful inward thrust ; 
 that in a Gothic aisled nave, for instance, the outward thrust of the high 
 vault is neutralised by the inward thrust of the flying buttress; so that the 
 lofty pier, pressed on one side by the high vault, and on the other by the 
 flying buttress, cannot budge an inch to north or south. This cannot be so. It 
 
 Westminster Nave and South Transept. 
 
 is true, of course, that the arcuated vault has an outward thrust. But a flying- 
 buttress has little inward or upward thrust. Yet it has been supposed to be able 
 to deliver a counter-thrust ; as if it were a sort of hydraulic ram, set in motion 
 by some hidden force, and always pushing inward at the clerestory wall. It is 
 true, indeed, that if the flying buttress be set almost horizontally, as at St Remi, 
 Reims, Notre Dame de Chali ms-sur-Marne, Canterbury choir, New Shoreham, and 
 
3 8o FUNCTION OF FLYING BUTTRESS. 
 
 certain other examples of twelfth-century Gothic, the weight of the upper part 
 of the flying buttress will press against the clerestory wall. And if the flying 
 buttress be heavily loaded with masonry, as were some of the early examples, 
 e.g. at New Shoreham, at Boxgrove, at St Martin, Laon, the pressure against 
 the wall will be considerably increased. But if we look at the buildings of the 
 thirteenth century — the culminating period of Gothic engineering — we shall find 
 that the flying buttress is set much more vertically — e.g. at Bath — and is often 
 of exceedingly light construction — e.g. at St Urbain, Troyes — and that in these 
 thirteenth-century flying buttresses the inward pressure is practically a negligible 
 quantity. Think of a ladder set up against a window. If it be set up at a 
 considerable angle it will certainly break the glass, more especially if the ladder 
 be a heavy one. If it be set up nearly vertically, it will, perhaps, not break the 
 glass. The former case is that of such flying buttresses as those of New Shore- 
 ham, Chichester, and Boxgrove ; the latter of such as Sherborne and St Urbain 
 cle Troyes. 
 
 But the main function of the flying buttress is not to originate thrusts itself. 
 What it was meant to do, and does effectively, is to transmit thrusts. Apart 
 from the comparatively slight pressure which the weight of its upper part exerts 
 against a clerestory wall, the function of a flying buttress is mainly that of a stay. 
 Of its two components, the bar above and the arch beneath, it is the former 
 which is of primary importance. The primary function of the arch is, not to 
 transmit, still less to produce thrusts, but simply to support the bar. That the 
 inward thrust of the later flying buttresses, if it exist at all, must be but incon- 
 siderable, is evident from the fact that the flying buttresses at Melrose have not 
 pushed in the clerestory walls, although the high vault has fallen. 
 
 Quite as remarkable as the engineering skill which invented the flying 
 buttress was the artistic instinct that beautified it. What can there be in 
 building "construction that is a priori more unpromising, as a subject for 
 architectural treatment, than a shore of masonry, built up on the outside of 
 a wall to prevent it from being thrust out by the pressure from within ? I do 
 not know what the modern architect would do as an artist if as a constructor 
 he found it necessary to employ such a member. In the absence of appli- 
 cable precedents he would be apt to conclude that so ugly an appendage to 
 his building would not do to show, and to conceal it behind a screen-wall nicely 
 decorated with pilasters.* But the builders upon whom the use of this member 
 was imposed, not having enjoyed the advantage of a classical education, saw 
 nothing for it but to exhibit the shore and to try to make it presentable by making 
 it expressive of its function. Their early efforts were so ' uncouth ' that the 
 modern architect, if he had seen the work at this stage, would have been con- 
 firmed in his conclusion that the shore was architecturally intractable. The 
 mediaeval builders kept at work at it, master after master, and generation after 
 generation, until at last they made it speak. Made it speak? They made it 
 sing, and there it is, a new architectural form, an integral part of the most 
 complicated and most complete organism ever produced by man, the Gothic 
 cathedral." f 
 
 * As Sir Christopher Wren did at St Paul's. 
 
 t Montgomery Schuyler in Architectural Record, iv. I, u. 
 
OPPOSITION OF THRUSTS. 38] 
 
 Part II. 
 Opposing Thrusts. 
 
 So far we have dealt with two cases of abutment. The first is that ol 
 the Lady Chapel of LICHFIELD (369), where the outward pressure of the 
 loaded arches of the vault is stopped by the inert resistance of the but- 
 tresses. The second is that of the presbytery of Lichfield, where the outward 
 thrusts are transmitted by flying buttresses to the aisle buttresses, by whose 
 inert resistance they are stopped. But there is a third method of parrying a 
 thrust. It is to bring to bear an opposing thrust. Suppose that a stream is 
 crossed by a bridge of three arches of equal span. Here the two outward 
 thrusts of the central arch are exactly neutralised by the inward thrusts of the 
 two outer arches. The same is the case with the range of arches on either side 
 of an aisled nave. All the way down the nave there are thrusts and counter- 
 thrusts neutralising one another, except in the case of two arches, the western- 
 most and the easternmost. As to these, the outer thrust of the westernmost 
 arch is stopped by a big buttress built on to the west front in the axis of the 
 pier-arcade ; or else by a western tower. But in a cruciform church, the thrust 
 of the easternmost arches is exerted, about midway up, against the western piers 
 of the crossing. This is therefore a very dangerous point in a mediaeval church ; 
 and at this point many churches gave way ; e.g. Winchester, Lincoln, Chichester, 
 El)'. Various remedies — none quite satisfactory — were employed to parry this 
 danger. One was to weight the four piers of the crossing by a tower ; another 
 was to construct a solid stone screen between the two eastern piers of the 
 crossing, as at York, Canterbury, Ripon ; another was to build horizontal stone * 
 girders (" strainer-arches " 1 across from pier to pier, as at Salisbury, Canterbury, 
 Rushden ; f or else a St Andrew's cross, as at Wells, and originally at Glaston- 
 bury, where the grooves of the cross may still be seen in the piers of the crossing. 
 
 Another weak point is that the inward thrust of the aisle-vault, r.g. EF, 
 has nothing to stop it adequately, unless the pier be made excessively thick, 
 as e.g. in NORWICH (238). It is true that the pier is loaded by the whole 
 weight of the triforium and clerestory wall, and even by a pinnacle B in addition 
 (382. 1 j. Nevertheless, as Sir Christopher Wren observed, man}' of these piers 
 incline inward ; showing that the inner thrust of the aisle-vault is not fully 
 stopped by the weights superimposed on the pier. It, however, we look 
 again, it is seen that there is a thrust CD from the high vault of the nave 
 which runs in the opposite direction to that, EF, of the aisle-vault. Obviously, 
 if we could bring these thrusts into proximity, CD might be employed to 
 neutralise EF. And if so, the happy result would follow, moreover, that we 
 could then dispense with the flying buttress G. 
 
 How is it to be done? We cannot lower the nave, so as to get the thrusts 
 of its vault lower down. But we can raise the aisle, as in Fig. 2 ; and then the 
 
 * Or a "strainer-beam," as at Pembridge. 
 t Illustrated in Churches of Norlhanls, 1S1. 
 
;S2 
 
 OPPOSITION OF THRUSTS. 
 
 thrust AB will partially neutralise the thrust CD, even without the aid of the 
 pinnacle E. But plainly the thrust A of the wider and heavier vault will push 
 the pier outward in defiance of the thrust CD ; from this point of view, therefore, 
 the weighting pinnacle E is essential.* 
 
 The above is a case of the opposition of unequal thrusts. But if we follow 
 the precedent of the bridge over the stream, and establish an opposition of equal 
 thrusts, then there will be no need of the pinnacle E. To do so, however, the 
 nave and its aisles, like the three arches of the bridge, must be of equal height 
 and equal span. There will be no need then, nor any place, for flying buttresses. 
 However, since the aisles now rise as high as the nave, there can be no clerestory 
 lighting. All the side-light must be derived from the aisle windows. But since, 
 ex hypotkesi, we have made the aisles as lofty as the nave, there is room in the 
 aisle-walls for very tall windows ; and the church will be flooded with light as if 
 it had a clerestory; as may be seen in the TEMPLE CHOIR (35.1). Thus, 
 without injuring the lighting of the church, we have got rid of our dangerous 
 
 ally, the flying buttress. Practically, 
 what we have built is not a nave and 
 aisles ; but three parallel naves. 
 
 And, of course, the system is just 
 as applicable to two parallel naves as 
 to three ; f as is seen in the highest 
 grandeur in the vaulted church of the 
 Jacobins at Toulouse. In the last half 
 of the twelfth and in the thirteenth 
 century it is the special mark of one of 
 the most highly individualised schools 
 of French Gothic, the Plantagenet style 
 of Anjou ; examples of which are 
 Poitiers Cathedral, begun 1163, the 
 Hospital of St Jean, and the Church of 
 St Serge, Angers. It is also a special characteristic of much German Gothic ; e.g. 
 Marburg, Wetzlar, Minden, Soest, Landshut, Thorn, Munster. 
 
 In England an early and important example is the choir of the TEMPLE 
 CHURCH (35- 1 ), London, finished in 1240. Here the three transverse arches, 
 though of the same height, are not of the same span ; their haunches accordingly 
 are heavily loaded. And both here and in Bristol choir each pier receives a 
 vertical load of masonry. In the Temple choir each of the three naves has its 
 own gabled roof. This leaves between the three roofs two gutters running the 
 whole length of the choir. Infiltration may occur from these ; and they may 
 be choked with snow. At BRISTOL (35.4) all three naves are spanned with a 
 single roof, giving a much superior drainage system. 
 
 A similar system to that of the Temple choir is sometimes adopted where 
 two or more aisles are built on one side of a clerestoried nave, as in Oxford 
 
 * Cf. the system of Romanesque high aisles in Poitou, Classification of Romanesque, 277. 
 
 t In many districts, e.g. in Kent and Sussex, the system of double or triple nave is common 
 in parish churches; e.g. at Wingham, Kent, and ST Lawrence (213) in the Isle ofThanet, 
 in preference to nave and aisles. 
 
OPPOSITION OF THRUSTS. 383 
 
 Cathedral choir. So again, the retrochoir of Winchester,*' c. 1200, consists of 
 three, that of St Saviour's, Southwark, of four naves, side by side, of equal 
 height. At SALISBURY (154,458, 173), the builder has amused himself by dividing 
 the narrow Lady Chapel into three naves ; and although all three have one 
 span-roof in common, he has built three eastern gables instead of one, to 
 draw attention to the disposition of the interior. 
 
 In other cases also, where it was desired to keep the vault low, instead of 
 vaulting an apartment hall-fashion, it was divided into two naves: e.g. at 
 Fountains the undercroft west of the cloister originally carried the dormitory of 
 the lay brethren ; and was divided into two avenues instead of one. So also 
 over Fountains chapter house was the dormitory of the monks ; the chapter 
 house was kept low, therefore, by being divided into three avenues. But even 
 when the building was a lofty refectory, the monks never designed it with a 
 clerestoried nave. If it was large, it was divided into two naves, as at Fountains, 
 or three ; and sufficient light was obtained from windows in the side and end 
 walls. We cannot help concluding that the clerestoried nave, with its appanage 
 of flying buttresses, was looked upon as a necessary evil, at any rate in this 
 country, and was seldom adopted where adequate height, breadth, and light 
 could be obtained in a simpler way. 
 
 * Illustrated in Dehio, Plate 601. 
 
Chapter XXV. 
 
 THE DRAINAGE OF THE ROOFS. 
 
 External Roofs — Corbel-Tables — Cornices — Parapets — Battlements — Gargoyles — Spouts. 
 
 External Roofs. 
 
 As we saw above, the enemy which the mediaeval builders had most to dread 
 was fire. To this they opposed their system of vaults ; and by the presence of 
 the peculiar form of vault that was affected from the twelfth century onward — 
 the vault with diagonal ribs — the whole construction and the whole aspect of the 
 church, internal and external, was dominated. But there was another enemy, 
 which did not come to the attack with the momentary fury of a conflagration, 
 but which yet was so unremitting, subtle, and destructive in its operations, that 
 it was quite as formidable as fire ; this was rain and snow. This peril too had 
 to be provided against ; and the provisions against the destructive action of 
 snow and rain, assisted by frost, profoundly modified the external aspect of 
 the churches. 
 
 In the vast structures of ancient Rome, the groined vaults were never 
 sheltered by an external roof of wood, but either by plates of metal or by large 
 tiles ; or else they were simply coated with fine cement. In the Baths of 
 Caracalla, the valleys between the groins were filled up with cement so as to 
 form a level pavement ; this was covered by a mosaic of coloured marbles, 
 forming a magnificent terrace. In S. Maria degli Angeli, originally a portion of 
 the Baths of Diocletian, and elsewhere, the external is precisely the same as the 
 internal disposition of the groined vaults, the hollows between the groins forming 
 excellent valleys for the rain to run down.* 
 
 These Roman vaults, whether groined, barrel, or domical, were enormously 
 thick and massive ; and though they were copied, with their absence of pro- 
 tective roofs, in the Romanesque of the middle and centre of France,f they were 
 replaced in Burgundian Romanesque by barrel vaults, and in Northern Gothic 
 by intersecting ribbed vaults so thin as to be but mere shells in comparison with 
 the massive vaults of Auvergne, Provence, and ancient Rome. Such thin light 
 shells were quite incapable of bearing the load of superposed masonry forming 
 the external slopes of the vault ; and if they had been capable, the thrusts of a 
 groined vault so loaded would have been increased beyond all reasonable power 
 of resistance. Therefore, both in the light barrel vaults of Cluny (1089-1095), and 
 
 * See Choisy's Roman Building. The same system prevailed originally in several of the 
 earlier Spanish cathedrals ; traces of the old roofing-system are still visible here and there, 
 t See Classification of Romanesque, 276 seq. 
 
385 
 
 © 
 
 cass^gHriSifsae; 
 
 
 f;^-^ ff\-ff f*-*^ -ff-©^: 
 
 4iK_;^s,^ \.',z^C 1k-wk7 \*^£-< t ' ' 
 
 r ==~' -1 ' I. NfgP -- 
 
 
 
 
 [ 
 
 
 
 II 1 1 1 q^ 
 
 
 1 
 
 1 
 
 1 1 1 
 
 
 1 o 
 
 I I I I 1- 
 
 Scale: of Feet fog N°- 5 6789. 
 
 i. West Walton 
 
 2. Romsey. 
 
 3. Romsey. 
 
 4. Grantham. 
 
 5. Ensham. 
 
 6. Fountains Nave. 
 
 7. Netley Choir Aisle. 
 
 8. St Mary's, York. 
 
 9. Bridlington N. Aisle. 
 
 2 I: 
 
<86 
 
 Wsiww^^ ^^y3^ f ^^khvk}:±;w& 
 
 i. Beverley Minster. 3. Warmington Spire. 
 
 2. St Mary, Scarborough. 4. St Mary Redcliffe, Bristol. 
 
 5. Westminster, St Erasmus Chapel. 
 
387 
 
 1. St Mary, Beverley. 
 
 2. New College, Oxford. 
 
 3. Bishopstone. 
 
 4. Lavenham. 
 
388 
 
 i. Heckington. 
 
 2. Winchelsea. 
 
EXTERNAL ROOFS. 389 
 
 in all the Gothic architecture of France and England, the protective covering 
 of the vault had necessarily to take the form of an independent roof* 
 
 In the Romanesque churches the wooden roofs, in all probability, were 
 usually covered with tiles. But there are records of the use of metal also. The 
 great and wealthy pilgrim-church of St Martin de Tours was famous for roofs of 
 tin. Lead was in frequent employment from the earliest times. Wilfrid is 
 said by Eddius to have roofed York Minster c. 669, " artifiose plumbo puro 
 tegens." When Ernulph's choir at Canterbury (1093-1130) was burnt down, 
 Gervase tells us that the sheet-lead melted; in 1706 two large glue-pots were 
 filled with lead picked out of the joints in the pavement of Ernulph's choir.-f 
 Gervase also tells us that William the Englishman covered the roofs of the 
 choir-aisles of Canterbury with lead in [184. But probably the most common 
 material was tiles ; or in stone districts, flat stone slates. In wood districts 
 shingles were commonly used. Imitations of these in stone are common in the 
 Gothic spires of Normandy ; e.g. Vernouillet. 
 
 The greater the amount of rain and the greater the amount of snow, the 
 sharper it is desirable to make the pitch of the roofs. Thus the Greek and 
 Roman temple with its comparatively low-pitched roof gives way to a high- 
 pitched roof in the Early Gothic of England and France ; while, in Norway, 
 with its heavy snowfall, the roofs, e.g. at Borgund, are made excessively steep ; 
 the snow must be got rid of at once, before it consolidate and its weight break 
 down the roof. A second consideration is the nature of the material which is 
 used as a roof-covering. If covering of tiles or stone slabs or shingle or thatch 
 is employed, it is usual to give the roof a high pitch ; as with these materials 
 it is desirable that the rain should drain off rapidly. But sheets of lead, alter- 
 nately expanding and contracting under the heat of day and the cold of night, 
 tear themselves from their fastenings unless the roof be considerably flattened. 
 (To minimise this, e.g. in covering timber spires with lead, the strips of lead 
 are often set diagonally.) And as lead came into greater use as a roof-covering 
 in later Gothic, probably because of the increasing exploitation of lead-mines 
 and its greater cheapness, the tendency of English Gothic was to lower the pitch 
 of the roofs ever more and more. It is curious that in France the tendency 
 was the very reverse ; from the beginning of the eleventh on to the sixteenth 
 century French roofs became more and more acute ; especially in Northern 
 France ; where it is not rare to see gables that have been heightened again and 
 again, each time that the roof has been reconstructed.* In England, on the 
 other hand, we see almost even-where weatherings on the towers showing that 
 originally there was a roof, sometimes more than one, of sharp pitch, before the 
 existing low-pitched roof was constructed; e.g. Pershore and TEWKESBURY 
 390). There are indeed few exteriors which have not lost much of their 
 beauty from their roofs being of a lower pitch than they used to be; but 
 generally the weather-mold remains to show the original pitch; e.g. in 391 
 
 * For such a roof, one of timber independent of the vault, there was a sixth-century pre- 
 . edent at Ravenna. There the vaults were sometimes constructed of hollow pots ; e.g. the 
 dome of S. Vitale ; and in such cases it was necessary to protect them with an independent 
 wooden roof. 
 
 + Willis' Canterbury, 108. + Enlart's Manuel, 50. 
 
39Q 
 
 Tewkesbury from W. 
 Southwell Transept and Choir. 
 
 Selby Choir. 
 Tiverton. 
 
CORBEL-TABLES. 39] 
 
 ABC shows the original steep roof. In time the ends of the rafters i;.\, BC, 
 decayed. Sometimes, instead ®f replacing these decayed timbers, it was thought 
 sufficient to cut off a foot or two from the ends of the old rafters. Thus the 
 rafters, being shortened, only reach to I>, and the roof is lowered. Then the 
 ends decay again, and the sooner, now that the rafters are old ; another piece 
 is cut off, and the roof sinks to E. But this is not all. Supposing ABC to be a 
 gable with a lofty window in it ; the window will probably be decapitated when 
 the roof and gable are lowered to D ; certainly, when the}- arc lowered to E. 
 This was often the case, till recent restorations ; e.g. the east windows of 
 Dorchester Abbey Church,* and at Great Haseley ; + it is still so at Whaplode. 
 
 In general our English roofs, till about the end of the thirteenth century, 
 had a pitch from about 45' to 50°. Exceptionally sharp, and exceptionally 
 effective, is the steep pitch of the roofs of Lincoln Minster. Here there is a 
 consummate harmony between the high-pitched roof and the acute lancets of 
 windows and arcade. Still more so, where, as at Salisbury, and originally at 
 Lincoln, the lancet window leads the eye up to the acute span-roof, and the 
 high-pitched roof to a soaring spire. Thus in Late Gothic the French pre- 
 served, and we often lost, the two great excellences of a mediaeval exterior ; the 
 high gable and the high-pitched roof. The later English church only too often 
 became, to the eye, four walls with a hole in the middle ; 
 little or no roof was visible at all : e.g. the fourteenth- 
 century naves of York and HULL (96), contrasting with 
 the earlier transepts, which have high-pitched roofs ; St 
 George's, Windsor ; King's College Chapel, Cambridge ; 
 and hundreds of parish churches. The culminating sky- 
 line of the roof was gone. Nevertheless the eye was not 
 satisfied without some emphatic upward termination of 
 the wall ; and this was provided by the greater elabora- 
 tion of the later battlement and parapet. The pinnacle too, which had originally 
 been constructional in function, now frequently was used merely decoratively ; 
 to distract attention, with the aid of parapet and battlement, from the fact that 
 the roof has disappeared. The elaboration of these late wall crestings is well 
 seen in York choir, Peterborough eastern chapels, Lavenham, Tiverton chantry, 
 King's College Chapel, Cambridge (365, 199 , and in Henry the Seventh's Chapel, 
 Westminster. Of these we ma)- perhaps fairly say, that though the roof has 
 disappeared, its absence is hardly felt. 
 
 Corbel-Table. 
 
 When it has been decided what relation the slope of the roof shall bear 
 to the amount of rain and snow on the one hand, and the properties of the 
 roofing material on the other, there still remains the important question how 
 to get rid of the drainage from the roof with the minimum of injury to 
 the walls, window-tracery, and doorways. If the rain be allowed to drip 
 down the walls, the surface will soon disintegrate, peel, and decay. This has 
 
 * Illustrated in Parker's Dorchester, 104, 105. 
 
 + Illustrated in Parker's Gothic Churches near Oxford, 12, 13. 
 
39 2 CORBEL-TABLES. 
 
 to be prevented by all means. One remedy is to make the roof overhang the 
 walls as far as possible. If the roof is one of thatch, as still in many churches 
 of East Anglia, the overhanging thatch, with deep shadows below, provides a 
 pleasing and adequate protection. But in very many churches in the same dis- 
 trict, covered with lead, e.g. Martham, Wiggenhall St Mary the Virgin, Worsted, 
 NORWICH ST Stephen's (228), the rafters project beyond the wall, and sheets of 
 lead overhang the rafters ; following no doubt the precedent of the indigenous 
 thatched roof of reeds. Elsewhere, usually, the two or three upper courses of 
 the wall are made to project beyond the lower ones ; forming what is called 
 a " table " or a " cornice." The blocks composing the lower course of such a 
 table require supports. These supports take the form either of projecting 
 stones in the wall (" corbels "), or of arches. The former are the more common 
 in English Romanesque ; the latter in the Romanesque of Lombardy and 
 Germany. Of the table with a horizontal base, Jumieges, Cerisy, Lessay, 
 Caen St Nicholas, the Abbaye-aux-Dames, in Normandy, and in England 
 NORWICH APSE (160), STEYNING (359), WEST WALTON (385. 1 ), ST MARY'S, 
 SCARBOROUGH (386.2), WELLS (373), and BEVERLEY MINSTER* (386.1), furnish 
 examples. Usually, but not always, the corbel is placed beneath each joint 
 of the base -course of the table. The origin of the corbel-table has been sought 
 in the projection of the tie-beams of a wooden roof beyond the walls ; in some 
 districts, e.g. Auvergne, the decoration of the corbel is evidently motived by 
 shavings of wood.f But corbel-tables were in use in ancient Rome and in 
 Ravenna ; e.g. in the mausoleum of Galla Placidia.^: And no doubt they were 
 invented over and over again, wherever the need was felt of giving the tiles the 
 greatest possible projection beyond the face of a wall. 
 
 Secondly, as in the Basilica of Junius Bassus,§ the projecting table may rest 
 on small arches. And the arches may either be supported by corbels or by 
 pilasters. The arches in Romanesque work are usually semicircular, and in 
 Early Gothic trefoiled ; e.g. at ROMSEY (385.2, 3), under the broach spire of 
 WARMINGTON (386.3), Ely facade, Carlisle Cathedral, Lincoln presbytery, 
 BOXGROVE {m). In the Norman work of Winchester and Ely the arches 
 of the corbel-table are segmental. Or the lower end of the corbels of the 
 arches may be rounded; producing a flow of curve called the nebule ; e.g. at 
 Peterborough ; Cambridge, St Sepulchre's ; SOUTHWELL transept (390). 
 
 Thirdly, where bricks only are employed, the necessary projection may be 
 obtained as at St Albans, || by causing each upper course of the table to project 
 somewhat beyond the lower course. If stone be employed, the same method 
 may be employed ; but the stone courses admit of moldings, as at Shottes- 
 brooke. 
 
 Fourthly, the arches supporting the table may rest not on corbels, but on 
 pilasters ; as in the transepts of Tewkesbury .If 
 
 * The Beverley cornice no longer carries the roof. 
 
 t See Viollet-le-Duc, Architecture, illustrated in iv. 309. 
 
 X Dehio, Plate 31, 8. § Dehio, Plate 31, 1 1. 
 
 || A similar treatment occurs in the mausoleum of Galla Placidia, at Ravenna. 
 
 I A similar arrangement occurs c. 410 in the Baptistery of St Orso at Ravenna; and in 
 the fifth or sixth century in the Syrian church of St Simeon Stylites. Illustrated in Scott's 
 Essay, 62. 
 
PARAPETS. 
 
 393 
 
 A late example of this system of dripping caves is that of Shottesbrooke 
 Church ; where the rain has dripped from the eaves since 1337 till quite recently, 
 without serious injury to the flint facing- of the walls. In France, too, dripping 
 eaves are found throughout the whole Gothic period, where economy had to be 
 studied.* Indeed, the preference for the open-air system of drainage is a rational 
 one. It is a great advantage to have all parts of the drainage-system visible 
 and accessible; whereas leakage in a gutter or a spout may go on for months 
 without being detected, and may do irreparable mischief to the masonry. It 
 provides no troughs for snow or leaves or dirt to accumulate, blocking up the 
 drainage. It is true that drip from eaves soon ruins the face of masonry if it is 
 of a freestone easily disintegrated by frost ; but if it is of a hard stone, e.g. like 
 the flint facing of Shottesbrooke, the mischief done is practically inappreciable. 
 
 Parapet. 
 
 Nevertheless, the system has practical disadvantages. In the first place, 
 in the case of a clerestoried church, if the roofs be of tiles, then in the case 
 of heavy rainfall, the rain falling on to the tiles 
 of the aisle-roofs from the height of the central 
 roofs will certainly find its way through the aisle- 
 roof and ruin any vault beneath it. And masses 
 of snow slipping from the central roofs on to the 
 lower roofs, in its fall will injure, if it does not 
 break them down. So will slates or tiles or sheets 
 of lead dislodged from the high roofs. In a town 
 where the church rises from a street or square, 
 these flying missiles are dangerous to life and 
 limb. Moreover, repairs of the slightest damage 
 to the roofs will be difficult and costly: especially 
 of the high roofs. Ladders will have to be erected 
 
 from the ground to the top of the clerestory wall ; the slaters or tilers engaged 
 in repairing the high roofs may slip, and life be lost. Nor is it possible, in 
 order to inspect the condition of the roofs, to walk round on the top of the 
 walls of the clerestory or the aisles ; for, with dripping eaves, the tiles or lead of 
 the roof project over and beyond it. 
 
 To remed\' all this inconvenience, the two or three upper courses of the 
 walls were no longer built solid, but were built in three parts ; an inner wall, 
 an open channel, an outer wall.f 
 
 In the diagram, 3 is part of a tie-beam 
 I, from which the rain is dripping; 4 is the inner wall; 5 the gutter; 6 the 
 
 Gutter and Parapet. 
 
 2 of a rafter, covered with lead, 
 
 * Choisy's Histoire, ii. 371. 
 
 + This construction is clearly seen above ; and the foot of Shottesbrooke spire, 395. 
 The outer wall and part of the channel arc seen at Heckington and Winchelsea, 388 ; Beverley 
 St Mary and Lavenham, 387 ; fountains, Netley, St Mary's York, and Bridlington, 385. In 
 England a simpler arrangement than that shown above is usual ; the coping, 8, of the inner 
 wall, 4, being omitted; and the lead, 1. being brought down to the gutter. See Pugin's 
 Examples, i. 4, Merton ; and -/. A. Sketch Book, New Series, xi. 18, Westminster. 
 
394 PARAPETS. 
 
 parapet ; 7 the solid wall ; 8 the " larmier," or " coping," so designed as to prevent 
 the rain from dripping down the wall-face. Supposing that the solid wall is 6 
 feet thick, it may become at the top an inner wall, say 3 feet thick ; a hollow or 
 channel, 2 feet broad ; and an outer wall or parapet, 1 foot thick. Over the 
 inner wall, which may be 2 or 3 feet high, are dripping eaves ; the lead or 
 tiles no longer drip over the outer face of the solid wall, but over the inner 
 wall only, and into the channel, which thus becomes a longitudinal gutter 
 running the whole length of the wall. This gutter is carefully constructed ; 
 sometimes of a hard, non-porous stone ; and to prevent infiltration, is covered 
 with lead at the bottom and at the sides. 
 
 In later Gothic the walls become thinner (at Wells, c. 1175, the aisle- 
 walls are 5 feet 3 inches thick) ; and there was hardly room on the top of 
 the wall for inner wall (French, bahut), gutter, and parapet. Therefore it was 
 usual to build the parapet, not in the plane of the outer face of the wall 
 beneath, but projecting beyond it.* To support such a projecting parapet, 
 corbels or arches might be employed ; but usually sufficient projection was 
 got by using long projecting blocks for the bottom course of the parapet ; 
 the lower edge of these blocks was then molded. So a parapet, when it 
 projects, is but an enlarged edition of the ancient corbel-table ; with this 
 essential difference, that it no longer carries the lead or tiles. And this is 
 the form the earliest parapets take; e.g. in FOUNTAINS NAVE (385.6), 1135 or 
 1 147; Whitby choir ; NETLEY CHOIR AISLE (385.7); SALISBURY (458.4); Ely 
 presbytery, 1140, and Lincoln presbytery, 1160, in both of which a plain solid 
 parapet rests on trefoiled arches, like those of ROMSEY (385) and Beverley 
 nave, c. 1 340, a very late example.f 
 
 In many cases, e.g. in the western bays of WELLS CHOIR (373), the 
 old system of dripping eaves has been converted into the new one ; a four- 
 teenth-century parapet having been superposed on the corbels of 1175, both 
 in the aisle and clerestory. So also in BEVERLEY CHOIR (386.1), the corbel- 
 table containing tooth ornament is thirteenth-century work ; the parapet is 
 a century later. In LINCOLN NAVE at present the walls are crowned with a 
 fourteenth-century parapet ; their original condition is shown in Mr Sharpe's 
 restoration (115). 
 
 The transition from one system to the other is well seen at SOUTHWELL 
 (390) ; where the Norman transept has a corbel-table ; the Gothic clerestory 
 of the choir a corbel-table and parapet, both of the same date ; and the Gothic 
 aisle a parapet \ without corbels. 
 
 Where, however, as in the diagram, the parapet did not project, a horizontal 
 string, 9, commonly called a " cornice," was usually constructed at the level 
 where the solid wall ended and the parapet began. This upper string was 
 molded ; and in its hollow are often found the ball-flower, as in the clerestory of 
 
 * This is well seen in the sections of parapets on 3S5.6 ; e.g. fountains nave. The 
 parapet of netley (385.7) is a remarkable example of porte-A-faux. 
 
 + The choir of Canterbury has an enriched lead parapet. The most ancient example of 
 the parapet in existence seems to be that of the eastern aisle and clerestory walls of the Norman 
 north transept of Ely, where a small molded parapet rests on a corbel table ; illustrated in 
 Stewart's E/y, Plates 9, 10. 
 
 \ This parapet may be not original. 
 
PARAPETS. 
 
 395 
 
 WELLS CHOIR (396), square-leaved flowers, masks, and grotesques, as at 
 
 GRANTHAM (385.4), ENSHAM (385.5), IIKCKIM ,T( )N (3XS.1 1, \YI \< II HLSKA < 3SN.2 /, 
 LAVENHAM (387.4), ST MARY REDCLIFFE, BRISTOL (386.4), Mold, Gresford, 
 Wrexham, TIVERTON, and BRIDLINGTON (390, 124). 
 
 It is interesting to note that the same causes which in the end almost 
 universally replaced the system of corbel-table and dripping eaves by that of 
 parapet and gutter, led also to the building of parapet spires, e.g. SHOTTES- 
 BROOKE and LOUTH 61 I j, instead of the earlier broach spire. 
 
 The gutter system of roof drainage was in constant employment in the 
 Greek and Roman temples. But it does not follow that the Romanesque 
 builders owed it to the survival of Roman building traditions. The break is too 
 long. From Roman times it seems to have disappeared from use till the twelfth 
 century. The chief exceptions are in the south-west of I 7 ranee ; where at 
 Poitiers Notre Dame, and Chauvigny, there are parapets which may be of 
 the eleventh century.* Neither in France nor England did the parapet and 
 gutter system come into general use till the 
 days of Gothic, i.e. till the thirteenth century. 
 With us it appears first at Ely ; then in the 
 Cistercian churches; e.g. the naves of FOUN- 
 TAINS (385. 6) and Kirkstall ; Roche choir, 
 Kyland nave. 
 
 We may suggest that in its origin it was 
 not due at all to drainage difficulties. For 
 military reasons every castle wall and every 
 tower wall had a parapet, with a path at the 
 back of it, by which the defenders could pass 
 along, sheltered by the parapet, to reinforce 
 any spot where attack was imminent. The 
 advantages of the continuous and protected 
 path may well have struck the church 
 builders, especially where, as in many districts 
 of South-western France, the churches became 
 fortresses. It was only necessary to add the 
 solve completely the problem of roof drainage. 
 Romanesque parapets, e.g. those of Notre Dame, Poitiers, are so very lofty that 
 it is plain they were motived by military reasons. As to the embattled 
 parapet of England, there can be no doubt that it was military in origin. 
 
 At first, e.g. in Fountains nave and Lincoln choir, the parapet was regarded 
 simply as something of practical value, and it did not occur to make it a 
 decorative element of the exterior, except so far that its cornice and coping 
 were molded. These plain parapets are in use at later periods also; e.g. HEDON 
 (359) ; and Patrington, fourteenth century. 
 
 But, as with the buttress and the flying buttress, what engineering skill had 
 
 invented, artistic instinct soon proceeded to beautify. In some of the earlier 
 
 examples of the decorated parapet, the parapet remains solid, but is ornamented, 
 
 e.g. at BEVERLEY MINSTER (386.1), with a diaper of square-leaved flowers, or as 
 
 * Choisy's Histoire, ii. 173, -°4- 
 
 Shottesbiooke. 
 
 bahut " or inner wall, to 
 At any rate, the earliest 
 
396 
 
 PARAPETS. 
 
 in ST MARY'S CHURCH, BEVERLEY (387. i ), with quatrefoils studded with pellets 
 (the beauty of the moldings of the cornice should be noticed) : or in the 
 north nave of Lichfield, with foliage. Both the Beverley parapets are of the 
 fourteenth century. At this time tiny figures, or groups of figures, are some- 
 times perched on the parapet, as on that of the chapter house of York ; 
 Beverley nave ; and SELBY CHOIR (86). Howden choir also had a decorated 
 solid parapet, with ball-flower in the cornice. 
 
 But, obviously, a parapet tells much more against the sky-line when it is 
 perforated. Of these pierced parapets one favourite design is constructed with 
 undulatory curves, taken as it were out of the favourite window of the fourteenth 
 century, that with reticulated tracery (618). Exceedingly beautiful is the 
 flow and counterflow of curves thus produced. SELBY CHOIR (86), HECKINGTON 
 
 (388.1), BISHOPSTONE, WILTS (387.3), 
 have fine parapets of this type. 
 
 In France the tracery of parapets 
 and windows corresponds invariably ; 
 period to period.* It was not always so 
 in England. As we have seen, it was 
 not till late, not indeed till the fourteenth 
 century was well advanced, that the 
 practice became general among us of 
 decorating the parapet. But by that time 
 we had discarded, or nearly so, the geo- 
 metrical window tracery which had come 
 into vogue about 1240, and had remained 
 in vogue till about 131 5. By the 
 latter date flowing curves had largely 
 supplanted rigid geometrical curves in 
 our window tracery. To be con- 
 sistent, therefore, we ought to have per- 
 forated our fourteenth-century parapets 
 with fourteenth-century patterns, as 
 Wells. m Selby clerestory, and with nothing- 
 
 else. But we chose to be incon- 
 sistent. We did not like to leave, unused in the parapet, the wealth of 
 beautiful geometrical design which had been wrought out in thirteenth-century 
 windows. And so, in spite of architectural propriety, the fourteenth-century 
 parapet was more often designed after the fashion of the late thirteenth than of the 
 fourteenth century, as in the transept of St Mary Redcliffe. A special favourite 
 among the pierced parapets was a triangle containing trefoils. It is seen in 
 WELLS PRESBYTERY (373); west of which it is superposed on a twelfth-century 
 corbel-table, as above. Tewkesbury apse and MALMESBURY NAVE (375) have 
 charming examples of the same character. So has the fifteenth-century choir of 
 st MARY REDCLIFFE, BRISTOL (3/6); borrowing its design from the fourteenth- 
 century transept (386.4). The liking for geometrical patterns never wore out ; 
 though they are often combined with vertical lines, as in window tracery of the 
 
 * Choisy's Histoire, ii. 396. 
 
 
PARAPETS. 
 
 yj7 
 
 same date: e.g. the clerestory of St George's, Windsor; and the Lady Chapel 
 of St Mary Redcliffe. Fine examples of late pierced parapets arc common in 
 the West of England ; e.g. Bridgwater and Taunton St Mary's. 
 
 It should be added that the pierced parapet sometimes found its way into 
 the interior also to protect the clerestory passage ; artistically, it was valuable 
 where the triforium arcade was insignificant, as at Exeter or BRIDLINGTON 
 (125); or was non-existent, as in Lichfield presbyter)' and SELBY CHOIR ( 390). 
 It is omitted, and is missed, in TEWKESBURY CHOIR (165). 
 
 The artistic value of 
 the parapet to the ex- 
 terior can hardly be 
 overestimated. For the 
 exterior it per f o r m s 
 precisely the same func- 
 tion as the longitudinal 
 ridge rib of the vault 
 (336) for the interior. 
 It stands for unity. 
 Buttresses, flying but- 
 tresses, pinnacles cut up 
 the building into so 
 many blocks, indepen- 
 dent of each other. This 
 independence of the 
 several blocks of nave, 
 choir, transept, menaced 
 by the longitudinal ex- 
 panse of ground-course 
 and string, is finally 
 checked by the powerful 
 horizontal summit-line 
 of parapet or battle- 
 ment : above all, when 
 the parapet or battle- 
 ment is pierced. The 
 predominance of the 
 vertical line in a Gothic 
 design is largely sub- 
 dued by strength of ground-course, string, and parapet ; it is only when it 
 breaks loose into tower and spire, that it succeeds in overpowering the 
 long horizontal lines of parapetted nave and choir and transept. Much has 
 been written of the principle of " vertically " as the characteristic feature 
 of Gothic design. Too much has been made of it. It is sure that even 
 in Early Gothic, as at SALISBURY (458), the builders saw the danger of 
 the undue predominance of the vertical line; and scored the building from 
 east to west, and from north to south, with the most powerful horizontal lines 
 of ground-course, string, and parapet that they could devise. Of these horizontal 
 
 petition, on. iitrie h 13 
 
 Louth, East Cable. 
 
3 9 8 
 
 BATTLEMENTS. 
 
 lines the parapet was the most useful ; for it gave not one horizontal line, but 
 three ; the molded coping, the molded cornice, and the broad band between 
 them. And when that broad band was pierced, a strength of horizontal line 
 was devised that no eye, viewing the exterior, could neglect. And so, in the 
 first fervour of enjoyment of the new design, many a church received a beautiful 
 fourteenth-century parapet ; pierced sometimes with flowing, more often with 
 geometrical patterns. The nave of Lincoln Minster has a fine example, 
 with niched pinnacles. 
 
 Particularly beautiful was the parapet as applied to the slopes of the gables 
 of the roof ; e.g. the east front of LOUTH (397) ; the south transept of Lin- 
 coln ; and the unique parapet of the east front of CARLISLE (128). 
 
 Battlement. 
 
 This form of parapet was clearly borrowed from fortified walls and towers. 
 In it merlons alternate with embrasures or crenels. Through the lower part, 
 the embrasure, the soldier got a view of the besiegers ; behind the merlon he 
 
 Selby Choir. 
 
 stood for shelter. The merlon was pierced with arrow-slits (loopholes or 
 eyelets). In a fortification these loopholes were splayed internally to give 
 wider range to the defender's arrows. In ecclesiastical battlements loopholes 
 occasionally occur, as in Ripon choir ; but they are merely ornamental, being 
 usually cut square.* 
 
 The battlement was in use among the Greeks and Romans ; it is rare in 
 our church-work till the fourteenth century ; after that date, to a large extent, 
 it ousted the horizontal parapet. It is a special feature of Late English as 
 opposed to French Gothic. It gives a sharp and effective outline against the 
 sky ; but its contour is harsh and angular ; and it interferes, every few feet, 
 with the flow of the main horizontal line of the building. Nevertheless there 
 was a craze for it : and much trouble was taken with its ornamentation. In 
 its simplest and earliest forms only the horizontal lines of merlon and embrasure 
 were molded ; e.g. in the porch of ST MARY, BEVERLEY (365). Often the 
 sides of each were molded also ; as at TIVERTON (390) and Gresford. Some- 
 times the upper surfaces and sides were chamfered ; and both the upper and 
 
 * Not always ; in Oundle and Kettering towers they are splayed. — R. P. B. 
 
GARGOYLES. 
 
 399 
 
 the lower edge of the chamfer were molded, as in ST ERASMUS' CHAPEL, 
 WESTMINSTER (386.5). Usually the face was left plain; but in late work il 
 was often panelled; e.g. in the clerestory of NORWICH CHOIR (160). In East 
 Anglia, the battlements are very frequently panelled in patterns of black flint ; 
 as at STRATFORD ST MARV's, SUFFOLK (355 . A still richer example is seen 
 at LAVENHAM (387.4). Sometimes both merlon and embrasure were perforated 
 in very unsubstantial fashion; as in YORK CHOIR 100 ; KING'S COLLEG1 
 CHAPEL (199); and BATH ABBEY i},^). Rarely, the faces were covered with 
 carving ; as at TIVERTON (390 ), with the arms and emblems of John Green, 
 clothier, who founded a chantry and rebuilt the church at the end of the fifteenth 
 century. Most gorgeous of all were the combinations of enriched parapet, 
 battlement, and pinnacle which crowned the lordly towers of Somerset; e.g. 
 TAUNTON ST MARY (60/J. In these miles of plain battlements there is great 
 monotony; seldom, as in the eastern chapels of Peterborough, and king's 
 COLLEGE CHAPEL (199) is trial made of a change of form. Elsewhere several 
 other forms of battlement occur ; e.g. in Scotland ; and at Verona. 
 
 And having in the end monopolised to a very large extent the cresting of 
 the walls, the battlement found its way into the interior 
 also as a decorative ornament, wherever a horizontal line 
 could be crowned with it ; in the cornice of a monument 
 or of a screen ; and with great inappropriateness, in the 
 transoms of the windows and even the capitals of piers ; 
 as at LAVENHAM (446) and King's Cliffe. 
 
 In France the battlement occurs, for military reasons, Woodford. 
 
 in cathedrals of the south, like Beziers, where the wars of 
 
 the religious made every church a fortress. It is very rarely used decoratively ; 
 the cathedrals of Chalons-sur-Marne and Troves are the only important excep- 
 tions. It was tried and rejected. 
 
 Gargoyle. 
 
 The gutter, gargoyle, and " larmier" or coping were all in use in Greek and 
 Roman work ; but were re-invented by the Gothic architects of the thirteenth 
 century. 
 
 When such a gutter as that shown on page 393 has been constructed on 
 the top of a wall, it is necessary to pierce the foot of the parapet at intervals 
 to let off the water that collects in the gutter. It is desirable to throw the 
 water as far away as possible from the face of the wall and the windows; 
 therefore these openings are often made through buttresses. Hut it is necessary 
 to protect the faces of the buttresses also from the falling water. So in the 
 opening a perforated block is fixed, of as much projection as is safe. (More- 
 over, in fourteenth-century work, the buttresses are sometimes constructed 
 more vertically than before, in order to escape, as much as possible, damage 
 from the falling water.) These long projecting blocks are usually carved 
 in grotesque shapes ; e.g. bat>, monsters, men baling out a boat, occur in 
 SKLISY CHOIR 39S . Sometimes only a lead spout is u>ecl, as at \Y( >oi>i'< H<1 > ; 
 sometimes a gargoyle with a lead spout in its mouth. In some cases no 
 
400 
 
 GARGOYLES. 
 
 doubt the grotesques may have been symbolical ; the idea being that the 
 Church overcomes and converts to good uses even the most monstrous forms 
 of evil. 
 
 One difficulty still remained. It was that the fall of the water from the 
 gargoyles of the clerestory down on to the aisle-roof far below in time must 
 damage that roof, if it were covered with tiles. One remedy was to cover the 
 aisle-roof with lead. Another remedy, where flying buttresses existed, was to 
 utilise each as an aqueduct. This is well seen in the flying buttress added c. 1337 
 to the thirteenth-century choir of SOUTHWELL MINSTER ; originally a spout or 
 gargoyle projected from the foot of the pinnacle. Such a flying aqueduct 
 ought to rise to the level of the gutter, as in the Southwell example. But the 
 vaulting thrust, which it is the function of the flying buttress to transmit, is 
 mainly felt at a considerably lower level. The head of the flying buttress, there- 
 fore, ought to be placed about the level of the one at Southwell on the right ; 
 otherwise great damage may result. Indeed at Famagusta in Cyprus, flying 
 
 buttresses, set too high, in order that they might 
 act as aqueducts, did actually ruin the church ; 
 and they would have done so in Evreux nave, 
 if the\ r had not been reconstructed.* In 
 Chichester nave the flying buttresses serve as 
 aqueducts, yet are placed at the proper level 
 to transmit the thrusts of the high vault, It 
 follows that the head of each is considerably 
 below the level of the clerestory gutter. Vertical 
 pipes, therefore, are passed through the wall at 
 intervals from the gutter to the head of the 
 flying buttress. Such pipes, however, are easily 
 choked, and in such a case great damage may 
 be quickly done to the masonry by infiltra- 
 tion. In the presbytery of Ely the upper flying 
 buttress served as an aqueduct, and the rain from 
 it originally passed through a hole made in the 
 centre of each buttress just below the pinnacle. 
 But when the presbytery was remodelled in the fourteenth century, a gar- 
 goyle was constructed at the side of the buttress, so as to do away with the 
 risk of infiltration into the inner masonry of the pinnacle.f But on the whole 
 this utilisation of the flying buttress did not find much favour in England. 
 In France, on the other hand, it was worked out to a complete solution. At 
 first two flying buttresses were superposed ; compare those of WESTMINSTER 
 ABBEY (379) ; the upper one forming the aqueduct, with its head at the 
 level of the clerestory gutter. Later, as at Amiens nave, Auxerre, Bordeaux, 
 Nicosia, Famagusta, Cologne, Eu, Abbeville, there was placed on the flying 
 buttress proper a balustrade, and on the balustrade an inclined bar, the upper 
 surface of which was hollowed as a channel for the rain. A flying buttress 
 of this type, but of decorative origin only, and with a double bar, is seen 
 in LOUTH SPIRE (611). In England the clerestories were usually some 50 feet 
 * Enlart's Histoire, 516. + See section in Stewart's Ely, 4. 
 
 Southwell Choi 
 
SPOUTS. 
 
 401 
 
 lower than those of France; and so the fall of water from the upper gargoyles 
 was less dangerous ; we were therefore satisfied usually to cover the aisle-roofs 
 with lead, and leave the flying buttresses to their proper function. 
 
 KetterinL 
 
 Spouts. 
 
 But, it may be asked, why did not the builders save themselves all these 
 complications by employing spouts, horizontal and vertical ? One reason was 
 that they could not cast iron spouts. But as a matter of fact 
 the>- did try vertical spouts of lead to carry off the rain in 
 the gutter ; e.g. in the fifteenth-century work at KETTER- 
 ING CHURCH and in King's College Chapel, Cambridge. 
 Earlier still, in 1241, an order was given that lead spouts 
 should be put up from the top to the bottom of the White 
 Tower of London, which had been newly whitewashed, that 
 the whitewashed walls might not be injured. Lydgate, in his poem, "Troy," 
 1555, speaks of houses 
 
 " With spoutes thorough and pipes as they ought 
 From the stone worke to the canell raught." * 
 
 It was not then that they were unacquainted with spouts, but probably because 
 they knew the mischief of choked spouts and preferred an open-air system 
 under observation and easily kept in order, that they adhered to the system 
 of gutter, parapet, and gargoyle, or else of dripping eaves. 
 
 Willis' Architectural Nomenclature, 38. 
 
Chapter XXVI. 
 THE PROTECTION OF THE WALLS FROM RAIN. 
 
 Ground-Courses — Strings — Dripstones — Hood-Molds — Labels. 
 
 So far we have spoken of the methods adopted to drain the roofs. But it is 
 necessary to protect from the drip of the rain the walls also, especially at their 
 foot, and the heads of windows and doorways. For these was contrived a 
 system of ground-courses, string-courses, dripstones, and hood-molds. 
 
 Ground or Basement Course. 
 
 For the real stability of a building, as well as for the satisfaction of the eye, 
 which demands something to soften the abrupt transition from the horizontal 
 ground to the vertical wall, the walls require a spreading base. In a Greek 
 temple this took the form of steps. In early Romanesque work the pro- 
 jecting basement was usually composed of one or more projecting rectangular 
 courses. But rain dripping from the eaves on to the level top of these soon 
 wore down a hollow, and moreover bespattered and injured the surface of the 
 wall adjacent. To get rid of this inconvenience, therefore, a 
 straight chamfer or chamfers was employed ; e.g. in FOUNTAINS 
 NAVE (679.1). But care had to be taken in the use of this 
 chamfer. If a joint occurs in the chamfer-plane, it will be seen, 
 e.g. at A, A, A, that it will be what is technically called " feather- 
 edged " ; * the dotted triangles will disintegrate early. It will be 
 seen that in WHITBY (679.3) and FOUNTAINS (679.4) choirs a 
 feather-edged joint occurs : but not in the other examples. 
 Secondly, still further to prevent the rain from getting into the joints, as 
 many of them as possible are placed under a projecting member. This 
 practice was long in coming to maturity. In Fountains nave every joint 
 is exposed ; but protection is given to two joints at HEXHAM (679.2), 
 to one at Whitby and Fountains choir ; to five at RIEVAULX (5), to 
 eight at BRIDLINGTON (6). So also, except in the two lowest courses, 
 every joint is protected in the fourteenth-century ground-course of WEL- 
 BOURNE (7), and the fifteenth-century one of KETTERING (8). The great 
 amount of projection given in a fourteenth-century basement is well seen at 
 TIDESWELL opposite. Care also had to be taken against excessive undercutting 
 and excessive projection of the protecting members. The fragile upper course 
 of the Whitby basement should be contrasted with those of Welbourne, which 
 have the " scroll-molding." Basement-courses are exposed to an exceptional 
 * Pugin's True Principles, 1 5. 
 
GROi'xn-rnt'ksi'.s. 
 
 •i '■ 
 
 amount of wet ; that of Whitby, in a rainy district, is not good construction. 
 Again, no hollows should be provided where the rain can lodge ; there are two, 
 however, at Whitby ; and at Hexham, a particularly bad example, there are no 
 less than four. Especially is it important that the rain should not be allowed 
 to trickle down from one member to another till it reaches the ground. It 
 does so in Fountains nave, in the two upper courses at Hexham, in the four 
 lower courses at Whitby. But at Rievaulx two of the projecting rolls, and at 
 Bridlington four, are so designed beneath that the drip is cut off by a hollow 
 {coupe-larme or " throat" , and is compelled to leave the face of the stone. 
 
 Then there was the question what projection a basement ought to have, 
 and what height, in proportion to the height and solidity of the walls. In the 
 thirteenth centurv and onwards the basements grew greatly in height and bulk. 
 
 
 B^L; u-> ■*'-■■ -Cr- ~; r- 
 
 4 
 
 Tideswell. 
 
 Then came in artistic considerations ; how best to contrast vertical line, chamfer, 
 and curve; and how to get beauty of curve. In the fourteenth century the 
 straight chamfer gave way greatly to the subtle ogee curve ; e.g. at TIDESWELL. 
 Especially was it sought more and more to get nice gradation of high light, 
 half light, and shadow. The ground-course of Fountains nave is shadowless; 
 there is little shadow at Whitby or Fountains choir ; but there are two broad 
 bands of shadow at Rievaulx and Welbourne, and three at Bridlington and 
 Kettering. In these protected recesses panelling might be introduced ; e.g. in 
 Cawston tower. In the last fifty years of English Gothic bands of quatre- 
 foils were often introduced; as in the DIVINITY SCHOOL, OXFORD (492), and 
 KETTERING (95). Sir G. Scott* was of opinion that the English basement- 
 
 Lectures, i. 164. 
 
4 04 GROUND-COURSES. 
 
 molds of the thirteenth century were peculiarly excellent. " I have never seen 
 any in France to equal many of our own in the quality of appearing eminently 
 fitted to support the whole structure, or in the artistic arrangement of their 
 parts." Mr E. Sharpe says,* perhaps with more justice, "in the use and design 
 of the moldings of the basement-course the Curvilinear period (1315-1360) 
 surpasses all others in England. Here, as well as in the set-offs of the buttresses, 
 the ogee comes in, with its fine sweep, with great effect. The smallest churches 
 of this period derive a dignity from this bold spreading external feature." But 
 grand examples of fifteenth and sixteenth century basement-molds also occur; 
 especially in the towers of Norfolk and Suffolk. For other illustrations of 
 basement-courses, see LINCOLN, SALISBURY, and SOUTHWELL (115, 458, 359). 
 
 The moldings of the basement of the buttress were usually the same as 
 those of the basement of the wall. See plates of buttresses (352-357). 
 
 String. 
 
 A string or string-course is a projecting course of masonry, usually 
 horizontal, whether external or internal. In thin Anglo-Saxon walls of rubble.t 
 the horizontal strings are probably bonding-courses, used to strengthen the weak 
 and bad masonry of the wall ; and may be derived from surviving traditions of 
 Roman craftsmanship, or from observation of the bonding-courses of long tiles 
 in existing Roman walls at Leicester, Pevensey, Richborough, and elsewhere. 
 
 In the earliest large Norman churches, Bernay and Jumieges, J string- 
 courses may be seen just coming into existence again. At Jumieges there is no 
 external string at all ; and in the interiors of both there is no string to separate 
 triforium and clerestory ; while in Bernay and WINCHESTER (261) transepts there 
 is no string to separate pier-arcade and triforium. After this the different stories, 
 external and internal, were carefully marked by strings. But very soon, even 
 in the Abbaye-aux-Hommes, Caen, and NORWICH (31), before the eleventh 
 century was out, strings were introduced which did not demarcate vertical 
 stages of the building, but were purely decorative.^ In the twelfth-century 
 aisles of Rochester, Romsey, Peterborough, Northampton St Peter's, Southwell 
 porch, the external abaci of the window shafts were continued to form 
 additional strings. In the Abbaye-aux-Hommes and Norwich an additional 
 string was got by continuing horizontally the dripstone of the lower window 
 of the aisle. At Cerisy, Winchester, and Durham, the external abaci of the 
 clerestory windows became strings, as they did internally at Winchester and 
 Dunfermline. The importance given to the string at Norwich, begun 1096, is 
 very remarkable. It is employed in greatest abundance and richness in the 
 west front of Ely. The great artistic value of the string is perhaps only realised 
 when it is absent, as in Malvern nave and GRESFORD (214) ; without it the 
 wall is bare and desolate. 
 
 * Lincoln Excursion, 133. t Baldwin Brown, 296. 
 
 X Illustrated in Ruprich- Robert, Plates 11, 12, 13. 
 
 § A string which separates stones is called in French a bandeau; one which does not 
 is a cordon. 
 
GROUND-COURSES. 405 
 
 In Gothic, after c. 1260, the decorative string or cordon is of less import- 
 ance; for the simple reason that the windows of aisle and clerestory extend 
 nearly or quite from buttress to buttress. In the facades it is especially useful 
 in emphasising their stages. The beautiful eastern transept of BEVERLEY 
 MINSTER (176), instead of being naturally divided by the windows into three or 
 four stages, is divided by four strings into five stages, and gains greatly in 
 apparent height; for who can imagine that a building five stages high is not 
 lofty? The east front of ELY (464), which is divided into four stages only, 
 looks low in comparison. The west tower of WYMONDHAM (589) has six 
 strings and seven stages. The eastern transept of Canterbury has eight strings. 
 
 Romanesque strings are often carved ; e.g. with zigzag or billet, as at 
 Norwich; the Southwell strings are exceedingly rich. In Gothic the string is 
 generally molded. 
 
 Like parapet and basement-course, the string has great value in binding 
 into one whole the disconnected units of the building, severed externally by- 
 buttress, flying buttress, and pinnacle, and internally by pier and vaulting shaft. 
 Still further to reduce the disuniting power of the buttress the string is often, 
 especially in the thirteenth century, carried round it. At Salisbury the greatest 
 possible emphasis is given to string, basement-course, and parapet, to curb as 
 much as possible the aspiring verticality of the design ; nowhere are the 
 buttresses scored so deep and broad with horizontal lines. 
 
 One main use of the string is to act as a drip-course. Indeed in such 
 an example as NORWICH (160;, the string is but a horizontal continuation of 
 the dripstone of the window ; or, to put it the other way, the dripstone is but 
 the arched form of the string. String and dripstone, therefore, are very fre- 
 quently the same in profile. So also a string is often but a continuation of the 
 abacus; and ma)- be identical in form. Indeed very similar conditions dictate 
 the form of the coping of the parapet, the string, the dripstone of the window, 
 the hood-mold of the doorway, the set-offs of the buttress, and the basement- 
 course ; and all naturally tend to take similar profiles. 
 
 What has been said, therefore, above of the molding of the basement-course 
 is largely applicable to the string. It is undesirable that its upper surface should 
 be horizontal ; for that, in an exterior, affords lodgment for snow, and in the 
 interior to dust ; and if an external string be flat-topped, the face of the wall 
 near is bespattered with rain. Such strings, therefore, as that of FURNESS 
 (681.3) are objectionable. Vet, strange to say, the flat top was now and then 
 reverted to in later Gothic; e.g. 683.17. Again, the string should not pro- 
 ject so much or be undercut so much as to be fragile: as at NETLEY (681.16), 
 BRIDLINGTON (681.18 , and ST MARY'S, YORK (683.3). And if it is an external 
 string, it should be undercut to cut off the drip. The "throat" is usually 
 present in thirteenth-century work, as at Netley, BRIDLINGTON (683.1), and 
 HOWDKN (683.5) ; but is often absent from fourteenth-century and later strings : 
 it appears, however, in examples 683. IO, 12, 17, 18. Finally, if a black 
 band of shadow is desired beneath it, it must be deeply undercut, as at Netley ; 
 but if the hollow is to be filled with foliage, it should be broad ; deep enough to 
 protect the foliage, but not so deep as to hide it from sight ; as it does at 
 Tiverton ; a hollow "casement" strikes the happy mean. 
 
4 o6 HOOD-MOLDS. 
 
 Norman strings are commonly square or semi-hexagonal. The strings 
 of FOUNTAINS and FURNESS (68 1) are of mid-twelfth century. A little 
 later, c. 1170, the ROCHE string (4) is a pointed " bowtell," instead of being 
 semicircular ; and that of BYLAND (6) ceases to be flat-topped. JERVAULX 
 (9), a few years later, has a filleted bowtell. WHITBY (11), NETLEY (16), 
 BRIDLINGTON (18), TINTERN (683.2), and ST MARY'S, YORK (3) strings 
 have the characteristic narrow, deep, semicircular hollow, and date from the 
 middle to the end of the thirteenth century. Throughout the whole century 
 the scroll-molding came more and more into fashion ; in the fourteenth century 
 the scroll or some variant of it was still more common (e.g. AUSTREY, 683.12), and 
 its lower curve was an ogee. Sometimes, as in 683.17, the coupe-larme is re- 
 tained ; but the special feature is the presence of ogee curves or " wave-moldings." 
 In later Gothic strings also the ogee or wave curve or the double ogee is rarely 
 absent ; but the special characteristic is that they are usually flat-topped, and 
 that the hollow becomes a broad shallow "casement," as in 683.15 and 13.* 
 
 Dripstone, Hood-mold, Label. 
 
 These terms are often used indiscriminately of the strings which surmount 
 openings, whether windows, doorways, or pier-arches. When the opening is the 
 outer side of a window or doorway, the term dripstone is appropriate. But if the 
 opening is within the building, where there is no rain to drip, it is better to call 
 the string a hood-mold, not a dripstone. The term label may be dispensed with 
 altogether; strictly it is applicable only to a rectangular dripstone over a doorway 
 or window."]- Weather-mold and water-table are synonymous with dripstone ; 
 and may also be dispensed with. 
 
 Just as strings are desirable to keep the drip of the rain off the walls, so 
 dripstones are necessary, and to a much higher degree, to prevent the rain from 
 dripping into the moldings of the window-arch and its tracery and the leading 
 of its panes ; or on to the moldings of a door-arch and the door itself. The 
 main function, therefore, of string and dripstone being identical, and string and 
 dripstone frequently being continuous, they are often similar in profile. The drip- 
 stones, strings, and abaci at Binham, WEST WALTON (432), and Netley are of 
 much projection : at Binham and West Walton, in the thirteenth century, the 
 excellent Barnack stone was employed. Two dripstones are given from AUSTREY 
 (683.11, 12), fourteenth century : the second has the scroll molding, the first is a 
 filleted scroll ; both have an ogee curve beneath ; they should be compared with 
 strings 10 and 9. 18 is a late fourteenth-century hood-mold from COTTINGHAM. 
 
 Sometimes, in early work, carved dripstones and hood-molds occur : e.g. in 
 Norman work in a window in St John's, Chester ; and in the eastern arch of 
 LLANDAFF CHOIR (580). Hood-molds of billet, zigzag, &c, are seen over the 
 windows of Ely, Peterborough, and Norwich ; over IFFLEY WEST DOORWAY 
 
 * For a fuller treatment of the moldings of strings, dripstones, and hood-molds, see Paley's 
 Manual of Gothic Moldings, Section 11 and Plate 16. In studying strings, and indeed moldings 
 in general, it must be borne in mind that each architectural province, e.g. that of Northants and 
 the surrounding districts, has its special characteristics ; some molds are exceedingly frequent ; 
 others, common elsewhere, hardly occur at all. 
 
 + Cf. Brandon's Analysis, ii. 35, note. 
 
HOOD-MOLDS. 407 
 
 (574); over the pier-arches of Waltham, Steyning, and St Alar-ant at Cliffc. 
 The practice continues into the thirteenth century; the clerestory windows 
 of Lichfield nave, c. 1270, having dripstones with the tooth ornament. 
 
 Usually the dripstone follows the curve, if any, of the opening, lint from 
 the fourteenth century an ogee dripstone, usually crocketed and terminating in 
 a finial, frequently surmounts the pointed arch of a doorway, as at ST MARY 
 BEVERLEY (365) and CLEV (85); and occasionally of a window, as in the 
 Lady Chapel of Lichfield, and in LEVERINGTON PORCH (84;; in Llandaff aisle 
 both the head of the window and the dripstone are ogee arches. In the fifteenth 
 century and onward the dripstone {label) of the doorway is often horizontal ; as 
 at sale (575) and KETTERIXi, (95). 
 
 In Early Xorman work, e.g. at WINCHESTER (261) and Malvern, both 
 dripstone and hood-mold are often omitted. So also over the pier-arches 
 of the thirteenth-century church at Yarmouth ; and often, at all periods, in 
 small churches. In man}- of the latest and finest churches also it is omitted ; 
 e.g. at Wrexham and GRESFORD (214); and in large Norfolk churches such as 
 CAWSTOX (552), Worsted, Trunch, and St Andrew's, Norwich. These Norfolk 
 churches are built of flint and rubble ; and it may be that it was desired to 
 economise freestone, which had to be brought from a distance.* But as the 
 clerestory string is also omitted, and all the arches are chamfered but not 
 molded, it may well be that the builders purposely designed the interiors with 
 a plain base as a foil, omitting dripstone and string and molding of arch, that 
 nothing might arrest or detain the gaze upward to the magnificent open roofs 
 which all the above churches possess. 
 
 Both dripstone and hood-mold require some kind of finish. In early 
 work this is sometimes bungled, as in the chapels of Kirkstall transept and 
 the doorways of Glastonbury Lad}- Chapel. In Early Norman work, e.g. 
 Chester St John's, the abacus is sometimes prolonged for the purpose. 
 Or the ends of the dripstone die into the buttress on either side ; as in 
 the cloisters of Salisbury, Westminster (489), and NORWICH (506). Or 
 one hood-mold meets another, and mitres into it ; somewhat artlessly at St 
 Mary's, Shrewsbury ; late twelfth century. Frequently, when the windows are 
 small, the dripstone " returns," i.e. is continued as a string, but at right angles. 
 This treatment is naturally most common in the twelfth and thirteenth century ; 
 e.g. Southwell porch and WELLS CHOIR (i73). But it may be seen also in late 
 work ; e.g. in the clerestory of Gloucester nave, and the Lady Chapel of St Mary 
 Redcliffe. In the fourteenth century this treatment is less common ; at this 
 time the dripstone very commonly terminates in a couple of heads.t 
 
 But, far more frequently, both hood-mold and dripstone terminate in some 
 kind of corbel ; it is called a "stop." Much ingenuity and fancy is put into 
 these "stops," and they should be specially noted. Their variety is so great 
 that they hardly admit of classification. { A Norman example from Malmes- 
 
 * But the omission of internal strings and hood-molds is just as common in the Late ( ,othic 
 of Somerset, where good stone was plentiful. 
 
 + See Rickman, 203, 266 ; and Paley's Gothic Architecture^ 176.5. 
 
 \ The "buckle" is worth note; a corbel producing by its shadows a reminiscence of the 
 human face. 
 
4 cS HOOD-MOLDS. 
 
 bury* also appears above the windows of Southwell porch. Leaf-scrolls are 
 common in the Early English work ; e.g. at Stanwick ; f and on a piscina in 
 Strumpshaw Church. t Heads, animals, and grotesques occur at all periods ; 
 the example from Merton College § is c. 1280. In fourteenth and fifteenth 
 century work, instead of being continued into a string, the dripstone has often 
 but a short return ; a truncated string ; as at St Martin's, Canterbury. || Or this 
 bit of string is curled up to form a circle, or extended so as to form a square, 
 rhombus, or hexagon ; as at Chippenham."! When heads are employed here or 
 elsewhere, they are often a clue to date ; each period having its characteristic 
 headdress ; e.g. in the reign of Edward I. a common stop of the dripstone is the 
 wimpled headdress.** 
 
 It has been pointed out above that a dripstone is necessary for practical 
 reasons outside, but not inside a building. In strict logic, therefore, it should 
 not be employed as a hood-mold over pier -arches. The French, more 
 faithful to the logic of Gothic construction than ourselves, almost always omit 
 it. But the eye greatly desiderates that the demarcation between support and 
 load shall be emphasised. ff It would seem that if it is right to mark off the 
 supporting pier from the supported arch by a capital, it is just as right to mark 
 off the supporting arch from the supported wall by a hood-mold. 
 
 * Glossary, 98. + Glossary, 98. % Colling's Details, i., E.E., 2. 
 
 § Colling's Details, i., E.E., 2. || Glossary, Text, p. 188. 
 
 1 Glossary, 98, and Brandon's Analysis, ii., Perp., 26, 29. 
 
 ** " Corbels, date of headdress. This is worth noticing, because the date of a church may 
 thus sometimes be ascertained. The principal varieties appear to be these : — 
 
 "About the year 1300, the wimple or handkerchief round the neck and chin. 
 
 " 1350. The net, confining the hair back from the forehead. 
 
 " 1380. The hair itself is braided in a square plait on each side the forehead. 
 
 " 1410. The crespine, resembling the latter style, but covered with a veil, which reaches to 
 the shoulders. This, towards 
 
 " 1430, takes a lunar or horned shape, which grows more and more outrageous till about 
 1470. 
 
 " 1480. The butterfly ; where the hair is worn in a net behind, with pinners, on wire, like 
 wings. 
 
 " 1 500. The kennel, or triangular forehead dress. 
 
 " Maidens wear their hair loose, without any headdress ; and widows, in late times, have 
 a gorge at the chin " {Handbook of English Ecclesiology, p . 1 1 9). 
 
 tt An arch without a hood-mold is as a fair face without eyebrows. — R. P. B. 
 
Chapter XXVII. 
 
 ROMANESQUE CAPITALS. 
 
 Function of Capitals — Cubical and Scalloped Capitals— Interlacing, Figure, and Storied 
 Capitals — Semi-Naturalistic Capitals. 
 
 The main function of the abacus and capital is to provide for the sprino-in^ 
 of the arch a surface of greater area than is provided by the pier. 
 
 In Byzantine work, when the supports were but slender columns, it was 
 sometimes the practice, e.g. in S. Vitale, Ravenna, to superpose another cap 
 (French, dosserei). In Romanesque and Gothic this additional upper capital 
 is not found. 
 
 The depth of the capital is or should be proportionate to the diameter 
 of the shaft or column. This is well seen in the capitals of Lincoln choir ; 
 and earlier still, in those of CHICHESTER RETROCHOIR (245). 
 
 From an artistic point of view, capitals are valuable in two ways. First, 
 the)- mask the awkward junction where the curving lines of the arch meet 
 the vertical lines of the pier. Secondly, where the plan of the pier is different 
 from the section of the arch, the junction of the two different sets of moldings 
 would be most unpleasant. But, where, as often in late work, e.g. at TENBY 
 (410), pier and arch have the same moldings, the capitals may be omitted ; 
 the moldings running continuously and uninterruptedly from the base of one 
 pier up to and round the arch and down to the base of the next pier.* Or, 
 sometimes in late work, as in CHIPPING NORTON (548), the inner order of 
 the arch alone is supported by a capital : the outer order of the arch runs down 
 to the ground. These two methods of procedure may be right logically ; 
 both are nevertheless objectionable. If there is what Willis calls a "dis- 
 continuous impost," then, as he says, " the arches appear ready to slip down 
 the sides of the pier, having nothing to rest upon." The capitals are omitted 
 in the thirteenth-century vestibule of the chapter house of Chester Cathedral ; 
 it is rare to find so early an example. It is very common abroad in Late 
 Gothic ; e.g. the cathedrals of Orleans, Louvain, and Antwerp. 
 
 CnncAL ok Cushion Caps. 
 
 In Norman and Lombardic work the two chief capitals employed are the 
 Corinthian and the cubical. The latter, seen in Ernulph's work at CAXTKR- 
 BURY ''430,1, is often asserted to be the earliest; but both in Normandy and 
 
 * In the Lady Chapel of LONG MELFORD (665.12), though the moldings of pier and 
 arch are identical, capitals are employed. 
 
4io 
 
 CUBICAL CAPITALS. 
 
 Tenby. 
 Ickleton. 
 
 England the Corinthian is 
 more often found till c. 1090. 
 Of the Romanesque pro- 
 vinces of Europe, those of 
 Lombardy, Germany, Nor- 
 mandy, and England use the 
 cubical capital most. In 
 Lombardy the abacus is often 
 very massive, and practically 
 becomes a dosseret. Ex- 
 amples in the Brera Museum 
 from the destroyed Church 
 of Aurona have been at- 
 tributed by De Dartein to 
 the eighth century. At any 
 rate in the eleventh century 
 they may be seen in abund- 
 ance, e.g. in S. Stefano, 
 Bologna, and the adjoining 
 cloister ; in the crypt of S. 
 Miniato, Florence; at 
 Aquileia (1019-1025); in S. 
 Abbondio, Como; in S. 
 Theodore, Bologna, 1190; 
 in Pavia they are not seldom 
 executed in bricks. In Ger- 
 many they are very com- 
 mon ; e.g. at Speyer, Augs- 
 burg, Bamberg, Ratisbon. 
 They are common in Nor- 
 mandy, and not infrequent 
 in Anjou and Saintonge ; 
 they occur also in the Primi- 
 tive Romanesque crypt of 
 Nevers Cathedral. 
 
 As to the origin of the 
 cubical capital there has been 
 much controversy. It is very 
 common, executed in wood, 
 in the Romanesque churches 
 of Norway;* which is natural 
 enough, as it is a form easily 
 obtained from wood. And 
 as the Normans came from 
 Scandinavia, Ruprich-Robert 
 
 * Several of these are illustrated 
 by Ruprich-Robert, vol. ii. 
 
CUBICAL CAPITALS. 411 
 
 predicates a Scandinavian origin for the cubical cap. On the other hand we may 
 urge that Scandinavia did not receive Christianity till the time of Olaf Trygvason, 
 c. 1000; and that consequently any church architecture there must be consider- 
 ably later than the time when the Normans conquered and settled in Normandy. 
 Again, the existing timber churches of Norway, from the richness and character 
 of their ornament, would seem not to have been built till well into the twelfth 
 century. Also, the stone cubical capital of Normandy is broader than the column 
 or shaft : and is moreover executed in a separate block. Hut in the Scandinavian 
 examples capital and post are all in one piece ; the capital is formed in the 
 simplest way by squaring the upper part of the round post into four faces ; thus 
 the four flat sides of the capital do not project, but are " in retreat." And even 
 if the Scandinavian theory explained the Normandy capitals, it would still fail 
 to explain the great multitude of cubical capitals in Lombardy, German} - , Anjou, 
 Saintonge, where no Scandinavian influence can be predicated.* 
 
 Another theory is that these capitals are of Byzantine origin. They occur 
 at S. Vitale, Ravenna ; and this church was imitated by southern workmen, 
 whom Charlemagne had sent for, at the end of the ninth century at Aix-la- 
 Chapelle. Other examples occur in St Mark's, Venice, commenced in 1063. 
 Examples occur as early as the seventh century, at Parenzo,f on the Adriatic ; 
 where the fact that the capital bears a dosseret, may point to Byzantine 
 influence.} From Ravenna also it may well have passed to Lombard}-, and 
 from Lombardy to France. 
 
 But it is quite as probable that this capital was not borrowed at all in 
 Northern Europe, but was a result of a peculiar process of craftsmanship. 
 The Romans themselves used the lathe in turning shafts ; and the lathe was 
 in constant use among our own Anglo-Saxon builders, as it was from the 
 eleventh century in France ; § e.g. the monolithic columns of the choir of St 
 Etienne, Nevers, dedicated in 1099 ; monoliths turned in the lathe are in great 
 use also in the choirs of the Romanesque churches of Auvergne. In Berry and 
 Poitou also they are common ; and sometimes there may still be seen the hori- 
 zontal grooves produced by the lathe. Now the shape of a cubical capital, the 
 penetration of a cube and a sphere, is precisely what would suggest itself to 
 craftsmen using lathes. It is probable then that in districts where the lathe 
 was in use for turning shafts, the cubical capital was reinvented ; and that from 
 these districts it spread to their neighbours." 
 
 Where the face of the cubical capital was left plain, it seems to have been 
 common to paint it ; painted capitals survive at Jumieges and St Georges de 
 
 * See Brutails, 45. + Dehio, Plate n. 
 
 I It should be mentioned, however, that the dosseret also occurs in S. Stefano Rotondo, 
 Rome, consecrated 467, where no Byzantine influence can be suspected. 
 
 S Viollet-Ie-I)uc, Architecture, iii. 495. 
 
 II Or it maybe regarded as the natural way— which would occur to many districts inde- 
 pendently— of accommodating a square to a circle. " If a cubical block of stone be placed on 
 a round shaft the diameter of which is less than a side of the square, and if now all the surface 
 material be cut away at the bottom, so that the large square above gradually changes and 
 diminishes into the circle beneath, we get the form of the new impost-capitals" (Lethaby's 
 Mediaval Art, 39). 
 
 IT See Brutails, 81. 
 
412 
 
 1. Ely Transept. 
 
 2. Malmesbury Nave. 
 
 3. Lessay Nave. 
 
 4. Steyning Nave. 
 
 5. St Davids Nave. 
 
 6. Wells Nave. 
 
 7. Steyning" Nave. 
 
 8. Dore Choir. 
 
SCALLOl'HD CAPITALS. 
 
 413 
 
 Boscherville.* In the eastern aisle of the Norman transept of ELY 412.1) 
 cubical caps may be seen with painted foliage: and others where painted foliage 
 has been rendered, later on, in stone. In very many cases plain cushion caps 
 have been carved later on ; specimens in all stages of development may be seen 
 in the crypts of CANTERBURY (193) and WESTMINSTER (41 5.1). Also capitals 
 rudely carved in the eleventh century were sometimes recut in the twelfth ; to t lu- 
 great confusion of archaeologists ;+ e.g. in Romsey choir and Porchester Church, f 
 
 Scalloped Caps, or Subdivided Cushions. More often, instead of painting 
 or carving the flat faces of a cubical capital, another process was adopted, which 
 was to be very fertile in results. It was to subdivide each face into two, three, 
 or more cushions. When the cushion cap is much subdivided, as at MALMESBURY 
 (412.2), it is usually well advanced in the 
 twelfth century. But this is not always so. 
 Fully developed scalloped capitals, possibly 
 imitations of Norman work, occur in the 
 Anglo-Saxon belfries of Gee and Brace- 
 bridge ; and at Branston § in an arcade on 
 the west wall of the tower, side by side with 
 Corinthian and cubical capitals. In Nor- 
 mandy, according to Ruprich-Robert, the 
 oldest examples go back to the last years of 
 the eleventh century. Those of Lessay are 
 not necessarily much later. They occur also 
 in St Georges de Boscherville, as well as 
 in the Abbaye-aux-Dames. In Norwich 
 Cathedral, begun 1096, capitals similar to 
 those of LESSAY ^41 2.3) occur in the triforium 
 of the apse, and in the eastern bay of that of 
 the south nave." 
 
 By the middle of the twelfth century the- 
 scallops had become so numerous that the 
 whole capital often became a fringe of tiny 
 cones. These have been termed by Mr 
 Sharpe coniferous capitals. The spaces be- 
 tween the cones were filled up with inverted 
 cones, beads, &c. ; and thus a very rich and 
 beautiful capital was elaborated out of the heavy, unsightly cushion ; e.g. 
 STEYNIXG ** (412.4). Very fine scalloped and coniferous capitals occur at 
 Ramsey Abbey, Sutton St Mary, Tilney All Saints, and Walsoken.-f-f By 
 
 * Dehio, Plate 312. 
 
 + See Anthyme St Paul's pamphlet on the subject. 
 
 I In WINCHESTER LADY CHAPEL is a cubical Norman capital carved into a mask; 
 probably in the fifteenth century. 
 
 S Illustrated in Baldwin's Brown's Arts in Early England, ii. 161. 
 
 || Ruprich-Robert, Plate 93. ^ Illustrated in Rickman, SS. 
 
 ** For other illustrations see Sharpe's Ornamentation of the Transitional Period; and 
 Sharpe's Lincoln Excursion. A beautiful variant is one in which the cones imitate the folds of 
 drapery. 
 
 tt Those of Walsoken are illustrated in Colling's Details, Norman, i. 6. 
 
 Winchester Lady Chapel. 
 
4 i4 SCALLOPED CAPITALS. 
 
 analogy with those of Sutton St Mary, the date of which is definite, probably 
 none of the above are earlier than 1 1 80. 
 
 Towards the end of the period, e.g. at ST DAVID'S (412.5), begun 1180, 
 and specially in those churches which were put up in the West of England 
 Gothic, a pretty modification was adopted ; the pier looks like the stump of a 
 willow tree that has been pollarded. The cones now are incurved and seen to 
 be growing out of the pier ; and the upper ends are cut off sharp, as it were 
 by a knife. The capital with incurved cones ; * e.g. in the west nave of Worcester, 
 the east transept of Hereford, the arch leading into the north aisle of Lichfield 
 choir, under the tower of Cheltenham, and in several churches of the Welsh 
 border ; is specially characteristic of the late Transitional or Early Gothic of the 
 twelfth-century school of the west country .f 
 
 The next step, as in several examples at ST DAVID'S (412.5) and WELLS 
 (412.6), is for the cone to become a little hollow trumpet; to bud and blossom 
 and bring forth leaf and flower. Exquisitely beautiful and consummate in 
 execution as is this capital from the western bays of the nave of Wells — nothing 
 better was ever done in English art than the capitals of Wells — there is good 
 reason to believe that it is not later than the year 1 206. 
 
 The next capital is from STEYNING (412.7). Here the carver has retained 
 the filling-in between the cones, but has omitted the cones. This capital is seen 
 also at Icklesham, near Hastings. 
 
 Interlacing Capitals. 
 
 The richer eleventh and twelfth century capitals, e.g. those of ELY TRANSEPT 
 (412. 1), ST PETER'S, NORTHAMPTON (415.6), OXFORD CATHEDRAL (417.1), and 
 ABBEY DORE (412.8), often contain interfacings ; sometimes of flat bands, some- 
 times of pipes, sometimes of foliage, sometimes monsters or human beings inter- 
 twined after the fashion of those on the shafts of the doorways of IFFLEY (256), 
 NORTHAMPTON (256), SHOBDON (415.3) and Kilpeck. As interlacings are 
 very common in the Early Irish missals and crosses, and also in those of the 
 Anglo-Saxons, some have attributed to these a Celtic, and others an Anglo- 
 Saxon origin. But as they occur also very frequently in Byzantine work of 
 the sixth century, e.g. in the screens of S. Vitale, Ravenna, and again in the 
 eighth century, J both the Irish and Anglo-Saxons may have got their interlacing 
 patterns in the way of patterned stuffs and ivories imported from Constantinople. 
 It is hardly necessary, however, to fetch in the Byzantines ; for interlacings are 
 particularly common in the Roman mosaic pavements which existed in every 
 province of the Roman Empire, and undoubtedly furnished patterns largely for 
 Early Christian art throughout Europe. If we take a broader survey, we shall 
 find interlacings in many a savage tribe which never heard either of Byzantium or 
 of Rome. It is one of the oldest and most widely-spread patterns in decorative 
 art. Its motif is evidently the plaited basket. Wherever basket work was in 
 
 * Mr Sharpe in Transitional Ornament illustrates the incurved cone from Ely Infirmary, 
 Llanthony, Worcester west nave, triforium of Romsey nave, Bibury. 
 + See on the triple shafting of the same school, page 245. 
 X Numerous examples are illustrated by Cattaneo. 
 
415 
 
 " ■ virjjv-^-jurgsSi-fffl 
 
 
 1,2. Westminster. 4. Winchester Crypt. 
 
 3. Shobdon. 5. Oxford Cathedral. 
 
 6. Northampton St Peter. 
 
416 FIGURE CAPITALS. 
 
 use, and some peoples have been so skilled in the art of plaiting that even their 
 drinking vessels are made of basket work, pretty patterns consonant with the 
 material have arisen, and have been transferred from the plaited vessel to 
 decorative art generally. 
 
 The capital from Oxford, with interlacing pipings, is one of those to which 
 an impossible Anglo-Saxon origin has been attributed; it is c. 1170: a very 
 similar one may be seen in the southern clerestory of the nave of the Abbaye- 
 aux-Hommes.* In the twelfth-century Gothic capitals of the West of England 
 capitals of intertwining leafage were in great favour ; e.g. in Wells nave. 
 
 Figure Capitals. 
 
 In Continental Romanesque it is not uncommon to find capitals composed 
 of writhing and intertwining monsters ; a fine example is seen in S. Zeno, 
 Verona. In our Norman work, e.g. at Shobdon,f masks and grotesques occur; 
 and occasionally, even in Gothic, a capital may be found composed, inartistically 
 enough, of heads or grotesques; e.g. Hampton Poyle^ and Oakham. In ST 
 MARY MAGDALENE, TAUNTON (417.4) all the capitals are composed of angels 
 with outspread wings. In France what are called " storied capitals " are by no 
 means rare ; a fine set of the twelfth century is seen in St Denis, Amboise ; and 
 others in the Cluny Museum, illustrating the fall of Adam and Eve, and other 
 scriptural subjects. Here, however, except at Wells, we were not very successful 
 in figure-sculpture ; in the north porch of Wells is a complete set of " storied " 
 capitals, now unfortunately much decayed. 
 
 Semi-Naturalistic Capitals of the Transitional Period. 
 
 Of these the most important is the so-called WATER-LEAF {feuille d'eau) ; it 
 is a broad, thick, flabby leaf terminating in a very small volute ; and this volute 
 curls inward instead of outward. It is well seen at WALSOKEN (417.2), near 
 Wisbech. It was astonishingly fashionable throughout Europe in the middle 
 and latter part of the twelfth century ; in England its vogue was limited to a 
 quarter of a century; from c. 1165 to c. 11 90; therefore, like the flattened 
 lower roll of the base, 451, it is a valuable criterion of chronology. § It was 
 
 * Another at Kirkton, c. 1165, illustrated in Sharpe's Transitional Ornament. 
 
 t Shobdon Church, near Leominster, was founded 1 141 to 1 150 by Oliver de Merlemond (see 
 Norman-French charter in Monasticon Anglicanum, vi. 345), on his return from a pilgrimage to 
 St James of Compostella in Spain. It is recorded that on his way back he stayed at the 
 monastery of St Victor in Paris ; and when his church was finished, he sent for two monks from 
 that monastery to serve it. In going and returning on the pilgrimage he probably passed 
 through Souillac ; for the same intertwining monsters and human beings are found there on 
 capitals and shafts as at Shobdon and Kilpeck, near Hereford. Such work is so exceptional in 
 England that it may be that he brought or sent for carvers from the west of France for his 
 church ; and perhaps the same men executed the work at Kilpeck. Casts of both may be seen 
 in the South Kensington Museum and the Crystal Palace. Shobdon Church is pulled down ; 
 the doorway is re-erected in the park (Rickman, 98). 
 
 % Glossary, i., Plate 51. 
 
 S E. Sharpe in Lincoln Excursion, 113; he calls it the "Transitional volute." It is 
 fully illustrated in his Ornamentation of the Transitional Period. 
 
4i; 
 
 1. Oxford Cathedral Choir. 
 
 2. Walsoken Nave. 
 
 3. Le I'uy Cloister. 
 
 4. Taunton Nave. 5. Dore Choir. 
 
 6. Steyning Nave. 
 
 7. Canterbury Crypt. 
 
 2 I) 
 
4 i8 SEMI-NATURALISTIC CAPITALS. 
 
 especially a favourite with the Cistercians ; e.g. the pier-arcades of Furness 
 and By-land. By the principles of their Order, they were debarred from the 
 richly carved capitals of the later Romanesque : and at first confined themselves 
 almost wholly to scalloped and coniferous capitals, as at Kirkstall, Fountains, 
 and Buildwas, or to the water-leaf: perhaps justifying themselves in the em- 
 ployment of the latter by the fact that it was " good construction," as indeed 
 it was. For the cubical capital of Norman work, with its convex curve, was 
 replaced in the Transitional period by one with a hollow, concave curve ; 
 and as the abacus still remained square in plan, its four corners overlapped 
 the circular bell below, and were dangerously unsupported. The water-leaf 
 form gave support just at the point where it was needed, and that with the 
 minimum waste of stone. A grand example of the water-leaf supports the 
 refectory pulpit of Rievaulx. The same reason that adapts the water-leaf 
 for use as a capital makes it equally serviceable as a " griffe " or spur ; e.g. 
 at BARNACK (695.9) : so a ^ so at Laon, Troyes, and Paris cathedrals.* 
 
 Its origin is obscure. It may be but the lower band of leafage, seen in 
 the capitals of the crypts of the Abbaye-aux-Dames and CANTERBURY (417.7), 
 with the addition of the Ionic volute, incurved for the constructional reasons 
 given above ; themselves reminiscences of such Corinthian capitals as that of the 
 Choragic monument of Lysicrates at Athens. 
 
 The only other leaf form in common use during the Transitional period 
 was the plantain. Viollet-le-Duc, in his illustrations, v. 490, shows how closely 
 the curves of young leaves of plantain are rendered in capitals at Notre Dame, 
 Paris, and Montreale, Yonne. They occur, however, even in Norman work ; 
 e.g. at EASINGTON, DURHAM (428) ; and the analog}' in form and position 
 between these and the leafage of the upper band of the Tower of the Winds 
 at Athens is so striking, that in spite of chronology and geography one can 
 hardly doubt that this is a case of survival of ornamental detail. Nothing 
 seems to have more persistent vitality than a decorative form. 
 
 The fern leaf, uncurling just as it protrudes from the ground in spring, 
 now and then occurs ; e.g. in York crypt ; choir aisle of Abbey Dore ; OXFORD 
 CATHEDRAL (415.5); Canterbury choir; all Transitional examples.f On the 
 remarkable capital from STEYNING (417.6), on the right, may be seen three 
 fern leaves, separated by incurved cones. Occasionally the leaf of the common 
 aru/n (Arum viaculatnvi) with berries occurs ; as in the west doorway of 
 Ledbury,} and in the spandrels of Wells triforium. In the Ledbury example 
 the berries are surrounded by a spathe as in nature, and overhung by a kind of 
 foliated hood. 
 
 The most beautiful examples of Transitional foliage occur in the superb 
 capitals of the eastern work of CANTERBURY CHOIR § (428) ; unsurpassed 
 in this or any other country ; the only parallels we have are those at Broad- 
 
 * Illustrated in Viollet-le-Duc, Architecture, vi. 51. 
 
 t A fine fern cap from Soissons Cathedral is illustrated in Parker's Glossary, Plate 48. One 
 occurs in the cloister of Le Puy Cathedral, probably eleventh century-. See also illustration in 
 Viollet-le-Duc, Architecture, vol. v. 486. 
 
 I Illustrated in Colling's Mediceval Foliage, Plates 7, 24. 
 
 § On the leafage of Canterbury choir, see Colling's Mediceval Foliage, pp. 32, 33, 34. 
 
SEMI-NATURALISTIC CAPITALS. 
 
 419 
 
 water Church* and Oakham Castle, and one or two in Oxford Cathedral 
 nave. Great was the promise of the Transitional capital, c. 11S0; leaf after 
 leaf, blossom and fruit were being experimented with ; we bade fair to have 
 foliated capitals varied in design as nature herself. Hut all came suddenly to 
 an end. All were abandoned, in the end, for a non-naturalistic capital ; the 
 famous Early Gothic capital, which was elaborated into such forms of utmost 
 beauty as we cannot but admire at Wells, Lincoln, Salisbury; but a poor 
 exchange for the infinite diversity of vegetative form that grew beneath the 
 chisels of the craftsmen of Broadwater and CANTERBURY (423.7)/)- 
 
 * Capitals from Broadwater are illustrated in the Building News, Oct. 1, 1869. They 
 differ from the Oxford and Canterbury capitals in admitting animal forms into the foliage ; 
 animal forms are very plentiful also in the eastern bays of Wells nave. 
 
 t Equally grand are the Transitional capitals of the cathedrals of Paris, Lisieux, Senlis, and 
 Yezelay choir, illustrated in Yiollet-le- Due's articles on Chapiteau and Sculpture ; and of 
 St Laumer, Blois, illustrated in Building News, March 27, 1S68. 
 
Chapter XXVIII. 
 
 FOLIATED CAPITALS. 
 
 Roman Corinthian and Composite Capitals, and Norman Versions of them — Welding of 
 the Volute and the Scroll Capital in the Early Gothic Capital of Stalked Conven- 
 tional Foliage — Naturalistic Capitals — Undulatory Foliage. 
 
 Of the three capitals in use in ancient Greece, the Doric was seldom employed 
 in the early Christian churches at Rome, and the Ionic far more rarely than 
 the Corinthian or its variant, the Composite. These two, the Corinthian and 
 the Composite, were produced by thousands in every province of the Roman 
 Empire, even in remote Britain. In some provinces Roman work survived in 
 great abundance, and in a high state of preservation, e.g. in Provence and 
 Languedoc, and in Burgundy ; and in these districts, naturally, Corinthian 
 capitals were rendered with great fidelity ; and indeed sometimes, in the schools 
 of Provence, Toulouse, and Burgundy, quite equalled or even surpassed their 
 prototypes. In Normandy and Britain it was otherwise ; Roman work had never 
 been so abundant as in the highly civilised provinces of Southern Gaul ; and 
 the destruction of what there was of it had been far more complete. Therefore 
 the Corinthian capital, as it appears in Norman and English hands in the 
 eleventh and twelfth centuries, is necessarily but archaic and rude. Nevertheless 
 this capital, which we may call the Corinthianesque, is of very considerable 
 interest not only for its high and ancient lineage, but from the predominating 
 influence which it will be found to have exerted over the development of carved 
 foliage, which is to a large extent but the history of the decomposition and 
 recomposition of the ancient Corinthian capital. 
 
 But, first of all, it must be borne in mind that the Corinthian capital of 
 Greece was not one, but diverse. That of the TEMPLE OF ZEUS (425) at 
 Athens may be regarded as the normal type ; nevertheless such capitals as 
 those of the Choragic monument of Lysicrates and the TEMPLE OF THE WINDS 
 (425) at Athens have also to be taken into account. But our mediaeval builders 
 knew nothing of Greek capitals ; only Roman work ; and the Romans had not 
 hesitated to bend to their purposes the Greek forms. In the first century of our 
 era, on the whole they adhered to the type of the Temple of Zeus ; e.g. in the 
 portico of the Pantheon ; the chief exception being that since the Corinthian 
 volutes looked, and were, too small and weak to support the angles of the 
 abacus, they often enlarged them ; changing the Corinthian into what is called 
 the Composite capital : one in which the upper half is Ionic, the lower half 
 Corinthian.* Of these Composite capitals those in the BATHS OF DIOCLETIAN 
 
 * On early types of Composite capitals in Egypt and Phrygia see Anderson and Spiers, 
 98 and 154 
 
42 1 
 
 1. Newport St Woolos 
 
 2. St Nicholas, Caen. 
 
 3. Dore Choir. 
 
 4. Youlgreave. 
 
 5. Newport St Woolos. 
 
 6. < Ixford Cathedral. 
 
 7. Oxford Cathedral. 
 
 8. New Shoreham. 
 
42: 
 
 io 
 
 i. West Walton. 
 
 2. Lincoln Choir. 
 
 3. West Walton. 
 
 4. Llanidloes. 
 
 5. Llanidloes. 
 
 6 Llanidloes. 
 
 7. Llanidloes. 
 
 8. Dore Retrochoir. 
 
 9. Dore Retrochoir. 
 10. Dore Retrochoir. 
 
i. New Shoreham. 5- Yarmouth. 
 
 2. New Shoreham. 6. Lincoln, St Mary-le-Wigford. 
 
 3. Whaplode. 7- Canterbury, Saint's Chapel. 
 
 4. Tilney All Saints'. 8. Oxford Cathedral Transept. 
 
 9. Lincoln Choir. 
 
424 
 
 1. Wells Choir. 
 
 2. Wells Transept. 
 
 3. Wells Nave. 
 
 4. St Margaret's, Lynn. 
 
 5. Llandaff Nave. 
 
 6. Llandaff Nave. 
 
 7. St Mary's, Shrewsbury, Nave. 
 
 8. Lichfield Nave. 
 
CORINTHIAN CAPITALS. 
 
 425 
 
 below may be regarded as typical ; they occur in great numbers throughout 
 Italy and the south of France (Anderson and Spiers, 150). 
 
 Now if we decompose these Greek and Roman Corinthian capitals, we find 
 the following elements. (1.) Supporting the abacus are the geometrical spirals 
 or volutes. These may be greatly enlarged ; and they may be covered with 
 
 :mmM. 
 
 ■tv* l{ '-v %vr '< f "'~ t£3- 
 
 wWmMWMM 
 
 3CM.E OF i_ 
 
 Temple of the Winds, Athens. 
 
 Choragic Monument of Lysicrates, Athens. 
 
 Temple of Zeus, Athens. 
 Baths of Diocletian, Rome. 
 
 foliated ornament. (2.) Midway between the volutes, on the same or on a higher 
 level, may be a rose. This is reproduced exactly in a granite capital, probably 
 Roman, in the eleventh century cloister of LE PUY CATHEDRAL (417.3)- (3-) More 
 often, instead of a rose, there is symmetrical leafage, as in the Choragic monu- 
 ment. This is sometimes called the anthemion or honeysuckle; but it is just as 
 much like palmdeaves, and may be called a palmette. (4.) The volutes issue 
 
426 
 
 CORINTHIAN CAPITALS. 
 
 from scrolls. These scrolls may be spirals, repeating on a smaller scale the 
 volutes above, as in the Choragic monument ; or scrolls of foliage, as in the 
 Temple of Zeus. These leaf-scrolls are much emphasised in many Norman 
 capitals, and are of great importance as one of the origins of the Gothic foliated 
 capital. (5.) Then comes one band or more of acanthus. (6.) Below the acanthus, 
 in the Choragic monument, is the so-called water-leaf. (7.) In the Temple of 
 the Winds, an exceptionalexample, there are no volutes : the upper part of the 
 capital consists entirely of a row of simple pointed leaves. These remind one 
 strongly of the lotus caps of Egypt, and the leaf may be that of the lotus. It 
 is often assumed to be that of the plantain. 
 
 The most striking part of the capital is the band of acanthus. As to this, 
 it is by no means certain that it is acanthus at all ; in Greek capitals, where it 
 has sharply pointed leaves, it has been assumed to be the Acantlms spinosus ; 
 but it is just as much like the sea-holly, Eryngium maritimum ; and still more 
 like the leaves of the common artichoke. But it is so much conventionalised 
 that it does not resemble any known leaf exactly. In Roman examples, e.g. at 
 Tivoli and Preneste, up to about the end of the first century of our era, the 
 acanthus was carved, as in the Choragic monument, with sharply pointed lobes 
 and angular notches between the lobes, and with " eyes " between each group of 
 
 „ lobes. And it was this Greek form 
 IwivMn £\iI'/JL °f the acanthus which was employed 
 
 in Byzantine work ; e.g. in St Sophia, 
 Constantinople; in the Syrian 
 churches ; in the Golden Gate at 
 Jerusalem ; in St Mark's, Venice. 
 It has been assumed, therefore,* that 
 wherever the pointed form of acan- 
 thus occurs, there we must admit the influence of the patterns of Byzantine 
 stuffs and ivories. Sir G. Scott recognises this form at St Front, Perigueux, 
 c 1 130; St Denis, 11 40; St Germain-les-Pres, Paris; in England only in the 
 work of William of Sens at Canterbury, and the north-west doorway of the 
 west front of Lincoln. In both these, however, the acanthus is decidedly of 
 the blunted type ; and no evidence of Byzantine influence in England. 
 
 In Roman work the acanthus is supposed to be the Acanthtcs mollis ; but 
 there is really very little likeness between the two. In some instances, as in 
 the arch of SEPTIMIUS SEVERUS (426.3), with which compare the Canterbury 
 capital (428), it is very much like the oak leaf. When formed as at the 
 PANTHEON (426.2), it is asserted by Sir William Chambers and Mr R. N. 
 Wornum to be simply a bunch of olive or laurel leaves. There seems to be 
 only one common characteristic of all the varieties of the Roman acanthus ; viz. 
 that the tip of the lobe is never acutely pointed, as it is in Greek work.f 
 
 After the second century the Roman carvers, weary of monotonous correct- 
 ness, altered the Corinthian and Composite capitals in the most fantastic fashion. 
 The volutes become crouching dragons; or eagles take the place of rose or 
 
 * See Sir G. Scott in Archceological Journal, 32, 360 ; and the numerous illustrations in 
 his Lectures, i. S5. 
 
 t On the varieties of acanthus employed, see Anderson and Spiers, 149. 
 
 Acanthus. 
 
VOLUTE CAPITALS. 
 
 427 
 
 palmette ; or the whole capital, if for some temple of Bellona or Mars Ultor, 
 may be composed of greaves, lances, cuirasses, and helmets ; as in the capitals 
 re-used in S. Lorenzo, Rome. And it was this, the late and debased Corinthian- 
 esque capital, which was being produced in every province of the Roman Empire, 
 when the first Christian churches were built ; and which was most imitated in 
 Northern Europe, the last provinces to be conquered ; where Corinthian capitals 
 of the best period were few or non-existent. That being so, we need not be 
 surprised at the vagaries of the Corinthianesque capital in Normandy and 
 England. 
 
 It occurs in the midst of undoubted Anglo-Saxon work ; e.g. at Scartho, 
 Great Hale, and Branston ; in the belfry windows of St Peter-at-Gowts and St 
 Mary-le-Wigford, Lincoln ; in Sompting tower, where unfortunately the artist 
 has got the volutes on one capital, and the 
 acanthus on the other. Where such work 
 occurs, it seems to be late ; due to contact 
 with and direct imitation of Norman 
 Romanesque.* 
 
 In the eleventh century in Normandy 
 there is no attempt to render the intricacies 
 of the acanthus band, with the exception 
 of the capitals of Bernay, Bayeux, and 
 Rucqueville, which are so exceptional that 
 they must be by the hand of craftsmen 
 from outside Normandy. + In the normal 
 capital the volutes are present ; the bell is 
 convex ; and between the volutes is usually 
 a sort of "console"; e.g. in the arches of 
 the central tower of st Nicholas, caen 
 (421.2J, c. 1083. In St John's Chapel in 
 the Tower of London, the console takes 
 the form of a tau. In the twelfth 
 century the console hardly ever appears. 
 Usually the bottom part of the capital 
 is bare ; but sometimes, as in the crypts 
 
 of the Abbaye-aux-Dames and CANTERBURY (193). an(1 m the doorway of 
 ST WOOLOS (421. 1), Newport, Monmouth, there is a band of leafage. This may 
 represent the lower leaves seen in the Choragic monument of Lysicrates ; but, 
 more probably, it is acanthus leafage cut as well as the carver knew how ; 
 not serrated, and usually without the veining. In Rome itself, in an upper story 
 of the COLOSSEUM '430), the very same type of foliage occurs; and here also, 
 it is probably acanthus, roughly blocked out because it was to be seen only from 
 a distance. Such rude caps are common in our early crypts ; there are especially 
 fine ones at Lastingham. % In those portions of our Norman churches which 
 were built in the eleventh century these Corinthianesque capitals are common; 
 
 ' See, however, Baldwin Brown's Arts in Early England, 161, 1S0. 
 
 t Ruprich-Robert, pp. 245,65, 116. 
 
 t Illustrated in Britton's Arch. An/., vol. v., Plate 6. 
 
 Barton-le-Street. 
 
42t 
 
 VOLUTE CAPITALS. 
 
 e.g., in the eastern limbs of Nor- 
 wich and Chichester and the tran- 
 sept of Ely. It was not till the 
 twelfth century that the cubical 
 capital predominated.* The two 
 photographs from BARTON - LE- 
 STREET (427) and EASINGTON 
 illustrate very well the fantasies of 
 Norman work, as seen in many a 
 remote village church. The latter, 
 with its laurel or plantainf leafage, 
 appears to be a direct descendant 
 of the capital of the Tower of the 
 Winds.} In Barton-le-Street both 
 capitals have interlacing ornament ; 
 one contains a mask, the other the 
 favourite beak-head. In the capital 
 on the left the anthemion appears below, and a symmetrical leaf-scroll above.§ 
 
 After the middle of the twelfth century such rude work becomes rare ; 
 but classical remini- 
 scences still abound ; 
 e.g. the capital from 
 ABBEY DOREi] (421.3) 
 has a small anthemion 
 below, and a large 
 symmetrical leaf- 
 scroll above. The 
 capital from OXFORD 
 CATHEDRAL CHOIR 
 (421.6), I I 54-I 180, 
 also has the an- 
 themion ; the large 
 volutes at the angle 
 are foliated ; and 
 what is very impor- 
 tant, the rest of the 
 bell is ringed round 
 with a row of smaller volutes. The next step is seen in the capital from the 
 NAVE (421.7); where the discovery has been made that a capital can be 
 
 Canterbury, Capital of Saint's Chapel. 
 
 * It is not intended to assert that the cubical capital was not largely employed in the 
 eleventh century also ; it is the predominant cap in the eleventh-century work of Winchester 
 and Durham. 
 
 t Capitals of plantain or laurel leaf occur at Battle, Tillington, Hurstmonceaux, Irthington, 
 Fountains, and Abbey Dore. 
 
 + Illustrated in Sharpe's Tra/isitionat Ornament. 
 
 S Cf. Pugin's Gothic Ornaments, Plate 76. 
 
 || Cf. Northorpe and Laughton in Sharpe's Transitional Ornament. 
 
VOLUTE CAPITALS. 429 
 
 composed with volutes only* These foliated volutes go by the name of 
 crockets. All sorts of experiments with them ma)- be watched in NEW 
 SHOREHAM choir (421.8). The cap of the cylinder is very artless; below 
 is a symmetrical classical ornament ; above is a band of crockets, all turned 
 sideways, as if whirled round by a current of air. This " wind-blown " capital 
 was a great favourite in the latter days of the twelfth century ; e.g. in St 
 Leonard's, Stamford. The capital of the OCTAGONAL PIER (423.1) has quite a 
 Gothic appearance ; nevertheless there are two small examples of the anthe- 
 mion below; and above each arris of the pier is an unmistakable foliated volute 
 or crocket. The capitals of the eastern respond are CROCKETS (423. 2) and 
 nothing else. (The classical leaf-scroll in the arches should be noted.) From 
 this last capital it is not a long step to those of the retrochoir of the neigh- 
 bouring cathedral of CHICHESTER (245), c. 1186: or from these to the crocket 
 capitals of GREAT YARMOUTH (423.5) and ST MARY-LE-WTGFORD, LINCOLN 
 (423.6). Thus then there is a lineal descent to be made out from the geometrical 
 volutes of the Corinthian capital f of ancient Rome to the famous thirteenth- 
 century capital of Early English and French Gothic.^ An example from CAN- 
 TERBURY (428) is illustrated ; it is a magnificent specimen of the French 
 Transitional Composite capital ; but no direct imitations of these are to be 
 found ; except only at Oakham, § and perhaps Broadwater. 
 
 So far we have spoken of the foliage of the Early Gothic capital as derived 
 from the foliated volute of the crocket. But there is a good deal of foliage into 
 which the form of the volute does not enter at all ; e.g. the lower band of the 
 capital over the octagonal pier of NEW SHOREHAM (423.1); those from CWM 
 iiir, now at Llanidloes (422.4,5); three from ABBEY DORE (422); two from 
 LINCOLN (422.2 and 423.9) ; one from WHAPLODE west NAYE (423.3). These 
 contain leafage with three or five lobes ; in early work usually of three only. 
 This capital is held to be a conventionalisation of the Herba benedicta or 
 clover ; || but the resemblance is not very marked ; and if it were, it is in the 
 highest degree unlikely that all Western Europe should hit upon such a motif 
 almost simultaneously. Whatever the motif was, it must have been some- 
 thing that was in the hands of everybody already. This we hold was the 
 leaf-scroll of the ancient Corinthian capital, e.g. that of the TEMPLE OF 
 ZEUS (425), which was still in full use in the last years of the twelfth cen- 
 tury ; e.g. all round the late Norman trefoiled doorway in ELY CLOISTER (430) 
 
 * Compare the Barnack capital (695.9), a "d those at Polebrook (Ncnc Valley, 12). All 
 Saints', Stamford, and Lichfield nave have good specimens. 
 
 t So also Enlart's Manuel, 558 : " Le crochet derive directement des volutes du chapiteau 
 corinthien." 
 
 t It should be added that though it is convenient to designate this the crocket capital, it is 
 erroneous to suppose that it is derived from the Gothic crocket proper. For the latter hardly 
 came into existence till c. 1 192 in the piers of Lincoln choir transept; while we have seen 
 crocket capitals occur earlier than this, and more highly developed, at Oxford and New 
 Shoreham. There are also several large capitals of this type in the Trinity Chapel, Canterbury. 
 In France the crocket capital received excessive development at times ; e.g. in Le Mans choir 
 and the Collegiate Church of Semur-en-Auxois. 
 
 s' Colling's Mediuval Foliage, Plate 19. 
 
 I! Others, with VioIlet-le-Duc, regard it as an abstraction of leaf-form in general. But that 
 seems a somewhat metaphysical conception to exist in the brain of a stone-mason. 
 
430 
 
 LEAF-SCROLL CAPITALS. 
 
 Ely Cloister. 
 
 are scrolls of trilobed leafage. Still more instructive is the work in the arches 
 of NEW SHOREHAM CHOIR (273) ; where the classical and Gothic scrolls may 
 be seen side by side. These very scrolls reappear at TILNEY ALL SAINTS' 
 
 (423.4), c. 1 180. In 
 the upper panel * 
 at NEW SHORE- 
 HA M the bottom 
 scroll is, if anything, 
 more classical than 
 •^f^f^^^j that of the Arch of 
 
 Septimius Severus 
 (426.3); while above 
 is a seven-lobed 
 leaf, which would be 
 at home in any 
 thirteenth-century capital. In the lower 
 panel is a compound leaf composed of two 
 trefoils ; and even marked by the bold pro- 
 jecting midrib which is so characteristic of 
 our Early Gothic foliage. With the New 
 Shoreham scrolls should be compared the 
 Early Gothic scrolls (431). On the whole 
 we may conclude that the Early Gothic 
 leaf capital is the result arrived at by a 
 long series of modifications of the combination of a simplified form of the 
 Roman acanthus scroll, with a foliated form of the Roman Composite volute. 
 
 In some cases the leaf-scroll, in others 
 the volute predominates ; e.g. in OXFORD 
 TRANSEPT (423.8), as in St Leonard's and 
 All Saints', Stamford, there are two bands 
 of small leaf-scrolls, and the volute type 
 is hardly perceptible at the corners. In the 
 two small capitals from the arcading of 
 St Hugh's work at LINCOLN (422.2, 423.9) 
 there is no trace at all of the volute ; 
 while in the capitals from the chancel arch 
 of WEST WALTON (432) and the chapter 
 house of Salisbury, and that from the arch 
 leading into the Lincoln choir aisle, c. 1240, 
 the volute is predominant. In Elygalilee, 
 the two forms occur side by side, as they 
 do in Auxerre choir ; the volutes on the 
 right, the scrolls on the left capital. 
 
 All these capitals are usually desig- 
 nated " stiff-leaved capitals." And in the early examples they are stiff indeed ; 
 e.g. at NEW SHOREHAM, Whaplode, and Cwm Hir. But nothing can surpass the 
 * For the classical scroll see Brandon's Analysis, Semi-Norman, 2. 
 
 New Shoreham. 
 
 Canterbury Choir. 
 Rome, Colosseum. 
 
LEAF-SCROLL CAPITALS. 
 
 43' 
 
 free flow of curve and abandon of the later capitals ; e.g. those of the ( Ihapel of the 
 Nine Altars, Durham. In all the special characteristics are (i) that the foliage 
 is conventional ; (2) that the stalks are always present. The object was that 
 the foliage should seem to grow out of the shaft as naturally as the branches 
 grow out of a tree trunk. 
 
 As to the treatment of the stalks, considerable differences prevail. Some- 
 times the stalk has a broad base ; and may indeed be the plantain leaf again, 
 expanding above into a volute. This is well seen in volute caps in the Ely 
 galilee ; so also in the tower arch of St Mary's, Stamford.* Somewhat similar 
 are the sturdy stalks of the Durham capitals. There can be no doubt that 
 
 I. Salisbury. 
 
 Stone. 
 
 4. Wells. 
 
 massiveness of stalk greatly adds, both in reality and to the eye, to the 
 adequacy of the volutes to perform their functions as corbels to the abacus. 
 It is far more common in French than in English Gothic. More often, our 
 stalks are somewhat flat and characterless ; and in early work, especially in 
 west country Gothic, as at Cwm Hir, Abbey Dore, Llandaff, they are often 
 mere slender pipings or ribs, barely perceptible on the bell. 
 
 The extent of projection given to the volutes is a mark of the amount of 
 skill reached by the carvers ; foliage of great projection does not occur in early 
 work ; though the reverse is not necessarily true ; e.g. compare the capitals of 
 * Illustrated in Nene Valley, Plate 73. 
 
43^ 
 
 EARLY GOTHIC CAPITALS. 
 
 Wells choir with those of the transept and nave which were built later. So 
 again the little capitals (1190) from Lincoln cling lovingly to the bell; great 
 is the advance in such a capital as that of WEST WALTON (422.1). From their 
 small projection the capitals of Cwm Hir and Abbey Dore may well have been 
 executed before the twelfth century was complete. 
 
 The direction which the foliage takes should be noted. In the twelfth- 
 century capitals and a little later it usually tends to be vertical ; e.g. the Lincoln 
 arcading ; Cwm Hir ; Abbey Dore ; Wells ; Llandaff. But the early foliage 
 sometimes swirls round ; e.g. at NEW SHOREHAM (421.8) ; ABBEY DORE (422.8) ; 
 St Leonard's, Stamford ; Jedburgh ; Moulton ; and the porch of St Mary Red- 
 
 cliffe, Bristol ; * less often in 
 later work, e.g. Southwell 
 choir, Lichfield nave. In 
 Lincoln presbytery it runs 
 round both ways. In the 
 end the favourite design was 
 for the foliage to rise up to 
 support the abacus, and then 
 to fall over in heavy clusters, 
 as at West Walton and the 
 doorway of Lincoln choir 
 aisle. 
 
 So again, the character 
 of the scrolls employed is 
 significant. It is only in 
 early work that such small 
 stiff trefoils, quatrefoils, and 
 cinquefoils are employed as 
 those of Cwm Hir ; Holy- 
 rood ; Lincoln arcading ; 
 Whaplode ; Haverfordwest 
 doorway ; Abbey Dore ; the 
 corbels of Ryland, Rievaulx, 
 and WHITBY (433) choirs ; 
 the vaulting shafts in the 
 south aisle of New Shore- 
 ham ; south doorway of 
 Valle Crucis nave ; Oxford 
 Cathedral ; f Lichfield choir ; Deerhurst ; Broadwater, j 
 
 Again, in early work, if the stalks are slender, they are often allowed 
 to intersect ; as in the New Shoreham vaulting shafts ; the corbels of WHITBY 
 CHOIR (433) ; § Valle Crucis, south doorway; Jervaulx chapter house.; So 
 also, where the early carvers have enough skill, the foliage itself twists and inter- 
 laces ; e.g. at St John's, Chester ; Whitchurch ; Shrewsbury St Mary's ; New 
 
 * Illustrated in Sharpe's Transitional Ornament. + Illustrated in Prior, Gothic Art, 14S, 149. 
 % Illustrated in Colling's Mediceval Foliage, 20. S Illustrated in Sharpe's Parallels, Plate 118. 
 || Illustrated in Prior, Gothic Art, 110. 
 
 West Walton. 
 
WEST OF ENGLAND CAPITALS. 
 
 433 
 
 Shoreham ; Chichester; Wycombe, Hospital of St John; Byland.* Especially 
 wonderful are the capitals in the west nave of WELLS (412.6); never were 
 capitals tangled into such exquisite grace. 
 
 The West of England Capitals deserve separate notice. Till 
 recently the importance of the early and remarkable school of the west had not 
 been recognised.+ The earliest capitals probably are those in the south nave of 
 St Mary's, Shrewsbury, and the western bays of Worcester nave, which may be 
 as earl}- as 1 160-1 170 : in both of these they are still mixed up with Romanesque 
 motifs. Then come a most remarkable group in the very middle of Wales, 
 in the CHURCH OF LLANII1L0ES ('552). Here wall-pieces of the roof rest 
 on broken bits of vaulting-shaft ; capitals, piers, and arches have been largely 
 transposed ; the piers are designed to carry an aisle-vault, which is not there ; 
 it is plain that the whole arcade was not built on the spot, but was trans- 
 ferred from some other building. This can hardly be other than Cwm Hir, some 
 20 miles distant. It was probably the largest Cistercian monaster}' in England or 
 Wales, with a nave 250 feet 
 long ; even surpassing Byland. 
 Capitals found in situ at Cwm 
 Hir are precisely the same in 
 type as those at Llanidloes. 
 The abbey was dissolved in 
 1536; and the Llanidloes roof, 
 as the shields at the ends of 
 the beams record, was put up in 
 1 542. Probably, the Llanidloes 
 people, wishing to add a north 
 aisle to their nave, transferred 
 one bodily, except the vault, 
 from Cwm Hir. \ 
 
 The work at ABBEY DORE 
 (151.4) is of two dates: the 
 original short choir, and its sub- 
 sequent eastern extension; 412.8, 417.5, 421.3 are from the choir; the rest are 
 from the ambulatorv. The earlv capitals in Wells Cathedral are nearly all of the 
 
 Whitby. 
 
 wlu 
 
 volute type. They fall into four groups: first, those of the choir, c. n; 
 tall volutes are mingled with short knobby ones (424.1); secondly, those 
 of the transept, where human figures are introduced, admirably carved (424.2); 
 thirdly, those in the east nave, where animal forms abound ; fourthly, those in 
 the west nave (412.6), where animal forms are rare; some of these last are 
 early in type, e.g. 424.3 ; others are of most intricate design, and difficult 
 
 * Prior, 142, 146. 
 
 + See account of Wells Cathedral in English Cathedrals Illustrated. It is one of the 
 special features of Mr Prior's History of Gothic Art in England that he has insisted on the 
 recognition of this as a distinct school of Early English (iothic. 
 
 J S<& Montgomeryshire Collations, xxiv. 395-417- Giraldus Cambrensis says that Cum Hir 
 was founded by " Robert, son of Stephen." This was the Robert Fitzstephen who commanded 
 the first English invasion of Ireland in 1 170 : he was uncle to Henry II. 
 
 2 E 
 
434 
 
 NATURALISTIC CAPITALS. 
 
 execution.* Both in Wells and Llandaff Cathedrals the capitals are unusually 
 tall and slender. Another common characteristic, not a pleasant one, is that 
 the necking or astragal is sometimes omitted, as in Glastonbury choir, Here- 
 ford east transept, ST DAVID'S (412.5). Another characteristic of the LLANDAFF 
 CAPITALS (424.5, 6) is the great height of the stalks and the smallness of the 
 space allotted to the foliage ; and the fact that the stalks are so diminutive. 
 To the Llandaff type belong the beautiful capitals of the north arcade of ST 
 MARY'S, SHREWSBURY (424.7). 
 
 Naturalistic Capitals. — The reign of conventional foliage was long. 
 Commencing about 11 70 in such examples as New Shoreham and Wells, its 
 vogue lasted till about i28o.f By the middle of that time it had on the 
 
 Southwell Chapter House. 
 
 whole, with such exceptions as the capitals of Durham eastern transept, settled 
 into a fairly uniform type. And with uniformity came monotony. After c. 
 1220 it was superseded to a considerable extent, e.g. at Salisbury, Beverley, 
 Westminster, and the Temple choir, London, by the molded capital. 
 
 As we have seen above, experiments had been made with capitals of more 
 or less naturalistic type in the last half of the twelfth century ; e.g. the campanula 
 and ground ivy are carved in Canterbury choir. So again % in one of the western 
 
 * Capitals from Whitchurch, Dorset, are illustrated by Mr Prior. The interior of Whit- 
 church is drawn in the Spring Gardens Sketch Book, vi. 65. 
 
 + Naturalistic foliage occurs in abundance in the diapered spandrels of the arcade round 
 the chapter house of Westminster ; illustrated vajohn of Gaunt Sketch Book, i. 72. 
 
 % Colling's Mediccval Foliage, Plate 15, p. 33. 
 
NATURALISTIC CAPITA I. S. 
 
 435 
 
 capitals from CWM ink (422.7), one knot of foliage may be seen expanding 
 into a lily bloom. A thoroughly naturalistic capital occurs in the Lincoln 
 chapter house, c. 1225, which it is difficult to reconcile with the thoroughly 
 conventional foliage round it.* In the Angel triforium of the presbytery the 
 two forms may be seen side by side again and again ; here conventional, there 
 real leafage; c. 1260. But when the east side of the cloister was built, r. 1296, 
 the capitals become reproductions of natural foliage. EXETER CATHEDRAL 
 1241), commenced c. 1280, is a treasure-house of beautiful naturalistic foliage, 
 especially the corbels of the vaulting-shafts, and the bosses of the vaults 
 (297). lichfield nave (424.8), like Lincoln presbytery and ST Margaret's, 
 LYNN (424.4), has admirable specimens both of conventional and naturalistic 
 leafage: the latter of more advanced character than at Lincoln : a cap of lilies 
 is shown. Early lilies, of charming type, are seen in the north porch of 
 
 
 Southwell Chapter House. 
 
 Bridlington and at Warmington. Later on, all kinds of leaves and berries 
 are copied. Perhaps the most beautiful work is the leafage of the shrines of St 
 Frideswide and ST THOMAS CANTELUPE ( 1 87) in the cathedrals of Oxford and 
 Hereford : the former contains maple, columbine, the greater celandine, oak, 
 sycamore, ivy, vine, fig, hawthorn, and bryony. SOUTHWELL CHAPTER HOUSE 
 434) is richest of all in this work. Here the skill of the craftsmanship is 
 wonderful. If a finger be placed at the back of the leaves, it will be found 
 that they are finished off quite smooth behind. In one instance there is an 
 animal behind the leaves, which can only be seen by looking up from 
 
 underneath. 
 
 •' In the elder days of Art 
 
 Builders wrought with greatest care 
 Each minute and unseen part : 
 For the gods see everywhere.'' 
 
 * Colling's Mediaval Foliage, p. 40. Mate 33, S. 
 
436 UNDULATORY FOLIAGE. 
 
 No two capitals or bosses or spandrels are alike. In the middle order of 
 the arch of the doorway the foliage stands quite free of the hollows, and 
 is attached only to their edges ; and the hollows themselves are as cleanly- 
 chiselled as any of exposed moldings.* Another fine set of capitals representing 
 the Seasons, c. 1310, may be seen in the pier-arcade of Carlisle choir, f 
 
 Another characteristic of these capitals is that the idea of growth out of the 
 shaft or column is no longer present. The stalks which the thirteenth- 
 century craftsmen had been so careful to insert were now usually omitted. 
 This had begun in the previous period. Instead of stalks, branches were 
 carved ; branches cut off by the knife from some neighbouring tree, as may be 
 seen in the SOUTHWELL SPANDRELS (435), and then transmuted into stone. 
 But more often it was rather tendrils than branches that were carved ; the capital 
 or the boss became a globular mass of interlacing tendrils and leafage like a 
 bunch of mistletoe. That being so, the outline of the capital was changed. The 
 thirteenth century had been careful to follow the graceful curve of an inverted 
 bell ; now the foliage became a globular mass, all beauty of outline lost : e.g. 
 the capitals of Selby choir. 
 
 UNDULATORY FOLIAGE. — After a time the fashion changed once more. 
 Such work as that at Southwell must have been tedious and costly in the extreme. 
 What was worse, every one had learnt the trick. So the tours de force of 
 undercut foliage were abandoned, except here and there, as in the capitals of 
 Selby and Wells choirs, and the wonderful work at the back of the reredos of 
 Beverley Minster. \ And a curious change set in as regarded the contour of the 
 leaves. Ogee curves and ogee arches had invaded the windows, the parapets, 
 and the moldings ; it was left to introduce them into every carved leaf; to give 
 each curve its just counter-curve. 
 
 First of all, the undulatory movement had caught the crockets and finials ; 
 then it appears in the diapers of the stone screens of Lincoln and Southwell 
 Minsters ; then the capitals caught the infection. It was as if the craftsmen 
 had grown weary of copying healthy foliage ; and picking up some leaf stung 
 by a fly, had set to work to reproduce diseased foliage for the sake of the 
 bulbous swellings that gave the admired ogee curve. 
 
 At first the difference was slight. All sorts of leaves are treated naturalisti- 
 cally in the Lincoln diaper, c. 1325; but each has the undulatory curve. Still 
 naturalistic, but more undulatory, are the misericordes of the stalls of Wells ; 
 a little later in date. Hardly naturalistic at all, but undulatory in the extreme, 
 are the corbels of St Stephen's Chapel, Westminster, 1348. The tendency 
 was henceforth to restrict the reproduction to those leaves which natural^ 
 assume an undulatory form ; e.g. the maple, the oak, but above all, the vine 
 leaf. Indeed, as time went on, carved leafage was limited almost wholly to 
 running patterns of grape and leaf; hundreds of examples of which are to 
 be seen beneath the cornices of stone monuments and wooden screens in the 
 work of the fifteenth and sixteenth centuries. 
 
 Frequently neither branch nor tendrils occur ; the capital consisting of 
 nothing but isolated leaves without any visible connection with the bell ; 
 
 * Livett's Southwell, m. t They were much restored in 1S03. 
 
 X Illustrated in Colling's Foliage, Plates 52, 53. 
 
UNDULATORY FOLIAGE. 
 
 437 
 
 gummed on, as it were. Not infrequently, as in the capitals ,,1" the lofty piers of 
 Hull nave, the foliage consists of quite small, isolated leaves studded on the 
 bell. On the whole, the tendency was towards larger leafage. Minute work 
 occurs in Chester nave; and in the beautiful rose capital' of PATRINGTON 
 below, but in the other two capitals of PATRINGTON larger leaves arc em- 
 ployed. In the capitals from WELLS RETROCHOIR both large and small 
 leafage occur side by side. Even in York nave, c. 1300, large-leaved capitals 
 
 1. Patrington Tower Pier. 
 
 2. Patrington Xave. 
 
 3. Patrington Transept. 
 
 4. Wells Retrochoir. 
 
 5. Salhouse. 
 
 6. Tiverton. 
 
 were already employed. The Berkeley Chapel in Bristol Cathedral, c. 1 330, is 
 particularly remarkable for the large size of the leafage. The capital from 
 SALHOUSE above well illustrates a common type of large-leaved capital; 
 similar capitals occur to the end of the Gothic period. In a small village church, 
 such as Salhouse, such large leaves are out of scale ; but they are very properly 
 employed in the soaring piei arcades of the greater churches of the Perpen 
 dicular period. On the whole, however, the tendency in the late parish churches 
 
438 UNDULATORY FOLIAGE. 
 
 was to prefer the molded capital ; good specimens of foliated capitals are 
 rare, except in Somerset and Devon.* The example from TIVERTON (437.6) 
 contains the passion-flower. The later examples, in most cases, show a great 
 degeneration from the wonderful work that had been wrought in the thirteenth 
 and fourteenth century. Somewhat monotonous, doubtless, fourteenth-century 
 work is (and from its uniformity and monotony valuable as a chronological 
 criterion) ; but criticism is dumb before the exquisite design and consummate 
 execution of such works as the Beverley reredos ; the Heckington seclilia ; 
 and that superb treasure-house of fourteenth-century craftsmanship, the LADY 
 CHAPEL OF ELY (269). All this wealth of beauty perished in one moment, on 
 the awful advent of the Black Death in 1349. When we began to build and 
 carve again, it was in very different fashion, in sober Perpendicular. What 
 English art might have developed into may be seen in the exuberance and riot 
 of French Flamboyant. 
 
 The reign of purely naturalistic foliage was short ; and being peculiarly 
 characteristic of the period c. 1280 to c. 131 5, it is often a useful criterion of 
 chronology. As specimens of craftsmanship these carvings are simply con- 
 summate. Every capital, corbel, spandrel, boss, is a masterpiece. Nevertheless 
 it was a mistake, and stands on a far lower level than the conventional designs 
 which it superseded. It was a mistake to attempt to compete with nature :f 
 man's best efforts can be no more than a coarse reminiscence of nature ; the 
 humblest herb that grows in the cranny of a wall has a beauty that man 
 cannot emulate. The more successful he is, the more faithful, true, and 
 exact his reproduction of leaf, berry, or bloom, the more he invites comparison 
 with nature, and must suffer from the comparison. The more painting and 
 sculpture resemble camera work, the less artistic they are. In all this work, 
 workmanship had gone ahead of design. Design indeed there was none. What 
 credit there was belonged to the mason ; and to nature, who fashioned the leaf, 
 the fruit, and the bloom. 
 
 * In Devon a wreath of flowers and foliage, sometimes with ribbons, is common round the 
 piers in place of capitals. 
 
 t Cf. Barry's Lectures, 104. 
 
Chapter XXIX. 
 
 THE MOLDED CAPITAL. 
 
 Part I. — Tin: Abacus. 
 
 The abacus is the projecting slab which rests on the capital. It was made 
 to project for two reasons; sometimes to enable a small capital and pier to 
 support the springings of two broad arches ; sometimes that its projection might 
 give supports for the planks or centering, while the arches were being built. 
 
 In plan the Early Xorman abaci were square. But soon after 1089 
 cylindrical piers of the West of England have circular abaci ; e.g. Malvern, 
 TEWKESBURY (297;, Gloucester, Shrewsbury. At DURHAM (239), in 1093, 
 cylindrical piers have octagonal abaci ; so also at Buildwas. In the eastern 
 aisle of ELY TRANSEPT (41 2.i ), soon after 1090, the abacus is cruciform; so 
 also in Peterborough choir and Melbourne. Good examples of the subdivided 
 abacus occur also in Xorwich * (commenced in 1096 , in the eastern pier of the 
 triforium arcade, and in the piers of the nave arcade. It is remarkable to find 
 abacus and capital subdivided so early, each part having its own special function. 
 At Peterborough one part carries the roofing shaft ; another the transverse rib 
 of the aisle-vault ; two more its two diagonal ribs ; the remaining four carry the 
 two orders of each pier-arch.+ 
 
 The upper surface of the Xorman and Transitional abacus, up to 
 c. 1 180, was always horizontal, and its face was vertical; i.e. it was square- 
 edged, not rounded. In Normandy in the eleventh century the under surface 
 was always either a straight or a hollow chamfer.^: The hollow chamfer 
 occurs in our Anglo-Saxon abaci ; e.g. Wing nave and Deerhurst. Good 
 examples of both forms of chamfer occur in the west front of Lincoln. 
 Xot very much was done by the Normans § in molding the abacus ; molded 
 capitals they did not employ at all in their Romanesque. The molded capitals 
 of the Gloucester type are rather imposts than capitals. 
 
 The use of an abacus square in plan differentiates not only English 
 Romanesque from English Gothic; but also English from French Gothic. 
 The circular abacus is no doubt more in harmony with the general elimination 
 
 * Both are illustrated in Britton's Norwich, Plate 21. 
 
 t " These Peterborough vaults are, for the epoch, designed and executed with more science 
 and precision than those of the Domaine Royal of France or of Champagne or of Burgundy or 
 of Central France. When this construction is compared with the contemporary work in France, 
 one is astonished at the science and experience of the Anglo-Norman builders ; who already, 
 at the beginning of the twelfth century, were in a position to construct vaults with diagonal 
 ribs, and who divided their capitals into as many parts as there were arches to receive" 
 (Viollet-le-Duc, Architecture, iv. 101, 103). 
 
 I Ruprich-Robert, i. 124. $ Advanced specimens are seen at NEWPORT (421. 1, 5). 
 
440 
 
 1. Canterbury East Crypt. 
 
 2. Ketton. 
 
 3 Wisbech. 
 
 4. Shrewsbury Abbey. 
 
 5. Trunch. 
 
 6. Etchineham. 
 
THE ABACI'S. 441 
 
 (if rectangular forms from Gothic generally, And in some situations, e.g. in a 
 re-entering angle over an angle shaft, the square abacus looks awkward, and 
 room can hardly be found for it. On the other hand, it is useful, when set 
 obliquely, to point out the direction of the diagonal ribs of a vault; it allows 
 what is sometimes desirable, the retention of square with circular moldings in 
 the arch; or, as at Wells, in the pier; and its square form presents a much 
 more vigorous and emphatic stop to the moldings of arch above and pier below-. 
 Our abolition of the square abacus was by no means an unmitigated gain.* 
 
 The history of the English molded capital commences about the middle 
 of the twelfth century: and as the earliest examples seem to occur in Cistercian 
 abbeys — Buildwas, Fountains, Kirkstall, Furness, Byland, Jervaulx, in which the 
 Cistercian objections to carving were at first carefully respected — it may well 
 be clue in part to the efforts of the English Cistercians to find a simple and 
 effective substitute for the enriched Norman capitals which were being carved 
 elsewhere. Outside these abbeys it seems not to occur till it is employed 
 between 1154 and 1181 by the Secular Canons of Ripon, who had the Fountains 
 capitals hard by to study; and by the Benedictines of Canterbury. In CAN- 
 TERBURY T334J it occurs chief!)' in the eastern part of the crypt, which was 
 not commenced till the departure of the French architect, William of Sens, 
 in 1 179. 
 
 In Norman days the earl}' abacus was square-edged, as at CANWICK 
 (685.1;, with a deep vertical face. The first step to lighten this clumsy 
 form was to cut an angular nook (quirk) at the bottom of the face: as at 
 l-IARMSTON (2). So also at WHAPLODE (3), where the depth of the face 
 is further reduced by cutting a larger quirk.f But the Norman abacus often 
 had beneath it a hollow chamfer instead of a straight slope ; introduce this, and' 
 we arrive at the abacus of BUILDWAS, FOUNTAINS, KIRKSTALL (20, 21, 22). 
 Add a small roll {dead) beneath the hollow, and we have the fully developed 
 Transitional abacus, as it appears at ASWARBY, FURNESS, and BYLAND (5, 
 2T,, 24). The next step was to round off the upper edge of the abacus; 
 and, secondly, to undercut it with a small hollow to get a narrow band of intense 
 shadow : these two steps were probably taken in some districts just before I 190 : 
 e.g. at HORBLING and WHAPLODE (9,8). Finally, lighten the upper member 
 and give it more projection, and we arrive at the simplest type of Early 
 Gothic abacus, that of LINCOLN ( 12). 
 
 The final emergence of the Early Gothic type of abacus, c. I 190, is marked 
 by another very important change ; its plan was changed, not merely its profile. 
 Hitherto, most eleventh and twelfth century abaci had been square on plan ; 
 now, for a long period, except over octagonal piers, the Early Gothic abacus was 
 to be circular on plan. This then is the outcome of the evolution ; a light abacus, 
 circular in plan, with its edge rounded, i.e. molded ; and deeply undercut. 
 
 * See Sir G. Scott's Lectures, i. 156. 
 
 t This quirk is often useful in distinguishing true Norman work from later forms. 1 he 
 square-edged abacus may be late ; the quirked abacus not.— K. I'. B. 
 
442 MOLDED CAPITALS. 
 
 Part II. — Molded Capitals. 
 
 Below the semi-Norman abacus came the capital, which, as in Norman work, 
 was cut in a separate block ; see DEEPING ST JAMES (685). At the foot of this 
 block, at its junction with the shaft, was the astragal or necking. This was usually 
 semi-hexagonal, as at Deeping St James ; or semicircular, as at FURNESS (23). 
 Between the abacus and astragal there was usually a bare hollow chamfer. So 
 that, starting from the top, in Transitional work, the usual order is abacus, side- 
 hollow, necking. These hollow-necked capitals are of very ancient date outside 
 England. They occur in the nave of St Etienne, Nevers (choir dedicated in 
 1099);* they are common quite early among the Mendicant Orders, who had 
 taken them from the Cistercians.f English examples are seen at Ripon, Byland, 
 Furness, Holme, Cultram, Old Malton, Darlington, Roche, Hartlepool, in the 
 North of England ; Navenby, Deeping St James, and Market Deeping, in 
 Lincolnshire ; and even in the Early Gothic work of Rochester Cathedral. In 
 the Deeping caps only the lower part of the bell has the hollow chamfer. It 
 is possible that the bare hollow bell may have been intended to be painted, 
 as were some of the Norman cushion capitals, j 
 
 Thirteenth -Century Capitals. — I. As we saw above, these capitals, 
 on plan, are usually circular. This distinguishes them from the earlier capitals, 
 Norman and Transitional, which are usually square on plan. Of the later 
 capitals, those of the fourteenth century are generally circular ; the later ones 
 circular or octagonal. II. Again, up to c. 1180, the upper edge of the abacus 
 is always square; but from c. 1180 to c. 1360, it is nearly always rounded ; § 
 after c. 1360 it has often a straight chamfer. III. As for the profile of the Early 
 Gothic abacus, it varies immensely, as in the Lincolnshire examples : more than 
 a score forms might be described. Nevertheless it is quite distinct from the 
 fourteenth-century capitals : in which the abacus nearly always has a scroll 
 molding whose inner surface is an oblique ogee curve, and beneath this a 
 small roll. The nearest approach to this is such a cap as that of Stickney and 
 Lincoln ; || where, however, the ogee curve lies horizontally, not obliquely, and 
 below it is not a roll, but a hollow.H IV. The one mark of marks, however, 
 is the presence of a hollow cut beneath the abacus to give a narrow band of 
 shadow. In the later capitals this hollow is replaced by a small roll. In this 
 respect capital and base to some extent reflect one another. The undercut 
 abacus of the thirteenth century answers to the waterholding base ; and the 
 fourteenth-century capital to the base with triple roll.** In some cases, however, 
 the hollow is not marked very decisively ; sometimes it may even be absent, or 
 
 * Illustrated in Dehio, Plates 309 and 460. t Dehio, ii. 579. 
 
 J The abaci in the south doorway of Barnack have red painted chevrons, still clearly 
 visible.— R. P. B. 
 
 S The fourteenth-century capitals at Oakham have square-edged abaci. — R. P. B. 
 
 || Illustrated in Paley's Moldings, x. 9, 12. 
 
 T There is, however, one among the capitals of Threckingham south doorway, c. 1250, 
 which is of fourteenth-century type ; illustrated in Bowman and Crowther, Plate S. 
 
 ** These last, however, are not quite contemporaneous ; the triple roll of the base came into 
 general use some time before the fourteenth-century capital. 
 
MOLDED CAPITALS. I , j 
 
 its place may be supplied by a quirk.* And it must be remembered that the 
 hollow beneath the abacus occurs, not infrequently, in the fourteenth century; 
 e.g. in the dated examples of Hingham and Harleston; and even in later work, at 
 Lovvick.t V. Since the hollow is set under the abacus, a large space intervenes 
 between it and the necking, e.g. in DURHAM EAST TRANSEPT 689.3). - Ny " t 
 infrequently, therefore, two projecting members are set in this space ; and thus 
 what is called a double capital is formed, e.g. warmixgton and WIGGENHALL 
 ST MARY (689.6, 8). In later capitals the hollow does not commence till much 
 lower down, and therefore the double cap becomes rare. VI. The midroll 
 or rolls vary in form; but the most common is a roll with a fillet or fillets: 
 as at Warmington and Wiggenhall. VII. The astragal or necking is almost 
 as protean as the abacus : it was evidently a time of experiment ; the builders 
 were trying all sorts of experiments in light and shade. VIII. A band of 
 carved ornament was often inserted; either nail-bead, as at KKTToX (440.2 , 
 or tooth. 
 
 Fourteextii-Cextuky Capitals (c. 1315 to <-. 1360). — A great gulf 
 separates the early from the late molded capitals.^ In the first place the 
 double capital of the thirteenth century becomes rarer ; but a dated example, 
 1325, occurs at Harleston, § at Yaxley, Harringworth, and elsewhere. In 
 late examples the tendency is to reduce the number of projecting members 
 to two. Secondly, the capitals have less spread than before; that from 
 TILTEY (691.13) is notably narrow. Thirdly, the upper member is no longer 
 an abacus, cut out of a separate block ; and separated from the midroll by a 
 deeply upward cut hollow; though there are exceptions; e.g. Hingham, 1316- 
 1359. As the upper member is no longer an abacus, it is better to give it a 
 distinct name — "upper roll." Fourthly, the position and direction of the upper 
 hollow is changed. Apparently it was thought better to begin it lower down, 
 so as to make the upper band of shadow more central ; moreover it was cut 
 sideways instead of upward. Thus the " upper roll " becomes a much broader 
 member than the thirteenth-century abacus had been. 
 
 Normally, the typical capital, shown diagrammatically in the example from 
 CHIPPING HILL (444), consists of five parts, viz. three projecting members, 
 separated by two hollows. (1.) Almost always the upper edge, as in all the 
 thirteen capitals illustrated, has the scroll-molding. Beneath the scroll, as in 
 all the thirteen examples, is almost always an ogee curve running obliquely. 
 Beneath this is usually a small semicircular roll; HECKINGTON TOWER (691.5), 
 exceptionally, replaces the ogee and roll by a double ogee. Beneath this i-, 
 generally a vertical fillet, in a plane with the arch above, as in all the thirteen 
 examples. (2.) Then comes the small upper hollow, cut sideways. TILTEY 
 (13) is an exception. (3.) Then comes the midroll; which takes all sorts of 
 
 * I'aley's Moldings, x. 16, 17, 34 ; and Bowman and Crowther's Churches, Frampton, 5. 
 
 T Nem Valley, Plate 60. 
 
 \ Only the first thirteen capitals on page 691 belong to this period. The three Rushden 
 capitals, 14, 15, 16, are transitional in character, and were probably executed after 1360. 
 They diverge in several respects from the typical fourteenth-century capital, and may be left 
 out of account for the present. 
 
 § Illustrated in Churches of Northants, 270. 
 
 I! Illustrated in Paley's Moldings, 78. 
 
444 
 
 MOLDED CAPITALS. 
 
 eccentric shapes ; but most often consists of a scroll with a fillet beneath, as, 
 e.g. in Swayton, Sleaford, Heckington nave, Leadenham, Asgarby, Holbeach ; 
 or of some form of ogee roll with central fillet, as in Heckington chancel. 
 (4.) Then comes the bell ; a long shallow hollow, sometimes undercut above, 
 to hold shadow. (5.) Finally, the astragal or necking also is almost always 
 a scroll ; as in all the thirteen examples, except LEADENHAM (8). It will be 
 seen that the characteristic par excellence is the scroll-molding ; in the capitals at 
 Swayton, Sleaford, Heckington nave, Chipping Hill, x\sgarby, and Holbeach, it 
 occurs thrice. Three examples of double capitals are illustrated from Rushden ; 
 their general effect is similar to that of the TRUNCH CAPITAL (440.5). The 
 characteristic of the fourteenth-century caps is their remarkable uniformity ; at 
 this period alone there was one normal standard design of molded capital. 
 
 Fifteenth and Sixteenth 
 
 L 
 
 >--■ 
 
 >--4 
 
 Century Capitals. — These, on the 
 
 other hand, vary so much that it is 
 
 .\ difficult to characterise any one as a 
 
 ^ normal type. I. One great difference 
 
 is in their plan. Hitherto most Nor- 
 man capitals had been square in 
 plan, and Gothic capitals circular ; 
 unless the pier itself was octagonal, 
 when it usually had an octagonal cap. 
 The chief exceptions occur when the 
 shafts are small ; as in arcading ; e.g. 
 octagonal caps occur over semi- 
 circular shafts in the Gothic arcading 
 of Peterborough west front, and of 
 the walls of Fountains choir and 
 Histon transept, near Cambridge.* 
 When, however, an octagonal capital 
 occurs over a cylindrical pier or 
 column, the presumption is that it is 
 of the fifteenth or sixteenth, or at any 
 Chipping Hill. ra -te late in the fourteenth century. 
 
 II. Also the number of projecting 
 members in the capital tends to diminish. In the thirteenth century there 
 were three, and not unusually four rolls ; in the fourteenth century usually 
 three; but in late work, as at TERLING (693.1) and CHELMSFORD PORCH (3), 
 often only two. Out of thirty-four late capitals figured in Paley's Moldings, 
 about one-half have two; one-half three rolls. III. The top edge of the 
 upper roll, in the fourteenth century, e.g. at CHIPPING HILL, had been almost 
 invariably a scroll. This hardly ever occurs afterwards. If the upper rolls of 
 Terling and Chipping Hill be compared, it will be seen that each has an inverted 
 ogee on a small roll on a vertical fillet on a hollow ; but there is one important 
 difference. It is that, e.g. at TERLING (1) and INGATESTONE (6), the curved 
 top edge has been straightened. Probably 50 per cent, of Perpendicular capitals 
 
 * Paley's Moldings, 79. 
 
MOLUKD CAPITALS. 
 
 •445 
 
 start with this straight slope* A second type, seen at ST Mary's, OXFORD 
 (693.2), and in CHELMSFORD RORCII (3), resembles a keel-mold whose 
 point has either been rounded off or replaced by a fillet; in this capital the 
 upper slope has the ogee curve instead of being straight. In the third type 
 the upper member is a roll ; and beneath it is a fillet set obliquely ; r.g. r.oCKt.xc 
 (7) and CHELMSFORD CHANCEL (4). Fourthly, the profile of the upper 
 member may even be square, like that of a Norman abacus: and various other 
 forms occur. IV. The midroll, where it occurs, varies greatly. V. (1) The 
 astragal is often a simple semicircular roll, as in thirteenth-century work ; e.g. 
 Terling and ST ALPHEGE, CANTERBURY (446;. Another earl)- type, the semi- 
 hexagon, or a variation of it, may occur. f Or it may be a repetition of the 
 second type of upper roll ; e.g. in Chelmsford porch, upper roll and astragal are 
 identical ; so also at ELTHAM 
 (5). Or the upper edge of 
 the last type ma) 7 have a 
 straight chamfer, as at Hock- 
 ing ; or a hollow chamfer, as 
 
 in CHELMSFORD CHANCEL 
 
 (4). VI. Very often Per- 
 pendicular capitals arc nar- 
 rower and smaller than their 
 predecessors. 
 
 Origin of Molded 
 
 CAPI T a LS. — T he mol d ed 
 capital is probably of com- 
 posite origin. It was not 
 uncommon to leave the detail 
 of Romanesque capitals to be 
 painted ; as may be seen in 
 the eastern aisle of Ely 
 transept. And the reaction 
 against sculptured imagery 
 and artistic extravagance in 
 church work which was pro- 
 moted among the Cistercians by St Bernard led his Order to abandon foliated 
 capitals, and to replace them either by a simple scalloped capital, as at 
 Kirkstall, or by a naked bell, as at Fountains. It was, however, a con- 
 siderable step further to mold this bell ; especially when, as in Canterbury 
 choir, it was of hard marble from Bethersden, Petworth, or Purbeck. But 
 William of Sens " had lathes (torneumatd) constructed very scientifically to shape 
 stones " at Canterbury ; they could be utilised not merely to round the smaller 
 marble shafts, but to mold the capitals. We may, then, probably attribute the 
 origin of moldings on a hollow-necked capital to the use of lathes. \ In some 
 
 * The same chamfered edge occurs c. 1195 at Barnack and Orpington. 
 
 + So far as I have observed, the semihexagon is equilateral in Norman, but not in late 
 Gothic— R. I', n. 
 
 % These had been used in England long before the Conquest. The balusters of Jarrow, St 
 Albans, Worcester, &c, could not have been molded except by lathes. 
 
 Wells 
 
446 
 
 MOLDED CAPITALS. 
 
 cases indeed the marks left by the counterpoint in the process of turning may 
 be detected.* Suggestions came no doubt also from molded abacus, molded 
 band, and molded base. In Lincoln choir the molded bands of the vaulting- 
 shafts are in immediate juxtaposition to the molded abaci of the capitals. And 
 the upward appeal of the molded base is for a molded, not for a foliated capital. 
 Value of Molded Capitals. — These are almost wholly wanting in 
 French architecture, except when the shafts are very small, as in the mullions of 
 windows. Here, on the other hand, we have many great interiors in which the 
 capitals of all the piers are molded : Rochester is perhaps the earliest example ; 
 Rievaulx ; Beverley ; Salisbury ; Boxgrove ; the Temple choir ; St Saviour's, 
 Southwark ; Southwell ; Westminster ; Bridlington. The Cistercians revenged 
 
 Canterbury, St Alphege. 
 
 Lavenham. 
 
 themselves in Rievaulx choir for the prohibition of foliated capitals by the 
 unusual refinement of their moldings. The molded is, architecturally, more 
 logical than the foliated capital. The vertical lines of a molded pier demand to 
 be stopped above by the transverse lines of a molded capital and abacus, as they 
 are stopped below by the transverse lines of a molded base, and midway often by 
 those of a molded band. It may be urged also that if the capital is foliated, 
 so also should the base. And indeed, while the " griffe " was yet in use, it fre- 
 quently took the form of the same leafage as was employed in the capital ; e.g. in 
 Transitional piers both capital and plinth may be found with the characteristic 
 water-leaf; in thirteenth-century piers, e.g. in Romsey nave, with leaf-scrolls. 
 * Paley's Moldings, 76 ; and his Gothic Architecture, 98. Cf. Prior, Gothic Art, 156. 
 
Chapter XXX. 
 
 THE BASE, PLINTH, SPUR. 
 
 The Base. 
 
 No member of a Gothic building perhaps obtains so little notice as the base. 
 But of all molded members it is the most important to the archaeologist ; for 
 every period has its own favourite, strongly marked, and characteristic form. At 
 first, indeed, it received little attention from the Norman builders ; but even 
 before the eleventh century was complete, they may be seen experimenting with 
 their bases, especially those of the smaller shafts, molding them and carving 
 them ; e.g. in the apses and apsidal chapels of Gloucester, Norwich, and Lessay. 
 And indeed the base was worth attention ; for it is near the eye, and every 
 detail can be seen close at hand and fully appreciated. When this was 
 recognised, the base rapidly grew in dimensions and importance. At Durham 
 (239J, in GLOUCESTER NAVE (99), even so late as the middle of the 
 twelfth century in ST JOHN'S, CHESTER (448.1), the base is so insignificant 
 that at a distance it is hardly noticeable. But by the middle of the thirteenth 
 century the base had received much spread, e.g. at Ripon and TINTERN 
 (697.7 ) i indeed, had reached its maximum. For a practical objection to this 
 far-spreading base could not but be felt; it occupied far too much floor space. 
 So by the end of the thirteenth century it begins to retreat inward ; and in 
 the fifteenth and sixteenth century it reaches a minimum of projection, as in 
 the nave piers of BEVERLEY ST MARY (697.20). On the other hand, in late 
 work, by compensation, it is glorified in height ; it rises in the fourteenth century, 
 and in the fifteenth and sixteenth still more ; set base upon base, plinth upon 
 plinth, as in the five last bases ('697). No doubt also a practical reason 
 contributed very largely to the heightening of the later bases ; it was that 
 fixed seats had become common ; there are great numbers of them left in the 
 East Anglian churches, e.g. Tuddenham ; Dennington ; Wiggenhall St Mary 
 the Virgin and St Germans; KlLKHAMPTox, somerset (568); in these the 
 base could not be seen, unless it was set on a tall plinth so as to overtop the 
 benches.* The height then of the base is not controlled by the dimensions of 
 the pillar, as in Greek, Roman, and Renaissance work ; however high and 
 massive the pillar of a Gothic arcade may be, its base does not rise higher 
 
 * When Sir Christopher Wren built his London churches after the Great Fire, he followed 
 the same principle. The pews of the day were exceedingly high ; but he kept his bases above 
 their tops. Where the pews have been cut down, as at St Stephen's, Walbrook, and his naked 
 plinths have been exposed, the result is disastrous. 
 
448 
 
 tag t- mm- GHr 
 
 m H 
 
 i. St John's, Chester. 
 
 2. Southwell Choir. 
 
 3. Louth Chancel. 
 
 4. Tewkesbury Porch. 
 
 5. Ash. 
 
 6. Cirencester Nave. 
 
THE BASE. 
 
 449 
 
 than the tops of the benches. In classical work, however, where the: piers are 
 of great size, the bases have to be raised proportionately ; so that very logically, 
 but inartistically, they are sometimes above the heads of the observer ; e.g. 
 at St Peter's, Rome. Such gigantic bases could be appreciated only if then- 
 were giants about. An exceptionally early example of a tall Gothic base is 
 figured from the TEMPLE CHOIR, c. 1240 (689.4). 
 
 A great amount of thought was put into this humble member of the building. 
 In the first place, it was necessary that the base should have some relation to 
 the capital. . The first Norman caps were immensely heavy and clumsy ; e.g. 
 PETERBOROUGH '3 1 S ; and the bases ridiculously small. In logic, since the 
 base has more to carry than the capital, it ought to be at least as large. But a 
 big base is an awkward obstruction to an interior, and so the cap was usually 
 larger than the base ; this is well seen in the cap and base from the Temple, and 
 those from St Mary's, Oxford, c. 1492. That raises another question. Ought 
 a molded base to reflect the design of that of the molded capital, or vice versa? 
 It certainly does in the two examples cited above. Occasionally indeed the 
 base is simply the cubical cap inverted. No doubt also facility of execution 
 had a great deal to do with determining the profile of a base. Viollet-le-Duc 
 (695) illustrates the successive steps by which an early molded base may 
 be got out of two square blocks. In all the five diagrams, AB is supposed 
 to be left square, to serve as a plinth. First, a cylinder is formed out of 
 the upper square block ; and then the upper part of it is cut back so as 
 to form the smaller cylinder HE. Second, the groove F is cut back all round 
 the small cylinder. Third, the angles at G and II are chamfered off.* Fourth, 
 the nicks I, K, L, M are cut round the two cylinders. Fifth, IK, KL, EM are 
 rounded off; so that KE forms a hollow or " scotia " ; while IK, LM are rolls 
 or " bowtells." Evidently, such bases as those of BOXGROVE NAVE (3, 4) 
 might easily be obtained by such a process ; so also the characteristic Early 
 Gothic bases, e.g. at Fountains. And a base, that could be obtained so 
 simply, would be preferred to one more difficult to work.f Moreover, the 
 knowledge of how to produce such a base would be a workshop secret or 
 recipe. The mediaeval masons, no doubt, had a large number of trade secrets 
 of this sort pertaining to their craft, which they guarded as jealously as a 
 modern workman does. And, like a modern workman again, having got with 
 some trouble into the habit of doing a thing in a particular way, the natural 
 indolence of human nature would cause them to persist in that way. Traditions 
 of craftsmanship are not lightly broken with. They pass on from one genera- 
 tion to another ; from father to son and grandson. So there is nothing 
 surprising in the fact, e.g. that the WATERIIUEDING cash u which came 
 into use c. 1 1 50 was still not wholly discarded c. 1260; i.e. it was the heritage 
 of more than three generations of craftsmen. 
 
 There was still the artistic side of the question ; beauty of form, good 
 proportion, and above all, play of light and shade had to be secured. The 
 last — the production of shadow effects — must have been a great worry. If 
 
 * Some-times the upper roll, as in the second Boxgrove example (4), sometimes the lower 
 roll, sometimes both, are left chamfered, not rounded off. 
 
 t This is true, of course, not only of bases, but of all molded work. 
 
 2 E 
 
4SO 
 
 THE BASE. 
 
 you look up at a molded cap or abacus or string or rib or arch, the shadows 
 are on the underneath side ; and as you are looking from the underneath side, 
 you can see and enjoy them. In a molded base or in a ground-course, the 
 shadows are also on the underneath side of the rolls ; but unless you lie on 
 your back on the pavement, you cannot see them. It was distinctly a puzzle 
 then, how to make base-shadows visible from above. For about a hundred 
 years, as above, the problem was solved by the waterholding base. A bad 
 solution it was from every point of view but the artistic. If such a base is 
 placed outside a building, e.g. under the shafts of a doorway, the hollow fills 
 with rain and snow ; and the base soon must disintegrate ; indeed few of 
 these external bases have survived. While, if the base be inside the church, 
 it fills up with dust and dirt, and where then are the shadow effects? It 
 is very seldom indeed that one finds the mediaeval builder putting art in 
 front of construction ; and the fact that he did so in the case of the water- 
 holding base, is a witness to the great importance that he attached to play 
 of light and shadow. By the middle of the thirteenth century, however, the 
 hollow fills up ; and for about a century the base is often comparatively shadow- 
 less, e.g. at Southwell and TINTERN (697.7). But before this a new idea had 
 struck the builder. It was to let the base oversail the plinth ; as at TINTERN 
 (8); then the base shadows the plinth below ; and most of the shadow can be 
 seen from above. Here again the builder prefers art to construction ; for it is not 
 good construction to make the plinth narrower than the base. An early ex- 
 ample of this treatment is seen in SALISBURY CLOISTER (1); in late work the 
 narrow octagonal plinth beneath the overhanging lower roll of the base is 
 exceedingly common. It is well seen at LOUTH (23), BEVERLEY ST MARY 
 (20), ARUNDEL (22), and COLCHESTER (24). 
 
 The plinth or pedestal too has its special history, dictated partly by 
 practical, partly by artistic reasons. In Norman work it is generally square. 
 Next, the edges were chamfered off, and the square plinth became octagonal. 
 The octagonal plinth grew more and more common till about 1250; it is 
 still retained in Westminster nave. For about another century the plinth more 
 often follows the outline of the pier, in order to take up as little floor space 
 as possible. In the following period, c. 131 5 to c. 1360, the octagonal plinth 
 comes into general use again ; the slenderness of these late plinths causing 
 the octagon to waste but little space. In later work the plinth is almost always 
 octagonal ; e.g. at LOUTH and CIRENCESTER (448.3, 6). 
 
 We may now attempt to follow out the history of the base in chronological 
 order : — 
 
 Bases of the Eleventh and Twelfth Centuries.— In Early 
 Norman work the moldings of the base are often but mere surface scratches ; e.g. 
 Hereford choir aisles. Sometimes there is nothing but a straight chamfer, as 
 in the Caen abbeys ; or chamfers and a roll. More often the chamfer is hollow ; 
 and there may be two hollow chamfers superposed ; as at Jumieges. Or there 
 may be a hollow chamfer on a roll, as in Peterborough nave and at CASTLE 
 HEDINGHAM (695.5). Rarely the "pudding-base," a single clumsy roll, is 
 found ; it may possibly be a survival of the Tuscan base of Rome ; it occurs 
 in Oxford Cathedral south transept, and is not uncommon as a base or a 
 
THK RASE. 
 
 45' 
 
 capital in Anglo-Saxon work ; e.g. in Worcester slype. Insignificant as the 
 Norman bases usually are, it is evident that in some places a good deal of 
 attention was being given to them even in the eleventh century. There are 
 many varieties of base in the crypts of Rochester and Worcester;* in CANTER- 
 BURY CRYPT (193) many of Lanfranc's bases were recut and improved in the 
 twelfth century. 
 
 The most interesting base of all, and one that is by far the most common 
 on the Continent in Romanesque, owing to the greater survival there of Roman 
 examples, is the famous ATTIC base ^695. 1 ) of ancient Greece; which con- 
 sists in the main of two rolls, the bottom one having the greater projection, 
 separated by a hollow or scotia. The examples from TEWKESBURY PORCH 
 (448.4; are grotesque enough ; nevertheless, the)' are, as far as the mason 
 knew how, Attic bases. But in the later Norman examples from CASTLE 
 HEDINGHA.M (695.5) ar| d BOXGROVE (3, 4), the resemblance is very close. 
 Then comes an important change. In the brilliant sun of Greece, the side- 
 hollow gave an adequate shadow effect. Here it was ineffective. So the 
 hollow begins to be cut more downward than sideways. It only remains to 
 undercut deeply, and the Attic is transformed into the Early Gothic "water- 
 holding " base. 
 
 This waterholding base occurs at Kirkstall 4 - c. 11 55, 
 and continued in vogue for more than a century. But 
 one curious difference is almost always present between 
 the bases of c. 1 155 to c. 1 190 and later ones ; it is that 
 in the earlier work the lower roll is generally elliptical in 
 section ; it is somewhat flattened, as if it felt and partially 
 succumbed under the heavy weight it has to bear. 
 This is well seen in the base from ST JOHN'S, CHESTER 
 (448.1); and in the profiles of those from JERVAULX 
 and FURNESS (695.8, 6). The flattening of the lower 
 
 roll is very well marked in many bases of Canterbury Choir, as in those of Sens. 
 A good example is illustrated from P.ARNACK (695.9;, where the foot-ornament 
 or griffe consists of the flabby leaf which occurs within the same limits of time 
 as the flattened lower roll 416;. 
 
 Bases of the Thirteenth Century.— From c 1190 till it disappears 
 towards the end of the thirteenth century, the waterholding base usually has 
 lower rolls of semicircular instead of elliptical section; e.g. at west WALTON 
 (432), Whitby, Fountains, Netley, Bridlington. The great spread of these 
 bases is seen in the example from WHITBY '687.1 1. At this period both a 
 double-molded capital and a double-molded base is not uncommon ; as in ASH 
 CHURCH, KENT (448.5), where the plinth is not octagonal, but follows the plan 
 of the base and pier. In CANTERBURY CHOIR the deep bands of shadow are 
 emphasised ; base and plinth are doubled ; and the plinths are concentric with 
 the bases and shafts. It should be noticed that, almost invariably, the large 
 spreading lower roll is worked out of the block (with which it usually stands 
 
 * Six from Worcester are illustrated in Rickman, 8;. See Ruprich-Robert, Plates 220-226, 
 for the variety of bases in the Abbaye-aux-Dames, Caen. 
 t Illustrated in Sharpe's . In//. Parallels, Plate 99. 
 
 Canterbury Choir. 
 
45 : 
 
 THE BASE. 
 
 flush, i.e. in the same plane) by a quirk, i.e. an angular nook ; * e.g. at Whitby, 
 Fountains, Netley, Bridlington. 
 
 About the year 1240 the waterholding base begins to be largely supplanted 
 by the base with double or triple roll, which is well seen in SOUTHWELL CHOIR 
 (448.2) ; it abounds in Westminster. It is still in great vogue in the first half 
 of the fourteenth century : e.g. at BOTTISHAM (697.19). An early example 
 is seen in the TEMPLE CHURCH, c. 1240 (689.4), where it has a triple roll. The 
 same type appears in the massive piers, c. 1290, of the tower of Merton College 
 Chapel, Oxford. Less often, there are two, not three rolls, as in WINCHELSEA 
 SEDILIA (697.3). The practical objections to the waterholding base and the 
 difficulty and cost of working it probably led to its disuse ; and the new base 
 may be regarded as an improved version of it ; the two rolls being brought close 
 together, as in 3, or the hollow being filled up with an intermediate roll, making 
 
 three in all. Or it may 
 be of a more ancient 
 descent ; for there is a 
 Romanesque triple roll 
 base, which occurs 
 sporadically ; e.g. in the 
 ninth-century tower of 
 S. Satiro, Milan ; -j- in 
 the crypt of St Benoit- 
 sur-Loire ; % in the 
 crypts of Rochester 
 and Worcester ; and 
 the western doorways 
 of Lincoln. Other 
 peculiarities connected 
 with the triple or 
 double roll base are 
 its great projection ; 
 early examples are 
 often as broad as they 
 are high ; that it 
 usually stands on a concentric and not on an octagonal plinth : e.g. Merton ; and 
 that in later examples it often oversails the plinth. 
 
 Bases of the Fourteenth Century. — The following century is charac- 
 terised by ogee arches and ogee curves all over the church. For a time the 
 triple roll base held its ground, as at BOTTISHAM (697.19) ; but in the end the 
 contours of the base of this period came to consist almost entirely of reversed 
 ogees ; or wave-moldings, of which there are two in the Bottisham example. 
 Moreover, it began to contract ; and as it contracted in breadth, it rose in height. 
 Thirdly, the octagonal plan of plinth was reverted to. All these three charac- 
 teristics are retained in fifteenth-century work, but with certain additions. The 
 bases of the fourteenth-century period may be regarded as transitional between 
 
 * Paley's Moldings, 86. t Illustrated in Rivoira, i. 276. 
 
 { Illustrated in Enlart's Manuel, 368. 
 
 ,'.,",,'7 y„"„\ 
 
 1, 2. Heverley Minster, Back of Reredos. 
 3. Rochester, Doorway to Chapter House. 
 
THE BASE. 
 
 base and the later ba 
 
 45. 3 
 
 the triple roll base and the later base; i.e. they are but imperfeel specimen; 
 of the latter;* so long as they retain the triple roll, and have considerable 
 spread, they attach themselves to the earlier type ; but when they abandon the 
 triple roll, when they introduce ogee or undulatory curves, when they become 
 narrow and tall, when they rise from octagonal plinths, when the base oversails 
 the octagonal plinth, they are tending fast to the later type. In the example 
 from the back of BEVERLEY REREDOS 452 it is only necessary to elongate 
 base and plinth, diminishing the spread, to have a close approximation to the 
 typical late base. An example of the ogee curve occurs at LEADENHAM (697.14) ; 
 and of the wave-molding at SLEAFORD i;, and BOTTISHAM (19); all these 
 have considerable spread. The bases from the doorway to the CHAPTER HOUS1 . 
 ROCHESTER (452;, are thoroughly late in type; they are generally attributed to 
 Bishop Hamo cle Hythe, who died in 1352. Caveler assigns them to his 
 successor, John de Sheppey (1 352-1 360 !. 
 
 SCALE. OF l 
 
 St George's, Windsor. 
 
 Fifteenth and Sixteenth Century Bases. — With these, as with their 
 ground-courses, the later builders were peculiarly successful ; the best examples 
 are exquisite in proportion ; and have a delightful flow of curve, gradation of 
 shadow, and nice contrast of octagon and circle. Their tall and slender pro- 
 portions distinguish them from most of their fourteenth-century brethren. The 
 base of the nave of BEVERLEY ST MARY (697.20) may be regarded as fairly typical. 
 There are four members — A, the base ; B, the upper octagonal plinth ; C, the 
 molded plinth, separated from 1: by a hollow chamfer, d, and from D by a quirk, 
 /; and i>, the lower octagonal plinth, the lower part of which is omitted in the 
 diagram. The base A oversails the upper plinth, I;. The roll of the molded 
 plinth, C, is flush with the lower octagonal plinth, D. The base, A, has a roll, a, 
 at the top ; the isolation of this annular roll is very characteristic of the style. 
 Usually, as at Beverley, it is semicircular ; but it may be a debased form of 
 
 * On the other hand, the later molded caps are degenerate specimens of the perfect ones 
 of the fourteenth century. 
 
454 
 
 THE BASE. 
 
 the scroll-molding ; or, as at COLCHESTER (697.24), a debased roll and fillet. At 
 the bottom of the base A is another roll, c. The space between is usually filled 
 in, if there is room, with a " bracket " (double ogee), or some modification of 
 a bracket, b. As a whole, the base is bell-shaped. Some such profile, 
 contracted in small examples, heightened in large ones, is the normal one. 
 Sometimes, as in ARUNDEL NAVE (22), the molded plinth, C, is omitted. 
 In a tall example, such as the bases of the piers of LOUTH STEEPLE (23), an 
 additional octagonal plinth, E, may be added. With the Beverley base should be 
 compared the bases of LOUTH CHANCEL (448.3), ROCHESTER DOORWAY (452), 
 and ST MARY'S, OXFORD (693.2). 
 
 As was pointed out above (244), some of the Late Gothic piers contain 
 
 Base and Capital of Pier, Aldeburgh, Suffolk. 
 
 a large number of columns and shafts, and the builders took much pleasure 
 in composing for each, however small, a fully developed base; each strictly 
 proportioned in breadth to the diameter of the shaft or column which it sup- 
 ported. An elaborate and early example occurs in Gloucester south transept, 
 1 330- 1 337. Sometimes, as in St Mary's, Oxford, and St Mary Redcliffe, Bristol, 
 all these bases, big and little alike, are on the same level. More often, as at 
 ST GEORGE'S, WINDSOR (453)* and the ROCHESTER DOORWAY (452), their 
 height as well as their breadth is proportioned to their load ; and thus they 
 become of several different heights. In such a case the upper member of all — 
 
 * The Cirencester piers and bases are illustrated on 544. 
 
THE SPUR. 455 
 
 what we have called the necking — may be on the same level, and the plinths 
 therefore descend to different levels ; and the builders found much amusement 
 in providing for them supports at different heights (see especially the Rochester 
 example). At Fotheringhay,* less happily, the smallest bases occur at the 
 lowest level. In Flamboyant work, all over Europe, complicated bases occur ; 
 care being often taken in them to give independent support to every rib of the 
 vault, and even to the moldings of the rib. + 
 
 SPUR. — This is variously styled the spur, claw, foot ornament ; French 
 griffe. It is used to fill up the unoccupied corners of a square plinth on which 
 rests a circular base. When the plinth became octagonal, or concentric with the 
 base, the spur naturally soon disappeared, as there was no longer room for it. 
 Griffes are found in Roman work in a similar position; the griffes at SpalatoJ 
 are very similar to those of st Peter's, Northampton (663. ij, being geometri- 
 cal in form. Pliny describes griffes of lizards and frogs ; with which may be 
 compared those in the crypt of St Peter's, Oxford. The)- are common in 
 Lombardic Romanesque ; e.g. in S. Ambrogio, Milan. S In our Romanesque 
 they increase in number with the general enrichment of Norman work towards 
 the middle of the twelfth century. Transitional examples may be seen at 
 BARNACK (695.9) ; ar "d York crypt ; thirteenth-century examples at Stockbury ; 
 ST CROSS, WINCHESTER (695) ; Romsey west nave ; and Salisbury ; in the last 
 they even occur on a circular plinth. In PYance griffes occur at Sens Cathedral 
 in the fourteenth century ; in the fifteenth century at Pont-sur-Yonne ; and in 
 the sixteenth at St Martin d'Auxerre and Chatel-Censoir. Griffes were especially 
 favoured in Burgundy.* 
 
 * Plate 59 in Nene Valley. 
 
 t See illustrations of piers of St Severin, Paris ; St Lambert, Minister ; and S. Juan de los 
 Reyes, Toledo ; in Dehio, Plate 567. 
 
 X Illustrated in Choisy's Histoire, i. 546. § Ruprich-Robert, Plate 1S7. 
 
 || Illustrated in Viollet-le-Duc Architecture, vi. 52. *I Enlart's Gothique en Italic, 279. 
 
Chapter XXXI. 
 
 THE ORIGIN OF WINDOW TRACERY. 
 
 On Windows — Balustered Windows — Norman Windows — Dimensions of Opening — Early 
 Glazing — Lancet Windows, Single and Grouped — Origin of Plate Tracer}' — Early 
 Bar Tracery — Disuse of Lancets. 
 
 Romanesque Windows. 
 
 One of the earliest types of window is that in which two or more lights are 
 separated by a shaft or by a baluster. It is very common in the Campanili of 
 Rome and Ravenna, and in Northern Italy ; it is equally common in the 
 Primitive Romanesque of Germany ; in Anglo-Saxon and in Norman work. 
 It occurs almost wholly in towers, e.g. ST MARY BISHOP HILL JUNIOR, YORK 
 (457.1); the baluster not being well adapted for glazing. In all these windows 
 the difficulty arises that the tower-wall is thick ; and that when two arches are 
 pierced through it, it is difficult to support their inner springing on the abacus 
 of a shaft. The favourite Anglo-Saxon way out of the difficulty was to place 
 on the shaft a long impost running the whole thickness of the wall, as at St 
 Mary Bishop Hill Junior, York. This was not good construction, as the pro- 
 jecting ends of the impost were without support, and might snap off. So the 
 Normans usually constructed a recessed arch under each of the arches, and if 
 necessary another ; i.e. they built the two window arches in two or more 
 recessed orders ; and thus the thickness of arch to be supported was little 
 greater than that of the shaft. 
 
 From the Anglo-Saxon method of construction it followed that the shaft 
 was placed centrally in the wall ; it is designated a midwall shaft. In the 
 Norman method, it was possible, according to the way the recessing was 
 managed, either to place the shaft centrally or not. Almost always, it was 
 placed towards the outer face of the wall ; probably because it could thus be 
 seen better, and become an effective architectural feature.* 
 
 The normal Romanesque window, however, is that which had been in use 
 both in Roman and Byzantine work ; an oblong opening, under a semicircular 
 arch. Usually the glass was set rather close to the face of the wall ; and, in 
 order to admit as much light as possible, the opening had a broad internal 
 splay. So that, though the window arch was small, the rear-arch was large. 
 But the external reveal was soon increased in depth, e.g. at NORWICH (160), 
 partly for decorative purposes ; in order to admit a recessed order and jamb 
 shafts ; partly, perhaps, to protect the glass. So again at STEYNING (458) 
 it will be seen that in the clerestory the glass is set well back from the face 
 
 * Both methods are illustrated in Baldwin Brown's Arts in Early England, ii. 64. 
 
45; 
 
 i. St Mary Bishop Hill Junior, York. 
 
 2. Romsey. 
 
 3, 4. Oxford St Giles. 
 5, 6. Netley. 
 
 ~. Winchester Castle Hall. 
 
 8. Winchester St Cross, 
 
 g. Stanwick Tower, 
 
 lo. Salisbury Cathedral Transept. 
 
453 
 
 EARLY WINDOWS. 
 
 of the wall. And occasionally well-marked instances of Norman windows 
 occur, splayed both internally and externally ; e.g. in the south cloister wall of 
 Hereford, and the west cloister wall of Norwich. 
 
 Steyning from N.E. 
 Salisbury from E. 
 
 Towyn Nave. 
 Salisbury from S.E. 
 
 The size of Romanesque windows varies curiously. Originally they were 
 large, as in the early Christian basilicas, and at St Generoux and the Basse 
 CEuvre, Beauvais ; and at Brixworth, which is probably seventh century. Later, 
 however, in many countries they become mere arrow-slits. Perhaps the earliest 
 
EARLY WINDOWS. 459 
 
 examples of these are the windows of the Ravenna mausoleums of Galla Placidia, 
 c. 440, and Theodoric, sixth century ; where the size of aperture is evidently 
 reduced for the sake of security. Isolated examples occur in the churches of 
 S. Victor and S. Agatha, Ravenna, and of Bagnacavallo. An early example of 
 the small window is at Alliata, ninth century.* In Norman work, some of these 
 openings, e.g. at Nateley, Hants, and Stow, Norfolk, do not exceed a few inches 
 in breadth, though 2 or 3 feet high. In many cases the aisle-windows are 
 small, the clerestory windows large; e.g. at Secqueville in Normandy ;+ St 
 Margaret at Cliffe; STEYNING (458) ; at Steyning the former are 1 foot broad, 
 the latter 2 feet 9 inches. At TOWYN, NORTH WALKS (458), the clerestory 
 windows also are diminutive. Where the aisles are loft)', as in cathedrals, e.g. 
 Norwich, Ely, Peterborough, the aisle-windows can be set high and therefore can 
 be large. From these facts we would suppose that where the aisles were low, 
 the windows were made diminutive in order to prevent pirates or thieves in 
 general from scrambling through. It was for the same reason perhaps that 
 Anglo-Saxon windows in general were set so very high in the wall ; e.g. at 
 Wing and Worth. 
 
 But it may be that peculiarities or difficulties of glazing had something 
 to do with the diminution of the window area. It is difficult to say when 
 glazing came into ordinary use. It was much used in the sixth century, in 
 Byzantine churches ; Wilfrid is said by Eddius and William of Malmesbury 
 to have "provided with glass the windows of York Minster, which formerly 
 derived their light from the transparency of linen or of boards pierced with many 
 holes." J Bede tells us that Benedict Biscop, founder of the monastery 
 of W'earmouth, sent to Gaul in 675 for workers in glass. But it does not 
 follow that, even in the eleventh and twelfth and thirteenth centuries, every 
 church, great or small, glazed its windows. Indeed we know that it was not so. 
 At Waltham and Darenth no provision was made originally for fixing panels of 
 glass. A shutter probably was designed to close upon the rebate, which some- 
 times may be seen worked externally in the masonry, as at Southease, Sussex. § 
 In the thirteenth - century chancel of Clymping, Sussex, all the windows, 
 including the triplet of lancets in the east end, are rebated externally, and 
 retain the hooks on which the shutters hung. : At the end of the twelfth 
 century glass was still expensive. Even in the great Norman abbey of 
 Peterborough, after the nave was finished in 1190, forty of the windows 
 remained filled with reeds and straw to keep out the rain for another twenty 
 
 * Cattaneo, 239, 240. t Ruprich-Robert, Plate 37, Ms. \ Willis' York, 3. 
 
 S "At the restoration of Great Yarmouth Church in 1870 the transept windows were found 
 never to have been glazed, but only provided with shutters" (A. W. Morant in Norfolk and 
 Norwich Arch. Sue, vii. 215). It must be borne in mind, however, that the absence of a 
 glazing groove or rebate does not necessarily mean that a window was not glazed. We know 
 that in houses, till quite a late date, the glass was often inserted in wooden frames, which were 
 carried about by the rich from house to house. In Ely transept the glass was originally fixed 
 in wooden frames, which were wedged into the window openings (Stewart's Ely, 44). 
 
 I] Brandon's Analysis, 12. The great church of Constantine at Tyre, described by 
 Bishop Eusebius, had not glass, but pierced wooden panels (Scott's Essay, n). S. Mimato, 
 Florence, eleventh century, has transparent slabs of alabaster. In Anglo-Saxon windows 
 "the actual opening for light is at times cut in a thin slab of stone or a plank of wood built into 
 the wall at the centre of its thickness" (Baldwin Brown's Arts in Early England, ii. 93). 
 
4 6o LANCET WINDOWS. 
 
 years.* It may be, therefore, that sometimes the windows were reduced in area 
 in the poorer churches because of the expense of glazing. 
 
 De Dartein, ii. 475, points out that in Lombardy the smallest windows are 
 those of the twelfth century ; whereas in the Byzantine and Basilican churches 
 the breadth is about one-half of the height, in these it is one-third, one-fifth, or 
 even one-tenth. He suggests another explanation ; viz. that the difference in 
 size of the windows is due to a difference in the methods of glazing ; that in the 
 early windows a wooden or stone framework was placed to contain the glass, and 
 therefore they had to be broad ; whereas in the later windows the glass was fixed 
 with iron and lead directly in the stone jambs. Cattaneo objects to this ; and 
 suggests that the builders designedly plunged the churches into gloom to 
 produce effects of mystery and awe ; but that is to look at mediaeval building 
 through the spectacles of a modern aesthete, and to ignore the fact that the 
 old men at all periods strove above all things to make their churches bright 
 and cheerful. 
 
 LANCET WINDOWS. — For hundreds of years the heads of windows remained 
 semicircular; at last the heads were pointed, producing the "Lancet" type of 
 window. The pointed arch had been in use for some time in constructional 
 arches ; but there was a considerable reluctance and delay in employing it 
 where the construction did not demand it, as in a window or doorway, or where 
 the arch was merely a decorative one, as in the arcading of an aisle ; e.g. at 
 FOUNTAINS (101), RIPON (102), and New Shoreham the arches of the pier-arcade 
 were pointed, to facilitate the vaulting ; but the arches of the windows of 
 Fountains and Ripon are semicircular, as are those of the aisle arcading of New 
 Shoreham. Indeed to eyes that were habituated to the consecrated semicircle, 
 so repugnant was the new pointed arch, that in the aisle of Glastonbury nave, 
 though the window-arch is pointed, its rear-arch remains semicircular. But 
 the desire for harmony in the end proved too strong ; not merely harmony 
 with the pointed arches of the pier-arcade, but with the pointed arches which 
 spanned the aisle, and above all with the pointed wall-arches of the aisle-vault. 
 Indeed the immediate juxtaposition of wall-arch and window-head made 
 identity of shape inevitable^ 
 
 As regards the proportions of the lancet window, there was at first a 
 considerable diversity of treatment. Those of Canterbury choir, designed 
 by William of Sens, naturally were of the French type ; somewhat broad in 
 proportion to their height. A still greater excess in width is seen in the contem- 
 porary windows of Glastonbury and Wells. Those of Wells are no less than 
 5 feet 4 inches broad, with a height of only 13 feet 5 inches. Usually the 
 precedent neither of Wells nor of Canterbury was followed. Everywhere in the 
 thirteenth century the English lancet tended to differentiate itself from the 
 French by its great loftiness and narrowness. And being slender, any single 
 lancet was a thing of beauty in itself: while an)' single lancet of Canterbury 
 or Wells or France was without grace. At Bottesford is a lancet 15-J- feet high 
 and only 8 inches broad. It has been said, indeed, that the breadth of the 
 
 * Craddock's Peterborough, 52. 
 
 t Nevertheless, for a time, as at Byland and elsewhere, round-headed windows were set 
 under the pointed wall-arches of the aisle vault. See Plate 21 in Sharpe's Arch. Parallels. 
 
461 
 
 < 
 
 > I 
 
 .M *±,i'bl*<*ki., t i^*tL.miJt,*m 
 
 I 
 
 
 
 ■3 
 
 6 
 
 
 
462 
 
 Temple Choir. 
 
LANCET WINDOWS. 
 
 4 r >3 
 
 English lancet varies inversely with its height* In England, therefore, it 
 was seen that the graceful single lancet could be composed into even yet more 
 graceful groups of lancets: and that group could be contrasted with group. 
 And so the history of the window passed in England through a whole stage 
 of development which was summarily shortened in France. France arrived at 
 window -traceiy rather 
 earlier than we ; on the 
 other hand, they ex- 
 perienced little of the 
 delights of Lancet- 
 grouping. 
 
 In BOXGROVE 
 CLERESTORY (373) 
 
 single lancets appear in 
 the clerestory ; in the 
 aisles of LINCOLN 
 (no) and in the 
 chapter house the 
 lancets are arranged in 
 pairs ; in the CLERE- 
 STORY OF LINCOLN 
 (i 1 5) and SALIS- 
 BURY (45 8 ) they are 
 in triplets. And as an 
 odd number of lights 
 was better adapted 
 for filling up a gable, 
 three, five, or seven 
 lancets in a group are 
 more common than 
 four or six. Southwell 
 choir, however, has a 
 fine example of two 
 sets of four lancets 
 superposed ; and six 
 occur in Repton 
 Church. So again, 
 the five gigantic lan- 
 cets of the NORTH 
 
 TRANSEPT OF YORK 
 
 11, the famous Five 
 Sisters, are not gradated in height. And these groups were nicely contrasted ; 
 e.g. in ELY EAST FRONT (464) a quintet of gradated lancets has above and 
 beneath it a triplet uniform in height. In BEVERLEY TRANSEPT (176) a gradated 
 is superposed on a non-gradated triplet. + 
 
 * Brandon's Analysis, iS. 
 
 t Perhaps the most beautiful groups of all are those of the lancets of the choir of Brecon 
 Priory Church, illustrated in Scott's Lectures, i. 2S5. 
 
 YVestwell, East End. 
 
464 
 
 LANCET WINDOWS. 
 
 As in Norman windows, the glass was usually placed near the external 
 face of the wall. Outside, therefore, there was little room for ornament. But 
 there was all the more room inside ; and windows plain externally like those 
 of the aisles of Worcester and CARLISLE (498) are often glorious within. 
 The contrast is well seen in many a window on the chalk of Kent or Sussex, 
 e.g. STONE CHURCH (513). Indeed, where freestone was scarce or dear, it was 
 naturally reserved as far as possible for internal use. 
 
 The lancet window seems to have come into use about 11 70; and though 
 its supremacy was challenged more and more, first by plate-tracery, and then 
 by bar-tracery, it was in constant employment for something like a century.* 
 
 Ely East Front. 
 
 Hereford Lady Chapel. 
 
 Traceried Windows. — Both Norman and lancet windows were at first 
 employed singly on the flanks of the church, but in groups in the end walls. 
 We find pairs of lancets in ST GILES', OXFORD (457.3, 4) ; and again in the 
 aisles of Lincoln choir and nave, where they are separated by a minor but- 
 tress (115). At WILEY (466) two advances are seen. Three instead of two 
 lancets are put side by side. Secondly, the central lancet is made higher 
 than the two side lancets : this is an important step ; it is to introduce the 
 principle of grouping. The same principle is recognised in a different way in 
 Lincoln clerestory ; by the central lancet being made the broadest of the three. 
 A remarkably early example of this treatment occurs in the west wall of ROMSEY 
 
 * Here and there it occurs much liter; e.g. in the fourteenth century at Ottery St Mary 
 and in the south clerestory of St Albans nave. In the latter it is assimilated to the thirteenth- 
 century work in the west of the nave. 
 
PLATE-TRACERY 
 
 465 
 
 south transept* (457.2); which, even if not part of the original work, can 
 hardly be later than the middle of the twelfth century. Another is in tin- 
 west front of GLASTONBURY LADY CHAPEL, 1185. Both are round-headed. 
 Such a combination was inevitable after the 
 treatment, so very frequent, of the inner 
 arcade of the Norman clerestories, e.g. at 
 ELY (57), Waltham, and PETERBOROUGH 
 (161). 
 
 But it was usual to protect the windows 
 from the drip of the rain by a dripstone. 
 Sometimes each window has its own inde- 
 pendent dripstone ; but in one of the St 
 Giles' windows, and in the Lincoln clere- 
 story, though each window has still its own 
 dripstone, all the dripstones are linked up. 
 This goes far to diminish the independence 
 of each window of the group. The next 
 step is still more important. It is to supply 
 the several windows of the group with a 
 common dripstone. Sometimes, as in the 
 aisle-window of WARMINGTON (461), each 
 window has its independent dripstone and 
 a share in a common dripstone. More often, 
 as at NETLEY (457.5, 6), and WINCHESTER 
 (457-7. 8), they lose their independent 
 dripstone, and have only the common one. 
 But the result is a blank space of masonry ; 
 at Netley a blank spandrel. To ornament this, an aperture is pierced 
 through ; a circle, a quatrefoil, or the like. And so in this gradual but inevit- 
 able way arises what is called Plate-Tracery. The evolution may be well 
 seen by comparing the two WARMINGTON WINDOWS (461) ; that of the aisle, 
 which was built first ; and that of the BELFRY (467). At SALISBURY (170) best 
 of all one can watch the transition from the grouped lancets of the lower parts 
 of the church to the plate-tracery of the gables of the transepts. 
 
 The above is the main origin of English Plate-Tracery ; nevertheless it is a 
 mistake to regard it as the only origin. Even in the Anglo-Saxon towers of 
 Northumberland it was by no means uncommon to pierce the spandrel of the 
 baluster windows of the belfry with a quatrefoil or some such aperture ; e.g. at 
 Billingham.t But these remote examples of Anglo-Saxon plate-tracery seem to 
 have escaped recognition or to have been ignored. Again, in two bays of PETER- 
 BOROUGH CHOIR (468), 1117-1140, the spandrels of the triforium arcade are 
 pierced with little circles. When such a triforium arcade as that of Rievaulx 
 or WHITBY (114) was designed, c. 1200, plate-tracery was inevitable. Again, 
 when two lancets were set in the same bay of an aisle, e.g. at CARLISLE (498), 
 it was only natural to pierce the spandrel between the tops of the lancets and 
 
 * Cf. the west front of patrixrourne (218). 
 
 t See illustrated paper by Mr C. C. Hodges in Reliquary. Jan. 1894, on the Pre-Conquest 
 Churches of Northumbria. 
 
 2 G 
 
 Glastonbury Lady Chapel, \\". Front. 
 
466 
 
 Wilev, Wilts. 
 
467 
 
 SCALE 9 1 2 $ 2 &0F FEET 
 
 \fy arming ton tower. 
 
PLATE-TRACERY. 
 
 the wall-arch. A very early example of this is seen in the south transept of 
 Kirkstall, where each bay is lighted by a pair of windows with a circle above 
 them.* 
 
 Again, many an early west or east front had a wheel window in its gable ; 
 e.g. Barfreston, PATRIXBOURNE (218), Iffley, Castle Hedingham ; above two or 
 three of the usual type of oblong round-headed or lancet windows. To turn 
 
 such a combination into a great 
 plate-tracery window all that 
 was necessary was to frame the 
 combination under a common 
 arch.f If the end walls of 
 Chartres, with a circular window 
 resting immediately on a group 
 of lancets, be examined, it will 
 be seen that here certainly is 
 the origin of the early and re- 
 markable plate-tracery windows 
 of Chartres aisles and clere- 
 story. \ So also the super- 
 position of a big circular window 
 on a pair of pointed arches in 
 the cloister of Laon § was sure 
 to lead to plate-tracery. || 
 
 Of plate-tracery windows 
 the earliest we possess is the 
 Norman window of ST MAURICE, 
 YORK, which may be c. 1160, 
 and the Transitional one of the 
 porch of ST MARY'S, SHREWS- 
 BURY, c. 1 1 80; next come those 
 of the Bishop's Hall at Lincoln, 
 begun before 1200 and finished 
 before 1224 ; those of the Win- 
 chester Castle Hall,1T finished 
 in 1234; and those of the 
 Bishop's Hall at Wells, not later than 1239. Other examples of plate-tracery 
 are illustrated from the belfry of OXFORD CATHEDRAL (512); from the aisle 
 
 * The same thing occurs at Notre Dame de la Couture, Mans ; where the aisleless nave is 
 vaulted in great squares, and two lancets are set in each bay, and the spandrel above them is 
 pierced with a circle. Cf. Dehio, ii. 9, 5S0. 
 
 t At Sporle, near Swaffam, a triplet lancet is seen under a foliated circle (Paley's 
 Gothic Architecture, 158). 
 
 I So also Dehio, loc. cit. ; and Lethaby's Med. Art, 171. 
 § Illustrated in Viollet-le-Duc, Architecture, iii. 429. 
 
 || In each side gable of Salisbury lady chapel (458) a cusped circle occurs over a pair 
 of lancets. The Lady Chapel was commenced in 1220 and finished in 1225. A good illustra- 
 tion of these windows is given in Parker's Glossary, Plate 231. 
 
 II Illustrated in Rickman, 165. 
 
 Peterborough Triforium. 
 York, St Maurice. 
 
 Peterborough Triforium. 
 
 Shrewsbury, Porch of 
 
 St Mary's Church. 
 
PLATE-TRACERY. 460 
 
 of STONE CHURCH, KENT (5 1 3 ; from BEELEIGH CHAPTER HOUSE; and 
 from ST JAMES', BRISTOL (516). 
 
 In France plate-tracery windows of great size and complexity were designed 
 at Laon c. 1170; at Chartres c. 1194. These windows, filled with rich twelfth- 
 century stained glass, must have made a great impression on the French builders 
 elsewhere. But it was not everywhere that such quarries could be found as 
 those near Laon and Chartres and in Burgundy, producing slabs that were 
 neither brittle nor difficult to work. Still there was a desire to have the new 
 windows ; and the question would arise, how to obtain the circles, jewelled in 
 ruby and azure, of Laon and Chartres, without possessing the materials requisite 
 for the production of plate-tracery. In France and in England the same and 
 even better effects were got by substituting Bar for Plate Tracery. Hitherto 
 the geometrical patterns had been got out of slabs of freestone, almost as if one 
 were cutting patterns out of a board with a 
 fret-saw. Now the stone was first worked 
 into curved bars ; which, when put together, 
 gave the circle, the triangle, or what not. 
 This great revolution followed fast on the 
 Chartres plate-tracery of 1 194. Orbais is 
 said to have had bar-tracery by 1 200;* the 
 eastern chapels of Rheims Cathedral had it 
 c. 1 2 12. 
 
 In England bar-tracery did not arise 
 quite so early ; we were engrossed in com- 
 posing triplets, quintets, sextets of lancets. 
 It is usually said that our first bar-tracery is 
 that of Westminster choir, begun 1245 ; 
 which, with the large circles in the window- 
 heads, may be admitted to be of French 
 origin. But the important windows in the 
 west front of KIN HAM 471 must not be 
 overlooked. They are now much dilapidated, 
 but may be restored with certainty from the 
 drawing by John Coney in Dugdale's Monasticon. The central window is 
 one of eight lights : divided into pairs ; each pair is contained by a detached 
 pointed arch, with a quatrefoil in its head. Kach pair of these arches is again 
 included in a detached pointed arch, which has in its head a sexfoliated circle. 
 Finally, these carry a centrepiece consisting of an octofoliated circle. Now 
 Matthew Paris + expressly says that " Prior Richard de Parco built the front of 
 the church from the foundation to the roof." He was Prior from [226 to 1244. 
 Consequently it would appear that we have at Binham large windows of bar- 
 tracery which are anterior to those of Westminster. The early date assigned 
 to the Binham work is borne out by comparison with other examples. Of all 
 our early bar-tracery windows the most important is the east window ot 
 LINCOLN PRESBYTERY (i//;, which was begun soon after 1256 and consecrated 
 in 1288. So great and magnificent a window- must have had predecessors ; and 
 * By Demaison. t I am indebted for the reference to Mr W. H. St John Hope. 
 
 fieeleigh Chapter House. 
 
47° 
 
 EARLY BAR-TRACERY 
 
 its prototype may well be the west window of Binham ; the number of lights is 
 the same, the skeleton framework of the tracery is the same, practically every 
 detail is the same except that the circle in the window arch is not octofoliated, 
 but contains nine foliated smaller circles. Another important early example 
 
 is the east window of the choir of 
 NETLEY (471), which was begun in 
 1239 by the executors of Peter de 
 Roche, Bishop of Winchester, with 
 funds bequeathed by him for the pur- 
 pose. This window has but four lights ; 
 otherwise the design is precisely the 
 same as that of Binham. Finally, we 
 know that the presbytery of St Paul's, 
 London, was consecrated, wholly or in 
 part, in 1240, in the presence of King 
 and Legate ; and that here also were 
 windows of fully developed bar-tracery. 
 Putting the evidence together, we may 
 fairly come to the conclusion that large 
 bar-tracery windows were in use in 
 England not later than the year 1240. 
 A drawing of one of the aisle-windows 
 also of BINHAM WEST FRONT is given ; 
 the lower half lighted the vaulted aisle, 
 the upper half the triforium.* The bar- 
 tracery windows of the west front of the 
 abbey of Valle Cruris, near Llangollen, 
 also appear to be of very early date.f 
 The abbey is said by Dugdale to have 
 been founded c. 1200. 
 
 So far we have traced the change 
 from grouped lancets to plate-tracery, 
 and from plate to bar tracery. But it 
 may be asked, why did we abandon 
 the combinations of lancets, so beautiful 
 and effective in Salisbury and else- 
 where? The chief reason doubtless was that we wanted bigger windows, so as 
 to make the churches less gloom}' ; the stained glass of the period being as yet 
 thick and opaque. Another, it may be, was the desire for harmony in bay design. 
 
 The ruins of Binham Abbey are 4 miles from New Walsingham, Norfolk. The facade is 
 built in admirable stone, and is of consummate design. For drawings and plan see Britton's 
 Architectural Antiquities, iii. 72 ; Colling's Details, i. 23, 24, and ii. 34 ; Parker's 
 Glossary, Plates 123 and 146. The whole of the work is very similar to that of West Walton 
 Church, near Wisbech, which may be by the same builders, where is an exquisite Early 
 Geometrical window of the Binham type ; cf. also the tooth ornament in Colling's Details, i. 22 
 and 24. West Walton is illustrated in Colling's Details, i., Plates 9 to 22. 
 
 + The exterior and interior of the west front of Valle Crucis are illustrated by H. H. 
 Hughes in the Architectural Association Sketch Book, New Series, viii. 35. 
 
 Binham Aisle, West Window. 
 
EARLY BAR-TRACERY. 4; , 
 
 With one bay and two windows, as in Salisbury aisles, or three as in Salisbury 
 clerestory, there was a lack of harmony. What was wanted was one hay, one 
 window. Here and there, too, the builders may have been feeling their way to 
 the realisation of the ideal of the Gothic building, as one whose 'vault could be 
 made to stand without support of walls: an ideal realised early in such chapter 
 
 SCALE OF FEET 
 
 Binham, West End. 
 
 Netley, East End. 
 
 houses as those of Westminster and Salisbury. In such a building it wa^ un- 
 necessary to confine the glass within the jambs of narrow lancet windows ; the 
 whole space from buttress to buttress could, if desired, be glazed. But if glazed, 
 the glass must have mullions and bars of stone to support it ; in other words, there 
 must be tracery. From one cause or other the traceried window was inevitable. 
 
Chapter XXXII. 
 
 GEOMETRICAL TRACERY. 
 
 We now come to what is to most the outward and visible sign of Gothic 
 architecture, the traceried window. Traceried windows were divided by 
 Rickman into Decorated and Perpendicular, and the division still stands 
 in the text-books. It is thoroughly misleading, however. A threefold 
 division is necessary. The so-called Decorated needs to be subdivided into 
 Geometrical, and what Mr Sharpe calls Curvilinear, Professor Freeman Flowing 
 Tracery.* In practice this has always been admitted. Brandon, Bloxam, 
 Bowman and Crowther, Parker, Sir G. G. Scott, constantly subdivide Decorated 
 tracery. Geometrical tracery they variously style Early Decorated, or Early 
 Middle Pointed, or Early Geometric Pointed ; Curvilinear tracery they style 
 Late Decorated, or Pure Decorated, or Flowing Middle Pointed. All these 
 clumsy periphrases are avoided by adopting the terms Geometrical and 
 Curvilinear. 
 
 The former came into use in windows of the former rank not later than 
 1240 in Binham, Netley, and Old St Paul's. It remained in vogue, according to 
 Mr Sharpe, till about 1315. But many windows of purely Geometrical character, 
 e.g. those with Kentish tracery or the ball-flower, were constructed for another 
 generation or so. It should be borne in mind also that though Geometrical 
 patterns without admixture of Flowing or Rectilinear patterns are rare after 
 1 3 1 5 > y & t in such admixtures they are common to the last. In fact in Tudor 
 days there was quite a strong reaction in favour of the employment of purely 
 Geometrical patterns into window-tracery, basement-courses, parapets, &c. ; 
 especially the quatrefoil ; e.g. KING'S COLLEGE, CAMBRIDGE (473). 
 
 Early Geometrical Tracery. — But we can go further. We can sub- 
 divide Geometrical tracery into Early and Late work.f The first is seen in 
 the BINHAM (470), NETLEY (471), and RUDSTON (508) windows, in SALIS- 
 BURY CLOISTER (458) ; in SELBY CHOIR AISLE (86) ; in all these the main 
 centrepiece consists of a foliated circle. So it does in WESTMINSTER CHOIR 
 (63) and chapter house. j In other windows the main centrepiece is still a 
 circle ; but instead of being merely foliated, it contains a ring of smaller circles ; 
 e.g. the east windows of LINCOLN PRESBYTERY (177) and TINTERN CHOIR 
 (475); this choir was finished in 1287. 
 
 Late Geometrical Tracery.— Of this there are three marks. First, 
 the Geometrical figures employed — trefoils, quatrefoils, &c. — are not always 
 bounded by circles, as in the earlier tracery ; but the lines of the foliations 
 
 * Unfortunately Freeman added another, quite unnecessary, division ; Flamboyant. 
 
 t To all these divisions and subdivisions exceptions occur. 
 
 % Canvas was bought for the windows of the chapter house in 1253. 
 
A71 
 
 King's College Chapel. 
 North Creake. 
 
 Exeter Lady Chapel. 
 Merton College Chapel. 
 
474 
 
 i. Gedney Chancel. 2. Ledbury Nave. 
 
 4- Ledbury, St Catherine's Chapel. 
 
 3. 'Wells Lady Chapel. 
 5. Hull, S. Chancel (Curvilinear). 
 
GEOMETRICAL TRACERY. 
 
 475 
 
 form the main lines of the figures, e.. 
 (546), the aisles of York nave ; at 
 Secondly, not only circles, but othei 
 e.g. trefoils, quatrefoils, lozenges, 
 spherical triangles, and spherical 
 squares. Thirdly, in the span- 
 drels of Late Lancet arcades, 
 e.g. at Binham, a long-lobed, 
 pointed trefoil had been common. 
 This was introduced into the 
 tracery c. 1290 ; and, where 
 present, is an infallible sign of 
 Late Geometrical work ; e.g. it 
 appears in the great east windows 
 of Ripon and GUISBOROUGH 
 (476), c. 1290 to c. 1300, and in 
 the west windows of Tintern 
 and HOWDEN (72). The Guis- 
 borough window, the drawing of 
 which is from a restoration by 
 
 taller even than the Curvilinear 
 windows of York and Carlisle. 
 A Late Geometrical window of 
 singular beauty remains in what 
 was formerly the chapel of the 
 Bishop of Ely, in Ely Place near 
 the Holborn Viaduct, London, 
 1 290- 1 300. The choir of EXETER 
 (473) is 130S to 1327. The 
 window in the Lady Chapel, like 
 all windows with five, seven, or 
 nine lights, was difficult to design. 
 It has seven lights: the central 
 one being taller and broader, and 
 independent of the other six. 
 These six are under detached 
 arches. Each pair of these arches 
 is contained in two intersecting 
 and detached arches ; these two 
 arches are placed under an en- 
 gaged pointed arch, in the head 
 
 '-. in the clerestory of HOWDEN NAVE 
 
 Easby, GUISBOROUGH (476;, Exeter.* 
 
 Geometrical patterns are employed ; 
 
 '.- -it ■ r 
 
 Tintern Choir. 
 
 >f which is a circle. The centrepiece is a 
 large circle, in which are six trefoils, three round-lobed and inverted, three 
 pointed and normal in position. 
 
 We may again subdivide the Late Geometrical windows ; first, into those 
 with centrepieces ; secondly, into those without. Of those with centrepieces we 
 * See illustrations in Sharpe's Windows, pp. 77, 7S, 82, and Plates 17 and 24. 
 
476 
 
 GEOMETRICAL TRACERY. 
 
 have first, those in which the 
 centrepiece is a circle ; e.g. 
 GUISBOROUGH and WELLS. 
 CHAPTER HOUSE ( I 2 3 ). 
 
 There is an elaborate speci- 
 men in the chapel of MER- 
 TON COLLEGE, OXFORD (473), 
 c. 1290. It has seven lower 
 lights,of which the central light 
 is treated independently. On 
 each side of it are three 
 pointed arches ; these are set 
 beneath two intersecting 
 pointed arches, the outer one 
 of which is engaged. These 
 two intersecting arches are 
 again set under an engaged 
 pointed arch. The centrepiece 
 is a wheel. 
 
 Secondly, the centre- 
 piece may be some other Geo- 
 metrical figure ; not a circle. 
 At NORTH CREAKE (473) it 
 is a round-headed trefoil. In 
 the west window of howden 
 (72) — the transom is not 
 original — it is a cusped quatre- 
 foil inscribed in a spherical 
 square, of which the two upper 
 sides are formed by the 
 window arch. In ST CATHE- 
 RINE'S CHAPEL, LEDBURY 
 (474.4), it is a quatrefoil 
 studded with ball-flower. 
 
 Thirdly, the centrepiece 
 
 may have what is called 
 
 Kentish Tracery. It appears 
 
 in all the chancel windows of 
 
 CHARTHAM CHURCH (477), 
 
 Kent, near Canterbury ; and in 
 
 the Infirmary Chapel and St 
 
 Anselm's Chapel in Canterbury 
 
 Cathedral ; also in Whitby 
 
 Nave ; Billingborough ; Great 
 
 Bedwyn, Wilts ; Lyddington, Berks ; Capel St Mary, Suffolk. It has been 
 
 variously described. Really it is nothing but a cusped quatrefoil or trefoil 
 
 with straight or curving spikes projecting from between the lobes. But, if a 
 
 Guisborough. 
 
GEOMETRICAL TRACERY 
 
 477 
 
 quatrefoil, instead of its normal position, it lies diagonally. It is strong in con- 
 struction, and handsome and unusual in design. It occurs so often in con- 
 nection with ogee arches that probably it is usually later than 131 5. 
 
 Then, there are several interesting types of Late Geometrical windows without 
 centrepieces. The first contains two or more lights, whose pointed arches are 
 continued to the main arch of the window so that they intersect. This inter- 
 secting tracer}-, when without cusps, as at Barholme and LEDBURY (474.2 j, 
 Herefordshire, is not graceful. From its cheapness many windows which 
 had fallen out of repair were rebuilt by the churchwardens in this fashion 
 during the seventeenth and eighteenth centuries, 
 e.g.iw Leominster aisle and Chichester choir aisle. 
 More often it is cusped as at Dorchester, and 
 the interstices also may be filled in with small 
 trefoils, quatrefoils, and circles : as at GEDNEY 
 (474.1), DURHAM (514), and SOLIHULL (626). 
 
 The second is a group of gradated lancets, 
 with bar-tracery. It is seen in its simplest 
 form in the clerestory of the choir of St 
 Albans,* begun c. 1257; and in ST MICHAEL'S, 
 OXFORD (478). It is more common in Late 
 than in Early Geometrical work ; in Late work 
 it is generally cusped, as in the choir of MILTON 
 ABBEY, DORSET (478) ; begun after the fire 
 of 1309. Naturally these windows are triplets, 
 or quintets, as the great window of the north 
 transept of YVimbornej which, like that of 
 Milton Abbe)', contains the long-lobed pointed 
 trefoil. At Ottery, however, the east window 
 has eight lights. 
 
 A variant of this, with elongated mullions, 
 was in use well into the middle of the four- 
 teenth century in the West of England ; e.g. in 
 Tewkesbury choir ; \ Berkeley ; Portbury ; 
 Somerset; LEDBURY (474.2); and Ottery St 
 Mary. All these types persist well into the 
 fourteenth century. 
 
 A third type, without centrepiece, is one 
 whose tracery is entirely formed of foliated patterns ; trefoils, quatrefoils, and 
 the like. It seems usually to be late ; for it is found surmounted with an ogee 
 dripstone in Lichfield Lady Chapel, and in proximity to undoubted Curvilinear 
 work in the LADY CHAPEL OF WELLS (474.3), which was finished in 1326. So 
 again in HEDON WEST NAVE (480.3), it appears next to an undoubted curvi- 
 LINEAR WINDOW (474.3). Good examples are the four-light windows of Chester 
 choir, and the three-light ones of the aisle of GLOUCESTER NAVE (360). 
 
 Fourthly, yet another set of windows seems to have been in use till well into 
 
 * Illustrated in Scott's Essay, Plate 23. 
 
 t Illustrated in Freeman's Windows, Plate n, 52. J Illustrated in Glossary, Plate 239. 
 
 Chartham. 
 
473 
 
 GEOMETRICAL TRACERY. 
 
 Oxford, St Michael. 
 
 the fourteenth century ; those studded with ball-flower. These also are charac- 
 teristic of the West of England. The most remarkable are the Leominster 
 windows ; * each of which contains 820 ball-flowers ; 
 while in the GLOUCESTER WINDOWS (360), 1318-1329, 
 if a horizontal line be drawn below the spring of 
 the arch of each window it cuts through 32 bands of 
 ball-flower ; and there are no less than 1,400 ball- 
 flowers in each window. Most beautiful of all windows 
 of this type are those of ST CATHERINE'S CHAPEL, 
 LEDBURY (4744). 
 
 The above classification by no means exhausts the 
 variety of Late Geometrical tracery, and its chrono- 
 logies must only be taken to be true in a general way ; 
 but at any rate it may serve to characterise the leading 
 types. The great treasure-house of it is the Cathedral 
 of Exeter; begun c. 1270 and finished c. 1350; with the later tracery largely 
 assimilated to the earlier. 
 
 It is curious that though the French commenced to produce bar-tracery 
 somewhat before ourselves, indeed at the very begin- 
 ning of the thirteenth century, they did not develop 
 its later Geometrical forms any sooner than we did, 
 nor to anything like the same extent. It was not till 
 the last quarter of the thirteenth century,f that in 
 addition to circles they began to employ other forms 
 such as triangles and lozenges. It is plain that the 
 development of Geometrical window-tracery went on 
 pari passu in the two countries; neither j borrowing 
 of the other. We may add that the French to the 
 very last retained their liking for the simpler and 
 earlier types of tracery, viz., those with circular centre- 
 pieces. In France they occur frequently in the four- 
 teenth century ; e.g. in the choir chapels of Notre 
 Dame, Paris, 1320 ;§ and of St Ouen, Rouen. The 1 ? 3 f ?*?!?? 1° FEET 
 chief difference was that the fourteenth-century tracery Milton. 
 
 was made excessively exiguous ; a triumph of skill on 
 
 the part of the mason. The chief centrepiece, other than a circular, in large use 
 in France was one of a pair of spherical triangles, supporting a third. 
 
 * Illustrated in Sharpe's Windows, Plate 21. + Enlart's Manuel, 5 28. 
 
 % Except probably at Westminster ; whose tracery, however, made no disciples. 
 § Illustrated in Viollet-le-Duc, Architecture, vi. 338. 
 
Chapter XXXIII. 
 
 CURVILINEAR TRACERY. 
 
 All are characterised by the presence of flowing curves and ogee arches ; and 
 usually date between 131 5 and 1360. They may be arranged in three classes. 
 
 Class I. — In these the main arch of the window is subdivided into two 
 detached sub-arches ; i.e. the outer sides of which are not the same as the sides 
 of the main arch of the window. These detached sub-arches are usually ogee 
 arches. If it be a large window with 4, 6, or 8 lights, the ogee sub-arches 
 are usually placed side by side; e.g. in the west window of YORK (82), where 
 the ogee sub-arches each contain four lower lights. In such a window there 
 is abundance of room for a centrepiece, which is usually large ; at York it is 
 unusually important, being double ; it consists of a heart-shaped figure, carrying 
 a vesica. These contain three leafed stems, which are repeated on either side of 
 the heart. Xo centrepiece is so successful as the leafed stem ; its beauty is 
 well seen in the "Bishop's Eye" of LINCOLN (484); and, somewhat confused, 
 in the east window of CARLISLE (128). 
 
 But if there are 5, 7, or 9 lights the ogee sub-arches intersect in very 
 charming fashion, as in the east window of HULL (81), thus binding 
 together the otherwise disconnected lateral compartments. The intersecting 
 ogee arch became one of the favourite motives of our Late Gothic ; it occurs 
 constantly in the woodwork of the screens ; and is employed not unsuccess- 
 fully so late as A.D. 1636 in the window of Jesus College Chapel, Oxford. 
 
 It should be noted that there is always a difficulty in designing tracery 
 for 5, 7, or 9 lower lights. The easiest way out of the difficult)- is to treat 
 the central lower light independent!} 7 , as in the seven - light windows of 
 EXETER (473) and Merton College, Oxford. So at Carlisle the separate 
 treatment of the central light leaves four lights on each side, which admit 
 of symmetrical treatment. But sometimes instead of dividing a window, say 
 of seven lights, into 3, 1, 3 ; it was divided into 2, 3, 2. Almost always 
 trouble followed, e.g. in the Late Geometrical west window of Tintern, which 
 has seven lights, the central sub-arch rose higher than the other two, and 
 had to be truncated to make room for the circular centrepiece.* So it was 
 also in Curvilinear work. The great windows of Heckington, HAWTON (483), 
 and Selby have seven lights arranged as at Tintern ; but the unfortunate 
 corollary is that the great ogee intersecting sub-arches have each one short 
 and one long leg. 
 
 * Illustrated in Sharpe's Arch. Parallels, Plate 53. 
 
480 
 
 i. Haverfordwest. 
 4. Hunstanton Porch. 
 
 2. Merton Vestry. 
 5. Old Walsingham. 
 
 3. Hedon Nave. 
 
 6. Tideswell Transept. 
 
48 1 
 
 
 
 Snettisham Weit Window. 
 
 2 II 
 
CURVILI \F..\R TRAC'KRY. 
 
 4«3 
 
 Detached pointed sub-arches are much less common in this class ; but 
 fine examples occur at Chipping Norton; Thurnham ; Plympton St Mary's; 
 Exeter west nave ; and St Saviour's, York.* 
 
 CLASS II. — In these the main arch of the window is divided into two 
 engaged sub-arches ; i.e. of which the outer sides coincide with those of the 
 main arch. As the main arch is pointed, it follows that the sub-arches also 
 will be pointed, and 
 not ocree arches. 
 
 If the window 
 has an even num- 
 ber of lights, the 
 engaged sub-arches 
 are generally placed 
 side by side : as in 
 HULL CHANCEL 
 (474.5). In that 
 case there is not 
 much room for a 
 centrepiece, which 
 therefore becomes 
 small and uninter- 
 esting ; e.g. in the 
 south of the chancel 
 of Hull it consists 
 merely of foliations: 
 so also in the south 
 transept of ST MARY 
 REDCLIFFE, BRIS- 
 TOL (376). 
 
 If the window 
 has an odd number 
 of lights, the central 
 light is usually 
 treated independ- 
 ently, as at Carlisle. 
 In such a case there 
 is room for a large 
 centrepiece. The 
 east windows of the 
 aisle of Hull chancel, 
 on the other hand, 
 
 have fixe lights without a centrepiece ; 
 of uninteresting foliations.f 
 
 Class III.— This is a miscellaneous collection of all windows whose skeleton 
 does not contain two detached or engaged sub-arches; e.g. in TIDESWELL 
 
 * Illustrated in Freeman's Window Tracery, 59, &5> 66, 6 7, 105. 
 t Illustrated in Sharpe's Windows, Plate 52. 
 
 Hawton. 
 
 for which is substituted a collection 
 
4§4 
 
 CURVILINEAR TRACERY. 
 
 TRANSEPT (480.6), the five bottom pointed arches are united into four inter- 
 secting pointed arches, and the two central of these into one ogee arch. At 
 Hedon, Salford, Hingham, Old Walsingham, the three lower arches retain their 
 independence altogether ; so do the four lower arches in Westminster cloister : 
 illustrated below. 
 
 I. Of this class the most common window is the Reticulated ; so called 
 
 from the net-like appearance of its tracery. It 
 
 Lincoln South Transept. 
 
 probably the earliest of all 
 Curvilinear tracery. For 
 it appears in the vestry 
 of MERTON COLLEGE 
 CHAPEL, OXFORD (480.2), 
 the foundations for which 
 the College accounts show 
 to have been dug in 13 10.* 
 And indeed if a window 
 have three or more lights 
 whose heads are ogee 
 arches, and these ogee 
 curves be continued till 
 they meet the window arch, 
 Reticulated tracery results 
 inevitably. And each 
 pattern, moreover, except 
 the uppermost one, has two 
 ogee arches, one right way 
 up, the other inverted. 
 Naturally, in the mania 
 prevailing in the fourteenth 
 century for the ogee arch, 
 Reticulated tracery came 
 early into favour. There 
 is a superb specimen in 
 WESTMINSTER CLOISTER 
 (489) ;f in this every ogee 
 pattern contains Kentish 
 tracery. 
 
 II. Flamboyant tracery 
 developed in a very simple 
 way out of Reticulated. 
 The first thing necessary 
 
 to change a Reticulated into a Flamboyant window is to give an acuter point to 
 the ogee heads of the lower lights. As a corollary, the wavy curves of these ogee 
 arches will prolong themselves up to the window arch much more vertically than 
 before : which is very satisfactory. For the more the curves of flowing tracery 
 tend to the horizontal, as in the east window of Shottesbrooke, % the weaker they 
 
 * See Rickman, 237. t A Reticulated window is shown at FRAMPTON (61S). 
 
 I Illustrated in Butterfield's Shottesbrooke. 
 
CURVILINEAR TRACERY. 4 s^ 
 
 are to the eye, and the weaker they are in reality. The more vertical, the 
 stronger and the more satisfactory.* The French saw tin's clearly ; and flowing 
 curve and countercurve, always essaying the vertical, are the note of their later 
 tracery. A good Flamboyant window is in essence the same as a Perpendicular 
 one ; except that the tracery, instead of rising in straight supermullions, flickers 
 to right and left. Secondly, in the best windows, the upper How of curve is 
 not a mere " repeat " of the ogee curves of the arches of the lower lights; all 
 sorts of subtle curves occur, as in the Beverley reredos : so that the patterns 
 cease to be sym metrical ogees. 
 
 Of Flamboyant windows that from CHIPPING NORTON (487.4) is of 
 typically English type. + In the window from the south aisle of the nave 
 of 1 1 EDON (480.3;, the aspiring effort of the curves is somewhat retarded by the 
 depressed ogee arches beneath. At SALFORD, WARWICKSHIRE 4S7.2 , the 
 pointed arches of the outer lower lights are a blemish ; for the flow of their 
 inner curves is suddenly cut short. 
 
 The finest examples of our Flamboyant tracer}- are to be found in arcadings 
 and doors : e.g. in the stone screens of Selbv choir, probably copied from 
 woodwork;* and of BEVERLEY MINSTER (4.86J ; where it will be seen that 
 only the central pattern is a symmetrical ogee. 
 
 It has been the fashion to include the few examples we possess in Fngland 
 of Flamboyant tracery in the sweeping condemnation of the magnificent Flam- 
 boyant windows of France. This is the merest Chauvinism. They are both 
 beautiful and practical. No other Flowing window equals them in beauty of 
 curve, except perhaps the leafed stem windows of the south aisle of Beverley 
 nave ; while their design in some cases, e.g. in the aisle-windows of Alencon, is 
 almost as well adapted for the reception of stained glass as in the best of our 
 late Perpendicular windows, such as that of the DIVINITY SCHOOL, OXFORD 
 
 (492). 
 
 CLASS IV. — The rest, like the TIDESWELL WINDOW (480.6), defy classi- 
 fication. Many of them arc hybrids; half-curvilinear; half-geometrical; or 
 exceptional forms, such as the roue tournante of HL'XSTAXTON PORCH (4X0.4). 
 
 Origin of Curvilinear Tracery. — It may not be uninteresting to turn 
 back to some of the beautiful Geometrical tracery, such as that of Guisborough, 
 Exeter Lady Chapel, and Merton College Chapel ; and endeavour to see whether 
 there is any practical or artistic defect in it ; or whether it was merely abandoned 
 from a desire for freshness and novelty. There can be little doubt that the 
 latter motive was not without weight ; for great as is the variety of simple 
 Geometrical forms, it is not infinite ; one comes at last to an end of the com- 
 binations of circles, trefoils, quatrefoils, spherical triangles and squares, and the 
 like ; repeats occur ; monotony sets in. So, no doubt, it had been some seventy 
 years before, when the beautiful Lancet groups were abandoned for Geometrical 
 tracery. Another reason, however, is suggested by Mr Sharpe ; viz., that these 
 rigid Geometrical forms left awkward interspaces (spandrels) between themselves 
 
 * That is why the leafed stem centrepiece is so satisfactory. 
 
 + Others are illustrated in Freeman's Window Tracery, Hates 44, 45, from Jersey ; South- 
 well ; Bolton Priory; St Mary Magdalen, ( )xford ; Etchingham ; Hawton. 
 \ Prior, Gothic Art, 393. 
 
486 
 
CURVILINEAR TRACERY 
 
 48; 
 
 and the window arch, and between one another. In such a window as that of 
 Exeter Lady Chapel these spandrels are very numerous ; they arc of all sorts of 
 shapes, difficult in foliation and difficult to glaze. There is no doubt some truth 
 in this ; in many of the later Curvilinear windows; e.g. those of Tideswell and 
 Selby, and especially in Flamboyant windows, such as those of Chipping Norton 
 and Salford, the difficulties of the treatment of the spandrels are reduced to a 
 minimum, or even are triumphantly surmounted. But in others, as in OLD 
 WALSINGHAM (480.5) and Westminster cloister, the spandrels are even more 
 
 I. Little Addington. 
 
 4. Chipping Norton. 
 
 2. Salford. 
 
 3. Caldicott. 
 Reticulated Tracerv. 
 
 awkwardly treated than even in the Geometrical period. Indeed every Reticulated 
 window is fringed with the scraps and fragments of patterns cut off by the 
 window arch. This was felt at the time ; for we find several Geometrical windows 
 of the type of that of HAVERFORDWEST (480.1), and Curvilinear ones as at 
 CALDICOTT, Monmouth ; in which the scraps and fragments are got rid of 
 by omitting the window arch altogether, or by making it conform to the curves 
 of the ogee pattern. But another practical consideration may be suggested. 
 These windows were meant to tell the story of saints and angels. But neither 
 
CURVILINEAR TRACERY. 
 
 saints nor angels are known to be triangular or circular or square, or in any way 
 geometrical in shape. Such a window, then, as that of the Exeter Lady Chapel 
 is largely wasted, as a means of preaching Scripture History ; practically the 
 whole head of the window is useless to the glassman. If, however, we turn to 
 a later type, such as that of Selby or York ; or still better to the Flamboyant 
 windows of Fiedon, Salford, and Chipping Norton, we shall see how much 
 better the glass craftsman can now utilise the tracery lights ; they almost seem 
 to invite the hierarchy of heaven to tabernacle in their niches with outstretched 
 wings. From the point of view of the glass craftsman — and he was beginning 
 to be a bigger person than the architect — the best Flowing tracery was a great 
 improvement on the best Geometrical. 
 
 Nor can it be denied that Flowing tracery is in itself far superior to 
 Geometrical. Who would care to gaze day after da}' into combinations of 
 square, circle, triangle, rather than graceful intertwining flowing curves? And 
 of course, artistically the Flowing tracer}- is on an altogether higher plane ; in 
 the window of Exeter Lady Chapel there is rule and compass, in the Carlisle 
 window human brain. We may add, too, that in Geometrical tracery, unity 
 
 cannot be reached in the design : the 
 sub-arches maintain their independence 
 of the centrepiece, and the centrepiece 
 of the sub-arches ; they do not fuse. 
 But in a good Curvilinear window, 
 especially if it have Flamboyant tracery, 
 mullion fuses into tracery without the 
 slightest breach of continuity. Not that 
 all the windows by any means reach 
 this high standard, e.g. at TIDESWELL 
 (480.6) the pointed arches and the 
 Flamboyant tracery are discordant. In 
 WESTMINSTER CLOISTER (489) there is 
 an awkward break between the pointed 
 arches of the lower lights and the Re- 
 ticulated tracery above. The same awk- 
 wardness occurs at the same place in the CARLISLE WINDOW (128); and, in 
 addition, five Geometrical patterns, quatrefoiled circles, are intruded ; arresting 
 the flow of curve : as also in the BEVERLEY REREDOS (486). There was many 
 a divagation in directions where no road was. Nevertheless in many a Cur- 
 vilinear window mullions and tracery at last touched and fused. " The trunks 
 of the mullions, all springing from one soil, rose up to a certain height, and then 
 shot themselves out into ramifications of the most intricate and delicate net- 
 work, exhibiting a variety of combinations which baffles enumeration ; the 
 branches climbing and twisting one into the other in a maze full of entangle- 
 ment, yet without confusion ; the whole composition nevertheless, in its utmost 
 license and seeming extravagance of fancy, subjected to the strict and inviolable 
 laws of primary truth." * 
 
 Setting Out of Flowing Tracery.— So far we have spoken of Flowing 
 
 * English Review, iv. 417. 
 
 Waltham. 
 
CURVILINEAR TRACERY. 
 
 489 
 
 as opposed to Geometrical tracer)-. 
 As a matter of fact, both are 
 Geometrical ; in both the curves 
 are drawn by a pair of compasses. 
 The difference is that the curves 
 of the former are simple, and of 
 the latter compound. One reason, 
 no doubt, why the Reticulated 
 variety appeared so early was 
 that though all its curves are 
 ogee curves, and therefore com- 
 pound, yet the)' can be set out 
 in the simplest possible way. For 
 example (4S7), in such a window 
 as that of Merton vestry, if the 
 base be taken as a radius, the 
 window-arches are obtained. 
 Then on the base set three semi- 
 circles, and let these be made to 
 support four circles, and the four 
 to support three circles, and the 
 three to support two circles, all 
 circles of the same size ; then 
 Reticulated tracery results at once. 
 Even quite complicated tracery 
 is easy to draw, if you know how ; 
 e.g. the circular windows at WAL- 
 THAM ABBEY (488) : to obtain i 
 centres for the curves, all that is g 
 necessary is to construct an 
 octagon with long and short sides 
 alternating ; each long side twice 
 as long as each short one. So 
 again in the windows at OLD 
 WALSINGHAM (480.5), and the 
 door at HOLBEACH ('583), a 
 pretty pattern common in East 
 Anglia is got by intersecting 
 octagons. But, as time went 
 on, in the subtle curves of Flam- 
 boyant, as in the Beverley reredos 
 and the west window of SNETTIS- 
 IIAM (481), the draughtsman 
 perhaps trusted to eye and wrist 
 more than to rule and compass. 
 
 Hybrid Windows. — Geo - 
 metrical and Curvilinear patterns are 
 
 Westminster Cloister. 
 Hingham S. Aisle. 
 
 often mixed ; unskilfully, as in the case 
 
490 CURVILINEAR TRACERY. 
 
 of the five circles in the Carlisle window, and the six in the Beverley reredos ; 
 or skilfully, as in the window at HINGHAM (489) and the south transept of 
 Hull* In both these the skeleton of the window is purely Geometrical ; and 
 the Geometrical patterns are filled up with flowing lines : but the two elements 
 are kept entirely distinct, and are on different planes of tracery. If the sub- 
 arches consist of two ogee arches, the centrepiece may consist of a great circle ; 
 as at MINCHINHAMPTON (286), in TEWKESBURY CHOIR (165), Nantwich and 
 Exeter,-)- and in a great window of six lights at Chipping Norton ; the 
 inner upper curves of the ogee arches forming part of the circumference of 
 the circle. An unhappy jumble of Geometrical and Curvilinear tracery is seen 
 at Mildenhall,} the east window of Bristol Cathedral, and the south transept 
 windows of St Mary Redcliffe. 
 
 ALTERNATION. — Both in the Geometrical and Curvilinear periods a range of 
 windows in aisle or clerestory was usually of one uniform design ; e.g. in Lincoln 
 presbytery, Beverley nave ; SELBY CHOIR (86). Sometimes, however, while 
 uniform in dimensions and in the number of their lower lights, the tracery was 
 diversified. This is especially the case at Exeter, where the windows differ 
 all the way down the church, though each corresponds to a window on the 
 opposite side of the church ;§ passing from Late Geometrical tracery in the 
 east choir to assimilated Curvilinear in the western nave. Instances of alternation 
 occur also in purely Curvilinear work ; e.g. the four windows on the side of 
 Patrington chancel are of two patterns. The clerestory windows of Tewkesbury 
 are much diversified. All four in the Latin Chapel of Oxford Cathedral are 
 different. So also in the clerestory of Norwich choir windows of Perpendicular 
 alternate with windows of Flowing design. 
 
 Size of Windows. — The side windows of the Gothic cathedrals soon grew 
 to their full dimensions, practically occupying the whole space from buttress to 
 buttress ; e.g. in Lincoln presbytery, c. 1260, and Exeter choir, c. 12S0. Perhaps 
 the greatest amplitude was given to the window first in the chapter houses, e.g. 
 Westminster and Salisbury. In the latter, as in Amiens nave and in most of 
 the Gothic of the He de France, the wall arch of the vault and the rear arch of 
 the windows are fused together ; and Gothic construction is carried out to its 
 logical issue ; the vault plainly resting not on walls but on detached piers. 
 Fortunately we were not often so ultra-logical ; exceedingly unsatisfactory 
 within and without is such an excessive attenuation of the supports ; indeed 
 the preponderance of the voids over the solids is the crying sin of Late Gothic 
 architecture. Equally objectionable is the failure to provide the beautiful 
 window tracery with a foil of blank stone around it. How disastrous it would 
 be to the design of SELBY CHOIR (86) to extend the windows from buttress to 
 buttress ! 
 
 * Illustrated in Sharpe's Windows, Plate 29. 
 
 t Sharpe's Windows, Plates 48 and 36. 
 
 X Illustrated in Paley's Gothic Architecture, 178. 
 
 $ For twelve of these windows see Britton's Exeter, Plate 12. 
 
Chapter XXXIV. 
 RECTILINEAR TRACERY. 
 
 Reasonableness of Rectilinear Tracery — Transitional Types — Premature Rectilinear 
 and Belated Curvilinear Tracery — Characteristics of Rectilinear Tracery. 
 
 From the graceful tangles of flowing tracery we pass to the Rectilinear 
 "gridiron"; from poetic fancy to plain prose. It is an extraordinarily sudden 
 and complete revolution in the formation of window tracery ; it is, moreover, 
 one into which none of the Continental Gothic styles passed, but is wholly 
 English. What brought it about? It has usually been held sufficient to say- 
 that in later English art a degradation and debasement in taste set in ; and 
 that this is its most conspicuous outward and visible sign. We humbly venture 
 to doubt it. Who that is an Englishman can regard the parish churches of East 
 Anglia, the choirs of Gloucester, York, and Norwich, the nave of Winchester, 
 the three Royal Chapels, but as a most precious heritage of mediaeval art, to 
 be treasured among the greatest artistic triumphs of our race? Yet in all these 
 beauty of window tracery is almost wholly sacrificed and abandoned. Never- 
 theless it was not sacrificed and abandoned without a very sufficient reason. 
 Beauty of tracery was sacrificed for something which in the later days of Gothic 
 was treasured more and more, and for which there was nothing which clerk and 
 craftsman were not content and glad to sacrifice. And that was Stained Glass. 
 Ever will our later English art be misunderstood and unappreciated so long 
 as we refuse to visualise to ourselves what was that art's ideal. That ideal 
 was to build no more the House of God in stone, but to transmute the stone 
 into glass ; to rear to the honour of God a spiritual house of coloured light ; 
 fusing the very window jambs and mullions in the spreading effulgence 
 radiating from the multi-coloured stainings of the glass. To get the glory of 
 the lantern church there was nothing that our later builders were unwilling to 
 sacrifice ; and the first thing to go was Flowing tracery. 
 
 It was in Gloucester Abbey that this great artistic revolution was initiated 
 and consummated. At Gloucester the glass,* the window tracery, the vault, all 
 received a special English character which was to differentiate them for ever 
 from the art of the Continent. 
 
 First, a great change was made in the glass.+ It had been thick and 
 
 * Winston in his memoir on the east window of the choir minimises the importance of the 
 change made in glass at Gloucester ; as is pointed out by Mr Lewis Day, p. 17S. 
 
 t From the middle of the fourteenth to the later years of the fifteenth century the tendency 
 was more and more in the direction of light ; "until our later Gothic windows become, in many 
 
49: 
 
 £i.iii£ 
 
 Gloucester Choir. 
 Oxford Divinity School. 
 
 St George's, Windsor, Nave. 
 Sail, East Window. 
 
RECTILINEAR TRACERY. 493 
 
 opaque; now it was made thin and translucent. It had been a mass ofcolourj; 
 now colour was reduced to a minimum ; a splash of rub}- here, of azure there, 
 a little purple or green, some golden stain, that was all ; the greater part of the 
 glass was uncoloured ; a silvery white. The pictures too were harmonised 
 with their architectural home; each figure was set in a tall canopy of silver 
 glass, itself fitted into a stone niche of tracery. It is only necessary to compare 
 the heavily coloured glass in the clerestory of the choir of Tewkesbury with 
 the contemporary glass in the clerestory and east window of the neighbouring 
 choir of Gloucester to realise how great the change was. It became possible 
 now to use glass with the utmost freedom ; it no longer dimmed and obscured. 
 The light of Gloucester choir was as far from the dim, religious gloom that 
 the modern ecclesiologist adores as from the light of common day.* 
 
 For the new glass there seems to have been a consuming enthusiasm. 
 Lincoln, Ripon, Guisborough had led the way with great windows in their 
 eastern fronts. In Carlisle and York we have seen the windows grow yet 
 further in area. But Gloucester outdid all ; she took out all her four 
 end walls; first that of the SOUTH TRANSEPT (495), then that of the 
 CHOIR (135), then that of the north transept, then that of the nave; and 
 replaced all four by glass. This began while everybody else in England was 
 putting in Flowing tracery ; and indeed when Gloucester herself was erecting 
 the monument, of thoroughly fourteenth -century type, to Edward II., which 
 still stands in her choir. For the remodelling of the south transept took place 
 between 1330 and 1337; and the vaulting of the choir and the glazing of its 
 great east window seem to have been finished by 1350.! 
 
 But certain corollaries were inevitable. It is not possible to construct a 
 window arch of the vast height and breadth of the east window of c.LOUCESTER 
 ('496; — one, moreover, that has to take its share in supporting the weight of the 
 
 instances, not so much coloured windows as windows of white and stain enclosing panels or 
 pictures in colour. Even in these pictures very often not more than one-third of the glass was 
 in rich colour. And not only was more white glass used ; but the white itself was purer and 
 more silvery, lighter, and at the same time thinner" (p. 53). "The gradual dilution of the deep, 
 rich, early colour is noticeable throughout the fourteenth century. Towards its close the glass 
 craftsman halts no longer between two opinions, between light and colour. He conceives his 
 window pretty generally as a field of white, into which to introduce a certain amount of rich 
 colour; not often a very large amount. As a rule, perhaps, not more than one-fourth of the 
 area of a fifteenth-century window was colour ; for, in addition to the white of the canopy, there 
 was commonly a fair amount of white in the draperies ; and the flesh was now always represented 
 by white. The typical Perpendicular window, then, is filled with shrine work in white ; enclos- 
 ing figures, or figure subjects, into which white enters largely (the flesh, and some of the drapery, 
 often a good deal, is sure to be white), upon a background of colour" (p. 1S1). Windows, a 
 Book about SI, lined .aid Painted Glass, by Lewis F. Day. Later, however, eg. in the windows 
 of Fairford Church, rebuilt c. 1490, and in those of King's College, Cambridge, contracted for 
 in I5i6and 1526, the English tradition disappears, probably owing to Flemish influence ; an 
 immense amount of colour is used ; enamelling is resorted to ; the windows become opaque once 
 more, and the interiors dark. See the photograph of King's College Chapel ; it was taken in 
 bright sunshine at mid-day in August. The invasion of Flemish glass is seen just as much in 
 France ; e.g. at Brou-en-Bresse. 
 
 * To see the new glass at its best, one should visit the ante-chapel of All Souls' College, 
 Oxford, and St Martin's, York. 
 
 t See Willis' paper on Gloucester, and Winston on the east window. 
 
494 RECTILINEAR TRACERY. 
 
 vault and the roof — without taking precautions to ensure its stability, in the 
 first place, lest the jambs bulge in, horizontal bars (transoms) have to be 
 inserted from side to side. Secondly, instead of the tracery being independent 
 of the window arch, as it is in the CARLISLE WINDOW (507), it must be so 
 designed as to support the window arch, so that it may not be crushed in. 
 For such support Flowing tracery, with its flickering ogee curves, is inadequate. 
 The mullions, instead of curving, must be straight ; as many of them as 
 possible must rise straight up to the window arch. Nay, even secondary 
 mullions (supermullions) must rise from the apex of every arch in the tracery, 
 as in the window of the DIVINITY SCHOOL, OXFORD (492), to give yet addi- 
 tional vertical support. If that is not enough, secondary transoms may be 
 introduced even into the tracery of the window head. The result is, a stone 
 gridiron. But the stone gridiron is inevitable if, instead of the eight lights 
 of York or the nine of Carlisle, Gloucester insists on having fourteen, ST 
 GEORGE'S, WINDSOR (492), fifteen lower lights. 
 
 And whatever the artistic demerits of a gridiron, one may venture to 
 imagine that it would not be without practical merits in the eyes of the 
 draughtsman and the mason. How easy to draw ! And how easy to execute, 
 with its frequent repetitions of identical parts ! Think of the Carlisle window, 
 the curves of which required the plotting out of 263 different circles ; and of 
 its jointing, so complicated and artificial ; while in the Gloucester and Windsor 
 windows nearly all the joints are horizontal or vertical ! Such a window was 
 as easy to execute as it was to design. No doubt it appealed powerfully to the 
 practical mind of the Englishman that so very much more window could now be 
 had for the same money. 
 
 But what of the glassman, now most important of craftsmen ? He must 
 simply have been delighted with Gloucester's new east window. Instead of the 
 circles and triangles of Geometrical tracery, and the writhing quatrefoils and 
 ogees of Curvilinear, he had got now precisely what was wanted ; for every 
 little figure in his stained glass there was a little rectangular niche. And here 
 we may perhaps believe the new type of tracery reacted on the design of 
 the glass. Rectilinear tracery lends itself readily to tier 1 upon tier of priest 
 and prophet and king ; martyr, saint, and angel, one in each niche ; it does not 
 lend itself so well to story-telling, to picture-groups. And so design in glass 
 grew much more simple ; very much for the better of the art ; for only the 
 very simplest pictorial treatment is right in glass ; glass should not try to be 
 canvas. And so our best late windows, e.g. those at Gloucester and All Souls', 
 become not as it were national picture galleries, but national portrait galleries.* 
 And how much better the portraits are hung in such windows than in the 
 Flowing windows ! Imagine a portrait gallery with half the portraits hanging 
 askew ; some to the right, some to the left ; a pain to the eyes. So it was 
 with pictured saint and martyr ; in the west window of York the)- flopped some 
 to right, some to left. All this indecorum was subdued by putting them in 
 rectangular niches, row upon row, each standing upright decorously on his 
 
 * It is not to be denied, however, that the picture-group still found admirers ; e.g. the 
 windows of Malvern, the east window of York Minster, and several windows in the parish 
 churches of York. 
 
495 
 
 Gloucester, South Transept and Crossing 
 
GLOUCESTER CATHEDRAL, THE PRESBYTERY. 
 
RECTILINEAR TRACERY. 497 
 
 feet* Such an arrangement could not but find approval with canon, monk, 
 craftsman, and devout layman alike. 
 
 May we not even go further, and argue that Geometrical and Flowing 
 tracen- was hardly worth the trouble and money it cost? The windows of 
 Lincoln presbytery and Beverley nave are beautiful indeed from without. But 
 are they beautiful from within ? Mr Ruskin has drawnf specimens of Geometrical 
 tracer}-, illustrating how exquisite is its beauty. So it is. But it is only under 
 the conditions shown in Mr Ruskin's drawing that this or an}' other tracery is 
 beautiful. Those conditions are that it be silhouetted either against a dark back- 
 ground or against the sky. Standing outside a church before its windows are 
 glazed, the white stone of the tracery is projected against a dark background. 
 See how beautiful under such a condition are the cloister windows of SALISBURY 
 and NORWICH (458, 506). Even Rectilinear tracery looks lovely against the 
 darkness. But if we stand inside a church, the tracery is only projected against 
 a bright sky when the church is a ruin or when it is building ; i.e. when 
 the glass has been knocked out of the window or before it has been put in. 
 What can be more beautiful than the tracer}-, albeit Rectilinear, of the ruined east 
 window of Melrose, or than that of St Lawrence, Evesham, as it was when Britton 
 drew it?^ There are indeed three distinct stages in the inner history of the tracer}' 
 of every window. It looks its best before it is glazed at all ; it retains some of its 
 beauty when glazed in transparent white glass ; it loses it almost wholly when 
 filled with stained glass, and viewed from within. If that be so § — it is only 
 necessary to stand within a church with stained windows and good tracery to 
 test the truth of it — may we not say that the old men were not altogether 
 ill-advised to concentrate their attention on the inside rather than on the outside 
 of the churches ; not on the tracer}- which was effective externally, but ineffective 
 within, but on the stained glass ; which, ineffective externally, was, within, a sea 
 of glory ? 
 
 GLOUCESTER (59) did not confine her innovations to her end walls. She 
 took down also the Norman clerestory of her choir, with its ineffective, small, 
 isolated single windows, and replaced it by a towering clerestory that was 
 practically one continuous sheet of glass: a clerestory that is to be regarded 
 not so much as a series of independent windows, as a single choir-long con- 
 tinuous window with piers for mullions. The precedent of the Gloucester 
 clerestory was followed soon after, timidly, at WINCHESTER and CANTER- 
 BURY (90); boldly at York, Malvern, Norwich. And what Gloucester had 
 done with its clerestory, other churches hastened to do with their aisles ; these 
 also soon became sheets of glass. So that the end walls were glass, the clerestory 
 walls glass, the aisle walls glass. 
 
 It was not, however, for a whole generation that the new art came to its 
 throne. While Gloucester choir was building, the people at Tewkesbury, 
 Beverley, Carlisle, El}', Lichfield, Malmesbury, St David's, Selby, Wells, 
 Worcester, developed into ever-growing splendour fourteenth-century art. But 
 
 * Equally indecorous is the disposition of the saints round the archivolts of the great 
 Erench doorways. 
 
 t Seven Lamps, Plate 3, p. 105. i Architectural Antiquities, v. Plate 68. 
 
 § Mr Ruskin elsewhere admits that "glass spoils all traceries " {Seven Lamps, 170). 
 
 2 I 
 
493 
 
 Carlisle Choir Aisle. 
 
RECTILINEAR TRACERY. 499 
 
 in 1349 came the Black Death: after that there was neither the money noi the 
 heart for another Ely Lad}- Chapel and Kly choir. The simplicity of Glouce tei 
 stone-craft, the superiority of Gloucester glass, made their appeal at the right 
 moment to a sobered England. The new art spread like wildfire through the 
 length and breadth of England ; for the future the architecture of England 
 was to be conditioned by glass and not by stone. Cathedral church and abbey 
 church were content to sit at the feet of Gloucester. Above all, Gloucester set 
 the fashion to all the \a.ter parish churches of England ; in particular to Norfolk 
 and Suffolk * and the glorious churches of the Fenland. 
 
 In 1327 Edward II. had been murdered at Berkeley. The Bristol 
 Augustinians dared not risk the animosity of the Court party by giving harbour 
 to his corpse. Brave Abbot Thokey conveyed it to Gloucester and buried it in 
 the old Norman choir. Miracles were wrought ; pilgrims came by thousands 
 from every part of the realm ; so vast were their offerings that the Gloucester 
 monks, so the chronicler tells us, might have rebuilt, had they willed, the whole 
 abbey church. They contented themselves at first with remodelling the south 
 transept and the choir. The works were commenced soon after 1330. Round 
 and above the new shrine of the King and Saint the works proceeded, till c. 1350 
 the vaults were up, and the windows glazed. The scaffoldings were removed, 
 and the world saw at last the revelation of a new art world ; the glories of 
 Gloucester choir. And so it was that Gloucester choir, being in the fourteenth 
 century so great a centre of English Christianity, was visited and inspected 
 more than any other church in the country ; and when the new glass and the 
 new vaults were at length revealed to view, the fame of them passed at once to 
 the furthest ends of the land. Every pilgrim's tale, c. 1350, would be of the new 
 vision in English art, the choir of Gloucester. 
 
 Transitional Windows. — It was not everywhere, however, that the con- 
 ditions of the new style were accepted without demur, as they were in the east 
 window of Gloucester and the west window of Winchester ; where the tracer)- is 
 almost wholly Rectilinear ; i.e. composed of two sets of straight lines, vertical 
 and horizontal, perpendicular to one another. In a considerable number of 
 the earlier windows, the greater part of the tracery is still composed of curves. 
 A charming example is seen at Houghton-le-Dale ; t and at St Michael's, Cam- 
 bridge.^: Other well-known examples are the east windows of the choirs of 
 Bristol^ and wells (127). Perhaps the finest example is the east window 
 of the fourteenth-century chancel of HULL (81). At first sight, with its two 
 great intersecting ogee arches and ogeed lower lights, it looks as thoroughly 
 Curvilinear as its rival, the west window of York. But on closer inspection it 
 will be seen that the vertical lines are there in abundance ; and that in many 
 cases these spring from the apices of arches, which are then said to be super- 
 mullioned. (The transoms are probably not original. J In the NAVE (96), 
 
 * At LONG MELFORD (547) there are no less than 74 traceried windows in the church 
 proper. Including the chapels, the vestries and the porch, but excluding the Lady Chapel, 
 there are 07 (Lauriston Conder's monograph, 26). 
 
 t Illustrated in Rickman, 223. I Illustrated in Freeman, 209. 
 
 § A five-light geometrical window with its sub-arches supermullioned is illustrated in 
 Freeman, 82, 94. 
 
Soo CURVILINEAR AND RECTILINEAR TRACERY. 
 
 which is not much later and is but a Rectilinear version of the chancel, the 
 principles of Rectilinear tracery are fully accepted, except that there are no 
 transoms in the bottom lights. 
 
 Juxtaposition of Curvilinear and Rectilinear Tracery. — We 
 now come to three puzzling sets of windows which appear to break the rules of 
 chronological sequence that have been set forth hitherto. The first set comprises 
 those which occur in a fourteenth-century building, the construction of which 
 was postponed ; so that when they were put up, a change of style had come 
 about in window tracery, and the new window was given Rectilinear tracery. 
 A curious early example of this is the east window of Evington, which has 
 supermullions and transoms in the tracery ; * i.e. it is a Rectilinear window ; 
 nevertheless the windows near it are of early Curvilinear design. It will be 
 found that a large number of Rectilinear east windows, e.g. that of PAT- 
 RINGTON (133), and some west windows also, e.g. in Ely Lady Chapel, occur in 
 buildings of undoubtedly Curvilinear date. It would seem that the end windows 
 were sometimes built later than the side windows.^ We may suggest that in 
 these cases, when all was complete except the end window, the Black Death 
 may have arrived, 1349,+ and stopped the works: and that when they were 
 resumed the style had changed, and the window was filled in with Rectilinear 
 tracery. Whether this be the true explanation or not, there must be some 
 reason for the common occurrence in end windows of Rectilinear tracery in 
 jambs and arches which have Curvilinear moldings ; as at Patrington. So also 
 in the fine Curvilinear church of Nantwich§ the end windows of the chancel and 
 of one transept are Rectilinear. 
 
 A second set comprises examples of what we may call PREMATURE 
 Rectilinear Tracery.[j Though Mr Sharpe gives c. 1360 as the commencing 
 date for Rectilinear tracery, we have seen that it was evolved in the south 
 transept of Gloucester between 1330 and 1337, and fully developed in the choir 
 by c. 1350. At Gloucester, therefore, Rectilinear tracery did not follow on 
 Curvilinear tracery. The Rectilinear tracery of the south transept followed 
 on the heels of the Late Geometrical windows of the south aisle of the nave ; 
 built by Abbot Thokey, 13 18 to 1329. At Gloucester^ the Curvilinear period 
 of tracery is largely missing. What was done at Gloucester would certainly 
 be copied here and there elsewhere ; e.g. in Cheltenham Church, where win- 
 dows Rectilinear and Curvilinear in character occur side by side, and where 
 the former have been erroneously supposed to be some thirty years later 
 than the latter. And no doubt the same was the case elsewhere. For in the 
 fourteenth century Gloucester was the greatest centre of pilgrimage in England ; 
 and its architectural doings could not fail to be reported far and wide, and here 
 and there to be copied. In cases, therefore, where we see a Rectilinear and 
 
 * Illustrated in Brandon's Analysis, 252. 
 
 + Perhaps the end wall was left open for the removal of scaffolding, &c. 
 
 I Ely Lady Chapel is said to have been finished in 1349 ; but the east window was not 
 inserted till 1373 ; the west window later still (Stubbs' Ely, 147). 
 
 § Illustrated in Bowman and Crowther. 
 
 || A five-light window at the east end of the south aisle of Warmington has no tracery ; the 
 mullions rise straight up to the head. Its date is not later than c. 12S0. — R. P. B. 
 
 IT Beautiful Flowing tracery occurs, however, in the upper aisle of the choir. 
 
CURVILINEAR AND RECTILINEAR TRACERY. 
 
 Vil 
 
 a Curvilinear window side by side, they may be absolutely contemporaneous; 
 
 the former having been executed between 1315 and 1360 on the lines of the 
 work at Gloucester. 
 
 But there is an important and very large set of windows which are to be 
 explained on other lines, and which we may call BELATED CURVILINEAR. 
 It is usually assumed that c. 1360 all the world unanimously set to work 
 to produce Rectilinear window tracery. This was certainly not so. In 
 some districts, especially in those far distant from Gloucester, Flowing tracery 
 held its own to the end of the fourteenth and probably well into the 
 fifteenth century. Not everybody was willing to sacrifice the exquisite 
 curves of Flowing tracery for the giassman's gridiron. This was especially 
 the case in Norfolk and Suffolk ; which at that time were hardly an integral 
 part of England ; but severed from the mainland by rivers and fens more com- 
 pletely than Ireland is now from England. Fast Anglia was an island, with 
 insular and independent architectural traditions. There fourteenth-century grace 
 found its last refuge. Thus at El}', Bishop Barnet, 1367 to 1373, made three 
 windows on the south side and two on the north side of the presbytery ; * 
 and these are of Flowing design.+ So in Wimmington, Bedford, the brass of the 
 founder gives 1391 as the date of his death; but " the details are remarkably 
 pure Curvilinear."^ So also the beautiful tomb in Durham Cathedral of Bishop 
 Hatfield, who died in 1 38 1 , is of fine Curvilinear design. So also the Church of 
 Ftchingham, Sussex, built in 13S6, is mainly Curvilinear. § In East Anglia the 
 retention of Curvilinear work is so common as to be characteristic. Thus in 
 North Walsham Church, which was rebuilt after the destruction caused by the 
 Norfolk rising of 1381, the east windows are thoroughly Curvilinear; and in 
 the porch, which is probably later still, Curvilinear and Perpendicular windows 
 alternate. So they do in the clerestory of the choir of Norwich Cathedral, 1361. 
 The two Wiggenhalls, St Alary Magdalen, and St Mar)* the Virgin, are 
 Curvilinear churches; so also is W'alpole St Peter's; but all three have 
 Perpendicular tracery in their windows. So also the windows of Terrington 
 St Clement's have Rectilinear tracery ; but the piers and arches are Curvi- 
 linear ; it seems unlikely that the arcade was built in the fourteenth century 
 without a clerestory ; or that, if it had a clerestory, a new one was substituted 
 for it in the fifteenth century. The same combination of Curvilinear arcade with 
 Rectilinear windows is seen at Bacton " and Fresingfield, Suffolk; and Wor- 
 stead, Norfolk. At Worstead, the hammerbeam roof and clerestory seem to 
 
 * Anglia Sacra, i. 664. 
 
 t In nearly all the aisle windows of Ely presbytery the masons copied the tracery without 
 disguise from that which they found in one of the windows of Bishop Hotham's time, 1316-1337 
 (Stewart's Ely, 122). 
 
 I Brandon's Parish Churches, 93. 
 
 § In the Battlefield Church, built after the battle of Shrewsbury in 1406, three of the 
 windows have Curvilinear tracery Mr W. A. Webb). 
 
 [| They are illustrated on pages 101, 107 of Parish Churches, by Mr Brandon, who 
 assumes that the W'alpole clerestory was originally Curvilinear, but rebuilt in Rectilinear 
 style. He acknowledges, however, that it is very improbable that a fourteenth - century 
 clerestory can have become ruinous in so short a time, so as to require renewal. 
 IT Illustrated in Brandon's Parish Churches, 124, 11S, 36. 
 
S02 
 
 CURVILINEAR AND RECTILINEAR TRACERY. 
 
 have been built in 1378 with timber granted by the Prior of Norwich ; here 
 also it is easier to suppose that the Curvilinear arcade was built at the same 
 time than to imagine that the arcade was put up c. 1 340 and roofed ; and that 
 in 1378 this roof, which would still be sound, was pulled down and replaced by 
 the present hammerbeam roof. So also at Rickenhall, Suffolk, a Rectilinear 
 east window is " identical in date and workmanship " * with the other windows, 
 
 IhWP^I 
 
 4. Ashby St Leger. 
 
 which are of early fourteenth-century type : it seems more natural to accept 
 the latter as Belated than with Mr Brandon to regard the former as Premature. 
 For anticipations of style are rare, and require to be supported by the most 
 positive documentary evidence ; but survivals must always be a common 
 result of the conservatism, indolence, and obstinacy of the average human 
 
 Brandon's Parish Churches, 45. 
 
RECTILINEAR TRACERY. -,,. 
 
 mind.* The existence of " survivals " rises to a certainty in such an example as 
 ROXWELL CHURCH (502), near Chelmsford. Here there are three windows, 
 of which one would say that 1 is Curvilinear, 2 Transitional to Rectilinear, 
 3 Rectilinear. But the stones of all three windows come from the same layer 
 in the same quarry. Also they were wrought at the same time ; for if 3 were as 
 much later than 1 as it appears to be, some sixty years, the stones in it would 
 be in a higher state of preservation than those of 1 ; but after a minute 
 examination of every stone, both inside and out, no difference can be detected. + 
 
 Characteristics of Rectilinear Tracery.— As has been pointed 
 out above, transoms were a necessity in windows of the vast span of the east 
 window of Gloucester. In this they occur in the lower lights. But they occur 
 also in the tracery, e.g. SALE (492), HULL (96), and CARLISLE (498) ; this is 
 especially common in later work. At first the transom was a mere molded 
 horizontal brace; but it was soon ornamented; in later work usually with 
 miniature battlements or a cresting of Tudor flower. Or it was cusped in a 
 very beautiful way, as in the ruined east window of Melrose and in Prince 
 Arthur's chantry at Worcester and at ASHBY ST LEGER (502). 
 
 Transoms had been in common use in domestic work from the thirteenth 
 century, e.g. in the halls of the Bishops of Lincoln and Wells, and in the 
 Winchester Castle Hall. The object of them was to enable a casement to be 
 inserted in that part of the lower light which was below the transoms. So 
 also in a chancel the lower part or the whole of a side window was sometimes 
 transomed, in order to permit the insertion of a shutter. 
 
 Transoms were frequently inserted later to strengthen the long mullions of 
 earlier windows; e.g. the west window of HOWDEN ("2) and the east window 
 of HULL (81). 
 
 Another characteristic of Rectilinear windows is that the heads of the 
 lower lights are almost always cinquefoiled ; whereas in earlier windows the 
 trefoiled head was common. 
 
 The quatrefoil also, which in the thirteenth century had had four equal 
 arms, was to go through one more metamorphosis. In Curvilinear tracery it 
 had had to take the form of a curved scimitar ; in Rectilinear work the blade- 
 is straightened, and we get the form of the dagger.% Two examples of this are 
 seen in the window from Ashby St Leger. In an early and simple form four 
 examples occur in the west window of Hull ; cf. William of Wykeham's work 
 at Winchester ; Edington Church ; Headcorn Church ; Merton College Chapel, 
 and St Mary's Church, Oxford.§ 
 
 As has been pointed out above, the primary motive of the design of 
 Rectilinear windows was the provision of suitable accommodation for stained 
 glass: beauty of tracery had become but secondary. Nevertheless artistic 
 instinct did not cease to work ; and in spite of the limitations imposed by the 
 
 * Flowing tracery is particularly common in windows of fifteenth-century towers. I >n the 
 retention of fourteenth-century design in East Anglia, see Mr J. L. Andre in Archaological 
 Journal, xlvi. 377, 389. 
 
 + Hadfield, 4. 
 
 % But this straight dagger appears also in late Geometrical tracery, e.g. SOLIHULL (620). 
 
 S Glossary, Plates 252, 253, 254. 
 
5 o4 RECTILINEAR TRACERY. 
 
 dominance of Rectilinear patterns, many of the windows have a distinct charm 
 of their own. Among the most successful may be mentioned the west windows 
 of BEVERLEY ST MARY (366), Shrewsbury Abbey, and HULL (96), all early 
 examples, and reminiscent of the great traditions of fourteenth-century design, 
 and of the north porches of Spalding and HEREFORD CATHEDRAL (203) ; both 
 late work. The most disastrous change was the multiplication of the transom 
 in the lower lights ; which is as common in large early windows such as the 
 east window of Gloucester and the west window of Winchester as in large late 
 examples such as St George's, Windsor, and Henry the Seventh's Chapel, 
 Westminster. It is only necessary to turn to the west window of Hull nave 
 to see how much a window gains by the absence of transoms in the lower lights. 
 Again, when arches occur in the tracery, they should not be cut through by 
 straight lines. This occurs in the east window of SALL (492) ; but it is 
 carefully avoided at Hull, and also in Sail transept.* The fewer curves there 
 are in a Rectilinear window, the less is the risk of awkward collisions between 
 the straight line and curve. Another objectionable practice, especially common 
 in Tudor work, was the introduction of geometrical patterns in the middle of 
 the rectilinear bars. It is seen at its worst, perhaps, in the lower windows of 
 KING'S COLLEGE CHAPEL, CAMBRIDGE (473). 
 
 * Note in the divinity school, oxford (492), how well adapted for stained glass figures 
 are the batement lights between the mullions, supermullions, and jambs ; cf. the Hull window ; 
 also that the tracery does not commence at the spring of the arch, as it usually does in 
 Geometrical and Curvilinear windows, and in the Hull window, but considerably below it : 
 this is almost unavoidable when the window is set under a depressed four-centred arch. To 
 this is given the name of Drop Tracery. 
 
Chapter XXXV. 
 
 WINDOW CONSTRUCTION. 
 
 Functions of Tracery — Planes of Tracery — Cusping and Foliation — Scoinson Arch — 
 Inner Arcade — Mullions — Circular Windows — Sound Holes — Low Side Windows. 
 
 So far we have traced the evolution of Gothic tracery from combinations of 
 lancets to the intermixture of rectilinear and geometrical patterns in Tudor 
 days. It remains still to deal with the functions and planes of tracery, cusping, 
 the rear-arch, moldings, and circular windows. 
 
 FUNCTIONS OF TRACERY. — The functions of tracery are partly construc- 
 tional, partly decorative. Its chief constructional function is, with the aid of 
 the iron bars fixed in the sills and jambs, to enable the glass to withstand 
 wind pressure. In early work the weight of superincumbent wall is carried 
 by the window-arch without the aid of the tracery ; which indeed in early 
 windows almost invariably was constructed only after the window-arch was 
 complete,* and might be removed and has often been removed without en- 
 dangering the stability of the arch. When, however, the arch is of a weak 
 form, e.g. when it is an ogee arch, the construction would be unsafe but for 
 the presence of the tracery. In large Rectilinear windows also, as has been 
 pointed out in 494, the mullions have real constructional value in streng- 
 thening the window-arch. In these the main mullions almost become piers, as 
 in the east windows of Gloucester, Sherborne, Warwick, and the west window 
 of Leominster ; and are sometimes strengthened outside by a buttress. But 
 even when it is not employed constructionally, the eye desiderates that the 
 tracery shall at any rate appear adequate to support the arch. This is all the 
 more so because from within the mullions are thinned to a surprising extent 
 by the irradiation of the light. Hence under no circumstances is attenuated, 
 wiry tracery, such as that of the east window of Shottesbrooke, endurable. 
 
 Praxes of Tracery. — But because it should be massive, it does not 
 follow that the whole of the tracery of the window need be equally massive. 
 One of the most beautiful features of the more highly developed mediaeval 
 windows is the alternation of tracer}- of two, three, four, or even five + thick- 
 nesses, to which the name of orders has been given. It is necessary that leg, 
 arm, and backbone should be thick ; but not toes, fingers, ribs. So in the 
 window of NORWICH CLOISTER (506) the main skeleton is carefully differenti- 
 ated from the minor work, the filling in of quatrefoils. The principle of sub- 
 ordination and recessing of orders had been worked out long ago, even in 
 Norman days, in the arch (272) ; it produced equally beautiful results when 
 
 * Sharpe's Windows, 39. t I.e. if the orders of the foliation be included. 
 
506 
 
 WINDOW CONSTRUCTION. 
 
 applied to window tracery.* The west window of HOWDEN (72) may be taken 
 as a characteristic example of subordination of orders. In this the window- 
 arch, the tracery bars, and the centrepiece each contain two orders of moldings ; 
 the foliation also contains two orders, the first of which is identical with the 
 second order of the tracery, and the second of which is formed by soffit-cusps. 
 Still more elaborate is the great east window at CARLISLE (128). The choir 
 of Carlisle Cathedral was lengthened eastward after the great fire of 1292. The 
 jamb shafts and the moldings of the jambs and the window-arch are of earlier 
 date than the tracer)'; the former, therefore, may be c. 1300, the latter con- 
 siderably later ; indeed the tracery was not glazed till 1380- 1384, as the arms 
 remaining in the glass show. Both this and the west window of YORK (82) 
 
 Norwich Cloister. 
 
 Ely Transept. 
 
 are 26 feet wide in the clear ; the Carlisle window is 5 1 feet high from the sill 
 to the top of the tracery, while the York window is 2 or 3 feet lower ; on the 
 other hand the tracery of the York window is more than 2 feet higher than that 
 of Carlisle. It is a curious fact that in the Carlisle window more than half of the 
 internal tracery was left unfinished. As is seen in 507, the tracery consists of 
 86 pieces ; some of these, e.g. 20 to 26, are very large, being between 4 and 5 
 feet long. They are beautifully jointed, and are arranged in such a way that 
 any single stone can be removed for repairs without endangering the stability of 
 
 It should be noted that the nomenclature is reversed. In an arch the innermost and 
 narrowest order is spoken of as the first order ; in tracery the outermost and thickest order is 
 the first. Also it should be remembered that the orders of the window-arch are one thing, and 
 the orders of the tracery another ; students should describe the two sets of orders separately. 
 
WINDOW CONSTRUCTION. 
 
 507 
 
 the tracery, except 
 such as 1, 7, 26 to 
 20. The tracer) - , un- 
 like that of York west 
 window, is nowhere 
 supported by the 
 window-arch. 
 
 Fig. 2 s h o w s 
 the disposition of the 
 three orders of 
 tracery. The princi- 
 pal or outer order is 
 confined to the 
 window-arch, the sides 
 of the sub-arches and 
 the centrepiece, and 
 the head of the circle 
 in the lower part of 
 the centrepiece. The 
 middle order com- 
 prises all the lines in 
 the centrepiece, ex- 
 cept the head of the 
 circle below and the 
 three openings in the 
 apex of the arch ; 
 also all the lines 
 shown in the left sub- 
 arch, and the corre- 
 sponding lines in the 
 right sub-arch. The 
 inner order comprises 
 the three openings in 
 the apex of the arch 
 at the top of the 
 centrepiece; the 
 tracery in the span- 
 drels ; and the whole 
 of the tracery shown 
 in the right sub-arch 
 except that which 
 has been specified 
 as being of the 
 middle order; also 
 the corresponding 
 lines in the left sub- 
 arch. 
 
 Carlisle East Window. 
 
So8 
 
 WINDOW CONSTRUCTION. 
 
 The curves of the window are struck from 263 centres.* 
 
 In the richer thirteenth and early fourteenth century work not only the 
 jambs but the mullions were often faced with tiny shafts, provided with proper 
 shaft and base ; and these were continued as roll moldings in the curving bars 
 which were the prolongation of the mullions upward. This roll is seen at 
 RUDSTON, and in the aisle of the choir at Selby ; f in the aisle of the nave 
 it is omitted. But charming as this roll molding always is, it is in practice 
 objectionable ; for in horizontal lower portions of it, e.g. of the great circle in 
 
 GUISBOROUGH WINDOW (476), 
 . , ' --'-i- T the rain lodges, and, if frost super- 
 
 venes, serious damage is done. 
 Often, therefore, e.g. at LITTLE 
 ADDINGTON (487. 1), the roll 
 molding is omitted, and the 
 tracery is chamfered so as not 
 to retain the wet. The same 
 change took place in France, j 
 
 Cusping and Foliation. — 
 In such a window as that of 
 RUDSTON, the four lower lights 
 are set under pointed arches. 
 The projecting points high up on 
 either side of these four arches are 
 called cusps, and the arcs on either 
 side of the cusps are called foils 
 (Latin, folia ; French, feuilles).\ 
 In each of these four lower lights 
 there are two cusps and three foils. 
 Each, therefore, is a trifoliated 
 pointed arch. || On page 510 
 there are illustrated both a quatre- 
 foil and a quatrefoiled circle. 
 Each pair of lower lights is in- 
 cluded in a detached pointed arch, in the head of which is a circle containing 
 four cusps and four foils or lobes ; it is therefore a quatrefoliated circle. The 
 two great detached pointed arches carry a circle which has six cusps and 
 six foils, and is therefore called a sexfoliated circle. 
 
 The origin of the foliated or cusped arch has been held to be arcading 
 composed of trefoiled arches."' If a trefoiled arch be included within a pointed 
 
 * Billing's Carlisle, Plates iS, 19 ; text, 59 to 64. t See Plate on page 86. 
 
 % Compare the window in Amiens choir figured in Dehio, Plate 577, 3, and text 2, 580, and 
 that of St Gervais in Choisy's Histoire, ii. 383, 379. 
 
 § What Rickman called foils are really cusps. In Parker's Glossary, also, 157, cusp is 
 erroneously given as the equivalent oifeuille. 
 
 || Such an arch is pointed, and must therefore be distinguished from a trefoiled arch. So 
 also a quatrefoil by itself is nothing but a quatrefoil ; but a circle containing a quatrefoil is a 
 quatrefoiled circle, as in the small circles of the Rudston window. 
 
 IT Willis' Middle Ages, 41 ; Sharpe's Windows, 32 ; Paley's Gothic Architecture, 160. 
 
 Rudston. 
 
WINDOW CONSTRUCT!* >N. 
 
 soy 
 
 arch, a trifoliated pointed arch is producer]. But there is no evidence whatever 
 that such a combination was in common use.* Moreover, the trefoiled arch only 
 appears in arcades in the second half of the twelfth century; whereas a 
 cusped window is already seen in the Norman work of the ruins of Castle 
 Rising ; it is distinctly cusped by pendent knobs projecting out of the arch. 
 Late Norman doorways also were sometimes trefoiled: e.g. the so-called 
 Prior's doorway at ELY (39) ; and the Transitional doorway of the north 
 transept of Ripon.f The function of cusps is asserted to be that of thickening 
 and strengthening a curved bar at the point where pressure is concentrated. \ It 
 may be doubted, however, whether cusps are not mainly, perhaps wholly, inserted 
 for decorative reasons. Certainly many cusps are set where they can be of no 
 constructional value. 
 
 Cusping was in use earlier than tracer)-. It is common in lancet windows 
 of the second quarter of the thirteenth century ; e.g. at WINNAL MAGDALEN, 
 HAMPSHIRE. When, therefore, in the Early Geometrical de- 
 signs two or more lancets were put together into one window, 
 it was natural to cusp both these lower lights, and the circular 
 centrepiece which they carried. 
 
 Cusps are of much importance as chronological § evidence. 
 In the first place, earl)' cusps spring from the soffit, and not as 
 the later cusps do, from the chamfer plane of the arch, i.e. from 
 the slope of the side. Soffit-cusps are seen clearly at Little 
 Addington, Winnal Hampden, Rudston. It is usually asserted 
 that the soffit-cusp is invariably earl)', and the chamfer-cusp 
 invariably late. I] But this is to exaggerate, 
 of HOWDEN (72), which can hardly be earlie 
 soffit-cusps ; probably, however, they hardly ever occur after the 
 first quarter of the fourteenth century. On the other hand, 
 chamfer-cusps are sometimes found in the first half of the 
 thirteenth century.1T Secondly, as in 510.8, in early examples of cusped circles, 
 a circle may be drawn passing through the points of all the cusps ; e.g. inside 
 the centrepiece and small circles of RUDSTON (508) : whereas in later examples, 
 the various foils are arcs not of one but of several circles. Thirdly, in early 
 examples all the foils or lobes of a trefoil, quatrefoil, &c, are parts of equal 
 circles disposed in a ring.** Fourthly, early thirteenth-century cusps are plain ; 
 
 The west window I O xyT 
 ier than 1 300, has j \ }*^ m ±j— t 
 
 Winnal 
 Magdalen. 
 
 * Paley, loc. tit., instances a lancet window in the tower of Clipsham, Rutland ; and the 
 windows of Stanton St John, illustrated in Glossary, Plate 226, as formed in this way. 
 
 t In France cusped doorways commenced early, especially in the South of France ; from 
 which they spread to Spain ; where they had a great vogue. Cf. the doorways at Dorat and 
 Rosiers, illustrated in Enlart's Manuel, 361, 305, 311. 
 
 \ Garbett's Principles, 176. So also Choisy's Histoire, ii. 380, 3S1, 382. So essential 
 a member did Mr Ruskin regard the cusp in Gothic construction that he defined Gothic as 
 "Foliated Architecture"; which is much as if one should define man as a being which hath 
 buttons. 
 
 § It must be borne in mind that such minute distinctions as the following are only true in 
 a general way. 
 
 || Rickman ; Brandon; Paley; Glossary. 1" Sharpe's Windows, 34. 
 
 ** Brandon's Analysis, 21. 
 
SI0 WINDOW CONSTRUCTION. 
 
 late cusps are sometimes molded, as in LINCOLN PRESBYTERY. Fifthly, 
 the easiest way of constructing a soffit-cusp was to cut a block to shape and to 
 insert it in a groove made in the arch or circle for its reception. Sir G. Scott 
 notes that at Westminster the whole of the windows of the church and cloister 
 had detached cusps. In a circle a set of cusps so fixed would often keep 
 one another in position. But if placed in the arch of a lower light, they would 
 drop out, if the frost found its way in. So the cusps of the lower lights were 
 sometimes cut out of the solid : while the cusps of the circles were detached 
 blocks fixed in a groove. In the end moisture and frost generally penetrated 
 
 i. Lincoln Presbytery. 
 
 4- 
 
 Ouatrefoiled Circle. 
 
 7. Ely. 
 
 2. Lincoln Presbytery. 
 
 5- 
 
 Quatrefoil. 
 
 8. Meopham 
 
 3. Ripon. 
 
 6. 
 
 Chamfer Cusp. 
 
 
 the groove, and the cusps of the circles also dropped out. Thus at Raunds * 
 the solid cusps of the lower lights survive, whereas the detached cusps of the 
 circles dropped out ; as also from the circles of the windows of the north aisle 
 of Chichester nave."|" In many late thirteenth-century windows circles which 
 now are uncusped were formerly cusped ; cusps, which had dropped out, not 
 having been renewed by the churchwardens, because of the expense or because 
 they increased the difficulty of glazing. In some examples, however, there are no 
 signs of grooves, and it is possible that the circles were always uncusped ; e.g. in 
 the west window of Grantham north aisle. £ Sixthly, the earliest cusps ended in 
 a point, as in the lower lights of Rudston, or in a blunt square tip, as in the 
 centrepiece, or in a round knob, as in Westminster chapter house. But this 
 tip was very often ornamented ; e.g. with conventional foliage, as in LINCOLN 
 
 * Illustrated in Sharpe's Windows, Plate 9. 
 \ Illustrated in Sharpe's Windows, Plate 10. 
 
 t They have been restored. 
 
WINDOW CONSTRUCTION. 
 
 5" 
 later on. with masks 
 
 ( 'lisps 
 
 in t Ik- 
 
 presbytery (510.2), and WARMINGTON (467);* 
 grotesques, and angels. 
 
 Cusp design reached its highest development in the fourteenth century. 
 were now themselves cusped ; i.e. compound replaced simple cusping ,'• % 
 choir screen of Exeter. Exquisite effects were thus gained in the rich work of 
 the fourteenth century; e.g. in the cusping of the shrine of St Etheldreda 
 ELY, and the PERCY SHRINE.f BEVERLEY (269), where tin- tips of the 
 cusps are carved into figures of the angelic choir, playing on instruments of 
 music. In Rectilinear work this exuberance of fancy died away, except here 
 
 Eh 
 
 Pedestal ot hhnne. 
 
 and there, as in HOWDEX CHAPTER HOUSE (137.), and in some of the rich work 
 of the Tudor period, such as Bishop West's Chapel at ELY (143). There is 
 much beautiful cusping in woodwork of all periods ; e.g. in the Winchester 
 choir stalls;* and the Rectilinear rood-screens of Dickleburgh and Ranworth, 
 Norfolk. § 
 
 The great west window of Binham on page 471 should be shown with foliated cusps. 
 
 t Really this is the tomb of Lady Eleanor Fitz Alan, wife of Henry, first Lord Percy of 
 Alnwick ; she died in 1328. Mr Longstaffe has pointed out that the canopy cannot have been 
 finished before 1340 ; for a shield of France and England quarterly occurs on the south side. 
 
 X Illustrated in Colling's Gothic Ornaments, i., Plate 77. 
 
 >j Illustrated in Colling's Gothic Ornaments, ii. 19. 
 
512 
 
 WINDOW CONSTRUCTION. 
 
 Rear-arch and Inner Arcade. — In the eleventh and twelfth centuries 
 the walls were often of poor construction, the inner and outer facings only 
 being of dressed stone {ashlar), while the middle was filled in with rubble. 
 Walls so constructed had necessarily to be thick. And in order to get 
 as much light as possible through such thick walls, the early windows, e.g. 
 RIPON AISLE (102), TINTERN, and STANWICK (457.9), were given a very broad 
 internal splay, so that the inner arch of the window was usually four or 
 five times as broad as the outer. But the splay was often not so extensive 
 above as at the sides ; * for it was not desired to project the light upward into 
 the roof, but downwards to the congregation. While, therefore, the window- 
 arch in the Lancet period was acutely pointed, the rear-arch might be 
 obtusely pointed or segmental. Sometimes, indeed, as in the south aisle of 
 St Albans, the rear-arch was actually lower than the window-arch. Such a 
 window consists of three parts : the outer or window-arch, the inner or rear-arch 
 
 Tintern Aisle. 
 
 Oxford Cathedral Belfry. 
 
 (Professor Willis' Scoinson arch), and the space between, the rear-vault : all 
 three parts are well seen in the Tintern window. Obviously the great thick- 
 ness of the wall, and the consequent breadth of the rear- vault and its supporting 
 walls, offered great scope to the designer. And so long as windows were built in 
 this fashion, with the glass near the external face of the thick wall, the}' could be 
 set in magnificent frames, as seen from within the church. Sometimes, but not 
 often, the rear-vault itself was vaulted with diagonal ribs ; or it was molded, 
 as in the east window of Ripon ; and most elaborately of all, in the east 
 window of NETLEY (699.4) ; most often, it was left plain, as at Stanwick. Its 
 jambs, as at Netley, might be beautified with detached shafts in correspond- 
 ence with the orders of the molded rear-vault above. Much care was given 
 also to the rear-arch. Sometimes, as at Stanwick, it was chamfered ; often it 
 was richly molded, as at Rievaulx, Guisborough, Netley, and TINTERN.f These 
 
 * York transept is an exception ; here the window and rear arches are concentric. 
 t Here w = window ; R = rear-vault ; S = Scoinson arch. 
 
WINDOW CONSTRUCTION. 
 
 513 
 
 moldings sometimes die into the jambs; sometimes they descend to the sill; 
 sometimes they are stopped by a corbel ; sometimes, as at Tintern and Netle) 
 they are carried to the sill by a shaft or shafts. In later examples, e.g. in a 
 fourteenth-century window in the triforium of Gloucester, and in the- north 
 aisle of the choir of Worcester, in thirteenth-century work, the arch may be 
 cusped. 
 
 But there was a still more delightful way of turning the plain splay into 
 a thing of beaut}- ; this was to 
 construct minor arches beneath 
 the rear-arch. This had been 
 done long before in many a clere- 
 story ; e.g. in Norman work at 
 Winchester and ELY (506; ; and 
 in the thirteenth century at 
 Boxgrove. So also, for the sake 
 of strength, belfry windows in 
 towers, e.g. at oxford (512), had 
 been at times built double. So 
 arose a fashion of constructing 
 two windows ; the outer one glazed, 
 the inner one unglazed. Charming 
 examples remain in Salisbury tran- 
 sept, Worcester choir aisle, and 
 Stone, Kent. In the late thir- 
 teenth century the inner arcade to 
 the window was still employed in 
 rich examples, as in Lincoln pres- 
 bytery, and Durham eastern tran- 
 sept. And in the best work, both 
 here and in France, e.g. at Stone, 
 DURHAM (514), and the glazed 
 triforium of St Ouen, Rouen, care 
 was taken that the inner and outer 
 tracery should be different in 
 pattern.* One of the latest ex- 
 amples of the double window is 
 in the presbytery of YORK (199). 
 But here the whole arrangement 
 is reversed ; it is the outer tracery 
 that is unglazed, the inner that is 
 
 glazed. The result is that the window proper is in a line with the inner face of 
 the clerestory wall, and from within the effect is flat and shadowless. Therefore, 
 
 * So those ingenious decorators, the Arabs, wishing to combine the beauties of two kinds 
 of ornament, often do so without inconsistency by placing them on the same surface, but giving 
 them different degrees of relief, or different colours, so that one appears superposed in front of 
 the other, without interfering with it. The eye can follow each separately, as the ear follows 
 the bass or treble of a complex piece of music (Garbett's Principles, 47). 
 
 2 K 
 
 Stone. 
 
5H 
 
 WINDOW CONSTRUCTION. 
 
 when the choir was built a few years later, the outer unglazed arcade was 
 abandoned, and the clerestory windows were set further away from the inner 
 face of the wall. A similar change of design occurs in St Urbain de Troyes, 
 where the windows both of choir and nave have double arcades of tracery ; 
 but in the choir the inner arcade is glazed, in the nave the outer. 
 
 So far we have spoken as if the custom of setting the window close to the 
 outer face of the wall was universal. This was not so. There were many excep- 
 tions ; e.g. at Glastonbury and Wells, and at Byland, Whitby, Lincoln, SOUTH- 
 WELL (390), the windows were considerably recessed externally. Such an exterior 
 gains greatly by the shadowy recesses of its fenestration. But the reason for the 
 
 arrangement was probably not 
 an aesthetic one. It was no 
 doubt simply the practical reason 
 that the glass, when recessed, 
 was better protected from the 
 rain. And as usual, in all good 
 Gothic, the practical reason pre- 
 vailed ; and the windows were 
 more and more recessed exter- 
 nally. In Curvilinear windows, 
 as at SLEAFORD (701.3), the glass 
 is usually placed exactly in the 
 centre of the wall, and the mold- 
 ings of the inner and outer arch 
 are usually the same ; in Geo- 
 metrical windows this is rarely 
 the case. In some Perpen- 
 dicular windows, e.g. in the tower 
 of LOUTH (611), the window is 
 recessed very deeply externally, 
 and very fine shadow effects are 
 produced. But what such win- 
 dows gained externally, they lost 
 internally ; especially now that 
 the walls were less thick than 
 they had been in the thirteenth 
 century. The loss, however, of 
 internal moldings such as those of Ripon and Netley may have been a matter of 
 indifference to the fifteenth-century designer ; who seems to have thoroughly 
 appreciated the fact that black-and-white work would be thrown away in the 
 blaze of colour of his windows ; just as an etching is killed by juxtaposition 
 with an oil painting. 
 
 MULLIONS (Monials). — In moldings of the mullions one practical improve- 
 ment was made. Some of the earlier mullions were nearly equally broad both 
 ways ; e.g. in the east windows of Lincoln and Guisborough. But to resist 
 wind pressure they needed to be longest in a direction perpendicular to the walls. 
 The later ones are generally so designed; e.g. BERE (515.13-15). Artistically 
 
 Durham Eastern Transept. 
 
Maldon. 
 
 6, 7. Feeling. 
 
 8, 9. George Inn, Glastonbury. 
 
 10, 11, 12. White < nine. 
 
 13, 14, 15. Here. 
 
5 i6 
 
 WINDOW CONSTRUCTION. 
 
 the window gains much by depth of mullion. For the glass cuts off half of each 
 mullion from view ; so that with mullions of the earlier type there is little room 
 for shadow. 
 
 Tracery bars are continuations of mullions ; and their moldings are some- 
 times repetitions, sometimes elaborations of those of the mullions, as at MALDON 
 
 (SiS)- 
 
 In the Maldon window, which is dated c. 1276 by Hadfield but appears 
 
 later, the centrepiece is a great circle ; B is the profile of the mullions of 
 
 the lower lights and of the tracery bars ; C of the outer order of the circle ; 
 
 F of its inner order and of the curved spokes ; E of the triangles and minor 
 
 tracery. 
 
 In the Curvilinear window of TILTEY (699), half plans are given of the 
 moldings of the mullions, tracery, and cusps. 
 
 If, however, the windows of small plain churches be inspected, such richly 
 molded work will rarely be found; e.g. at FEERING ( 515.6, 7), though it is of 
 
 St James', Bristol. 
 
 Boyton. 
 
 the same period as Tiltey. Still plainer are the mullions, tracery, and cusps 
 of the WHITE COLNE window (515. 10-12); though it also is of the fourteenth 
 century. At all periods merely straight chamfers may occur, as in the Recti- 
 linear east window of BERE (515. 13-15) ; the moldings of the south window of 
 this chancel are of a type common in Late Gothic. 
 
 Circular Windows. — A small circular window or oculus was a natural 
 ornament to set in a gable, and occurs in Early Christian basilicas, and in our 
 Norman and Transitional work ; e.g. Daren th ; Iffley ; and St Cross, Winchester. 
 (1.) The smaller examples, and even the large circle in Canterbury south tran- 
 sept, are without tracery. (2.) Sometimes, however, as at Barfreston, PATRIX- 
 BOURNE (218), PETERBOROUGH WEST FRONT (lI2), and BEVERLEY MINSTER 
 (176), tracery is inserted in the form of diminutive shafts, each with capital and 
 base ; a window with such shafts or spokes is called a wheel window. (3.) 
 Or plate-tracery may be inserted, as in ST JAMES', BRISTOL, and Beverley 
 
WINDOW CONSTRUCTION. 
 
 5«; 
 
 transept. In England the example on the largest scale is in LIN< OLN NORTH 
 transept (69), the Dean's Eye fit faces the Deanery). We have nothing, 
 however, to compare with the immense plate-tracery circles of Laon and 
 Chartres. (4.) Later come those with Geometrical bar-tracery, as at UOYTON, 
 WILTS (516). (5.) Next come Curvilinear windows, with Flowing tracery, of 
 which we illustrate examples at WALTHAM (488) and at LINCOLN 484), the 
 Bishop's Eye, facing the Bishop's Palace. (6.) Rectilinear examples are rare : 
 one is the east window of St Margaret, Lynn. 
 
 The rose window was not developed to the same extent as in France. 
 Nor as a rule did we put it in so conspicuous a position ; employing it usually 
 in the transepts only. In the west front of Byland, however, there is a ruined 
 rose of great span; and a vast rose, as big as the largest of France, those 
 of the transepts of Notre Dame, Paris, seems to have occupied the east front 
 of Old St Paul's. 
 
 In France the rose was first put under a circular 
 arch ; as at Laon, Chartres, Paris west rose, Braisne, 
 Mantes, Abbaye d'Ardenne. Then it was set under a 
 pointed arch, as at Reims. Then it was inscribed in a 
 square, and the spandrels were pierced, as in Paris and 
 Sees transepts. This was the stage apparently reached 
 in Old St Paul's.* Then the French set it on a tier 
 of lower windows, so that the rose ceases to be pre- 
 dominant, and becomes merely the centrepiece of a 
 vast window-composition covering the whole end of 
 such transepts as those of Rouen and Sens Cathedrals. 
 
 Circular windows occasionally occur in the clere- 
 story ; e.g. in Southwell nave and Ledbury chancel ; 
 and frequently in East Anglian parish churches ; more 
 rarely in the triforium, as in Waltham Abbey, WEST- 
 MINSTER (379), Hereford north transept. 
 
 Other forms of window occur. Especially common 
 is the Vesica Piscis, e.g. in the gable of the south 
 transept of UEVERLEY (176). The clerestory windows 
 spherical triangles. 
 
 BELFRY Windows. — In Somerset it is common to fill in the lower part of 
 belfry windows with perforated stonework instead of louvre boards ; as at 
 HUISH EPISCOIT and ISLE ABBOTS (S9l)- In the next stage lower, rectangular 
 openings called "sound holes," are usually employed in the towers of Norfolk; they 
 are generally filled with flowing tracery + e.g. Worstead ; HOLME (59O. Really 
 they are not sound holes ; for they light the ringers' chamber ; not the belfry. 
 
 Low SIDE WINDOW. — This occurs usually on one of the sides of a chancel : 
 its lower part, or the whole of it, closed with a shutter; e.g. GEDNEY (474.1 . 
 Its ritualistic use is still uncertain. The chief theories are these— (1.) It may 
 be a Leper's window ; this is highly improbable. (2.) A lamp may have been 
 placed within to scare away ghosts. (3.) Confessions may have been heard 
 
 * Sec Hollar's print in Prior, 345. 
 
 t A collection of "sound holes" was illustrated in the Builder, Aug. 18, 1900. 
 
 jit A jfc 
 
 ""7* « »'<** 
 ^ _ 
 
 ¥. 
 
 Huish Episcopi. 
 of Lichfield nave are 
 
Si8 WINDOW CONSTRUCTION. 
 
 through it of persons not allowed to enter the church. (4.) A small sanctus bell 
 may have been rung from it, to apprise the neighbourhood of the Elevation of 
 the Host. See illustrated articles by J. H. Parker in Archaeological Journal, 
 vol. iv., Dec. 1S47 ; J. Piggott in Reliquary, 1868, vol. ix. 9; P. M. Johnston in 
 Transactions of St Pauls Ecclesiological Society, vol. iv. 263 ; J. J. Cole in Journal 
 of Archaeological Institute, March 1848; Rev. J. P. Hodgson in Archceologid 
 Aeliaua, 1901. 
 
Chapter XXXVI. 
 
 THE TRIFORIUM. 
 
 Definition— Sham Triforium — Four-storied Interiors — Windowed Triforium Blindstory 
 —Walled Triforium— Absorption of the Triforium into the Pier Arcade or the 
 Clerestory — The Transparent Triforium — Use of the Triforium. 
 
 Defimtii ix.— The term triforium strictly applies only to the arcade which is 
 seen in many vaulted churches below the clerestory and above the pier-arches. 
 But it is often used, not of the arcade, but of the space at the back of the arcade. 
 So that it means sometimes the triforium arcade, sometimes the triforium 
 chamber.* We will speak first of the triforium chamber; secondlv, of the 
 triforium arcade. 
 
 TRIFORIUM CHAMBER.— This is the space between the vault of the aisles 
 and a lean-to roof of sharp pitch which is constructed in order to protect the 
 masonry of the vault from the weather. If, however, the vault of the aisle be a 
 half barrel f 284;, as in St Trophime, Aries, there will be no space between 
 the half barrel and the roof, and consequently no triforium chamber.+ We 
 have no example of an aisle so vaulted. 
 
 In Southwell NAVE (520) is seen the exterior of the north aisle of the 
 nave with a lean-to roof. Below is shown the vaulting of this Norman aisle, 
 which has diagonal ribs. In the interior of the nave one looks through the 
 triforium-arcade, which consists of one semicircular arch in each bay, and has 
 a glimpse of the lean-to roof of the aisle. In the example from CHICHESTER 
 NAVE ("520), the whole of the triforium-chamber of the south aisle is seen 
 from end to end, from the distant east end where it opens into the south transept 
 to the west end where it opens into the south-west tower. The interior of a 
 lofty triforium-chamber is illustrated from Norwich nave (371 1. 
 
 SlIA.M TRIFORIUMS. — Before going further, we may clear the ground of 
 
 * The etymology of the word is obscure. It has been derived from trim? fores, "a triple 
 opening." But the triforium has usually a single, double, or quadruple opening ; very seldom 
 a triple one. Mr Edward Bell, however, notes that Gervase, who first uses the term, applies it 
 at Canterbury not only to the triforium proper, but to any passages in the thickness of the wall, 
 e.g. in front of the clerestory. Now the arcade of a Norman clerestory is often a triple one ; 
 e.g. at Chichester and Ely. He therefore suggests that triforium was originally applied to the 
 triple arcades of the clerestory passage, and was afterwards extended to any wall passages or 
 thoroughfares. Willis, however {Nomenclature, 6 r, 3, and 70, 4), points out that opus triforiatum 
 was applied to perforated work in lock plates, brass fenders, &c, in which figures of plants and 
 animals were produced by piercing plates of metal. So also Enlart's Manuel, 255 : " Triforium 
 vient de l'adjectif francais trifore ou trifoire, sorti du latin transforahem et qui signifie reperce 
 ou ajoure ffrequent dans le-> inventaires de joyaux). Le triforium e^t, en efifet, un chemin de 
 ronde dont la paroi est ajouree." S.> that originally it means any passage in the thickness 
 of the wall as well as the passage provided by a triforium chamber. 
 
 + See Classification of Romanesque, 278. 
 
520 
 
 Southwell from N.W. 
 Southwell, N. Aisle of Nave. 
 
 Southwell Nave from S.W. 
 Chichester, Triforium of Nave. 
 
Waltham Abbey. 
 Cerisy Nave. 
 
 Shrewsbury Abbey. 
 
 St Saviour's, Southwark, Choir. 
 
C22 
 
 Chester Cathedral Cloister. 
 Malmesbury Nave. 
 
 Norwich Cathedral from S.W. 
 Ely Choir from S.W. 
 
523 
 
 York S. Transept 
 Lichfield Nave. 
 
 Canterbury Choir. 
 Hereford X. Transept. 
 
524 
 
 Chester Cathedral Presbytery. 
 Wells Nave from E. 
 
 Chester Cathedral N. Transept. 
 Tintern from N.E. 
 
5^5 
 
 Bayeux Choir from S.W. 
 Oxford Cathedral Transept. 
 
 Bristol St Mary Redcliffe. 
 St David's Nave. 
 
520 
 
 Clermont-Ferrand Choir. 
 Ely Presbytery from S.E. 
 
 Shrewsbury Abbey. 
 
 Ely Presbytery from S.W. 
 
SHAM TRIFORM MS. 
 
 certain so-called triforiums,* which lack cither the vault below, or the lean-to 
 roof above, or both. Of the first, that of Rochester nave is an example; it 
 has the triforium-arcade, but the aisles arc unvaulted. It was never intended 
 to have a triforium-chamber, for a passage is provided from bay to hay in the 
 thickness of the wall. The nave of WALTHAM (521) also, c. H30, has no 
 aisle to the vault, as appears from the foreground, and therefore no triforium- 
 chamber. But at the back of each pier and on the aisle wall opposite to each 
 pier are the marks left by supports for a vault which have been torn aw aw So 
 that originally there was a genuine triforium here. .\ French example of a 
 sham triforium is seen in the archaic nave of Vignory. 
 
 To a second class belong those triforiums which have above them not a 
 lean-to roof, but a vault. Of this second class we have two examples. One is 
 ST JOHN'S CHAPEL (283) in the Tower of London, c. 1080; where above the 
 groined lower aisle is a longitudinal barrel vault. The other is in GLOU- 
 CESTER CHOIR (282), where above a groined lower aisle is a half barrel f 
 vault. In the Romanesque and Gothic of the twelfth century in France 
 the so-called triforium was frequently covered with a groined or with a ribbed 
 vault. This occurs even in the eleventh century at Jumieges. Considerable- 
 confusion has been caused in French and English text-books by speaking of 
 these as triforiums ; they are both upper vaulted aisles. \ 
 
 A third class includes those which have neither lean-to roof nor vaulted 
 aisle, but a triforium-arcade ; e.g. Malton, Dunstable, Tutbury, and SHREWSBURY 
 ABBEY (521). This has come about accidentally. Originally there was a 
 clerestory. When this became ruinous it was pulled down ; the lean-to roof of 
 the aisle was lowered, and the triforium-arcade was glazed ; e.g. the triforium- 
 arcade of SHREWSBURY ABBEY (526) was glazed after it had lost its clerestory ; 
 recently the clerestory has been rebuilt. At Binham it was the aisle that became 
 ruinous and was pulled down; leaving a triforium-arcade, but no triforium- 
 chamber. At PONT AUDEMER ( 1 3 1 ), in Normandy, a charming example of 
 Flamboyant, it seems to have been found impossible to erect a clerestory; and 
 so what was meant to be a triforium-arcade was utilised as a clerestory. 
 
 FOUR-STORIED INTERIORS. — Of this we have but an imperfect example in 
 the south transept of Westminster. What would normally be occupied by a 
 western aisle is here represented by one of the walks of the vaulted cloister ; 
 and above that is the muniment room ; above which are triforium-chamber 
 and clerestory; four vertical compartments in all. There was hardly room for 
 four stories in English interiors. Our churches are much lower than those of 
 France ; our highest vaulted interior (excepting Westminster, the unusual height 
 of which is no doubt due to French influence , that of Salisbury, is but 84 feet 
 high. From the beginning the French interiors were very lofty ; in the end 
 Amiens reached 144 feet, Beauvais 150 feet, Cologne 155 feet. In the twelfth 
 
 * " Certaines eglises ont de fausses tribunes " (Enlart's Manuel, 256). 
 
 + Gloucester choir, with -rained vault to its lower aisles and demi-berceau to its upper 
 aisles, may possibly reproduce the original dispositions of the nave of the Abbaye-aux-Hommes, 
 at Caen. 
 
 X "Les tribunes Iateralesne sont autre chose qu'un second ct.i^e des bas-cotes (Enlarts 
 Manuel, 254). 
 
528 
 
 FOUR-STORIED INTERIORS. 
 
 I.aon. 
 
 century it seems to have been thought 
 in France that clerestory walls so 
 lofty and loaded with vaults required 
 to be steadied ; and so aisles were 
 built up in two stories ; both of them 
 vaulted, and the upper vaulted aisle 
 giving valuable abutment to the clere- 
 story wall* Jumieges, consecrated 
 1067, has superposed aisles; here, 
 however, the nave was not vaulted, 
 and the upper vaulted aisle was there- 
 fore hardly necessary ; it does not 
 appear again in the remaining 
 churches of Anglo-Norman Roman- 
 esque. Tournay nave is a fine ex- 
 ample ; also Romanesque. Such an 
 internal elevation found great favour 
 in the last half of the twelfth century 
 in the Transitional and Early Gothic 
 of the He de France and Champagne ; 
 but seldom elsewhere ; e.g. St Ger- 
 mer, near Beauvais ; Montier - en - 
 Der ; Notre Dame, Chalons - sur - 
 Marne; LAON nave; Noyon; Senlis ; 
 Soissons south transept ; Mantes ; 
 Meaux, originally ; Notre Dame, 
 Paris, now only in the bays adjoining 
 the transept. These are some of the 
 most impressive interiors in Gothic 
 architecture ; multiplicity and com- 
 plexity of parts, so effective in in- 
 creasing the apparent length of the 
 Gothic churches, was equally success- 
 ful in magnifying their apparent 
 height ; nothing later ever surpassed 
 the choir of St Remi, Reims ; or its 
 probable prototype, Notre Dame, 
 Chalons-sur-Marne.f But for this 
 
 * " If the walls of the tribunes have them- 
 selves vaults or transverse arches, the tribunes 
 provide abutment to the transverse arches or 
 to the vaults of the central aisle. . . . The 
 principal use of the tribunes is to provide 
 abutment by their vault to the central vault " 
 (Enlart's Manuel, 255, 256). 
 
 + Parts of Notre Dame, Chalons-sur- 
 Marne, are 1 157- n 83 ; parts of S. Remi are 
 1 170-1 190 (Enlart's Manuel, 628). 
 
HEIGHT OF TRIFORIUM. 529 
 
 a heavy price had been paid. The cost of giving the aisles additional height 
 and an extra set of vaults must have been enormous. Moreover the pier- 
 arcade was usually so curtailed as to be low and squat. So also was the 
 clerestory. This perhaps was the most serious defect of all ; for it meant to 
 cut off much of the main source of light from the Early Gothic church, already 
 plunged in gloom with its k\v insulated exiguous windows obscured by the 
 opaque stained glass of the day. This objection was taken seriously to heart. 
 Hardly was Notre Dame, Paris, finished, when the pitch of the aisle roofs was 
 lowered, thus abolishing the triforium chamber and its arcade of circular aper- 
 tures ; and the clerestory windows were brought clown lower to increase the 
 amount of light ; only the bays near the transepts being left unaltered, for fear 
 of endangering the stability of the great arches of the transept. And so the 
 greater part of Notre Dame now has a curious three-story elevation ; viz. lower 
 vaulted aisle, flat-roofed upper vaulted aisle, clerestory ; without any triforium 
 at all. While at Meaux, extensive repairs being necessary, the vault of the lower 
 aisle was pulled out, without endangering the stability of the church ; leaving 
 the aisle covered with what had originally been the vault of the upper aisle/' 
 At Eu, near Dieppe, the authorities evidently were halting between two opinions ; 
 for some of the bays are three, others four stories high internally.f Still more 
 undecided was the state of mind of the architect of Rouen Cathedral ; for he 
 built up the nave wall in four stories ; pier-arcade of the lower aisle, arcade of 
 the upper aisle, triforium-arcade, and clerestory. But as a matter of fact there- 
 is only one aisle ; not two aisles superposed ; the lower vault never having been 
 built, though the supports for it are there. % This period of hesitancy ended in 
 the adoption by everybody of a three-story interior. In Gothic architecture 
 Sens had already led the way in 1 140 ; Lisieux, c. 1 160-1 188, set the fashion to 
 Normandy, with the one exception of Rouen Cathedral ; Soissons and Chartres 
 were both designed by the end of the twelfth century with three stories only. 
 In England, as we have seen, three stories had always been the normal arrange- 
 ment. This now became universal throughout Western Christendom in churches 
 whose aisles were covered with ribbed vaults. 
 
 HEIGHT OF Triforium. — As regards, however, the relative proportions of 
 these three stories, and especially as to the prominence to be given to the 
 triforium, considerable differences in practice prevailed. We may suggest that 
 these differences arose mainly out of the difficulties of lighting the Norman 
 interiors. To get more window-area, some preferred to increase the dimensions 
 of the aisle-windows ; others, to introduce an entirely new row of windows in the 
 triforium chamber, immediately above the windows of the aisles. Where the 
 latter course was preferred, it became necessary to make the triforium-arcade 
 lofty, so as to obstruct the new source of light as little as possible. This meant 
 a lofty triforium-arcade, and a corresponding low pier-arcade. In England many 
 of our Norman churches, e.g. WINCHESTER (261), Waltham, ELY (57), Peter- 
 
 * Viollet-le-Duc, Architecture, i. 19S. + Section in Choisy's Histoirc, 11. 434- 
 
 J One curious result was that in order to provide a passage, where the floor of the upper 
 aisle ought to have been, between the different bays of the intended upper aisle, he had to con- 
 struct a pathway at the back of each pier on shafts resting on the capitals which had been 
 intended to support a vault of a lower aisle. Illustrated in Viollet-le-Duc, Architecture, vi. 18, 9. 
 
 2 I. 
 
53o HEIGHT OF TRIFORIUM. 
 
 borough, Wymondham, Binham, NORWICH (330), following the precedents of 
 the ABBAYE-AUX-HOMMES and CERISY LA FORET * (319, 52 1), made the 
 three stories nearly equal in height, assigning to each about one-third of the 
 internal elevation. Even where the Norman church has passed away, the 
 Norman proportions may sometimes still be recognised in the Gothic work 
 which supplanted it ; e.g. at ELY (522), the fourteenth-century choir has a 
 tall triforium, to harmonise with the thirteenth-century presbytery east of it ; 
 and this presbytery again has a tall triforium to harmonise with that of the 
 Norman choir which originally existed to the west of it (117, 526). So in 
 Worcester nave we may conjecture that in the westernmost Transitional bays 
 of the nave the triforium was built tall, because the triforium was tall in 
 the eastern bays of the Norman nave then standing. Then in the thirteenth 
 century the proportions of the Norman and Transitional nave were retained in 
 the Gothic presbytery and choir ; while in rebuilding the western bays of the 
 nave in the fourteenth century, these were made to agree fairly as to pro- 
 portions with those of the choir to the east and the Transitional bays to the 
 west of them. So also in Romsey nave. But even in Gothic churches that 
 were built de novo the tradition of the tall Norman triforium here and there 
 lingered long; e.g. at ST SAVIOUR'S, SOUTHWARK (521). An elevation, how- 
 ever, so proportioned is never quite satisfactory ; for a successful interior it 
 is essential that the pier-arcade shall be lofty and dignified. To gain height 
 for the ground story, and still more for the clerestory, became more and more 
 the object of the mediaeval builders. 
 
 But in the churches of the Benedictine, Cluniac, and Cistercian monks, and 
 of the Augustinian and Premonstratensian canons, there was a special cause 
 which seriously affected the proportions of the interiors ; at any rate of the nave. 
 It was that, attached to one wall of the nave, generally the south wall, as at 
 Worcester, but sometimes, as at Gloucester and Chester, the north wall, was one 
 of the covered walks of the cloister. Aisle-windows, therefore, pierced in this 
 wall would look into the cloister and not into the open air ; the light obtained 
 would be but borrowed light and of very little service. At Chester, therefore, 
 it was not thought worth while to pierce this wall at all ; and this, the north- 
 west corner of the nave, remains to this da}- plunged in perpetual gloom. 
 At Y\ orcester the monks could not resign themselves to gloom. They raised 
 the aisle-walls, and inserted their windows high up so as just to clear the cloister 
 roof. In Gothic days the same course was taken at CHESTER + (522). But 
 still this half of the nave was but imperfectly lighted. The most drastic 
 remedies for this defect were adopted by the monks of Tewkesbury and 
 Gloucester. They had designed their choirs with low pier-arcades. But they 
 designed their naves with positively Brobdingnagian piers. ^ A similar design 
 
 * The vault of Cerisy is modern ; lath and plaster. The balustrade in front of the Caen 
 triforium is not part of the early work. 
 
 + One of these later inserted windows may be seen on the left of the photograph of the 
 interior of Chester nave, ineffectually lighting the north aisle. At Leominster windows were 
 built, one in stone, the rest in wood, so high above the cloister and the roof of the north aisle as 
 to be half dormers. 
 
 1 See illustrations (26) and (297). 
 
THE WINDOW!-'.]) Tk I h'( )k [ PM. s ,, 
 
 is seen in the great abbey church of Tournus, where also there is a cloister 
 walk alongside of the nave.* With aisles so lofty as these, tall windows could 
 be had, which would be able to clear the cloister walk ;+ and at the same time 
 north and south aisles could be designed, if desired, symmetrically. Neverthe- 
 less, even with these lofty pier-arches, the naves of Tewkesbury and Gloucester, 
 and those of Durham, St Georges dc Boscherville, the Abbaye-aux-Darm 
 where also the piers are of considerable altitude, are badly lighted ; and this hold 
 attempt at a solution of the lighting problem found little favour for a Ion- time. 
 It had, however, its importance in familiarising people even in the eleventh 
 century with tallness of ground story ; ami no doubt had its share in bringing 
 about the ultimate improvement of side-lighting in lofty Gothic aisles, such as 
 those of Bristol and Canterbury. 
 
 THE Windowed TriFORIUM.— In such churches as the above neither 
 side-lighting across the aisles nor top-lighting from the clerestory was at all 
 adequate. In Gothic days, especially in Gloucester choir and its imitators, it 
 was clerestory lighting that was to be developed ; in Norman days the builders 
 were always intending to vault their naves, though they very seldom accom- 
 plished it ; and they may well have been unwilling to pierce their clerestory 
 walls with bigger windows or more windows, weakened as they were already 
 by having a clerestory passage ("545 ) constructed in their thickness. But for 
 the numerous windows in their end walls, e.g. in the apse of CERISY (161), 
 and the transept of PETERBOROUGH (161), they would have been but im- 
 perfectly lighted. But there was yet one other method of illumining the central 
 darkness beside the employment of end windows ; and this method in Norman 
 and English Romanesque was in constant employment from the first. This was 
 to insert windows at the back of the triforium. The object of these was not so 
 much to light the triforium chamber, as to transmit light across the triforium 
 chamber into the central area.* To get these windows, of course, it was neces- 
 sary to raise the aisle-walls considerably; but so important was it considered 
 to get an additional source of light, that the builders did not shrink from the 
 great additional expense involved. § At Norwich (522), so late as the fifteenth 
 century, the triforium light was still considered to be of great value ; and to 
 obtain more of it, the whole of the aisle-walls were raised, in order to allow the 
 insertion of large Gothic windows on the top of the original Norman windows. 
 In the choir triforium the new windows are exceptionally large (160). 
 
 In foreign churches where there was an upper vaulted aisle, this was 
 lighted by windows in its outer vail ; e.g. in Lombardy at S. Ambrogio, Milan, 
 and S. Michele, Pavia ; in Normandy at Jumicges in the eleventh century; and 
 in the twelfth century at Laon ; St Germer ; Paris Cathedral ; Soissons south 
 
 * Illustrated in Classification of Romanesque, p. 275. Tewkesbury nave was dedicated in 
 1 123. Gloucester nave is apparently later than that of Tewkesbury. 
 
 t One of them, filled with later tracery, is seen on the left in the photograph of 
 GLOUCESTER NAV V. ( 1 60). 
 
 { So also at Notre Dame, Paris, the vault of the upper aisle was tilted up, " comme une 
 sorte d'entonnoir, par oil la tumiere plotigeait jusque an centre de la grande nef (Choisys 
 //istoirc, ii. 430). 
 
 § This must have been very great ; calculate what it cost to raise all the aisle-walls of 
 Ely, say 1,500 feet long, by 15 feet in height. 
 
532 THE BLINDSTORY. 
 
 transept, and others. Though this was well lighted, it was accessible only by 
 narrow staircases in the thickness of the wall, and was therefore not intended 
 for worship. So it was also in the three-storied Norman churches ; e.g. the 
 triforiums of the Abbaye-aux-Hommes and CERISY (160) have windows of con- 
 siderable size at the back. This was the arrangement in the upper aisle of 
 the choir of GLOUCESTER (135), and probably of Tewkesbury ; also at St Albans, 
 Winchester, Peterborough, and ELY (34.1, 57), Dunfermline, NORWICH (31).* 
 In NORWICH NAVE (522), indeed, light from the triforium was judged so 
 valuable that the back wall was raised still higher, and a range of large traceried 
 windows was inserted. So that in the external elevation of this nave, counting 
 from the top, one sees, first, the clerestory windows ; second, the Gothic windows ; 
 and, under them, third, the Norman windows, now blocked, of the triforium ; 
 fourthly, blank arcading, in which one broad low window has been inserted, 
 on the left ; fifthly, the upper windows of the cloister ; sixthly, its lower 
 unglazed windows. In such naves as those of Ely and Peterborough, the 
 lighting problem indeed is completely solved. Nowadays they are over- 
 lighted ; but when they had their stained glass the lighting must have been 
 admirable. 
 
 To get this source of light, however, as far as possible, in its entirety, it was 
 necessary that the triforium-arcade should consist of single cavernous arches. 
 And this, which we may call the logical arrangement, is what we do find in 
 many examples; e.g. in the ABBAYE-AUX-HOMMES (319) ; at NORWICH (330), 
 where it is by no means so unsightly as usual ; at Binham ; at Wymondham ; 
 at Blyth ; at Carlisle ; at Dunfermline ; in GLOUCESTER CHOIR (47) ; Selby 
 nave, in part. At ST ALBANS (14) the single cavernous arch occurs in the 
 nave ; but not in the transepts, where there was plenty of light from the end 
 windows. Where, however, as in some few of the above, there is no back 
 window, the rationale of the single cavernous arch does not exist. 
 
 The Dark Triforium or Blindstory. — So far we have dealt with the 
 triforium as it was affected by the lighting problem. But there was another 
 and a very important problem that the Norman builders had to solve. They 
 were anxious to light their churches well ; they were also anxious to vault 
 them. So it was all over Romanesque Europe ; everywhere the same two 
 problems. Sometimes, in a sunny clime, the builders cared less for the 
 lighting ; and vaulted their churches after the fashion of Poitou or Auvergne.f 
 Under the grey sky of Normandy and England, for a considerable time, 
 vaulting was subordinated to lighting. When the Abbaye-aux-Hommes and 
 Norwich were built, the lighting difficulty was thoroughly overcome. But 
 the more thorough the success in triforium-lighting, the harder the vaulting 
 problem became. For in such churches as these, the nave-vaults they desired 
 
 * At Bernay, Caen St Nicholas, Mt. St Michel, St Georges de Boscherville, Lessay, Ber- 
 nieres, St Gabriel, the Abbaye-aux-Dames, Ouistreham, Romsey, Rochester, and elsewhere, the 
 windowed triforium was not adopted, perhaps because of its great expense. It is common to 
 find the Norman windows at the back of the triforium enlarged and filled with tracery in Gothic 
 days ; e.g. in Peterborough transept, and in Norwich transept and CHOIR (160), where the 
 square-headed window on the left lights the triforium. 
 
 + See Classification of Romanesque. 
 
THE BLINDSTORY. 533 
 
 so much to have, would have to spring from between the great cavernous 
 arches of the triforium ; which may have seemed unsound construction. So 
 we may well believe that there was an opposition school who insisted on 
 filling up these single arches more or less with minor arches, which ihould 
 act as straining-arches* to keep the supports of the vaults from bulging 
 inwardly or outwardly. The naves of Cerisy and Ely represent a compromise. 
 Both have back windows to the triforium ; but the front arch is filled in with 
 two minor arches. These, however, are so lofty, and the shaft between them 
 so slender, that little obstruction was caused to the light (521, 57). At 
 Romsey the tympanum above the minor arches is pierced ; but as there is no 
 back window, the arrangement is objectless. At SOUTHWELL NAVE (520), the 
 Romsey arrangement was intended ; here there are back windows ; so small, 
 however, that they can barely be distinguished in the photograph of the exterior 
 (520;. Most often the tympanum is solid, and the shaft substantial ; some- 
 times, as in Hereford choir, quite massive. With the subdivision of the triforium- 
 arcade, logically, there should go the omission of the back windows.f 
 
 Of these subdivided triforium arches we may distinguish two main classes : 
 first, those triforiums in which there are one or two containing-arches in 
 each bay ; secondly, those in which there is no containing-arch. 
 
 Single Containing-arch. — Of this species of triforium-arcade, Durham, 
 Rochester, Ely, Chichester, and originally YValtham Abbe}-, are Romanesque 
 examples (8, 57, 521, i, also MALMESBURY (522; and St Bartholomew's, 
 Smithfield. Beautiful Gothic parallels are seen in the presbytery and CHOIR 
 OF ELY (117, 522J, whose triforium should be compared with that of the 
 NORMAN TRANSEPT (506). But there is a practical objection to the use of a 
 single containing-arch ; viz. that it causes the triforium to take up too much 
 space, at the expense of clerestory lighting ; as is plainly seen in Ely choir and 
 presbyter)-, and at Hexham and Whitby and Bridlington. This was ingeniously 
 obviated in the nave of RIPON, by allowing the arcade to occupy only the 
 central part of the bay (102). Still more objectionable is the design of Salis- 
 bury, YORK TRANSEPT 523 , and the north side of the nave of BRIDLINGTON 
 (125); where there are three sets of arches; the outer containing-arch, two 
 intermediate arches, and four lower arches ; this takes up far too much room ; 
 the result is a truncated clerestory. At Pont Audemer there is an exquisite 
 Flamboyant triforium of this character f 131) ; but no clerestory at all. At 
 Salisbury the clerestory retains reasonable height ; but the triforium suffers 
 greatly ; being so squeezed together that its outer arches are of an unpleasant 
 segmental form. 
 
 Two Containing-arches. — If, however, instead of one, there are two 
 containing-arches, the height of the triforium can be greatly reduced, as is 
 
 * " Reste a prevenir la flexion qui se tend a se produire dans le sens transversal {i.e. from 
 pier to pier). La solution consiste a rendre les piles solidaires deux a deux a l'aide d'arceau.x 
 ou de traverses d'entretoisement " (Choisy's Histoire, 313). 
 
 t In speaking of the back windows of the triforium, one excludes those which are so 
 small that they were plainly intended not to light the nave, but merely to throw sufficient light 
 on the aisle-vault to facilitate the execution of any necessary repairs ; e.g. those of Lincoln 
 nave and presbytery ; and of \Yaltham nave- 
 
534 THE BLIXDSTORY. 
 
 seen in GLOUCESTER NAVE (26). This it was that became the favourite de- 
 sign of the English Gothic triforium. It is seen in the CHOIR OF CANTER- 
 BURY (106), as in that of Sens ; it takes exquisite form in Rievaulx choir, in 
 LICHFIELD NAVE (523), in Lincoln presbytery, in WESTMINSTER (119). In 
 Westminster the triforium, like the vault, is purely English ; no Frenchman 
 would have designed that black belt of shadow.* In an unusual form, two 
 containing-arches above six lower arches, it appears in that beautiful architec- 
 tural freak, the NORTH TRANSEPT OF HEREFORD (523). 
 
 No Containing-arch. — Several Xorman churches, e.g. the Abbaye-aux- 
 Dames, have simply an arcade of several small round-headed arches in each bay 
 of the triforium. Translated into Gothic, this gives the pointed arcades of the 
 triforiums of Byland, St David's, Xew Shoreham, Canterbury Trinity Chapel ; 
 and the trefoiled arcades of that of Guisborough." \\Tth the addition of a wall 
 behind the arches, the Byland design appears in St John's, Chester, and ST 
 SAVIOUR'S, SOUTHWARK (521); and that of Guisborough in the PRESBYTERY 
 OF CHESTER CATHEDRAL (524). The merit of this form of triforium is that 
 there being no containing-arches at all, it is greatly curtailed in height ; and 
 the space thus gained can be added to the clerestory, to the great improvement 
 of the lighting. 
 
 Continuous Arcade. — In all of the last class vertical lines are drawn 
 between the bays in the form of vaulting shafts or roofing shafts. But in the 
 Early Norman TRANSEPT OF CHESTER CATHEDRAL (524), and still more at 
 WELLS (524), no distinction is made between the bays. At Wells the vaulting 
 shafts are stopped abruptly, and the arrangement of the arches of the triforium 
 does not correspond either with that of the pier-arcade or of the clerestory. The 
 object ma}- have been by the uninterrupted flow eastward of small arches passing 
 count to give the appearance of length to this short nave : certainly it looks far 
 longer than it is. 
 
 Y\ ALLED TRIFORIUM. — There was yet one other treatment of the triforium ; 
 for which something was to be said. If, as in the examples recently cited, the 
 triforium was an open blindstory, being unlighted it was of no ritualistic use. 
 Moreover, when its roof was covered with tiles, cold air was introduced, and there 
 were down draughts on to the heads of monk or canon in the stalls below. Also 
 it became a receptacle of dust ; and this was blown into the church. But the 
 most serious objection to an open triforium was a constructional one : its 
 openings weakened the supports of the vault and roof. To the early Cistercians 
 in particular these objections must have seemed very serious ; they loved sound 
 construction and they hated decorative extravagance. So in their twelfth-century 
 churches, e.g. FOUNTAINS (101) and Kirkstall naves, and even in thirteenth- 
 century TINTERN (524), the triforium had no arcade in front of it, but simply 
 a plain, solid wall.;; BOXGROVE (318) is a curious example of this treatment; 
 owing to the presence of the big containing-arch over every pair of bays, one 
 
 * Perhaps he would not have wished to expose the roof timbers of the aisle, which at 
 times, as in Lincoln presbytery, are unpleasantly conspicuous. 
 
 t Illustrated in Sharpe's Parallels. 
 
 1 This plain wall occurs also in the fine sixteenth-century church of St Pierre, Auxerre ; 
 there it may have been intended to be frescoed. 
 
ABSORPTION OF TRIFORIUM INTO PIER-ARCADK. 
 
 forgets that there is a triforium at all. In our late churches, too, the triforium 
 is out of sight and out of mind ; merely a solid wall panelled, as in ST MARY 
 REDCLIFFE (525), and ST GEORGE'S, WINDSOR 330). In those early church 
 on the other hand, which walled in the triforium chamber, the wall was deco 
 rated with single or with superposed arcades. In St John's, (Hester; St 
 Saviour's, Southwark ; Chester and St Albans choirs, the arcade is single; 
 and the blank wall makes its presence painfully felt. But in Worcester and 
 Beverley choirs there are two arcades, and the wall ceases to be obtrusive. At 
 Worcester both arcades are pointed ; at Beverley, on the other hand, a delightful 
 contrast is obtained by constructing the back arcade low and pointed, while the 
 front arcade is tall and trefoiled.* But neither the triforiums of Beverley and 
 Worcester nor of Amiens nave and Clermont Ferrand can compare with the 
 blindstories of the Gothic of England and Normandy: e.g. BAYEUX (525; and 
 Pont Audemer. Those of our interiors which possess a blindstory have a 
 charm unknown to French interiors with the walled triforium. In such a nave 
 as Westminster or Lichfield, above is the dazzling light of the clerestory ; below 
 is the subdued light of the lower nave ; between the two, separating them, and 
 contrasting with them, is the black band of triforium gloom. No such grada- 
 tions and contrasts of light and gloom are to be had in the characteristic 
 French interiors. In BEVERLEY CHOIR (5 I) alone, the masterpiece of thirteenth 
 century Gothic, one hardly misses the shadowy blindstory ; so profuse is the 
 employment of dark Purbeck shafting. 
 
 Absorption of the Triforium into the Pier-arcade. — Not even 
 
 yet had all the possibilities of triforium design been exhausted. In the Priory of 
 St Frideswide, now the Cathedral, of OXFORD ('525, 27J, the whole height of the 
 nave wall is under 42 feet. Now if this height had been divided up into three 
 equal stories, the ground story would have been only 14 feet high, and the 
 whole internal elevation would have been crushingly low and squat. But by 
 heightening the piers so that their capitals are on a level with the floor of the 
 triforium, and by constructing the arches + from these capitals above instead of 
 under the triforium-arcade, a pier-arcade has been got which occupies not one- 
 third, but nearly two-thirds of the total height, and which is really dignified and 
 effective ; so much so that this low church appears quite lofty and imposing. 
 The first example we have of this ingenious design, with triforium as it were 
 suspended from the inside of the pier-arcade, is seen in the easternmost bay of 
 Romsey nave. } This design was also adopted at Dunstable Priory; also at 
 Jedburgh, § where it is employed much less clumsily. In the great church ot 
 GLASTONBURY (536J, c. I 1 85, it produced an interior of real grandeur. 
 
 Absorption of the Triforium into the Clerestory.— The Glaston- 
 bury design was an artistic triumph ; in practice it had one fatal defect ; viz. that 
 the height of the real pier-arcade was cramped and curtailed : whereas it was 
 
 * The Beverley design is plainly inspired by the aisle arcading in St Hugh's work at 
 Lincoln. 
 
 + Of course the real pier-arch is the plain lower arch corbelled into the piers half-way up. 
 
 \ It is amusing to sec the number of experiments in triforium design in Romsey transept 
 and nave. 
 
 S Illustrated in Fergusson, ii. 421. 
 
536 ABSORPTION OF TRIFORIUM INTO CLERESTORY. 
 
 desirable to have the pier-arches as lofty as possible. For this the reason was 
 not primarily an artistic one ; everything essential in Gothic design was dictated 
 primarily by a religious or constructional motive, only secondarily by artistic 
 considerations. It was the practical reason that if the pier-arches were high, the 
 aisle-walls could be built high, and lofty windows could be inserted in them, and 
 
 so one of the main sources of light, 
 the side-lighting, could be vastly 
 improved. The remedy for this was 
 to make room for the heightened 
 pier-arches by crushing the triforium 
 and clerestory together into one 
 story. Of the two the triforium 
 suffered most ; for the clerestory 
 was another very important source 
 of light, and its height was by no 
 means to be curtailed. A Norman 
 example of this blend of triforium 
 ,and clerestory was to be seen in St 
 Botolph's, Colchester ; * founded in 
 1 102; where, however, advantage 
 was not taken of the design to 
 heighten the piers. The next ex- 
 ample is seen at ST DAVID'S (525), 
 c. 1 190; designed for sexpartite 
 vaulting ; with which may be com- 
 pared the twelfth-century church of 
 Ouistreham, near Caen. This ar- 
 rangement seems to have been 
 particularly popular in our Western 
 Gothic. It appears at Llanthony ; 
 Christ Church, Dublin, and Per- 
 shore ; f and at Dore. Outside the 
 western district it is seen in South- 
 well choir. Ultimately it greatly 
 influenced English design ; such 
 interiors as those of Chester and 
 BRIDLINGTON (125) naves; the pres- 
 byteries of Wells and Lichfield, 
 York and SELBY (390) choirs ; 
 Winchester and Canterbury naves, 
 are all attempts to reduce the 
 triforium to a minimum in order to magnify the pier-arcade and clerestory. It 
 is particularly interesting at Winchester to see how vastly the nave has gained 
 by the transformation of a low into a tall pier-arcade (261, 90) ; and of course 
 the lighting has been improved also. An interesting foreign parallel is seen 
 
 * Illustrated in Britton's Arch. Ant., vol. i. 
 t Both illustrated in Prior, Gothic Art, 1S4, 5. 
 
 Glastonbury. 
 
THE TRANSPARENT TRIFORIUM. 
 
 stories liis/1 
 
 537 
 xit in the nave 
 
 at Semur-en-Auxois, where the transept is three 
 triforium and clerestory are blended together. 
 
 Transparent Triforium.— In France the development of the triforium 
 took an entirely different direction. Here also there was the same desire to 
 utilise the triforium by causing it to provide an additional source of light. 
 But whereas we had been putting windows at the back of the triforium, the 
 
 French succeeded in putting them in front 
 design is seen in CLERMONT FERRAND 
 there is in front an open arcade ; 
 and, as is usual in French churches, 
 a solid wall behind it. If this 
 wall were removed or pierced, no 
 light would be got ; for behind it 
 there is but the darkness of the 
 triforium chamber beneath the 
 lean-to roof. If, however, the 
 upper part of the rafters be 
 omitted, as shown in the drawing 
 of ST DEXIS, the roof can be recon- 
 structed as a double or span roof; 
 i.e. with two slopes instead of one. 
 In that case if openings are made 
 in the front wall of the triforium, 
 the) T will look into the open air. 
 These openings can now be filled 
 with tracer}- and can be glazed ; 
 and there will now be in front of 
 the triforium two sets of traceried 
 arches, one unglazed, the other 
 glazed. The change is seen in ST 
 DEXIS NAVE, which was rebuilt 
 1231-1280.* Hut there was a 
 serious objection to the new span 
 roof over the aisle. Its inner side 
 sloped towards the clerestory wall 
 at C ; and it was difficult to arrange 
 for the drainage of this lengthy 
 longitudinal gutter. Two methods 
 found favour. One, which was 
 adopted about the same time in 
 the new span roof of the 
 the 
 bay 
 
 The 
 
 526 , 
 
 older and normal triforium 
 Auvenmc. In this i hi lir 
 
 St Denis Nave. 
 
 to construct 
 parallel to 
 
 Amiens choir, 1 240- 1 269, was 
 aisles, not longitudinally, i.e. 
 ength of the nave or choir, but in separate hipped roofs over each 
 Each of these roofs is called a pavilion. Transverse gutters were 
 provided between the pavilions ; and being short, were easily kept in order. 
 Very many of the later French churches have aisles roofed 
 The nearest approach we have to it is in the nave 
 
 Enlart's Manuel, 642. 
 
 in this fashion, 
 of SHREWSBURY 
 
538 THE TRANSPARENT TRIFORIUM. 
 
 ABBEY* (526). The second method was much simpler. It was to retain 
 the lean-to roof, but to make it nearly flat ; covering it with lead ; or, as 
 at Narbonne and Limoges, with stone slabs. Of the depressed roof we have an 
 interesting example in BRIDLINGTON NAVE (124); where the north aisle 
 has an acute, the south aisle a flattened roof. In RIPON CHOIR (540) the 
 aisles originally had lean-to roofs of sharp pitch. Later, perhaps early in the 
 fourteenth century, these roofs were flattened, as shown in the photograph ; and 
 the triforium-arcade was glazed. So also in the aisles of the choir of ST CROSS, 
 Winchester (104), the roofs have been flattened. Two bays also of ELY 
 PRESBYTERY ( 1 1 7) have been treated in similar fashion. The presbytery is 
 dark ; and the shrine of St Etheldreda, a great resort of pilgrims, was badly 
 lighted. The original appearance of the triforium is illustrated on 526. Now 
 two bays on either side of the presbytery are glazed. The thirteenth-century 
 triforium of the presbytery had windows at the back : the windows and parapet 
 of these two bays of the triforium are still there, but the windows have lost 
 their glass. The roof has been flattened, and what was formerly an open 
 arcade in front of these two bays of the triforium is now glazed. 
 
 These, however, are exceptions. It was in the He de France and in 
 those districts which copied its architecture that the glazed triforium reached 
 its highest development; e.g. in the choirs of Amiens, 1240-1269; Beauvais, 
 1247-1272; St Urbain, Troyes, 1262-1369; St Ouen, Rouen, 1318-1339; Sees, 
 c - L3S3; girdling their interiors round with an additional band of gleaming, 
 sparkling stained glass. Burgundy, however, refused the transparent triforium ; 
 Auxerre and Nevers Cathedrals, Semur and Notre Dame, Dijon, remained 
 faithful to the blindstory ; as to a large extent did the Normandy churches also. 
 
 Then came in the passion which attacked the Gothic architects as much 
 in France as in England ; the passion for harmonising and simplifying. The 
 French architect when he looked up in his beautiful church, St Denis, Amiens 
 choir, EVREUX (539), saw at its greatest height a magnificent range of stained 
 glass windows lighting the clerestory, and below them another range, with similar 
 tracery and similar glass, the new transparent triforium ; window above and 
 window below ; comparing with one another, contrasting, both of them, with 
 the pier-arcade below. Evidently the internal elevation was throwing itself 
 into a tremendous contrast ; a contrast between light and shade ; between the 
 upper blaze of colour and the shadowed choir beneath. Artistically the eleva- 
 tion had arranged itself ; it was become an elevation of two instead of three 
 stories. It was simply pier-arcade versus window system. The triforium con- 
 structionally and artistically had effaced itself. The progress of effacement, once 
 commenced, went on with rapid stride. The first step to recognise the essential 
 unity of the two upper members of the internal elevation was to run the mullions 
 of the clerestory window, as had been done already in Amiens nave and Clermont 
 Ferrand, down to the sill of the triforium. This was to recognise that the 
 clerestory and triforium were no longer two distinct members but one. And 
 so the builders gradually became familiar with the idea of regarding triforium 
 and clerestory windows as being but one window. And it was not long 
 before they perforated the spandrels of the triforium windows and removed 
 * Here, however, the transverse roofs over the aisles are a modern addition. 
 
SINGLE-STORIED INTERIORS. 
 
 5.39 
 
 all traces of solid wall between triforium and clerestory, as at Sei . making 
 the whole one lofty window, descending from the ridge of the vault all the 
 way to what had been the sill of the triforium arcade. This disastrous plunge 
 into logic was first taken by the architect of St Urbain dc Troyes; the 
 harbinger and precursor of the developments of Flamboyant Gothic. For a 
 long time no one followed his lead : the triforium glazed or unglazed is the 
 greatest glory of a Gothic interior ; and many a triforium was still erei ted i ven 
 in the fifteenth and sixteenth centuries; e.g. in Evreux choir, PONT AUDI Ml i 
 (131), and Abbeville nave. I5ut in the end logic prevailed. Late in the 
 fifteenth and in the sixteenth century, e.g. at St Gervais and St Riquier, the 
 members of the internal ele- 
 vation dwindle down to two ; 
 there is practically nothing 
 left but tall clerestory and 
 tall pier-arcade: of the tri- 
 forium nothing remains but 
 here and there a balustrade. 
 
 S IN GLE - S TO RY IN- 
 TERIOR. — One step more re- 
 mained to work out the logic 
 of the triforium design. It 
 was taken with reluctance. 
 We saw above the gradual 
 upward leap of the pier. In 
 Gloucester and Tewkesbury 
 naves it rose to vast height 
 in order to allow a win- 
 dowed aisle high above the 
 cloister roof. At Romsey, 
 Oxford, Dunstable, Jedburgh, 
 Glastonbury, it rose to the 
 sill of the clerestory. Hut 
 one step remained. It was 
 to convert the wall-arch or 
 formerets of the high vault 
 into the arch of the pier. 
 
 This meant that the primary object of the piers was to support, not the arches 
 below, but the high vault. The pier-arcade was to be subsidiary; the vault- 
 arcade was to be the main feature of the interior. And this was right. For 
 the dominant factor in a Gothic building is or ought to be the vault; plinth, 
 base, pier, capital, abacus, buttress, flying buttress, pinnacle, all are con- 
 ditioned by the vaults. All these had become subservient to the vault long 
 ago; the pier had remained recalcitrant. Now its turn had come. So far 
 as it was a support of the arches of the ground story, it was to be made 
 unobtrusive ; so far as it was a support of the vault, it was to be given the 
 utmost emphasis. Few, however, learnt the new lesson thoroughly. In 
 " Illustrated in Viollet-lc-Duc, Architecture, 9. 25S, n. 
 
 ~J 
 
 ^^H 
 
 1 
 
 1 ' 1 
 
 pi 
 
 ■ ri 
 
 
 Bg 1 
 
 B^t 
 
 
 « 
 
 
 • 
 
 A i 
 
 if 
 
 1 ! 
 
 
 
 
 
 Mill U. !!■ M 
 
 ■Sm> 
 
 
 
 Evreux. 
 
540 
 
 SINGLE-STORIED INTERIORS. 
 
 LICHFIELD NAVE (523), the vault-arch is there, in the clerestory ; but it has 
 no supports. In ST MARY REDCLIFFE, BRISTOL (525), the arch of the ground 
 story is allowed to cut into the high pier (we will call it the vault-pier') : and 
 it is given capitals ; i.e. an independent existence. At Winchester, the wall- 
 arch of the vault is carried, not by a vault-pier, but by the ancient Norman 
 vaulting-shafts retained in William of Wykeham's transformation of the 
 nave (90). In CANTERBURY NAVE (go), the supports of the diagonal and 
 transverse arches are strongly differentiated from those of the wall-arches 
 of the vault, one does not see why ; their upward flow is broken by bands ; and 
 
 the piers of the ground story 
 are emphasised by capitals 
 and bands : it is a bungled 
 version of Winchester 
 nave. GLOUCESTER (59), 
 however, the cradle of our 
 later Gothic, even in 1350, has 
 the vault-pier fully developed, 
 though one desiderates greater 
 emphasis of the wall -arch 
 of the vault. The lesson 
 taught by Gloucester in her 
 choir, and, still more per- 
 fectly, in her fan -vaulted 
 CLOISTER (344), was more or 
 less elaborated in the 
 three Royal Tudor Chapels ; 
 King's College, Cambridge ; 
 Henry the Seventh's Chapel, 
 Westminster ; best of all in 
 ST GEORGE'S, WINDSOR 
 (330). Here there is no 
 question that the piers belong 
 to the vault and not to the 
 ground story ; they bring the 
 vault right down to the 
 ground ; the elevation is no 
 longer one of three or two 
 stories, but of one. And so a 
 magnificent unity was attained at last in the English interiors, whether vaulted 
 like the Royal Chapels, or covered with wooden roofs, as at Worstead, Cawston ; 
 St Andrew's, and ST STEPHEN'S, NORWICH (228) ; and, most effectually of all, 
 at CHIPPING NORTON (548). 
 
 TRIFORIUM ROOFS. — So far we have been regarding the rise, decline, and fall 
 of the triforium from inside the church. Let us go outside. As we have seen, 
 there was a widespread desire to diminish the height of the triforium-chamber. 
 This was often done, as at Ripon, Ely, and Narbonne, by partly lowering the slope 
 of the aisle-roof as in B, e.g. the south aisle of Bridlington ; and St Ouen, Rouen; 
 
 Ripon Choir and Transept. 
 
USE OF TRIFOR1UM. -.,, 
 
 or as in c, Ripon and Narbonne (558.1). From this followed a whole train of 
 consequences more or less unexpected. First, a roof, flattened so much, must be 
 covered with lead, and no longer with tiles. Secondly, there is incongruity 
 between the flat pitch of the aisles and the steep pitch of the double or span roof 
 of the nave; e.g. in SELBY choir (86). Here steps in the principle of har- 
 mony. This suggests that the steep roofs in the centre should be lowered 
 to correspond with the depressed roofs of the aisles. In many wood-roofed 
 churches, where the roof was of the late tie-beam type, e.g. GRESFORD (214), 
 the high roof had been flattened to the angle of a cambered tie-beam 562) ; 
 so that there was precedent for the change. Moreover, when once the depressed 
 four-centred transverse-arch had been adopted in vaults instead of the acutely 
 pointed transverse-arch, it had been found possible to construct the vaults nearly 
 flat ; e.g. in the fourteenth century in Ely Lad)- Chapel, and in the sixteenth 
 century in St George's, Windsor. The result was that in many a late church, 
 whether unvaulted as in HULL CHANCEL and NAVE (96), or vaulted as in BATH 
 ABBEY (373) and St George's, Windsor, both the high roof of the central aisle 
 and the low roofs of the side aisles were constructed nearly flat, and covered 
 with lead. But a roof so depressed disappears from view. The interior of 
 the church is represented simply by a series of holes between some high and 
 low walls. So that, all to get rid of the triforium, we have sacrificed what were 
 the crowning glories of a Gothic exterior, the lofty gables of its end walls and 
 the skyline of its acutely pointed roofs. The Gothic church has sunk down to 
 the level of a Greek temple. And just as the Greeks had found it necessary to 
 concentrate attention on the cresting of the temple, elaborating it into architrave, 
 frieze, and cornice, so the Gothic builders tried to divert attention from the 
 disappearance of roof and gable and the loss of skyline, by elaborating, more 
 than ever before, the cresting of their walls with parapet, battlement, and 
 pinnacle; as at ST George's, WINDSOR* (492); king's college chapel, 
 CAMBRIDGE (199); and above all, in HENRY THE SEVENTH'S CHAPEL, WEST- 
 MINSTER (378) ; by way of compensation. So that the florid exterior of a 
 Tudor church is not without its meaning and apology. 
 
 USE OF TRIFORIUM. — As we have seen, the primary origin of the triforium 
 chamber lay in the fact that a ribbed vault over an aisle was originally 
 protected from the weather by a lean - to roof of sharp pitch. But being 
 there, it may be asked, was no use made of it? In most cases it was not 
 used at all. It cannot have been used, if it was a triforium with a wall in 
 front, as in Tintern or Beverley ; or if there was no wooden floor on the 
 vaults, as at Lincoln ; or if there were no large windows at the back to light 
 it, as in Gloucester nave ; or if the newel staircase leading to it was so 
 narrow that only one person could ascend at a time. Sometimes, however, 
 it was turned to practical account. It provided a passage all round the 
 church at mid -height; which would be convenient, e.g. when tapestry f was 
 to be hung down over the nave arches on festal days. Then again an open 
 triforium-arcade diminished the weight resting on the pier - arches below. 
 
 * Probably the pinnacles originally supported beasts rampant in metal, 
 t In Winchester nave the hooks may still be seen from which was hung the tapestry at the 
 wedding of Philip of Spain and Queen Mary. 
 
542 USE OF TRIFORIUM. 
 
 But there is evidence that occasionally ritualistic use was made of the 
 triforium chamber. This was certainly the case in the upper aisle of the 
 choir and transept of Gloucester. Here there are broad staircases ; and, 
 before its eastern apse was lopped off in the fourteenth century, there was a 
 broad chamber all round the apse on the first floor. Moreover this upper 
 aisle is spacious ; is floored ; is open to the choir in front, and well lighted 
 by windows at the back. At Gloucester its ritualistic use is quite clear. The 
 eastern apses of its transept and the tangential apses of the choir are each 
 three stories high. To the lowest chapel in each apse access is gained from 
 the crypt ; to the intermediate chapel from the ground floor of the church ; 
 but to the uppermost chapel only from the upper aisle. Here then the object 
 of the upper aisle clearly is to provide access to chapels on the first floor, and 
 a road from one chapel to another ; as well as space for a congregation, the 
 altared apse itself serving as sanctuary to the priest and his acolytes. In the 
 transepts of CERISY-LA-FORET (199), the Abbaye-aux-Hommes, St Georges de 
 Boscherville, Winchester, and originally Ely, return aisles run across each tran- 
 sept end, providing again a nave for a congregation gathered in front of the 
 eastern apse of the transept. In WESTMINSTER ABBEY (119), the triforium 
 chamber is unusually spacious and well lighted ; and was perhaps designed to 
 accommodate numerous spectators on grand occasions, such as coronations and 
 royal funerals.* We have one more bit of evidence about a ritualistic use of 
 the triforium. At Winchester, by way of protest against the heavy contri- 
 bution imposed by the bishop on the monastery to recoup himself for the 
 sums exacted from him by the king, the monks went round the triforium 
 with banners and cross reversed ; from which we may perhaps argue that 
 they sometimes went round the triforium in procession with banners and cross 
 in normal position. 
 
 * Scott's Gleanings, 25. 
 
Chapter XXXVII. 
 THE CLERESTORY. 
 
 Growth of the Clerestory — The Clerestory Passage 'The Wall Passage. 
 
 DEFINITION. — The term was formerly applied to any window or to any traceried 
 opening* in a church; e.g. in an aisle, tower, cloister, or screen. It is now 
 restricted to the high windows in an aisled nave; or to the range of wall in 
 which the high windows are set. 
 
 Growth of the Clerestory.— Up to the end of the twelfth century 
 the clerestory received little development. In the greater Norman churches 
 the clerestory walls were almost always hollow, and there may well have 
 been some reluctance to weaken the inner and outer shells of the wall more 
 than necessary by windows. Secondly, where vaulting was contemplated, it 
 was desirable to keep the clerestory wall as solid as possible. So in all the 
 Norman churches, and even as late as Wells nave, and WHITBY CHOIR, c. I2IO 
 (i 14), there is but a single window in each bay of the clerestory wall. But 
 in Byland, c. ii/O, in Wenlock, c. 11S0, Rievaulx choir, c. 1230, and South- 
 well choir, c. 1230, there are two clerestory windows in each bay. Also triplets 
 of lancets appear ; first, three lancets graduated in breadth in LINCOLN CHOIR 
 (1 10), begun 1 192; then graduated in height, in SALISBURY (458); and Netley. 
 Then comes the invention of bar-tracery, c. 1240; enabling clerestory windows 
 to be immensely enlarged. Of the big broad clerestory window LINCOLN 
 PRESBYTERY (56) offers an early example, 1256-1280; then come Bridling- 
 ton, Guisborough, Exeter, and York nave. The lesson of the broadening of 
 the window had been learnt in the chapter houses ; e.g. of Westminster, 
 Salisbury, Southwell, York, Wells. Having made the clerestory windows 
 stretch from buttress to buttress, all that remained for the builders was to 
 give them a vast increase in height. Here, as usual, GLOUCESTER CHOIR 
 (S9J led the way (492;. Gloucester clerestory was copied in Norwich choir 
 after the fall of the spire in 1360. Winchester and Canterbury followed 
 Gloucester with timidity (90); York choir with less reserve. Finally came 
 the vast clerestories of ST GEORGE'S, WINDSOR (330); and BATH ABBEY 
 (573)- Ys we have seen above, expansion of the clerestory windows 
 was necessary to get more light ; but it was carried yet further, and 
 beyond all architectural bounds, by the furore for stained glass. In some 
 instances, however, artistic instinct rebelled against the undue predominance 
 of the voids over the solids; the windows were not extended from buttress 
 to buttress ; the restraint which had been so beautiful in the Tintern of 
 1269, still made itself felt in the designs, so much more solid and satis- 
 * It follows that it is incorrect to say that it means a window clear of, i.e. above, the aisles. 
 
544 
 
 Chichester Nave. 
 Cirencester Nave. 
 
 Peterborough Nave. 
 Hedon Nave. 
 
THE CLERESTORY. -,- 
 
 factory, of the choirs of Selby and Howden and the nave of Beverley; and 
 here and there in fifteenth-century work also. On the whole, however, the 
 glass had its way. 
 
 CLERESTORY PASSAGE.— One of the things particularly desirable was the 
 provision of means of getting at the clerestory windows without the trouble 
 and expense of putting up ladders and scaffolding on each occasion when 
 repairs were needed. In France it seems to have been more usual to construct 
 a passage for this purpose outside the windows. This was rarely done in 
 England. York presbytery is a late exception ; in the choir, added imme- 
 diately afterwards, the passage was made internal.* In Normandy and England 
 the rule was to construct the clerestory passage inside the church. + 
 
 No Clerestory PASSAGE. — The practice, however, was not universal, 
 even in England. There is no clerestory passage at Blyth or Leominster. The 
 early Cistercians, sticklers for sound building construction above everything, 
 built their clerestory walls solid, i.e. without an internal passage, e.g. in the 
 naves of fountains (ioi), Kirkstall, and Buildwas ; which is the more note- 
 worthy, because none of them were intended to be vaulted. At Dore* also, 
 c. i 190, and Tintern, 1269, which were vaulted throughout, the clerestory walls 
 are solid. But even with the Cistercians the English practice in the end 
 prevailed; e.g. at Byland, 11 70, which was unvaulted ; and in Rievaulx choir, 
 which was vaulted. New Shoreham, c. 1200, follows the early Cistercian 
 practice.^ Later on, the nave of Lichfield and the north transept of Hereford, 
 owing to the peculiar shape of their clerestories, have to dispense with an 
 internal passage (523;. In our Late Gothic the clerestory passage is, with some 
 exceptions, abandoned ; e.g. in GLOUCESTER CHOIR (492) ; Winchester and 
 Canterbury naves ; St Mary Redcliffe, Bristol ; St George's, Windsor ; and 
 Bath Abbey. 
 
 Clerestories of the Lesser Churches. — From the descriptions that 
 have come down to us, it would seem likely that the greater Anglo-Saxon 
 churches had clerestory lighting ; but all such churches have perished. It was 
 
 * In the illustration on 199, the passage is external on the right of the transept, internal on 
 the left. 
 
 t E.g. in the eleventh and twelfth centuries the Abbaye-aux-Hommes ; Cerisy; Lessay ; St 
 Georges de Boscherville : St Albans: Winchester transept; CHICHESTER (313): ELY (273) ; Dur- 
 ham; Norwich; Christchurch, Hants: Selby nave ; PETERBOROUGH (161); GLOUCESTER (26) ; 
 MELBOURNE (203) ; Carlisle nave ; Romsey ; Southwell nave. Late in the twelfth century 
 clerestory passages are seen at ST DAVID'S (525); Wells; Glastonbury; Llandaff; Canterbury 
 CHOIR (106); Ripon transept; Malmesbury ; St John's, Chester ; Wenlock ; Oxford Cathedral ; 
 Byland. In the first half of the thirteenth century they occur at Lincoln ; Ely presbytery ; St 
 Saviour's, South walk ; Beverley Minster; Boxgrove ; Christ Church, Dublin; Rochester; 
 Rievaulx ; Southwell ; Pershore ; in the second half in the north transept of Hereford and the 
 presbyteries of Lincoln and Chester; at Exeter; Bridlington; Carlisle. In the fourteenth 
 century in Ely, Selby, Tewkesburv choirs ; and Chester nave. In later Gothic in York choir ; 
 Ripon nave, c. 1502 ; Winchester presbytery, 1500-1528. More instances might be added ; but 
 enough have been adduced to show how widespread was the practice, and that it was noL 
 altogether abandoned even in the last days of Gothic architecture. 
 
 \ Dore choir seems not to have received a high vault till the fourteenth century. 
 
 § The solid was of course far more ancient than the hollow clerestory wall ; it appears at 
 Bernay and Jumieges in the second quarter of the eleventh century. 
 
 2 M 
 
546 
 
 THE CLERESTORY, 
 
 not till the fourteenth century was well advanced that it became normal for the 
 parish churches, first in the towns, and then in the villages, to have clerestories.* 
 Nevertheless there never was a time when some few parish churches were not 
 built with clerestories cathedral-fashion. Of these none can be more rude and 
 archaic than Towyn, where the windows are mere slits in the wall (458). 
 Well on in the twelfth century, at STEYNING (458), the aisle windows are 
 
 mere slits, probably 
 
 "if t ^fffHSIt Qf / >-''i- J £-l ^ or defensive reasons ; 
 but those of the clere- 
 story are large ; as 
 also are those of St 
 Margaret at Cliffe, St 
 Woolos, Newport, and 
 St Peter's, Northamp- 
 ton. In the thir- 
 teenth century clere- 
 stories become more 
 common ; e.g. Dar- 
 lington. HEDON 
 (544), late in the 
 same century, has a 
 clerestory still mode- 
 rate in size ; so also 
 W a r m i n g t o n and 
 Barnwell St Andrew's. 
 In the fourteenth- 
 century church of 
 Hingham also the 
 clerestory is moderate 
 in size. Then the 
 same cause which 
 brought about the 
 upward leap of the 
 clerestory of the 
 vaulted choir of 
 Gloucester, came into 
 play in the parish 
 churches ; e.g. in 
 the collegiate nave 
 of Howden, and that 
 most magnificent of all our parish churches, BOSTON (222), whose founda- 
 tions were laid in 1309. In Boston and at HOLBEACH (133), the design of 
 which is very similar to that of Boston, two windows, instead of one, are 
 inserted in each bay of the clerestory, which thus becomes an almost unbroken 
 sheet of glass. And it is this, the Howden double window, which is seen ir 
 
 In Herefordshire, however, clerestories are the rule rather than the exception; and few 
 are so late as the fourteenth century.— R. A. D. 
 
 Howden Nave. 
 
THE CLERESTORY. 
 
 547 
 
 many of the noblest churches of East Anglia, such as Snettisham ; St 
 Nicholas, Lynn; St Peter Mancroft and ST STEPHEN'S 
 and far away, as at Leighton Buzzard ; and Southam, 
 the single window remains to the ^n<\ the 
 
 NORWICH 228 ; 
 Warwick ; though 
 ie more common ; e.g. at 
 (139) and CIRENCESTER (544. In this last church 
 for light and stained glass that 
 an additional clerestory was con- 
 structed over the north aisle as 
 well as a great window above 
 the chancel arch ; the latter 
 one more reminiscence of 
 Gloucester choir ; cf. CHIPPING 
 NORTON (548).* 
 
 Moreover, vast numbers of 
 churches constructed clerestories 
 above ancient pier-arcades, and 
 rebuilt their end walls in glass, 
 making themselves Gloucesters as 
 far as in them lay. Stratford-on- 
 Avon is a fine example ; with great 
 Perpendicular west window, and 
 with Perpendicular clerestory on 
 the top of thirteenth-century pier- 
 arcade. 
 
 In comparison with these bril- 
 liantly illuminated interiors, 
 churches without a clerestory must 
 have seemed indeed gloomy and 
 old-fashioned. Nevertheless im- 
 portant examples occur ; e.g. the 
 great thirteenth-century church 
 of Yarmouth ; the fourteenth- 
 century churches of Grantham, 
 PATRINGTON f 1 33), Xantwich, 
 and NORTH WALSHAM 562). 
 The absence of the clerestory 
 at North Walsham is clue to 
 the peculiar form of roof 
 adopted. 
 
 In but few parish churches a 
 wall passage occurs. At Rye, c. 
 
 1200, there are wall passages both in the clerestory and in the aisle walls At 
 Deeping St James, originally a Benedictine church, there is on each side a con- 
 tinuous passage inside the clerestory windows, opening into the nave by thirteen 
 
 * Rather the window or windows over the chancel arch is one of the characteristics of 
 West of England Gothic ; a pair of small windows is common in Herefordshire in the thirteenth 
 century. 
 
 Lon» Melford. 
 
54 S 
 
 Chipping Norton. 
 
THE CLERESTORY. 
 
 549 
 
 lancet arches.* In the conventual unaislcd chapel of Xiin Monkton there is a 
 continuous passage all round in front of the northern, western, and southern 
 windows.f There was a similar passage at Ripon in the thickness of the lower 
 part of the twelfth-century unaislcd nave. At St Margaret's, Lynn, the 
 thirteenth-century choir had a continuous passage in front of the clerestory 
 windows. In the fifteenth century the inner arcade of this was raised in order 
 to insert taller windows ; but several of the old shafts and foliated capitals 
 were re-used. 
 
 * This is fully illustrated in Nene Valley, Plates [3, 14. 
 
 t Illustrated in Churches of Yorkshire. 
 
Chapter XXXVIII. 
 
 ROOFS. 
 
 Roofs and Vaults — Trussed Rafter Roofs — Steep Roofs with Tie Beams — King and 
 Queen Posts — Depressed Roofs with Tie Beams — Arch Braced Roofs — Hammer- 
 beam Roofs — Aisle Roofs — Wooden Vaults — Material and Construction. 
 
 Roofs over Vaults. — In the vaulted buildings of ancient Rome, whether 
 the vault took the form of a dome, a barrel, or a groined vault, e.g. the Pantheon 
 and the BASILICA OF MAXENTIUS (290), the vault was of brick or concrete ; 
 and was so thick and massive, that the only external covering employed was 
 plates of bronze or lead. And this substantial method of construction was at 
 first imitated in the Romanesque dome and barrel-churches of Perigueux and 
 Auvergne.* But when, in order to lessen the thrust, the vaults were constructed 
 in thin shells, as in S. Vitale, Ravenna, in Cluny, and in Gothic-ribbed vaults, it 
 was usual to protect them from the weather by timber roofs. These external 
 roofs seem at first to have rested on the vaults,! as they now do in part in 
 Wells choir ; thus imposing on the vault a dangerous weight, which it was not 
 intended to bear. As Gothic construction improved, however, care was gene- 
 rally taken to prevent any of the weight of the roof from falling on the vault. \ 
 Such external roofs are, of course, hidden by the vault; and so their form was 
 regulated by constructional motives only. 
 
 Roofs of Unvaulted Churches. — But both here and in France the vast 
 majority of churches were not vaulted at all. In France most of the smaller 
 churches, even those which were built with exceptional care and finish, were con- 
 structed without vaults even in the middle of the thirteenth century ; this was 
 especially the case in Northern France, and still more in the Netherlands. § 
 Even in French parish churches that were vaulted, those parts were rarely 
 vaulted which had aisles, and which therefore would have required flying but- 
 tresses. || The vaulting was usually confined to the unaisled choir, and to the 
 central tower if there was one. This was the case in England also ; where in 
 the twelfth century there was a marked tendency, afterwards largely abandoned, 
 to vault the chancel ; e.g. Tickencote, Lastingham, Darenth, CASSINGTON (215.5), 
 Hemel Hempstead. The great majority of English parish churches, however, 
 
 * Classification of Romanesque, 279. 
 
 t The roofs still rest on the vaults in the Baptistery of Alliata, and at St Thomas near 
 Almenno (De Dartein), 466. 
 
 % Not always. In Winchester nave parts of the vault were actually built in the four- 
 teenth century round the ancient tie-beams. See Colson's account of the restoration of 
 this roof. 
 
 § Enlart's Manuel, 496. || Choisy's Histoire, ii. 47S. 
 
55' 
 
 Woodbastwick Nave. 
 Paston Nave. 
 
 C.edney Nave. 
 Sail Nave. 
 
552 
 
 Llanidloes from S.W. 
 Paston Barn. 
 
 Trunch from W. 
 Cawston from W. 
 
;;-, 
 
 Km il - 
 
 8. Patrington. 
 
 13. Trunch, Cornice. 
 
 12. Trunch, Hammerbeam. 
 14. Trunch, Purlin. 
 
 15. Trunch, Principal. 
 
tt#2& 
 
 '^'i 1 ." 15?" ', • --■ ■- 
 
 Stuston. 
 Lympenhoe. 
 
55; 
 
 fiBCtion of Priiiupal 
 I 
 
 Section of Collai 
 [ ] 
 
 Capel St Mary. 
 
Principal and Couraicm Rafters 
 
 Norwich St Stephen. 
 
557 
 
 mm 
 
 
 Westminster Hall. 
 
553 
 
 PITCH OF ROOFS. 
 
 i. Principal. 
 
 2. 
 
 Purlin. 
 
 3. Collar. 
 
 4. Arched Brace. 
 
 5- 
 
 Wall Post. 
 
 6. Wall Plate 
 
 7. Strut. 
 
 8. 
 
 Rafter. 
 
 
 
 III. 
 
 
 1. 
 
 Principal. 2. Purlin. 
 
 4. Arched Brace 
 
 5- 
 
 Wall Post. 6. 
 
 Wall Plate. 
 
 7- 
 
 Longitudinal Strut. 8. 
 
 Rafter. 
 
 9- 
 
 Tie Beam. 10. 
 
 King Post. 
 
 11. 
 
 Strut. 12. 
 
 Ridge Piece. 
 
 I. Lowering of Roof. 
 III." Adderbury. 
 
 I. Leicester St Mary. 
 IV. Faringdon. 
 
 were altogether without 
 vaults.* Moreover, very 
 few of the cathedral or abbey 
 churches were vaulted till 
 late in the twelfth century ; 
 some, like Peterborough,-^ 
 have remained throughout 
 unvaulted to this day. 
 Therefore it was a matter of 
 primary necessity in England 
 to work out a system of 
 ornamental roof construc- 
 tion. At first, as in all our 
 mediaeval architecture, the 
 problem was approached 
 only from the engineering 
 side. It was probably not 
 till the beginning of the 
 fourteenth century that the 
 builders, having made their 
 roofs satisfactory from the 
 point of view of building 
 construction, found them- 
 selves at liberty to develop 
 the roof from the artistic 
 side. An early example of 
 the new type of roofs may 
 be that of the barn of Easby 
 Abbey ;} other examples c. 
 1300 are to be seen at Pole- 
 brook, Sparsholt, Kidding- 
 ton,§ ADDERBURY, and St 
 Martin's, Leicester. || 
 
 Pitch of Roofs. — 
 The pitch of a roof depends 
 
 * In later days it was common 
 to vault the western tower : and 
 in East Anglia the porches also ; 
 e.g. the north and south porches 
 of Sail church. 
 
 t With the exception of the 
 eastern chapels added afterwards, 
 and probably its Norman apse. 
 
 I G. E. Street in Prior, 355. 
 
 § Glossary, Plates 173 and 
 176.' 
 
 || Illustrated in Brandon's 
 Roofs. 
 
TRUSSED RAFTER ROOFS. 
 
 559 
 
 partly on climate, partly on roof-covering. In a climate like Norway, where the 
 snowfall is excessive, the most essential thing is to get the snow away at once 
 before it consolidate and break down the roof by its weight ; the pitch, therefore, 
 of the roofs of Borgund and Hitterdal is excessively sharp. The pilch of an 
 English roof is chiefly governed by the nature of the roof-covering. While tiles 
 or thatch were chiefly employed, the pitch of the roofs was moderately sharp. 
 But as lead came more and more into use, it became possible to reduce the pitch 
 more and more, till indeed many later roofs are nearly flat. Though the Norman 
 roofs have nearly all disappeared, we can recover their pitch from the weatherings 
 of many an old roof, left on a western or central tower, as at TEWKESBURY 
 (390) ; from these it would appear that an angle of 90 degrees was usual. 
 Early Gothic roofs are usually steep, at Lincoln remarkabl) so ; but they arc- 
 rarely equilateral ; i.e. with an angle of 60 degrees. Roofs of low pitch are 
 most common in late work ; but they occur late in the thirteenth century at 
 Warmington and Polebrook ; and soon after at St Martin's, Leicester, and at 
 Wymington. At no period, however, was the roof of steep pitch disused.* 
 
 ROOF Construction. — The simplest form of roof is that in which the 
 rafters are not supported by a framework of more massive timbers ; these are the 
 trussed rafter or unframed roofs. But in large roofs the rafters are usually 
 supported by a framework or truss of more substantial character ; comprising 
 principal rafters and purlins ; and usually tie-beams or collars or both, to which 
 various other members may be added. In the present chapter all the roofs 
 except the trussed rafter roofs are framed or trussed roofs ; i.e. roofs with framed 
 trusses. 
 
 Trussed Rafter Roofs. — The earliest large roofs probably had tie- 
 beams ; and when they received the full apparatus of wall-piece, arched brace 
 struts, principals, and ridge-piece, as at Gedney, they were very heavy. More- 
 over, in many districts no doubt timber of sufficient scantling for big tie-beams 
 was not to be had. What was wanted was a roof that could be constructed of 
 small timber. This was got by constructing a roof of common rafters only ; to 
 prevent these from spreading diagonal ties (553.5) were added; all the timbers 
 being halved and pegged together with wooden pins. The common rafters were 
 placed from i! to 2 feet apart ; their scantling averages 5 by 4 inches. Such a 
 roof was comparatively inexpensive. It probably was in early use over vaults; 
 for which it gave greater headway than a tie-beam roof. The foot of the rafter 
 often projected beyond the outer face of the wall, especially in Last Anglia, 
 in order to give good "dripping eaves" (393); and therefore could not be 
 fixed to a central wall-plate. Therefore, as in STUSTi IN PORCH (554), other short 
 pieces of timber, Sole-pieces, are laid on the wall-plate at right angles to it; 
 and the rafters are pegged to their projecting ends. Still further to attach the 
 rafters to the sole-pieces, and to prevent the former from spreading, upright 
 struts (553.IO), are inserted ; pegged to the rafter and to the inner end of the 
 sole-piece. Such a roof has six sides, and may be distinguished as a six-sided 
 rafter roof. It will be seen that it is much simpler in construction than the 
 tie-beam roof; dispensing as it does with ridge-piece, purlins, principals, and 
 tie-beam. The defect of it is the absence of longitudinal ties ; when not 
 
 Brandon's Roofs, 1 2. 
 
S6o 
 
 TRUSSED RAFTER ROOFS. 
 
 boarded, the rafters are generally out of the perpendicular, inclining, in a nave, 
 either to east or west. At Ely it was found that the trussed rafter roof of the 
 presbytery, an exact counterpart of that of the nave, had pushed out the east 
 wall so much that it overhung its base two feet ; the wall was got back into 
 position by means of screws. 
 
 In the above six-sided rafter roofs, in spite of the four ties (553.5), there 
 is still much risk of spreading, and they are not fit for use over wide spans. 
 Usually, therefore, they are strengthened by an upper beam called a collar beam,* 
 
 Wimbotsham: 
 
 or Collar ; which is most commonly placed above the diagonal ties (553-6). 
 Such a roof may be called a seven-sided rafter roof, e.g. PASTON (551). Owing 
 to the additional strength given by the collar, this roof was sometimes used 
 over wide spans; e.g. in Reedham Church, 31 feet span ; and Ely nave. To 
 keep out draughts, this roof also was often boarded over, as in Ely nave and 
 WIMBOTSHAM. Sometimes, as at Heckington, the chancel had boarded, the 
 nave open rafters. 
 
 This is also called a Wind-beam ; because another of its functions is to resist wind 
 pressure. There may be two or more collars, as at Wimbotsham. 
 
TIE-BEAM ROOFS. -m 
 
 More rarely, the diagonal ties intersect the collar beam, as at i.\ MP] nhok 
 (554), and FILBY* Norfolk (553.7). 
 
 Large numbers of trussed rafter routs of the above three types still remain ; 
 they are an interesting feature in the Canterbury parish churches. They are 
 said not to have been used till the thirteenth century. They never afterwards 
 went entirely out of use. 
 
 STEEP Rooks with TlE-BEAMS.— The earliest roofs of considerable span 
 probably had tie-beams. A small roof might be constructed as in 553.1, 
 with rafters only. But such a roof would tend to spread. To prevent this, 
 two precautions were taken. One was t<> lay on or build into each wall a longi- 
 tudinal wall-plate. (553.2); into this the feet of the rafters were fastened. 
 Secondly, across the nave, at right angles to the wall-plates, tie-beams were placed, 
 and the ends of these tie-beams were pinned to the wall plates. This prevented 
 the wall-plate from being dislodged by the outward thrust of the rafters. 
 Such a roof system stopped these thrusts. But there were serious objections 
 to it, especially when employed over considerable spans. Beams of such 
 length and scantlings as those of Winchester nave were difficult to obtain ; at 
 Winchester they were obtained by a permission employed, not quite honestly, 
 to cut down a royal forest. + With the bad roads and unbridged rivers of those 
 days, the cost of transport also was excessive. And it must have been very 
 difficult to raise them into position. Moreover, if it was wished to vault the 
 church, the apex of the vault was prevented from rising higher than the tie- 
 beam. But the worst of all was that a long tie-beam tends to sag or droop in 
 the centre: a tendency which was aggravated in man)- roofs by the super- 
 position of a king-post (553-3). 
 
 KlXG-POST Roof. — We have said that there was a tendency of the 
 rafters of a nave to spread outwardly; /'.<•. north to south. But they also 
 might rack from east to west. To prevent this, the tops of each pair were 
 fastened to a longitudinal beam, called the ridge-piece. % To support this ridge- 
 piece, upright posts, called king-posts, were sometimes set up in the centres 
 of the tie-beams. Usually, to the upper part of the king-post were attached 
 four struts ; two passing up to the riclge-piece, and two to a pair of rafters ; 
 as at YVOODBASTWTCK >; (55 1). Such a construction made matters worse still ; 
 additional weight from the roof being brought down on to the tie-beam just 
 at its weakest point, the centre. Nowadays we should peg the king-post to the 
 tie-beam as well as to the ridge-piece. These king-post roofs were very 
 common in Kent ; it will often be found that the king-post has been removed 
 from the tie- beam ; probably because it made it sag. 
 
 QUEEN-POST R00E. — A less objectionable roof is one in which there are two 
 uprights (5 5 34). At north WALSHAM (562), c. 1390, there is a king-post in 
 the aisle roof, queen-posts in the nave roof. By a further extension of this 
 
 * Illustrated in Brandon's Parish Churches, 37. t Britton's Winchester, 38. 
 
 \ See ridge-piece in 564; and sections of ridge-pieces in 555, 55''. 
 
 § King-posts might also be placed on collar beams. 
 
 || That is, in medieval construction the king-post was in compression ; in modern con- 
 struction it would be in tension. A member is said to be in tension when a chain or wire can be 
 substituted for it. 
 
 2 \ 
 
562 
 
 TIE-BEAM ROOES. 
 
 system, a row of uprights was sometimes set up on later tie-beams and collars ; 
 as at ST NICHOLAS, LYNN (214); and LOUTH (213); harmonising very well 
 with the tracery of the Perpendicular windows. A singularly beautiful roof of 
 this type is to be seen in the Trinity Chapel, Cirencester Church.* 
 
 Depressed Roofs with Tie-beams. — The more work was thrown on the 
 tie-beam, the more it tended to sag. To stop this in later roofs, various remedies 
 were adopted. 
 
 I. One was to camber \ it ; i.e. so to cut it that it was slightly arched at the 
 centre; as at WOODBASTWICK (SSl), GEDNEY (551), LONG MELFORD (214). 
 Almost all later tie-beams are more or less cambered. The camber may 
 perhaps average 1 inch in every 5 feet ; in a modern roof perhaps 1 inch in 20 
 feet. 
 
 II. A second method was to construct arched braces under the tie-beam 
 to support it. Sometimes the curved braces meet the tie-beam in the centre, 
 
 North Walsham. 
 
 and form a complete arch, as at ST MARY'S, LEICESTER; ADDERBURY (558); 
 LONG MELFORD; ST NICHOLAS, LYNN (214); and Croydon Palace.^ Sometimes 
 they support the tie-beam only towards the extremities. In the latter case 
 four-centred arches may be formed ; harmonising with the four-centred arches 
 of doorways and windows ; see illustrations of LOUTH (213); GRESFORD (214) ; 
 and MELBOURNE (203). 
 
 In some few examples, e.g. Morton Church, and Pulham,§ the arched 
 braces are framed into, and look as if they pass through, the tie-beam, with very 
 unpleasant effect. 
 
 III. But in later examples, such as Boston, Holbeach, St Peter Mancroft 
 and St Stephen's, Norwich, the clerestories were perforated with numerous and 
 large windows, between which little masonry or pier was left. To balance the 
 
 * In Westminster hall great use is made of these upright struts (557). 
 t A cambered beam is defined as one whose upper surface has been cut to a slight slope 
 from the middle towards each end. 
 
 J Illustrated in Pugin's Examples, i. 38. § Illustrated in Brandon's Roofs, 13 and 21. 
 
ARCH-BRACED ROOFS. 
 
 563 
 
 heavy tie-beam roofs on the tops of these slender piers would have been bad 
 construction. So a corbel was inserted lower down in the wall, e.g. at LOUTH 
 (213), and on this an upright post or wall-post (or pendant post) was placed, on 
 which the end of the tie-beam rested. Then, by the insertion of a strut, the 
 main weight of the roof was brought considerably lower down. Sometimes the 
 wall-post is omitted ; to the detriment of the construction and of artistic effect; 
 
 as in HULL CHANCEL (8l). 
 
 IV. Such a roof, however, was still very heavy, with its combination of 
 tie-beam, struts, and principal rafters, as at <;i-;r>XKY (551). The remedy 
 was to make the tie-beam serve also as principal rafter. This was a great 
 economy of wood and of weight, dispensing as it did at once with principals 
 and struts. The result was of course a very depressed roof, as at GRESFORD 
 (214); Haverfordwest; St Mary's, Shrewsbury; Lavenham ; Long Mel ford ; 
 Raunds and Rushden.* The gradual absorption of the principal rafter into the 
 tie-beam is well seen in early examples ; e.g. Kiddington and Sparsholt.t and 
 the south aisle of St Martin, Leicester,} which are all early in the fourteenth 
 century ; in all of them the tie-beam is exceedingly massive. 
 
 Purlins. — One more difficult}- about roof construction remains to be noted. 
 It is that the rafters, being slight, tended to bulge inward under wind pressure 
 and the weight of the tiles or lead by which the)- were covered. To obviate 
 this, the principal rafters or Principals over the tie-beams were made exception- 
 ally strong; and into these one or more horizontal timbers were framed (355, 
 SS8j. These are called Purlins. Then on the outside of the purlins were 
 laid the outer rafters ; which are called Common Rafters. Another service- 
 rendered by the purlins is that the) - prevent the rafters of a nave racking from 
 east to west. In Leicester St Mar}', Adderbury, North Walsham, Brinton, 
 there is one purlin on each side of the roof; at Gresford there are three. 
 
 ARCH-BRACED ROOFS. — As we saw above, the later tie-beams were sup- 
 ported by arched braces ; e.g. at Louth and Gresford, whose lower ends rested 
 on the same corbels as the wall-posts. It was only necessary to support the 
 collar instead of the tie-beam in this way, and a new and simple form of roof 
 was obtained. § At PATRINGTON (553.8), c. 1340, the principle is fully admitted; 
 the collar being very small and set very high up. 
 
 In such an example as Patrington the collar is reduced to a minimum ; the 
 next step was to abolish it altogether. At Tenby it was felt a dangerous thing 
 to do, the risk of spreading being great ; so the arched braces were constructed 
 exceedingly massive. Sometimes, however, little substance was put into them, 
 as in UGGLESHALL(S72)and BRINTON, NORFOLK ('564). At Tenby and Outwell 
 no precautions are taken against the spreading of the roof, which rests on 
 the aisle wall. At Brinton, however, the arched brace is fastened to a wall- 
 post ; and the thrust of the rafters therefore cannot dislodge the wall-plate 
 unless it also dislodges all the courses of wall against which the wall-post rests. 
 
 * Illustrated in Churches of Northants, 59, 185. 
 
 t Glossary, Plate 176. J Illustrated in Brandon's Roofs, Plate 8. 
 
 § These roofs are styled by Mr Brandon Collar-braced. The term is objectionable; 
 because not infrequently arch-braced roofs were constructed without collars ; e.g. brinton 
 (564). 
 
564 
 
 HAMMERBEAM ROOFS. 
 
 Thus the wall-post not only serves to bring the weight of the whole roof lower 
 down, as explained above, but also to stop the thrust of the rafters.* One 
 of the most audacious roofs of this type is that of SALL, NORFOLK (551) ; in 
 which the arched braces are minimised to a most dangerous extent.f 
 
 HAMMERBEAM ROOFS. — Arch-braced roofs are often assumed to be a 
 simplification of hammerbeam roofs. Chronology seems to be against this. 
 
 No hammerbeam roof is said 
 to exist earlier than that of 
 Westminster Hall, finished in 
 T 399-+ But arch-braced roofs 
 existed more than half a century 
 earlier than this. That of Pat- 
 rington transept can hardly be 
 later than 1340 ; earlier still was 
 the arch-braced roof of the 
 Worcester Guest Hall, now re- 
 moved to Holy Trinity Church, 
 near Shrub Hill Station ; this 
 roof was constructed in 1320. 
 Hammerbeam roofs did not 
 come into general use till late 
 in the fifteenth century.§ The 
 only real difference between an 
 arch-braced roof and a hammer- 
 beam roof is that in the former 
 (553.10) the sole-piece does 
 not project beyond the inner 
 face of the wall ; while in the 
 latter (553.1 1) it is elongated, 
 and may project considerably. 
 If in the first diagram a curved 
 brace be made to spring, not from A, 10, but from B, 1 1, the end of the sole-piece 
 elongated, the arch-braced roof will be converted to a hammerbeam (555). 
 Another difference, but of minor importance, is that advantage is taken of the 
 elongation of the sole-piece, CB, to erect a second upright or strut, EF, as shown 
 
 * That the wall-piece is not always put in to carry weight is clear from the fact that it 
 sometimes is employed without any corbel to support it ; in such cases its object can only be 
 to prevent spreading ; e.g. in the hammerbeam roofs of CAPEL st mary (555), Freslingfield, and 
 Bacton ; and the arch-braced roofs of Pulham and Starston (illustrated in Brandon's Roofs, 
 Plates 17, 31, 35, 27, 29) ; and UGGLESHALL (572). 
 
 + Vast spans, however, may be successfully roofed in this way. At Vicenza, the Basilica, 
 built 1314, has a roof of the shape of a pointed arch, with a span of 70 feet. The great hall at 
 Padua has a similar roof, with a span of about 85 feet; the ribs have a scantling of about 13 
 inches square. 
 
 % But in Carlisle choir the hammerbeams remain of 1375 to 1400 ; see Billing's Carlisle, 
 Plate 37; those of Boston must be still earlier; the aisle of Hingham, 1316-1359, has a 
 hammerbeam roof; but this maybe contemporaneous with the nave roof of 1664, destroyed 
 in a recent " restoration." 
 
 § Brandon's Roofs, 2 1 , note. 
 
 Brinton. 
 
HAMMHRBKAM ROOFS. 
 
 565 
 
 in 553.11, which still further prevents the rafters from spreading. The main 
 feature of the hammerbeam roof is the elongation of the sole-piece : in fa< I the 
 hammerbeam is nothing else than an elongated sole-piece. 
 
 This elongation ma)' be seen slowly going on. In many cases it is not 
 easy to say whether the roofs ought to be styled arch-braced or hammerbeam . 
 
 W'ymondham Nave. 
 
 e.g. CAPEL ST MARY (555). If the hammerbeam be but of slight projection 
 as at Capel St Mary and LLANIDLOES (552), it is comparatively ineffective. It 
 is only when it projects boldly some 5 or 6 feet, as at TRUNCH, CAWSTON (552), 
 and KNAPTON (567), that full advantage can be taken of its decorative capacities. 
 Another unfounded assumption finds the origin of the hammerbeam in a 
 
566 HAMMERBEAM ROOFS. 
 
 tie-beam with the centre cut out. For this there is nothing to be said except 
 that occasionally tie-beam and hammerbeam do occur in alternation, as in 
 PASTON BARN, NORFOLK (552); Weston Church, Lincolnshire; and else- 
 where. The real object of the elongation of the sole-piece is to provide a 
 couple of curved braces (of which the upper brace descends either from the 
 collar (555) or from the ridge (556) to the hammerbeam, and the lower from the 
 hammerbeam to the wall-post) in order that the rafters may be prevented from 
 spreading. It will be seen also that while some of these roofs omit the collar, 
 and others the struts, the hammerbeam and the curved brace are always present ; 
 evidently these are the essential elements. 
 
 We may arrange the hammerbeam roofs in order of complexity.* First, 
 some have neither collar nor struts ; e.g. WYMONDHAM NAVE (565) and Pal- 
 grave ; the upper part of the arched brace being strengthened so as to serve 
 at once as collar and brace. f Second, some have struts but no collar ; e.g. 
 CAWSTON (552), TRUNCH (553), and ST STEPHEN'S, NORWICH (556). Thirdly, 
 others have collar but no struts; e.g. CAPEL ST MARY (555); here the 
 hammerbeam has little projection. J Fourthly, the single hammerbeam roof, 
 with full complement of collar and struts ; e.g. Freslingfield, Upwell, Blakeney, 
 and Bury St Edmunds. Fifthly, the double hammerbeam roof; in which the 
 arched braces rise to the collar not in one, but in two flights. In these the 
 principal rafter is tenoned into the lower hammerbeam, but the upper hammer- 
 beam into the principal; e.g. Bacton;§ Grundisburgh ; || Woolpit ; *\\ Tostock ; 
 St Margaret's, Ipswich ; Weatherden ; Rattlesden ; all in Suffolk ; Swaff- 
 ham, Tilney All Saints, and KNAPTON (567), in Norfolk;** MARCH (566), in 
 Cambridge. 
 
 Outside East Anglia examples of single or double hammerbeams occur 
 chiefly in or near London ; e.g. Eltham Palace ; f f Beddington Hall, near 
 Croydon; Hampton Court Palace ; jj Westminster Hall ; §§ Middle Temple 
 Hall. I! [ They occur but sporadically elsewhere ; e.g. in the Law Library to the 
 north of Exeter Cathedral ; and at Llangollen and Gilgen, Denbighshire ; and 
 at Llanidloes. They remained in use till the seventeenth century, well into 
 the Renaissance period ; e.g. Hampton Court, 15 30; Middle Temple, 1570; 
 Wollaton, 1580; Trinity College, Cambridge, 1604; and Wadham College, 
 Oxford, 1612.HH 
 
 Cornice. — In East Anglia the cornice is more developed than elsewhere. 
 Owing to the local custom of dispensing with gutter and parapet, the rafters 
 were set far back from the inner surface of the wall ; e.g. in the trussed rafter 
 
 * Chronologically the simplest roofs may be the least ancient, 
 t Brandon's Roofs, Plates 20, 22. J Brandon's Roofs, Plate 17. 
 
 § Brandon's Roofs, Plate 35. || Brandon's Analysis, Plate 26. 
 
 1" Brandon's Parish Churches, 49. 
 
 ** The latter retains much of its colour. A reproduction of this roof in colour is given in 
 Brandon's Roofs, Plate 38. 
 
 tt Pugin's Examples, i., Plate 46. \% Pugin's Specimens, ii., Plate 8. 
 
 §§ Pugin's Specimens, i., Plate 32 ; and Viollet-le-Duc, Architecture, iii. 32, 33, 34. 
 III! Illustrated in Weale's Quarterly Papers, ii. 
 Ill" Gotch's English Renaissance, 161. A fine modern hammerbeam roof is illustrated at 
 HOLBEACH (133). 
 
MARCH, ROOF OF NAVE. 
 
HAMMKkBKAM ROOKS. 
 
 6; 
 
 roof of STUSTON porch (S54). This left a triangular space between the struts 
 and the sloping rafters. It was easy to board over the front of this space; 
 and in East Anglia a cornice of two or three parts is often thus formed in 
 
 Knapton. 
 
 different varieties of late roofs; sometimes carved, e.g. in the hammerbeam 
 roof of KNAPTON with square flowers and angels with outspread wings, some- 
 times molded. The cornice, instead of being vertical, was often inclined 
 
HAMMERBEAM ROOFS. 
 
 forward to be more directly opposite to the line of vision ; as in the cornice of 
 the roof of CAPEL ST MARY shown in section on 555. 
 
 Not only on the cornices, but at the ends of the bammerbeams, at the 
 foot of the wall-pieces, and erect on the collar beams, were often set angels 
 with outstretched wings, as at MARCH (566), Woolpit, Knapton, and Swaffham ; 
 some holding shields charged with the instruments of the Passion, others labels 
 with devout scriptures, others instruments of music ; spirits, as it were, arrested 
 midway in their heavenward flight ; hovering over, worshipping with the faithful 
 
 gathered below. 
 
 Arcuated 
 ROOFS. — Artisti- 
 cally, the note of 
 the later churches 
 was Unity. This 
 is just as clear at 
 March and Swaff- 
 h a m as in ST 
 GEORGE'S, WIND- 
 SOR (330). Be- 
 low were the 
 arches of the 
 pier- arcade ; 
 above, those of 
 the clerestory 
 windows ; all set 
 longitud inally. 
 Now, set trans- 
 versely, arch rose 
 upon arch in 
 many a roof as 
 well. Such a roof 
 in form was be- 
 come arcuated 
 instead of tra- 
 beated ; roof and 
 church, wood and 
 stone, were ar- 
 cuated alike ; the 
 whole was indeed 
 
 "a house at one with itself." See the illustrations of TRUNCH (552), WVMOND- 
 HAM (565), GEDNEY (551), KILKHAMPTON, NEW WALSINGHAM (571). 
 
 LONGITUDINAL BRACES. — As we saw above (560), there was a tendency 
 of the rafters to deviate from a perpendicular plane. Except in trussed rafter 
 roofs this was effectually guarded against by the provision of purlins and ridge- 
 pieces. But longitudinal arched braces are often inserted as well, even where 
 constructionally they are little needed. Thus at KNAPTON (567) they are in- 
 serted between the wall-pieces, and might be employed with perfect decorative 
 
 Kilkhampton. 
 
VARIOUS ROOFS. 569 
 
 propriety in such a position as a repeat of the heads of the clerestory windows 
 immediately below; as the)- are at TRUNCH (552; and ST STEPHEN'S, NORWK 11 
 (556). But they also occur between the principal rafters;* e.g. at Kllhaui, 
 Worcester Guest Hall, Bury St Edmunds. Lastly, if the span from one tie- 
 beam or collar to the next be considerable, a longitudinal arch may be erected 
 from the centre of one beam to the centre of the next, in order to support the 
 ridge-piece halfway ; as at ADDERBURY fSSSj. 
 
 Sometimes a horizontal strut was provided between the collar beams; as at 
 Charney, Sutton Courtney,! WIG< .enhall ST MARY (571 ) ; where it passes 
 below the collar and is supported by king-posts ; or as at FARINGDON (558), 
 where it is unsupported from below. Cf. Kingsland and Horton Kirkby. 
 
 COLOUR. — Wherever windows contain fragments of ancient glass, there the 
 roofs seem almost invariably to have been coloured. In Palgrave Church and 
 Ely transepts the colours are chiefly black, white, and red. In Long Melford 
 the rafters of the flat roofs are painted red ; the panels between them blue, 
 powdered with gold stars. The gorgeous roof of the nave of Knapton is mainly 
 yellow, green, and red. \ 
 
 Hybrid Roofs. — All the above — steep and depressed roofs with tie-beams, 
 trussed rafter roofs with six or seven sides, arch-braced roofs without or with 
 collars, hammerbeam roofs — are simple types. But frequently two or more 
 forms of construction are combined in one roof. Thus Outwell aisle, Weston 
 nave, and Paston barn have alternately hammerbeams and tie-beams. WOOD- 
 BASTWICK (551) has a combination of king-post roof and boarded rafter roof 
 with seven sides. Sandridge, Herts, has a trussed rafter roof with the addition 
 of a tie-beam to each twelve rafters. Clymping, Sussex, has a similar roof.§ 
 Pulham St Mary has an arched brace rising to the ridge-piece, and pinned 
 into the tie-beam. The roof of St Mary, Leicester, has arched brace and 
 collar ; but, having no ridge-piece, it partakes also of the character of a trussed 
 rafter roof." The roof of Solihull, Warwick, is a combination of trussed rafter 
 and arch-braced roof with collar. Inside the rafters and collar are fixed arched 
 braces. When a roof of this sort is boarded over, it produces a type which from 
 its frequency requires a special name. It is a Barrel roof. It occurs in Carlisle 
 choir, instead of the hammerbeam roof originally planned. Barrel roofs are 
 especially common in Somerset and Cornwall; e.g. KILKHAMPTON (568). It 
 will be noticed that in this church the aisles also have barrel roofs. A 
 room at Charney, and the hall of Sutton Courtney, Berks, and the church of 
 Wiggenhall St Mary, have a combination of king-post and seven-sided rafter 
 roof. The tie-beams at Higham Ferrers have both king-post and queen-posts.** 
 The roof of WESTMINSTER HALL ft (55/), with the vast span of nearly 69 
 
 * Here they have value as against wind-pressure. + See Glossary, Plates 171, 175. 
 
 \ For coloured plates of the above, see Brandon's Roofs; Paly rave, Plate 22 ; Knapton, Plate 
 38 ; and Colling's Gothic ( Imaments ; Long Melford, Plate 36 ; Bury St Edmunds, Plate 40. 
 
 § Illustrated in Brandon's Parish Churches, 75. II Illustrated in Brandon's Roofs, 21. 
 
 IT Illustrated in 558.2. ** Glossary, Plate 174. 
 
 t+ Westminster Hall is 68 feet wide internally : the height of the walls is 42 feet, and of 
 the roof 48 feet. Each truss is about 19 feet 10 inches wide, and has to carry, besides its own 
 weight, the timber and slates necessary to roof 1,359 feet of floor. The pitch of the roof is 52'. 
 It is of English oak, and was constructed in 1397. Professor Roger Smith in Builder, 48, 479. 
 
S70 
 
 AISLE ROOFS. 
 
 feet, is a combination of hammerbeam roof and arched-brace and collar roof, 
 with king and queen posts and numerous upright struts. 
 
 CEILINGS. — A few original ceilings remain. St Gregory's, Sudbury, has a 
 flat molded ceiling in the chancel. The most elaborate of all is that of the 
 nave of st DAVID'S (525). 
 
 AlSLE ROOFS. — If the aisle was broad, as at Wymondham, Outwell, and St 
 Martin's, Leicester, it often had a roof of double slope; i.e. a Span, Saddle, Gable, 
 or Compass roof. If the nave had a clerestory, the aisle roof was flattened in 
 order not to interfere with the clerestory lighting". If there were no clerestory, 
 all three span roofs might be equal in height ; e.g. at Old Basing, which may be 
 seen on the left near the railway before entering Basingstoke from the east. 
 A church so roofed corresponds in section with vaulted churches of the type of 
 the choirs of the TEMPLE CHURCH (35.1). 
 
 Usually the aisle had 
 a lean-to roof; i.e. of one 
 slope only. In many of 
 the earlier thatched 
 churches and barns the 
 span roof of the naves was 
 continued downward so as 
 to shelter the aisles as well 
 as the nave ; e.g. EAST- 
 WOOD (224) ; TOOT BALDON 
 (225).* This brought the 
 eaves nearly down to the 
 ground, and only the most 
 diminutive aisle windows 
 were possible. So in such 
 examples the slope of the 
 roof over the aisle is 
 generally a little flatter 
 than that of the nave ; in 
 order to get more headway 
 for the windows of the 
 aisle. The principal rafters of the aisle roof were often strengthened with struts 
 or with curved braces ; as at NEW WALSINGHAM (571). 
 
 Both in aisles and naves the rafters were often boarded over, and the 
 spandrels of the trusses have traceried panels ; e.g. in the aisle of NEW WAL- 
 SINGHAM and the chancel of Sail. 
 
 Hammerbeam roofs occur over the aisles of Upwell and Hingham. 
 Wood Vaults. — A considerable number of vaults were executed in wood, 
 either through distrust or dislike of stone vaulting, or inability to find money for 
 it ; especially in the North of England. At Selby and York heavy pinnacles 
 had been erected to receive the thrust of stone vaults transmitted by flying 
 buttresses ; but no stone vaults and no flying buttresses were ever erected (390, 
 10). Wood vaults appear c. 1300 in St Albans presbyter)-, and earlier still 
 * Compare Brandon's Roofs, Plate 2. 
 
 New Walsinsfham. 
 
ROOF CONSTRUCTION. 
 
 over Warmington Church. Winchester choir exemplifies their use as late as 
 15 10 to 1528. 
 
 St Peter Mancroft, Norwich, has a hybrid roof.* The lower part is a ribbed 
 vault in wood ; the upper part an arch-braced roof; so also at Framlingham. 
 
 Whether vaulted or semi-vaulted in wood, such roofs are objectionable as 
 being a reproduction in one material of forms which arose out of the nature of 
 another. 
 
 Material and Construction. — These roofs appear to have been con- 
 structed of oak-. It has often been asserted, e.g. by Evelyn and by Sir Christopher 
 Wren, that the sweet or edible Spanish chestnut was also employed, but papers 
 read before the Royal Institute of British Architects in 1 S 5 8 by Mr Wyatt 
 Papworth and in 1S78 by Mr Thomas Blashill show that the assumption is 
 
 Wiggenhall St Mary the Virgin. 
 
 New Walsinsdiam. 
 
 doubtful. The sweet chestnut is not an indigenous tree here; and it is very 
 
 unlikely that there were forests of it in mediaeval times. Neither Sir Gilbert 
 
 Scott nor Viollet-le-Duc came across any use of chestnut in the course of their 
 
 restorations. Mr Fowler, in a paper on Maulbronn Abbey, Wurtemberg read 
 
 before the Institute, mentions that the south transept retains its original roof and 
 
 ceiling of fir. , 
 
 The different timbers were always morticed and tenoned together, and fixed 
 
 with wooden pins ; no iron ties or straps, or even nails, being used in any part. t 
 
 According to Viollet-le-Duc,; none of the oak employed was of great 
 
 age ; sixty, eighty, or one hundred years at the most : it was cut before the heart 
 
 * Illustrated in Hickman, 280. t Brandon's Roofs, 32. X Architecture, 11. 215. 
 
S7- 
 
 OLDEST ROOFS. 
 
 had begun to decompose. First the tree was barked, and then cut. It was cut 
 in winter ; and then only during a certain moon. It was squared and piled, well 
 ventilated, under cover ; and seasoned for several years. 
 
 Great care was taken to ventilate the roofs. The tie-beams were laid on a 
 wall-plate ; not directly on the wall ; and as much of the surface of the timber 
 as possible was exposed to the air. When well ventilated, and protected from 
 damp and from contact with stone, the duration of well-seasoned oak is almost 
 illimitable. 
 
 Utrsrleshall from E. 
 
 Southwold Nave. 
 
 Oldest Roofs. — Of the Norman roofs perhaps the best preserved is that 
 above the Bishop's Palace at Hereford. At Winchester that of the south transept 
 is the original eleventh-century roof; and in the nave some of the Norman tie- 
 beams and rafters seem to have been re-used by Wykeham and his successors ; * 
 the length of the tie-beams in the nave was 45 feet ; they were 12 inches thick 
 and 20 inches deep. The ceiling of Peterborough nave is said, improbably, to be 
 Norman. The roof of Adel Church is described as Norman. t To the thirteenth 
 century belong the tie-beams of Old Shoreham ; which have tooth ornament. \ 
 
 * Colson, 8, 14, 16. 
 
 X Illustrated in Glossary, text 278. 
 
 t Assoc. Soc. Reports, xix. 1 10. 
 
Chapter XXXIX. 
 
 THE DOORWAY AND PORCH. 
 
 Lintel — Tympanum— Single and Double Doorways Norman and Gothic Doorways 
 Western Doorway — Priest's Doorway — Porch — Parvise— Doors -Rear Arch. 
 
 TRABEATED DOORWAYS. — The form of a doorway is in a measure controlled 
 by that of the door. And since it is as easy to construct an oblong door as 
 it is difficult to construct one with an arched head ; and since also there is 
 a difficulty about opening a door with an arched head (584), the doorways 
 of the Greek temple and the modern tenement are alike square-headed. This 
 square-headed opening may be bridged by a beam (Latin, Irabs) of wood or by 
 a lintel of stone. 
 
 But a long block so used will crack in the middle if the superimposed 
 weight is great. To protect the lintel, therefore, by carrying off as much 
 weight as possible to the right and left, the Romans turned over it what is called 
 a discharging or relieving arch ; which, however, they usually concealed under 
 the plaster or marble veneering of the wall.* So also in the fifth-century 
 Baptistery of Ravenna the lintel of the square-headed doorway is protected by 
 a discharging arch of brick, hardly visible in the wall. In s. ETIEXNE, NEVERS 
 (574), consecrated in 1099, a doorway of this simple type survives. The next 
 step is seen in the tenth century at the Baptistery of Biella ; where a single 
 slab is introduced between the lintel and discharging arch.t So that now the 
 doorhead consists of three elements — lintel, tympanum, arch. The tympanum 
 offered an excellent field of sculpture ; and became very popular in some dis- 
 tricts. Many of our Norman doorways have carved tympanums ; often Our 
 Lord is represented in majesty, with hand uplifted in the act of benediction of 
 all who enter the church ; as at Malmesbury, North Newbald, in the so-called 
 monks' doorway in Ely nave; and the west doorway of Rochester '574. 
 In Erench Gothic, e.g. at Chartres, the sculptured treatment of the doorway 
 received marvellous development. With us the art of sculpture, which in Wells 
 west front had shown equal promise with that of Chartres, fell more and mure 
 into the background. With the exception of the Last Resurrection sculptured 
 on the south doorway of Lincoln presbytery, England, like Italy, abandoned the 
 Erench form of doorway, with tympanum peopled with statuary. \ 
 
 * Willis' Nomenclature, 115. The accentuation of the Roman relieving-arch over a lintel 
 is found in the second century A.D. in Asia Minor and Syria (Anderson and Spiers, 259). 
 
 t Illustrated in Cattaneo and Rivoira. 
 
 \ Dehio, ii. 584, is of opinion that the design of these doorways ii non-architectural, and 
 moreover does injustice to the statuary ; but that, though disagreeable, it is very rich and 
 grand ; and was based not on artistic, but on religious motives. 
 
574 
 
 Iffley, W. Doorway. 
 Nevers, St Etienne. 
 
 Rochester, W. Doorway. 
 Beverley Transept. 
 
Hales. 
 Hinham. 
 
 Sail. 
 
 Trunch Chancel. 
 
5/6 
 
 Gimingham. 
 Tiverton. 
 
 Sherborne. 
 Lavenham. 
 
DOORWAYS. 
 
 i77 
 
 In all these doorways, as has been pointed out, there was risk of fracture 
 of the lintel. Sometimes this was guarded against, as at St Ftienne de 
 Xevers, by making the lintel thicker in the centre. Or a horizontal arch 
 was constructed ; i.e. a coursed lintel; sometimes with joggled joints. Or the 
 span of the opening was reduced by inserting corbels under the ends of the 
 lintel; this expedient is not employed in Normandy till the twelfth century, 
 except at Due y.* 
 1" he most obvious 
 way out of the diffi- 
 culty was to put a 
 pillar {trumeau) under 
 the centre of the 
 doorway, as was done 
 in the west doorway 
 ofVezelay; forming a 
 double doorway. A 
 later treatment of the 
 door deserves men- 
 tion ; it is character- 
 istic of Champagne ; 
 and consists in glazing 
 the tympanum ; e.g. 
 Reims Cathedral. 
 
 Single a n d 
 Double Doorways. 
 — But from the first 
 it seems to have been 
 not uncommon to 
 omit both lintel and 
 tympanum, as at 
 IFFLEY (574). It is 
 curious that Poitou 
 and Saintonge, al- 
 though sculpture was 
 very advanced, and 
 the tympanum offered 
 a fine field for it, 
 largely abandoned the 
 lintel and tympanum 
 type of doorway .+ 
 
 In the thirteenth century a central pillar or trumeau is 
 duced and a doorway with double arch is for the first time produced. Fine 
 examples appear at Higham Ferrers and St Cross, Winchester. These double 
 doorways are particularly characteristic of the chapter houses; e.g. Lincoln, 
 Southwell, Wells, York. Fine double doorways lead into the transepts .if 
 Lichfield; in the west front the inner doorway is double, the outer single. 
 * Ruprich-Robert, 124. + Viollet-le-Duc, Architecture, vii. 408. 
 
 2 1 1 
 
 I ill. \ South Doorway. 
 
 sometimes intro- 
 
578 
 
 NORMAN DOORWAYS. 
 
 In France the trumeaux of great numbers of these doorways were removed 
 in the eighteenth century ; probably because they were serious obstructions, 
 e.g. to carrying the shrine of a local saint, or to holding the baldachino over 
 the head of the bishop in processions. For similar reasons probably they' 'went 
 more and more out of fashion in England after the thirteenth century. 
 
 Warmington West Doorway. 
 
 NORMAN DOORWAYS. — On their doorways the Romanesque builders lavished 
 all their resources ; especially valuable to them was the principle of recessing 
 the orders. Arch was built upon arch to gain strength ; and when all the 
 strength required was gained, they still went on building more arches, as 
 a field for decorative carving ; till sometimes, as in the doorways of Malmesbury 
 
GOTHIC DOORWAYS. S79 
 
 and S. Fermo, Verona, the arch consisted of no less than eight orders. Even 
 in the thick Norman walls there was not always room for arches of such 
 great breadth. So frequently the wall was built additionally thick all round 
 the doorway; e.g. at Iffley. This projecting space of wall was .sometimes 
 carried up to a gable; as at SEMPRINGHAM (40) and Kirkstall.* lint 
 the gabled doorway never received here the great development given to it 
 in France. We have nothing comparable with the five solid gables of tin- 
 western doors of Bayeux ; with those of Bourges, pierced with plate-tracery ; 
 still less with the gables of open-work tracery of the Porte de la Calende, 
 Rouen Cathedral. Even more superb arc the fifteenth and sixteenth century 
 gabled doorways of Abbeville and Tours ; of the transepts of Sens, Beauvais, 
 and Senlis ; of St Maclous, Rouen ; and the sculptured gables of the cathedral 
 of Troyes. 
 
 Of Norman doorways very numerous examples survive. Windows of the 
 eleventh, twelfth, and thirteenth centuries had to be taken out to insert 
 larger ones. But the doorways were generally large enough. At Birkin, 
 when an aisle was added in the fourteenth century, the Norman doorway 
 was taken down and reset ; so also at Kenilworth. In the same way 
 thirteenth-century doorways were reset at Louth and Leominster when the 
 aisles were widened. The doorway described by William of Worcester as 
 on the north side of St Stephen's, Bristol, is now on the south side.+ 
 
 The doorway of the last half of the twelfth century usually has a semi- 
 circular head ; but in the larger examples, e.g., Ketton, it may be flanked by 
 a blank arcade of pointed arches. 
 
 GOTHIC DOORWAYS. — In the first half of the thirteenth century full 
 advantage is taken of the recessing of the orders, just as much as in Norman 
 work ; e.g. at Elgin. This was the period that took most delight in shafting 
 and in wealth of marble monoliths. Now, too, the double doorway had its 
 vogue. Aided by profusion of tooth ornament and of small, deep, delicate 
 moldings, and by exquisite foliage, as at West Walton, SKELTON (78), and 
 WARMINGTON. (578), the doorways have surpassing charm. No more beautiful 
 examples exist than the doorways and arcading of the west front of BINHAM 
 (575) ; executed in an admirable stone. 
 
 Later in the century doorways are more commonly single than double. 
 In the richer examples, e.g. the west doorway of York, the pointed arch is 
 usually surmounted by a straight-sided gable. 
 
 Doorways of the fourteenth century often employ an ogee instead of a 
 straight-sided hood-mold ; e.g. at CLEY, NORFOLK (85). In this rich example 
 crockets are used profusely, and the hold-mold rises into a floriated finial ; 
 the cusping is compound ; the foliage bulbous.^ 
 
 In later Gothic a special characteristic of the doorways is that the arch is 
 
 * Late in the thirteenth century this straight-sided gable plays a very important part, 
 appearing not only over doorways but over windows, niches, and arcades ; e.g. in York nave 
 and chapter house, and howden west front (72). 
 
 t See Willis' Nomenclature, 6. 
 
 \ A rich doorway, of somewhat unusual type, at Heckington, is illustrated in Bowman and 
 Crowther, C/iu?r//cs, Plate 31. 
 
S8o 
 
 DOORWAY ARCH. 
 
 Tenby. 
 
 Llandaff Choir Arch. 
 
 often framed in a horizontal 
 hood-mold {i.e. a label) ; e.g. at 
 KETTERING (95), Fotheringhay, 
 Cawston, Tiverton, Sail. BEVER- 
 LEY ST MARY (365) is an excep- 
 tion ; but it is very early in the 
 period.* In the spandrels at 
 SALL (575) are censing angels ; 
 the door is original. 
 
 Form of Doorway ' Arch. 
 — In Norman work this is usually 
 semicircular, but other forms may 
 occur ; e.g. for want of space the 
 semicircular arch becomes seg- 
 mental in the south transept of 
 Southwell. So also in the thir- 
 teenth century, where a doorway 
 has to be placed under a window, 
 it often becomes semicircular 
 instead of pointed ; e.g. in the 
 transept of BEVERLEY (574). 
 Even in the fourteenth century, 
 semicircular arches may occur ; 
 e.g. at Badgeworth, Gloucester, 
 where the doorway is ornamented 
 with ball-flower. 
 
 Doorways with pointed arch 
 occur till the end of Gothic, e.g. 
 at Tiverton, Sail, Kettering ; 
 though in late work a depressed 
 four-centred arch is frequent. 
 
 Trefoiled arches occur in 
 Norman doorways ; e.g. the so- 
 called Prior's doorway at Ely ; 
 Mathon, Hereford ; and Nately, 
 Hants. About 1170, in Ripon 
 north transept and Byland west 
 front. About 1250, at WARMING- 
 TON (578) ; it is rare in later 
 work. In and after the fourteenth 
 century the doorway sometimes 
 has an ogee arch ; e.g. the west 
 doorway of TENBY. In some 
 districts, e.g. Northants, it is 
 extremely common. 
 
 * A rich example from lavenham 
 (576) is illustrated in Brandon's Ana- 
 lysis, Perpendicular, Plates 6 and 7. 
 
WEST DOORWAYS. 
 
 ;8i 
 
 West Doorways. — 
 The side entrance for the 
 laity was in one of the 
 western bays of the nave. 
 In most churches there were 
 doorways both on the north 
 and the south side of the 
 nave. Probably this was 
 for processional * purposes. 
 In most of the smaller and 
 earlier parish churches there 
 was no western doorway ; 
 e.g. in none of the eighteen 
 Lincolnshire churches visited 
 by Mr Sharpe in 1870 was 
 there a western doorway ; 
 except at Helpringham. 
 But almost all the larger 
 churches* had a large 
 western doorway. This was 
 the great ceremonial door of 
 the church, and in many 
 churches remains so to this 
 day ; only opened on some 
 great function, as the visit 
 of the sovereign or the bishop 
 or for a wedding or a funeral. 
 In mediaeval days it took a 
 leading part in the important 
 ceremonies of Palm Sunday, 
 which may still be seen 
 abroad ; e.g. at Troyes. On 
 that day half the choir is 
 outside the west door, and 
 sings the " Attollite Portas " 
 — " Open ye the gates that 
 the King of Glory may come 
 in." Then comes the ques- 
 tion from the choir within, 
 "Who is the King of Glory?" 
 Finally the great doors are 
 opened, and the procession 
 enters, symbolising Christ's 
 entry into Jerusalem. 
 
 * A recess for a tall processional cross is found in many churches of East Suffolk, usually in 
 the thickness of one of the walls of the western tower ; e.g.sX St Margaret, Lowestoft, and at Cromer. 
 
 t Romsey is an exception ; but it was a nunnery church. 
 
 
 „J£Z 
 
 .■■N^HAf 
 
 Whissendine. 
 
5 82 
 
 PORCHES. 
 
 In later days nearly all the larger parish churches seem to have had 
 western doorways. This led to difficulties. For the later builders built usually 
 their western windows as large as possible, to get all the western light they 
 could for the nave.* The result was that there was no room for a doorway 
 of scale proportionate to the magnitude of the west front or west tower in 
 which it was placed. Attempts to cope with this difficulty may, however, 
 occasionally be seen, e.g. at WHISSENDINE (581), Hickling, and Ingham. 
 
 Priest's Doorway. — On one side of the chancel, usually the south, was 
 placed a doorway, usually small and simple, but sometimes, as at Simonburn, 
 highly ornate, to give the priest independent access to the chancel ; e.g. TRUNCH 
 (575) ; where in curious fashion its porch is formed of a split buttress. 
 
 PORCH. — Just as the nave was more secular than the chancel, so the porch 
 
 was more secular than the 
 nave ; e.g. at Patrington "forty 
 pounds was to be paid on 
 March 25, 161 3, within the 
 south porch of the church " t 
 (215.11). 
 
 A considerable number 
 of Norman porches remain ; 
 e.g. Castle Ashby, Durham, 
 Kelso, Southwell, Mal- 
 mesbury, SHERBORNE (576), 
 Adel, Witney, Morwenstow, 
 Balderton. Beautiful ex- 
 amples occur in all the 
 Gothic periods, j The porch 
 
 at LEVERINGTON (84) shows 
 
 fourteenth-century design 
 at its best. In East Anglia 
 especially, when everything 
 had been done to make church 
 and tower as glorious as 
 possible, the parishioners set 
 to work to build a porch : 
 e.g. GIMINGHAM (576) and 
 LAVENHAM (576). And as this was, comparatively, a small expense, it was 
 often built more sumptuously than the church itself; e.g. while the church walls 
 might be of undressed or of split flints, the porch walls were of squared flints ; 
 and sometimes, as at North Walsham, Norfolk, the flints were carefully 
 arranged in zones of different colours ; often also the porch was traceried 
 with patterns of different coloured flints. And while the church was roofed 
 in wood, the porch was often vaulted, as in both the porches of Sail. 
 
 * Wellingborough, illustrated in None Valley, Plate 104, is an exception ; the west window 
 being a small traceried circle. 
 
 T Mem. of Patrington Manor Court. 
 
 % See especially the E.E. porch of West Walton (illustrated in Colling's Details, i. 14). 
 
 Boxford. 
 
PORCHES. 
 
 583 
 
 BEVERLEY ST MARYS 
 (365) has a compara- 
 tively simple, TIVER- 
 TON a late and florid 
 porch (S76). HERE- 
 FORD indulged in the 
 luxury of a double 
 north porch ; the inner 
 porch built c. 1288, the 
 outer c. 1520 (203 ). 
 
 In the churches of 
 the monks and of the 
 Regular Canons there 
 was a cloister on one 
 side of the nave ; and 
 their entrance into the 
 church was from the 
 cloister. Consequently 
 in these churches the 
 doorway of the laity 
 was on the opposite 
 side to the cloister. 
 Now the normal posi- 
 tion of the cloister was 
 south of the nave. 
 Therefore the majority 
 of the doorways of the 
 laity are on the mirth 
 side ; e.g. at Bayham, 
 Bolton, Durham, 
 Lanercost, Malvern, 
 Peterborough, Romsey, 
 Selby, Tewkesbury, 
 Waltham, West- 
 minster, Whitby, Wor- 
 cester. On the other 
 hand the cloister some- 
 times had to be placed 
 on the north side ; in 
 these cases the lay 
 doorway is on the 
 south ; e.g. Canterbury, 
 Chester, Dorchester, 
 Gloucester, Malmes- 
 bury, Monkton, Sher- 
 borne. Some, however 
 porch ; e.g. Fountains, Byland, Glastc 
 
 of the monastic churches had a 
 
 nbury, Ely, St Albans, Winch 
 
 narthex or western 
 ter ; others 
 
5 8 4 
 
 PORCHES. 
 
 had merely western doorways ; e.g. Norwich, Rochester, Netley. But some 
 of the cathedrals served by Secular Canons also indulged in the luxury of a 
 cloister ; in such cases the side doorway again was on the side opposite to the 
 cloister ; at Salisbury, Wells, Hereford, Lincoln ; in all of which it takes the form 
 of magnificent porches; at Lincoln* on the south side, the other three porches on 
 the north. Turning to the other greater churches of the Secular Canons, we find 
 the entrance on the north side ; e.g. Beverley ; Christchurch ; Chester St John's ; 
 Southwell ; YVimborne ; Ripon ; all of which, except Ripon, have north porches. 
 Exceptional are Chichester, which has three porches, south, north, and west ; 
 Lichfield, which has doorways to the north and south transepts ; and York, the 
 side entrance to which is in the south transept ; while in Benedictine Westminster 
 it is in the north transept. 
 
 As to the porches of the parish churches, if the church was large or a town 
 church, or if there were habitations on both sides of the church, there were often 
 both north and south doorways, as at Louth and Sail. If the village was to the 
 south of the church,+ as for some mysterious reason it seems more often to have 
 
 been, the doorway would be 
 on the southern side of the 
 nave ; otherwise, on the 
 northern. 
 
 The porch was often 
 two stories high. At Wed- 
 more, Bruton, and Cirencester 
 the porch is three stories 
 high ; at Cirencester it was 
 the property of the trade 
 gilds. This upper room is 
 sometimes called a parvise. 
 It was used for various pur- 
 poses. At Southwell it has 
 a chimney. In such cases it 
 may have been the room of 
 the sacristan, who was bound 
 in some cases, e.g. at Southwell, to be on guard all night in the church. Some- 
 times it has a piscina ; and then may have been an altared chapel ; as at 
 Fotheringhay and Sail. Sometimes there is a squint ; blocked up or still left 
 open. At Warwick the south porch had a room built over it in 149 1 as a 
 library. I After the Reformation it was often used as a church library. Or, 
 as at Patrington, it became a registry of the parish books ; § but originally 
 probably the sacristan's chamber. 
 
 Lutton. 
 
 Hieham Ferrers. 
 
 * There seems to be no mediaeval authority for the employment of the term Galilee of this 
 Lincoln porch. 
 
 + The south aisle is generally of more importance than the north ; more often it is broader ; 
 and therefore more often has a span roof; the porch is more often on the south ; when there 
 is but one porch, it is generally on the south ; as also are the church-well or brook, the lych- 
 gate, and the yew trees (Brandon's Open Timber Roofs, 30). 
 
 1 Leland in Glossary, 343. § Churches of Yorkshire. Patrington, 10. 
 
DOORS. 
 
 WOODEN PORCHES. — In timber districts many charming exampli ol 
 wooden porches and lichgates remain: e.g. at BOXFORD (582), Aldham, and 
 Warblington.* 
 
 DOORS of all periods survive. The doors at Castor and skm PR] \< ,1 1 a \i 
 (40) are Norman; that of Castor retains the lock and key; round the edge 
 is an inscription that the Rector, Richard Beby, had it made; that was about 
 1 1 33.+ The doors of Sempringham and of York chapter house arc of deal. 
 The great western doors of Peterborough are early thirteenth century ; and 
 inside have tooth ornament. Late in the thirteenth century art' the doors of 
 York chapter house and St Margaret's Chapel, Herts. Fourteenth-century 
 doors remain at Milton, Kent; North Mimms, Herts; Stoke, Suffolk ; Wells; 
 HOLBEACH, Lincolnshire, and CI.KY, Norfolk, are illustrated in 583 and 85. 
 Many fifteenth-century doors remain, e.g. at SALL (575). Doors at St Mary's 
 Beverley, Helmingham, North Petherton, Stratford-on-Avon, and Tempsford, 
 are illustrated in Colling's Details ; and an earl} - example from St Albans 
 in Brandon's Analysis. 
 
 Hinges survive of all periods ; e.g. Norman at Sempringham ; fourteenth- 
 century at Cley. See especially J. Starkie Gardiner's Ironwork, vol. i. 
 
 REAR-ARCH. — -When a doorway with an arched head was set near the front 
 of the wall, it was necessary to make the rear-arch segmental in form; e.g. 
 HICHAM FERRERS (584) ; in order to allow the door to open. 
 
 * Illustrated in Glossary, Plate 164. t Paley's Gothic Architecture, 197. 
 
Chapter XL. 
 
 TOWERS. 
 
 Early Towers — Central Towers — Western Towers — Position of Towers — Tower Groups — 
 Plan of Towers — Round Towers — Tower Construction — Tower Roofs — Tower 
 Design. 
 
 Antiquity of Towers. — Both among the Greeks and the Romans towers were 
 in use. In the south of the Peloponnesus at Messena, at Phigalia, in the isle of 
 Andros, and at Paestum, there are well-constructed towers, both square and 
 round, several of which are in a fair state of preservation.* Many Roman 
 towers remain ; e.g. a round tower at Perigueux ; a square tower outside 
 Autun ; the Pharos at Dover. It has been supposed, however, that bells of large 
 dimensions and belfries to carry them were not employed in Western Europe till 
 late. This, however, is a mistake. Fleuryf has proved definitely from documents 
 that bells of very large dimensions, and towers to hold them, were in use in the 
 sixth century. To that century perhaps \ belong several of the towers of 
 Ravenna ; most of them round, some square. A mosaic of 432 to 440 in S. 
 Maria Maggiore, Rome, shows two churches near which are round towers. On 
 the doors of S. Sabina, Rome, which are early in the fifth century, is shown a 
 church with two western towers. § Two low western towers were built in the 
 Syrian churches of Tourmanin || and Qualb Louzeh. 
 
 Of Romanesque towers one of the oldest is that of S. Satiro, Milan, A.D. 
 879 ; of campanile type ; of four stories, with central-balustered window, pilaster 
 strips at the angles by way of buttresses, and a rudimentary kind of nebule 
 corbel table under each of the three strings.lf There are many fine campaniles 
 in Rome ; they have been given early dates ; ** but are so highly developed in 
 design that none of them are likely to be earlier than the eleventh century, not 
 even their lowest stages. ff The Irish round towers also are usually greatly 
 over-dated. Their date is practically certain; viz. from 890 to 1 238.JJ In 
 France we have documentary evidence of the existence of church towers in very 
 early times ; e.g. at St Martin, Tours, consecrated 477 ; one of grandiose design 
 at Nantes, consecrated by St Felix, 550 to 600 ; one at Narbonne, so lofty that 
 
 * See illustrations in Laloux's L 'Architecture grecque, 263, 266. 
 
 + La Messe. % Cattaneo, 235. 
 
 § Enlart's Manuel, 124. || Illustrated in Fergusson, i. 427. 
 
 1 Illustrated in Cattaneo, 236. Rivoira, 51, regards Viterbo and San Ambrogio, Milan, as 
 possessing the earliest examples of church towers ; both of the ninth century. 
 
 ** Mr J. Tavernor Perry in a fully illustrated paper in the. Journal of the R.I.B.A., Feb. 21, 
 1898, dates them from the beginning of the ninth to the end of the eleventh century. 
 
 tt Cattaneo, 157. \\ See Miss Stokes' Ecclesiastical Architecture of Ireland. 
 
S3; 
 
 Oxford St Michael. 
 Manorbier. 
 
 Ely Cathedral from S.W. 
 Hereford Cathedral from N'.W. 
 
588 
 
 Magdalen College, Oxford. 
 Glastonbury St John. 
 
 Canterbury Cathedral. 
 Wells Cathedral. 
 
589 
 
 
 I 
 
 
 
 Cawston. 
 
 Meiton College, Oxford. 
 
 Lavenhani. 
 Wymondhani. 
 
590 CENTRAL TOWERS. 
 
 Alaric II. had the upper stage demolished because it blocked the view from his 
 palace; two others, one at Bordeaux, built c. 550; the other at Notre'Dame, 
 Paris, appear to be indicated by the descriptions of the poet Fortunatus.* Pope 
 Stephen II., 752 to 757, built a tower at the Vatican, expressly for bells. 
 
 Of Anglo-Saxon towers we possess a great number ; a few of which incline 
 to the tall and slender campanile type ; e.g. ST MICHAEL'S, OXFORD (587) ; 
 and several in Lincolnshire and the North of England ; e.g. Monkwearmouth. 
 At Brixworth only the lowest stage of the Anglo-Saxon tower appears to be 
 contemporaneous with the nave, i.e. c. 680. It seems probable that no Anglo- 
 Saxon tower is earlier than the end of the ninth century ; f and that Anglo- 
 Saxon central towers, such as that of St Mary in the Castle, Dover, are but little 
 anterior to the Norman Conquest. 
 
 Central Towers. — The central tower is a natural outcome of the cruci- 
 form plan ; as facilitating the roofing of a church with transepts, especially 
 when the transepts are lower than the nave and choir (see 196). At Clermont- 
 Ferrand there was built in the year 510, in a church of secondary rank, a central 
 tower " with a great luxury of materials and profiles." Several others of like 
 date are described by Fortunatus of Poitiers. \ We possess an illustration of the 
 primitive abbey of St Riquier ; in which the central tower is crowned by an open 
 pavilion or louvre. § 
 
 In Anglo-Saxon days it clearly existed at Ramsey Abbey, A.D. 969. || And 
 cruciform Anglo-Saxon churches with central towers still exist at St Mary's in 
 the Castle, Dover ; Norton ; Stow ; Breamore ; and St Mary's, Guildford. 
 Anglo-Saxon central towers also were built in churches without transepts ; 
 e.g. at Barton-on-Humber. In Normandy churches of this last type are rare, 
 but there is one at Englesqueville.^ In England Norman examples occur at 
 Studland in Dorset ; Basing, South Hayling, and Sopeley, Hants ; Bredon and 
 Iffley in Mid-England, and at Fairford, Kempsford, Tong. But a central tower 
 is naturally more common over a cruciform church. In Normandy in the 
 eleventh century there was hardly a cruciform church but had a central 
 tower. In the twelfth century, in Normandy, there are a few examples of 
 Norman western towers ; the central tower was much more common. In 
 England Norman central towers were less common.** In some cases perhaps a 
 central tower fell, or threatening to fall was pulled down, e.g. at NEWARK (178). 
 Often it seems not to have been rebuilt ; j~f but, instead, a western tower was 
 built. Wherever aisles were being added or broadened, and an open arcade 
 was substituted for a solid wall, the abutments of the central tower were rendered 
 less stable ; and there was risk of a collapse. Nevertheless, in the cathedrals 
 and abbey churches with few exceptions, Llandaff is one, central towers con- 
 
 * Anthyme St Paul, Histoire Monwnentale, 60, and M. De Rossi in Revue de PArt Chretien, 
 1890, page 5. 
 
 + Baldwin Brown, Arts in Early England, 286, 246. 
 % A. St Paul in Planat's Encyclopedic, article Architechire Romane. 
 § Illustrated in Enlart's Manuel, 173. 
 
 || See description quoted in Scott's Lectures, ii. 33. IT Ruprich-Robert, 101. 
 
 ** For list of towers which fell, see Builder, xix., 1861, April 13. 
 
 tt At St Mary Redcliffe, Bristol, and at Great Wilbraham, Cambridge, the lower courses 
 of the old central tower may still be seen beneath the roof of the crossing. 
 
CENTRAL TOWERS. 
 
 59" 
 
 tinued to be employed; Bath Abbey has one of the latest.* In the parish 
 churches, on the other hand, the central tower went more and more out of 
 use; except in some of the greater churches, such as patkington 612), 
 HULL (96), Nantwich, Doncaster, Rotherham. 
 
 Holme. 
 
 Isle Abbots. 
 
 In Continental Romanesque the practice varied. Where it was insisted 
 * Kings College Chapel, Cambridge; St George's, Windsor; Eton College Chapel ; 
 Henry the Seventh's Chapel, Westminster, were indeed built without towers . but they are all 
 chapels ; and the absence of a tower was at times a distinguishing mark of a chapel. 
 
592 CENTRAL TOWERS. 
 
 that the whole church should be made fireproof by vaulting, the broad space 
 over the crossing was vaulted with a dome. But the dome itself required to be 
 protected from the weather ; and so on the same arches which supported the 
 dome there was often run up a central tower ; which not only protected the 
 dome from the weather, but also weighted and steadied its haunches, as a 
 pinnacle does a buttress. These towers over domes were sometimes square, 
 as at Tournus ; more often octagonal, and crowned with octagonal spires, as at 
 Paray-le-Monial.* But we had no crossing-domes in England, till the present 
 St Paul's ; and consequently no such use of central tower. Spain developed 
 a fine series of central domes, not masked by towers, at Salamanca, Zamora, 
 and Toro. 
 
 In Normandy, however, a much earlier type of central tower was followed, 
 that of St Riquier, surmounted by an open louvre. As to the object of this 
 arrangement, there can be no doubt whatever. In the sixth century Gregory 
 of Tours and Fortunatus of Poitiers, speaking of the central lantern towers of 
 Clermont, Nantes, Narbonne, and Paris, designate them turris domus altaris 
 or domus ctrae : i.e. the tower was placed over the high altar; and was pierced 
 with windows, that on this, the most important spot in the whole church, a 
 flood of light might pour from above, especially at the most sacred moment 
 of the Mass, that the Elevated Host might be conspicuous afar, even to the 
 remotest recesses of the nave. It has been urged that this central lantern of 
 wood may be a survival of the tradition of the lanterned domes of S. Sophia, 
 Constantinople, and S. Vitale, Ravenna ; but chronology is against the hypo- 
 thesis. These two Byzantine churches were built in the sixth century ; but 
 there was a central lantern at St Martin, Tours, as early as 477. Again, if the 
 central lantern emanated from or through Ravenna, we should expect to find 
 many early examples of it in Northern Italy. But it is just in Italy that they do 
 not occur ; except in the basilica of the Crucifix at Spoleto. In Romanesque 
 times they became common in many districts ; e.g. in Lombardy, Germany, 
 Auvergne, Burgundy, Laonnais, Normandy, England, Germany, Spain. Early 
 examples in France are Germigny - des - Pres, consecrated in 806 ; St Martin, 
 Angers, which is variously dated at 819 or 1020; Ouerqueville in Normandy, 
 and St Saturnin, tenth century.f 
 
 It is quite possible that the churches on which these louvred central towers 
 were erected were in the shape of the crux commissa (195); i.e. with apse 
 joined directly on to the central tower without the interposition of a choir. 
 In such a church the clergy would be seated round the apse, and the altar would 
 stand on the chord of the apse ; as it still does at S. CLEMENTE, ROME (3). 
 Such an interior, with the dramatic contrast of shadowy nave and brilliantly 
 lighted crossing, would be exceedingly impressive ; and we may be permitted 
 to regret that it has passed out of mediaeval and out of modern practice. Here 
 and there, in Continental churches, such an interior has come about, somewhat 
 fortuitously ; e.g. at St Eusebe, Auxerre, where the contrast between the 
 brilliant Late Gothic choir with clerestory soaring out of sight and the low and 
 ■dark nave is strikingly effective. 
 
 * Classification of Romanesque, 283. 
 
 + See Enlart's Manuel, 123, 124 ; and A. St Paul's Histoire, 60. 
 
CENTRAL TOWERS. 
 
 But the lantern that was effective in a crux commissa was less effa tivc in a 
 crux immissa, i.e. where there was a choir east of the transept. At first, indeed 
 the Norman choirs were short; those of Ccrisy, Lincoln, Selby, and Can- 
 terbury were but two bays long; and the central lantern was still of some 
 service in illuminating the altar and the stalls in front of it. But most of tin- 
 Anglo-Norman churches were planned, like Norwich and St Albans, with choirs 
 four or more bays long. And in Gothic days these choirs were elongated still 
 more, as at Ely, Lincoln, and Old St Paul's, till the altar was far away from 
 the transept, and unapproachable by any light from a central lantern. This 
 once recognised, we find monks and canons shutting up the central lanterns. 
 In Gothic days indeed the fenestration of end walls and side walls had been 
 increased to such a vast extent that supplementary lighting from a central 
 lantern was supererogatory. Accordingly at Winchester the Norman lantern, 
 in spite of its beautiful arcading, was shut off from view by a wooden vault ; 
 at Wells a fan vault was put under the late twelfth-century tower ; at Gloucester 
 the lierne vault of the choir was continued westward till not a vestige of central 
 tower appears within the church. Beautiful as is the internal effect of a central 
 tower, it seems to have been held of little account in later Gothic days. Not 
 that the liking for a central tower, open to the transept, ever wholly disappeared ; 
 beautiful late examples remain ; e.g. York Minster, Hedon, Howden, East 
 Dereham.* And in many vaulted lanterns, care was taken to place the vault 
 above, not below, the lantern ; e.g. at LINCOLN (3-8), Coutances, Evreux, and 
 Canterbury. 
 
 But for this feature, of supreme artistic value, even more externally than 
 internally, a heavy price had to be paid. For its supports, the four isolated legs 
 on which its arches rest, in such a cathedral as Winchester or Wells, it was 
 necessary to have piers of enormous area, very largely obstructive to the chief 
 use of the church ; viz. the view of the elevation of the Host by the largest 
 possible number of worshippers. Rather than have such huge obstructions the 
 canons of Bourges and Llandaff dispensed with a transept altogether ; while 
 those of St Leu d'Esserent, Le Mans, Barcelona, EXETER (i 54.4), and Ottery 
 St Maryj instead of one central tower, built a couple of towers, one at the 
 end of each transept. In most cases, however, the central tower was erected ; 
 by the cautious only so high as to provide a ring of windows ; as at BEVERLEY 
 MINSTER (176), WESTMINSTER (379), Winchester, and Bristol; by the bolder 
 so as to be of imposing height and bulk; as at St Albans, St David's, NEW 
 SHOREHAM (573), and TEWKESUUKY (390); by the venturesome crowned with 
 a wooden spire, as originally at Ripon, Hereford, and Lincoln ; by the rash 
 surmounted by a spire of stone, as at Norwich, Chichester, and Salisbury. 
 Being built moreover with a thick core of rubble and mortar, + they were most 
 unnecessarily heavy, exerting a pressure of from 5 to 20 tons to the square 
 
 * Both at Canterbury and York the original Norman central towers are still there, cased up 
 in later towers. The original Norman arches of the crossing were taken out and replaced by 
 the present Gothic arches without taking down the old towers resting on them. 
 
 + St Stephen's, Vienna, also has transeptal towers ; but only one completed. 
 
 I This has often decomposed into a fluid powder. See Cottingham's report on the 
 condition of the central piers at Hereford. 
 
 2 V 
 
594 CENTRAL TOWERS. 
 
 foot ; * yet sometimes with foundations of little depth or spread. The result was 
 that some, like Wells and York, tore away from nave, transept, and choir, burying 
 themselves deeper into the ground ; some collapsed soon after they were built, 
 e.g. the first Norman central tower of Winchester and the first Gothic tower of 
 Lincoln ; from others their wooden spires had to be removed, as at Hereford, 
 Lincoln, and Ripon ; Chichester central tower telescoped within the memory of 
 living men ; -f- others, like those of St David's, Sherborne, Peterborough, St 
 Albans, Hereford, Rochester, Salisbury, have had to be vastly strengthened and 
 under-pinned or completely rebuilt. The following is the diameter of some 
 central tower piers : — 
 
 Canterbury - - -12 feet. Worcester 9 feet. 
 
 York - - - 10 „ Peterborough - - 7 „ 
 
 Winchester - - - 10 „ Salisbury - - - - 7 „ 
 
 Norwich - - - 10 „ Rochester- - - - 6 „ 
 
 At Winchester, after the fall of their first tower, an area of 78 J feet was 
 given to the piers of the new tower, which rests on arches of 30 feet span ; at 
 Peterborough, the area was only 38-i feet, though the span of the arches was 35 
 feet. I In France, in the Domaine Royale, all the constructional objections to a 
 central tower existed in an aggravated form. Their cathedrals were nearly 
 twice as high as those of England ; any central tower, to dominate roofs so vast 
 in height, must rise at least to 500 or 600 feet, like the new western spires of 
 Cologne Cathedral. That was to impose enormous weight on the arches and 
 piers beneath. The builders shrank from it. So in the Domaine Royale they 
 were satisfied, as at Amiens and Notre Dame, Paris, to crown the crossing with 
 a light openwork spire of wood : a mere pinnacle in scale in comparison with 
 the bulk of these enormous cathedrals. At Beauvais, indeed, a spire was 
 erected of the height of 486 feet; but it fell down in 1573 after a life of but 
 five years. 
 
 In some cases there was no western tower at all ; but simply a central 
 tower. When this is placed midway, as at Worcester, the effect is not quite 
 happy. Where the nave is considerably longer than the choir, as at Gloucester 
 and Bath, the tower groups much better. But, even at Salisbury, the effect is 
 somewhat parochial ; from association one distinguishes a cathedral from a 
 parish church by its plenitude of towers. Contrariwise, when one comes across 
 a parish church with two or three towers ; as St Margaret's, Lynn ; Ottery ; 
 Beverley Minster ; Melbourne ; one expects them to be something more than 
 parochial ; usually they are or have been collegiate. § 
 
 * See paper by Mr S. B. Beale in Builder, Jan. 11, 1890. The tower of St David's is 
 estimated to weigh 4,000 tons. 
 
 + At Gloucester the S.W. tower fell between 1164 and 1 175 ; at Worcester the " nova turris 
 fell in 1 175 ; at Evesham the central tower fell in 1213 ; at Dunstable two western towers fell in 
 1221 ; at Worcester two small towers were blown down in 1222 ; the tower of St Radegund, 
 Cambridge, fell in 1270 ; the central tower of Ely in 1321 ; the central tower of Selby in 1690; 
 the single western tower of Hereford in 1806" (see Willis' Chichester, p. ii.). 
 
 % Craddock's Peterborough, 131. 
 
 S Examples of churches which now at any rate have only a central tower are Bath, Cart- 
 mel, Gloucester, Hexham, Malvern, Milton, New Shoreham, Norwich, St Asaph, St David's, 
 Salisbury, Sherborne, Tewkesbury, Winchester, Worcester. 
 
WESTERN TOWERS. 
 
 595 
 
 Western Tower. — In Kngland the majority of our churches have now and 
 always have had a western tower. The list of central towers of Anglo-Saxon date 
 now surviving is very short ; though it must be remembered that a central 
 
 Taunton St Mary Magdalen 
 
 tower may have formed a characteristic feature of their greater churches, now 
 all destroyed. But of Anglo-Saxon western towers over eighty still remain ; of 
 which about fifty are in East Anglia, Lincolnshire, Yorkshire, Durham, and 
 
596 WESTERN TOWERS. 
 
 Northumberland.* In the Romanesque of Normandy, on the other hand, a 
 western tower is quite exceptional. f But of the Romanesque churches built in 
 England a very large number have western towers. Whence came it ? Not 
 from Normandy, for there it was little employed. It must be a continuation of 
 the tradition of Anglo-Saxon building ; one of the very few instances in which 
 Anglo-Saxon Romanesque may be held to have influenced that of Norman 
 England. And as time went on, and as central towers fell into disrepair or 
 collapsed, the tendency was ever, more and more, in all except the largest 
 churches, to construct the tower westward. 
 
 The origin of the Anglo-Saxon western tower has been much disputed. It 
 may be but an upward development of the fore-porch or narthex,+ as at Corbridge, 
 Monkwearmouth, South Elmham, Boarhunt, and Daglingvvorth. It has been 
 derived from Germany, where the western tower is characteristic. But immense 
 tower-groups are much more characteristic of German Romanesque than the 
 single western tower ; and we have none of the former. It is hardly likely that 
 the latter should have had immense influence, the former none. Moreover, here 
 and there outside Germany, where German influence did not exist, square western 
 towers are common ; e.g. in Western France. On the whole there seems to be 
 little likelihood that the Anglo-Saxon western tower was borrowed from 
 Germany. 
 
 By far the most common position of the western tower is central ; i.e. at 
 the end of the nave. Sometimes the aisles are continued along its northern 
 and southern sides ; sometimes only its eastern side is engaged ; in either case 
 much of the light of the western window is lost. Externally, as at Louth, 
 Grantham, and NEWARK (627), where the former and more objectionable 
 position is adopted, and at LAVENHAM (589) and Boston, where the tower 
 is clear, it makes the nave appear low, while it crushes the aisles into 
 insignificance. Much better is the effect when it is placed unsymmetrically. § 
 Singularly effective is the north-west position of the towers of St Mary Redcliffe, 
 Bristol, and of TERRINGTON ST CLEMENT (92) ; in the latter of which the 
 dignity of the facade is greatly enhanced by the increase of breadth ; the tower 
 in this instance being detached from the aisle. More often, as in Heme, Kent,] 
 where also the tower is north-west, it forms the west end of the aisle. Here 
 also it is very effective. Fleet Church, Lincolnshire,^ has a south-western 
 tower, clear of the aisles. At Donington, Lincolnshire,** and at All Saints, 
 Maidstone, the south-west tower is added to the second bay from the west of 
 the south aisle, and forms a most dignified porch to the church. But there 
 is hardly any position in which a tower does not occur ; e.g. south of the 
 chancel ; or south of the crossing, as at Whaplode ; or serving as a south 
 transept ; or as a lichgate, as at WEST WALTON (597). 
 
 * Baldwin Brown, Arts in Early England, ii. 158. t Ruprich-Robert, 97. 
 
 % See Peers in Archaeological Journal, Dec. 1901, p. 429. 
 
 § So also Dehio, ii. 562, who instances St Pierre, Rouen. On the other hand Mr James 
 Cubitt wrote that a north-west or south-west position is the worst of all. 
 || Illustrated in Brandon's Parish Churches, 7. 
 IT Illustrated in Brandon's Parish Churches, 51, 53. 
 ** Illustrated in Brandon's Parish Churches, 51, 53. 
 
597 
 
 West Walton. 
 
598 TWO WESTERN TOWERS. 
 
 Occasionally, the tower is detached altogether, as at Beccles, West Walton, 
 Ledbury, and a good many churches in Gloucestershire and Herefordshire.* 
 Wooden campaniles also are not infrequent in well-timbered districts such as 
 Essex and the West of England. 
 
 Where the towers were central, as at Worcester and Salisbury, or much pierced 
 with openings, as at Chichester, it was not uncommon to have an additional 
 tower, detached, serving as campanile. Thus East Dereham has a central 
 lantern tower and a detached campanile. Chichester preserves its campanile ; 
 those of Salisbury and Worcester have been destroyed. 
 
 A few Continental examples of eastern towers have been noted ; e.g. the 
 great church of the Jesuits at Antwerp ; and some at Namur. There is one 
 at Polesworth. 
 
 TWIN TOWERS. — In the larger Romanesque churches north of the Alps, in 
 spite of numerous exceptions,"f a common arrangement was to have two western 
 towers. Of these one was intended to hold bells. Two western towers are 
 shown in an Exeter seal of 1 133, and may be those of the Anglo-Saxon cathedral 
 restored c. 1020. \ In the larger Romanesque churches both of Normandy and 
 England, two western towers are common ; e.g. the Abbayes aux-Hommes 
 and aux-Dames ; SOUTHWELL (520); Melbourne; Canterbury; DURHAM (28); 
 LINCOLN (562). But it not seldom happened that the completion of the nave 
 to the west was postponed ; and when at length it was taken in hand again, no 
 towers were erected in the facade. Whatever be the cause, it is a fact that many 
 churches of great magnitude are without western towers. Some, however, which 
 formerly possessed them, may have lost them. This may have been the case at 
 Winchester, Gloucester, and Norwich ; where great changes were made in the 
 facades in the fourteenth and fifteenth centuries. Such facades have a parochial 
 appearance, unworthy of churches of cathedral rank. Omission of western 
 towers is, however, frequent ; e.g. Binham, Chester Cathedral, Gloucester, Malvern, 
 Rochester, Romsey, Tewkesbury, Worcester, Bolton, St David's, St Saviour's 
 Southwark, Salisbury, Glastonbury, Whitby, Howden, Wenlock. The case of 
 Byland, Buildwas, Fountains, Furness, Jervaulx, Kirkstall, Netley, Rievaulx, 
 Roche, Tintern, Valle Crucis, is different. Like the great church of PONTIGNY 
 (S99)i near Auxerre, which still exists complete, they were all Cistercian, and 
 therefore omitted towers in conformity with the austere statutes of the Order. § 
 
 In most cases, if there are two towers, they are placed at the west end of the 
 aisles. In that position they have decided constructional value. For if there be a 
 central tower, its pressure tends to thrust the arcades of the nave out of the vertical. 
 Against the western towers these thrusts are powerless. Nevertheless, in a con- 
 siderable number of cases the towers were so placed as to flank the aisles and not 
 the nave; e.g. Bury St Edmunds and Wells. This was the arrangement intended 
 
 * In the fine Fenland churches the tower very frequently is isolated from the church ; e.g. 
 Sutton St Mary ; Fleet ; Whaplode ; Terrington St Clement and St John ; West Walton. 
 
 t In Lombardy the typical Romanesque facade, such as that of S. Michele, Pavia, has no 
 towers at all. 
 
 % Illustrated in Baldwin Brown, ii. 243. 
 
 § A statute of 1 154 says : " Let there be no stone towers for bells ; nor yet wooden ones of 
 inordinate height, such as to disgrace the simplicity of the Order." 
 
WESTERN AND CENTRAL TOWERS. 
 
 ;(J9 
 
 at St Albans. (Ripon, too, had nankin- towers to its unaisled nave; when 
 aisles were added in the sixteenth century, the towers became axial to the aisles. 
 Where there was a 
 western transept ( see 
 204), this was a natu- 
 ral arrangement. At 
 ELY (587), where there- 
 is a western transept, 
 its end towers become 
 mere turrets owing to 
 the presence of a big 
 central western tower. 
 
 It would seem that 
 in the tenth and 
 eleventh centuries a 
 common position was 
 to the north and south 
 of the choir ; e.g. S. 
 Ambrogio, Milan, and 
 S. Abbondio, Como. 
 This tradition may sur- 
 vive in Ernulph's 
 Canterbury, which had 
 towered chapels, on 
 one side that of St 
 Anselm, on the other 
 side that of St Andrew. 
 And again in the 
 twelfth-century YORK 
 Minster; on the founda- 
 tions of whose choir- 
 towers are built the 
 eastern transepts illus- 
 trated on 199. 
 
 Western and 
 Central Tower. — 
 We possess a con- 
 siderable number of 
 churches in which a 
 central tower is com- 
 bined with a single 
 western tower occupy- Beverle y Minster from S.W. 
 
 ,, r ' Pontigny from b. W. 
 
 mg the centre of the 
 
 facade; e.g. ELY (587;; WYMONDHAM (589); Wimborne ; Christ Church, 
 Hants ; Shrewsbury Abbey ; Waltham ; Malmesbury ; Bangor ; Lewes.* They 
 * Mr St John Hope's paper, read to Royal Arch. Institute, August 1S83. The western 
 tower of Hereford, which fell in 1786, is said to have been added in the fourteenth century. 
 
6oo 
 
 TOWER GROUPS. 
 
 are not uncommon abroad ; e.g. St Martin, Tours ; St Benoit-sur-Loire ; Poissy ; 
 St Germain des Pres ; Roeskilde. They are so common in the Romanesque 
 of the school of Auvergne ; e.g. Notre Dame du Port, Clermont ; Issoire ; 
 Brioude ; Le Dorat ; Le Puy ; St Savin ; that one is tempted to try to connect 
 our English examples with that early and important school. But the monk 
 Reginald, writing in the twelfth century, says that the Durham Cathedral of 
 999 had one central and one western tower ; and we have an equally clear 
 statement about the Ramsey Abbey of 970.* Here then, as in the case of 
 the western tower without central tower, the planning may be simply a con- 
 tinuation of indigenous Anglo-Saxon tradition. 
 
 TOWER GROUPS. — Where all three towers have been carried up, as at 
 LINCOLN (44), SOUTHWELL (520), Wells, Canterbury, York, DURHAM (28), one 
 
 obtains masses, in their grouping as 
 impressive as ever man wrought ; impres- 
 sive even among the works of nature. 
 LICHFIELD (frontispiece) alone retains 
 its spires ; in Lichfield alone we see the 
 glory complete of the English Gothic 
 exterior as it was designed to be. The 
 glorification of the tower went even 
 further abroad. In Germany there are 
 many instances of four towers ; e.g. 
 Paulinzelle ; St Castor, Coblentz ; the 
 Apostle's Church, Cologne ; in France, 
 Vdzelay ; St Benoit-sur-Loire ; St Mar- 
 tin de Tours. There were five towers 
 designed at Angouleme Cathedral, and 
 Deols, Indre. . At Cluny, Tournai, Lim- 
 burg, Speyer, Worms, Laach, there were 
 six. Rouen, Reims, and Laon Cathe- 
 drals were intended for seven towers ; 
 five of them, like those still surviving 
 at Tournai, were to be central. Chartres 
 was designed for eight. Except at 
 Peterborough, we were less ambitious. 
 PLAN OF TOWERS. — The most usual form is the square. But since the 
 transepts are sometimes narrower than nave and choir, the central tower is 
 often oblong ; at BATH (373), it is markedly so. 
 
 Towers with eight t sides are not uncommon; e.g. Uffington ; Coxwold. 
 Sometimes an octagon is superposed on a square ; e.g. Ely west tower ; 
 Fotheringhay ; Boston; Irthlingborough ; Lowick. At Cartmel a square is set 
 diagonally on a square. ' All Saints', Maldon, has a triangular tower. 
 
 Round towers are common in Norfolk and Suffolk ; and several examples 
 
 * See Scott's Essay, 105 ; and Baldwin Brown, Arts in Early England, ii. 242. 
 
 t An octagonal tower occurs in the eleventh century at Tordouet, Calvados ; Caumont, 
 Statistique Monu men tale, v. 809 ; Ruprich-Robert mentions five examples of the twelfth century 
 in Normandy, i. 164. 
 
 Thorineton. 
 
PLAN OF TOWERS. 
 
 
 occur in Essex, Cambridge, and Sussex* The)- have been regarded as survivals 
 of Roman forms; such as that of the Roman Pharos at St Mary in the l 
 Dover, which however is octagonal; Roman lighthouses were still common 
 in the time of Rede. Round towers of the sixth century still remain at 
 Ravenna;! e.g. S. Apollinare in Classe. In a mosaic of the fifth century in S. 
 Maria Maggiore, Rome, round towers are shown in the neighbourhood of two 
 churches. In the ninth-century plan of St Gall round towers are shown, from 
 the end of the ninth to the beginning of the thirteenth century the Irish were 
 building round towers. 
 
 We may indicate another source. Newel staircases vis or vice , 
 circular staircases winding round a central column, were employed by the 
 Romans, and still survive inside the Columns of Trajan and Antoninus at Rome 
 and of Theodosius at Constantinople. A tenth-century example exists in the 
 Palatine Chapel, Aix. j Numerous examples of 
 slender, round towers, containing staircases, exist 
 all over Europe ; e.g. BECKLEY. In Anglo- 
 Saxon work a newel staircase is constructed in the 
 thickness of the wall of the tower of Great Hale ; 
 while at Brixworth, Brigstock, Houghton-on-the- 
 H ill, and Broughton a half or three-quarter round 
 turret is built for it in front of the west wall 
 of the tower. A similar turret seems to have 
 existed at North Elmham. § The resemblance 
 between such staircase turrets as that of Beckley, 
 and such towers as those of S. Giovanni, Ravenna, 
 St Maurice, Epinal, and the Irish round towers, 
 is so close that a causal connection seems pos- 
 sible.*" 
 
 But for the broader English round towers a 
 simpler explanation suffices. They occur almost 
 wholly in chalk districts, where for outer facing 
 nothing is available but flint. It was easier to 
 construct flint-cased towers circular. So the towers 
 were made round. If they had been made square, 
 it would have been necessary to import freestone quoins from a distance. 
 
 Tower CONSTRUCTION.— In the Romanesque and in the earlier Gothic 
 towers there was an enormous waste of material. It has been estimated that in 
 the western towers of Coutances there is an excess of some 1,000 tons. Still 
 more remarkable is the case of the Leaning Tower of Pisa.** In this the wall of 
 the lower story is 13.I feet thick, that of the upper stories 9 feet. With this we 
 
 * See illustrated paper in Archmologia, vol. 23. 
 
 t Rivoira denies the existence of any church towers anterior to the ninth century. 
 % Enlart's Manuel, 19. 
 
 S Baldwin Brown, ii. 175, 211. In Normandy also Romanesque staircase turrets occur, 
 some in the thickness of the wall, some in turrets Ruprich-Robert, 163. 
 II Illustrated in Miss Stokes' Early Christian Art in Ireland, ii. 54. 
 IT Minarets are but a decorative treatment of the staircase turret. 
 ** For the calculations see paper by Mr S. 11. lie.de, in Builder, Jan. iS, 1S90. 
 
 Beckley, Oxford. 
 
602 TOWER CONSTRUCTION. 
 
 may compare the tower of St Mary Redcliffe, Bristol, A.D. 1292, which, according 
 to William of Worcester, tapers from 7 to 5 feet ; and that of Louth, which 
 is 7 feet thick at the bottom ; 6 feet 2 inches at the lower windows ; 3 feet 9 
 inches at the belfry chamber ; it is 20 feet 1 1 inches square inside. The whole 
 steeple is 300 feet high above the street. It is estimated that the Leaning 
 Tower brings a pressure of j\ tons to the square foot on soil which is only 
 capable of sustaining 4! tons.* Had the tower been bonded into the cathedral, 
 the latter must have been dragged down.t It is perhaps because they feared 
 this, that the Italian architects usually detached their heavy campaniles from 
 the churches. £ In England and France, as masoncraft improved, it was felt to 
 be as unnecessary as it was dangerous to build towers with walls so thick. In 
 some cases the builders went further still. Anticipating the methods of con- 
 struction of the domes of Florence Cathedral and St Peter's, Rome, the tower 
 walls were built in two thin shells with a hollow space between ; e.g. the 
 central towers of Lincoln and Salisbury. § While at Hereford and WELLS (588), 
 where the fourteenth-century tower was built on weak Norman piers, the 
 inner skin of the tower was constructed, girder-fashion, of vertical and hori- 
 zontal stone beams. Similar construction occurs in the central towers of 
 Coutances and St Etienne le Vieux, Caen. 
 
 TOWER ROOFS. — Norman tower roofs were apparently often hipped to a 
 point ; i.e. were low square spires, as at Priestholm || or Puffin Island, and at 
 Thaon in Normandy. Sometimes saddle-back roofs occur ; as at Wadenhoe ; 
 Claydon ; Little Claydon ; Icomb ; Tinwell ; and along the coast of South 
 Wales. f These are more common in Normandy than in England. We did 
 not adopt the fashion common in late French Gothic of a roof hipped to a 
 ridge. A few examples of stone roofs occur on towers. 
 
 Tower Design. — Occasionally an English tower is slender and tall ; e.g. 
 MANORBIER (587) ; but unless it be detached, like most Italian campaniles, 
 it challenges disastrous comparison with the bulk of the church. On the other 
 hand, provided that it have breadth, mass, bulk, it need not be tall. Indeed 
 the characteristic tower of the smaller village-churches of England is one 
 broad and low;** but little overtopping the nave roof: e.g. those of Gresford, 
 ST MARY'S, BEVERLEY (603), TERRINGTON ST CLEMENT (92). The con- 
 trast is well seen at Oxford ; where the slender tower of MAGDALEN (588) 
 is satisfactory because it is detached ; while the short, massive, engaged 
 tower of MERTON (589) is equally effective. In a tower what is most 
 desiderated is not grace, but strength. The central towers of York, HERE- 
 FORD (587), and WELLS (602) are all impressive ; but it is not by their 
 height. So also it is mainly to their bulk that the satisfactory effect is due 
 
 * On the weight-bearing capacities of soils see " Foundation " in R. Sturgis' Dictionary 
 of Architecture and Building. 
 
 + Louth steeple, which is admirably built, has sunk 7 inches. 
 
 % So also in the oozy soil of the Fenland the towers are very often detached ; see list 
 in 598, note. 
 
 § Sir Charles Anderson's Lincoln, 120. || Illustrated in Baldwin Brown, ii. 163. 
 
 1" A fourteenth-century example, with spirelet rising from the saddle, occurs at Brentingby, 
 Leicester. 
 
 ** A tower for eight bells would have to be broad. 
 
WELLS CATHEDRAL FROM S.E 
 
TOWKR DESIGN. 
 
 ' ! 
 
 of such Norman towers as TEWKESBURY 390), Exeter, St Albans. But if with 
 bulk there be combined height, the tower reaches the maximum of impressive- 
 
 ness ; as at NEW SHOREHAM ( 
 
 Fountains 
 
 GLOUCESTl I [32); LAVENHAM 
 
 (589);* Wymondham ; Cawston ; Cromer; Southwold ; Hedon ; 
 the king of towers, the Victoria Tower of the Houses of Parliament. 
 
 As regards central towers, the most important factor internally is not 
 height, but breadth and illumination. If a central tower be high and narrow, 
 like that of CANTERBURY (588), however graceful externally, its interior cannot 
 be seen except with the sensation of looking up from the bottom of a well. 
 If it be badly lighted, like that of LINCOLN (328), then its interior, being 
 barely visible, cannot tell. The value of breadth and illumination are well 
 
 Trunch. 
 
 Beverley St Mary's, from S.E. 
 
 seen in York crossing. At ELY (45 I the breadth of the central octagon is 
 even greater than at York ; and floods of light stream down from its many- 
 windowed flanks. Nowhere are there contrasts so sharp, so dramatic, of light 
 and darkness as at Ely. The low, brilliantly lighted lantern of Dereham also 
 deserves study. 
 
 Externally, the fenestration of towers, whether central or not, is of the highest 
 importance. Unless the windows be deeply recessed, there is no adequate 
 play of light and shadow : the windows do not look windows, being indis- 
 tinguishable from the panelling. Many of our towers fail to tell at a moderate 
 distance owing to the shallowness of the fenestration ; e.g. the central tower 
 of Wells. 
 
 * Lavenham tower is just three times as high as the clerestory wall of the nave. 
 
604 
 
 TOWER DESIGN. 
 
 When a block of masonry has been raised high in the air, one's first anxiety 
 is as to its stability. One's anxiety is roused if the voids, i.e. the windows, are 
 numerous and large, as at Howden and HULL (96) ; more still, if one side 
 of the tower has a huge hole in the wall, as at Shrewsbury Abbey. One is 
 reassured if the solids greatly preponderate ; and one does not object in the 
 least if the ground story, as at Magdalen College and Manorbier, is plain 
 solid wall. It may be doubted whether the richest towers of Somersetshire 
 are more impressive than the gaunt mountain masses of East Anglia, four- 
 square to the winds ; such as HOLME (591) and CAWSTON (589); or than the 
 fortress donjons of Kent and Bedford, with slits for windows ; e.g. SOUTHFLEET. 
 It is in the towers of the west country that most attention is paid to the 
 fenestration ; e.g. ST MARY MAGDALEN, TAUNTON (595) ; Bishop's Lydiard ; 
 Huish Episcopi ; st John's, Glastonbury (588); Wrington; St Stephen's, 
 
 Bristol ; St John's, Cardiff. 
 These bell towers naturally 
 fall into three divisions ; the 
 ground story, which contains 
 both state doorway and west 
 window ; the first floor con- 
 taining the windows of the 
 ringers' chamber ; and the 
 second floor containing the 
 belfry windows. But many 
 of the towers are so lofty ; 
 e.g. St Mary Magdalen, 
 Taunton ; that an additional 
 mid-story has to be inserted. 
 This additional story is 
 usually a duplicate of that 
 of the ringers' chamber. In 
 some of the best examples. 
 e.g. ST JOHNS, GLASTON- 
 BURY (588), and EVERCREECH 
 (595), it seems to have 
 been felt that the voids were in superabundance ; and the intermediate story 
 or stories has been kept solid or nearly so. All these towers are crowned 
 with magnificent combinations of panelled or pierced parapets, battlements, 
 and pinnacles (607). All were built in what used to be styled the "debased" 
 period of English architecture. 
 
 In that masterpiece of Late Gothic, the steeple of LOUTH (611), great 
 depth of shadow is gained in the windows by setting them back towards the 
 inner face of the tower wall. How great is the gain may be seen by contrasting 
 this steeple with the tower of BOSTON (222). 
 
 Midway in many Norfolk towers ; e.g. North Walsham, Fakenham, Cromer, 
 HOLME (591); square openings filled with Curvilinear or Rectilinear tracery 
 occur; erroneously called "sound holes." 
 
 In several Somersetshire towers the upper windows are filled with perforated 
 
 Southfleet. 
 
TOWER DESIGN. 
 
 
 stone tracery; e.g. ISLE ABBOTS (59] and HUISH EI'ISCOP] (51;). II,- 
 where it is more usual to employ sloping louvre boards sel horizontally. The 
 treatment of these boards profoundly affects the character of a tower, hi 
 little village churches in France the louvre boards arc few, broad, and have -real 
 projection ; consequently there is a depth of shadow rare with us. 
 
 Very greatly, too, the character of a tower depends on the treatment of the 
 angles. How they shall be treated depends on what is wanted. If it is to 
 overawe, the tower may rise in sheer vertically from pavement to cornice, like 
 the campanile in the piazza of St Mark's, Venice, and that of Manorbier.* 
 What Mr Ruskin said of the transept wall of Beauvais, that there are few such 
 lofty precipices even in the Alps, is still more true of the unbuttressed type of 
 tower. But the corners are the weak points of towers. We did not like to lose 
 the help of buttresses. So, here and there, we built buttresses, but built them 
 practically vertical ; e.g. in Magdalen College, Oxford ; the Angel Tower of 
 Lincoln ; Whissendine, Rutland ; Wymondham. If that was not enough, if we 
 hankered after "the solemn frown of projection," corbel table and cornice could 
 be built out, and on these over-sailing parapet, battlement, and pinnacles could 
 be poised; as in the west towers of YORK (821; Dundry and St Stephen's, 
 Bristol ; and, above all, in the Kreisker at St Pol de Leon, Brittany. -f- 
 
 But we may wish, not to overawe, but to reassure. Nothing reassures so 
 much as appearance of stability. And that appearance is given by the pyramidal 
 form. So sometimes the walls of the tower " batter," i.e. slope backward ; as 
 in the lowest stage of many church towers of Herefordshire. In Gothic design 
 each stage is usually vertical, but each stage may be drawn back a little from 
 the stage below ; e.g. Bishop's Lydiard, Somerset ; Kettering, North Hants ; 
 Moulton, Lincolnshire; Ingatestone, Essex. \ Or, as at TRUNCH, NORFOLK 
 (603), both expedients may be resorted to. Such a tower, tapering upward, 
 seems to be vertical, unless the process is overdone, as at Trunch and St 
 Peter's, Oxford. An absolutely vertical tower appears to the faulty lenses of 
 our eyes actually to overhang ; just as do Flamborough cliffs, seen from a boat 
 below. The Greeks, aware of this, sloped backward the columns of their temples. 
 
 We may set the tower buttresses diagonally, as at Trunch ; one buttress 
 doing the work of two. But the effect is not satisfactory unless the projection 
 of the buttresses be slight, as at Chipping Campden and Gloucester Cathedral ; 
 or unless the tower has an octagonal spire, as at St Sepulchre's, Northampton, 
 and Edenbridge, Kent ; in which case the diagonal buttresses carry down the 
 oblique sides of the spire. In later work, as in the central towers of Ashford, 
 Cricklade, BATH ($73), and CANTERBURY (588), and at WYMONDHAM (589) 
 the buttresses may be hexagonal or octagonal ; but nothing equals the 
 older arrangement, as seen at ST JOHN'S, GLASTONBURY (588), and in the 
 Somerset churches generally. In these last a fine effect is got sometimes by 
 setting the two corner buttresses some little distance from the corner ; thus 
 the quoins peer through and present gratis an additional strong vertical line ; 
 
 * In chalk districts there was naturally a tendency to omit buttresses ; e.g. SOUTHFLEET 
 (604), Kent ; because of the expense of freestone quoins brought from a distance, 
 t Illustrated in Enlart's Manuel, 570. 
 \ Described by Sharpe in Lincoln Excursion, 144 ; and Buckler in Churches of Essex, 100. 
 
606 TOWER DESIGN. 
 
 e.g. St Stephen's, Bristol ; Dundry ; Evesham ; Wisbech ; Titchmarsh.* 
 Whatever the form of buttress employed, undue spread must be guarded 
 against ; otherwise to the horizontal line of nave and choir we shall have 
 opposed in the outline of the tower, not a vertical, but a sloping line. The 
 buttresses must not be developed so far as to destroy or impair the verticality 
 of the tower. 
 
 Another critical point is the relation of the buttress to the pinnacle. A 
 solution which is as easy as it is satisfactory is to make buttress and pinnacle 
 one ; as in Auxerre Cathedral ; Bath ; Canterbury, Wells and York central 
 towers ; Whiston ; St Cuthbert, Wells ; Ashford ; Cricklade ; Magdalen 
 College, Oxford. Another method, not so satisfactory, is to bring the tower 
 design to an end at the cornice ; then to commence a fresh design for pinnacles 
 and battlements independent of all below; as in GLOUCESTER (132) and 
 HEREFORD (587) central towers ; and ST MARY MAGDALEN, TAUNTON (607). 
 A similar fault is failure to carry the buttress, as at TRUNCH (603), or the 
 buttresses, as at CAWSTON (589), up to the cornice ; they are pulled up with 
 a jerk, as it were. It is common in the Somerset churches ; e.g. St Mary 
 Magdalen, Taunton ; where an attempt at compensation is made in the 
 form of a flimsy little buttress hanging on to the corner of the pinnacle. This 
 weakness, however, is less observable in perspective than in elevation. 
 
 The different stories of the tower are usually marked off by strings. Should 
 they pass round the buttresses ? As to that the practice varied, even in the same 
 district. Thus at Wrington no strings pass round the buttresses ; cf. Moulton. 
 At St Mary Magdalen, Taunton, only the string beneath the belfry stage passes 
 round ; severing in rather unfortunate fashion the belfry story and pinnacles 
 from the rest of the tower below. l\X Dundry and St Stephen's, Bristol, all the 
 strings pass round. 
 
 But buttresses themselves are usually divided into stages by set-offs. 
 Should these stages and set-offs correspond with the stories and strings of 
 the tower? Sometimes they correspond exactly; eg. at St Mary Mag- 
 dalen, Taunton ; where notice, however, the difference in the treatment of the 
 quoins ; and still more minutely at Wisbech ; where even the hood-molds 
 of the windows become strings, each running into its own set-off. At 
 KETTERING (623) every other set-off corresponds with a string. On the other 
 hand, at St Neots, the set-offs are studiously independent of the strings ; cf. 
 Moulton.f 
 
 But the most important question is the treatment of the skyline. Shall we 
 be satisfied with a strong horizontal line, as in Giotto's tower at Florence ; 
 Trunch ; Cawston ; % Howden ; Lavenham; St Nicholas at Wade; Wymondham; 
 New Shoreham ; emphasising it to the utmost by projection of cornice ; or 
 above the cornice subdivide the single huge vertical mass of the tower into four, 
 eight, or sixteen spirelets ; cresting and crowning the tower with a galaxy of 
 
 For Towers and Spires consult Wickes' three folios of illustrations. 
 
 + Of the seven stages of the western towers of SOUTHWELL (520), the four lowest are 
 determined by the strings and roof of the nave. 
 
 \ In some cases it may be that the tower has a horizontal skyline only because it is 
 unfinished ; e.g. York central tower and Wells western towers. 
 
6o7 
 
 012345 io 
 
 SCflLF OF 1 ■ ' ' ' ' 
 
 Taunton St Mary Magdali n. 
 
60S TOWER DESIGN. 
 
 pinnacles, parapets, and battlements. Of the two types the pinnacled type had 
 by far most admirers. 
 
 But as to pinnacles also there are rival schools ; some adopting a tall, 
 some a small pinnacle. If the tower be tall and slender, as at Magdalen College ; 
 Ashford ; Cricklade ; Tideswell ; Canterbury; a tall pinnacle sits well. If it be 
 short and bulky, short pinnacles are more usual ; e.g. St Mary's, Beverley ; 
 Hull ; Gresford ; Chipping Campden ; Terrington St Clement's. On the whole 
 the tendency is to a rather small pinnacle ; the less the pinnacle is emphasised, 
 the more important looks the belfry stage. In Somersetshire the schools are 
 equally divided ; e.g. at Taunton the pinnacles of St James are insignificant, 
 while those of St Mary Magdalen are large and important.* 
 
 As to the application of ornament, we may discriminate four types of 
 tower; one, without ornament; a second, with equally diffused ornament; a 
 third, with ornament concentrated in the ground story and the belfry story ; a 
 fourth, with ornament concentrated at the summit. Of the first, Manorbier, 
 Wymondham, and Southfleet are examples. Where freestone is costly and 
 the church is cased in flint, there will naturally be little or no ornament. 
 Even when built, like Cawston, in an admirable freestone, it will follow the 
 local tradition in its sparing of ornament. To the second type belong the 
 towers of the eleventh century in Normandy, and of the twelfth and thirteenth 
 in England, covered with bands of arcading in the same fashion as the 
 facades of Ely and Lincoln ; e.g. ELY (587), DURHAM (28), Castor, LINCOLN 
 (562), NORWICH (522), TEWKESBURY (390), exeter (377), St Albans, of the 
 twelfth century ; and of the thirteenth century Whaplode, WEST WALTON 
 (597), Raunds, SUTTON ST MARY (613); and St Mary, Stamford. These 
 are large and enriched examples ; there are also simple and effective 
 designs of smaller thirteenth-century towers, such as Madley, Herefordshire. 
 Several of the greater towers of the fourteenth century also are smothered, 
 in ornament ; e.g. the central towers of Hereford and WELLS (602). And in 
 the fifteenth and sixteenth centuries many a great tower drips with panelled 
 tracery from top to bottom; e.g. Evesham; Gloucester; Canterbury; WREXHAM 
 (609) ; St Peter Mancroft, Norwich ; most of all, perhaps, Eye Church, Suffolk, 
 which is panelled in flint ; ornament is everywhere ; not a quiet spot anywhere 
 to rest the eye ; for want of which, window tracery, parapet, and pinnacle fail 
 to tell. In the Somersetshire towers generally a far higher level of design is 
 reached. At St John's, Glastonbury ; North Petherton ; Chewton Mendip ; the 
 Taunton churches ; and many a fine tower elsewhere ; e.g. at St Neots, Hunt- 
 ingdon ; Whissendine ; Wrexham ; and at Terrington St Clement's ; the inter- 
 mediate story or stories is treated more simply ; the main body of ornament 
 being concentrated on the ground story and the belfry story. Of these two 
 the belfry story is usually the more decorated ; and its richness is enhanced 
 yet further by elaboration of pinnacle, parapet, and battlement. The fourth type 
 is seen in its simplest elements in the Italian campaniles ; the ornament is 
 reserved for the summit ; or if there be any below the cornice, it rises in 
 
 * The original form of the pinnacles has often been departed from in " restorations " ; e.g. 
 the present pinnacles of Worstead tower, Norfolk, are far taller than those shown in Neale's 
 Churches of Great Britain, 1824, vol. i. So also those of Bath Abbey, now being removed. 
 
609 
 
 
 I a 
 
 
 
 
 
 : 
 
 - ■- 
 
 1 '.-->, 
 
 
 . 
 
 
 
 - :;<r/-^fir~;i ■ :--- ' EPA* ' 
 
 ■ ■• v y-'# ->*-■ - -•!-•! ; '• San i 5 ' 
 
 iil 
 
 
 
 * .. . ■T ; " i 
 
 
 
 Wrexham. 
 
 2 i I 
 
610 TOWER DESIGN. 
 
 richness as it nears the summit. This is the principle of design in the 
 towers of St Stephen's, Bristol, and Magdalen College, Oxford. In these the 
 glories of the summit are all in all ; nothing is allowed to detain the gaze in its 
 upward flight. Truly these late builders knew how to design. Nor had they 
 said their last word in towers ; there were Louth and Coventry spires to come. 
 Their very success with the tower nevertheless curtailed the building of spires. 
 In the thirteenth and fourteenth centuries it almost seems as if every tower 
 was designed for a spire. The fifteenth-century tower needed no spire ; and 
 spires grew infrequent, except in certain districts, e.g. Northants ; while towers 
 were erected by hundreds. 
 
Chapter XLI. 
 
 THK SPIRE. 
 
 Origin and Classification of Spires— Spire Angles — Bands — Plate and Bar Tracery 
 — Fenestration— Plan — Height— Proportions — Entasis— Spire Design- Spire Con 
 struction. 
 
 ORIGIN" OF THE SPIRE. — Of all the members of a Gothic building the spire is 
 the most original, the most religious, and the least essential. The pier, the arch, 
 the capital, the base, the vault, the dome, all find their ancestry in a distant 
 past. The spire was something new, that Greek and Roman knew not of. 
 Not till the twelfth century was well advanced in France, not till it had neared 
 its close in England, did the low pyramidal roof rise to that " sky-y-pointing 
 pyramid," the spire. It was a concession, the chief concession in Gothic 
 architecture, to religious symbolism. To Wordsworth, its lines, all converging 
 upward and heavenward, were a reminder of the direction that his thoughts and 
 prayers and life should take — 
 
 " Watching with upward eye the tall spire grow 
 And mount at every step — with living wiles 
 Instinct — to rouse the heart and lead the will 
 By a bright ladder to the world above." 
 
 Or, as he puts it in one of his sonnets — 
 
 " Spires whose silent ringer points to heaven." 
 
 Artistically, its special value lies in the strength of contrast which is afforded 
 between the vertical lines of the spire and the prolonged horizontal lines of nave 
 and chancel. Equally effective is it whether central, as at Patrington and 
 Salisbury, converting all the projected masses of the buildings into one con- 
 verging pyramid ; detached, as at Fleet, Ledbury, and St Mary Redcliffe ; * or 
 giving the noblest of facades to the parish church, as at Grantham, Newark, 
 and Louth. 
 
 In respect of the spire Gothic architecture abandoned one of its leading 
 principles ; and perpetrated a piece of " constructed decoration." The spire has 
 been defended as being a necessary roof to the tower. At first indeed it was 
 but a tower roof; either in timber; or else in stone corbelled inwards, as at 
 Priestholme. But it soon became much more than a roof; and in the open- 
 work spires of the Continent, Freiburg, Strasburg, Antwerp, it forgot to be a 
 roof altogether. 
 
 Something of the character of the spire existed doubtless in quite early 
 times. The earliest representations of French churches, e.g. that of the abbey 
 * To the eye the Bristol steeple is detached : though really it is engaged in the north aisle. 
 
6l2 
 
 Ledbury. 
 
 Tilney All Saints'. 
 
 Bridgwater. 
 Patrington from S.E. 
 
6i 3 
 
 Barnack. 
 
 Oxford Cathedral. 
 
 Sutton St Mary. 
 North Luffenham. 
 
614 
 
 r-^fe 
 
 mm m 
 
 ill rr *?^p- I i 
 
 J 4j §§&■ r-A 
 
 ; z 
 
 '-'■-"•-.. : - 
 
 Southam. 
 
 De.iford. 
 
6i S 
 
 3J( IIWIltgflJlMlltll L III 
 
 AJHft#i ;: fe • t* ^f sir 
 
 King's Sutton. 
 
 I [olbeach. 
 
6i6 
 
 ib ■■■'■■■■■■.■■ : 
 
 ^>fe*feyX~ 
 
 Binsham. 
 
 Desboroueh. 
 
EARLY SPIRES. 617 
 
 church of St Riquier, show us central towers with superposed louvres. And in 
 various Anglo-Saxon manuscripts; e.g. j n Caedmon's metrical paraphrase of 
 Scripture history* the exact form of a spire is represented. But all those pro- 
 bably were in wood. It is not till the latter half of the eleventh century that a 
 low spire appears in masonry. This is at Thaon, Calvados.- 1 - In England early 
 examples are seen at PenmonJ and Priestholm or Puffin Island, Anglesea. 
 A spire is shown in the conventual seal of Kenilworth.§ Turner's Liber Stiuli- 
 orum has a careful drawing of the western spires of SOUTHWELL (520) ; they 
 have lately been restored. All these, however, arc still roofs; not spires. 
 Thaon spire is quite low; the interior is still tied with wooden beams. 
 In France, however, the stone roof of the tower leaped almost at once 
 into a spire ; and by the middle of the twelfth century probably, great 
 spires had been constructed ; e.g. at St Germain, Auxene, and at La 
 Trinite de Vendome ; the latter of which rises 262 feet from the ground. In 
 England it is possible that the spire of BARNACK^f (613) may have been built 
 in the last years of the twelfth century. Another early spire is that of OXFORD 
 CATHEDRAL (613J. This also is a spire of obtuse angle; and it is of a type 
 somewhat rare in England, but common in France, e.g. in the south-west 
 spire of Chartres ; viz. without parapet, but with pinnacles and dormers : in 
 this there may be French influence, and the date may be early. The New 
 Romney spire also was designed for pinnacles. Another early spire, hardly 
 clear of Transitional detail, is that of SUTTON ST MARY** (613), Lincolnshire. 
 This, however, is a timber spire, covered with lead. 
 
 The above spires, however, did not produce schools. It was not till the first 
 half of the thirteenth century that indigenous schools of English spire builders 
 arose. One built in timber, not in Sutton fashion, but as at NEWHAVEN (17); 
 or else, as at ETTON (620J, they copied in stone the forms which had arisen 
 out of timber technique. The second school arose in Northamptonshire and the 
 neighbouring districts, wherever that beautiful freestone, the Oolitic limestone, 
 was found on or just below the surface of the ground. Its spire was of stone ; 
 the angle of the spire, at first, was obtuse ; it had numerous windows , these, at 
 first, were large ; it had no parapets, and it had broaches ; e.g. Sleaford, Rauceby, 
 Warmington, Polebrook, FRAMPTON MS, NORTH LUFFENHAM (613). This 
 long was the premier school in England. The third school, which in the end 
 ousted the rest, was that of the parapetted spire, usually with pinnacles ; and 
 sometimes with flying buttresses. 
 
 It was not till the fourteenth century that spire building was fully developed. 
 Then it was first that great spires were erected ; e.g. Salisbury ; Grantham and 
 its imitator Newark ; St Mary's, Oxford ; St Mary's, Stamford.-H By this time 
 
 * Archceologia, xxiv., Plate 83. + Illustrated in Pugin's Normandy, Plate 22. 
 
 X Illustrated in Parker's Inirodiictzon, 81. •? Illustrated in Bloxam, 103. 
 
 || Illustrated in Viollet-le-Duc, Architecture, iii. 357, 56. 
 
 w Prior, p. 370 ; who suggests a late twelfth-century date also for the spires of Oxford 
 Cathedral and Sutton St Mary's ; and the base of New Romney spire. 
 
 ** See measured drawings in Spring Gardens Sketch Book. 
 
 tt The towers of Ketton, Melton Mowbray, and St Mary, Stamford, are very similar, and 
 are probably by the same hand. The spires of Ketton and St Mary, Stamford, are a century 
 later in date than their towers, and show an even greater similarity to one another. — R. P. B. 
 

 fet 
 
 Be 13 II IS II 
 
 n <SS 
 
 Frampton. 
 
CLASSIFICATION OF SPIRES. 619 
 
 many a chancel had been lengthened 1225;; and the tower, which had been 
 dominant before, had become inadequate. 
 
 Still more was this the case in the fifteenth century, when the passion raged 
 for elevating clerestories ; to many a tower was added a spire, because the nave- 
 had received a clerestory ; e.g. Burton Latimer, Northants. And these later 
 spires, e.g. Coventry and Louth, were as noble as any. 
 
 But as in its beginning, so to the end, spire building was curiously sporadic 
 and local. Pass by rail from Spalding to Lincoln, or down the Nene valley, and 
 spires cut the horizon far as the eye can reach. In other districts, e.g. Norfolk 
 and Somersetshire, the tower is equally predominant. On the whole, as was 
 natural, spire building was in vogue most where there was the best craftmanship, 
 i.e. just where there was the best and most abundant freestone ; or where there 
 was the cheapest transport ; all places, whose rivers or drains put them within 
 reach of Barnack, Ketton, and Ancaster quarries, could afford to build spires. 
 
 Classification of Si-irks. — English spires are very varied in design; 
 nevertheless they ma}' be reduced to a few main types. One feature distin- 
 guishes them from a large class of Continental designs ; viz. that our builders 
 kept in view the sound Gothic principle, that the load should be distinct from the 
 support. They did not mix up pier and arch, except in a kw late examples, 
 eg. TENBY (410) ; and they did not mix up tower and spire. Abroad, however, 
 it is frequently quite impossible to say where the tower ends and the spire 
 begins ; e.g. at Le Puy, Senlis, Freiburg, Antwerp, and St Stephen's, Vienna. 
 In England there is seldom any such difficulty ; there is a distinct line of 
 demarcation between tower and spire, and on that line we can base a simple 
 system of classification. It is this. Some spires have no path round their foot ; 
 we may call them Pathless spires. The rest have a path round ; we might call 
 them footpath spires ; but as the path is always guarded by a parapet or a 
 battlement, we may as well avoid a barbarism, and call them Parapetted spires. 
 
 PATHLESS SPIRES: Class I. — Of these the first division consists of timber 
 spires ; which are of course plentiful where oak was abundant and stone scarce ; 
 e.g. in the Sussex Weald and much of Kent and Essex. These are without 
 windows; e.g. at Shere, Tangmere, Merstham, NEWHAVEN (17), Plumpton, 
 Southwick, Bourn, Walsingham, Wickham Market, Ryton, Chesterfield, Harrow, 
 Hemel Hempstead, Godalming. The normal type is well seen at Newhaven. 
 Another type, less common, is that which, as at Irchester, starts as a square 
 spire, but almost immediately becomes octagonal. The Chesterfield spire is 
 curiously warped ; probably it was built of unseasoned timber, and has twisted 
 round on the sunny side.* Many wooden spires have perished by fire, lightning, 
 or decay; or have had to be pulled down; e.g. the triplets of Lincoln and 
 Ripon. The wooden spire on the central tower of Lincoln was 523 feet high, 
 a landmark far out at sea : that of Old St Paul's was probably 500 feet high. 
 Ely had one western spire of wood ; Durham two : Hereford a central spire. 
 Worcester, Rochester, Finchall Abbey, Malmesbury, Tewkesbury, also had 
 spires. 
 
 The curious form which is characteristic of most timber spires is due to the 
 fact that some distance above the base of the spire there is a timber collar, which 
 *See paper read at the R.I.B.A. by Mr Coldwell ; and Builder, 1S55, 13, 4°- 
 
620 
 
 BROACH SPIRES. 
 
 is upheld by timbers sloping up to it from the four angles of the tower : on this 
 collar much of the upper part of the spire rests. A similar construction pre- 
 vailed in the wooden spires of France ; e.g. the 
 central spire of Notre Dame, Paris. 
 
 The second division consists of those which 
 imitated in stone the characteristic form of a 
 timber spire; e.g. ETTON ; Crick;* Bythorne; 
 and Lostwithiel.t This curious nonlithic de- 
 sign was infrequent. 
 
 Class II. — The next type, the broach 
 spire, was common in the thirteenth and still 
 more in the fourteenth century. In its native 
 district, in and near Northants, it lingered till 
 the fifteenth century ; e.g. at Brampton, Bar- 
 rowden, and Stanion. At first the broach — 
 the pyramidal mass of masonry, covering a 
 squinch, and leaning against an oblique side 
 of the spire — was low or of moderate size ; 
 e.g. at NORTH LUFFENHAM (613). Strictly, 
 however, the broach ought to have a slope 
 intermediate between the vertical line of the 
 tower and the inclination of the spire ; and 
 very soon the broach was made much taller to 
 the great improvement of the general effect ; 
 e.g. at ketton (621), at Aumsby ; Walcot; 
 Anwick ; % Aldwinkle ; § Threckingham ; § and 
 above all, at EWERBY || (630). Late broaches, 
 however, are often exceedingly small and low ; 
 especially in the typical spire of Gloucester- 
 shire ; e.g. Leckhampton. Outlying examples 
 of broach spires are Hemingborough, York- 
 shire ; Kidwelly, Carmarthen ; St Cuthbert's, 
 Cornwall. In France the broach occurs de- 
 tached ; e.g. at Nesle ; and St Leu d'Esserent. 
 At NORWICH (622), instead of broaches, the 
 spire has buttresses ; with somewhat unhappy 
 effect. Just as the timber spire-form was 
 copied in stone ; so the stone broach was 
 copied in wood ; e.g. at Braunton, Devon.1l In 
 the broach spire the window plays a prominent 
 part. It should be added that in mediaeval 
 documents the term broacli is usually applied 
 to a whole spire of any type. 
 Sometimes, by way of variation, these spires substitute a pinnacle for the 
 
 Etton. 
 
 Johnson's Reliques, Plate 76. 
 Johnson's Reliques, Plates 15, 13, 22 
 Bowman and Crowther. 
 
 t Wickes' Spires, Plates 21, 23. 
 § Brandon's Parish Churches, 59. 
 IT Illustrated in Prior, 372. 
 
BROACH SPIRES. 
 
 621 
 
 broach. The earliest example is probably that of oxford CATHEDRAL (613). 
 While the broach spire has two or more tiers of windows, this type has but one. 
 This window is large and is placed between the pinnacles. Just as the broach 
 or pinnacle stops the thrust of the oblique sides of the spire, so the big gabled 
 windows bring down more vertically the thrusts of the cardinal sides. It is a 
 type of spire somewhat infrequent in England ; but common in early French 
 work ; e.g. at St Germain, Auxerre ; Chartres ; the Abbaye-aux-Hommes, and 
 St Pierre, Caen; Bayeux ; Bernieres ; Vernouillet ; Limay ; * Migneville ; * 
 Plessis-le-charmant ; * Beaulieu.* It is the typical Normandy spire of the 
 thirteenth century ; and was probably the earliest type of French spire. 
 
 Ketton from S.W. 
 
 Witney from S. 
 
 In a few broach spires the experiment was made of allowing the pinnacle 
 to grow up out of the broach, like a pin out of a pincushion; e.g. WITNEY, f 
 Wollaston, SOUTHAM (614), Wellingborough. The same design occurs at 
 Bayeux. 
 
 Class III. PARAPETTED SPIRES— To all the above spires, however, there 
 is one grave objection. It is that they are difficult to repair. But if a space 
 round the edge of the tower wall be kept clear, it can be used for hauling up 
 material from the ground ; and if it be protected by a parapet, ladders set up 
 
 * Illustrated in Johnson's Specimens of Early French Architecture. 
 
 t That of Witney is similar to, and probably not much later than, the neighbouring spire 
 of Oxford Cathedral. 
 
62: 
 
 PARAPETTED SPIRES. 
 
 against the sides of a spire in construction or under repair cannot slip at the 
 foot. So a pathway was left round the spire. And to get as much breadth as 
 possible for this pathway, it was sometimes set on a corbel table or on a cornice ; 
 so that the parapet overhung the face of the tower. Such an oversailing parapet 
 is well seen in the fourteenth-century spire of SHOTTESBROOKE (395) ; still 
 better in the spire of Autun Cathedral and at the Kreisker in St Pol-de-Leon, 
 Finisterre.* 
 
 (i.) Of the varieties of parapetted spires the first comprises those which 
 have neither pinnacle nor broach ; e.g. WINGHAM, near Canterbury ; LEDBURY ; 
 BRIDGWATER (6 1 2). Simple as it is, the sharp contrast between tower and 
 spire is very effective. It is exceedingly common in the village churches. 
 
 Norwich Cathedral. 
 
 Winsrham. 
 
 An early example of the parapetted spire is the south-western spire of PETER- 
 BOROUGH (112), which is probably c. 1330.! Good examples are seen at 
 Donington, Lincoln ; Bredon, Worcester ; Bramford, Suffolk.^ 
 
 Then we have variants in which the timber type of spire, the broach spire, 
 and the pinnacle spire are surrounded by a parapet. 
 
 (ii.) The timber type of spire, worked in masonry, with a parapet, occurs 
 at St Sepulchre's, Northampton ; and DENFORD, NORTHANTS (614) ; but is not 
 common. 
 
 * Illustrated in Enlart's Manuel, 570. 
 
 + Deene in Northants has a parapetted spire of the second half of the thirteenth century. — 
 R. P. B. 
 
 1 Illustrated in Brandon's Parish Churches, 53, 109, 125. 
 
PARAPETTKD SPIRES. 
 
 623 
 
 (iii.) To this division belongs the parapctted spire which has broaches, but 
 not pinnacles. At BINGHAM (616), the broach is emphasised; the parapet is 
 unimportant. At HOLBEACH (615) the parapet is emphasised; the broach is 
 less important. Naturally this design is rare ; for the broaches, unless small, 
 obstruct the pathway. 
 
 (iv.) This hybrid form combines 
 broach, pinnacle, and parapet. Here 
 again the broach obstructs the path- 
 way ; moreover pinnacle and broach 
 harmonise ill. Nevertheless it occurs 
 in a few early and important spires ; 
 e.g. Grantham and NEWARK (627), 
 c. 1340; and in the fifteenth century 
 at All Saints', Stamford. 
 
 (v.) Still less common and less 
 successful is the parapetted spire in 
 which the pinnacle sticks up out of 
 the broach; e.g. at DESBOROUGH 
 (616). This has a parapet ; but no 
 passage round. 
 
 (vi.) In the end the broach dis- 
 appeared. The way was then clear 
 for the final development of the 
 English spire with pinnacles, simple 
 or compound. 
 
 Many shapes of pinnacle were 
 experimented with. In the Abbaye- 
 aux-Hommes the pinnacles are tri- 
 angular. At Newark they are 
 hexagonal. At Shipton they are 
 circular. All these are exceptional. 
 Most often the)- are square or 
 octagonal ; square in harmony with 
 the tower ; or octagonal in harmony 
 with the spire. In and near Northants 
 there was a special local cult of pin- 
 nacles in the form of a polygonal 
 turret embattled in harmony with 
 the battlemented parapets below ; 
 e.g. KETTERING ; Exton ; Byfield ; 
 Oundle; All Saints', Stamford; 
 Leverington. 
 
 The position of the pinnacle or the turret varies, 
 tinuation of the buttresses of the tower. Or it was 
 tower, as at Witney. Less often, as at Adderbury ; 
 and Leverington, it nestles up to the spire. 
 
 But when the tower was high, and its walls consequently thick, there was 
 
 Ketterint 
 
 Sometimes it is a con- 
 set at the corner of the 
 St Nicholas, Gloucester; 
 
624 
 
 PARAPETTED SPIRES. 
 
 Newcastle. 
 
 ful ; e.t 
 
 more room at the angles, 
 and the builders could in- 
 dulge in pairs of pinnacles ; 
 a lower pinnacle at the angle 
 or over the buttress, a taller 
 pinnacle behind, leading the 
 eye up to the spire.* At 
 Peterborough the outer 
 pinnacle is a square set 
 diagonally ; the inner a 
 triangle. At Chichester 
 there are two octagons. At 
 Salisbury the outer pinnacle 
 is octagonal, the inner 
 square. At ST MARY'S, 
 OXFORD t (631), there is 
 the richest cluster of all. 
 
 (vii.) At St Mary's, 
 Oxford, we saw the inner 
 pinnacles climbing up the 
 tower. More easily to lead 
 the eye "by a bright ladder" 
 to the spire above, there was 
 substituted for the inner 
 pinnacle a flying buttress. 
 This, however, involved the 
 builders in much trouble. 
 The eye demanded that the 
 flying buttress should be 
 massive ; to abut the spire. 
 But the builders seem to 
 have thought that if it were 
 massive and heavy, it would 
 push out the pinnacle on 
 which its foot rested. \ So 
 they made the flying but- 
 tress almost always light 
 and flimsy ; often designed 
 with weak curves ; and 
 usually set at too flat an 
 angle. Unsatisfactory ex- 
 amples are only too plenti- 
 
 COVENTRY (635); PATRINGTON (6 1 2), where the pinnacle straddles 
 
 * Plans of pinnacles are given in Mr James Cubitt's paper on " Spires" ; Building News, 
 May 29, 1885. 
 
 t For plan of the springing of this spire see monograph on St Mary's, Oxford, by Mr T. G. 
 Jackson, 84. 
 
 1 It has done so at Higham Ferrers. 
 
PARAPETTKI) SPIRES. 625 
 
 on two ridiculously weak legs; Rushden ; Moulton ; Billingborough ; Higham 
 Ferrers; Fleet; Loughton-en-le-Morthen ; Thaxted ; Helpringham ; Hanslope. 
 At WILBY (629) the steeple has an octagonal drum; and both spire and 
 drum have flying buttresses; these are set at a proper angle. At KING'S 
 SUTTON (615) and Patrington the inner pinnacle is retained. In the former 
 spire there is a double flight of flying buttresses ; from the outer to the inner 
 pinnacle, and from the latter to the spire. At Patrington the inner pinnacles 
 are grouped into a coronal for the spire ; it may have been to maintain the 
 importance of this coronal that the outer pinnacles were made so insignificant. 
 At LOUTH (611) alone was the difficulty of the flying buttress design com- 
 pletely surmounted. Here the architect, on the one hand, strengthened his 
 pinnacles, and on the other perforated his flying buttresses ; getting in them the 
 appearance of mass without weight. 
 
 (viii.) The flying buttresses once introduced, the next step was to let 
 them converge ; thus crushing the spire proper out of existence ; the new 
 composition consisting simply of four flying buttresses converging to a centre, 
 on which was poised a pinnacle. It was a chef-d'oeuvre of building construction 
 if not of art. It occurs in NEWCASTLE CATHEDRAL (624). In Scotland it is 
 seen in the Cross Steeple, Glasgow ; King's College, Aberdeen ; and St Giles', 
 Edinburgh,* and formerly existed at Haddington and Linlith- 
 gow. It was repeated by Sir Christopher Wren in St Dunstan- 
 in-the-East, London. 
 
 Summary. — Setting aside the less frequent or exceptional 
 or hybrid forms, we arrive at the conclusion that there are 
 three chief types of English spire. The first is the timber 
 type, whether executed in wood or in stone ; the second is the 
 broach spire. Both are without pathway and parapet ; both R , j L ' 
 
 occur mainly in the thirteenth and fourteenth centuries. The 
 third is the parapetted spire without pinnacles ; or with pinnacles, and with or 
 without flying buttresses. It is the parapetted spires that are most character- 
 istic of the later Gothic of England. 
 
 SPIRE ANGLES. — In many cases the builders were quite satisfied to leave 
 the spire to tell its own tale ; e.g. at North Luffenham and Barnack. But at a 
 distance the sharp edges or arrises of an octagonal spire hardly tell sufficiently ; 
 instead of being octagonal, the spire appears to be circular. Now it had been 
 customary early in the thirteenth century to decorate the sides of turrets with 
 slender shafts ; e.g. at Peterborough ; and to continue these shafts up the con- 
 verging arrises of the pinnacle which crowned the turret. In imitation of this, 
 many spires were built with their edges emphasised by a slender roll ; e.g. 
 Witney ; Peterborough ; St Mary's, Oxford ; Wingham ; Ketton ; Bridgwater ; 
 Ledbury ; Coventry. For this roll there was substituted in some fourteenth cen- 
 tury spires a band of the ballflower characteristic of the period ; e.g. at Salisbury 
 and the western spires of Lichfield. The crocket was also greatly in vogue 
 during and after the fourteenth century : and many a spire employed this 
 beautiful ornament to emphasise its edges ; e.g. Higham Ferrers, Walcot, 
 Wimmington, King's Sutton, Oundle, Moulton, Louth, Norwich. It was especially 
 * For measured drawings see Building News, Sept. 7, 1877. 
 2 R 
 
626 
 
 STRINGS AND BANDS. 
 
 commended to the builders in that it afforded ready access to all parts of the 
 spire in case repairs were needed, without the expense of putting up scaffolding ; 
 
 a function which it performs to 
 the present day. At RUSHDEN 
 (625) there can be no doubt 
 that the crockets were purposely 
 designed to give foothold to the 
 steeplejack. 
 
 Strings and Bands. — In 
 many cases the sides of the spire 
 also were ornamented ; some- 
 times with strings or bands, 
 sometimes with spire-lights. 
 The western spires of LICHFIELD 
 and that of SOLIHULL are not- 
 able instances of multiplicity of 
 strings ; with these we may 
 compare the central spire of 
 Semur-en-Auxois. Other ex- 
 amples, where they are used 
 less emphatically, are Denford, 
 Southam, and St Sepulchre's, 
 Northampton. Sometimes, in- 
 stead of a string, there was a 
 broad band of carved ornament ; 
 King's Sutton has one small 
 band ; Salisbury and the rebuilt 
 spire of St Mary Redcliffe, 
 Bristol, have three. It may be 
 doubted whether these horizontal 
 lines do not lessen the apparent 
 height of the spires ; that may 
 be the reason why they were not 
 employed more frequently. 
 
 Plate-tracery. — Rarely 
 with us, e.g. SOLIHULL, War- 
 wickshire ; Fleet, Lincolnshire ; 
 but frequently in France ; e.g. 
 St Pierre and St Sauveur, Caen; 
 the sides of the spire were per- 
 forated with bands of quatrefoils. 
 It has been asserted * that the 
 object of these quatrefoils, as of 
 spire windows, is to lessen wind 
 pressure, by allowing the wind 
 to pass through the spire. The reverse is the case. Openings in the sides 
 * Paley's Gothic Architecture, 259. 
 
 Solihull. 
 
SPIRE WINDOW'S. 
 
 C)2J 
 
 of the spire give the wind a 
 better grip. 
 
 Spire Windows or Spire- 
 lights or Lucarnes. So far as 
 openings in the sides of a spire 
 have practical value, it is to light 
 the interior,* thus facilitating re- 
 pairs. But their value is mainly 
 decorative. In many spires they 
 were eschewed ; e.g. Barnack, 
 Ledbury, Bridgwater, Patring- 
 ton. In spires of the type of 
 that of Oxford Cathedral there 
 was usually but one tier of large 
 windows ; set at the foot of the 
 spire ; so also at Bloxham and 
 St Mary's, Oxford. 
 
 The earliest fenestrated 
 spires had two or three tiers of 
 windows ; e.g. Rauceby, War- 
 mington, Sleaford, Polebrook, 
 Anwick, North Luffenham. 
 There was quite a furore for 
 windows in the district of the 
 Northamptonshire broach spire ; 
 at first they could hardly be too 
 numerous or too large. The 
 windows were sometimes elabo- 
 rated to the detriment of the 
 spire. All the above examples 
 and many more, have the win- 
 dows so large that their vertical 
 lines seriously break up the con- 
 verging lines of the spire. In 
 the florid western spires of LICH- 
 FIELD (frontispiece I windows are 
 in excess. But at NEWARK and 
 in later spires, such as Louth 
 and Coventry, better design 
 came in. The number and the 
 dimensions of the windows were 
 greatly curtailed. At Aloulton 
 and Coventry the upper windows 
 are quite small ; at Louth win- 
 dows play a still less important 
 part. In the later spires too, e.g. 
 at Aloulton, the upper windows 
 
 * Rarely, as at Weobley, Herefordshi 
 in the spire. 
 
 mm* 
 
 -io'.v : 
 
 iilll 
 
 
 
 Newark, 
 re, and Ringstead, Northants, the belfry windows are 
 
628 
 
 SPIRE WINDOWS. 
 
 were still further curtailed in order that their rapid diminution upward might 
 make the spire look taller ; even the crockets diminish in size upward at 
 
 Moulton, for the same reason. 
 
 Sometimes, as at LICH- 
 FIELD, King's Norton, Grant- 
 ham, the windows were set in 
 tiers, not alternating, on every 
 face ; but this was rare. More 
 often, especially in broach spires, 
 e.g. Ketton, EWERBY (630), Wol- 
 laston, Threckingham, Bing- 
 ham, Newark, Walcot, An wick, 
 Holbeach, North Rauceby, St 
 Mary Stamford, Billingborough, 
 Moulton, the windows are set 
 alternately ; the lowest and 
 upper four occupying the 
 cardinal sides of the spire, and 
 the intermediate tier the oblique 
 sides. Usually the oblique sides 
 are without windows ; all the 
 windows being set on the 
 cardinal sides of the spire ; as 
 at Kettering, Whittlesea, Rush- 
 den, Market Harborough, By- 
 thorne, Desborough, Oakham, 
 Denford, Barnwell, Luddington, 
 Ringstead, Warmington, Hig- 
 ham Ferrers, FRAMPTON (618). 
 Bar-tracery Spires. — 
 Abroad, with marvellous skill, 
 the spire was sometimes com- 
 posed wholly, or nearly so, of 
 bar-tracery ; e.g. at Chartres ; 
 Antwerp; Strasburg; Freiburg; 
 St Marie de l'Epine, Chalons- 
 sur-Marne ; St Stephen's, 
 Vienna ; Batalha ; Burgos. It 
 might perhaps be pleaded in 
 extenuation of this treatment 
 of stone as if it were wood or 
 metal, that in these openwork 
 spires wind pressure is reduced, 
 and less burden imposed on 
 the supports. 
 A spire may be the same on plan as its tower. Thus 
 
 Stanwick 
 
 Plan of Spires — 
 the Irish Round Towers 
 
 and Southese Church, Sussex, have a conical capping 
 
PLAN OF SPIRES. 
 
 629 
 
 or spire. The early square Norman towers, e.g. Thaon and Priestholme, had low- 
 square spires; which also'appear later in many Sussex churches,,-.^. Rodmel) 
 and Slaugham. In the Early Gothic of France* square spires, either hipped to 
 a point or to a ridge, had a much greater vogue ; especially in Normandy and 
 the South, where more than twenty 
 examples survive, some mounting to 
 the fifteenth century. So again an 
 octagonal tower may support an 
 octagonal spire, as at STANWICK 
 (628) and Wickham Market. 
 
 But in spire design contrast was 
 found more telling than harmony. 
 So the round tower of Piddinghoe, 
 Sussex, carries an octagonal spire. 
 But experiment proved that by far 
 the best combination was that of an 
 octagonal spire and a square tower : 
 and to this type nearly all our 
 steeples conform. 
 
 A sort of compromise, however 
 — and like most compromises, not 
 very satisfactory — was to introduce 
 the octagonal form in the upper part 
 of the tower ; so that the steeple 
 consisted of three stories ; the square 
 tower, the octagonal drum, and the 
 octagonal spire. It appears at Bar- 
 nack and Exton,+ Rutland ; WILBY, 
 Northants ; Graft ham, Hunts ; Blox- 
 ham, Oxfordshire; Masham, York; 
 Patrington ; Norwich ; Coventry. In 
 the greater number of these the drum 
 fails to extricate itself sufficiently 
 from the tower ; making it difficult 
 to see where the tower ends and the 
 spire begins; or else it is so smothered 
 in panelling and pinnacles, that its 
 presence is hardly perceptible. 
 
 A more important consideration 
 is the height the tower should have 
 in proportion to that of the spire. 
 As to this, great diversity of practice 
 prevailed. On the whole, we seem 
 
 to have got our towers too squat. How far superior is the eftect of a tall 
 
 spire, when seated on a tall tower, is seen at once by inspecting such examples 
 
 as Louth, Coventry, Salisbury, Patrington, which are among the finest steeples 
 
 * E.g. Tournai Cathedral and St Gille, Caen. + Illustrated in YVickes' Spires, 24. 
 
KEg E3 
 
 10 5 fO 20 
 
 W ■?<>« 
 
 Ewerby. 
 
PROPORTIONS OF S1TRKS. 
 
 63 1 
 
 in England. It was in Normandy that 
 this principle was most fully recognised ; 
 it is to the height, slenderness, and light- 
 ness of their towers that such steeples as 
 Bernieres and Vernouillet owe their sur- 
 passing charm. Sometimes our towers 
 were crushing!)' squat in proportion to 
 the needle-like spires which they carried ; 
 e.g. BRIDGWATER (612) ; Hemingbor- 
 ough, York ; Walcot, Lincoln. While at 
 Louth, Newark, Kettering, and Ketton, 
 tower and spire areabout equal in height,* 
 at Hemingborough the spire is twice as 
 high as the tower. But just as the spire 
 may be too tall for the tower, so the 
 tower may be too tall for the spire. This 
 was naturally the case with many of 
 the earlier spires ; e.g. Oxford Cathedral, 
 Barnack, Sleaford ; and with several later 
 spires in the Norfolk and Lincolnshire 
 marshland ; e.g. Walsoken, Fleet, and 
 tilxev ALL .saixts' {6\2). But almost 
 at once the spire soared upward at a 
 sharp angle ; e.g. at Ketton and Salis- 
 bury ; and earlier still in France ; e.g. St 
 Germain, Auxerre. The change from a 
 somewhat obtuse to a sharp angle was 
 so rapid that there must have been some 
 other than aesthetic reasons for it. Pro- 
 bably the builders calculated that the 
 acuter the angle of the spire, the less it 
 tends to thrust out the corners of the 
 tower on which it rests ; moreover an 
 acute spire is less affected by wind pres- 
 sure. The later spires on the whole vary 
 from an angle of fifteen degrees, e.g. 
 Chichester, to one of ten degrees, e.g. 
 Louth and Moulton : in later Gothic they 
 were generally slender. 
 
 * In the best broach spires of Northants 
 the spire and tower seem to be generally equal 
 in height, and the tower is generally from 3 to 
 3} squares high from the ground to the corbel- 
 table. Louth steeple is 300 feet high from the 
 street ; 294 feet from the floor ; the tower and 
 spire are each 147 feet high. At St Mary's, 
 Oxford, the tower is 87 feet 5 inches from the 
 floor ; the spire is 101 feet 1 inch high. 
 
 Oxford St Mary. 
 
632 ENTASIS OF SPIRES. 
 
 EXTASIS. — In some spires the sides are straight ; e.g. Kettering; St Mary's, 
 Oxford ; and Salisbury ; several have been shortened * at the top after damage 
 from storm or lightning, and have thus accidentally lost their straight outlines, 
 e.g. Grantham ; a few were built with slightly convex outline, like the Greek 
 columns. This entasis, as it is called, was given because a long upright line 
 appears hollow at or about the centre. The entasis f corrects this optical 
 illusion. It is quite slight in a spire, probably not amounting to more than 
 I inch in 60 feet in the best examples. This is greatly exceeded, however, 
 in the circular spires of the district of the Charente ; e.g. Saintes ; and in a few of 
 our own spires ; e.g. the Lincolnshire spires of Caythorpe, Welbourn, and Glinton ; 
 which have been opprobriously designated "sugarloaf" spires. It occurs in 
 the twelfth-century spire of St Germain, Auxerre ; and in the Normandy 
 spire of Bretteville. Other English examples are Gedling and Wittering. At 
 Louth the necessary correction is obtained by increasing the projection of the 
 crockets about one-third of the way up. 
 
 Unity of Tower and Spire. — In many cases the tower was completed 
 some considerable time before the spire was built ; e.g. Norwich ; Salisbury ; 
 St Mary's, Stamford ; Gedney ; Walsoken. This may have familiarised the 
 builders with the idea of designing the tower without reference to the spire. 
 It might chance that funds for adding a spire might never be forthcoming. It 
 was well therefore to design a tower which should be complete in itself: e.g. St 
 Martin's, Stamford, has an excellent Perpendicular tower ; but it has squinches ; 
 so that a spire was evidently intended. 
 
 But with some exceptions — Salisbury is one — the best spires were those 
 which were designed not from the tower upwards, but from the weathercock 
 downward ; e.g. Louth, Bingham, and Kettering are spires all the way to the 
 ground. I We may indeed go so far as to lay down that where we find a bad 
 tower produced by the removal of the spire, the presumption is that the 
 steeple is a success. This is true of Patrington ; St Mary's, Oxford ; Louth ; 
 Bloxham ; Heckington ; Billingborough ; § St Mary Redcliffe, Bristol ; and 
 more especially of the broach spires ; e.g. Bingham ; Aumsby ; Barnwell ; 
 Threckingham ; Walcot ; Wollaston. Nevertheless it must be admitted that 
 there are some steeples which would remain excellent as towers, if the spires 
 were removed ; e.g. Newark ; Coventry ; Moulton ; Whittlesea. 
 
 Distribution of Ornament. — Some of the earlier spires tended to be 
 florid. At Grantham the architect has given the spire double the usual number 
 of windows ; together with parapet, pinnacles, broaches, and crockets. Newark 
 is better ; there are not so many windows ; and rolls are substituted for 
 crockets. For a lesson in the use, not the misuse, of ornament we must study 
 
 * Others have been heightened at "restorations," e.g. Oundle. — R. P. B. 
 
 t The entasis of a spire may be due to a curious bit of masoncraft. "The stone-masons 
 who put up a tall pole in the axis of the spire they were building, and who diminished the height 
 of this pole from day to day, building always so that the slope of the spire was directed towards 
 the upper end of the pole, were giving the same entasis to the slope that a Greek artist gave to 
 his columns" (Sturgis' Dictionary of Architecture and Btiilding, iii. 267). 
 
 \ But this may be overdone ; as at St Sepulchre's, Northampton ; where the vast projection 
 of its diagonal buttresses makes the tower a mere annexe to the spire. 
 
 § Johnson's Reliques, 34. 
 
SPIRE DESIGN. 633 
 
 the composition of such spires as Patrington ; St Mary's, Oxford ; Louth ; and 
 Coventry. 
 
 PATRINGTON STEEPLE (634) is practically a three-storied elevation. The 
 lower story, i.e. the lower stage of the tower, is absolutely plain. So is the spire. 
 All this to show off better the ornament of the intermediate story, which 
 includes the upper stage of the tower and the pinnacled drum. In this steeple 
 all the ornament is concentrated at the mid-zone. 
 
 In ST MARY'S, OXFORD (63 1), the mid-zone is the pinnacled base of the 
 spire. Owing to the wealth of ornament here, it is possible to give more orna- 
 ment to the tower than at Patrington, and to provide the spire with mils. 
 
 At LOUTH (611) again the ornament grows in richness upward till it 
 culminates at the base of the spire. For the great west window is substituted 
 a pair of windows ; in the next story the windows have crocketted ogee drip- 
 stones ; then comes a galaxy of panelled pinnacles, big and little, battlements, 
 windows, and flying buttresses. Finally this wealth of ornament dies away in 
 the crockets converging to the summit. Louth is beyond compare. One of 
 the most remarkable of its characteristics is the vast height to which it seems 
 to soar; therein surpassing every spire. It is moderately high; 300 feet; but 
 it seems lofty beyond finite measurement. 
 
 If any prefer a more florid type, what can surpass ST MICHAEL'S, covextkv ? 
 (635 ). As the eight string-courses show, it is an elevation of nine stories. These 
 continually increase in richness and beauty from the doorway to the battlements 
 which crown the tower. From that point the decoration gradually dies away. 
 The sixth story, the octagonal drum, is a little less rich. The seventh and 
 eighth have panelling and windows ; the ninth has windows only. Again, the 
 nine stories are arranged in triplets ; the highest of the three lower stories 
 being the richest of the three. In the spire, which also has three stories, this is 
 reversed ; the lowest story being the richest. The three middle stories are the 
 richest of all ; and of these the central one has tiers of statuettes to give it 
 predominance over the other two. In the presence of such design as that 
 of Louth and Coventry, it is futile indeed to speak of our late Gothic as 
 " debased." * 
 
 Elevation and Perspective.— A special difficulty which confronted the 
 builders was that a square tower looks different, but an octagonal spire the same, 
 whether seen from a cardinal point or obliquely. A square tower is half as wide 
 again when seen diagonally as it is when seen full-face. Therefore a spire, in 
 relation to its tower, if right when seen full-face, is wrong when viewed diagon- 
 ally; and vice versa ; it cannot be right both ways. The point of view from 
 which a square tower is seen therefore affects the aspect of the spire. Seen from 
 north, south, east, or west, a spire seems more massive ; seen from the inter- 
 mediate points, it seems more slender.+ This seems occasionally to have been 
 recognised ; e.g. Helpringham Church is approached by a long road leading 
 straight eastward to the spire ; which, like Threckingh.am spire, from this point 
 
 * For an appreciation of Moulton steeple, see Sharpe's Lincoln Excursion, 99, 147. 
 
 t This is most so with a parapetted spire ; especially if it has turrets instead of pinnacles. 
 If it has pinnacles, and these nestle up to the spire, it looks equally well at an angle or seen 
 straight on ; e.g. Nassington. — R. P. B. See Mr T. G. Jackson on St Mary's, Oxford, 86. 
 
634 
 
 Patrinerton from W. 
 
635 
 
 1 
 
 3||i ! ||| T |i^!i T pT : 
 
 fk 
 
 
 r 1 
 
 « it- 
 
 i 
 
 I 
 
 $ — r 
 A' i 
 
 m 
 
 m 
 I $1 
 
 y ffii 
 
 I 
 
 
 IO 5 lO 20 JO 40 50 
 
 SCALE OF 
 
 Coventry St Michael. 
 
6^6 
 
 SPIRE CONSTRUCTION. 
 
 of view is seen to much greater advantage than when seen diagonally.* The 
 slender spire of Louth also looks adequate when approached from the only point 
 from which it was intended to be seen ; viz. the west. 
 
 Spire Construction. — When the spire superposed on a square tower is 
 octagonal, only the cardinal sides rest on the walls of the tower. For the 
 oblique sides therefore some support has to be devised. This support is 
 obtained by building concentric arches receding towards each angle of the 
 tower ; as is seen at Oxford Cathedral ; an early spire, where the squinches 
 are rude and archaic ; at SALISBURY ; SHOTTESBROOKE (395) ; COVENTRY 
 (635); and LOUTH (611). 
 
 Special precautions were sometimes taken to prevent the capstone being 
 dislodged by the movements of the vane. The interior of Salisbury spire is 
 filled with a timber frame, consisting of a central piece with arms and braces. 
 This entire frame, the arms of which were made to support floors which served 
 as scaffolds whilst the spire was building, is hung to the capstone of the spire 
 by iron cross bars, and by the iron standard of the vane, which is fixed to the 
 upper part of the central piece. f 
 
 Salisbury Cathedral Spire. 
 
 To diminish the weight on the tower as much as possible, the later spires 
 were built very thin. The south-west spire of Chartres, an Early Gothic 
 cathedral of exceptionally massive construction throughout, is 31 \ inches 
 thick at the bottom and 1 if inches at the top. That of Salisbury is 
 24 inches thick at the bottom and 9 inches at the top. But in the two great 
 spire-building districts — the Caen stone district of Normandy, and the Oolitic 
 limestone districts of England — these dimensions were considerably reduced. 
 The first gh feet of Kettering spire are 14 inches thick ; above, it is 6 inches. 
 The Normandy spires of the thirteenth century are very often only from 5 to 
 6 inches. + The thickness of the spire of St Mary's, Oxford, is about 14 inches, 
 clear of the spire lights and pinnacles ; and at the top, before it becomes solid, 
 from 6 to 7 inches. Oundle spire, after the first 9 or 10 inches, is about 7 inches 
 
 Sharpe's Lincoln Excursion, 77. 
 Ruprich-Robert, 163. 
 
 t Murray's English Cathedrals, 87. 
 
SPIRE CONSTRUCTION. 
 
 637 
 
 thick. Louth* spire is 10 inches thick below ; 5 inches thick above. The spire 
 of Moulton, Lincolnshire, is not more than 5 inches thick. 
 
 The courses of a spire might be laid horizontally ; or with the joints at 
 right angles to the slope of the spire. In the latter case the spire is a genuine 
 vault ; simply a tall elongated vault. + But being a vault, it has a decided 
 outward thrust, diminished indeed by the acuteness of the angle of most spin-,, 
 but still present, and needing to be guarded against. Moreover it is easy for 
 the rain to penetrate into its inclined joints. For these reasons the builders 
 seem to have preferred to build the spires in horizontal courses ; i.e. corbelled 
 inwardly. 
 
 * For Louth Church and spire see paper by Mr James Fowler in the Transactions of the 
 Lincoln Diocesan Arch. Society, 1S73. The Churchwardens' accounts of the building of the 
 spire have been printed in Archccologia, x. 70-98 ; see also Britton's Archil. Antiq.. iv. 1-7. 
 The spire cost £,2jg. 10s. 5d. 
 
 t Indeed in the French example of Loches, it is difficult to say whether the nave is roofed 
 with a couple of domed vaults or a couple of spires. Classification of Romanesque, 273. 
 
 ! If ill h/.Jf.J 
 Winchester. 
 
Chapter XLII. 
 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 [In the following list are included most of the larger churches ; and of the smaller churches those for 
 whose date documentary evidence exists. Where no authority is quoted, the date is fixed by architectural 
 evidence only, and must be regarded with suspicion. The documentary evidence also must be received 
 with caution; see paper by the author in the Journal of the R.I.B.A., 26th November 1898, on The 
 Comparative Value of Documentary and Architectural Evidence in Establishing the Chronology of the 
 English Cathedrals. Fuller quotation of authorities will be found in the seventh edition of Rickman's 
 Gothic Architecture. Additions to the list and corrections will be welcomed.] 
 
 ARLINGHAM, Gloucestershire. Parish Church. 1373. Bloxam, 227. 
 
 ARUNDEL, Sussex. Made Collegiate in 13S0. 
 
 ASHBOURNE, Derbyshire. Parish Church. Parts are 1235-1241. Inscription on 
 
 brass plate : facsimile in Rickman, 170. 
 ASHFORD, Kent. Parish Church. Rebuilt 1461-1490. Rickman, 30S. 
 AUGHTON, Yorkshire. Parish Church. Steeple, 1536. Rickman, 318. 
 AVENING, Gloucestershire. Parish Church, c. 1125. Messrs Carpenter and 
 
 Ingelow in Transactions of Bristol and Gloucester Archaeological Society, xiv. 5. 
 
 BALSHAM, Cambridgeshire. Parish Church. Chancel — 1390-1400. Inscription on 
 
 brass plate : facsimile in Rickman, 293. 
 BANGOR CATHEDRAL, Carnarvonshire. Secular Canons. Built, except the 
 
 Choir, 1 509-1 5 3 2. Rickman, 315. 
 BARTON-UNDER-NEEDWOOD, Stafford. Parish Church. 1517. Rickman, 316. 
 BATH, Somerset. Parish Church. Formerly Abbey Church of Benedictine Monks. 
 
 Begun c. 1500 ; nearly finished in 1539. Rickman, 313. 
 BATTLE ABBEY, Sussex. Benedictine Monks. Founded 1067. Refectory — First 
 
 half of XIII. Cent. Gatehouse — License to crenellate granted 1339. Rickman, 
 
 240. 
 BEAULIEU ABBEY, Hampshire. Cistercian Monks. Founded 1204; Church 
 
 hallowed 1246; Frater, now Parish Church, between these dates. Mr Harold 
 
 Brakspear. 
 BERKHAMSTED, Hertfordshire. Parish Church. First half of XIII. Cent. 
 BEVERLEY MINSTER (St John's), Yorkshire. Secular Canons. Choir and Transepts 
 
 — c. 1225 — c. 1245. Nave — c. 1320 — c. 1349. West Front — c. 1380 — c. 1430. 
 
 Mr John Bilson in Architectural Review, iii. 199. 
 BEVERLEY ST MARY, Yorkshire. Parish Church. Originally aisleless Church 
 
 of second quarter of XII. Cent. Aisles added to nave in second quarter of 
 
 XIII. Cent. Chapel on east side of north transept, east aisle of south transept, 
 
 and south aisle of chancel, of end of XIII. and beginning of XIV. Cent. North 
 
CHRONOLOGY OF ENGLISH CHURCHES. 639 
 
 arcade of chancel and north chapel, of second quarter of XIV. Cent. Nave 
 clerestory and west end, early XV. Cent., followed by chancel clerestory and 
 reconstruction of transepts, to middle of XV. Cent. Nave and tower reconstructed 
 after fall of tower in 1520. Mr John Bilson. 
 
 BINHAM PRIORV, Norfolk. Benedictine Monks. Nave — c. 1100. Was in exist- 
 ence 1093, but not fully endowed till 1101-1106. (Dugdale's Monasticon, iii. 
 341. West Front — Built by Richard de Parco, who was Prior 1 226-1 244. 
 Matthew Paris. 
 
 BIRKIN, Yorkshire. Parish Church. Granted to the Templars in 1152. 
 
 BISHOP AUCKLAND, Durham. Hall, now the Chapel— c. 1190. 
 
 BLYTH PRIORY, Notts. Benedictine Monks. Nave— Founded, and probably 
 begun, 1088. Monograph with measured drawings by Mr C. C. Hodges ; 
 Hexham, 1SS1. 
 
 BLYTHBURGH, Suffolk. Parish Church. 1442-1473. Suffolk Arch. Institute, iv. 
 233 and 422. 
 
 BOLTON PRIORY, Yorkshire. Austin or Black Canons. Parts of Nave, Tran- 
 sept, and Choir — Begun c. 1151. Parts of Nave, North Aisle, and Older Facade — 
 Early XIII. Cent. Part of Choir — Second quarter of XIV. Cent. New Front 
 and Tower — Inscription has date 1520. Builder, 4th May 1895. 
 
 BOLTON PERCY, Yorkshire. Parish Church. Built by its Rector, 1411-1423. 
 
 BOSTON, Lincolnshire. Parish Church. Foundations laid in 1309. Nave and 
 Chancel — Later. Assoc. Soc. Reports, x 175. 
 
 BOURN ABBEY, Lincolnshire. Austin Canons. Founded: 138. Assoc. Soc. Reports, 
 xx. 7. 
 
 BOXGROVE PRIORY, Sussex. Benedictine Monks. Choir— c. 1235. Same in 
 carving and style as the south chapels of Chichester nave and east chapel of north 
 side of nave. Mr E. S. Prior. 
 
 BRADSOLE, ST RADEGUND'S ABBEY, Kent. Premonstratensian or White 
 Canons. Founded in 1191. W. H. St John Hope in Arch. Cant., xiv. 140. 
 
 BRAY, Berks. Parish Church. Partly rebuilt, 129$ sea. Rich/nan, 235. 
 
 BRECON PRIORY, Brecknock. Benedictine Monks. Choir and Transepts— Early 
 XIII. Cent. Alterations — Second half of XIV. Cent. 
 
 BRIDLINGTON PRIORY, Yorkshire. Austin Canons. North Aisle, North Porch 
 and both Arcades of A r ave, c. 1250. Upper Work, North Side, c. 1270; South 
 Side, c. 1290. West End and SAV. of Nave, c. 14S0. Assoc. Soc. Reports, iii. 40. 
 
 BRINGTON, Northants. Parish Church. 1445-1457. Churches of Northants, 261. 
 
 BRINKBURN PRIORY, Northumberland. Austin Canons. Late XII. Cent. 
 
 BRISTOL ABBEY. Austin Canons. Became a Cathedral in 1542. Chapter House 
 —Third quarter of XII. Cent. (Rickman, 99.) Elder Lady Chapel — Early XIII. 
 Cent. Abbot David was buried in it in 1253. Choir — Built by Abbot Knowle, 
 129S-1332. {Abbot Newland's Roll.) Central Tower — c. 1450 — c. 1470. Mr G. 
 E. Street. 
 
 BRISTOL, St Mary at Redcliffe. Parish Church. Inner Porch — End of XII. 
 Cent. Porch and Tower — Begun 1292. {Rickman, 234.) South Transept — 
 Second quarter of XIV. Cent. ; finished in 1376 by William Canynges the elder. 
 The rest built chiefly by William Canynge the younger, who died in 1474. 
 William of Worcester in Canon Norris' Guide to the Church. 
 
 BRISTOL, St Stephen's. Parish Church. 1455. 
 
 BUCKLAND, Herts. Parish Church. 134S. Inscription reproduced in Rickman, 
 241. 
 
640 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 BUILDWAS ABBEY, Shropshire. Cistercian Monks. Founded 1135. Some of it 
 is before 1148 ; but the greater part immediately after. Mr Harold Brakspear. 
 
 BURNHAM ABBEY, Bucks. Austin Canonesses. Was founded 1277 by Richard, 
 King of the Romans, and the buildings are of this date, though there is no 
 documentary evidence of their erection. Mr Harold Brakspear. 
 
 BURY ST EDMUNDS, Suffolk Abbey. Benedictine Monks. Ruins — Some part 
 completed in 1095; consecration in 1096. {Rickman, 82.) Tower Gateway — 
 1121-1130. {Rickman, 92.) Gatehouse — After 1327. (Rickman, 239.) Dr M. R. 
 James in Camb. Antiq. Soc. Sz>o publications, No. 28. 
 
 BURY ST EDMUNDS, St Mary's. Parish Church. 1424-1433. Rickman, 299. 
 
 BYLAND ABBEY, Yorkshire. Cistercian Monks. The monks removed to Byland 
 in 1 1 77 from a few miles away; probably had begun building some few years 
 earlier. 
 
 CAMBRIDGE, Corpus Christi College. Old Court — 1352-1377; practically no 
 
 old work left. 
 CAMBRIDGE, Christ's College. First Court — Nearly finished in 1509; entirely 
 
 refaced in the XVIII. Cent. 
 CAMBRIDGE, Holy Sepulchre. Probably n 20-1 140. 
 CAMBRIDGE, Holy Trinity. Parish Church. Tower-piers— Late XIII. Cent. 
 
 Nave and Aisles — Late XIV. Cent. Transepts, North Porch, and Clerestory 
 
 of Nave— Late XV. Cent. 
 CAMBRIDGE, King's College. First stone of Chapel laid 1446. Parts in white 
 
 stone (magnesian limestone) are 1446-1461. The rest by Henry VII. and his 
 
 executors, 1508-1515. Fan Vault — Begun 1512. 
 CAMBRIDGE, Pembroke College. First Court — 1346 seq.; but entirely altered. 
 CAMBRIDGE, Peterhouse. Hall — Partly c. 1286; two doorways (restored) left. 
 
 North Quadrangle — 1424. West Quadrangle — 1431. Kitchen — 1450. Stone 
 
 Parlour — 1460. Courts refaced in the XVIII. Cent. 
 CAMBRIDGE, Queen's College. First Court and Far Side of Second Court — 
 
 1447 and 1465. 
 CAMBRIDGE, St Andrew the Less. The so-called "Abbey" Church. Early 
 
 XIII. Cent. 
 CAMBRIDGE, St Benedict. Parish Church. Tower and N.E. Angle of Nave — 
 
 Pre-Norman. Nave — XIII. Cent. Aisles — Rebuilt 1853 and 1872. 
 CAMBRIDGE, St Edward. Parish Church. Late XIV. Cent. ; except Tower, 
 
 some parts of which are XIII. Cent. Aisles — Added c. 1445. 
 CAMBRIDGE, St John's College. Gatehouse — c. 151 1. First Court — 1510-1520; 
 
 refaced. 
 CAMBRIDGE, St Mary the Great. Parish Church. Begun 1478. Nave 
 
 Roof— 1506. Tower — 1491 ; 1530- £ elf ?y Stage — 1593-1608. 
 CAMBRIDGE, St Mary the Less. Parish Church. 1340-1352. 
 CAMBRIDGE, St Michael. Parish Church. After 1323. 
 CAMBRIDGE, St Radegund. Benedictine Nuns. Chapel of Jesus College — 
 
 Early XIII. Cent. A. Gray, Camb. Antiq. Soc. 8vo publications. 
 CAMBRIDGE, Stourbridge Chapel. Hospital of St Mary Magdalen. — c. n 25. 
 CAMBRIDGE, Trinity College. Great GW?— 151S-1535. Chapel— c. 1564. 
 
 Great Quadrangle — Mainly 1593-1615. The dates of buildings in Cambridge are 
 
 taken from the Architectural History of the University of Cambridge, by Robert 
 
 Willis and J. W. Clark. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 641 
 
 CANTERBURY CATHEDRAL, Kent. Benedictine Monks. Parts in West 
 Crypt— Rebuilt by Archbishop Lanfranc, 1070-1077. Choir and Crypt— Rebuilt 
 by Ernulf, Prior 1096-1107, and Prior Conrad. Dedicated 1130. Choir again 
 rebuilt— William of Sens, 1175-117S; and William the Englishman, 1179-1184. 
 Black Princes Chantry — 13 70-1 3 7 9. Western Bays of Nave — 1 379-1 381. Rest 
 of Nave and West Side of Transepts — 13S2-1400. Cloister — 1397-1412. St 
 Michael's, Somerset, or Warrior's Chapel — Finished 1439 (Hope). South-west 
 Tower— 1440-1452. Lady or /Jean's Chapel— c. 144S-1455 (/Pope). Central 
 Tower — 1495-1503. Christ Church Gateway — 1517. See Professor Willis' 
 Canterbury. 
 
 CANTERBURY, St Augustine's Abbey. Benedictine Monks. Founded c. 597. 
 Begun by Abbot Scotland 1070-1087 ; his work finished in 1091. .SV Augustine's 
 Gateway — 1309. Mr IV. H. St John Hope. 
 
 CANTERBURY, Holy Cross. Parish Church. Removed to present site, 1380. 
 Rickman, 2 go. 
 
 CARLISLE CATHEDRAL, Cumberland. A house of Austin Canons was founded 
 and endowed in 1101. Nave and Transept — Probably soon after 1101. Became 
 a cathedral in 1133. Choir — Middle of XIII. Cent. Choir-piers, Triforium, 
 and Clerestory — Rebuilt after fire of 1292 ; prolonged up to springing of arch of 
 east window by 1322 : east window glazed 13S0-13S4. C. J. Ferguson in Builder, 
 6th May 1893. 
 
 CARTMEL PRIORY, Lancashire. Austin Canons. Founded in 11SS. South 
 Aisle of Presbytery — Second quarter of XIV. Cent. Nave — Rebuilt later. 
 
 CASTLE ACRE, Norfolk. Cluniac Monks. Founded before 1189 or in 1190. 
 W. H. St John Hope in N. and N. Arch Soc, 1894. 
 
 CASTOR, Northants. Parish Church. Tower and parts are 11 24. Inscription 
 over doorway : illustrated in Rick/nan, 93. 
 
 CATTERICK, Yorkshire. Parish Church. 141 2. The contract for building was 
 printed in 1834 by Rev. J. Raine. 
 
 CHATTKR1S PRIORY Cambridge. Benedictine Nuns. Probably rebuilt after fire 
 of 1310. Consecrated 1352. Rickman, 242. 
 
 CHELMSFORD, Essex. Parish Church. Repaired or rebuilt 14S9 sec. Rickman, 
 312. 
 
 CHEPSTOW PRIORY, Monmouth. Benedictine Monks. Begun between 1135 
 and 1 154, if not earlier. 
 
 CHESTER CATHEDRAL, Cheshire. Benedictine Monks. Founded and richly 
 endowed by Hugh Lupus in 1093. North Transept — 1093. East Bay of Choir 
 Aisle — c. 1200. Chapter House and Vestibule — First half of XIII. Cent. Choir — 
 Was rebuilding in 1283. South Transept and South Nave — Chiefly in the second 
 quarter of the XIV. Cent. Nave — Finished in XV. Cent. 
 
 CHESTER, St John. Secular Canons. Cathedral in 1075. Fragments West of 
 Nave — Soon after 1075. Nave — Ground story, c. 11 70; triforium, c. 1180; 
 clerestory, c. 1 200. 
 
 CHETWODE, Buckinghamshire. Austin Canons. Founded in 1244. Chancel. 
 Rickman, 172. 
 
 CHICHESTER CATHEDRAL, Sussex. Secular Canons. Choir— c. 1091 ; con- 
 secrated 1 108. Nave — c. 1114-1123. Lady Chapel — Finished before 11 75. 
 Retrochoir, Recusing and Vaulting — After fire of 11 86. Eastern Chapels, Library, 
 Vestry, South and North Porch — First quarter of XIII. Cent. South Chapels 
 of Nave, East Chapels of North Side of Nave, and Doorway of South Porch — 
 
 2 S 
 
642 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 Second quarter of XIII. Cent. West Chapels of North Side of Nave and West 
 Porch — Third quarter of XIII. Cent. Lady Chapel — 1288-1304. South Transept 
 — Remodelled 1305-1337. Spire (which fell down in i860) — Begun c. 1370; 
 finished c. 1470. Campanile — Begun c. 1370; still in progress in 1436. Cloisters 
 — XV. Cent. Professor Willis on Chichester, and E. Sharpe on New Shoreham, 
 Report by Mr Gordon M. Hills, and paper by Mr E. S. Prior on Chichester 
 Masoncraft in Proceedings of Harrow Architectural Club, 1904. 
 
 CHICHESTER, St Mary's Hospital. Hall and its Chapel— c. 1290. 
 
 CHIPPING CAMPDEN, Gloucester. Parish Church. Chancel— 1380-1401. Rick- 
 man, 290. 
 
 CHRIST CHURCH, Hampshire. Secular Canons; replaced by Augustinian Canons 
 in 1 150. Probably commenced by Flambard after he became Bishop of Durham 
 in 1099. (Sir G. G. Scott.) Lady Chapel — Completed before 1395 or 1406. 
 West Tower and Choir — XV. Cent. ; choir vaulted between 1502 and 1520. 
 Ferrey's Memorials of Christ Church, Twynham. 
 
 CIRENCESTER, Gloucester. Collegiate Church. Nave — Rebuilt c. 15 15. 
 
 CIRENCESTER ABBEY, Gloucester. Austin Canons. Foundations— Temp. Henry II. 
 
 CLEE, Lincolnshire. Parish Church. Chancel and Transepts — Consecrated in 11 92, 
 according to inscription cut on an earlier pier: facsimile in Rickman, 158. 
 
 COLCHESTER PRIORY, St Botolph, Essex. Austin Canons. Founded 1102. 
 Rickman, 89. 
 
 COLCHESTER ABBEY, Essex. Benedictine Monks. Consecration 1104. 
 
 COLLUMPTON, Devon. Parish Church. Lane's Chantry and Fan Vault — 15 10- 
 1528. Rickman, 316. 
 
 COLMOUTH, Bedford. Parish Church. Finished 1396. Rickman, 294. 
 
 COVENTRY, St Michael, Warwick. Parish Church. Tower — Finished 1394. 
 Spire — Begun 1432. Nave — Enlarged c. 1420. Bloxam, 268. 
 
 COVENTRY, Holy Trinity. Parish Church. Chancel — Rebuilt in 1391. East 
 end rebuilt in 1786. 
 
 CROWLAND or CROYLAND ABBEY Lincolnshire. Benedictine Monks. Choir 
 — Commenced in n 13. Facade — Early and late XIII. Cent., and XV. Cent. 
 Nave — Recast in XV. Cent. Tower — 1427. Builder, 1st September 1894. 
 Rickman, 307. 
 
 CROXDEN ABBEY, Stafford. Cistercian Monks. South Transept— First half of 
 XIII. Cent. 
 
 CRUMWELL, Notts. Parish Church. Tower — -1442. Rickman, 303. 
 
 CWM HIR ABBEY, Radnor. Cistercian Monks. Founded in 1143 by Robert Fitz 
 Stephen, who commanded in the first invasion of Ireland, 1170. Giraldus 
 Cambrensis. See " Llanidloes," and page 433. 
 
 DARLINGTON, Durham. Parish Church. Was building in 1192, and unfinished 
 
 in 1 195. Arch. Aeliana, xvii. 145. 
 DARTON, York. Parish Church. Chancel — Inscription, 15 17, on wall-plate. 
 
 Rickman, 316. 
 DAVINGTON PRIORY, Kent. Benedictine Nunnery. Founded n 53. Arch. 
 
 Cant., xxii. 275. 
 DEEPING ST JAMES PRIORY, Lincolnshire. Benedictine Monks, c. 11S0. 
 
 Sharpe's Nene Valley. 
 DEERHURST PRIORY, Gloucester. Benedictine Monks. Nave— Probably last 
 
 years of XII. Cent. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 643 
 
 DEVIZES ST MARY and DEVIZES ST JOHN, Wilts. Parish Churches. 
 Built by Roger of Sarum. May be as late as 1 139. Mr Harold Brakspear. 
 
 DONNINGTON, Lincolnshire. Parish Church. Nave — 1351. Hickman, 242. 
 
 DORCHESTER ABBEY, Oxfordshire. Austin Canons. Parts of all periods, 
 especially of the last half of the XIII. Cent. Dorchester Abbey, published by 
 Parker. 
 
 DORE ABBEV, Hereford. Cistercian Monks. Founded 1147. Choir— Probably 
 the last years of the XII. Cent. East Portion of Choir — c. 1200. 
 
 DOVER, St Martin. Austin Canons. Choir— Begun 1 1 3 1 ; finished 1140. 
 {Hickman, 96.) Refectory, 1131-1139. Plumptre in Arch. Cantiana, iv. 
 
 DUBLIN, Christ Church. Cathedral. Arrosian Canons. Choir and Transepts — 
 Probably begun soon after 1171, when Henry II. landed. Strongbow was buried 
 in it in 11 76, and a stone altar was given to it in 1180. Monograph by G. E. 
 Street. Nave — Completed in 1235. Sir T. Drew in Builder, 7th April 1S94. 
 
 DUBLIN, St Patrick. Cathedral. Patent Roll of 1225 speaks of preachers going 
 through Ireland to beg alms for the fabric. Bernard's St Patrick's Cathedral, 9. 
 
 DUCKLINGTON, Oxfordshire. Parish Church. North Aisle— c. 1335. Pick- 
 man, 238. 
 
 DUNFERMLINE ABBEY, Fife. Benedictine Monks. Probably built soon after the 
 accession of David I. in n 24. Macgibbon and Ross's Ecclesiastical Archi- 
 tecture of Scotland. 
 
 DUNMOYV PRIORY, Essex. Austin Canons. Choir— XIV. Cent. 
 
 DUNSTABLE PRIORY, Bedford. Austin Canons. Choir— Begun in 1 122. Pick- 
 man, 96. A''<7('«— Probably third quarter of XII. Cent. West Front — Probably 
 after the fall of the central tower in 1222. 
 
 DUNSTER, Somerset. Priory Church. Documentary evidence gives 1419 as date of 
 tower. Murray's Somerset, 475. 
 
 DURHAM CATHEDRAL. Benedictine Monks. Choir— Begun 1093; works had 
 reached the nave in 1099 ; body of St Cuthbert translated to the choir in 1104. 
 Nave — Completed by 1128; except its Vault, 1128-1133. Chapter House— 
 1133-1140. Galilee — c. 1175; or perhaps ten years earlier. Upper Part of West 
 Towers — Perhaps 1217-1226. East Transept — Begun 1242; finished 12S0. 
 Reredos— 1380. Central Tower— Probably begun c. 1470. Priors Kitchen— 
 Begun in 136S. Pick/nan, 24S. Mr J. Bilson's Beginnings; Canon Greenwell's 
 Durham Cathedral. 
 
 EDINGTON, Wilts. Priory Church. Dedicated in 136 1. Rickman, 243. 
 
 ELY CATHEDRAL, Cambridgeshire. Benedictine Monks. Begun 1083. Became 
 a cathedral in 1109. Choir— Destroyed. Ground Story of Transept— Begun 
 c. 1090. Upper Stories of Transept and Nave — c. 11 00-1130. Upper Part of 
 West Transept, Facade, and Infirmary -c. 1 170-1200. (E. Sharpe.) St Mary's 
 Parish Church— Ditto. Galilee— Built by Eustace, who was Bishop 1198-1215. 
 (Sir G. G. Scott.) Presbytery -1235-1252. Lady Chapel— Commenced in 1321 ; 
 unfinished in 1349. (Monasticon Anglicanum, i. 464.) Central Tower— Fell 
 in 1322. Octagon and Choir— Begun 1322. Prior Crauderis Chapel— 1325-6. 
 Choir Stalls— Begun 133S. Great Gatehouse— Begun 1396-7. Octagon of West 
 Tower— 1 40 1 or 1444. Bishop A /cock's Chapel— 1488. Bishop West's Chapel 
 — 1534. North-west Doorway of Transept— 1699. D. J. Stewart's Arch. Hist, oj 
 Ely Cathedral, 1868. 
 
644 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 ETCHINGHAM, Sussex. Parish Church. Brass of the founder, who died in 1387. 
 Rickman, 292. 
 
 ETON COLLEGE, Bucks. Chapel. Begun 1441 ; consecration 1443 ; unfinished 
 in 1463; finished 1522. Rickman, 303. 
 
 EVESHAM ABBEY, Worcester. Benedictine Monks. Tower — 1533. Walcot. 
 
 EWENNY PRIORY, Glamorgan. Benedictine Monks. Founded 1147; church 
 mainly of this date, but crenulated in the XIV. Cent. Mr Harold Brakspear. 
 
 EXETER CATHEDRAL, Devon. Secular Canons. Towers— 1112-1136. {Rickman, 
 90.) The lower part of the Lady Chapel with its companion chapels and the 
 retrochoir are certainly one work, c. 1270; the remodelling of the transepts is 
 probably before 1280; the above is the work of Branscombe, who was Bishop 
 1257-1280. Quivil, who was Bishop 1280-1291, began the presbytery, which was 
 ready for its roof in 1299 and was partly glazed in 1301 ; about the same time the 
 Lady Chapel was completed. Bitton, who was Bishop 1292-1307, nearly finished 
 the choir, and inserted a triforium arcade in the presbytery. Stapledon, who was 
 Bishop 1308-1326, probably built the crossing and the eastern bay of the nave; 
 the latter was not glazed till i3i7or 131 8. To Stapledon also belong the Bishop's 
 throne, the sedilia, and the choir screen. Grandisson, who was Bishop 1327-1369, 
 finished the nave c. 1350. The east window is Brantingham's (1370-1394). The 
 lower part of the screen of the west front is c. 1 345-1369; the upper figures are 
 1377-1399. The screens to the three eastern chapels have the arms of Stafford, 
 who was Bishop 1395-1419. The three chantries are later. Professor Lethaby 
 in Architectural Review, March and May 1903. 
 
 FAIRFORD, Gloucester. Parish Church, c. 1490. Leland in Carbonell's Fair- 
 ford, 2. 
 
 FINC HALE PRIORY, Durham. Benedictine Monks. Founded 1 196. New church 
 begun 1242, and probably finished c. 1265. Rickman, 171. 
 
 FORDINGTON, Dorset. Parish Church. Endowed in 109 1. Sculptured lintel illus- 
 trated in Bloxam, 86. 
 
 FOTHERINGHAY, Northants. Collegiate. Chancel — 1415. Nave — 1435 sea. 
 Rickman, 301. 
 
 FOUNTAINS ABBEY, Yorkshire. Cistercian Monks. Founded 1132. Nave— 
 Begun c. 11 35. Chapter House — c. 11 50-1 160. Presbytery — Begun c. 12 10; 
 finished c. 1220. Eastern Transept — Built by John of Kent, abbot 1220-1247. 
 Great North Tower — Built by Marmaduke Huby, abbot 1494-1526. Infirmary 
 and Cloister — Built for the most part by Abbot John of Kent. Mill — Before 1 147 ; 
 with later additions. Eastern Guest-house — c. 1150. Northern Half of Western 
 Range, and Western Gatehouse — c. 1160-1170. Warming-house, Frater, Kitchen, 
 Southern Half of Western Range, and Infirmary of lay-brothers — Late XII. and 
 early XIII. Cent. W. H. St John Hope in Yorkshire Arch. Journal, xv., 1900. 
 
 FURNESS, Lancashire. Cistercian Monks. Nave and Transepts — Probably begun 
 soon after the monks became Cistercian in 1148. Crossing-piers — Before 1 148. 
 Presbytery — Early XV. Cent. Western Tower — Late XV. Cent. Chapter House, 
 Nortlurn Pa?-t of Dorter Range, and Abbot's House — Early XIII. Cent. Chapel 
 without the Gates — Late XIII. Cent. Infirmary Hall, Chapel, &c. — Early XIV. 
 Cent. Mr W. Ff. St John Hope in Transactions of Cumberland and Westmorland 
 Archceological Society, xvi. ; and Mr Harold Brakspear in Transactions of lanca- 
 shire and Cheshire Arch. Soc, xviii. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 645 
 
 GLASTONBURY ABBEY, Somerset. Benedictine Monks. Lady Chapel- Begun 
 after fire of n 84 : dedicated 11 86. (Willis, 13.) Choir— Commenced 1184. 
 (Willis, 23.) Choir — Lengthened two bays by Walter Monington, abbot 1341- 
 1374. (Willis, 31.) 
 
 GLASTONBURY ST JOHN, Somerset. Parish Church. 1485. 
 
 GLOUCESTER ABBEY. Benedictine Monks. Became Cathedral in 1541. 
 Choir — Foundation stone laid in 1089; dedication in 1100. Nave—c. 11 20. 
 Vault of Nave — 1245. South Aisle of Nave — 1318-1329. Shrine of Edward 
 II- — 1329-1334. Work in South Transept— \ 331-1337. Choir and North 
 Transept — Remodelled 1337-1377. Choir Vault and Northern Stalls — Were 
 finished 1337-1357. The east window was probably glazed 1347-1350; accord- 
 ing to Winston in Archozological Journal, xx. North Transept — Remodelled 
 1368-1374. East Walk of Cloister up to Chapter House Doorway — 1351-1377. 
 Cloisters — Completed 1381-1412. West Front, two West Bays of Nave, and 
 probably South Porch — 1421-1437. Central Tower — Begun 1450-1457 ; finished 
 by the monk Tully, who became Bishop of St David's in 1460. Lady Chapel — 
 Begun 1 45 7-1472 ; finished 1472-1498. Mr W. H. St John Hope's Notes on 
 the Benedictine Abbey of St Peter at Gloucester, and Professor Willis, Arch. 
 Institute, i860. 
 
 GREAT BOORHAM, Surrey. Parish Church. 1341. Hickman, 240. 
 
 GREAT PONTON, Lincolnshire. Parish Church, c. 1519. Rickman, 317. 
 
 GUISBOROUGH PRIORY, Yorkshire. Austin Canons. Choir— After the fire of 
 1289; probably was building in 1309. 
 
 HARLESTON, Northants. Parish Church. Chancel— 1320. Nave— 1325. Churches 
 of Northants, 266 sea. 
 
 HARTLEPOOL, Durham. Parish Church, c. 1190. Arch. Aeliana, xvii. 201. 
 
 HAVERFORDWEST PRIORY, Pembroke. Augustinian Canons. Probably c. 1220. 
 Rickman, 162. 
 
 HAWTON, Notts. Parish Church. Chancel — Second quarter of XIV. Cent. Tower 
 — c. 14S3. Rickman, 310. 
 
 HAYLES, Gloucester. Cistercian Monks. Founded 1246. Church, with thirteen 
 altars, hallowed 125 1. New Work, in consequence of gift of the Holy Blood, 
 begun 1270 and finished 1277. Mr Harold Brakspear in Bristol and Gloucester 
 Transactions, xxiv. 126-135. 
 
 HEDON, Yorkshire. Parish Church. South Transept — End of XII. Cent. Chancel 
 and North Transept — First half of XIII. Cent. Nave — Probably commenced in 
 the last quarter of the XIII. Cent. ; its western bay probably not earlier than 1325. 
 Central Tower — 1427-1437. Hedon, by Mr J. R. Boyle; and G. E. Street in 
 Archccologia, xlviii. 
 
 HEREFORD CATHEDRAL Secular Canons. Choir— Begun 1079-1095 ; dedica- 
 tion 1 1 10. (William of Malmesbury .) East Wall of South Transept — Ditto. 
 Nave — Completed 1131-1148. {William of Wycumbe.) East Transept — 1186- 
 1199. (Sir G. G. Scott.) Lady Chapel — c. 1220. Choir Clerestory — c. 1250. 
 North Transept — c. 1260. Inner North Porch — c. 12 88- 1300. Remodelling of 
 Aisles — c. 1282-1350. Central Tower — 1325-1352 (Hills); 1 3 16-1327 (Scott). 
 Vaulting of South Transept — c. 1400. Vicars Cloister — c. 1490. Bishop Audley's 
 Chantry— c. 1500. Outer North Porch — c. 15 20-1 530. Cloisters — c. 1520. Fall 
 of Western Twer, Shorten lug and Remodelling of Nave by Wyatt — 1 786-1 796. 
 Gordon Hills in fouruat of British Archaeological Institute, 1871. 
 
646 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 HEXHAM ABBEY, Northumberland. Austin Canons. Choir— Probably c. 11S0— 
 
 e. 1210. Transept — Probably c. 1215 — c. 1230. Monograph by Mr C. C. 
 
 Hodges, with measured drawings. 
 HIGHAM FERRERS, Northants. The College. 1422 sea. Rickman, 298. 
 HILLESDON, Bucks. Parish Church. 1493. Rickman, 313. 
 HINGHAM, Norfolk. Parish Church. Built by Remigius, who was rector 1316- 
 
 1359. 
 HINTON CHARTERHOUSE, Somerset. Carthusian Monks. Founded 1232. 
 
 Chapter House, with Treasury over, and Fragment of Church. Rickman, 168. 
 HITCHIN, Hertford. Parish Church. Pier Arcade — Finished before 1302. 
 
 Rickman, 235. 
 HOWDEN, Yorkshire. Became Collegiate in 1265. Transepts, Nave, and West Front 
 
 — c. 1265 — c. 1310. Choir — c. 1310 — c. 1330. Chapter House — 1 380-1407. 
 
 Central Tower — Middle stage, first quarter of XV. Cent. ; upper stage, end of 
 
 XV. Cent. Mr John Bilson. 
 HULL, Holy Trinity, Yorkshire. Parish Church. Chapel to Hessle Church till 
 
 1661. Said to have been begun in 1285. Transept, with crossing, built first. 
 
 William Scale was buried "in the New Chapel" in 1327. Choir — Probably 
 
 unfinished in 1350. Work was being done in the Nave c. 1389 and in 1395. 
 
 Tower — Upper stages probably 1520-1529. See Mr J. R. Boyle's monograph 
 
 (Brown, Hull, 1890). 
 HURLEY, Berks. Benedictine Priory. 1082-10S9. {Rickman, 86.) Present church 
 
 not earlier than c. n 20. Mr IV. H. St John Hope. 
 HYTHE, Kent. Parish Church. Early XIII. Cent. Arch. Cant., xviii. 403. 
 
 IFFLEY, Oxford. Parish Church. Given to Kenilworth Priory c. 1160; probably 
 
 built soon after. Rickman, 92. 
 ILMINSTER, Somerset. Parish Church, c. 1490. 
 INGOLDMELLS, Lincolnshire. Parish Church. Aisles, South Porch, To7ver, and 
 
 Font — 1346. Rickman, 241. 
 IPSWICH, St Lawrence, Suffolk. Parish Church. 1420-1431. Rickman, 297. 
 IRON ACTON, Gloucester. Parish Church. Tower — c. 1430. Rickman, 299. 
 
 JEDBURGH ABBEY, Roxburgh. Austin Canons, c. 1175 — c. 119c Mr C. C. 
 
 Hodges' Hexham. 
 JERVAULX ABBEY, Yorkshire. Cistercian Monks. May be c. 1170 — c. 1190. 
 
 KENILWORTH PRIORY, Warwick. Austin Canons. Founded in 1122. Rick- 
 man, 92. 
 
 KETTON, Rutland. Parish Church. Indulgence in 1232 to all who would con- 
 tribute to the building or reparation. Rickman, 168. 
 
 KIRKHAM PRIORY, Yorkshire. Austin Canons. Founded 1 121. Outer Parlour 
 — c. 1 1 90. Presbytery — c. 1240. Gateway — Finished 1296. Mr C. C. Hodges' 
 Hexham; and Rickman, 92. 
 
 K1RKSTALL ABBEY, Yorkshire. Cistercian Monks. Removed to present site 
 in 1 152 ; the church and all the claustral buildings of the Abbey are recorded 
 to have been finished by the first Abbot, Alexander. 
 
 LACOCK ABBEY, Wilts. Austin Canonesses. Was founded 1232 by Ela, 
 Countess of Salisbury, and the buildings are of the date of the foundation. There 
 
CHRONOLOGY OF ENGLISH CHURCHES. 647 
 
 was a large Lady Chapel being erected 1320, of which a building agreement 
 exists, but the chapel is destroyed. Mr Harold Brakspear. 
 
 LAMBETH, Surrey. Chapel of Archbishop's Pala.ce. First half of XIII. Cent. 
 
 LANERCOST PRIORV, Cumberland. Austin Canons. Consecration in 1169; 
 probably only of the ground story of the choir. The works progressed slowly, 
 upwards and westwards, till the nave and west front were complete, c. 1250. Mr 
 E. R. Tate in Builder, 1st October 1898. 
 
 LASTINGHAM, Yorkshire. Benedictine Monks. Crypt — c. 10S0. The present 
 Parish Church is but the Choir and Crossing of the Monastic Church contem- 
 plated. Rickman, 82. 
 
 LAVENHAM, Suffolk. Parish Church. Vestry— 1470-1486. Porch— c. 1529. 
 {Rickman, 308.) Nave — Early XVI. Cent. Suffolk Archozological Institute, vi. 114, 
 and ix. 370. 
 
 LEEDS, St John, Yorkshire. Parish Church. Consecrated 1634. 
 
 LEOMINSTER PRIORV, Hereford. Benedictine Monks. Granted to Reading 
 Abbey in 1121. Nave — A consecration in 11 30. Lower Part of Tower — Late 
 XII. Cent. Inner Aisle ; and South Doorway, not in situ — Before the conse- 
 cration of 1239. Outer South Aisle — First half of XIV. Cent. See Archaeological 
 Journal, x. 
 
 LEWES PRIORV, Sussex. Cluniac Monks. Founded 1077. Enlarged 1091-1098, 
 and again in 1 142-1 147. See paper by Mr W. H. St John Hope in Archaeological 
 Journal, vol. xl. 
 
 LICHFIELD CATHEDRAL, Stafford. Secular Canons. Choir, Ground Story— c. 
 1 195. {Prior, iigand 157.) South Transept — c. 1220. {Willis.) North Transept, 
 Chapter House, and Vestibule— c. 1240. (Willis.) Xave— Second half of XIII. 
 Cent. West Front ami Towers — Probably c. 1280-1330. Lady Chapel — Begun 
 1310: unfinished in 1 33 1. {Rickman, 236.) Presbytery and Upper Choir — Second 
 quarter of XIV. Cent. South-west Spire — Ditto. North-west Spire and Upper 
 Part of Tower is a XV. < lent, copy of the south-west steeple. (J. T. Irvine.) Choir 
 Clerestory Repaired and Central Spire Rebuilt — After 1661. See Willis in 
 Archaeological Journal, 1 86 1 . 
 
 LILLESHALL ABBEY, Shropshire. Austin Canons. The Abbey was founded in 
 1 145. Presbytery, Crossing, and South Transept of this date remain, together 
 with part of Eastern Range. Mr Harold Brakspear. 
 
 LINCOLN CATHEDRAL. Secular Canons. West Front— Remigius' work, 
 commenced between 1072 and 1075; consecrated 1092. Lotver Portions of West 
 Towers a?id North and South Gables— c. 11 40. West Doorways — c. 1150. Choir 
 and Eastern Transepts — 1192 — c 1200. Great Transept — Probably followed 
 choir. Galilee Porch — c. 1230. Chapter House— Is mentioned 1220-1 235. Nave 
 
 -Timber for roof was bequeathed in 1233. Central Tower — Fel 
 
 in 
 
 ■5i- 
 
 1239, and was rebuilt. Presbytery— 1256— c. 1320: but St Hugh's body was 
 translated to it in 1280. Cloister, South Side— 1296. Upper Stage of Central 
 Tower — 1307. South Transept End— Probably after the death of Bishop 
 Dalderby in 1319. Upper Stories of Western Towers— before 13S0. Chantry 
 Chapel of Bishop Fleming, died 1 430-1 431. Chantry Chapel of Bishop Russell, 
 died 1494. Chantry Chapel of Bishop Longland, died 1547. See Sharpe's Lincoln 
 Excursion: and Venables in Archaeological Journal, xl. 
 
 LINCOLN, St Mary-le-Wigford. Parish Church. Was building in 122S. Rick- 
 man, 168. 
 
 LINDISFARNE PRIORY, Northumberland. Benedictine Monks. Some part was 
 
648 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 finished before c. 1128, but the Church was probably not completed till c. 1150. 
 Mr C. C. Hodges in Builder, 1st June 1895 ; and Mr John Bilson's Beginnings, 
 
 3IS. 3i6. 
 LITTLE KIMBLE, Bucks. Parish Church. Consecrated 131 7. Rickman, 237. 
 LITTLE MALVERN PRIORY, Worcester. Parish Church. Built by Bishop 
 
 Alcocke between 1476 and 1482. Bloxam, 273. 
 LLANDAFF CATHEDRAL, Glamorgan. Secular Canons. Remains of Norman 
 
 Choir — 1120-1133. Nave Arcades — c. 1190. {Prior, 157.) West Front — Early 
 
 XIII. Cent. Chapter House — 1244-1265. Lady Chapel — c. 1280. South-west 
 Tower — End of XV. Cent. See monograph by Mr J. H. James. 
 
 LLANIDLOES, Montgomery. Parish Church. Piers and arches probably brought 
 
 from Cwm Hir in 1542 ; the Abbey was dissolved in 1536. These piers and 
 
 arches appear to belong to the last years of the XII. Cent. 
 LLANTHONY PRIORY, Monmouth. Austin Canons. Last quarter of XII. Cent. 
 
 See Freeman in Archceologia Cambrensis, Series I., vol. i. ; and Prior, 157. 
 LONDON, Austin Friars. Augustinian Friars ; now Dutch Church. Rebuilding 
 
 1354; again c. 1475. Mr A. S. Walker in Builder, 4th April 1896. 
 LONDON, St Andrew Undershaft. 1520-1532. Stow. 
 LONDON, St Bartholomew's Priory, Smithfield. Austin Canons. 1123-1133. 
 
 Clerestory — Early XV. Cent. Architectural Review, i. 1. 
 LONDON, St Etheldreda, Ely Place, Holborn. Chapel of the Palace of the Bishops 
 
 of Ely. c. 1290 — c. 1300. Sir G. G. Scott in Lectures, i. 184. 
 LONDON, St Helen, Bishopsgate. The North Aisle was the Chapel of the Benedictine 
 
 Nunnery ; the South Aisle was Parochial. The screen between them was removed 
 
 c. 1538. Side Chapels— Probably XIV. Cent. Tall Arches of Nave— Probably 
 
 1475- 
 LONDON, Savoy Chapel. 1505 sea. Rich/nan, 313. 
 LONDON, Church of Knights Templars. Nave — Consecrated 1185. Choir — 
 
 Consecrated 1240. Rickman, 111 and 171. 
 LONDON, Tower of London. St John's Chapel. Last quarter of XL Cent. 
 LONG MELFORD, Suffolk. Parish Church. Chancel— Finished 1479. The seven 
 
 western bays of the Nave are inscribed 1481. South Aisle — c. 1484. Lady 
 
 Chapel — Finished c. 1496. Monograph by Lauriston Conder. 
 LOUTH, Lincolnshire. Parish Church. Chancel — c. 1400-1445. To'wer — c. 1445- 
 
 1500. Spire — 1501-1515. James Fowler in Transactions of Lincoln Archaeological 
 
 Society, 1873 ; and J. J. Creswell in do., 1897. 
 LOUTH PARK ABBEY, Lincolnshire. Cistercian Monks. Founded in 1139; the 
 
 foundations have been excavated. The work varies from n 39 to the XIII. 
 
 Cent, west front. Assoc. Arch. Soc. Reports, vol. xii. 
 LOW HAM, Somerset. Parish Church. Built by Sir Richard Hext, who died in 
 
 1624. Murray's Somerset, 444. 
 LUDLOW, Shropshire. Enlarged when it became Collegiate in the latter part of the 
 
 XIV. Cent. Tower and much rebuilding in late XV. Cent. Mr Thomas 
 Wright's Guide to Ludlow. 
 
 LUTON, Bedford. Parish Church. Chancel — 1430-1440. North Chapel — c. 1460. 
 Rickman, 299. 
 
 LYNN, Chapel of E_ed Mount, Norfolk. 1484-1485. Coiporation Records. 
 
 LYNN, St Margaret, Norfolk. Earliest Work — 1091-1119. North-west Tower — 
 Ordered to be built 1453. South Aisle of A T ave — Before 1483. North Aisle of 
 Chancel — Ordered to be rebuilt 1472. Clerestories — Rebuilt and new roofs, 1491. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 649 
 
 South Aisle of Chancel — Nearly finished in 1494. Spire on south-west tower 
 was blown down in 1741. Wooden central octagon, 132 feet high, was taken 
 down, and Nave rebuilt by 1745. Fen and Marshland Churches, ii. 
 LYNN, St Nicholas, Norfolk. Chapel to St Margaret. Was being pulled down in 
 1413. Built entirely by voluntary contributions before 1419. See Lynn St 
 Nicholas, by Mr E. M. Beloe, p. 151. 
 
 MAIDSTONE, All Saints', Kent. The Church was made Collegiate by Archbishop 
 Courtenay, 13S1-1396. The rebuilding was largely complete in 1395. Whichcord 
 in Weale's Quarterly Papers, iv. 
 
 MALLING, Kent. St Leonard's Tower. Built by Gundulf, Bishop of Rochester from 
 1077 to 1 roS. 
 
 MALLING ABBEY, Kent. Benedictine Nuns. Lower Part of West Front— -By 
 Bishop Gundulf, c. 1 077-1 108. Upper Part —c. 1150. Remains of Cloister — 
 c. 1 1 go. 
 
 MALMESBURY ABBEY, Wiltshire. Benedictine Monks. Nave— The Church was 
 not commenced by Roger, Bishop of Salisbury, who died 1139, but after the 
 monastery regained its independence in 1140 ; for William of Malmesbury, whose 
 History ends in 1142, does not mention building as going on in the Church. 
 (Bilson's Beginnings, 309.) Clerestory and Vault — Second quarter of XIV. Cent. 
 
 MALTON PRIORY, OLD, Yorkshire. Gilbertine Canons. Last quarter of XII. 
 Cent. 
 
 MALVERN PRIORY, Worcester. Benedictine Monks. Said to have been begun 
 c. 10S4. Rebuilding begun c. 1450; service resumed 1460. Completed between 
 1476 and i486. Tiles have dates 1453 and 1456. Builder, 2nd January 1897. 
 
 MANCHESTER CATHEDRAL, Lancashire. Made Collegiate in 1422 ; Cathedral in 
 1847. Choir and Chapter House — 1422-145S. Nave — Said to have been com- 
 pleted 1465-1481. Chapel of Holy Trinity — 1498. Jesus Chapel (Vestry and 
 Library) — 1506. St fames' Chantry or Ducie Chapel; also the Choir Stalls: and 
 St George's Chapel — 150S. Derby and Ely Chapels— c. 15 15. Lady Chapel — 
 Remodelled 151S-1535. Perkins' Manchester Cathedral, 6. 
 
 MELBOURNE, Derbyshire. Parish Church. First half of XII. Cent. 
 
 MEOPHAM, Kent. Parish Church. Built 1315 sea. Repaired 1381-1396. Rick- 
 man, 237 and 290. 
 
 MILTON ABBEY, Dorset. Benedictine Monks. Choir and South Transept— Com- 
 menced after fire of 1309. Vaults of Tower and Transept — 1481. Roland Paul 
 in Builder, 5th January 1901. 
 
 MINSTER, THANET, Kent. Benedictine Nuns. Early Norman vaulting remains ; 
 also Tudor work. 
 
 MINSTER, THANET, Kent. Nuns' Church. Also Parochial. Nave— Probably 
 third quarter of XII. Cent. Vaulted Chancel and Transept— First half of XIII. 
 Cent. 
 
 MONKTON, Pembroke. Benedictine Monks. Probably first half of XII. Cent. 
 
 NANTWICH, Cheshire. Collegiate. Chancel— 1327-1333. 
 
 NETLEY ABBEY, Hants. Cistercian Monks. Begun in 1239. Richnan, 170. 
 
 NEW ROMNEY, Kent. Parish Church, c. 1190 and first half of XIV. Cent. 
 
 Arch. Cant., xiii. 466. 
 NEW SHOREHAM, Sussex. ? Collegiate. Nave, Four Arches of Crossing, Lower 
 
 Part of To?i'er—c. 11 30. Choir, Pier-arcades, with North and South Aisles and 
 
6so CHRONOLOGY OF ENGLISH CHURCHES. 
 
 their Vaults, and Upper Part of Tower — c. 1175. Upper Parts of Choir — c. 1190, 
 1210. E. Sharpe's monograph; and Arch. Aeliana, xvii. 217. 
 
 NORTHAMPTON, St Peter. Parish Church, c. 1140-1160. Proceedings of Soc. 
 of Ant., 1902, 74. 
 
 NORTHAMPTON, St Sepulchre. Parish Church, c. iioo— c. iiio. Cox and 
 Serjeantson's Monograph. 
 
 NORTHLEACH, Gloucester. Parish Church. Nave— 1458. Poof— Soon after. 
 Rickman, 306. 
 
 NORTH WALSHAM, Norfolk. Parish Church. Probably rebuilt after destruction 
 in the rising of 1381. The heraldic bearings in the porch prove that this was 
 built before 1405. 
 
 NORWICH CATHEDRAL, Norfolk. Benedictine Monks. Choir, Transepts, East 
 Nave as far west as the Altar of the Holy Cross — Built by Herbert, who was 
 Bishop 1096-1119. Foundation Stone laid — 1096. West Nave and Tower — 
 Built by Eborard, who was Bishop 1121-1145. St Ethelbert's Gateway — After 
 riots of 1272. Chapel of Charnel House — 131 0-1325. Presbytery Clei-estory 
 — Probably after fall of spire, c. 1362. Erpingham Gate — 1416-1425. Vault 
 of Nave — After 1463-1472. Vault of Presbytery and alteration of pier-arches 
 — 1472-1499. Vault of Transepts, and Chantry Chapel of Bishop Nykke — 
 1501-1536. Chapter House, begun 1289; Cloister, begun opposite chapter house, 
 1297; east and south alley, 1299-1325 ; west alley, 1416-1426; north alley 
 finished, 1430. W. H. St John Hope. 
 
 OARHAM, Rutland. Hall of Ring's Manor. 1165-1191. Rickman, 104. 
 
 OTTERY ST MARY, Devon. Collegiate. Bishop Grandisson's rebuilding began 
 
 <"■ *337- 
 
 OXFORD, All Souls' College. South Quadrangle — Foundation 1438. Chapel — 
 
 Consecrated 1442. 
 OXFORD, Balliol College. Library — 1 431-1460. 
 OXFORD, Bodleian Library. Older Portion — 1613. 
 OXFORD, Brasenose College. Chapel — Consecrated 1666. 
 OXFORD CATHEDRAL. 1154-1180. Confirmation of charters to the Austin 
 
 Canons in 1158. A consecration in 1180. Chapter House — c. 1220. Lady 
 
 Chapel — c. 1220. Latin Chapel — Built before 1355. Choir Vault — c. 147S- 
 
 1503. {Rickman, 103, 244.) Became Cathedral 1546. 
 OXFORD, Christ Church. College. Hall, Kitchen, and Part of Great Quadrangle — 
 
 1528-1530. Staircase to Hall, Fan Vault — 1640. Great Quadrangle — Finished 
 
 1665. 
 OXFORD, Corpus Christi College. Chief Buildings and Chapel — 15 16-1520. 
 OXFORD, Divinity School. 1445-1480. 
 
 OXFORD, Exeter College. Quadrangle, Hall, &*c. — 1610-1618. 
 OXFORD, Jesus College. Founded 157 1. 
 OXFORD, Kettel Hall, Broad Street. 161 5. 
 OXFORD, Lincoln College. Hall, Library and North Quadrangle — 1436-1438. 
 
 South Quadrangle — 161 2. Chapel — Consecrated 1631. 
 OXFORD, Magdalen College. 1474- 1480. Chapel — Consecrated 14S0. South 
 
 Cloister — 1490. Tower — Completed 1505. 
 OXFORD, Merton College. Chapel — Probably commenced about 1277, soon after 
 
 the founder's death. Vestry of Chapel — 1310. Library — c. 1380. Transepts and 
 
CHRONOLOGY OF ENGLISH CHURCHES. 651 
 
 crossing completed, except the top of the Tower, in 1424. Tower -Finished 
 
 1450. Great Southern Quadrangle — 1610. 
 OXFORD, New College. Foundation Stone laid 1380. New Buildings finished 
 
 13S6. Cloisters and Bell Tower completed 1400. 
 OXFORD, Oriel College. South and West Sides of Quadrangle- -1620-1637. 
 OXFORD, St Aldate. Parish Church. South Aisle — 1335. Rickinan, 23S. 
 OXFORD, St John's College. Gateway, Tower and Part of West Front — 1437. 
 
 Library and First Quadrangle — 1597. Inner Quadrangle and Library -1631- 
 
 l6 35- 
 OXFORD, St Mary's Church. Tower — 1 280-1 290. Spire — First half of XIV. Cent. 
 
 Chancel — Begun 1462. Nave— 1490-1503. Porch — 16 37 . Mr T. G. /aekson's 
 
 monograph. 
 OXFORD, St Mary Magdalen. Parish Church. South Aisle — 1318-1337. Tower 
 
 — Completed 1505. 
 OXFORD, St Mary's Hall. Chapel, &c— 1640. 
 OXFORD, St Peter ix the East. Parish Church. Crypt and Chancel and Doorway 
 
 — c. 1 1 70. 
 OXFORD, Schools' Quadrangle. 16 10-16 18. 
 OXFORD, University College. West Quadrangle — Begun 1634. Hall — Completed 
 
 c. 1657. 
 OXFORD, Wadham College. 1610-1613. 
 
 The above chronology of Oxford buildings is taken from Parker's Visitors' Guide 
 to Oxford. 
 
 PATRINGTON, Yorkshire. Parish Church. Transepts— Probably first quarter 
 of XIV. Cent. Choir. Nave, and Tower — Second quarter of XIY. Cent. Spire 
 — Probably early XV. Cent. See Mr J. T. Micklethwaite's paper in Yorkshire 
 Arch. Journal, ix. 99. 
 
 PERSHORE ABBEY, Worcester. Bexedictixe Moxks. South Transept— Probably 
 late XI. Cent. Choir — The earlier work is 1223-1239. Upper Choir and Central 
 Tower — After the fire of 12SS. Rickinan, 165 ; and Arch. Soe. Reports, iv. 355. 
 
 PETERBOROUGH ABBEY. Benedictixe Monks. Became Cathedral in 1541. 
 Foundations laid in 1 1 1 7 or 111S. The monks entered the "nova ecclesia '' in 
 1 140 or 1 143. The work done probably included the Pier-arcade and Triforium 
 of the Choir ; of part of the Transepts : of the two Eastern Bays of the Nave : and 
 most of the South Aisle Wall. The Choir and Transepts were probably finished 
 ll SS' 1I lS '• a ' so three stages of the Central Tower and the six Central Bays of 
 the Nave except the Clerestory. The Clerestory of the Nave ; the West Bays of 
 the Nave ; and the Western Transept — Probably n 77-1 193. West Front — Pro- 
 bably built by Acharius, 1201-1214. {Mr J. T. Irvine.) Bell Tower — 1274- 
 1295. South-west Spire — Probably second half of XIY. Cent. Bishop's Gateway 
 — 1303. Porch in West Front — c. 1370. Eastern Chapels — Begun between 
 1438 and 147 1 ; finished between 1496 and 1528. Mr J. T. Irvine and Craddock's 
 Pete/borough Cathedral; and Journal of the R.I.B.A., 26th November 1898. 
 
 PORCHESTER, Hampshire. A Priory of Austin Canons was founded here in 1133. 
 
 PORTSMOUTH, St Thomas, Hampshire. Austin Canons. They settled here 
 between n 80 and 1200, and before 11S9 obtained the confirmation of a charter 
 for the Chapel of St Thomas the Martyr, "which they had begun to build. 
 Rickinan, 109. 
 
 POYNINGS, Sussex. Parish Church. 136S seq. Rickinan, 247. 
 
652 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 RAMSEY, Huntingdon. Parish Church, belonging to Benedictine Abbey. Last 
 half of XII. Cent. 
 
 READING ABBEY, Berkshire. Benedictine Monks. Ruins. Founded 1121. 
 
 RIEVAULX ABBEY, Yorkshire. Cistercian Monks. Fragments of Nave and 
 Transept — Built with stone brought by a canal cut c. 1145. Frater, Transepts, 
 and Choir are of Hollins stone brought by a canal, the ground for which was 
 acquired between 1193 and 1203. Mr H. A. Rye in Archceological Journal, 
 March 1900. 
 
 RIPON MINSTER, Yorkshire. Secular Canons. Became Cathedral in 1836. 
 Built 1154-1181. West Front — c. 1233. Fast Front — Probably 1288-1300. 
 Nave — Aisles begun in 1502 or 1503. Sir G. G. Scott in Archceological Journal, 
 December 1874. 
 
 ROCHE ABBEY, Yorkshire. Cistercian Monks. Founded in 1147, and soon largely 
 endowed with land. Probably third quarter of XII. Cent. Gordon Hills in 
 Archceological Journal, xxx. 421 ; and Assoc. Soc. Reports, xvii. 39, xviii. 35, 
 xix. 392. 
 
 ROCHESTER CATHEDRAL, Kent. Benedictine Monks. Parts of Nave and 
 Crypt — Gundulf's work, 1077-110S. Nave — Chiefly 1 1 1 5-1 130. Presbytery and 
 Easte?-n Transept — Begun c. 1200. Choir — Finished 1227. North Transept — 
 c. 1240-1255. South Transept — c. 1280 sea. Lady Chapel — c. 1512. Rochester 
 Cathedral, by Mr W. H. St John Hope. 
 
 ROMSEY ABBEY, Hampshire. Benedictine Nuns. The Eastern Parts are probably 
 not later than 1120. (Mr E. P. Loftus Brock in Builder, 5th October 1S95.) 
 Commenced c. n 10. (Mr Bilson's Beginnings, 307.) Western Nave and Facade 
 — c. 1220. 
 
 ROTHERHAM, Yorkshire. Collegiate. 1480-1501. 
 
 ROTHWELL, Northants. Parish Church. Last half of XII. Cent. 
 
 RYARSH, Kent. Parish Church. Tower — 1450. Rich/nan, 305. 
 
 RYE, Sussex. Parish Church, c. 1195. Prior, 156. 
 
 ST ALBANS ABBEY, Hertfordshire. Benedictine Monks. Became a Cathedral 
 in 1S77. Earliest portions built 1077-10S8. Dedicated and probably completed 
 1 1 15. West Front — Commenced 1 195-12 14. Completion of West Front, and 
 Western Bays of Nave — 1214-1235. Choir Clerestory — Begun 1257. Ante- 
 chapel — 1260-1326. Lady Chapel — 1291-1326. St Allan's Shrine — 1302-1308. 
 Rebuilding of East Bays of South Nave — After 1323. Cloister — 1323, 1335, 
 1349, 1360. Rood Screen — 1340-1350. Gatehouse — 1349-1396. Reredos — 1476- 
 14S4. Monographs by Mr James Neale and Messrs Buckler. 
 
 ST CROSS, Winchester, Hants. Hospital. Probably none of it is earlier than c. 11 60, 
 except the Sacristy. 
 
 ST DAVID'S CATHEDRAL, Pembroke. Secular Canons. Begun 11S0. Repairs 
 after fall of Central Tower in 1220. Lady Chapel — 1296-1328. Rood Screen — 
 1324-1347. Nave Roof — 1472-1509. Bishop Faughan's Chapel — c. 1509. See 
 Jones and Freeman's St David's. 
 
 SALISBURY CATHEDRAL, Wilts. Secular Canons. Begun 1220. Consecration 
 in 1225. Another consecration in 125S; finished 1266. Chapter House and 
 Cloister — 1 263-1 284. West Front — Probably late in XIII. Cent. Upper Tower 
 and Spire — Second half of XIV. Cent. Rickman, 164, 175, 239. 
 
 SARUM, OLD, Wilts. Cathedral. Consecrated 1092. Rebuilt 1115-1139. 
 Britton's Salisbiny, ii. Rickman, 90. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 653 
 
 SANDWICH, St Bartholomew, Kent. Chapel of Hospital. Founded by Thomas 
 Craythorn in 1190. North Aisle — Finished by Sir Humphrey Sandwich in 1230. 
 Rickman, 172. Archceologia Cantiana, vol. xv. 7. 
 
 SCARBOROUGH, St Mary, Yorkshire. Parish Church. Given by Richard I. to 
 the Abbey of Citeaux in 1 198. 
 
 SEDGEFIELD, Durham. Parish Church. Probably 1 242-1 2S0. Mr C. C. Hodges 
 in Archceologia Aeliana, \vi. 
 
 SELBV ABBEY, Yorkshire. Benedictine Mi inks. Founded 1069. East Nave 
 
 c. 1097-1123. Central Nave — c. 11231175. West Nave, West Front, and North 
 
 Porch — c. 1 1 75 — c. 1 190. Parts of Upper Nave and West Pro/it — c. 11 90 
 
 c. 1220. Part of Choir Aisles and Sacristy — c. 1280-1300. Choir — Completed in 
 second quarter of XIV. Cent. Mr C. C. Hodges in Yorkshire Arclneological 
 Journal, xii. 340. 
 
 SHERBORNE ABBEY, Dorset. Benedictine Monks. Begun in 1107. Choir— 
 Vaulted 1436-1459. Nave — Vaulted 1475-1504. R. H. Carpenter in Journal 
 of R.I. B. A., 19th March 1S77. 
 
 SHOBDON, Hereford. Parish Church. Ruins. 1141-1150. Rickman, 98. 
 
 SHOTTESBROOKE, Berkshire. Collegiate. Founded 1337. Monograph by Mr 
 Butterfield. 
 
 SHREWSBURY ABBEY, Shropshire. Benedictine Monks. Choir, now destroyed, 
 begun 10S3. Builder, xlviii. 740. 
 
 SKELTON, Yorkshire. Parish Church. Finished before 1247. Monograph by E. 
 Christian. 
 
 SKIRLAUGH, Yorkshire. Chapel. Shortly before 1405. Bloxam, 270. 
 
 SLEAFORD, Lincolnshire. Parish Church. Chancel — 1403. Rickman, 296. 
 
 SOUTHWARK, St Saviour or St Mary Overie. Austin Canons. Burnt in 1213. 
 Rebuilt by Peter de Rupibus, Bishop of Winchester from 1204 to 123S — East 
 Limb, probably up to crossing. North Transept, then South Transept — Probably 
 after 1273. Reredos — By Bishop Fox, of Winchester, c. 1500. Nave — De- 
 molished and rebuilt 1833 ; again demolished and rebuilt 1895. Monograph 
 by F. T. Dollman (1881). Became a cathedral 1st May 1905. 
 
 SOUTHWELL MINSTER, Notts. Secular Canons. Became Cathedral in 1SS4. 
 Commenced 1108-1114. Nave — Second quarter of XII. Cent. Choir was 
 building in 1233 ; a Chantry was founded in it in 1241. Chapter House — c. 1294. 
 (Guide to Southwell, by Rev. G. M. Livett.) In the glass are the arms of 
 Eleanor of Castile, who died in 1290. 
 
 SOUTHWOLD, Suffolk. Parish Church. Said to have been finished 1460-1461. 
 Porch— 1 488- 1 489. Suffolk Arch. Institute, viii. 413. 
 
 STAMFORD, St John. Parish Church. Finished in 145 1. Rickman, 305. 
 
 STAMFORD, St Leonard's Priory, Lincolnshire. Benedictine Monks. End of 
 XII. Cent. 
 
 STANTON HAROLD, Leicestershire. Parish Church. 1653. 
 
 STAVERDALE PRIORY, Somerset. Austin Canons. Nave, Choir, and Chantry 
 Chapel -Consecrated 1443. Rickman, 303. 
 
 STEWKLEY, Bucks. Parish Church. Granted to Kenilworth Priory, c. 1150. 
 Rickman, 92. 
 
 STEYNING, Sussex. Parish Church, c. 1160. Mr E. Sharpe. 
 
 STOKE GOLDINGHAM, Leicester. Parish Church. 1 275-1 290. Weale's 
 Quarterly Papers, i. 
 
 STONE, Kent Parish Church. Probably 1251-1274. Monograph by G. E. Street. 
 
654 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 STOW, Lincolnshire. Parish Church. Chancel — 1123-1148. Mr Bilson's Beginnings. 
 
 STRATA FLORIDA ABBEY, Cardigan. Cistercian Monks. 1 166-1203. S. W. 
 Williams' monograph. 
 
 STRATFORD ON AVON, Warwick. Collegiate. Choir— 1465-1491. Rickman, 
 308. 
 
 SUTTON ASHFIELD, Notts. Parish Church. Soon after 1391. Bishop Bucking- 
 ham's Mem., 379. 
 
 SUTTON ST MARY, or LONG SUTTON, Lincolnshire. Parish Church. Site 
 granted c. 1180. Fen and Marshland Churches, iii. 25. 
 
 SWAFFHAM, Norfolk. Parish Church. Chancel — 1474. Tower — Finished 1510. 
 
 TATTERSHALL, Lincolnshire. Castle and Church. Both 1433-1455. Rickman, 299. 
 TENTERDEN, Kent. Parish Church. Tower— 1462. Rickman, 308. 
 TERRINGTON ST CLEMENT, Norfolk. Parish Church, c. 1377-1399. Fen 
 and Marshland Churches, i. 37. Dr Seccombe in Norfolk Archeology, xii. ; 
 
 1389-1425- 
 
 TEWKESBURY ABBEY, Gloucester. Benedictine Monks. The Lordship of 
 Tewkesbury was granted in 1087 to Robert Fitz-Hamon, who founded the Abbey. 
 The monks entered their new Church in 1102. In 11 23 there was a consecration, 
 probably of the whole Church, including the Nave. Lady Chapel — First half of 
 XIII. Cent. Remodelling of Choir — Second quarter of XIV. Cent. Rickman, 89. 
 
 THORNEV ABBEY, Cambridge. Benedictine Monks. Part of Nave— 1085-1108. 
 Rickman, 85. 
 
 THORNTON ABBEY, Lincolnshire. Austin Canons. Became an Abbey in 1148. 
 Church — Begun in 1263. Chapter House — Begun in 1282; paved in 1308. 
 Gatehouse — 1382. Barbican — Perhaps 1520. Mr C. C. Hodges in Reliquary, ii. 1. 
 
 TIDESWELL, Derbyshire. Parish Church. John Foljambe, who "did many good 
 things in the building of the Church," died in 1358, and is buried in the Chancel. 
 
 TINTERN ABBEY, Monmouth. Cistercian Monks. Founded 1131. Conventual 
 Buildings — Rebuilt middle of XIII. Cent. Church — Begun 1269; and Presby- 
 tery, South Transept, Crossing, and two bays of the Nave finished 1288, when Mass 
 was said at the High Altar. Rest of Nave and part of North Transept not 
 finished till middle of XIV. Cent. Mr Harold Brakspear. 
 
 TONG, Shropshire. Collegiate. 1401-1411. Rickman, 295. 
 
 TUTBURY PRIORY, Stafford. Benedictine Monks. Founded in 1080; founder 
 interred in it in 1090. Mr C. Lynam in Archceological Journal, New Series, iii. 148. 
 
 TUXFORD, Notts. Parish Church. Building — 1473. Chancel — 1495. Rickman, 309. 
 
 TYNEMOUTH PRIORY, Northumberland. Benedictine Monks. Last years of 
 XII. Cent. Lady Chapel — c. 1400. Arch. Aeliana, 23, 29. 
 
 VALLE CRUCIS ABBEY, near Llangollen, Denbigh. Cistercian Monks. Said by 
 Dugdale to have been founded c. 1200. 
 
 WALSBERSWICK, Suffolk. Parish Church. Dedicated 1493. Unroofed 1695. 
 
 Suffolk Arch. Lnstitute, viii. 418. 
 WALTHAM ABBEY, Essex. Austin Canons. Nave — Not the work of King Harold 
 
 in 1062. The rebuilding of the Church was probably commenced after grants of 
 
 Henry I. in 1120. Building operations were going on in 1125 and 1126. See 
 
 Builder, 30th April and 21st May 1898. 
 WANBOROUGH, Wilts. Parish Church. Tower— 1435. Rickman, 301. 
 
CHRONOLOGY OK ENGLISH CHURCHES. 655 
 
 WARKWORTH, Northumberland. Parish Church. Vaulted Chancel— e. 1125. Mr 
 J. Bilson's Beginnings, 31C 
 
 WARWICK, St Mary. Colli-;. hate. Chancel -1381-1391. Beauchamp Chapel 
 Begun 1439- Nave — 1694-1706. Richman, 290, 302. 
 
 WAVERLEY ABBEY, Surrey. Cistercian Monks. Founded 112S. Monks' Infirmary 
 with Chapel— Before 1201, when the latter was hallowed. New Church Begun 
 1203. Five eastern altars hallowed 12 14. Three chapels in North Transept 
 1226. Three chapels in South Transept hallowed and Choir entered 1231. 
 Completed and hallowed 1278. Cott. MS. Vesp, A., xvi. 901; to r8o. Mr Harold 
 Brakspear. 
 
 WELLS CATHEDRAL, Somerset. Secular Canons, (i.) Begun by Reginald de 
 Bohun, who became Bishop in 1174 : and made a large grant for the fabric prior 
 to 1 180. A charter referring to the "admirable structure of the rising Church" 
 is attested by witnesses who appear elsewhere in 1206 and 1221. Probably he 
 built the Choir and East of the Transept. (2.) The rest of the Transept and the 
 Nave and North Porch seem to have been completed in two sections before 1206. 
 (3.) West Front — 1220 to consecration in 1239. Undercroft to Chapter House — 
 c. 1270. Staircase to Chapter House — c. 12S6-1300. Chapter House — Not finished 
 before 1319. Lady Chapel — Finished before 1326. Central Tower — Raised and 
 roofed c. 1 32 1. Retrochoir and Remodelling of Choir— 1329-1363. South-west 
 Tower — After 1386. North-west Tower — After 1424. Bishop's Palace — 1225- 
 1239. Ruined Hall of Palace — 12S0-1292. Gatehouse — 1340. Canon Church's 
 Chapters in the Early History of Wells. 
 
 WENLOCK PRIORY, Shropshire. Cluxiac Monks. Last quarter of XII. Cent. 
 
 WESTENHANGER, Kent. Parish Church. 1520. Rich/nan, 317. 
 
 WESTMINSTER ABBEY. Benedictine Monks. Parts underground are the work of 
 Edward the Confessor ; begun 1050. Choir — Dedicated 1065. (1.) Present Church 
 commenced 1245. In I2 S3 canvas was bought for the windows of the Chapter 
 House. Choir, Transept, and one bay west of the Crossing, the Revestry, and 
 so much of the Cloister as lay within the South Transept, with its east wall to 
 beyond the doorway of the Chapter House — 1245-1260. (2.) Second to fifth 
 eastern bays of the Pier-arcade and Triforium, first to fourth bays in the 
 Clerestory, also first to fourth bays in north walk of Cloister — 1260-1269. The 
 monks entered the choir in 1269. (3.) East walk of the Cloister carried to 
 the corner — c. 1330-1350. The Cloister was nearly finished in 1352. (4.) West 
 bays of the Nave probably commenced c. 1350, and went on slowly under 
 Richard II. (1377-1399); almost stopped in Henry IY.'s time (1399-1413); and 
 were nearly finished by Henry V. (1413-1422) ; except — (5.) The fifth and sixth 
 bays from the east and the westernmost bay of the Clerestory, and the west front. 
 In the vaulting Tudor badges appear — c. 1500. (6.) Henry the Seventh's Chapel 
 — 1500-1512. Mr I. T. Micklethwaite in Archmological Journal, li. 
 
 WESTMINSTER HALL. Built by William Rufus, and first used in 1099. Re- 
 modelled and new roof, 1 397-1 399. Rich/nan, 294. 
 
 WESTUTCK, Norfolk. Parish Church. Tower — 1460 and 1473. 
 
 WHISTON, Northants. Parish Church. Built by Anthony Catesby, who died 1553. 
 Mr R. P. Brereton. 
 
 WHITBY ABBEY, Yorkshire. Benedictine Monks. Destroyed by the Danes in 
 1 1 73, and rebuilt at the beginning of the XIII. Cent. 
 
 WIMBORNE MINSTER, Dorset. Secular Canons. Crossing— Early XII. Cent. 
 Central Tower and East Nave — Third quarter of XII. Cent. Presbyter) — c. 1220. 
 
656 CHRONOLOGY OF ENGLISH CHURCHES. 
 
 Lengthening of Transepts and Nave — XIV. Cent. Clerestory and Western Tower 
 — 1448-1464. Perkins' Wimborne Minster. 
 
 WINCHELSEA, Sussex. Parish Church. The earliest monument is that of Gervase 
 Alard, who was living in 1306. The Chancel appears to be of the first quarter of 
 the XIV. Cent. 
 
 WINCHESTER CASTLE. Shire or Great Hall. 1222-1235. Rickman, 165. 
 
 WINCHESTER CATHEDRAL, Hampshire. Benedictine Monks. Crypt and 
 Transept Ends — Begun 1079, consecrated 1093. Crossing and Inner Parts of 
 Transept — Built after the fall of the Central Tower in 1107. Central Tower — 
 Probably second half of XII. Cent. Eastern Part and Lady Chapel — Begun 
 1202; the west part of the Lady Chapel may be a little later. Presbytery — 
 Probably temp. Bishop Edington, 1345- 1366. Remodelling of the Nave — Begun 
 by William of Wykeham in 1371. Probably the West Front is of this date ; Rest 
 of Nave, 1394 to c. 1460. Reredos — c. 1480. Remodelling of Lady Chapel — After 
 1487 to c. 1500. Fox's work in Presbytery — 1500-1528. Choir Enclosure — 
 Dated 1525. Air IV. H. St John Hope. 
 
 WINCHESTER COLLEGE. 1387-1393. Chantry Chapel and Library in Middle of 
 Cloister — 1420. Rich man, 292. 
 
 WINCHCOMBE, Gloucester. Parish Church. 1456-1474. Rickman, 306. 
 
 WINDSOR CASTLE, Berkshire. Royal Chapel of St George. Begun by Edward 
 IV. c. 1473 ; but little done. South Transept — 1481-1502. Choir Vault — Begun 
 in 1507; but unfinished in 1519. Mr Ambrose Poynter, on heraldic evidence, 
 assigned the date of 1528 to the High Vault, and of 1537 to the Aisle Vaults 
 of the Nave. Rickman, 3 1 4. 
 
 WITHAM CHARTERHOUSE, Somerset. Carthusian Monks. Chapel of the 
 Frary or House of the Lay Brothers; now Parish Church, n 76-1 186. Rick- 
 man, 109. 
 
 WORCESTER CATHEDRAL. Benedictine Monks. Crypt— Begun 1084; Synod 
 held in it in 1092. West Bays of Nave — Probably c. n 70. (See Prior, 91 and 
 121.) Retrochoir — Probably after the fire of 1202 and the canonisation of 
 St Wolfstan in 1203: dedicated 1218. (Prior, 162, note.) Foundation stone of 
 Choir laid 1224. Five East Bays of North Nave — 1317-1324. Guest Hall — 1320. 
 Prater and Cloister — 1372. Sub-vault of Prater — Early XII. Cent. Tower — 
 Finished 1374- Vault of Crossing — 1376. Two Bays of North Nave and seven 
 of South Nave— May be c. 1350-1377. Vault of two Western Bays of Nave — 
 1377. Willis in Arch. Journal, xx. 
 
 WORKSOP PRIORY, Notts. Austin Canons. Probably c. 1 175. Assoc. Soc. Reports, 
 v. 219. 
 
 WYMINGTON, Bedford. Parish Church. Brass of founder, who died in 1391. 
 Brandon's Parish Churches, 93. 
 
 WYMONDHAM ABBEY, Norfolk. Benedictine Monks. Founded as a Priory 
 before 1107; became an Abbey in 1448. Central Tower — 1390-1409. Font — 
 c. 1410. Nave and North Aisle, Clerestory and Roofs — 1432-1445. West Tower 
 — 1445-1478. South Aisle — 1534- Rev. J. L. Petit's monograph. 
 
 YARMOUTH. Parish Church. Founded and built 1096-1119. Nave Arcade — 
 Probably c. 1190. Aisles and West Front — Perhaps c. 1230. A. W. Morant in 
 Norfolk and Norwich Arch. Soc, vii. 215. 
 
 YATTON, Somerset. Parish Church. Nave and Aisles — c. 1375. Finished in 1475. 
 Rickman, 309. 
 
CHRONOLOGY OF ENGLISH CHURCHES. 657 
 
 YORK CATHEDRAL. Secular Canons. Part of Crypt— 1154-1181. South Tran- 
 sept — 1230 — c. 1241. North Transept 1241-1260. Nave — 1291-1324; except 
 wooden vault of c. 1354. Part of Wist Front— West window glazed, 1338. 
 Chapter House — "The glass was executed before 1307." Vestibule After Chapter 
 House. Presbytery — c. 1361-1370. Choir— c. 1380 — c. 1400. East Window 
 Glazed — 1405-1408. Central Tower — c. 1400- 1423. South-west Tower — 
 1433-1447. North-west Tower — 1470-1474. See Guides to York, by A. 
 Clutton Brock, 20 ; by G. Benson, 84 : and Rickman, 168, 173, 181, 185, 240, 215, 
 234, 241, 244, 246, 256, 296, 273, 292, 308 : and Willis on York Minster* 
 
 YORR, St Mary's Abbey. Benedictine Monks. The first stone was laid in 1271. 
 The first stone of the columns was not laid till 1273. Finished before 1296. 
 Rickman, 175.T 
 
 YORR, St Martin-LE-Grand. Parish Church. Tower and Glass — 1437. Rick- 
 man, 302. 
 
 YORR, St Michael-le-Belfry. Parish Church. 1535-1545. Bloxam, 275. 
 
 * For the English Cathedrals, see the illustrated articles in the Builder, published in one 
 volume in 1894 ; Britton's English Cathedrals ; Browne Willis' Survey of the Cathedrals, 1727 : 
 and Carter's plans and drawings. 
 
 + For the English Abbeys, see the illustrated articles in the Builder, 1894 to 1901. 
 
 2 T 
 
658 ROMANESQUE PIERS. 
 
 Romanesque Piers of the Greater Churches. 
 
 1. Durham Choir. 
 
 2. Ely North Transept, West Aisle. 
 
 3. Winchester Nave. 
 
 4. Winchester North Transept, West Aisle. 
 
 5. St Albans Nave. 
 
 6. Peterborough, Minor Pier in Nave. 
 
 7. Norwich Nave. 
 
»$,, 
 
 Romanesque Piers of thf Greater Churches. 
 
66o GOTHIC PIERS. 
 
 Gothic Piers of the Greater Churches. 
 
 i. Kirkstall. 
 
 2. Roche. 
 
 3. Byland. 
 
 4. Wells. 
 
 5. York St Mary's. 
 
 6. Guisborough. 
 
 7. Bridlington. 
 
 8. Lichfield Choir. 
 
 9. Exeter Nave. 
 
 10. Canterbury Nave. 
 
 11. Bristol Choir. 
 
 12. St George's, Windsor (enlarged, page 255). 
 
I,,,] 
 
 © 
 
 © 
 
 ■' 
 
 d 
 
 © 
 
 •0 
 
 ,.- 
 
 SCALE OF FEET 
 
 Gothic Piers of the Greater Churches. 
 
662 PIERS OF PARISH CHURCHES. 
 
 Piers of Parish Churches. 
 
 i. Northampton St Peter. 
 
 2. Market Deeping'. 
 
 3. Skelton. 
 
 4. High am Ferrers. 
 
 5. Northborough. 
 
 6. Bottisham. 
 
 7. Lave n ham. 
 
 8. Coggeshall. 
 
 9. Long Melford Nave. 
 
66- 
 
 Fiers of Parish Churches. 
 
664 PIERS AND ARCHES. 
 
 Piers and Arches. 
 
 1. Northampton St Peter. 
 
 2. Warmington. 
 
 3. Market Deeping. 
 
 4. Barnack. 
 
 5. Stamford All Saints Nave. 
 
 6. Stamford All Saints Nave. 
 
 7. Stone. 
 
 8. Stamford All Saints Chancel. 
 
 9. Finedon. 
 
 10. Stamford St Martin. 
 
 11. Hutton. 
 
 12. Long Melford Lady Chapel. 
 
66s 
 
 12 e o 
 
 Scale of Feet 
 
 Piers and Arches. 
 
666 
 
 Arches — Sheet I. 
 
 i. Chichester Nave. 
 2. Heding-ham. 
 
 3. Warmington. 
 
 4, 5. Wawne. 
 
60/ 
 
 ARCHES — Sheet II. 
 i. Walsoken. 3. New Shoreham. 
 
 2. New Shoreham. 4. Byland. 
 5. Grimsby. 
 
668 
 
 Arches — Sheet III. 
 
 i. Temple Choir. 
 2. Lincoln Nave. 
 
 3. Temple Choir. 
 
 4. St Mary's Abbey, York. 
 
66g 
 
 Arches — Sheet IV. 
 
 i. Bridlington. 
 
 2. Lincoln Presbytery. 
 
 3. Tintern Choir. 
 
 4. Winchelsea. 
 
670 
 
 Arches — Sheet V. 
 
 1. Leadenham. 3. Northborough. 
 
 2. Helpringham. 4. Beverley St Mary Chancel. 
 
6-i 
 
 Z 
 
 j Feet 
 
 Ar< hes — Sheet VI. 
 i. Bottisham. 3. Kettering. 
 
 2. Chelmsford. 4. George Inn, Glastonbury. 
 
 5. St Mary, I Ixford. 
 
672 RIB-MOLDS. 
 
 Rib-Molds— Sheet I. 
 
 1. Gloucester Crypt. 
 
 2. Canterbury Treasury. 
 
 3. Birkin Apse. 
 
 4. Abbaye-aux-Hommes, Caen. 
 
 5. Kirkstall Nave Aisle, Transverse Rib. 
 
 6. Kirkstall Nave Aisle, Diagonal Rib. 
 
 7. Furness Nave Aisle, Transverse Rib. 
 
 8. Furness Nave Aisle, Diagonal Rib. 
 
 9. Furness Nave Aisle, Wall Rib. 
 
 10. Buildwas Chapter House, Transverse Rib. 
 
 11. Buildwas Chapter House, Diagonal Rib. 
 
6/3 
 
 Rm-MuLDs— Sheet I. 
 
 2 U 
 
674 RIB-MOLDS. 
 
 Rib-Molds— Sheet II. 
 
 i. Roche, Transverse Rib. 
 
 2. Roche, Diagonal Rib. 
 
 3. Byland Aisle, Transverse and Diagonal Ribs. 
 
 4. Byland Aisle, Wall Rib. 
 
 5. Jervaulx Aisle, Transverse Rib. 
 
 6. London, Aisle of Nave of Temple Church. 
 
 7. London, Aisle of Nave of Temple Church. 
 
 8. London, Aisle of Nave of Temple Church. 
 
 9. London, Aisle of Nave of Temple Church. 
 
 10. London, Aisle of Nave of Temple Church. 
 
 1 1. London, Aisle of Nave of Temple Church. 
 
 12. Whitby Choir Aisle, Transverse and Diagonal Ribs. 
 
 13. Whitby Choir Aisle, Wall Rib. 
 
 14. Rievaulx Choir and Aisles, Transverse and Diagonal Ribs. 
 
 1 5. Rievaulx Choir and Aisles, Wall Rib. 
 
675 
 
 Rib-Molds — Sheet II. 
 
676 RIB-MOLDS. 
 
 Rib-Molds— Sheet III. 
 
 i. Fountains Choir Aisle, Transverse and Diagonal Ribs. 
 
 2. Fountains Choir Aisle, Wall Rib. 
 
 3. Lincoln Galilee. 
 
 4. Bridlington Nave Aisle, Transverse and Diagonal Ribs. 
 
 5. Tintern Aisles, Transverse and Diagonal Ribs. 
 
 6. Tintern Aisles, Wall Rib. 
 
 7. York, St Mary's Abbey ; Nave Aisle, Transverse Rib. 
 S. York, St Mary's Abbey ; Nave Aisle, Diagonal Rib. 
 9. York, St Mary's Abbey ; Nave Aisle, Wall Rib. 
 
 10. Howden Choir, Transverse and Diagonal Ribs. 
 
 11. Howden Choir, Wall Rib. 
 
 12. Selby Choir Aisle, Transverse and Diagonal Ribs. 
 
 13. Selby Choir Aisle, Wall Rib. 
 
 14. Wells, Porch of Staircase of Vicars' Close. 
 
 15. Great Chalfield, Wilts; Porch of Manor House. 
 
 16. Great Chalfield, Wilts; Porch of Manor House. 
 
 17. Kenilworth, Octagonal Lobby. 
 
6 77 
 
 Rib-Molds— Sheet III. 
 
678 GROUND-COURSES. 
 
 Ground or Basement Courses. 
 
 i. Fountains Nave. 
 
 2. Hexham North Transept. 
 
 3. Whitby Choir. 
 
 4. Fountains Choir. 
 
 5. Rievaulx Choir. 
 
 6. Bridlington Nave. 
 
 7. Welbourne. 
 
 8. Kettering Tower. 
 
 9. Stoke Golding. 
 
 10. Merton College Chapel Transept, Oxford. 
 
 1 1 . All Souls' College Chapel, Oxford. 
 
 12. Magdalen College Chapel, Oxford. 
 
679 
 
 Ground-Courses. 
 
68o STRINGS. 
 
 Strings— Sheet I. 
 
 i. Fountains, Aisle of Nave. 
 
 2. Fountains, Clerestory of Nave. 
 
 3. Furness, Aisle of Nave. 
 
 4. Roche Triforium. 
 
 5. Roche Aisle. 
 
 6. Byland Triforium. 
 
 7. Hexham, Aisle of Choir. 
 
 8. Hexham, Triforium of Choir, 
 g. Jervaulx Aisle. 
 
 10. Jervaulx Aisle. 
 
 11. Whitby Triforium. 
 
 1 2. Whitby East End. 
 
 13. Whitby Aisle. 
 
 14. Fountains Aisle. 
 
 15. Rievaulx Triforium. 
 
 16. Netley Aisle. 
 
 17. Netley Aisle. 
 
 18. Bridlington Aisle. 
 
68 1 
 
 o z> e> 9 12 „ 
 
 ■ ■ . ■ INCHES 
 
 Strings — Sheet I. 
 
682 STRINGS. 
 
 Strings — Sheet II. 
 
 i. Bridlington Clerestory. 
 
 2. Tintern Clerestory. 
 
 3. St Mary, York, Aisle. 
 
 4. Guisborough Aisle. 
 
 5. Howden, Aisle of Choir. 
 
 6. Selby, Aisle of Choir. 
 
 7. Brandon's Analysis. 
 
 8. Brandon's Analysis. 
 
 9. Brandon's Analysis. 
 
 10. Brandon's Analysis. 
 
 1 1 . Austrey Dripstone. 
 
 12. Austrey Dripstone. 
 
 13. Frampton, Dripstone. 
 
 14. Brandon's Analysis. 
 
 15. Brandon's Analysis. 
 
 16. Brandon's Analysis. 
 
 17. Brandon's Analysis. 
 
 18. Cottineham. 
 
68 3 
 
 Strings — Sheet II. 
 
684 ABACI AND CAPITALS. 
 
 Moldings of Abaci and Capitals — Sheet I. 
 
 i. Canwick. 
 
 2. Harmston. 
 
 3. Whaplode. 
 
 4. Whaplode. 
 
 5. Aswarby. 
 
 6. Moulton. 
 
 7. Fulbeck. 
 
 8. Whaplode. 
 g. Horbling. 
 
 10. Weston. 
 
 n. Waddington. 
 
 12. Lincoln. 
 
 13. Lincoln. 
 
 14. Lincoln St Mary Wigford. 
 
 15. Threckingham. 
 
 16. Threckingham. 
 
 17. Lincoln. 
 
 18. Lincoln. 
 
 19. Lincoln. 
 
 20. Buildwas. 
 
 21. Fountains Nave. 
 
 22. Kirkstall. 
 
 23. Furness. 
 
 24. Byland. 
 
 25. Deeping St James. 
 
 26. Deeping St James. 
 
 27. Deeping St James. 
 
 28. Market Deeping. 
 
 29. Northampton St Sepulchre. 
 
68s 
 
 £ ^ ^ ^ T *v 
 
 ©r © \_ ©V ©\_ ©S © »- 
 
 L L. l_ ' 1_ |_ L. 
 
 © r © r © r © r © r © r © ' 
 
 < «i (^^ ^ 
 
 ® r © r ©> © r © p © r 
 
 .\i:\i i and Capitals — Sheet I. 
 
686 CAPITALS AND BASES. 
 
 [olded Capitals and Bases — Sheet II. 
 
 i. Whitby. 
 
 2. Fountains. 
 
 3. Netley. 
 
 4. Bridlington. 
 
 5. Tintern. 
 
68: 
 
 Capitals — Sheet II. 
 
638 CAPITALS. 
 
 Molded Capitals — Sheet III. 
 
 i. Doddinghurst, Shaft in Doorway. 
 
 2. East Thorpe Sedilia. 
 
 3. Durham Eastern Transept. 
 
 4. London, Temple Choir. 
 
 5. Great Baddow Pier. 
 
 6. Warmington, Shaft in West Doorway. 
 
 7. Warmington, Shaft in West Doorway. 
 
 8. Wiggenhall St Mary the Virgin, Shaft in North Doorway. 
 
 9. Wiggenhall St Mary the Virgin, Shaft in North Doorway. 
 10. Fairstead Sedilia. 
 
68 9 
 
 Capitals — Sheet III. 
 
 2 x 
 
690 CAPITALS. 
 
 Molded Capitals — Sheet IV. 
 
 i. Swayton Nave. 
 
 2. Sleaford North Transept. 
 
 3. Sleaford North Transept. 
 
 4. Helpringham Nave. 
 
 5. Heckington Tower. 
 
 6. Heckington Chancel. 
 
 7. Heckington Nave. 
 
 8. Leadenham Nave. 
 
 9. Chipping Hill, Piers of Nave. 
 
 10. Asgarby Nave. 
 
 11. Aswarby Nave. 
 
 12. Holbeach Nave. 
 
 13. Tiltey Sedilia. 
 
 14. Rushden, Piers of North Side of Nave. 
 
 15. Rushden, Piers of North Side of Nave. 
 
 16. Rushden, Piers of North Side of Nave. 
 
691 
 
 L^J — I 1 — 1 I 1 I I tiit 
 
 FEET 
 
 Capitals — Sheet IV. 
 
692 CAPITALS. 
 
 Molded Capitals — Sheet V. 
 
 1. Terling, Pier of Nave. 
 
 2. Oxford St Mary, Pier in Nave. 
 
 3. Chelmsford, Shaft in Porch. 
 
 4. Chelmsford, Pier in Chancel. 
 
 5. Eltham Palace, Shaft in Doorway. 
 
 6. Ingatestone, Pier in Nave. 
 
 7. Booking, Shaft in Doorway. 
 
6w 3 
 
 c 
 
 ^ 
 
 / /V J 
 
 to 
 
 ? 
 
 U 
 
 L 
 
 © 
 
 ex 
 
 'iSb 
 
 f, Y .V 
 
 XD 
 
 e 
 
 © 
 
 _i i i — i — i — i- 
 
 Y 
 
 c 
 
 ^L 
 
 c 
 
 7 
 
 FEET 
 
 Capitals — Sheet V. 
 
694 BASE-MOLDS. 
 
 Base-Molds — Sheet I. 
 
 i. Attic Base. 
 
 2. Waterholding Base. 
 
 3. Boxgrove Cloister. 
 
 4. Boxgrove Cloister. 
 
 5. Hedingham Castle, Window Shaft. 
 
 6. Furness. 
 
 7. Fountains. 
 
 8. Jervaulx. 
 
 9. Barnack. 
 
 10. Winchester St Cross. 
 
 1 1. Winchester St Cross. 
 
695 
 
 ® 
 
 J o J J 
 
 c 
 
 ) 
 
 ) 
 
 (' 
 
 c ~^ 
 
 
 J © U© © | © 
 
 Bases — Sheet I. 
 
696 BASE-MOLDS. 
 
 Base-Molds — Sheet II. 
 
 1. Salisbury Cloister. 
 
 2. Winchelsea Piscina. 
 
 3. Winchelsea Sedilia. 
 
 4. Winchelsea, Shaft of Window. 
 
 5. Winchelsea Pier. 
 
 6. Austrey Pier. 
 
 7. Tintern. 
 
 S. Tintern, Pier of Nave. 
 9. Great Gonerby Pier. 
 
 10. Wells Chapter House, Central Column. 
 
 11. Selby Choir, Arcading and Vaulting Shaft. 
 
 12. Guisborough, Arcading and Window Jamb. 
 
 13. Beverley St Mary, Shaft of Window in Flemish Chapel. 
 
 14. Leadenham Pier. 
 
 15. Billingborough Pier. 
 
 16. Osbournby Pier. 
 
 17. Sleaford Pier. 
 iS. Holbeach Pier. 
 
 19. Bottisham Pier. 
 
 20. Beverley St Mary, North Side, Pier of Nave. 
 
 21. Bristol, Colston House, Shaft of Fireplace. 
 
 22. Arundel, Pier of Nave. 
 
 23. Louth, Respond of Tower Arch. 
 
 24. Colchester Pier. 
 
Bases— Sheet II. 
 
698 WINDOW-MOLDS 
 
 Window-Molds — Sheet I. 
 
 1. Rievaulx, Lower Triplet of East End, Jamb and Bases. 
 
 2. Rievaulx, Arch and Abaci. 
 
 3. Netley, East Window, Jamb and Bases. 
 
 4. Netley, Arch and Abaci. 
 
 5. Stone, both Jambs, Bases, and Mullion. 
 
 6. Stone, Arch. 
 
 6a. Stone, Arch Enlarged. 
 
 7. Guisborough, Aisle of Choir, Jamb and Bases. 
 
 8. Guisborough, Arch and Abaci. 
 
 9. Tiltey, Mullions. 
 
699 
 
 Windows — Sheet I. 
 
7oo 
 
 WINDOW-MOLDS. 
 
 Window-Molds — Sheet II. 
 
 i. Bishopstone, Wilts, both Jambs and Mullion. 
 
 2. Wells, Vicars' Close, one Jamb and Mullion. 
 
 3. Sleaford, Arch and Mullion. 
 
 4. Maidstone All Saints, one Jamb and Mullion. 
 
 5. Maidstone All Saints, one Jamb and Mullion. 
 
 6. Oxford All Souls - ' Chapel, one Jamb and Mullion. 
 
7oi 
 
 1 O I 
 
 2 -> * 
 
 FEET 
 
 Windows — Sheet II. 
 
"02 DOORWAYS. 
 
 Molds of Doorways — Sheet I. 
 
 i. Rochester Cathedral, Jamb of West Doorway. 
 
 2. Arch of ditto. 
 
 3. Plan of ditto. 
 
 4. Codford St Mary, Wilts, Jamb and Arch of Chancel Arch. 
 
 5. Wenlock, Arch and Abaci of Doorway into Cloister. 
 
 6. Jamb and Bases of ditto. 
 
 7. Colchester Magdalen Chapel, Jamb and Base of Doorway. 
 S. Arch of ditto. 
 
703 
 
 Doorways— Sheet I. 
 
704 DOORWAYS. 
 
 Molds of Doorways — Sheet II. 
 
 i. Clare Church, Suffolk ; West Doorway; Jamb, Arch, and Abaci. 
 
 2. Wiggenhall St Mary the Virgin, Norfolk ; North Doorway ; Jamb, Arch, and Abaci. 
 
 3. Stamford, St Leonard's Priory ; West Doorway ; Jamb, Arch, and Abaci. 
 
 4. Middleton Cheney, Northants ; Doorway of South Porch. 
 
 5. London, St Etheldreda's, Holborn ; Arch of Doorway on South Side (c. 1380, Caveler). 
 
 6. Jamb, Bases and Abaci of ditto. 
 
 7. Warmington ; Doorway of South Porch ; Jamb, Arch, and Abaci. 
 
705 
 
 Doorways — Sheet II. 
 
 2 Y 
 
7 o6 DOORWAYS. 
 
 Doorways — Sheet III. 
 
 i. Houghton-le-Dale, Norfolk ; Jamb. 
 
 2. Oxford, Merton College Transept ; Plan of Moldings above Capitals. 
 
 3. Oxford, Merton College Transept ; Plan of Columns showing Capitals. 
 
 4. Oxford, Merton College Transept ; Plan of Columns showing" Bases. 
 
 5. Oxford, Magdalen College Chapel ; West Doorway. 
 
 6. Fotheringhay ; West Doorway ; Arch and Abacus. 
 
 7. Fotheringhay ; West Doorway ; Jamb and Bases. 
 
707 
 
 Doorways— Sheet III. 
 
INDEX TO ILLUSTRATIONS. 
 
 NOTE. — The full titles and descriptions of the books referred to in this list will be found in 
 the List of Authorities, pages xii-xvi. The illustrations to which no book is appended 
 as authority are from photographs and drawings. Professional photographers are 
 distinguished by the name of their town being given. 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 No. 
 
 in 
 
 Page. 
 
 Source. 
 
 Abbotsbury, Dorset 
 
 Exterior from N.W. 
 
 284 
 
 
 J. Norton. 
 
 Do. 
 
 Interior from VV. 
 
 2S5 
 
 
 do. 
 
 Achurch, Northants 
 
 Plan 
 
 2I 5 
 
 7 
 
 Architectural Society's 
 Dictionary. 
 
 Adderbury, Oxon. 
 
 Roof ----- 
 
 558 
 
 in. 
 
 Bloxam's Principles. 
 
 Adel, Yorkshire 
 
 Exterior from N.W". 
 
 220 
 
 
 Churches of Yorkshire. 
 
 Aldeburgh, Suffolk 
 
 Fifteenth-century base 
 
 454 
 
 
 F. Bond. 
 
 Do. 
 
 Fifteenth-century capital - 
 
 454 
 
 
 do. 
 
 Angers cathedral • 
 
 Interior from VV. 
 
 3iS 
 
 
 do. 
 
 Do. 
 
 Plan of nave vault - 
 
 30S 
 
 7 
 
 do. 
 
 Arundel church 
 
 Base of pier - 
 
 697 
 
 22 
 
 Colli ng's Details. 
 
 Asgarby 
 
 Molded cap of pier in nave 
 
 691 
 
 10 
 
 Sharpe's Lincoln. 
 
 Ash, Kent 
 
 Base in nave - 
 
 44S 
 
 5 
 
 F. Bond. 
 
 Ashby St Ledger, 
 
 Window - 
 
 502 
 
 
 FJrandon's Analysis. 
 
 Northants 
 
 
 
 
 
 Aswarby 
 
 Abacus - - - - - 
 
 6S5 
 
 5 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 Athens - 
 
 Molded cap of pier in nave 
 Capital of Temple of the Winds 
 
 691 
 425 
 
 1 1 
 
 do. 
 Mauch's Orders. 
 
 Do. 
 
 Capital of Choragic monument 
 
 425 
 
 
 do. 
 
 Do. 
 
 Do. 
 Attic base 
 Austp.ey, Warwick - 
 
 of Lysicrates 
 Acanthus from do. - 
 Capital of Temple of Zeus 
 
 Base of pier 
 
 426 
 
 425 
 695 
 697 
 
 1 
 
 1 
 6 
 
 Colling's Foliage. 
 Mauch's Orders. 
 
 do. 
 Colling's Details. 
 
 Do. 
 
 Buttress - 
 
 356 
 
 
 do. 
 
 Do. 
 
 Do. 
 
 1 lupstone 
 do. 
 
 683 
 6S3 
 
 1 1 
 12 
 
 do. 
 do. 
 
 Baginton, Warwick 
 Barnack, Northants 
 
 Plan 
 
 Pier arch and abacus 
 
 216 
 665 
 
 3 
 
 4 
 
 Parker's Churches. 
 Churches of the Nene 
 Valley. 
 
 Do. 
 
 Pier, cap, and base - 
 
 695 
 
 9 
 
 do. 
 
7io 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Barnack, Northants 
 
 Spire from S.W. 
 
 613 
 
 
 F. Bond. 
 
 Barton-le-Street, 
 
 Capital ----- 
 
 427 
 
 
 C. C. Hodges. 
 
 Yorkshire 
 
 
 
 
 
 Bath 
 
 Exterior from N.E. - 
 
 373 
 
 
 F. Bond. 
 
 Bayeux cathedral, 
 
 Interior of choir from S.W. 
 
 5 2 5 
 
 
 do. 
 
 Calvados 
 
 
 
 
 
 Beckley, Oxford - 
 
 Staircase turret 
 
 601 
 
 
 Parker's Glossary. 
 
 Beeleigh abbey, 
 
 Plate tracery - 
 
 469 
 
 
 Hadfield's Essex. 
 
 Essex 
 
 
 
 
 
 Bere, Essex - 
 
 Three mullions 
 
 515 
 
 
 do. 
 
 Beverley minster, 
 
 Base in plan and elevation ; 
 
 45 2 
 
 
 Colling's Details. 
 
 Yorkshire 
 
 from back of reredos 
 
 
 
 
 Do. 
 
 Corbel-table and parapet 
 
 386 
 
 I 
 
 do. 
 
 Do. 
 
 Doorway of transept 
 
 574 
 
 
 G. H. Lovegrove. 
 
 Do. 
 
 Exterior from S.W. 
 
 599 
 
 
 F. R. Taylor. 
 
 Do. 
 
 Exterior from S. (collotype) face 
 
 1 76 
 
 
 C. C. Hodges. 
 
 Do. 
 
 Interior of choir from S.E. 
 
 5 1 
 
 
 do. 
 
 Do. 
 
 Percy tomb from N. 
 
 269 
 
 
 F. Bond. 
 
 Do. 
 
 Pier in N.E. transept 
 
 244 
 
 
 do. 
 
 Do. 
 
 Plan ; omitting the later N.E. 
 chapel 
 
 i5 2 
 
 I 
 
 John Bilson. 
 
 Do. 
 
 Tracery at back of the 
 
 486 
 
 
 Colling's Details. 
 
 
 reredos 
 
 
 
 Beverley, St Mary, 
 
 Arch-mold of chancel 
 
 670 
 
 4 
 
 do. 
 
 Yorkshire 
 
 
 
 
 
 Do. 
 
 Base of shaft of window in N.E. 
 chapel 
 
 697 
 
 13 
 
 do. 
 
 Do. 
 
 Exterior from S.E. - 
 
 603 
 
 
 F. Bond. 
 
 Do. 
 
 Ground story of chancel - 
 
 254 
 
 
 Colling's Details. 
 
 Do. 
 
 Interior of nave from S.E. 
 
 253 
 
 
 F. Bond. 
 
 Do. 
 
 Parapet - - - - - 
 
 387 
 
 1 
 
 Colling's Details. 
 
 Do. 
 
 Pier, base, and cap, north side 
 of nave 
 
 697 
 
 20 
 
 do. 
 
 Do. 
 
 South porch - 
 
 365 
 
 
 F. Bond. 
 
 Do. 
 
 West front 
 
 366 
 
 
 F. R. Taylor. 
 
 Beziers, St Aphro- 
 
 Pier 
 
 236 
 
 
 F. Bond. 
 
 disie 
 
 
 
 
 
 Billingborough, 
 
 Base of pier - 
 
 697 
 
 15 
 
 Sharpe's Lincoln. 
 
 Lincolnshire 
 
 
 
 
 
 Bingham, Notts 
 
 Spire ----- 
 
 616 
 
 
 Johnson's Reliques. 
 
 Binham, Norfolk 
 
 West doorway 
 
 575 
 
 
 F. Bond. 
 
 Do. - 
 
 West window (the cusps should 
 be floriated) 
 
 47i 
 
 
 Rev. R. Corrie Castle. 
 
 Do. 
 
 West window of aisle (the 
 lower window lighted the 
 aisle, the upper the triforium) 
 
 470 
 
 
 Colling's Details. 
 
 Birkin, Yorkshire - 
 
 Rib in apse - 
 
 673 
 
 3 
 
 Bilson's Beginnings. 
 
 Bishop's Cleeve, 
 
 Pinnacle 
 
 363 
 
 
 Rickman's Styles. 
 
 Gloucester 
 
 
 
 
 
 Bishopstone, Wilts 
 
 Both jambs and mullion of 
 windows 
 
 701 
 
 1 
 
 Weale's Papers. 
 
 Do. 
 
 Parapet - 
 
 3 8 7 
 
 3 
 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 7ii 
 
 BOOKING 
 
 Boston, Lincoln- 
 shire 
 
 Do. 
 
 Do. 
 Bottisham, Cam- 
 bridge 
 
 Do. 
 
 Do. 
 Boxford, Suffolk - 
 Boxgrove priory, 
 Sussex 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Boyton, Wilts 
 Bredon, Worcester 
 Bridgwater, 
 
 Somerset 
 Bridlington priory, 
 Yorkshire 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 Brixton, Norfolk 
 Bristol cathedral - 
 Do. 
 Do. 
 Do. 
 Bristol, Colston 
 
 House 
 Bristol, St James - 
 Bristol, St Mary 
 Redcliffe 
 Do. 
 Do. 
 Brixworth, 
 Northants 
 Bucknell, Oxford- 
 shire 
 
 Molded cap of shaft in doorway 
 Exterior from S.E. (collotype) 
 
 Plan - - - - 
 
 South doorway of nave 
 
 Arch mold - - - - 
 
 Base of pier - 
 Plan of pier, base, 
 N. porch 
 Base in cloister 
 
 plinth 
 
 do. 
 Exterior from S.E. - 
 Interior from W. 
 Plan of vault of choir 
 Circular window 
 Pinnacles - 
 
 Exterior from S.E. - 
 
 Exterior of nave from N. 
 
 Exterior of nave from S. - 
 Interior of nave from N. - 
 Interior of nave from S. - 
 Ground-course of nave 
 Molded cap and base 
 Parapet of north aisle 
 Part of arch-mold - 
 Plan of pier - 
 Transverse and diagonal ribs in 
 
 aisle 
 String of aisle - 
 String of clerestory - 
 Roof 
 
 Plan of pier of S. aisle of choir 
 Transverse section of choir 
 Vault of Berkeley chapel 
 Vault of S. aisle of choir 
 I Ins.' of shaft of fireplace 
 
 Circular window in west front - 
 Exterior of S. transept and choir 
 
 Interior of choir from S.W. 
 
 Parapet - 
 
 Arch of nave - - - - 
 
 Exterior from N.E. - 
 
 
 No. 
 
 Page. 
 
 in 
 Page. 
 
 693 
 
 7 
 
 222 
 
 
 2 l6 
 
 4 
 
 129 
 671 
 
 1 
 
 697 
 
 663 
 
 19 
 6 
 
 582 
 
 6 95 
 
 3 
 
 69s 
 
 373 
 318 
 30S 
 S>6 
 
 3 6 3 
 612 
 
 124 
 
 1 24 
 
 '-5 
 125 
 679 
 687 
 
 38S 
 
 669 
 661 
 
 677 
 
 681 
 683 
 
 5 6 4 
 661 
 
 35 
 329 
 3 2 9 
 697 
 
 5 '6 
 376 
 
 5 2 5 
 386 
 
 274 
 j 2 1 
 
 I [adfield's Esst 1 . 
 
 G. W. Wilson Ltd.. 
 
 Aberdeen. 
 Architectural Society's 
 
 Dictionary. 
 J. J. Cresswell. 
 Brandon's Analysis. 
 
 do. 
 
 do. 
 Johnson's Reliques. 
 Colli ng's Details. 
 
 do. 
 F. Bond. 
 H. Bedford Pim. 
 F. Bond. 
 
 Parker's Glossary. 
 Rickman's Styles. 
 F. Bond. 
 
 Sharpe's Parallels. 
 
 do. 
 do. 
 do. 
 do. 
 do. 
 do. 
 do. 
 do. 
 
 do 
 
 iS do. 
 
 1 do. 
 
 Brandon's Roofs. 
 1 1 Britton's Cathedrals. 
 4 do. 
 
 F. Bond, 
 do. 
 21 Colling's Details. 
 
 Parker's Glossary. 
 F. Bond. 
 
 do. 
 Bloxam's Principles. 
 R. P. Brereton. 
 
 Parker's Churches. 
 
712 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 No. 
 
 in 
 
 Page. 
 
 Source. 
 
 
 Bucknell, Oxford- 
 
 Plan ----- 
 
 2I 5 
 
 6 
 
 Parker's Churches. 
 
 shire 
 
 
 
 
 
 
 Buildwas abbey, 
 
 Diagonal rib in chapter house - 
 
 673 
 
 II 
 
 Potter's Buildwas. 
 
 
 Shropshire 
 
 
 
 
 
 
 Do. 
 
 Transverse rib in chapter house 
 
 673 
 
 IO 
 
 do. 
 
 
 Do. 
 
 Vault of chapter house - 
 
 326 
 
 5 
 
 do. 
 
 
 Do. 
 
 Molded capital 
 
 685 
 
 20 
 
 Sharpe's Parallels. 
 
 
 Burford, Oxford- 
 
 Exterior from south ; Norman 
 
 206 
 
 
 Maurice B. Adams. 
 
 
 shire 
 
 central tower 
 
 
 
 
 
 Do. 
 
 Plan ----- 
 
 217 
 
 1 
 
 do. 
 
 
 Bury St Edmunds- 
 
 Plan ----- 
 
 r 5° 
 
 3 
 
 Dr M. R. James. 
 
 
 Byland abbey, 
 
 Arch-mold - 
 
 667 
 
 4 
 
 Sharpe's Parallels. 
 
 
 Yorkshire 
 
 
 
 
 
 
 Do. 
 
 Molded capital 
 
 685 
 
 24 
 
 do. 
 
 
 Do. 
 
 Plan of pier - 
 
 661 
 
 3 
 
 do. 
 
 
 Do. 
 
 String of triforium - 
 
 681 
 
 6 
 
 do. 
 
 
 Do 
 
 Transverse and diagonal ribs in 
 aisle 
 
 675 
 
 3 
 
 do. 
 
 
 Do. 
 
 Wall rib in aisle 
 
 675 
 
 4 
 
 do. 
 
 
 Caen, Abbaye-aux- 
 
 Interior from S.W. - 
 
 3 J 9 
 
 
 F. Bond. 
 
 
 Dames 
 
 
 
 
 
 
 Do. 
 
 Vault of S. aisle of nave - 
 
 2 93 
 
 
 do. 
 
 
 Caen, Abbaye-aux- 
 
 Interior from S.W. - 
 
 3 T 9 
 
 
 do. 
 
 
 Hommes 
 
 
 
 
 
 
 Do. 
 
 Rib in nave - 
 
 673 
 
 4 
 
 Bilson's Beginnings. 
 
 
 Do. 
 
 Vault of nave - - - - 
 
 315 
 
 
 F. Bond. 
 
 
 Caen, St Nicholas - 
 
 Capitals of pier of central tower 
 
 421 
 
 2 
 
 do. 
 
 
 Do. 
 
 Vault of choir 
 
 293 
 
 
 do. 
 
 
 Caldicott, Mon- 
 
 Window ----- 
 
 487 
 
 3 
 
 Freeman's Windows. 
 
 
 mouth 
 
 
 
 
 
 
 Cambridge, King's 
 
 Exterior from S.W. 
 
 199 
 
 
 F. Bond. 
 
 
 College chapel 
 
 
 
 
 
 
 Do. 
 
 Interior from E. (collotype) face 
 
 62 
 
 
 do. 
 
 
 Do. 
 
 Tracery - - - - - 
 
 473 
 
 
 do. 
 
 
 Do. 
 
 Vault ----- 
 
 333 
 
 
 do. 
 
 
 Canterbury cathe- 
 
 Ambulatory of eastern crypt - 
 
 334 
 
 
 do. 
 
 
 dral 
 
 
 
 
 
 
 Do. 
 
 Arcading of south aisle of 
 Ernulf's choir 
 
 43° 
 
 
 do. 
 
 
 Do. 
 
 Base of retrochoir, 1178 - 
 
 45i 
 
 
 Parker's Glossary. 
 
 
 Do. 
 
 Capital in Ernulf's crypt - 
 
 4i7 
 
 7 
 
 F. Bond. 
 
 
 Do. 
 
 Capital in eastern crypt - 
 
 44° 
 
 1 
 
 do. 
 
 
 Do. 
 
 Capitals in St Thomas' chapel - 
 
 423 
 
 7 
 
 do. 
 
 
 Do. 
 
 do. - 
 
 428 
 
 
 S. Gardner. 
 
 
 Do. 
 
 Central tower from S.E. - 
 
 588 
 
 
 F. Bond. 
 
 
 Do. 
 
 Interior of choir from S.E. 
 
 S 2 3 
 
 
 do. 
 
 
 Do. 
 
 Interior of choir 
 
 106 
 
 
 Sir G. G. Scott. 
 
 
 Do. 
 
 Interior of St Thomas' chapel - 
 
 107 
 
 
 W. S. Weatherley. 
 
 
 Do. 
 
 Internal elevation of one bay 
 of the nave 
 
 90 
 
 
 Willis' Canterbury. 
 
 
INDEX TO ILLUSTRATIONS. 
 
 7'3 
 
 Name of Place. 
 
 Description. 
 
 ['age. 
 
 No. 
 
 in 
 
 Sunn e. 
 
 
 
 ">:-, 
 
 Page. 
 
 S. ( iardner. 
 
 Canterbury cathe- 
 dral 
 Do. 
 
 Interior of Ernulf's crypt - 
 
 Plan of church in 1130 - 
 
 149 
 
 2 
 
 Si oil's Essay. 
 
 Do. 
 
 Plan of church in 1184 
 
 149 
 
 3 
 
 do. 
 
 Do. 
 
 Plan of pier of nave 
 
 661 
 
 10 
 
 Willis' ( 'anterbury. 
 
 Do. 
 
 Transverse section of choir 
 
 34 
 
 3 
 
 do. 
 
 Do. 
 
 Rib in treasury 
 
 673 
 
 2 
 
 Bilson's J'n^i 'linings. 
 
 Canterbury, St 
 
 Capital in nave 
 
 44 6 
 
 
 F. Bond. 
 
 Alphege 
 
 
 
 
 
 Canwick, Lincoln- 
 
 Abacus - 
 
 6S5 
 
 1 
 
 Sharpe's Lincoln. 
 
 shire 
 
 
 
 
 
 Capel St Mary, 
 
 Roof of nave - 
 
 555 
 
 
 Brandon's Roofs. 
 
 Suffolk 
 
 
 
 
 
 Carlisle cathedral - 
 
 East front (collotype) - face 
 
 12S 
 
 
 G. W. Wilson Ltd., 
 Aberdeen. 
 
 Do. 
 
 Do. 
 Do. 
 
 Cassington, Ox- 
 fordshire 
 Castle Rising, 
 
 Norfolk 
 
 Cawston, Norfolk 
 
 Do. 
 
 Do. 
 
 Cerisy-la-Foret, 
 
 Manche 
 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 
 Chartham, Kent - 
 Checkendon, Ox- 
 fordshire 
 Chelmsford - 
 
 Do. 
 
 Do. 
 Chester cathedral - 
 
 Do. 
 
 Do. 
 
 Do. 
 Chester, St John - 
 
 East window; setting out of the 507 
 
 tracery 
 
 East window; jointing of tracery 507 
 
 Interior of N. aisle of choir 1 498 
 
 from W. 
 
 Plan - 215 
 
 Interior from W. 162 
 
 Ground-course of tower - 10; 
 
 Interior from S.W. - 552 
 
 Towei from S.W. - 589 
 
 Plan. ("Zr? //('/ devait avoir 148 
 
 six trav'ees dans sa longueur, 
 
 en tout 257 pieds. La facade 
 
 et ks premieres travees de la 
 
 nef out ite detruites en 182Q." 
 
 — Ruprich-Robert, ii. 3.) 
 
 Exterior from S.E. - - - 160 
 
 Interior of nave from N.E. 521 
 
 (The triforium arcade has 
 
 been blocked up; the vault 
 
 is of lath and plaster.) 
 
 Interior of choir from W. 1 u 1 
 
 E. side of S. transept from N.YV. 199 
 
 Window- of chancel - - 477 
 
 Interior of apse 
 
 Arch-mold north of chancel - 671 
 
 Molded cap of shaft in porch - 693 
 
 Molded cap of pier in chancel 693 
 
 Interior of N. transept from S.W. 524 
 
 Interior of presbytery from S.W. 524 
 
 Exterior of N. aisle, nave, and 5; 
 
 cloister from N. 
 
 Plan of vault of chapter house - 324 
 
 Base and plinth 44S 
 
 hillings C arb.ch 
 Cathedral. 
 do. 
 do. 
 
 Parker's Churches. 
 
 F. Bond. 
 
 do. 
 
 do. 
 Ruprich-Robert's 
 L' Architecture 
 Normande. 
 
 F. Bond. 
 
 do. 
 
 do. 
 
 do. 
 H. Bedford I'im. 
 do. 
 
 Hadfield's Essex. 
 
 do. 
 
 do. 
 F. Bond. 
 
 do. 
 
 do. 
 
 do. 
 do. 
 
714 
 
 INDEX TO ILLUSTRATIONS. 
 
 Name of Place. 
 
 Chichester cathedral 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Chipping Hill, 
 
 Essex 
 Do. 
 Chipping Norton, 
 
 Oxford 
 Do. 
 Christ Church, 
 
 Hants 
 Cirencester, 
 
 Gloucestershire 
 Do. 
 Clare church, 
 
 Suffolk 
 Clermont-Ferrand 
 
 cathedral, Puy-de- 
 
 Dome 
 Cley church, 
 
 Norfolk 
 Cluny, Saone-et- 
 
 Loire 
 Do. 
 
 Codford St Mary, 
 
 Wilts 
 Coggeshall, Essex 
 
 Colchester, Essex 
 
 Colchester, Mag- 
 dalen chapel 
 Do. 
 
 Conisborough 
 castle, Yorkshire 
 
 Cottingham, York- 
 shire 
 Do. 
 
 Coventry, St 
 Michael 
 
 Deeping St James 
 Do. 
 
 Description. 
 
 Transverse section of nave ; on 
 the left before the fire of 
 1 1 86, on the right after the 
 fire 
 
 Vault and clerestory of nave 
 
 Triforium of nave from VV. 
 
 Clerestory passage of S. nave 
 from VV. 
 
 Pier and capitals in retrochoir - 
 
 Arch-mold in west of nave after 
 fire of 1 1 14 
 
 Molded cap of pier of nave 
 
 do. enlarged 
 Interior of nave from S.W. 
 
 Window - - - - - 
 
 Reredos of high altar 
 
 Interior of nave from S.W. 
 
 Base in nave - 
 
 Jamb, arch, and abaci of west 
 
 doorway 
 Interior of choir from N.W. 
 
 Doorway 
 
 South exterior, from model in 
 
 Cluny museum 
 Plan of church 
 
 Jamb and arch of chancel arch 
 
 Left, pier, arch, abacus ; Right, 
 
 pier, base, plinth 
 Base of pier - 
 Jamb and base of doorway 
 
 Arch of do. 
 Fireplace 
 
 Buttress - 
 
 String - 
 Spire 
 
 Page. 
 
 No. 
 
 in 
 
 Page. 
 
 Molded capital 
 do. 
 
 34 
 
 3i3 
 
 520 
 
 544 
 
 245 
 666 
 
 691 
 
 444 
 548 
 
 487 
 180 
 
 544 
 
 7°5 
 526 
 
 35 
 168 
 
 150 
 
 703 
 
 663 
 
 697 
 7°3 
 
 7°3 
 i 261 
 
 357 
 
 683 
 63S 
 
 685 
 685 
 
 Sou 
 
 Willis' Chichester. 
 
 F. Bond, 
 do. 
 do. 
 
 do. 
 
 Sharpe's New Shore- 
 ham. 
 Hadfield's Essex. 
 
 do. 
 H. Mace. 
 
 Freeman's Windows. 
 F. R. Taylor. 
 
 F. Bond. 
 
 do. 
 Colling's Details. 
 
 F. Bond. 
 
 Bowman & Crowther's 
 
 Churches. 
 F. Bond. 
 
 Walcott's Church 
 
 Arrangement. 
 Brandon's Analysis. 
 
 Hadfield's Essex. 
 
 Paley's Moldings. 
 Hadfield's Essex. 
 
 do. 
 
 Colling's Details. 
 
 do. 
 J. Oldrid Scott, 
 Maurice B.Adams, 
 and J. Drayton 
 Wyatt. 
 Nene Valley. 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 7'5 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 685 
 
 in 
 Page. 
 
 27 
 
 Source. 
 
 Deeping St James 
 
 Molded capital 
 
 Vene Valley. 
 
 Denford, Northants 
 
 Spire 
 
 6 14 
 
 
 Johnson's Religues. 
 
 Desborough, do. 
 
 do. 
 
 nil, 
 
 
 (1.,. 
 
 DOUDINGHURST 
 
 Molded cap of shaft in doorway 
 
 689 
 
 1 
 
 Hadlield's Essex. 
 
 Dore abbey, Here- 
 
 Plan of church 
 
 '5' 
 
 4 
 
 ( '. Henman and 
 
 fordshire 
 
 
 
 
 Roland VV. Paul. 
 
 Do. 
 
 Exterior from east - 
 
 1S2 
 
 
 K. Bond. 
 
 Do. 
 
 Capitals in choir aisle 
 
 412 
 
 S 
 
 do. 
 
 Do. 
 
 Capital in do. 
 
 417 
 
 5 
 
 do. 
 
 Do. 
 
 Capitals of do. 
 
 421 
 
 3 
 
 do. 
 
 Do. 
 
 Capital in retrochoir 
 
 422 
 
 8 
 
 do. 
 
 Do. 
 
 do. 
 
 422 
 
 9 
 
 do. 
 
 Do. 
 
 do. 
 
 422 
 
 1 
 
 do. 
 
 Dublin, St Patrick's 
 
 Buttress - - - - - 
 
 354 
 
 
 Colli ng's Details. 
 
 Durham cathedral - 
 
 Plan of Norman church - 
 
 U9 
 
 1 
 
 Billings' Durham and 
 John Bilson. 
 
 Do. 
 
 Transverse section of nave 
 
 34 
 
 2 
 
 Billings' Durham. 
 
 Do. 
 
 Interior of S. aisle of choir 
 from east 
 
 306 
 
 
 J. P. freeman. 
 
 Do. 
 
 Vault of S aisle of choir - 
 
 3i5 
 
 
 do. 
 
 Do. 
 
 Plan of compound pier in choir 
 
 659 
 
 1 
 
 Bilson's Beginnings. 
 
 Do. 
 
 Interior of nave from S.E. 
 
 8 
 
 
 I. P. freeman. 
 
 Do. 
 
 Plan of vaults in nave 
 
 308 
 
 2 
 
 F. Bond. 
 
 Do. 
 
 N. aisle of nave from E. - 
 
 ^39 
 
 
 do. 
 
 Do. 
 
 Interior of triforium of choir - 
 
 37° 
 
 
 do. 
 
 Do. 
 
 do. nave 
 
 37° 
 
 
 do. 
 
 Do. 
 
 Exterior from N.W. 
 
 2S 
 
 
 J. R. Edis, Durham. 
 
 Do. 
 
 Window in E transept - 
 
 514 
 
 
 Billings' Durham. 
 
 Do. 
 
 Molded cap in east transept 
 
 689 
 
 3 
 
 do. 
 
 Do. 
 
 Vault of prior's kitchen - 
 
 33S 
 
 
 do. 
 
 Do. 
 
 Plan of vault of prior's kitchen 
 
 3^4 
 
 10 
 
 F. Bond. 
 
 Easington, Durham 
 
 Capital ----- 
 
 42S 
 
 
 C. C. Hodges. 
 
 East Thorpe 
 
 Molded cap of sedilia 
 
 6S9 
 
 
 Hadfield's Essex. 
 
 Eastbourne, St Mary 
 
 Plan, with late sacristy at E. 
 end 
 
 216 
 
 5 
 
 W. Budgen. 
 
 Eastwood, Essex - 
 
 Exterior from N.W. 
 
 224 
 
 
 J. S. Collings. 
 
 Eltham palace 
 
 Molded cap of shaft in doorway 
 
 693 
 
 5 
 
 Pugin's Examples. 
 
 Ely cathedral - 
 
 Plan - - - - 
 
 153 
 
 4 
 
 Walcott. 
 
 Do. 
 
 Transverse section of nave 
 
 34 
 
 1 
 
 E. Hardy. 
 
 Do. 
 
 Internal elevation of one bay 
 of the nave 
 
 57 
 
 
 do. 
 
 Do. 
 
 Interior of transept 
 
 273 
 
 
 F. Bond. 
 
 Do. 
 
 do. 
 
 506 
 
 
 do. 
 
 Do. 
 
 Plan of pier in west aisle of 
 north transept 
 
 659 
 
 2 
 
 Bilson's Beginnings. 
 
 Do. 
 
 Capital in eastern aisle of cen- 
 tral transept 
 
 412 
 
 1 
 
 F. Bond. 
 
 Do. 
 
 West doorway from the cloister 
 into the south aisle of the 
 nave 
 
 39 
 
 
 Alan Potter. 
 
716 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Ely cathedral - 
 
 Leaf scrolls of east processional 
 doorway of nave 
 
 43° 
 
 
 Colling's Foliage. 
 
 Do. 
 
 Cusp of doorway, page 430 
 
 5io 
 
 7 
 
 Parker's Glossary. 
 
 Do. 
 
 West tower from S.W. 
 
 587 
 
 
 F. Bond. 
 
 Do. 
 
 Presbytery ; interior elevation 
 of three bays 
 
 117 
 
 
 C. Spooner. 
 
 Do. 
 
 Interior of presbytery from S.E. 
 
 526 
 
 
 F. Bond. 
 
 Do. 
 
 Vault of presbytery 
 
 3 2 7 
 
 
 do. 
 
 Do. 
 
 Interior of retrochoir from N.E. 
 
 247 
 
 
 G. H. Tyndall, Ely. 
 
 Do. 
 
 Piers of choir and presbytery - 
 
 251 
 
 
 S. K. Greenslade. 
 
 Do. 
 
 Corbel in presbytery 
 
 241 
 
 
 G. H. Tyndall, Ely. 
 
 Do. 
 
 Exterior of presbytery from S.W. 
 
 526 
 
 
 F. Bond. 
 
 Do. 
 
 East front - 
 
 464 
 
 
 do. 
 
 Do. 
 
 Interior of central lantern from 
 
 S.E. 
 
 45 
 
 
 E. Gunn. 
 
 Do. 
 
 Interior of choir from S.W. 
 
 522 
 
 
 F. Bond. 
 
 Do. 
 
 Vault of choir 
 
 3 2 9 
 
 
 do. 
 
 Do. 
 
 Pedestal of shrine of St Ethel- 
 dreda (Sir G. Scott) ; may be 
 the canopy of a tomb 
 
 S 11 
 
 
 E. W. M. Wonnacott. 
 
 Do. 
 
 Pinnacles of choir - 
 
 3 6 5 
 
 
 F. Bond. 
 
 Do. 
 
 Arcading of Lady chapel 
 
 269 
 
 
 do. 
 
 Do. 
 
 PriorCrauden'schapel, interior of 
 
 130 
 
 
 W. H. Bidlake. 
 
 Do. 
 
 Doorway of Bishop West's 
 chapel 
 
 i43 
 
 
 G. H. Tyndall, Ely. 
 
 Do. 
 
 Vault of Bishop West's chapel - 
 
 334 
 
 
 F. Bond. 
 
 Ensham, Oxford 
 
 Cornice ----- 
 
 385 
 
 5 
 
 Parker's Glossary. 
 
 Etchingham, Sussex 
 
 Capital in nave 
 
 440 
 
 6 
 
 F. Bond. 
 
 Etton, Northants - 
 
 Spire . - . - - 
 
 620 
 
 
 Brandon's Churches. 
 
 Evercreech, Somer- 
 set 
 Evreux cathedral, 
 
 Tower - 
 
 595 
 
 
 F. J. Allen. 
 
 Interior of choir from S.W. 
 
 539 
 
 
 F. Bond. 
 
 Eure 
 
 
 
 
 
 Ewerby, Lincolnshire 
 
 Spire 
 
 630 
 
 
 Bowman & Crowther's 
 Churches. 
 
 Exeter cathedral - 
 
 Plan - - - - 
 
 154 
 
 4 
 
 Britton's Cathedrals. 
 
 Do. 
 
 Transverse section of choir 
 
 35 
 
 3 
 
 do. 
 
 Do. 
 
 Interior of nave from W. 
 
 9 
 
 
 F. J. Allen. 
 
 Do. 
 
 Plan of vault of nave (two 
 bays) 
 
 3 2 4 
 
 7 
 
 F. Bond. 
 
 Do. 
 
 Plan of pier of nave 
 
 661 
 
 9 
 
 Harbottle Reed. 
 
 Do. 
 
 Pier and corbel in choir - 
 
 241 
 
 
 F. Bond. 
 
 Do. 
 
 Boss in choir aisle - 
 
 297 
 
 
 do. 
 
 Do. 
 
 East window of Lady chapel - 
 
 473 
 
 
 do. 
 
 Do. 
 
 Flying buttresses of choir 
 from E. 
 
 377 
 
 
 do. 
 
 Fairstead, Essex - 
 
 Molded cap of sedilia 
 
 689 
 
 10 
 
 Hadfield's Essex. 
 
 Faringdon, Berks - 
 
 Roof - - - 
 
 558 
 
 IV 
 
 Parker's Glossary. 
 
 Feering, Essex 
 
 Two mullions - 
 
 SIS 
 
 
 Hadfield's Essex. 
 
 Finedon, Northants 
 
 Pier, arch, abacus - 
 
 665 
 
 9 
 
 Nene Valley. 
 
INDEX TO 
 
 LUSTRA! K )\S. 
 
 717 
 
 Name of Place. 
 
 1 Ascription. 
 
 Page. 
 
 No. 
 
 in 
 
 Soun e. 
 
 
 
 
 Page. 
 
 F. Bond. 
 
 Florence cathedral 
 
 Plan of nave vault - 
 
 308 
 
 
 
 Florence, S. 
 
 Diaphragmatic roof system 
 
 2 35 
 
 
 Willis' Middle Ages. 
 
 Miniato 
 
 
 
 
 
 FOTHERINGHAY 
 
 Arch and abacus of west door- 
 way 
 
 70/ 
 
 1, 
 
 Nene Valley. 
 
 Do. 
 
 Jamb and bases of west door- 
 way 
 
 7°7 
 
 7 
 
 do. 
 
 Fountains abbey, 
 
 Flan of church 
 
 '5° 
 
 2 
 
 Sharpe's Parallels. 
 
 Yorkshire 
 
 
 
 
 
 Do. 
 
 Interior of nave from S.E. 
 
 IOI 
 
 
 do. 
 
 Do. 
 
 Base ----- 
 
 695 
 
 7 
 
 do. 
 
 Do. 
 
 Molded capital in nave ■ 
 
 685 
 
 21 
 
 do. 
 
 Do. 
 
 Ground course of nave 
 
 679 
 
 1 
 
 do. 
 
 Do. 
 
 Parapet of nave 
 
 3«5 
 
 6 
 
 do. 
 
 Do. 
 
 String of aisle of nave 
 
 6S1 
 
 1 
 
 do. 
 
 Do. 
 
 String of clerestory of nave 
 
 6S1 
 
 2 
 
 do. 
 
 Do. 
 
 Molded cap and base in choir 
 
 6S7 
 
 2 
 
 do. 
 
 Do. 
 
 Transverse and diagonal ribs in 
 aisle of choir 
 
 677 
 
 1 
 
 do. 
 
 Do. 
 
 Wall rib in aisle of choir - 
 
 677 
 
 2 
 
 do. 
 
 Do. 
 
 Ground-course of choir - 
 
 679 
 
 4 
 
 do. 
 
 Do. 
 
 String of aisle - - - - 
 
 6S1 
 
 14 
 
 do. 
 
 Frampton, Lincoln- 
 
 Exterior from S.E. - 
 
 61S 
 
 
 W. H. Bidlake. 
 
 shire 
 
 
 
 
 
 Do. 
 
 Buttress - 
 
 357 
 
 
 Colling's Details. 
 
 Do. 
 
 Dripstone 
 
 683 
 
 13 
 
 do. 
 
 Fulbeck 
 
 Abacus - 
 
 6S S 
 
 7 
 
 Sharpe's Lincoln. 
 
 Furness abbey, 
 
 Base 
 
 695 
 
 6 
 
 Sharpe's Parallels. 
 
 Lancashire 
 
 
 
 
 
 Do. 
 
 Molded capital 
 
 685 
 
 23 
 
 do. 
 
 Do. 
 
 String of aisle of nave 
 
 6S1 
 
 3 
 
 do. 
 
 Do. 
 
 Transverse rib in aisle of nave 
 
 673 
 
 7 
 
 do. 
 
 Do. 
 
 Diagonal rib in aisle of nave - 
 
 673 
 
 S 
 
 do. 
 
 Do. 
 
 Wall rib in aisle of nave - 
 
 673 
 
 9 
 
 do. 
 
 Gednev, Lincoln- 
 
 Interior of nave from S.E. 
 
 5Si 
 
 
 F. Bond. 
 
 shire 
 
 
 
 
 
 Do. 
 
 Chancel window 
 
 474 
 
 1 
 
 do. 
 
 Gimingham, Norfolk 
 
 S. porch 
 
 57" 
 
 
 do. 
 
 Glastonbury abbey, 
 
 Bay design 
 
 536 
 
 
 Gerald Horsley. 
 
 Somerset 
 
 
 
 
 
 Do. 
 
 West front of Lady chapel 
 
 465 
 
 
 F. Bond. 
 
 Glastonbury, 
 
 Arch-mold of doorway 
 
 67 1 
 
 4 
 
 Pugin's Examples. 
 
 George Inn 
 
 
 
 
 
 Do 
 
 Two mullions - 
 
 515 
 
 
 do. 
 
 Glastonbury, St 
 
 West tower from S.E. 
 
 588 
 
 
 F. Bond. 
 
 John 
 
 
 
 
 
 Gloucester cathe- 
 
 Exterior from S.E. (collotype) 
 
 
 
 
 dral 
 
 face 
 
 132 
 
 
 W. T. Alston. 
 
 Do. 
 
 Plan of crypt - - - - 
 
 192 
 
 T 
 
 Britton's Cathedrals. 
 
7 i8 
 
 INDEX TO ILLUSTRATIONS. 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 No. 
 
 in 
 
 Page. 
 
 Source. 
 
 Gloucester cathe- 
 dral 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Grantham 
 Great Baddow 
 Great Chalfield, 
 
 Wilts. 
 Do. 
 Great Gonerby, 
 
 Lincolnshire 
 Gresford, Cheshire 
 Grimsby 
 Guildford, St Mary 
 
 Guisborough abbey, 
 Yorkshire 
 
 Rib in crypt 
 
 Norman portions from S.E ; 
 with N.E. apsidal chapel and 
 interior elevation of Norman 
 choir in process of transfor- 
 mation 
 
 Exterior of N.E. apse 
 
 N. ambulatory of choir and 
 tomb of Edward II. 
 
 Vault of S. ambulatory of choir 
 
 Upper south aisle of choir 
 from W. 
 
 Interior of nave from W. (collo- 
 type) - face 
 
 Interior of N. aisle of nave 
 from W. 
 
 Vault of N. aisle of nave - 
 
 Exterior of S. aisle of nave 
 
 Western doorway into the clois- 
 ter from the north aisle of 
 the nave 
 
 Boss in west nave (Coronation 
 of the Virgin) 
 
 Interior of S. transept from N. 
 
 Vault of S. transept 
 
 Transverse section of choir 
 
 Interior elevation of three bays 
 of the choir 
 
 N. clerestory of choir 
 
 Interior of choir (collotype)/a« 
 
 Interior of north choir and east 
 window 
 
 Vault of choir 
 
 Plan of vault of choir; three 
 bays 
 
 Cloister (collotype) - - face 
 
 Lavatory in cloister 
 
 Cornice ----- 
 
 Molded cap of pier 
 
 Rib in porch of manor house - 
 
 do. 
 Base of pier - 
 
 Interior from N.E. - 
 
 Arch-mold in nave - 
 
 Interior of north aisle of 
 chancel ; two fifteenth cen- 
 tury windows 
 
 Buttresses - 
 
 673 
 !35 
 
 160 
 294 
 
 2 93 
 282 
 
 26 
 99 
 
 3i3 
 360 
 269 
 
 297 
 
 495 
 306 
 
 35 
 59 
 
 492 
 496 
 
 47 
 
 334 
 325 
 
 344 
 334 
 385 
 689 
 677 
 
 677 
 697 
 
 214 
 
 667 
 
 36 
 
 354 
 
 Parker's Glossary. 
 F. S. Waller. 
 
 W. T. Alston, 
 do. 
 
 F. Bond. 
 
 S. A. Pitcher, Glou- 
 cester. 
 
 W. T. Alston, 
 do. 
 
 F. Bond. 
 W. T. Alston, 
 do. 
 
 Oscar Clark. 
 
 W. T. Alston. 
 
 F. Bond. 
 
 Britton's Cathedrals. 
 
 do. 
 
 F. Bond. 
 W. T. Alston, 
 do. 
 
 F. Bond, 
 do. 
 
 W. T. Alston. 
 F. Bond. 
 
 Parker's Glossary. 
 Hadfield's Essex. 
 Pugin's Examples. 
 
 do. 
 Colling's Details. 
 
 F. Bond. 
 J. J. Cresswell. 
 Colling's Gothic 
 Or/ia?ne/its. 
 
 Sharpe's Parallels. 
 
INDEX TO ILLUSTRATIONS 
 
 7 19 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 No. 
 
 in 
 
 Source, 
 
 
 
 476 
 
 Page. 
 
 Sharpe's Parallels. 
 
 Guisbokough abbey, 
 
 East window of choir (restora- 
 
 
 Yorkshire 
 
 tion) 
 
 
 
 
 D,, 
 
 Jamb and bases of window of 
 choir aisle- 
 
 699 
 
 7 
 
 do. 
 
 Do. 
 
 Arch and abaci of do. 
 
 699 
 
 8 
 
 do. 
 
 Do. 
 
 Base of arcading and shaft of 
 umdow 
 
 (.97 
 
 12 
 
 do. 
 
 Do. 
 
 Plan of pier - 
 
 Oh, 
 
 6 
 
 do. 
 
 Do. 
 
 String of aisle - 
 
 683 
 
 4 
 
 do. 
 
 Hadleigh, Essex ■ 
 
 Plan ----- 
 
 2I 5 
 
 1 
 
 Buckler's Essex. 
 
 Hales, Norfolk. 
 
 Exterior from S.E. - 
 
 156 
 
 
 F. Pond. 
 
 Do. 
 
 N. doorway 
 
 575 
 
 
 do. 
 
 Harmston, Lines. 
 
 Abacus - 
 
 68s 
 
 2 
 
 Sharpe's Lincoln. 
 
 Haverfordwest, 
 
 Windows in S. aisle- 
 
 480 
 
 I 
 
 F. Bond. 
 
 Pembroke 
 
 
 
 
 
 Hawton, Notts 
 
 East front 
 
 4*3 
 
 
 H. Bedford Pirn. 
 
 Do. 
 
 Ground-course 
 
 355 
 
 
 ( olling's Details. 
 
 Heckington, 
 
 Pinnacle 
 
 3 6 4 
 
 
 Bowman & Crowther's 
 
 Lincolnshire 
 
 
 
 
 Churches. 
 
 Do. 
 
 Parapet ----- 
 
 38S 
 
 I 
 
 do. 
 
 Do. 
 
 Molded cap in tower 
 
 691 
 
 5 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 Molded cap in chancel - 
 
 691 
 
 6 
 
 do. 
 
 Do. 
 
 Molded cap in nave 
 
 691 
 
 7 
 
 do. 
 
 Hedingham castle, 
 
 Arch-mold 
 
 666 
 
 2 
 
 Hadfield's Essex. 
 
 Essex 
 
 
 
 
 
 Do. 
 
 Base of shaft of window - 
 
 695 
 
 5 
 
 do. 
 
 Hedon, Yorkshire - 
 
 Interior of nave from S.W. 
 
 544 
 
 
 F. Bond. 
 
 Do. 
 
 Exterior of S. aisle of nave 
 from S.W. 
 
 35<5 
 
 
 do. 
 
 Do. 
 
 West window on S. side of nave 
 
 480 
 
 3 
 
 do. 
 
 Helpringham, 
 
 Arch mold 
 
 670 
 
 2 
 
 Sharpe's Lincoln. 
 
 Lincolnshire 
 
 
 
 
 
 Do. 
 
 Molded cap in nave- 
 
 691 
 
 4 
 
 do. 
 
 Hemel Hempstead, 
 
 Exterior from N.W. 
 
 '9 
 
 
 F. Bond. 
 
 Hertfordshire 
 
 
 
 
 
 Hereford cathedral 
 
 Central tower and N. transept 
 from N.W. 
 
 537 
 
 
 do. 
 
 Do. 
 
 N. porch 
 
 203 
 
 
 do. 
 
 Do. 
 
 Interior of Lady chapel - 
 
 464 
 
 
 do. 
 
 Do. 
 
 Interior of N. transept from S.W. 
 
 523 
 
 
 do. 
 
 Do. 
 
 Pedestal of shrine, probably of 
 St Thomas of Hereford 
 
 .87 
 
 
 do. 
 
 Do. 
 
 Vault of S. transept 
 
 333 
 
 
 do. 
 
 Hexham abbey, 
 
 Ground-course of north transept 
 
 679 
 
 2 
 
 Sharpe's Parallels. 
 
 Northumberland 
 
 
 
 
 
 Do. 
 
 Strmg of aisle of choir 
 
 681 
 
 7 
 
 do. 
 
 Do. 
 
 String of triforium of choir 
 
 681 
 
 8 
 
 do. 
 
 Higham Ferrers, 
 
 Doorway- 
 
 534 
 
 
 Brandon's Analysis. 
 
 Northants 
 
 
 
 
 
 Do. 
 
 Plan of pier, base, plinth 
 
 663 
 
 4 
 
 Nene Valley. 
 
■J20 
 
 INDEX TO ILLUSTRATIONS. 
 
 Name of Place. 
 
 Hingham, Norfolk - 
 Holbeach, Lincoln- 
 shire 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Holme, Norfolk 
 Horbling 
 Houghton-le-Dale, 
 
 Norfolk 
 Howden, Yorkshire 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 
 Howell, Lincoln- 
 shire 
 
 Huish Episcopi, 
 Somerset 
 
 Hull, Holy Trinity 
 
 Description. 
 
 Do. 
 
 Do. 
 Do. 
 
 Hunstanton, 
 
 Norfolk 
 Hutton, Essex 
 
 Ickleton, Cambs. - 
 
 Do. 
 Iffley, Oxford 
 
 Do. 
 
 Do. 
 Ilkeston, Derbyshire 
 Ingatestone - 
 Isle Abbots, 
 Somerset 
 
 E. window of S. aisle 
 Spire from S.E. 
 
 Interior from S.E. - 
 
 Door - 
 
 Molded cap of pier in nave 
 
 Base of pier 
 
 Tower 
 
 Abacus - 
 
 Jamb of doorway 
 
 West front 
 
 Interior of nave from S.W. 
 
 Arcading of chapter house 
 
 Transverse and diagonal ribs in 
 
 choir 
 Wall rib in choir 
 String of choir aisle 
 Plan - 
 
 Belfry window 
 
 Interior of chancel. (The east 
 window has been frequently 
 repaired ; the lower transom 
 is not original, nor probably 
 the upper one. Rickman's 
 drawing shows the head of 
 the most lateral light on 
 either side more depressed 
 than at present. Cf. 93.) 
 
 S. exterior of chancel. (A 
 curvilinear window of the 
 fourteenth century.) 
 
 Interior of nave from N.W. 
 
 West front (collotype) - face 
 
 Window in porch - 
 
 Pier, arch, abacus - - - 
 
 Interior of nave 
 
 Doorway 
 
 S. doorway 
 
 West doorway 
 
 Shaft - 
 
 Ground course 
 
 Molded cap of pier in nave 
 
 Tower - 
 
 489 
 6i5 
 
 133 
 
 583 
 691 
 697 
 
 S9i 
 685 
 
 707 
 
 72 
 546 
 137 
 
 677 
 
 677 
 683 
 
 215 
 
 517 
 Sr 
 
 474 
 
 93 
 
 96 
 
 480 
 
 665 
 
 410 
 274 
 577 
 574 
 256 
 
 355 
 693 
 59i 
 
 No. 
 
 in 
 
 Page. 
 
 Source. 
 
 F. Bond. 
 Johnson's Reliques. 
 
 F. Bond. 
 
 Brandon's Analysis. 
 Sharpe's Lincoln. 
 
 do. 
 H. Bedford Pim. 
 Sharpe's Lincoln. 
 Pugin's Examples. 
 
 Sharpe's Parallels. 
 Sydney Vacher. 
 Colling's Gothic 
 
 Ornaments. 
 Sharpe's Parallels. 
 
 do. 
 do. 
 Brandon's Churches. 
 
 Parker's Glossary. 
 
 Rickman's Styles. 
 
 F. Bond. 
 
 H. W. Pigeon. 
 C. C. Hodges. 
 F. Bond. 
 
 Hadfield's Essex. 
 
 H. Bedford Pim. 
 
 do. 
 Colling's Foliage. 
 Colling's Details. 
 Hadfield's Essex. 
 H. Bedford Pim. 
 
INDKX TO ILLUSTRATIONS. 
 
 ■ ■ i 
 
 Name of Place. 
 
 Description. 
 
 Jervaulx abbey, 
 
 Plan of church 
 
 Yorkshire 
 
 
 Do. 
 
 base - 
 
 Do. 
 
 String of aisle - 
 
 Do. 
 
 String ----. 
 
 Do. 
 
 Transverse rib in aisle 
 
 Kells, Ireland 
 
 St Columba's church 
 
 Kenilworth - 
 
 Rib in octagonal lobby - 
 
 Kettering, 
 
 Spire ----- 
 
 Nortbants 
 
 
 Do. 
 
 West doorway 
 
 Do. 
 
 String round spout - 
 
 Do. 
 
 Arch-mold 
 
 Do. 
 
 Ground-course of tower - 
 
 Ketton, Rutland - 
 
 Exterior from S.W. - 
 
 Do. 
 
 Capital in nave 
 
 Kidlington, Oxford 
 
 Plan 
 
 KlLKHAMPTON, 
 
 Interior from N.W. - 
 
 Somerset 
 
 
 King's Sutton, 
 
 Spire ----- 
 
 Northants 
 
 
 Kirkstall abbey, 
 
 Plan of church 
 
 Yorkshire 
 
 
 Do. 
 
 East front of chapter house 
 
 Do. 
 
 Plan of pier of S. aisle of nave 
 
 Do. 
 
 Molded capital 
 
 Do. 
 
 Transverse rib in aisle of nave 
 
 Do, 
 
 Diagonal rib in aisle of nave - 
 
 Knapton, Norfolk - 
 
 Interior of nave from west 
 
 Do. 
 
 Cornice of roof 
 
 Do. 
 
 Hammerbeams of roof - 
 
 Laon cathedral 
 
 Internal elevation of nave 
 
 Lavenham, Suffolk - 
 Do. 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Leadenham, Lin- 
 colnshire 
 Do. 
 Do. 
 Ledbury, Hereford- 
 shire 
 Do. 
 Do. 
 
 Tower from S.W. 
 
 S. porch - 
 
 Exterior of S. aisle - 
 
 Battlement 
 
 Screens of chantry chapel of 
 
 Spring family 
 Capital in nave 
 Plan of pier, base, plinth 
 Arch-mold 
 
 Molded cap in nave 
 Base of pier - 
 Exterior from S.E. - 
 
 Window in St Catherine's chapel 
 Window in S. aisle - 
 
 
 No. 
 
 Page. 
 
 in 
 
 
 Page. 
 
 '53 
 
 3 
 
 695 
 681 
 681 
 
 8 
 
 9 
 10 
 
 6 75 
 
 5 
 
 219 
 
 
 677 
 623 
 
 17 
 
 95 
 
 
 401 
 67 1 
 679 
 621 
 
 3 
 
 8 
 
 440 
 2 16 
 
 2 
 
 6 
 
 56S 
 
 
 6i5 
 
 
 152 
 
 4 
 
 359 
 661 
 
 1 
 
 685 
 673 
 673 
 567 
 5 6 7 
 567 
 
 2 2 
 
 5 
 6 
 
 3 
 1 
 
 528 
 
 
 5§9 
 576 
 365 
 387 
 178 
 
 4 
 
 446 
 663 
 670 
 
 7 
 1 
 
 691 
 697 
 612 
 
 8 
 14 
 
 474 
 
 4 
 
 474 
 
 2 
 
 Source. 
 
 Sharpe's Parallels, 
 
 do. 
 do. 
 do. 
 do. 
 
 Pugin's Examples. 
 R. P. BrerUon. 
 
 Billings' Kettering. 
 do. 
 do. 
 do. 
 
 F. Bond. 
 
 do. 
 Parker's Churches. 
 
 G. K. Saunders. 
 
 Johnson's Reliques. 
 
 Sharpe's Parallels. 
 
 F. Bond. 
 Sharpe's Parallels. 
 
 do. 
 
 do. 
 
 do. 
 W. Davidson. 
 
 do. 
 
 do. 
 
 R. J. Johnson's Early 
 French Architecture. 
 F. Bond. 
 
 do. 
 
 do. 
 Brandon's Analysis. 
 F. Bond. 
 
 do. 
 Brandon's Analysis. 
 Sharpe's Lincoln. 
 
 do. 
 do. 
 F. Bond. 
 
 do. 
 
 do. 
 
722 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Leicester, St Mary 
 
 Roof ----- 
 
 558 
 
 II. 
 
 Bloxam's Principles. 
 
 Leicester St Nicholas 
 
 Interior of nave from S.W. 
 
 213 
 
 
 F. Bond. 
 
 Le Puy, Haute-Loire 
 
 Capital in cloister - 
 
 417 
 
 3 
 
 do. 
 
 Lessay, Manche 
 
 Interior from S.W. - 
 
 319 
 
 
 do. 
 
 Do. 
 
 Interior of choir from S.E. 
 
 3*9 
 
 
 do 
 
 Do. 
 
 Vault of choir 
 
 3 r 5 
 
 
 do. 
 
 Do. 
 
 Plan of vault of choir 
 
 308 
 
 3 
 
 do. 
 
 Do. 
 
 Vault of aisle of nave 
 
 2 93 
 
 
 do. 
 
 Do. 
 
 Capitals in west nave 
 
 412 
 
 3 
 
 do. 
 
 Leverington, Cam- 
 
 South porch (1884) 
 
 84 
 
 
 H. Needham Wilson. 
 
 bridge 
 
 
 
 
 
 Lichfield cathedral 
 
 Exterior from S.W. (collotype) - 
 
 Frontis 
 
 piece 
 
 H. Bedford Pim. 
 
 Do. 
 
 Interior of nave from S.W. 
 
 5 2 3 
 
 
 W. T. Alston. 
 
 Do. 
 
 Vault of west bays of nave 
 
 3 J 3 
 
 
 F. Bond. 
 
 Do. 
 
 Plan of vault of nave, two bays 
 
 3 2 4 
 
 9 
 
 do. 
 
 Do. 
 
 Capitals in nave 
 
 424 
 
 8 
 
 do. 
 
 Do. 
 
 Vault of S. transept 
 
 32S 
 
 
 do. 
 
 Do. 
 
 Plan of vault of S. transept 
 
 3 2 4 
 
 5 
 
 do. 
 
 Do. 
 
 Interior of choir from S.W. 
 
 244 
 
 
 T. Gough. 
 
 Do. 
 
 Plan of pier of choir 
 
 661 
 
 8 
 
 Scott's Lectures. 
 
 Do. 
 
 South of Lady chapel and choir 
 
 369 
 
 
 — . Bullock, Lichfield 
 
 Lincoln minster 
 
 Plan, c. 1250 - 
 
 !5 r 
 
 1 
 
 Beresford Pite 
 
 Do. 
 
 Exterior from N.E. (collotype) 
 
 
 
 
 
 face 
 
 44 
 
 
 C. C. Hodges. 
 
 Do. 
 
 Exterior from S.W. (collotype) 
 
 
 
 
 
 face 
 
 562 
 
 
 Alan Potter. 
 
 Do. 
 
 Exterior elevation of choir 
 
 1 10 
 
 
 Sharpe's Periods. 
 
 Do. 
 
 Exterior of S.E. transept from 
 S.E. 
 
 66 
 
 
 S. K. Greenslade. 
 
 Do. 
 
 Exterior of S.E. and S. transepts 
 from S.E. 
 
 1 12 
 
 
 T. Gough. 
 
 Do. 
 
 Nave, restoration of, exterior of 
 
 n<; 
 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 Exterior of N. transept and 
 cloister from N.E. 
 
 69 
 
 
 Alan Potter. 
 
 Do. 
 
 Presbytery, east front 
 
 177 
 
 
 J. P. Gibson, 
 Hexham. 
 
 Do. 
 
 Exterior of S. transept. (The 
 lancets belong to the work of 
 the first years of the thir- 
 teenth century.) 
 
 484 
 
 
 Alan Potter. 
 
 Do. 
 
 Transverse section of choir 
 
 34 
 
 5 
 
 Dehio and von 
 Bezold's Kirchen. 
 
 Do. 
 
 Interior elevation of choir 
 
 in 
 
 
 Sharpe's Periods. 
 
 Do. 
 
 Vault of choir 
 
 3 2 7 
 
 
 F. Bond. 
 
 Do. 
 
 Plan of vault of choir 
 
 324 
 
 2 
 
 do. 
 
 Do. 
 
 Pier in eastern transept - 
 
 249 
 
 
 Parker's Glossary. 
 
 Do. 
 
 Capital in arcading of choir 
 
 422 
 
 2 
 
 F. Bond. 
 
 Do. 
 
 do. 
 
 423 
 
 9 
 
 do. 
 
 Do. 
 
 Abacus - 
 
 685 
 
 12 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 do. 
 
 685 
 
 13 
 
 do. 
 
 Do. 
 
 do. 
 
 685 
 
 17 
 
 do. 
 
 Do. 
 
 do. 
 
 685 
 
 18 
 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 7^3 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 No. 
 
 in 
 
 Source. 
 
 
 
 
 Page. 
 
 
 Lincoln 1 minster 
 
 Abacus - - - - - 
 
 685 
 
 J 9 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 Interior of S. transept from 
 N.W. 
 
 199 
 
 
 F. Bond. 
 
 Do. 
 
 Sexpartite vault of south transept 
 
 30S 
 
 5 
 
 do. 
 
 Do. 
 
 Vault of nave - - - - 
 
 3 2 7 
 
 
 do. 
 
 Do. 
 
 Plan of vault of nave 
 
 3 2 4 
 
 4 
 
 do. 
 
 Do. 
 
 Plan of vault of aisles of nave - 
 
 308 
 
 8 
 
 do. 
 
 Do. 
 
 Vault of S.W. chapel of nave - 
 
 3 28 
 
 
 do. 
 
 Do. 
 
 Plan of vault of S.W. and N.W. 
 chapels of nave 
 
 308 
 
 9 
 
 do. 
 
 Do. 
 
 Arch-mold in nave 
 
 66S 
 
 2 
 
 W. G. Watkins. 
 
 Do. 
 
 Interior of chapter house 
 
 3-4° 
 
 
 A. Foster. 
 
 Do. 
 
 Plan of vault of chapter house - 
 
 3 2 5 
 
 1 
 
 do. 
 
 Do. 
 
 Rib in galilee 
 
 677 
 
 3 
 
 Sharpe's Lincoln. 
 
 Do. 
 
 Vault of central tower 
 
 328 
 
 
 F. Bond. 
 
 Do. 
 
 Interior elevation of one bay of 
 the presbytery 
 
 56 
 
 
 Sydney Vacher. 
 
 Do. 
 
 Vault of presbytery 
 
 3-7 
 
 
 F. Bond. 
 
 Do. 
 
 Plan of vault of presbytery 
 
 3 2 4 
 
 6 
 
 do. 
 
 Do. 
 
 Arch-mold in presbytery- 
 
 669 
 
 2 
 
 W. O. Watkins. 
 
 Do. 
 
 Molded cusp in presbytery 
 
 5 IQ 
 
 1 
 
 Sharpe's Windows. 
 
 Do. 
 
 Foliated cusp in presbytery 
 
 5 IQ 
 
 2 
 
 Parker's Glossary. 
 
 Do. 
 
 Arcading under S.W. tower 
 
 269 
 
 
 F. Bond. 
 
 '. Lincoln, St Mary-le- 
 
 Capital in nave 
 
 4 2 3 
 
 6 
 
 do. 
 
 Wigford 
 
 
 
 
 
 Do. 
 
 Abacus - 
 
 6S5 
 
 14 
 
 Sharpe's Lincoln. 
 
 Little Addington, 
 
 Window ----- 
 
 4S7 
 
 1 
 
 Churches of North- 
 
 Northants 
 
 
 
 
 ants. 
 
 Little Wenham, 
 
 buttress - 
 
 35 2 
 
 
 Brandon's Analysis. 
 
 Suffolk 
 
 
 
 
 
 Llanbelig, Car- 
 
 Stoup inside N. doorway - 
 
 204 
 
 
 F. Bond. 
 
 narvon 
 
 
 
 
 
 Llandaff cathedral 
 
 Eastern arch of presbytery 
 
 5S0 
 
 
 do. 
 
 Do. 
 
 Capital in nave 
 
 4 2 4 
 
 5 
 
 do. 
 
 Do. 
 
 do. 
 
 424 
 
 6 
 
 do. 
 
 Llanidloes, Mont- 
 
 Interior from S.W. - 
 
 55 2 
 
 
 do. 
 
 gomery 
 
 
 
 
 
 Do. 
 
 Capital in nave 
 
 422 
 
 4 
 
 do. 
 
 Do. 
 
 do. 
 
 422 
 
 5 
 
 do. 
 
 Do. 
 
 do. 
 
 422 
 
 
 
 do. 
 
 Do. 
 
 do. 
 
 422 
 
 7 
 
 do. 
 
 London, St Ethel- 
 
 Arch of doorway on south side 
 
 7°5 
 
 5 
 
 Caveler's Specimens. 
 
 dreda, Ely Place 
 
 (c. 1380, Caveler) 
 
 
 
 
 Do. 
 
 Jamb, bases, and abaci of do. - 
 
 7°5 
 
 6 
 
 do. 
 
 London, St John's 
 
 Plan 
 
 192 
 
 t 
 
 Scott's Lectures. 
 
 chapel in the 
 
 
 
 
 
 Tower of London 
 
 
 
 
 
 Do. 
 
 Interior - - - - - 
 
 2 §3 
 
 
 Parker's Glossary. 
 
 London, Temple 
 church 
 
 Transverse sections of choir 
 
 35 
 
 1 
 
 Billings' Temple 
 
 
 
 
 Church. 
 
 Do. 
 
 Lancet window of west gable of 
 choir 
 
 462 
 
 
 Caveler's Specimens. 
 
724 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Cage. 
 
 Source. 
 
 London, Temple 
 
 Arch-mold in western arch of 
 
 668 
 
 1 
 
 Billings' Temple 
 
 church 
 
 south aisle of choir 
 
 
 
 Church. 
 
 Do. 
 
 Arch-mold in round church 
 west of above 
 
 668 
 
 3 
 
 do. 
 
 Do. 
 
 Molded cap in choir 
 
 689 
 
 4 
 
 do. 
 
 Do. 
 
 Rib in aisle of nave 
 
 675 
 
 6 
 
 do. 
 
 Do. 
 
 do. 
 
 675 
 
 7 
 
 do. 
 
 Do. 
 
 do. 
 
 675 
 
 8 
 
 do. 
 
 Do. 
 
 do. 
 
 675 
 
 9 
 
 do. 
 
 Do. 
 
 do. 
 
 675 
 
 10 
 
 do. 
 
 Do. 
 
 do. - - - 
 
 675 
 
 1 1 
 
 do. 
 
 Long Melford, 
 
 Exterior elevation of aisle and 
 
 547 
 
 
 Lauriston Conder. 
 
 Suffolk 
 
 nave 
 
 
 
 
 Do. 
 
 Left, pier, arch, abacus of nave ; 
 Right, pier, base, plinth of 
 nave 
 
 663 
 
 9 
 
 do. 
 
 Do. 
 
 Pier, arch, and abacus of Lady 
 chapel 
 
 665 
 
 12 
 
 do. 
 
 Do. 
 
 Interior from S.W. - 
 
 214 
 
 
 F. Bond. 
 
 Louth, Lincolnshire 
 
 Exterior from N.E. - 
 
 139 
 
 
 — . James, Louth. 
 
 Do. 
 
 Spire, exterior and interior 
 
 
 
 
 
 face 
 
 611 
 
 
 James Fowler. 
 
 Do. 
 
 East front 
 
 89 
 
 
 |. |. Cresswell. 
 
 Do. 
 
 Parapet and cross of E. gable - 
 
 397 
 
 
 do. 
 
 Do. 
 
 Interior from S.W. (the roof is 
 modern) 
 
 213 
 
 
 F. Bond. 
 
 Do. 
 
 Base in chancel - 
 
 448 
 
 3 
 
 do. 
 
 Do. 
 
 Base of piers of tower 
 
 697 
 
 23 
 
 Paley's Moldings. 
 
 Lutton, Northants 
 
 Doorway 
 
 584 
 
 
 Parker's Glossary. 
 
 Lvmpenhoe, Norfolk 
 
 Roof of nave - 
 
 554 
 
 
 Brandon's Roofs. 
 
 Lynn, St Margaret, 
 
 Parclose screen N. of choir 
 
 162 
 
 
 F. Bond. 
 
 Norfolk 
 
 
 
 
 
 Do. 
 
 Capital in choir 
 
 424 
 
 4 
 
 F. Bond. 
 
 Lynn, St Nicholas - 
 
 Plan --..- 
 
 21 7 
 
 2 
 
 E. M. Beloe. 
 
 Do. 
 
 Interior from S.W. - 
 
 2 14 
 
 
 F. Bond. 
 
 Maidstone, All 
 
 One jamb and mullion of 
 
 701 
 
 4 
 
 Weale's Papers. 
 
 Saints' 
 
 window 
 
 
 
 
 Do. 
 
 do. 
 
 701 
 
 5 
 
 do. 
 
 Maldon 
 
 East window and four mullions 
 
 5'5 
 
 
 Hadfield's Essex. 
 
 Malmesbury abbey, 
 
 Exterior of S. nave - 
 
 375 
 
 
 T. Perkins. 
 
 Wiltshire 
 
 
 
 
 
 Do. 
 
 Interior of nave from N.W. 
 
 522 
 
 
 F. Bond. 
 
 Do. 
 
 Capital in nave 
 
 412 
 
 2 
 
 do. 
 
 Manorbier, Pem- 
 
 N. tower from N.E. 
 
 587 
 
 
 do. 
 
 broke 
 
 
 
 
 
 March, Cambridge 
 
 Hammerbeam roof (collotype) 
 
 
 
 
 
 face 
 
 566 
 
 
 C. C. Hodges. 
 
 Market Deeping, 
 
 Plan of pier, base, plinth 
 
 663 
 
 2 
 
 Nene Valley. 
 
 Lincolnshire 
 
 
 
 
 
 Do. 
 
 Pier, arch, hood-mold, abacus - 
 
 665 
 
 3 
 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 Name of Place. 
 
 I (escription. 
 
 Page. 
 
 No. 
 
 in 
 
 Page 
 
 28 
 
 Soun 1 
 Nene Valley. 
 
 Market Deeping, 
 
 Molded capital 
 
 685 
 
 Lincolnshire 
 
 
 
 
 
 M i i imi I'M , 1 »erb\ 
 
 Interior from east - 
 
 203 
 
 
 F. Pond. 
 
 shire 
 
 
 
 
 
 Do. 
 
 Exterior of S. side of chancel - 
 
 213 
 
 
 do. 
 
 Meopham 
 
 Cinquefoiled circle - 
 
 5 10 
 
 8 
 
 Brandon's Analysts. 
 
 Middleton Cheney, 
 
 Doorway of south porch - 
 
 7°5 
 
 4 
 
 1 . w. \\ oodhouse. 
 
 Northants 
 
 
 
 
 
 Milton abbey, 
 
 Window in choir 
 
 47« 
 
 
 Sharpe's Windows. 
 
 Dorset 
 
 
 
 
 
 MlNCHINHAMPTON, 
 
 Exterior of S. transept 
 
 286 
 
 
 ( '. VVontner Smith. 
 
 Gloucester 
 
 
 
 
 
 Do. 
 
 Interior of S. transept 
 
 287 
 
 
 W. E. Hodgkin. 
 
 Moulton, Lincoln- 
 
 Abacus - 
 
 685 
 
 6 
 
 Sharpe's Lincoln. 
 
 shire 
 
 
 
 
 
 Netley abbey, 
 
 Fair of lancets under common 
 
 457 
 
 5 
 
 Sharpe's Windows. 
 
 Hampshire 
 
 dripstone ; tympanum un- 
 pierced 
 
 
 
 
 Do. 
 
 Plate tracery in chapter house - 
 
 457 
 
 6 
 
 do. 
 
 Do. 
 
 East window - 
 
 47 1 
 
 
 Sharpe's Parallels. 
 
 Do. 
 
 Jamb and bases of east window 
 
 699 
 
 3 
 
 do. 
 
 Do. 
 
 Arch and abaci of do. 
 
 699 
 
 4 
 
 do. 
 
 Do. 
 
 Molded cap and base 
 
 687 
 
 3 
 
 do. 
 
 Do. 
 
 Parapet of choir aisle 
 
 385 
 
 7 
 
 do. 
 
 Do. 
 
 String of aisle - - - - 
 
 6S1 
 
 16 
 
 do. 
 
 Do. 
 
 do. 
 
 6S1 
 
 17 
 
 do. 
 
 Nevers, St Etienne, 
 
 Doorway of N. transept - 
 
 574 
 
 
 F. Bond. 
 
 Nievre 
 
 
 
 
 
 Newark, Notts 
 
 Spire 
 
 627 
 
 
 Wickes' Spiirs and 
 Towers. 
 
 Do. 
 
 Interior from S.W. (the piers of 
 the original central tower 
 remain) 
 
 178 
 
 
 F. Bond. 
 
 Newcastle-on-Tyne 
 
 Spire of cathedral - 
 
 624 
 
 
 J. P.Gibson, Hexham. 
 
 Newhaven, Sussex 
 
 Exterior from N.E. (south nave 
 and porch are modern) 
 
 17 
 
 
 H. Bedford Pirn. 
 
 Newport, St YVoolos, 
 
 Capital of west doorway - 
 
 421 
 
 1 
 
 F. Bond. 
 
 Monmouth 
 
 
 
 
 
 Do. 
 
 do. 
 
 421 
 
 5 
 
 do. 
 
 New Shoreham, 
 
 Exterior from S.E. - 
 
 37 3 
 
 
 do. 
 
 Sussex 
 
 
 
 
 
 Do. 
 
 Vault of choir 
 
 3*3 
 
 
 do. 
 
 Do. 
 
 Leal -■ roll in N. pier arch 
 
 43° 
 
 
 Brandon's Analysis. 
 
 Do. 
 
 Capital and spring of arch in 
 choir 
 
 ^73 
 
 
 F. Bond. 
 
 Do. 
 
 ( apital in choir 
 
 421 
 
 8 
 
 do. 
 
 do. 
 
 Dm. 
 
 do. 
 
 423 
 
 1 
 
 Do. 
 
 do. 
 
 423 
 
 2 
 
 do. 
 
 Do. 
 
 Arch mold in choir 
 
 667 
 
 2 
 
 Sharpe's New Shore- 
 ham. 
 
 Do. 
 
 do. 
 
 667 
 
 3 
 
 do. 
 
726 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 New Walsingham, 
 
 Roof of S. aisle 
 
 57° 
 
 
 Pugin's Examples. 
 
 Norfolk 
 
 
 
 
 
 Do. 
 
 Interior of S. aisle from W. 
 
 57i 
 
 
 F. Bond. 
 
 Northampton, St 
 
 Shaft ----- 
 
 256 
 
 
 Colling's Foliage. 
 
 Peter 
 
 
 
 
 
 Do. 
 
 Capital ----- 
 
 4i5 
 
 6 
 
 do. 
 
 Do. 
 
 Plan of pier, base, plinth - 
 
 663 
 
 1 
 
 Nene Valley. 
 
 Do. 
 
 Pier, arch, abacus - 
 
 665 
 
 1 
 
 do. 
 
 Northampton, St 
 
 Molded capital in nave - 
 
 685 
 
 29 
 
 do. 
 
 Sepulchre 
 
 
 
 
 
 NORTHBOROUGH, 
 
 Plan of pier, base, plinth 
 
 663 
 
 s 
 
 do. 
 
 Northants 
 
 
 
 
 
 Do. 
 
 Arch-mold - 
 
 670 
 
 3 
 
 do. 
 
 North Creake, 
 
 Windows in S. aisle 
 
 473 
 
 
 F. Bond. 
 
 Norfolk 
 
 
 
 
 
 North Luffenham, 
 
 Spire from N.E. - 
 
 613 
 
 
 do. 
 
 Northants 
 
 
 
 
 
 North Mimms, 
 
 Plan 
 
 2I 5 
 
 10 
 
 Brandon's Churches. 
 
 Hertfordshire 
 
 
 
 
 
 North Walsham, 
 
 Roof 
 
 562 
 
 
 Brandon's Roofs. 
 
 Norfolk 
 
 
 
 
 
 Norwich cathedral 
 
 Plan 
 
 148 
 
 4 
 
 Walcott's Church 
 Arrangement. 
 
 Do. 
 
 Exterior of north transept from 
 N.W. 
 
 3* 
 
 
 Messrs Coe, Norwich. 
 
 Do. 
 
 Spire from S.W. 
 
 622 
 
 
 F. Bond. 
 
 Do. 
 
 Nave, S. transept, and cloister 
 from S.W. 
 
 522 
 
 
 do. 
 
 Do. 
 
 Exterior of choir and S.E. apse 
 
 160 
 
 
 do. 
 
 Do. 
 
 Interior of N. aisle of nave 
 from VV. 
 
 238 
 
 
 do. 
 
 Do. 
 
 Plan of major pier in nave 
 
 659 
 
 7 
 
 Britton's Cathedrals. 
 
 Do. 
 
 Interior of S. triforium of nave 
 from W. 
 
 37' 
 
 
 F. Bond. 
 
 Do. 
 
 Interior of nave from S.W. 
 
 33° 
 
 
 do. 
 
 Do. 
 
 Vault of nave - - - - 
 
 33° 
 
 
 do. 
 
 Do. 
 
 East side of interior of south 
 transept 
 
 168 
 
 
 do. 
 
 Do. 
 
 Triforium of apse from N.W. - 
 
 273 
 
 
 do. 
 
 Do. 
 
 Tracery in west walk of cloister 
 
 506 
 
 
 do. 
 
 Norwich, St Stephen 
 
 Exterior from S.W. - 
 
 228 
 
 
 do. 
 
 Do. 
 
 Interior from S.W. - 
 
 228 
 
 
 do. 
 
 Do. 
 
 Roof of nave - 
 
 SS6 
 
 
 Brandon's Roofs. 
 
 Old Walsingham, 
 
 E. window of aisle - 
 
 480 
 
 5 
 
 F. Bond. 
 
 Norfolk 
 
 
 
 
 
 Osbournby, Lin- 
 
 Base of pier - 
 
 697 
 
 16 
 
 Sharpe's Lincoln. 
 
 colnshire 
 
 
 
 
 
 Ottery St Mary, 
 
 Screen of Lady chapel from E. 
 
 181 
 
 
 F. B. Dickinson. 
 
 Devon 
 
 
 
 
 
INDEX TO ILLUSTRATIONS. 
 
 Name of Place. 
 
 1 Ascription. 
 
 Page. 
 
 No. 
 
 in 
 
 Page. 
 
 Source. 
 
 OUNDLE 
 
 Window W. of S. aisle 
 
 461 
 
 K. P. Brereton. 
 
 Oxford, All Souls' - 
 
 Fan vault of corridor to chapel 
 
 344 
 
 4 
 
 Pugin's Examples. 
 
 Do. 
 
 One jamb and mullion of win- 
 dow in chapel 
 
 701 
 
 6 
 
 do. 
 
 Do. 
 
 Ground-course of chapel - 
 
 679 
 
 1 1 
 
 do. 
 
 Oxford, Christ 
 
 Vault of hall staircase 
 
 343 
 
 
 W. S. Weatherley. 
 
 Church 
 
 
 
 
 
 Oxford cathedral 
 
 Plan of church 
 
 r 5 2 
 
 3 
 
 Britton's Cathedrals. 
 
 (St Frideswide's 
 
 
 
 
 
 priory) 
 
 
 
 
 
 Do. 
 
 Spire ----- 
 
 613 
 
 
 F. Bond. 
 
 Do. 
 
 Window of belfry 
 
 5'2 
 
 
 Parker's Glossary. 
 
 Do. 
 
 Interior of choir and central 
 tower from N.W. 
 
 27 
 
 
 T. Maclaren. 
 
 Do. 
 
 Interior of N. transept from 
 S.W. 
 
 5 2 5 
 
 
 F. Bond. 
 
 Do. 
 
 Capital in choir 
 
 417 
 
 1 
 
 do. 
 
 Do. 
 
 Capital ----- 
 
 4-' 
 
 6 
 
 do. 
 
 Do. 
 
 do. - 
 
 421 
 
 7 
 
 do. 
 
 Do. 
 
 Capital in transept - 
 
 4 2 3 
 
 8 
 
 do. 
 
 Do. 
 
 Fern capital - 
 
 4i5 
 
 5 
 
 Parker's Glossary. 
 
 Do. 
 
 Interior of choir from S.W. 
 
 33 l 
 
 
 F. Bond. 
 
 Do. 
 
 Vault of choir - 
 
 33i 
 
 
 do. 
 
 Do. 
 
 Pendant in vault of choir 
 
 297 
 
 
 E. W. M. Wonnacott. 
 
 Oxford, Divinity 
 
 Plan (not to scale) - 
 
 347 
 
 
 F. Bond. 
 
 School 
 
 
 
 
 
 Do. 
 
 Interior of 
 
 33 1 
 
 
 do. 
 
 Do. 
 
 Vault ----- 
 
 33i 
 
 
 do. 
 
 Do. 
 
 Window ----- 
 
 492 
 
 
 do. 
 
 Oxford, Magdalen 
 
 Tower ----- 
 
 s ss 
 
 
 do. 
 
 College 
 
 
 
 
 
 Do. 
 
 Ground-course of chapel - 
 
 679 
 
 12 
 
 Pugin's Examples. 
 
 Do. 
 
 Jamb of west doorway of chapel 
 
 707 
 
 5 
 
 do. 
 
 Oxford, Merton 
 
 Tower 
 
 5S9 
 
 
 F. Bond. 
 
 College chapel 
 
 
 
 
 
 Do. 
 
 East window - 
 
 473 
 
 
 do. 
 
 Do. 
 
 Buttress on north side 
 
 35 2 
 
 
 Pugin's Specimens. 
 
 Do. 
 
 West windows of vestry - 
 
 4S0 
 
 2 
 
 F. Bond. 
 
 Do. 
 
 ( Iruund-course of transept 
 
 679 
 
 IO 
 
 Pugin's Examples. 
 
 Do. 
 
 Plan of moldings above capitals 
 of transept doorway 
 
 707 
 
 - 
 
 ^do. 
 
 Do. 
 
 do., plan of columns showing 
 capitals 
 
 707 
 
 3 
 
 do. 
 
 Do. 
 
 do., plan of columns showing 
 bases 
 
 707 
 
 4 
 
 do. 
 
 Oxford, New College 
 
 Parapet - 
 
 3S7 
 
 - 
 
 do. 
 
 OxFORD,Proscholium 
 
 Vault - 
 
 3 2 9 
 
 
 F. Bond. 
 
 Do. 
 
 Plan of vault - 
 
 308 
 
 10 
 
 do. 
 
 Oxford, Schools 
 
 do. 
 
 3 2 4 
 
 8 
 
 do. 
 
 tower 
 
 
 
 
 
 Oxford, St Giles - 
 
 Two pairs of lancets not under 
 common dripstone 
 
 457 
 
 3.4 
 
 Sharpe's Windows. 
 
728 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Oxford, St Mary - 
 
 Spire ----- 
 
 6 3 l 
 
 
 T. G. Jackson. 
 
 Do. 
 
 Arch-mold in nave - 
 
 671 
 
 5 
 
 Pugin's Specimens. 
 
 Do. 
 
 Molded cap of pier in nave 
 
 693 
 
 2 
 
 do. 
 
 Oxford, St Mary 
 
 Buttress on S. side, c. 1340 
 
 353 
 
 
 do. 
 
 Magdalen church 
 
 
 
 
 
 Oxford, St Michael 
 
 Tower ----- 
 
 587 
 
 
 F. Bond. 
 
 Do. 
 
 Window ----- 
 
 478 
 
 
 Parker's Glossary. 
 
 P aston, Norfolk 
 
 Interior of barn 
 
 5S 2 
 
 
 F. Bond. 
 
 Do. 
 
 Interior of nave from S.E. 
 
 55 1 
 
 
 do. 
 
 Patrington, York- 
 
 Plan - - - - 
 
 2I 5 
 
 1 1 
 
 John 0' Gaunt Sketch 
 
 shire 
 
 
 
 
 Book. 
 
 Do. 
 
 Exterior from west - 
 
 634 
 
 
 L. Barlow and H. A. 
 Paley. 
 
 Do. 
 
 Exterior from S.E. - 
 
 612 
 
 
 F. Bond. 
 
 Do. 
 
 Interior from W. 
 
 133 
 
 
 do. 
 
 Do. 
 
 Capital of tower pier 
 
 437 
 
 1 
 
 do. 
 
 Do. 
 
 do. nave 
 
 437 
 
 2 
 
 do. 
 
 Do. 
 
 do. transept - 
 
 437 
 
 3 
 
 do. 
 
 Do. 
 
 Arch-braced roof with collar 
 
 553 
 
 8 
 
 Churches of Yorkshire. 
 
 Patrixbourne, Kent 
 
 East front - - - - 
 
 218 
 
 
 J. W. Bignell. 
 
 Pershore, Wor- 
 
 Interior of choir from S.W. 
 
 75 
 
 
 H. Bedford Pirn. 
 
 cestershire 
 
 
 
 
 
 Peterborough 
 
 West front (collotype) - face 
 
 112 
 
 
 G. W. Wilson Ltd., 
 
 cathedral 
 
 
 
 
 Aberdeen. 
 
 Do. 
 
 Exterior of eastern chapels 
 
 3 6 5 
 
 
 F. Bond. 
 
 Do. 
 
 Plan of vaults 
 
 308 
 
 1 
 
 do. 
 
 Do. 
 
 S. aisle of nave from E. - 
 
 238 
 
 
 do. 
 
 Do. 
 
 Plan of minor pier in nave 
 
 659 
 
 6 
 
 Britton's Cathedrals. 
 
 Do. 
 
 Clerestory passage of N. nave 
 from E. 
 
 544 
 
 
 F. Bond. 
 
 Do. 
 
 Interior of N. transept from 
 S.W. 
 
 161 
 
 
 do. 
 
 Do. 
 
 Interior of S. aisle of choir 
 from W. 
 
 3i8 
 
 
 do. 
 
 Do. 
 
 Triforium of choir - 
 
 468 
 
 
 Scott's Gleanings. 
 
 Do. 
 
 do. 
 
 468 
 
 
 do. 
 
 Pont Audemer, 
 
 Interior ----- 
 
 J3 1 
 
 
 H. A. Prothero. 
 
 Normandy 
 
 
 
 
 
 Pontigny, Yonne - 
 
 Exterior from S.W. - 
 
 599 
 
 
 F. Bond. 
 
 Do. 
 
 Shrine of St Edmund of Can- 
 terbury from west 
 
 187 
 
 
 do. 
 
 Do. 
 
 do. from east, with staircase 
 
 187 
 
 
 do. 
 
 PUCKLECHURCH, 
 
 Buttress ----- 
 
 355 
 
 
 Colling's Details. 
 
 Gloucestershire 
 
 
 
 
 
 Rievatjlx abbey, 
 
 Transverse and diagonal ribs 
 
 675 
 
 14 
 
 Sharpe's Parallels. 
 
 Yorkshire 
 
 in choir and aisles 
 
 
 
 
 Do. 
 
 Wall rib in choir and aisles 
 
 675 
 
 15 
 
 do. 
 
 Do. 
 
 String of triforium - 
 
 6S1 
 
 •5 
 
 do. 
 
INDKX TO ILLUSTRATIONS. 
 
 7-') 
 
 Name of Place. 
 
 Rievaulx abbey, 
 Yorkshire 
 Do. 
 Do. 
 
 Ripon minster, 
 Yorkshire 
 Do. 
 
 Do. 
 Roche abbey, York- 
 shire 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Rochester cathe- 
 dral 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Rome 
 Do. 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Ro.m-.ey abbey, 
 Hampshire 
 Do. 
 
 Do. 
 
 Do. 
 Roxwell, Essex 
 
 Do. 
 
 Do. 
 
 Do. 
 Rudston, Yorkshire 
 RxjSHDEN, Northants 
 
 Description. 
 
 Do. 
 
 Jamb and bases of lower triplet 
 
 of lancets at east end 
 Arch and abaci of do. 
 Ground-course of choir - 
 Interior of choir (restoration of) 
 
 Exterior of choir and S. tran- 
 sept 
 Arch-mold of east window 
 Plan of pier 
 
 Transverse rib 
 Diagonal rib - 
 String of triforium - 
 String of aisle - 
 West doorway 
 
 Jamb of west doorway of nave 
 
 Arch of do. 
 
 Plan of do. - 
 
 Bases of doorway to chapter 
 house 
 
 Plan of Old St Peter's 
 
 Plan of S. Maria Maggiore 
 
 Interior of S. Agnese from W. - 
 
 Interior of apse of S. Clemente ; 
 bishop's chair and bench of 
 presbyters ; celebrant faces 
 the congregation 
 
 Basilica of Maxentius and Con- 
 stantine (Temple of Peace) ; 
 plan 
 
 Capital of Baths of Diocletian - 
 
 Capital of Colosseum 
 
 Acanthus from Arch of Sep- 
 timus Severus 
 
 Acanthus from Pantheon 
 
 Plan of church 
 
 Norman triplet in W. wall of S. 
 
 transept 
 Corbel table - 
 
 do. 
 Buttress (i 377-1404, Hadfield) 
 Window - - - - - 
 do. 
 do. 
 do. 
 Molded cap of pier on north 
 
 side hi nave 
 Molded cap of pier of north 
 side of nave 
 
 Page. 
 
 699 
 
 699 
 679 
 
 5 1° 
 
 51° 
 
 661 
 
 675 
 67S 
 681 
 681 
 
 574 
 
 7°3 
 7°3 
 703 
 45 2 
 
 147 
 
 148 
 
 t S 5 
 
 3 
 
 290 
 
 4^5 
 43° 
 426 
 
 426 
 151 
 
 457 
 
 3S5 
 385 
 353 
 502 
 502 
 502 
 5°S 
 691 
 
 6g 1 
 
 No. 
 
 1 
 
 Page. 
 
 Si IK] , ,-. 
 
 Sharpe's Parallels. 
 
 do. 
 do. 
 W. S. Weatherley. 
 
 C. Watson, Ripon. 
 
 Sharpe's Windows. 
 Sharpe's Parallels. 
 
 do. 
 do. 
 do. 
 do. 
 F. R. Taylor. 
 
 Caveler's Specimens. 
 do. 
 do. 
 do. 
 
 Hubsch's Monuments 
 do. 
 
 H. Bedford Pirn. 
 
 Scott's Essay. 
 
 Mauch's Orders. 
 Choisy's Histoire. 
 Colling's Foliage. 
 
 do. 
 Britton's Architec- 
 tural Antiquities. 
 do. 
 
 Parker's Glossary. 
 
 do. 
 Hadfield's Essex. 
 
 do. 
 
 do. 
 
 do. 
 Sharpe's Windows. 
 Nene Valley. 
 
 do. 
 
/o 1 - 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Rushdex, Northants 
 
 Molded cap of pier on north 
 side of nave 
 
 691 
 
 16 
 
 Nene Valley. 
 
 Do. 
 
 Spire crocket - 
 
 625 
 
 
 Churches of North- 
 ants. 
 
 St Alban's cathedral 
 
 Plan ----- 
 
 153 
 
 2 
 
 Lord Grimthorpe's 
 Guide. 
 
 Do. 
 
 Exterior (twelfth century) from 
 N.E. 
 
 98 
 
 
 Scott's Lectures. 
 
 Do. 
 
 Interior of nave from S.W. 
 
 
 
 
 
 (collotype) - - - face 
 
 14 
 
 
 F. R. Taylor. 
 
 Do. 
 
 Plan of pier in nave 
 
 659 
 
 5 
 
 Neale's St Albans. 
 
 Do. 
 
 Saint's chapel from S.W., with 
 pedestal of St Alban's shrine, 
 and watching-loft 
 
 186 
 
 
 G. H. Lovegrove. 
 
 Do. 
 
 Rood screen from W. 
 
 178 
 
 
 F. Bond. 
 
 Do. 
 
 Reredos of high altar 
 
 178 
 
 
 do. 
 
 Do. 
 
 Chantry chapel of Abbot Ram- 
 ryge ('• JS 20 ) 
 
 208 
 
 
 E. W. M. Wonnacott. 
 
 St David's cathedral 
 
 Interior of nave from S.E. 
 
 525 
 
 
 F. Bond. 
 
 Do. 
 
 Capitals in nave 
 
 412 
 
 5 
 
 do. 
 
 Do. 
 
 Pedestal of shrine of St David - 
 
 1S7 
 
 
 do. 
 
 Do. 
 
 Presbytery from W. 
 
 162 
 
 
 do. 
 
 St Dexis, Seine 
 
 Interior of nave from S.W. 
 
 537 
 
 
 Viollet-le-Duc's 
 Dictiomiaire. 
 
 St Gall 
 
 Plan ----- 
 
 194 
 
 
 Scott's Essay. 
 
 St Georges de 
 
 Exterior from E. - - - 
 
 160 
 
 
 E. Muff. 
 
 BOSCHERVILLE, 
 
 
 
 
 
 Seine-Inferieure 
 
 
 
 
 
 St Hilaire, St 
 
 Vault of nave - 
 
 328 
 
 
 F. Bond. 
 
 Florext, Maine- 
 
 
 
 
 
 et-l.oire 
 
 
 
 
 
 St Lawrence, 
 
 Exterior from S.W. - 
 
 2I 3 
 
 
 do. 
 
 Thanet 
 
 
 
 
 
 St Neots, Hunting- 
 
 Buttress ----- 
 
 356 
 
 
 Colling's Details. 
 
 don 
 
 
 
 
 
 Salford, Warwick - 
 
 Window ----- 
 
 487 
 
 2 
 
 Parker's Glossary. 
 
 Sai.house, Norfolk - 
 
 Capital of nave - 
 
 437 
 
 5 
 
 F. Bond. 
 
 Salisbury cathedral 
 
 Plan ----- 
 
 r 54 
 
 2 
 
 Britton's Cathedrals. 
 
 Do. 
 
 Exterior from N.E. (collotype) 
 
 
 
 
 
 face 
 
 170 
 
 
 G. W. Wilson Ltd., 
 Aberdeen. 
 
 Do. 
 
 Exterior from E. - - - 
 
 458 
 
 
 F. R. Taylor. 
 
 Do. 
 
 Cloister and nave from S.E. 
 
 458 
 
 
 W. Davidson. 
 
 Do. 
 
 Interior of Lady chapel - 
 
 173 
 
 
 F. R. Taylor. 
 
 Do. 
 
 Gable of S. transept 
 
 457 
 
 10 
 
 Parker's Glossary. 
 
 Do. 
 
 Bands of piers 
 
 249 
 
 
 Viollet-le-Duc's 
 Dictionnaire. 
 
 Do. 
 
 Leaf scroll 
 
 43 1 
 
 1 
 
 Colling's Foliage. 
 
 Do. 
 
 Base in cloister 
 
 697 
 
 1 
 
 Colling's Details. 
 
 Do. 
 
 Interior of spire 
 
 636 
 
 
 Parker's Glossary. 
 
 Sai.l, Norfolk 
 
 West doorway 
 
 575 
 
 
 F. Bond. 
 
INDEX TO ILLUSTRATIONS. 
 
 
 Name of Place. 
 
 I lescription. 
 
 Page 
 
 No 
 
 in 
 
 Sourci . 
 
 
 
 
 
 
 Page 
 
 
 
 Sall, Norfolk 
 Do. 
 
 East window - 
 
 Interior of nave from E. - 
 
 492 
 
 55' 
 282 
 
 
 F. Bond, 
 do. 
 
 
 Scarborough, St 
 
 Exterior from S.E. - 
 
 
 R. 11. Barker. 
 
 
 Mary 
 
 
 
 
 
 
 Do. 
 
 Interior from N.W. - 
 
 282 
 
 
 do. 
 
 
 Do. 
 
 Corbels - 
 
 386 
 
 2 
 
 ( filing's Details. 
 Sharpe's Parallels. 
 
 
 i Selby abbey, York- 
 shire 
 
 Exterior of north choir - 
 
 S6 
 
 
 
 Do. 
 
 Interior of choir from S.E. 
 
 39° 
 
 
 F. bond. 
 
 
 Do. 
 
 Gargoyles of choir - 
 
 398 
 
 
 Sharpe's Parallels. 
 
 
 Do. 
 
 Transverse and diagonal ribs 
 in choir 
 
 677 
 
 I 2 
 
 do. 
 
 
 I )o. 
 
 Wall rib in choir 
 
 677 
 
 13 
 
 do. 
 
 
 Do. 
 
 String of choir aisle 
 
 683 
 
 6 
 
 do. 
 
 
 Do. 
 
 Base of arcading and vaulting 
 shaft 
 
 697 
 
 1 1 
 
 do. 
 
 
 Sempringham, Lin- 
 
 South doorway 
 
 40 
 
 
 Brandon's Analysts. 
 
 
 colnshire 
 
 
 
 
 
 
 Sens cathedral 
 
 Interior of nave from S.W. 
 
 107 
 
 
 W. S. Weatherley. 
 F. bond. 
 
 
 Sherborne abbey, 
 
 Exterior of choir from S.E. 
 
 376 
 
 
 
 Dorset 
 
 
 
 
 
 
 Do. 
 
 Porch and S. nave - 
 
 576 
 
 
 do. 
 
 
 Do. 
 
 Vault of nave - - - - 
 
 333 
 
 
 do. 
 
 
 Do. 
 
 do. - - - - 
 
 346 
 
 
 R. H. Carpenter. 
 
 
 Do. 
 
 Arch from crossing into transept 
 
 262 
 
 
 do. 
 
 
 Shobdon, Hereford 
 
 Shaft and capital 
 
 415 
 
 3 
 
 Rickman's Styles. 
 
 
 Shottesbrooke, 
 
 Interior of spire 
 
 395 
 
 
 Butterfield's Shottes- 
 
 
 Berks 
 
 
 
 
 brooke. 
 
 
 Shrewsbury abbey 
 
 Interior of nave from S.W. 
 
 521 
 
 
 F. Bond. 
 
 
 Do. 
 
 Capital in western nave - 
 
 440 
 
 1 
 
 do. 
 
 
 Do. 
 
 Exterior of nave from N.- 
 
 526 
 
 
 do. 
 
 
 Shrewsbury, St 
 
 Plate tracery in S. porch - 
 
 468 
 
 
 Scott's Gleanings. 
 
 
 Mary 
 
 
 
 
 
 
 Do. 
 
 Capitals in north aisle of nave - 
 
 424 
 
 7 
 
 F. Bond. 
 
 
 Silchester, Hamp- 
 
 Basilican plan 
 
 -'5 
 
 1 
 
 W. H. St John Hope. 
 
 
 shire 
 
 
 
 
 
 
 Skelton, Yorkshire 
 
 South doorway 
 
 73 
 
 
 Christian's Skelton 
 C 'hurch. 
 
 
 Do. 
 
 Plan of pier, base, plinth 
 
 663 
 
 3 
 
 do. 
 
 
 Sleaford 
 
 Molded cap in north transept - 
 
 691 
 
 2 
 
 Sharpe's Lincoln. 
 
 
 Do. 
 
 do. 
 
 691 
 
 3 
 
 do. 
 
 
 Do. 
 
 Base of pier - 
 
 697 
 
 17 
 
 do. 
 
 
 Do. 
 
 Arch and mullion of windows - 
 
 701 
 
 3 
 
 Brandon's Analysis. 
 
 
 Snettisham, Norfolk 
 
 West window (western porch 
 below) 
 
 481 
 
 
 H. Bedford Pirn. 
 
 
 Solihull, Warwick 
 
 Spire ----- 
 
 626 
 
 
 Brandon's Churches. 
 
 
 Southam, Warwick 
 
 do. 
 
 614 
 
 
 Johnson's Reliques. 
 
 
 South fleet, Kent - 
 
 Exterior from N.W. 
 
 604 
 
 
 H. Bedford Pim. 
 
 
 Southwark, St 
 
 Interior of choir from N.W. 
 
 S 21 
 
 
 F. Bond. 
 
 
 Saviour's 
 
 
 
 
 
 
 Do. 
 
 Vault of retrochoir - 
 
 326 
 
 4 
 
 Pugin's Spec une -ns. 
 
 I Jo. 
 
 Ploughshare vaulting 
 
 311 
 
 
 Scott's Lectures. 
 
73- 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 
 Southwell minster, 
 
 Plan 
 
 L5 2 
 
 2 
 
 Weale's Papers. 
 
 Notts 
 
 
 
 
 
 
 Do. 
 
 Exterior from N.W. 
 
 520 
 
 
 F. Bond. 
 
 
 Do. 
 
 Exterior of choir from S.E. 
 
 359 
 
 
 do. 
 
 
 Do. 
 
 Exterior of S. transept and choir 
 
 39o 
 
 
 do. 
 
 
 Do. 
 
 Fourteenth-century flying but- 
 tress and pinnacle to thir- 
 teenth-century choir 
 
 400 
 
 
 Bell's Southwell 
 
 Minster. 
 
 
 Do. 
 
 Interior of nave from S.W. 
 
 520 
 
 
 F. Bond. 
 
 
 Do. 
 
 Interior of N. aisle of nave 
 from E. 
 
 520 
 
 
 do. 
 
 
 Do. 
 
 Base in choir - 
 
 448 
 
 2 
 
 do. 
 
 
 Do. 
 
 Capitals of doorway of chapter 
 house 
 
 434 
 
 
 A. J. Loughton, 
 Southwell. 
 
 
 Do. 
 
 Spandrels of arcading of chap- 
 ter house 
 
 435 
 
 
 Colling's Foliage. 
 
 
 Do. 
 
 Plan of vault of chapter house 
 
 3 2 5 
 
 2 
 
 Lord Grimthorpe. 
 
 
 Do. 
 
 Choir screen from S.E. - 
 
 179 
 
 
 F. Bond. 
 
 
 Southwold, Suffolk 
 
 Nave roof from west 
 
 57 2 
 
 
 Donald Gooding. 
 
 
 Stamford, All 
 
 Arch, abacus in nave 
 
 66 S 
 
 5 
 
 Nene Valley. 
 
 
 Saints 
 
 
 
 
 
 
 Do. 
 
 Pier, base, plinth in nave 
 
 665 
 
 6 
 
 do. 
 
 
 Do. 
 
 Pier, arch, abacus in chancel - 
 
 665 
 
 8 
 
 do. 
 
 
 Stamford, St 
 
 Jamb, arch, and abaci of west 
 
 705 
 
 3 
 
 J. Langham. 
 
 
 Leonard's priory 
 
 doorway 
 
 
 
 
 
 Stamford, St Martin 
 
 Left, pier, base, plinth ; Right, 
 pier, arch, abacus 
 
 665 
 
 10 
 
 Nene Valley. 
 
 
 Stanwick, Northants 
 
 Spire - 
 
 628 
 
 
 Brandon's Churches. 
 
 
 Do. 
 
 Splayed lancet 
 
 45 7 
 
 9 
 
 Churches of North- 
 ants. 
 
 
 Steyning, Sussex - 
 
 Exterior from N.E. - 
 
 458 
 
 
 F. Bond. 
 
 
 Do. 
 
 Interior of nave from S.W. 
 
 273 
 
 
 do. 
 
 
 Do. 
 
 Exterior of S. aisle of nave 
 
 359 
 
 
 do. 
 
 
 Do. 
 
 Capital in nave 
 
 412 
 
 4 
 
 do. 
 
 
 Do. 
 
 do. 
 
 41 2 
 
 7 
 
 do. 
 
 
 Do. 
 
 do. 
 
 417 
 
 6 
 
 do. 
 
 
 Stoke Golding, 
 
 Ground-course, c. 1330 - 
 
 679 
 
 9 
 
 Weale's Papers. 
 
 
 Leicester 
 
 
 
 
 
 
 Stone, Kent - 
 
 Aisle window - 
 
 5 r 3 
 
 
 Cresy's Stone Church. 
 
 
 Do. 
 
 Both jambs, bases, and mullion 
 of window 
 
 699 
 
 5 
 
 do. 
 
 
 Do. 
 
 Arch of do. - 
 
 699 
 
 6 
 
 do. 
 
 
 Do. 
 
 do. enlarged 
 
 699 
 
 6 a 
 
 do. 
 
 
 Do. 
 
 Pier, arch, abacus - 
 
 665 
 
 7 
 
 do. 
 
 
 Do. 
 
 Leaf scroll - 
 
 43 ' 
 
 2 
 
 Colling's Foliage. 
 
 
 Do. 
 
 do. 
 
 43i 
 
 3 
 
 do. 
 
 
 Stratford, Suffolk 
 
 Buttress, fifteenth century 
 
 353 
 
 
 Brandon's Analysis. 
 
 
 Stuston, Suffolk 
 
 Roof of porch 
 
 554 
 
 
 do. 
 
 
 Sudbury, St Peter, 
 
 Parclose screens of choir 
 
 162 
 
 
 F. Bond. 
 
 
 Suffolk 
 
 
 
 
 
 
 Sutton St Mary, 
 
 Interior of nave from the aisle 
 
 42 
 
 
 do. 
 
 
 Lincolnshire 
 
 
 
 
 
 
INDEX TO ILLUSTRATIONS. 
 
 ?},?•> 
 
 
 
 
 
 No. 
 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 
 'age. 
 
 Source. 
 
 Sutton St Mary, 
 
 Spire from E. - 
 
 ni ; 
 
 
 F. Bond. 
 
 
 Lincolnshire 
 
 
 
 
 
 
 Swayton 
 
 Molded cap in nave 
 
 691 
 
 1 
 
 Sharpe's Lincoln. 
 
 
 Tangmere, Sussex - 
 
 Plan - 
 
 2I 5 
 
 3 
 
 Brandon's Churches. 
 
 
 Taunton, St Mary 
 
 Tower from N.E. 
 
 595 
 
 
 F. J. Allen. 
 
 
 Magdalen, Somer- 
 
 
 
 
 
 
 set 
 
 
 
 
 
 
 Do. 
 
 Pinnacles of tower - 
 
 607 
 
 
 Colling's Details. 
 
 
 Do. 
 
 Capital in nave 
 
 417 
 
 4 
 
 F. Bond. 
 
 
 Temple Balsal, 
 
 Plan 
 
 2I 5 
 
 2 
 
 Brandon's Churches. 
 
 
 Warwick 
 
 
 
 
 
 
 Tenby, Pembroke - 
 
 Interior of nave from N. aisle - 
 
 410 
 
 
 F. Bond. 
 
 
 Do. 
 
 West doorway 
 
 580 
 
 
 do. 
 
 
 Terrington St 
 
 Exterior from S.W. 
 
 92 
 
 
 Johnson's Reliques. 
 
 
 Clement, Norfolk 
 
 
 
 
 
 
 Terling, Essex 
 
 Molded cap of pier in nave 
 
 6 93 
 
 1 
 
 Hadfield's Essex. 
 
 
 Tewkesbury abbey, 
 
 Central tower from W. 
 
 39° 
 
 
 W. G. Bannister. 
 
 
 Gloucestershire 
 
 
 
 
 
 
 Do. 
 
 Doorway of north porch - 
 
 260 
 
 
 do. 
 
 
 Do. 
 
 Bases in N. porch - 
 
 44S 
 
 4 
 
 do. 
 
 
 Do. 
 
 Interior of nave from S.W. 
 
 297 
 
 
 F. Bond. 
 
 
 Do. 
 
 Vault of nave - 
 
 33 2 
 
 
 W. G. Bannister. 
 
 
 Do. 
 
 Interior of choir 
 
 165 
 
 
 W. T. Alston. 
 
 
 Do. 
 
 Vault of choir 
 
 33° 
 
 
 F. Bond. 
 
 
 Threckingham, 
 
 Abacus ----- 
 
 685 
 
 is 
 
 Sharpe's Lincoln. 
 
 
 Lincolnshire 
 
 
 
 
 
 
 Do. 
 
 do. 
 
 685 
 
 16 
 
 do. 
 
 
 Thorington, Suffolk 
 
 Round tower - 
 
 600 
 
 
 F. Bond. 
 
 
 Tideswell, Derby- 
 
 Exterior of S. transept 
 
 4S0 
 
 6 
 
 do. 
 
 
 shire 
 
 
 
 
 
 
 Do. 
 
 Ground-course 
 
 4°3 
 
 
 do. 
 
 
 Tilney, All Saints', 
 
 Spire - 
 
 6i 2 
 
 
 do. 
 
 
 Norfolk 
 
 
 
 
 
 
 Do. 
 
 Capital in nave 
 
 423 
 
 4 
 
 do. 
 
 
 Tiltey, Essex 
 
 Molded cap of sedilia 
 
 691 
 
 1 J 3 
 
 Hadfield's Essex. 
 
 
 Do. 
 
 Mullions ; fourteenth century - 
 
 699 
 
 9 
 
 do. 
 
 
 Tintern abbey, 
 
 [ Interior of nave from N.E. 
 
 5 2 4 
 
 
 F. Bond. 
 
 
 Monmouth 
 
 
 
 
 
 
 Do. 
 
 East window of choir (restora- 
 tion) 
 
 47.; 
 
 
 Sharpe's Parallels. 
 
 
 Do. 
 
 Rear-arch of aisle window 
 
 5 T2 
 
 
 Sharpe's Windows. 
 
 
 Do. 
 
 Arch-mold in choir 
 
 669 
 
 3 
 
 Sharpe's Parallels. 
 
 
 Do. 
 
 Molded cap and base 
 
 687 
 
 5 
 
 do. 
 
 
 Do. 
 
 Base of pier in nave 
 
 697 
 
 8 
 
 Paley's Moldings. 
 
 
 Do. 
 
 Base - - - - - 
 
 697 
 
 7 
 
 do. 
 
 
 Do. 
 
 Transverse and diagonal ribs 
 in aisle 
 
 677 
 
 5 
 
 Sharpe's Parallels. 
 
 
 Do. 
 
 VVall rib in aisles 
 
 677 
 
 6 
 
 do. 
 
 
 Do. 
 
 String of clerestory - 
 
 683 
 
 2 
 
 do. 
 
 
 Tiverton, Devon - 
 
 S. porch 
 
 576 
 
 
 F. Bond. 
 
734 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Tiverton, Devon - 
 
 Battlements - 
 
 39° 
 
 
 F. Bond. 
 
 Do. 
 
 Capital of nave 
 
 437 
 
 6 
 
 do. 
 
 Toot Baldon, Ox- 
 
 Plan 
 
 216 
 
 2 
 
 Parker's Churches. 
 
 fordshire 
 
 
 
 
 
 Do. 
 
 West front 
 
 225 
 
 
 do. 
 
 Torcello cathedral 
 
 Plan 
 
 148 
 
 1 
 
 Hubsch's Monuments. 
 
 Tours, St Martin's 
 
 Plan of foundations of crypt 
 
 192 
 
 3 
 
 Comte Robert de 
 
 abbey 
 
 
 
 
 Lasteyrie. 
 
 Towyn, Merioneth - 
 
 Exterior from S. 
 
 453 
 
 
 F. Bond. 
 
 Trunch, Norfolk - 
 
 Exterior from S.W. - 
 
 603 
 
 
 do. 
 
 Do. 
 
 Priest's doorway 
 
 575 
 
 
 do. 
 
 Do. 
 
 Interior from S.W. - 
 
 552 
 
 
 do. 
 
 Do. 
 
 Capital in nave - - - 
 
 440 
 
 5 
 
 do. 
 
 Do. 
 
 Ham merbeam roof (details) 
 
 553 
 
 
 Brandon's Roofs. 
 
 Uggleshall, Suffolk 
 
 Nave from east 
 
 572 
 
 
 F. Bond. 
 
 Vicenza, SS. Felix 
 
 Pier and capital 
 
 236 
 
 
 Cattaneo's L' Archi- 
 
 and Fortunatus 
 
 
 
 
 tecture en Italic 
 
 Waddington - 
 
 Abacus ----- 
 
 685 
 
 11 
 
 Sharpe's Lincoln. 
 
 Walsoken, Norfolk 
 
 Capital in nave - 
 
 4i7 
 
 2 
 
 F. Bond. 
 
 Do. 
 
 Arch-mold - 
 
 667 
 
 1 
 
 Colling's Details. 
 
 Waltham abbey, 
 
 Interior of nave from S.E. 
 
 521 
 
 
 F. Bond. 
 
 Essex 
 
 
 
 
 
 Do. 
 
 Circular window 
 
 488 
 
 
 Brandon's Analysis. 
 
 Warmington, 
 
 Window of S. aisle - 
 
 461 
 
 
 R. P. Brereton. 
 
 Northants 
 
 
 
 
 
 Do. 
 
 Belfry window 
 
 467 
 
 
 Caveler's Warming- 
 ton Church. 
 
 Do. 
 
 Buttress 
 
 352 
 
 
 do. 
 
 Do. 
 
 Corbel table of spire 
 
 386 
 
 3 
 
 do. 
 
 Do. 
 
 West doorway 
 
 578 
 
 
 do. 
 
 Do. 
 
 Molded cap of shaft in west 
 doorway 
 
 689 
 
 6 
 
 do. 
 
 Do. 
 
 do. 
 
 689 
 
 7 
 
 do. 
 
 Do. 
 
 Doorway of south porch; jamb, 
 arch, and abaci 
 
 7°5 
 
 7 
 
 Nene Valley. 
 
 Do. 
 
 Pier, arch, hood-mold, abacus - 
 
 665 
 
 2 
 
 do. 
 
 Do. 
 
 Arch-mold - 
 
 666 
 
 3 
 
 Caveler's Warming- 
 ton Church. 
 
 Wawne, Yorkshire - 
 
 Arch-mold (different scale) 
 
 666 
 
 4 
 
 
 
 Do. 
 
 Part of arch-mold (to scale) 
 
 666 
 
 5 
 
 Colling's Details. 
 
 Welbourne, Lin- 
 
 Ground-course, c. 1330 - 
 
 679 
 
 7 
 
 Sharpe's Lincoln. 
 
 colnshire 
 
 
 
 
 
 Wells cathedral 
 
 Plan .... 
 
 T 54 
 
 3 
 
 Rickman's Styles. 
 
 Do. 
 
 Exterior from S.E. (collotype) 
 
 
 
 
 
 face 
 
 602 
 
 
 F. Bond. 
 
 Do. 
 
 Window in Lady chapel - 
 
 474 
 
 3 
 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 ; 
 
 Name of Pla 
 
 Wells cathedral 
 Do. 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 
 Do. 
 Do. 
 Do. 
 Do. 
 
 Do. 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 
 WENLOCK priory, 
 Shropshire 
 Do. 
 Westham, Sussex 
 
 Westminster abbey 
 church 
 Do. 
 Do. 
 Do. 
 Do. 
 
 Do. 
 
 Do. 
 Do. 
 
 Des 
 
 nption. 
 
 Exterior of choir from S.W. 
 Parapet of choir. (Below is a 
 
 twelfth-century string.) 
 Interior of central tower - 
 Interior of nave from S.E. 
 Bishop Bubwith's chantry (pb. 
 
 1424) in N. aisle of nave 
 Plan of pier - - - - 
 Interior of choir from W. 
 
 Vault of choir 
 
 Interior from east of Lady 
 
 chapel 
 Interior of Lady chapel from 
 
 S.W. 
 Vault of Lady chapel 
 Chapter house, interior ■ 
 
 Vault of undercroft of chapter 
 house 
 
 Vault of chapter house - 
 
 Base of central pier of chapter 
 house 
 
 Capitals in choir 
 
 Capitals in western nave - 
 do. 
 
 Capitals in west aisle of S. 
 transept 
 
 Capitals of retrochoir 
 
 Capitals of eastern nave, c. 133S 
 
 Leaf scroll 
 
 Rib in porch of staircase of 
 Vicar's close 
 
 One jamb and mullion of win- 
 dow in Vicar's close 
 
 Arch and abaci of doorway into 
 cloister 
 
 Jamb and bases of do. 
 
 Plan 
 
 P 
 
 Exterior of nave and S. transept 
 Parapet of St Erasmus' chapel - 
 Transverse section of choir 
 Internal elevation of one bay 
 
 of the choir 
 Interior from west - 
 
 Triforium 
 
 Plan of vaults of choir 
 
 Page. 
 
 373 
 
 396 
 
 588 
 
 5 2 4 
 
 20() 
 
 66l 
 
 ] 27 
 
 33 2 
 161 
 
 333 
 
 325 
 1- i 
 
 325 
 
 325 
 697 
 
 424 
 1 1 2 
 
 424 
 424 
 
 437 
 445 
 43 1 
 677 
 
 7 ° ' 
 
 703 
 
 7°3 
 
 2 1 (1 
 
 1 Si 
 
 379 
 
 3S6 
 
 55 
 55 
 
 63 
 
 119 
 
 324 
 
 No. 
 
 in 
 
 Page 
 
 F. Bond. 
 
 I )awkes & Partridge, 
 
 Wells. 
 
 ( '•. K. Saunders. 
 
 F. Bond. 
 
 Dawkes & Partridge, 
 
 Wells. 
 
 J. T. Irvine. 
 I )awkes iV 1'artridge, 
 Wells. 
 
 G. K. Saunders. 
 
 F. Bond. 
 
 do. 
 
 C. D. Sheldon. 
 Dawkes & Partridge, 
 
 Wells. 
 ( '. 1 ). Sheldon. 
 
 E. ( 1. B. Perkins. 
 Paley's Moldings. 
 
 F. Bond. 
 
 do. 
 do. 
 
 do. 
 
 do. 
 
 G. K. Saunders. 
 Colling's Foliage. 
 Pugin's Examples. 
 
 do. 
 
 Potter's Wenlock. 
 
 do. 
 
 E. Turner and W. 
 Ainslie in John 0' 
 Gaunt Sketch Book. 
 
 J. T. Micklethwaite. 
 
 F. R. Taylor. 
 Bloxam's Principles. 
 R. P. Pullan. 
 
 Sydney Vacher. 
 
 G. W. Wilson Ltd., 
 Aberdeen. 
 
 F. R. Taylor. 
 F. Bond. 
 
736 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 Westminster abbey 
 
 Circular window of south 
 
 344 
 
 5 
 
 Pugin's Specimens. 
 
 church 
 
 transept 
 
 
 
 
 Do. 
 
 Capitals in chapel of the Pyx - 
 
 415 
 
 I 
 
 Scott's Gleanings. 
 
 Do. 
 
 do. - 
 
 415 
 
 2 
 
 do. 
 
 Do. 
 
 Plan of aisle of Henry the 
 Seventh's chapel 
 
 94 
 
 
 Britton's Arch. 
 Antiquities. 
 
 Do. 
 
 Flying buttress and pinnacle of 
 Henry the Seventh's chapel 
 
 378 
 
 
 Pugin's Specimens. 
 
 Do. 
 
 Vault of Henry the Seventh's 
 chapel 
 
 347 
 
 
 Willis' Vaulting. 
 
 Do. 
 
 do. (collotype) - face 
 
 343 
 
 
 G. W. Wilson Ltd., 
 Aberdeen. 
 
 Do. 
 
 East walk of cloister 
 
 489 
 
 
 F. R. Taylor. 
 
 Do. 
 
 Vault of cloister 
 
 326 
 
 6 
 
 Viollet-le-Duc's 
 Dictionnaire. 
 
 Westminster hall - 
 
 Roof ----- 
 
 557 
 
 
 F. T. Dollman. 
 
 Weston, Lines. 
 
 Abacus ----- 
 
 685 
 
 10 
 
 Sharpe's Lincoln. 
 
 West Walton, 
 
 Tower ----- 
 
 597 
 
 
 H. J. Austin in John 
 
 Norfolk 
 
 
 
 
 o' Gaunt Sketch 
 Book. 
 
 Do. 
 
 Corbel table - 
 
 3S5 
 
 1 
 
 Colling's Details. 
 
 Do. 
 
 Capital of chancel arch - 
 
 422 
 
 1 
 
 F. Bond. 
 
 Do. 
 
 Capital and base of pier in nave 
 
 43 2 
 
 
 Colling's Details. 
 
 Do. 
 
 Capital of pier in nave 
 
 422 
 
 3 
 
 F. Bond. 
 
 Westwell, Kent - 
 
 East end of chancel 
 
 463 
 
 
 T. C. Yates. 
 
 Whaplode, Lin- 
 
 Capital in western nave - 
 
 423 
 
 3 
 
 F. Bond. 
 
 colnshire 
 
 
 
 
 
 Do. 
 
 Abacus ----- 
 
 685 
 
 4 
 
 Sharpe's Li?icoln. 
 
 Do. 
 
 do. - 
 
 685 
 
 3 
 
 do. 
 
 Do. 
 
 do. 
 
 685 
 
 8 
 
 do. 
 
 Whissendine, 
 
 Tower ----- 
 
 581 
 
 
 J. Langham. 
 
 Rutland 
 
 
 
 
 
 Whitby abbey, 
 
 Choir (restoration), interior 
 
 114 
 
 
 Sharpe's Parallels. 
 
 Yorkshire 
 
 elevation of 
 
 
 
 
 Do. 
 
 Arch of three orders 
 
 275 
 
 
 Parker's Glossary. 
 
 Do. 
 
 Foliated corbel 
 
 433 
 
 
 Sharpe's Parallels, 
 
 Do. 
 
 Ground-course of choir - 
 
 679 
 
 3 
 
 do. 
 
 Do. 
 
 Molded cap and base 
 
 687 
 
 1 
 
 do. 
 
 Do. 
 
 Transverse and diagonal ribs in 
 choir 
 
 675 
 
 12 
 
 do. 
 
 Do. 
 
 Wall rib in choir 
 
 675 
 
 J 3 
 
 do. 
 
 Do. 
 
 String of triforium - 
 
 681 
 
 1 1 
 
 do. 
 
 Do. 
 
 String of east end - 
 
 681 
 
 12 
 
 do. 
 
 Do. 
 
 String of aisle - 
 
 681 
 
 J 3 
 
 do. 
 
 White Colne 
 
 Three mullions 
 
 5*5 
 
 
 Hadfield's Essex. 
 
 Wiggenhall St 
 
 Interior from S.E. - 
 
 57i 
 
 
 F. Bond. 
 
 Mary the Virgin, 
 
 
 
 
 
 Norfolk 
 
 
 
 
 
 Do. 
 
 Jamb, arch, and abaci of north 
 doorway 
 
 705 
 
 2 
 
 Colling's Details. 
 
 Do. 
 
 Molded cap of shaft in north 
 doorway 
 
 689 
 
 8 
 
 do. 
 
INDEX TO ILLUSTRATIONS. 
 
 737 
 
 Name of Place. 
 
 WlGGENHALL St 
 
 Mary the Virgin, 
 
 Norfolk 
 Wiley, Northants 
 
 Do. 
 Wiley, Wilts - 
 
 WlMBOTSHAM, 
 
 Norfolk 
 Winchelsea, Sussex 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 Winchester cathe- 
 dral 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 
 Do. 
 -Do. 
 Winchester, Hos- 
 pital of St Cross 
 Do. 
 Do. 
 Do. 
 Do. 
 Winchester castle 
 
 hall 
 Windsor, St 
 George's chapel 
 Do. 
 Do. 
 Do. 
 Do. 
 Do. 
 Do. 
 Wing ham, Kent 
 Winnal Magdalen, 
 
 Hants 
 Wisbech, Cambridge 
 
 shire 
 Witney, Oxford 
 
 Description. 
 
 Molded cap of shaft in north 
 doorway 
 
 West front 
 Spire 
 
 Lancet triplet - 
 Roof of nave - 
 
 Parapet 
 
 Arch-mold 
 
 Base of pier 
 
 Base of shaft of window - 
 
 Base of sedilia 
 
 Base of piscina 
 
 Plan - 
 
 Plan of crypt - 
 Capital in crypt 
 Interior of N. transept from 
 
 S.E. 
 Plan of pier in west aisle of 
 
 north transept 
 Plan of pier in nave, after re- 
 
 modt lling 
 Interior elevation of one bay of 
 
 the nave 
 Interior of nave from \V. 
 
 (collotype) - - - face 
 Capital in Lady chapel - 
 Bench end in Lady chapel 
 Plan of church 
 
 Exterior from S.E. - 
 Plate tracery in N. porch 
 Spur - 
 do. - 
 Plate tracery - - - - 
 
 Wtst front 
 
 Interior of nave from S.E. 
 Vault of nave - 
 
 do. - 
 Plan of pier of nave 
 Enlarged plan of pier 
 liases, in elevation - 
 Spire from S.W. 
 Window 
 
 Capital in S. aisle - 
 
 Exterior from S. 
 
 3 * 
 
 
 .\... 
 
 Page, 
 
 in 
 
 
 Page. 
 
 689 
 
 9 
 
 270 
 629 
 466 
 560 
 
 
 388 
 
 2 
 
 669 
 697 
 697 
 697 
 697 
 
 4 
 
 5 
 1 
 3 
 
 2 
 
 154 
 
 1 
 
 192 
 
 5 
 
 4IS 
 261 
 
 4 
 
 659 
 
 4 
 
 659 
 
 3 
 
 90 
 
 
 342 
 
 
 413 
 637 
 2I 5 
 
 8 
 
 104 
 
 
 457 
 695 
 695 
 
 8 
 10 
 1 1 
 
 457 
 
 7 
 
 492 
 
 
 33° 
 
 
 332 
 
 
 34i 
 661 
 
 12 
 
 2 55 
 
 
 453 
 622 
 
 
 5°9 
 
 
 440 
 
 3 
 
 621 
 
 
 Colling's Details. 
 
 R. P. Brereton. 
 
 do. 
 Brandon's Analysis. 
 Brandon's Roofs. 
 
 Colling's Details. 
 do. 
 
 do. 
 
 do. 
 do 
 do. 
 Britton's Cathedrals. 
 
 do. 
 Willis' Winchester. 
 F. R. Taylor. 
 
 Bilson's Beginnings. 
 
 Willis' Winchester. 
 
 do. 
 
 F. R. Taylor. 
 
 G. H. Lovegrove. 
 
 do. 
 Dollman's Domestic 
 
 Architecture. 
 H. Bedford Pim. 
 Sharpe's Windows. 
 Parker's Glossary. 
 
 do. 
 Sharpe's Windows. 
 
 F. Bond. 
 
 do. 
 Alan Potter. 
 Willis' Vaulting. 
 
 E. Nash. 
 
 do. 
 do. 
 
 F. Bond. 
 Brandon's Analysis. 
 
 F. Bond. 
 
 H. Bedford Pim. 
 
738 
 
 INDEX TO ILLUSTRATIONS. 
 
 
 
 
 No. 
 
 
 Name of Place. 
 
 Description. 
 
 Page. 
 
 in 
 Page. 
 
 Source. 
 
 WOODBASTWICK, 
 
 Interior of nave from S.E. 
 
 S5 1 
 
 
 F. Bond. 
 
 Norfolk 
 
 
 
 
 
 Woodford, North- 
 
 Lead spout - 
 
 399 
 
 
 Churches of North- 
 
 ants 
 
 
 
 
 ants. 
 
 Worcester cathe- 
 
 Plan of crypt - 
 
 192 
 
 2 
 
 Britton's Cathedrals. 
 
 dral 
 
 
 
 
 
 Wrexham, Denbigh 
 
 Tower ----- 
 
 609 
 
 
 H. Phibbs. 
 
 Do. 
 
 Interior of chancel from S.W. - 
 
 214 
 
 
 F. Bond. 
 
 Wymondham abbey, 
 
 Exterior from S.W. - 
 
 589 
 
 
 do. 
 
 Norfolk 
 
 
 
 
 
 Do. 
 
 Roof of nave - 
 
 565 
 
 
 Brandon's Roofs. 
 
 Yarmouth 
 
 Capital in nave 
 
 423 
 
 5 
 
 F. Bond. 
 
 York minster 
 
 Plan ----- 
 
 153 
 
 1 
 
 Britton's Cathedrals. 
 
 Do. 
 
 West front (collotype) - face 
 
 82 
 
 
 C. C. Hodges. 
 
 Do. 
 
 Western towers and S. nave 
 
 366 
 
 
 F. Bond. 
 
 Do. 
 
 Exterior of choir, S.E. transept 
 and presbytery 
 
 199 
 
 
 do. 
 
 Do. 
 
 Exterior of north transept and 
 central tower 
 
 1 1 
 
 
 do. 
 
 Do. 
 
 Interior of S. transept from 
 S.E. 
 
 523 
 
 
 do. 
 
 Do. 
 
 Interior of nave 
 
 10 
 
 
 do. 
 
 York, St Mary's 
 
 Plan of pier - 
 
 661 
 
 5 
 
 Sharpe's Parallels. 
 
 abbey 
 
 
 
 
 
 Do. 
 
 Arch-mold - - 
 
 668 
 
 4 
 
 do. 
 
 Do. 
 
 Parapet ----- 
 
 385 
 
 8 
 
 do. 
 
 Do. 
 
 String of aisle - 
 
 683 
 
 3 
 
 do. 
 
 Do. 
 
 Transverse rib in aisle 
 
 677 
 
 7 
 
 do. 
 
 Do. 
 
 Diagonal rib in aisle 
 
 677 
 
 8 
 
 do. 
 
 Do. 
 
 Wall rib in aisle 
 
 677 
 
 9 
 
 do. 
 
 York, St Mary 
 
 Belfry window 
 
 457 
 
 1 
 
 Parker's Glossary. 
 
 Bishop Hill Junior 
 
 
 
 
 
 York, St Maurice - 
 
 Plate tracery - 
 
 468 
 
 
 Scott's Gleanings. 
 
 Youlgreave, 
 
 Capital of nave 
 
 421 
 
 4 
 
 F. Bond. 
 
 Derbyshire 
 
 
 
 
 
INDEX OF PLACES. 
 
 ABBAYE-AUX-HOMMES; AUX-DAMES. 
 See CAEN. 
 
 ABBEVILLE. Nave, tnforium, 539. 
 
 ABBEY CWM-HIR. Capitals and foliage, 
 429, 43°' 43 1 ' 43 2 . 435 : " Western " style 
 of capitals, 433 and note, 435. Gothic, 
 " Western school '' of, 105. Piers, "West- 
 ern," 245. Roll molds, 27S. See also 
 LLANIDLOES. 
 
 ABBEY DORE. Aisles, two eastern, 172. 
 Ambulatory, rectangular, 106, 171. Capi- 
 tals and foliage, 418, 428 and note, 429, 
 431, 432: interlacings on capitals, 414: 
 " Western " style of capital, 433. Chapels, 
 five eastern, 172 : low walls between, 200. 
 East front, 68. Flying buttresses, absent, 
 372. Got/iie, "Western school" of, 105. 
 Piers, "Western," 245. Screens and Wood- 
 , work, Tacobean, 203, note. Triforium, ab- 
 sorbedintoclerestory, 536. Vault, 545, note. 
 
 ABBOTSBURY. St Catherine's Chapel. 
 Stone roof, 284. 
 
 ABERDEEN. King's College. Spire 
 lantern or crown, 625. 
 
 ABINGDON. Aisles, 224. 
 
 ABINGER. No buttresses, 362. 
 
 ACHURCH. Transepts, 222. 
 
 ADDERBURV. Roof, 558: arched braces 
 of, 562 : longitudinal braces of, 569. 
 Spire, pinnacles, 623. 
 
 ADEL, 220. Norman porch, 582. Norman 
 roof 572. 
 
 AGLIATE. Basilica, 232. 
 
 AIRAINES. Vaulting, ridge ribs, 335, 339. 
 
 AIX-LA-CHAPELLE. Capitols, cubical, 
 411. Columns, 232, note. Vaults, groined, 
 2Q2, 295, note 3 : wall ribs, 300. 
 
 ALBI. Cathedral. Hall plan, 321 ; with- 
 out aisles, 229. Vault, groined, 291. 
 
 ALLIATA. Baptistery. R00J oh vault, 
 550, note. Church. Apses, three eastern, 
 163. Vaults, east bay of aisles and crypt, 
 groined, 292. Windows, small Roman 
 esque, 459. 
 
 ALMENNO. St Thomas. Roof on vault, 
 
 550, note. 
 ALPIRSBACH. Basilica, 232. 
 
 AMIENS. Cathedral. Altitude ami pro 
 portion, 61. Arches, ogee, north chapel, 
 270. Buttresses, flying, aqueduct on, 400. 
 Chapels, 131, note ; 170. Chapels between 
 buttresses, 197, notes. Fleche, 594. Height, 
 527. Piers, illogical, 241, note. J "lunacies, 
 363, 377. Porches, western, 205. Roofs 0] 
 aisles, 537. Spire, 594. Transepts, 197. 
 Triforium, choir, 537 : nave, 535 : glazed, 
 537, 53S. Vaulting, span of, 291 : ridge 
 ribs, 335. Windows, 490 : choir, 508, note : 
 clerestory, 57. 
 
 ANGERS. Cathedral. Aisleless, 228, 321. 
 Vault, high, in squares, 321 : ridge ribs, 
 335. La Trinite. Vault, quasi-sexpar- 
 tite, 321. St Martin. Tower, central 
 lantern, 592. 
 
 ANGOULEME. Domes, 13, 281. 
 
 ANIOU. Churches. High aisles or parallel 
 naves, counter-thrusts of vaults, 282. 
 
 AQUILEIA. Capitals and abaci, 410. 
 
 ARBONA. Vaulting, ridge ribs, 335. 
 
 ARDENNE. Aisled barn, 225. 
 
 ARLES. Roman Amphitheatre. Ceilings, 
 Roman, stone, 285. Orders, 274. St 
 Trophime. Arch, pointed, 264. Vault, 
 barrel, 13: half barrel in aisles, 13, 
 5'9- 
 
 ARUNDEL. Base, 450. 
 
 AS( JARBY. Molded capital, 444- 
 
 ASH. Capitals and bases, 451. 
 
74Q 
 
 INDEX OF PLACES. 
 
 ASHBY ST LEGER. Window, transom, 
 
 503 : dagger in tracery, 503. 
 ASWARBY. Abacus, transitional, 441. 
 ATHENS. Corinthian capitals, 420, 425, 
 
 426. Foliage, 418. 
 AUCKLAND. Pudsey's Hall. Marble 
 
 shafts, 250. 
 AURONA. Capitals and abaci, 410. See 
 
 BRER A. 
 AUSTREY. Buttress, 358. Dripstones, 
 
 406. Strings, 406. 
 AUTUN. Narthex, 202. Pilasters, 239. 
 
 AUVERGNE. Churches of. Arches, 
 pointed, 264. Choirs, monolithic shalts, 
 411. Corbel-tables, 392. Flying buttresses, 
 369. Polychromy, 250. Romanesque, 369, 
 411. Vaults, barrel, 550; construction, 
 289. 
 
 AUXERRE. Cathedral. Front, west, 
 doorway, 91. St Eusebe. Lighting, con- 
 trast of dark nave and light choir, 592. 
 St Germain. Crypt, or Cou/essio, 191. 
 Spire, 617, 621; sharp angle of, 631; 
 entasis, 632. 
 
 AVRANCHES. Apse, and circumambient 
 aisles, 168, note. 
 
 BACTON. Perpendicular window, 501. 
 
 BADGEVVORTH. Arch, semicircular, 
 
 Gothic fourteenth-century doorway arch, 
 261, 580. Ball flower ornament, 83, 261. 
 
 BAGINTON. Aisles, 224. 
 
 BAIAE. Piscina Mirabilis. Piers, 236. 
 
 BAKEWELL. Plan, cruciform extension, 
 
 222. 
 BALDERTON. Norman porch, 582. 
 
 BANGOR. Cathedral. Towers, western 
 
 and central, 599. 
 BAQUOZA. Basilica, 231. 
 BARCELONA. Transeptal towers, 593. 
 
 BARFRESTON. East end, Norman, rect- 
 angular, 28, 220. Fast front, 29. Niches, 
 84. Window, circular, 468, 516. 
 
 BARNACK. Abaci, painted (south door- 
 way), 442, note. Bases, with flattened 
 roll, 451 ; with water leaf griffe or spur, 
 418, 455. Capitals, 429, note. Pilaster 
 strips (Anglo-Saxon), 351. Spire, 617; 
 too short for tower, 631 : on octagonal 
 drum, 629. Quarries, 20. 
 
 BARNWELL. Tower, no buttresses, 362. 
 
 BARNWOOD. Norman chapel, north of 
 chancel, 225. 
 
 BARROWDEN. Broach spire, 620. 
 
 BARTON-LE-STREET. Capital, 428. 
 
 BARTON-ON-HUMBER. Choir, 221. 
 Tower, central, Anglo-Saxon, 590. 
 
 BARTON SEAGRAVE. South Transept, 
 222. 
 
 BASING. Tower, central, Norman, 590. 
 
 BATH. Abbey Church, 15. Arches, equi- 
 lateral, 259, note : four-centred, 267. Battle- 
 ments, perforated, 399. Buttresses, octa- 
 gonal, to tower, 605. Clei'estory, 259, note, 
 543 : no passage, 545. Flying buttresses, 
 374: with pierced spandrels, 377. Front, 
 east, 91 : west, 91. Internal elevation, 88. 
 Parapets, pierced, 399. Plan, aisled par- 
 allelogram, 179. Shaft-bands, 249. Tower, 
 central with octagonal buttresses, 605 : 
 effect of, 594 : pinnacles, 606 — altered in 
 restoration, 608, note : plan of, oblong, 
 600. Triforium, blank, 88. Vault, ng- 
 shafts, 240. Roman Baths. Compound 
 piers, 236. 
 
 BATTLE. Capitals, plantain or laurel leaf, 
 428, note. Plan, periapsidal, 167. 
 
 BATTLEFIELD. Windows, 501, note. 
 
 BAYEUX. Capitals, Norman, acanthus in 
 427. Doorways, gabled western, 579 
 Facade, western, 73. Hood-molds, 277 
 Moldings, "wolf-heads" in, 41. Spire. 
 621. Triforium (choir), transverse arches. 
 369 note : blindstory, 535. 
 
 BAYHAM. _/%;z,aisleless nave, 201. Eastern 
 transepts, 1S9. 
 
 BEAULIEU. Apse and ambulatory, 157. 
 
 BEAUVAIS. Cathedral. Spire, central, 
 collapse of, 594. Transept wall, 605. 
 Vault pier, $1 . Basse (Euvre. Windows, 
 large Romanesque, 458. St Ettenne. 
 Aisle, vaulting, diagonal ribs, 300. 
 
 BECCLES. Detached tower, 598. 
 
 BECKLEY. Staircase turret, 601. 
 
 BEELEIGH. Chapter house, plate tracery, 
 469. 
 
 BELLEFONTAINE. Vaulting, diagonal 
 ribs, 300. 
 
 BERE. Window mullions, 516. 
 
INDEX OF PLACES. 
 
 ; ii 
 
 BERNAY. Capitals, Norman, acanthus in, 
 427. Choir, short, 1.S4. Clerestory, no 
 passage, 545, note. Domes, 281. Moldings, 
 roll, 277. Piers, compound, 234. String 
 courses, 404, Triforium, 532. note. Vaults 
 of choir aisles, 292 ; no high vault, 290. 
 
 BERNIERES. Triforium, 532, note. /W/, 
 quasi-sexpartite, 32r. 
 
 BETHERSDEN. /tfarWe from, used at 
 Canterbury, 250. 
 
 BETHLEHEM. Basilica of the Nativity. 
 Apses, three parallel eastern, 159. 
 
 BEVERLEY. Minster. .-/«/«, double, 
 transept, 198. Arches, Gothic semicircular, 
 transept doorway, 261 : ogee, 270. Bases, 
 453. Capitals, molded, 434, 446. Corbel 
 table, 394 : horizontal base, 392. Cusps, 
 Percy tomb, 270, 511. Foliage, reredos, 
 back of, 436. Fronts, west, 91 : east, 68 : 
 transepts, 68, 71, 405. Parapets, 395; 
 choir, 394 ; nave, figures on, 396. Piers, 
 "Northern,'' transept, 253. Proportion, 
 54. Reredos, see Foliage, Tracery, and 
 Shafts. Sanctuary, unaisled, 176. Shafts, 
 250. Tombs, Fitzalan, 132: Percy, arch, 
 ogee, 270; cusps, 511; pinnacles, 364. 
 Towers, effect of, 594 : group of, 594 : 
 lantern, 593: west, 91. Tracery, Flam- 
 boyant, 4S5, 488, 490 : leafed stem, 
 485, 490. Transepts, aisles, double, 198: 
 eastern transepts, 1S9. Triforium, blank 
 •with double arcade, 535. Vaulting, nave, 
 material, 304 : height of spring of, 307 : 
 quadripartite, 82. Windows, circular, 516: 
 lancet (transepts), 463 : leafed stem (nave 
 aisles), 485, 490 : vesica piscis (north tran- 
 sept), 517 (see Tracery). St Mary's. 
 Arch-molds, 277 ; chancel, 280: nave, 280. 
 Bases, nave piers, 447, 453 : bases and 
 plinths, 450. Buttresses, gabled, 367. Door, 
 fifteenth century, 585. Doorway, 580. 
 Dripstone, ogee, crocketted with finial, 407. 
 Hood-mold, 406. Parapet, 396, 39S. Pin- 
 nacles, 362 : of porch, set diagonally, spear- 
 headed, 364 : of tower, 60S. Porch, 582. 
 West window, 504. 
 
 BEZIERS. St Aphrodisie. Battlements, 
 399. Piers, 236. 
 
 BIBURY. Capitals, 414, note. 
 
 BIELLA. Baptistery. Buttresses : (Roman), 
 35 1. Doorway, 573. 
 
 BILLINGHAM. Plate tracery, 465. 
 
 BINGHAM. Spire, 623: spire lights, ... 
 rangement of, 628. Tower and »/''". 
 unity of, 632. 
 
 BINHAM. {References, 170, note; Bar 
 tracery, 469, 470, 472. Cusps, 511, note. 
 Difference of work, 46. Dripstone, 406. 
 Growth, very slow, 46. Mortar, 24. 
 String courses, 406. Triforium arcade, 
 527: single arches of, 532. West /rout, 
 
 BIRKIN. Doorway, Norman, rebuilt, 579. 
 Pilaster strips, 358. 
 
 BISHOP'S CLEEVE. Pinnacles, 362. 
 
 BISHOPS TONE. Perforated parapet, 396. 
 
 BLYTH. Arches, square-edged, 2>?> Clere- 
 story, 29 : no passage, 545. Triforium 
 arcade, single arches, 532. Vault, low 
 spring of, 374. 
 
 BLYTHBURGH. Nave, angels as pin- 
 nacles, 366. Presbytery, lighting of, 226. 
 
 BOARHUNT. Tower, Anglo-Saxon, de- 
 velopment of, 596. 
 
 BOLOGNA. S. Petroxio. Vaulting, 321. 
 
 S. Stefaxo. Capitals and abaci, 410. 
 Piers, alternation of, 243 : piers and 
 columns, 243. S. Theodore. Capitals 
 and abaci, 4 1 o. 
 BOLTON. Priory. Transept arches, 
 pointed, 265. West front, 73. 
 
 BORDEAUX. Nave, hall plan, 321, note. 
 
 Tower, 590. 
 BO ROUND. Roofs, pitch of, 3S9, 559. 
 
 BOSCHERVILLE 
 
 DE. 
 BOSHAM. 
 
 237- 
 BOSTON. 
 story, 546 
 
 /-'lunacies, 
 
 See ST GEORGES 
 
 Piers, Anglo-Saxon, compound, 
 
 Advanced type, 132, 
 
 Clere- 
 
 windows, two in each bay, 82. 
 
 364. Plan, 80 : aisled nave, 
 223: unaisled sanctuary, 176. Tower, 
 position of, 596. 
 
 BOTHYVELL. Barrel Vault, 284. 
 
 BOTTESFORD. Windows, narrow lancet, 
 462. 
 
 BOTTISHAM. Arch moldings, 2S0. Bases, 
 
 45-- 
 BOURGES. Buttress chapels, 197, note. 
 
 Crypt, 194. Flying buttresses, no pinnacles, 
 
 363. Transepts, none, 593. Western 
 
 porches, 205. 
 BOXFORD. Porch, wooden, 3S5. 
 
742 
 
 INDEX OF PLACES. 
 
 BOXGROVE. Arch, straight sided, 259. 
 Bases, 449 : Attic, 451. Capitals, molded, 
 446. Choir, unaisled parallelogram, 179. 
 Clerestory, single lancets, 463. Corbel-table, 
 392. East front, 68. Flying buttresses, 
 55> 372, 374: heavy, 375, 3 80 : to P sur- 
 face of, 377, 380. Piers, alternating, 317, 
 note: "Southern," 246. Styles, melange of, 
 46. Triforium, 534. Vaults, square, 320. 
 
 BOYTON. Circular window, 5 1 7. 
 
 BRACEBRIDGE. Capitals, scalloped, 
 Anglo-Saxon belfry, 413. 
 
 BRADFORD-ON-AYON. Anglo-Saxon 
 Church. Pilaster-strips, 351. Transepts, 
 222. Barn. Roofs of Aisles, 225. 
 
 BRADSOLE. St Radegund's Priory. 
 Flying buttresses, 372, note. Unaisled 
 sanctuary, 176. 
 
 BRAMPTON. Broach spire, 620. 
 
 BRANSTON. Acanthus in Anglo-Saxon 
 work, 427. Capitals, Anglo-Saxon, scal- 
 loped, 413. 
 
 BRAUNTON. Wooden broach spire, 620. 
 
 BREAMORE. Tower, Anglo-Saxon, cen- 
 tral, 590. 
 
 BRECON. Priory Church. Lancets, 
 choir, 463. 
 
 BREDON. Pinnacles, 362. Porch, vault- 
 ing, material of, 304. Toiver, Norman, 
 central, 590. 
 
 BRENTINGBY. Tower, saddle-back roof 
 with spirelet, 602, note. 
 
 BRERA. Museum. Capitals and abaci, 
 from AURONA, 410. 
 
 BRESCIA. Older Cathedral. Groined 
 crypt, 292. Temple of Vespasian. Com- 
 pound piers, 236. 
 
 BRIDGWATER. Pierced parapets, 397. 
 Tower and spire, proportion of, 631. 
 
 BRIDLINGTON. Aisles, roofs, 538, 540. 
 Basement-courses, 402 : coupe-larme or 
 hollow throat, 404: shadows in, 403. But- 
 tresses, gabled, 367. Capitals, porch natu- 
 ralistic foliage, 435 : molded, 446. Clere- 
 story, interior pierced parapet, 397. Front, 
 west, towers, 91. Nave, no high vault, 82 : 
 triforium, north side, too large, single con- 
 taining arch, 533: south side, absorbed into 
 clerestory, 536. Strings, fragile undercut, 
 throat, 405, 406. 
 
 BRIGSTOCK. Tower, Anglo-Saxon newel 
 
 stair turret, 601. 
 BRINKBURN. Arches, pointed, 103. 
 
 Nave, one-aisled, 201. 
 BRINTON. Roof, arched braces, 563. 
 
 BRISTOL Cathedral. Aisles, high, with 
 skeleton vaulting, 340, 382. Berkeley 
 chapel, large leaf foliage, 437 : skeleton 
 vaulting, 340. Choir, former Norman, 
 square ended, 164, note: vaulting oppos- 
 ing thrusts, 340, 374, 382. Roof, three- 
 span, 570. Sanctuary, unaisled, 176. 
 Styles, melange of, 46. Tower, central, 593. 
 Transepts, 46. Vault, see Aisles, Berkeley 
 chapel, and Choir : web, thickness of, 304. 
 Window, east, 490, 499. City Schools. 
 Doorway, arch, four - centred, 267. St 
 James'. Piers, 236. Window, circular, 
 516. St Mary Redcliffe. Aisle, eastern, 
 172. Bases, 454. Capitals in porch, 432. 
 Clerestory, no passage, 545 : wall, thin- 
 ness of, 60, note. Construction, skeleton, 
 60. Flying buttresses, crocketted, 377. 
 Lady chapel, 172: east front, 9 1 . Para- 
 pets, perforated, 396. Porch, foliage of 
 capitals, 432. Remodelled in fifteenth cen- 
 tury, 138, 172. Spire, bands round, 626. 
 Tower, central, lower courses of, 599, 
 note : western, position of, 596 ; thick- 
 ness of walls of, 602. Transepts, 223: 
 altars in, 197 : double aisles, 198. Tri- 
 forium, panelled wall, 88, 535. Vault, 
 82 : vaulting pier, 540 : vaulting shafts, 240 : 
 web, thickness of, rubble work covered 
 with concrete, 304. Windows, curvilinear, 
 4.83, 490 : vertical tracery with geometrical 
 work, 397. St Stephen. Chancel arch, 
 absent, 227. Doorivay, removed, 579. 
 Tower, 610: battlements, 606 : buttresses 
 of, 606: design of, 610: oversailing 
 parapet, 605 : pinnacles, 606 : strings, 606. 
 
 BRITTANY. Architecture of, 132. 
 
 BRIXWORTH. Apse, Anglo-Saxon, 156: 
 Anglo-Saxon buttresses to, 351 : polygonal, 
 220: roof of, 219. Arches, compound, 
 Anglo-Saxon, 272: Anglo - Saxon pier 
 arcade, 234. Plan, threefold division, 
 221. Tower, Anglo-Saxon, 590: newel 
 staircase to, 601. Western transept, 203. 
 Windows, large Romanesque, 458. 
 
 BROADWATER. Foliage, character of, 
 432 : transitional, 418 and note ; 419, note. 
 
 BROU-EN-BRESSE. French Flamboyant, 
 Flemish influence on, 128, note. Stained 
 glass, 493, note. 
 
NDEX OF PLACES. 
 
 74 5 
 
 BROUGHTON. Piers, Anglo-Saxon com- 
 pound, 237. Tower, Anglo-Saxon, newel 
 staircase, 601. 
 
 BRUTON. Three-storied porch, 5S4. 
 BUCKNELL. Choir, 221. 
 
 BUILDWAS. Abaci, octagonal, 37, 439: 
 chamfered, 441. Chapter house, vaulting 
 of, rib moldings, 301 : French method 
 of filling webs, 335. Sanctuary, unaisled, 
 176. 
 
 BURGOS. Flying buttresses, used as aque- 
 ducts, 377, note. 
 
 BURGUNDY. Pointed arch, 264. 
 
 BURNHAM OVERY. Aisle, early Eng- 
 lish, south chancel, 226. 
 
 BURTON LATIMER. Spire, 619. 
 
 BURY ST EDMUNDS. Abbey. Area, 15, 
 
 note. Choir, 184. Masonry, 20. Nave, 
 1S4 : length of, 53: span of, 15. Peri- 
 apsidal plan, 167. Towers, western, posi- 
 tion of, 598. Transepts, 16: eastern 
 aisle to, with apses, 198: western transepts, 
 16. Western vestibule, 204. St Mary. 
 Hammerbeam roof, 566: longitudinal braces, 
 569- 
 
 YEAND. Abacus, transitional, 441. Ad- 
 vanced work, 103. Ambulatory, rectangu- 
 lar, 104, 179. Aisle, eastern, 172. Arch- 
 moldings, 279. Choir, long, 1 79. Clerestory 
 passage, 545. Doorway, trefoil headed, 
 580. Foliage, interlacing, 433. Front, 
 
 • west, 73. Galilee, 202. Mo/dings, 279. 
 Narthex, 202. Plan, simplified Cister- 
 cian, 171. Strings, 406. Transept, double 
 aisled, 198. Triforium arcade, no con- 
 taining arch, 534. Windows, rose, 517: 
 semicircular headed, 462, note. 
 
 BYTHORNE. Spire, 620. 
 
 CAEN. Stone, 20. Abbaye-aux-Dames. 
 Apses, three parallel eastern, 163. Arch, 
 elliptical, 262, 310: pointed, 265: semi- 
 elliptical, 310. Bases, Norman, 450. Capi- 
 tals, foliage, 427 : scalloped capitals, 413. 
 Chapels, elongated in transepts, 167. 
 Flying buttresses, internal, 370. Gallery, 
 between western towers, 204. Triforium, 
 arcade, no containing arch, 534: internal 
 flying buttresses, 370 : vault, thrust of, 
 376, note: windows not adopted, 532, 
 note. Vaulting, groined, nave, aisles and 
 crypt, 292 : of choir, 294 : high vault not 
 originally intended, 290 : rebuilt, quasi- 
 
 sexpartite, 321 : thi< kne \ of, ; 1 1 Abbayj 
 u \ I [ommi s. Area, 15. note. Bases, 
 Norman, 450, 451, note. Canterbury 
 Cathedral, modelled on, 98. Capital, 
 south clerestory, with interlaced pipings, 
 416. Ceilings, wooden, 312. Elevation, 
 interior, three-storied, 530. Flying but- 
 tresses, choir, 372, note. Front, west, 29: 
 towers, 73 ; with gallery between, 204. 
 Piers, alternating, 3 1 7, note: and detai hed 
 shafts (triforium), 237. Pinnacles of spire, 
 623. Principals, former stone, 235. Siring 
 courses, 404. Towers, western, 73: gallery 
 between, 204. Transepts, return aisles, 542 : 
 three parallel apses, 163. Triforium, piers 
 and detached shafts, 237 : use of, 542. 
 Vault, aisles groined, 292: high vault, 312 : 
 not originally intended, 290 : nave, upper 
 aisles of, half barrel, 285 : ribbed vaults, 13 : 
 rib moldings, 301 : voussoirs of ribs, 301 : 
 sexpartite, 320. Windows, triforium, 532. 
 St Etienne-le-Vieux. Central tower, 
 construction of inner shell of, 602. St 
 Gilles. Pier arcades, Gothic semicircular 
 arch, 261. St Nicholas. Apses, three 
 parallel eastern, 163. Capitals, volutes 
 and consoles in, 427. Narthex, 202, note. 
 Piers and detached shafts, 237. Transept, 
 return or end aisles, 19S. Triforium, 
 532, note. Vault, groined, 13, 292,316: 
 ribbed, 13. St Pierre. Licrne vault, 
 342. St Stephen'. Short choir, 184. 
 Transepts with return or end aisles, 198. 
 
 CALDICOTT. Curvilinear windows, 487. 
 
 CAMBRIDGE. Jesus' College Chapel 
 (St Radegund). Central tower, fall of, 
 594, note. King's College Chapel. 
 Aisleless, 228, 229. Buttresses, 361 : im- 
 portance of, 367. Buttress-chapels, 197. 
 Clerestory, 543. Construction, skeleton, 
 61. Down pipes, 401. Parapet, battle- 
 ment and pinnacles, 367, 541 : perforated 
 parapet, 399. Pinnacles, 367 : with ogee 
 cupolas, 366. Stained glass, 61, 493, note. 
 Tracery, geometrical, 472, 504. Vault, 
 fan, 346 : vault pier, 540. St Michael. 
 Transitional window, 499. St Sepul- 
 chre. Nebule corbel-table, 392. 
 
 CANTERBURY. Cathedral. Abacus, 
 chamfered, 37. Abutment, 54. Acanthus 
 leaf, 426. Aisles, choir, 389. Anglo- 
 Saxon columns, from Reculvers, 234. 
 Apses, plan of, 16S, 200. Arches, ogee, 
 270, note: strainer, 381: transverse in 
 triforium, 369. 375. Bases, 451. Bosses, 
 305. Buttresses, turretted, tower, 605. 
 
744 
 
 INDEX OF PLACES. 
 
 Canopies of tombs, 342. Capitals — 
 crocket capitals, 23, 429 and note : 
 cushion capitals, 409, 413 : foliage of 
 capitals, 418, 427, 434: molded capitals, 
 44 r. Chapels, apsidal, 65, 168: oblong 
 and rectangular, 168: radiating, 167: 
 Saint's, 183, 185, 186, 193: St Michael's 
 or Warrior's, 138: Trinity, 186, 232, note. 
 Choir, Ernulph's, 167, 187 : Lanfranc's, 
 short, 184, lengthened, 16, 185 ; lead from 
 former roof, 389 : choir rebuilt, 108 : 
 raised, with high flights of steps, 191. 
 Clerestory, 536, 543; no passage (nave), 
 545. Columns, 232, 234. Corona, 186. 
 Crypt, 191, 193, 256. Flying buttresses, 30, 
 54, 368, 369, 371, 375, 377. Fronts, east 
 transept, 71: west front, 91. Gothic, 
 juxtaposition of early and late, 46. Lan- 
 tern, 593. Nave, Norman, size of, 15. 
 Panelling, 608. Parapet, lead, 394, note. 
 Piers, central, 594 : crypt, 246 : nave, 540. 
 Pinnacles, 362, 606, 608. Pulpitum, 179. 
 Roofs, 389. Screens, 179, 182, 270, 381. 
 Shafts, 256: marble, 249, 250. Shaft-bands, 
 249. Stained glass, 479. Steps, 1 9 1 . Strings, 
 Norman, 405. Styles, melange of, 46. 
 Towers — Central tower, 593 and note; in- 
 terior effect of, 603. Choir towers, 599. 
 Western towers, 91. Transepts, 15 : aisle- 
 less, 200. Choir transepts, 16. Eastern 
 transepts, fronts of, 71. Lanfranc's, return 
 aisles, 198. Triforium, 519, note: choir, 
 two containing arches, 534: transverse 
 arches, 369, 375. Nave, 88 : absorbed into 
 clerestory, 536. Trinity chapel, no con- 
 taining arch, 534. Vaulting, material, chalk, 
 and tufa, 304. Choir, length of ribs, 16 ; 
 ribs ornamented, 301 : shafts, 240. Cloister, 
 337. Crypt, eastern, ribs, 268 ; spring of 
 vaulting, 343. Dean's chapel, fan vaulting, 
 344, 593. Lanfranc's tower, springers of 
 vault, 303. Nave — liernevaulting, 341 ; ribs, 
 301 ; shafts, 240, 540. Octopartite vault- 
 ing (treasury), 339. Sexpartite vaulting, 
 320. Tomb of Archbishop Stratford, 342. 
 Tower, central, fan vaulting, 344, 593 ; 
 springers, 303. Transepts — eastern, bosses, 
 305 • north-eastern, short voussoirs to ribs, 
 301. Transition vaults, height of springing 
 of, 307. Treasury, octopartite vaulting, 
 339; ribs, moldings of, 301. ]}'indows, 
 French lancets, 462 : Kentish tracery, 476, 
 Perpendicular, 497. Parish Churches. 
 Trussed rafter roofs, 561. St Augustine's 
 Abbey. Apse, periapsidal plan, 167. 
 Choir, 184. Norman nave, 184. Tran- 
 septs, apsidal chapels, 197. St Augus- 
 
 tine's Gateway. Vaulting, material, 
 304; sexpartite, 320. St Martin. Anglo- 
 Saxon buttresses, 351. Stop of hood-mold, 
 408. St Pancras. Anglo-Saxon work, 
 apse, 156; buttresses, 351 ; transepts, 222 ; 
 western porch, 202. St Peter. Groined 
 aisle, 30. 
 
 CANWICK. Norman abacus, 441. 
 
 CAPEL ST MARY. Hammerbeam roof, 
 564, note, 566 : cornice of, 568. 
 
 CARCASSONE. St Vincent. Hall plan, 
 321. 
 
 CARLISLE. Cathedral. Aisle (choir) 
 arcading, cinquefoil arch, 268 : vaulting, 
 335 : (nave) Norman pilaster strips, 358. 
 Buttresses (choir), gabled, 367. Capitals 
 (choir), naturalistic foliage, 436. Choir, 
 11, 179; no vault, 118, but barrel roof, 
 569 ; hammerbeams, 564, note. Corbel- 
 table,^^. Fast front, 82: parapet of, 398. 
 Gothic growth, 49; melange of styles, 46. 
 Nave and aisles, unvaulted, 20. Triforium, 
 one arched arcade, 532. Windows — curvi- 
 linear, 479, 488, 490, 494 : east, 506 : plate 
 tracery, 465. 
 
 CARTMEL. Lantern on central tower, 
 600. Unaisled sanctuary, 176. 
 
 CASAMARI. Cistercian plan, 16. Vault- 
 ing, ridge rib, 335. 
 
 CASHEL. St Cormac's Chapel. Barrel 
 vault, 284. The Seven Churches, 207. 
 
 CASSINGTON. Choir, 221: vaulted 
 chancel, 550; vaulted sanctuary, 221. 
 
 CASTLE ACRE. Apses, three parallel 
 eastern, 164 : apsidal chapels (transepts), 
 197. Alternation of piers, 243. Shaft 
 ornaments, 256, note. West front, 28, 73. 
 West toivers, 73. 
 
 CASTLE ASHBY. Norman porch, 582. 
 
 CASTLE HEDINGHAM. Arch, 279. 
 Bases, Attic, 45 1 : Norman, 450. Circular 
 window, 468. 
 
 CASTLE RISING. Pood screen and altars, 
 181, note. West front, 29. Window, 
 cusped, Norman, 509. 
 
 CASTOR. Norman door, with original lock 
 and key, 585. 
 
 CAWSTON. Pier arcade, no hood-mold, 
 407. Roof — hammerbeam, 565, 566: 
 lead covered eaves, 392 : longitudinal 
 braces, 569. Tower — basement-course, 
 panelled, 403 : buttresses not carried to 
 top, 606 : horizontal sky-line, 606 : little 
 ornament, 608. Vault, piers, 540. 
 
INDEX OF PLACES. 
 
 745 
 
 Spire, entasis, "sugar 
 
 CAYTHORPE. 
 loaf," 632. 
 
 CERISY-LA-FORET. Abacus strings, 404. 
 Aisles, square ended, 163, 1(14. Apse, cen 
 tral, 163, 164: lighting of, 175. Arches, 
 Norman, 276. Canterbury, modelled on, 
 98. Choir, short, 184. Elevation, interior. 
 three-storied, 530. East front, 28. Piers 
 and stone principals, 235 : piers and de- 
 tached shafts, 237. Precedent of, 100. 
 Transepts with return aisles, 198, 542. 
 Triforium — arcade, 533 : use of, 542 : 
 windows, 532. Vaults, groined (aisles), 
 292. 294, 295, note : high vault not origi- 
 nally intended, 290. 
 
 CHARTHAM. Kentish tracer, 476. 
 
 CHALONS-SUR-MARNE. Battlements, 
 399- 
 
 CHARTRES. Confessio or crypt, 191. 
 Interior elevation, three storied, 529. 
 Spires, 617, 621 : tracer)-, 628: walls of, 
 thickness of, 636. Towers, group of, 600. 
 1 r aulting, thickness of, 304 : pointed dia- 
 gonals, 322. Windmvs, circular, 517: 
 plate tracery, 517, 468, 469. 
 
 CHAUMONT. Lierne vaulting, 342. 
 
 CHECKENDOX. Apse, -with semi-dome, 
 20. 
 
 CHELMSFORD. Arch - moldings, 2S0. 
 Porch, molded capitals, 444, 445. 
 
 CHELTENHAM. Capitals with incurved 
 cones, 414. Windows, 500. 
 
 CHESTER. Cathedral. Aisle, nave, 
 north-west corner without windows, 530 : 
 wall raised for windows, 530. Apse, former, 
 167 : periapsidal plan, 167 : apsidal 
 chapels, 197. Arch, unmolded Norman, 
 276. Capitals, absence of, in chapter 
 house, 409 : (nave) foliage of, 437. 
 Chapter house, absence of capitals, con- 
 tinuous molding, 409. Clerestory, nave, 
 absorption of triforium into, So, 536 : 
 passage, 545, note. East front, 68. 
 Flying buttresses (nave), 369. Gothic, 
 "Western school"' of, original Gothic 
 choir, 105 : melange of styles, 45. Lady 
 chapel, 68 : no foundations, 25. Mold- 
 ings, continuous (chapter house), 409. 
 Plan, periapsidal, 167, 197. Saint's 
 chapel (St Werburgh's), 183. Styles, 
 melange of, 45. Transepts, one Roman- 
 esque, the other Gothic, 45 : south tran- 
 sept, double aisled, 198. Triforium— chain, 
 blank arcade, 535 : round-headed trefoiled 
 
 arch, 268 : nave, absorbed into 1 lerestory, 
 ■So, 536 : presbytery, arcade, in foili d 
 arches, no containing an h, 534 : transept, 
 north, continuous arcade, 534. Vaulting, 
 ribs, 300, note. Chapter house, tierce 
 rons i 337. M'): note: ridge rib, 339, noti . 
 
 Nave, vaulting shaft, 240. Windcnos, 
 
 metrical, 477: north aisle ol nave, 530. 
 St John. Abacus prolonged as finish to 
 dripstone, 407. Arches, square edged, 33: 
 unmolded Norman, 27(1. liases, 447 : with 
 flattened roll, 451. Dripstone or hood 
 mold, Norman, 406 : finish of, 407. Foli 
 age, interlacing, 432. Co/hi,; melange ol 
 styles, 46 : slow growth, 46. Triforium, 
 blank arcade, 535. no containing arch, 534. 
 Si Werburgh. South transept, no high 
 vault, 82. 
 
 CHICHESTER. Abutment, 30, 55. Aisle, 
 double, 156. Apsidal chapels, transept, 
 197. Arch, square-edged Norman, 276: 
 moldings, t,^' 2 7$ '■ arches in triforium 
 chamber, 30, 369, 370. Buttresses, aisles, 
 363 : polygonal, with pinnacles, 363. 
 Campanile or bell tower detached, 598. 
 Capitals — Norman Corinthianesque, 42S : 
 crocket capitals (retrochoir), 429: propor- 
 tion of, 409. Chapels — apsidal (transept), 
 197: buttress, 197: Lady chapel, 68: 
 rectangular eastern, 172. Cusps, nave 
 windows, 570. Flying buttresses (choir), 
 55 : (nave) as aqueducts, 400 : heavy, 
 374: superposed, 378: triforium, 371, 
 372. Foliage, interlacing;, 433. Front, 
 east, 68. Gothic — Melange of styles, 45. 
 Eastern limb extended in, 45. Retrochoir 
 rebuilt in early Gothic. 45, 109. Masonry, 
 20, 23, note. Moldings, arch, 33, 278 : 
 roll. 33. Pier, retrochoir, 246. Pin- 
 nacles, nave, aisle, 363 : of spire. 624. 
 Plan, periapsidal, 167. Porch, western, 
 205 : three porches, 584. Retrochoir, 45, 
 46, rog. Spire, 593, 598 : angle of, 631 : 
 collapse of, 594 : pinnacles, 624. Stalls 
 in nave, 185. Styles, melange of, 45. 
 Tower, detached, 59S : central, 593, 594, 
 59S. Transept, apsidal chapels, 197. 
 Triforium, 519: arcade, single containing 
 arch, 533 : flying buttresses, 371 : trans- 
 verse arches, 369, 370. Vaulting, ashlar, 
 299, note : ribbed (south aisle), 298 : wall 
 ribs, 300 : moldings, 301 : ridge ribs 
 (nave), 336 : spring of, height of, 307 : web, 
 French method of filling, 335. II indoles, 
 cusps, 510. 
 
 CHIPPING HILL. Capitals, typical four 
 teenth century, 443, 444. 
 
746 
 
 INDEX OF PLACES. 
 
 CHIPPING NORTON. Capitals and con- 
 tinuous moldings, 409. Clerestory, 547. 
 Vault piers, 540. Windows, east, over 
 chancel arch, 547 : Flamboyant, 485, 487, 
 488, 490. 
 
 CHIVY, near Laon. Stone prijicipals over 
 aisles, 235. 
 
 CHRISTCHURCH. East front, 91. Lady 
 chapel, 91. Moldings of pier arches, 33: 
 roll moldings, 277. Reredos, 181. ,S«»ff- 
 tuary, unaisled 176. Styles, melange of, 
 46. Tower, 599. Transept, apsidal chapels, 
 197. Triforium, choir, 88. Vault, Salis- 
 bury chantry chapel, 345. 
 
 CIRENCESTER. Bases, 454- Clerestory, 
 547. Plinth, octagonal, 450. Porch, three 
 stories, 584. 7?^^ Trinity chapel, 562. 
 Vault, fan, inner aisle, 346. Window, 
 east, over chancel arch, 547. 
 
 CITEAUX, 16. 
 
 CIVIDALE. Early groined vaulting, 292. 
 
 CLAIRVAUX, 157. 
 
 CLEE. Arch, Anglo-Saxon, compound, 272. 
 Scalloped capitals, Anglo-Saxon belfry, 413. 
 
 CLERMONT. Sixth-century transept, 196. 
 Central lantern tower, 592. 
 
 CLERMONT FERRANU. Central tower, 
 590. Triforium, 537. 
 
 CLEY, 132. Fourteenth-century door, 485. 
 Doorway, gable, hood-mold or dripstone, 
 4°7> 579- 
 
 CLUNY. Abbey. Aisles, double, 156. 
 Area, 15, note. Narthex, 202. JPeri- 
 apsidal plan, 169. Roofs, over vaults, 
 388. Transepts — apsidal chapels, 197 : 
 eastern transept, 189. Vault, barrel, 388. 
 Museum. Storied capitals, 416. 
 
 CLYMPING. Hybrid roof, 569. Shuttered 
 windows, 459. 
 
 COLCHESTER. StBotolph. ^^square- 
 edged Norman, 276. Bases, 450, 454. 
 Clerestory, absorption of triforium into, 
 536. Internal elevation, and triforium, 
 2 7i 536- West transept and towers, 204. 
 
 COLLUMPTON. South chantry chapel, 
 224 : fan vaulting, 346. 
 
 COMO. S. Abbondio. Apses, 163. Basi- 
 lica, 232. Capitals and abaci, 410. 
 Column-buttress, 351. Plan, "crux com- 
 missa " formerly, 195. Position of choir 
 towers, 599. 
 
 COMPTON. Norman screen, 226, note. 
 
 COMPTON BASSETT. Rood screen, 181, 
 note. 
 
 COMPTON MARTIN. Vaulted sanctuary, 
 221. 
 
 CONISBOROUGH. Keep. Fireplace, jog- 
 gled arch, 260. 
 
 CONSTANCE. Basilica, 232. 
 
 CONSTANTINOPLE. Aqueduct, near. 
 Pointed arches, 263. Basilica of St John 
 Studius, 231, 233. S. Sophia. Acanthus, 
 426. Arches, orders in, 274. Atrium, 
 201. Dome, 281. Groined vaults, 309. 
 
 CORBRIDGE. Tower, Anglo-Saxon, de- 
 velopment of, 596. 
 
 CORHAMPTON. Pre-Conquest church, 220. 
 
 COTTINGHAM. Hood-mold; 406. 
 
 COUTANCES. Central tower, 602. Vaulted 
 lantern, 593. 
 
 COVEHITHE. Presbytery, 226. 
 
 COVENTRY. St Michael's. South door- 
 way, round-headed trefoil arch, 268. Spire 
 — construction, 636: flying buttresses, 377, 
 624: lights, 627 : on octagonal drum, 629 : 
 squinches, 636. Tower and spire — distri- 
 bution of ornament, 633 : proportion of, 
 629: tower versus spire, 632. Churches. 
 Clerestory wall, skeleton construction, 60, 
 note. 
 
 COXWOLD. Octagonal tower, 600. 
 
 CREIL. Triforium, Romanesque interior 
 flying buttresses, 371. 
 
 CREULLY. Unaisled vaulted choir, no 
 buttresses, 362. 
 
 CREWKERNE. West front, 91. 
 
 CRICK. Spire, 620. 
 
 CROMER. Ceremonial western doorway, 
 with recess for processional cross, 581, 
 note. Pier, moldings, 256. 
 
 CROWCOMBE. Jacobean screens and fit- 
 tings, 203, note. 
 
 CROWLAND or CROYLAND. Founda- 
 tions, 25, Masonry, 20. 
 
 CROYDON PALACE. Arched braces of 
 roof, 562. 
 
 CROXDEN, 157. Periapsidal plan, 167. 
 
 CTESIPHON, Palace of. Pointed arches, 
 263. 
 
 CULLOMPTON. See COLLUMPTON. 
 
 CWM HIR. See ABBEY CWM-HIR. 
 
INDEX OF PLACES. 
 
 747 
 
 DAGLINGWORTH. Tower, Anglo-Saxon, 
 
 development of, 596. 
 
 DAMAS. Arch. Compound pier, 236. 
 
 DAMASCUS. Pkopye.ea. Arch on columns, 
 233, note. 
 
 DANA. Aisled basilica, horseshoe arch, 
 262. 
 
 DARENT. Priests chamber, 221. / aulted 
 
 sanctuary, 221. 
 
 DARENTH. Chance/, vaulted, 550. East 
 front, Norman, rectangular, 28. II indoles, 
 glazing of, 459. 
 
 DARLINGTON. Clerestory, 68, 546. 
 Marble shafts, 250 and note. 
 
 DEENE. Parapctted spire, 622, note. 
 
 DEEPING. See Market Deeping. 
 
 DEEPING ST JAMES. Capital, 442. 
 
 Clerestory passage, 547. "Northern'" pier, 
 
 253- 
 DEERHURST. Characler of foliage, 432. 
 
 Transepts, 222. 
 
 DENDERA. Compound arches, 272, note. 
 
 DENFORD. Spire, 622: band round, 626: 
 arrangement of lights, 628. 
 
 DENNINGTON. Bases and benches, 447. 
 
 DEREHAM. Porch, holy-water stoup, 207. 
 
 DESBOROUGH. Spire, 623. 
 
 DEVIZES. St Mary. Elliptical arch, 
 262. Semi-elliptical diagonals, 310. Nor- 
 man high vault, 314. St John. Norman 
 high vault, 314. 
 
 DEVONSHIRE. Screens, 1S1, note. 
 
 DIARBEKIR. Pointed arch, 263. 
 
 DIJON. St Benigne, 317, note. 
 
 DONCASTER. Central tower, 591. 
 
 DONTNGTON. Position of tower, 596. 
 
 DORCHESTER. Nave, one-aisled, 201. 
 Roof, lowered, 391. Window, east, head 
 cut off by roof, 391. 
 
 DORE. See ABBEY DORE. 
 
 DOVER. St Mary in the Cam it;. 
 Anglo-Saxon tower, 590. Transepts, 222. 
 St Martin-le-Grand. Periapsidal plan, 
 167. St Martin's Priory. Apsidal 
 aisles, 164. Rectangular choir, 98, 164. 
 
 DUBLIN. Christ Church Cathedral. 
 Clerestory, absorption of triforium into, 
 536. Gothic, "Western school" of, 105. 
 
 Mas, 
 
 tonry, 20. Pio r, " U estern, ■ 1 
 Purbeck shafts, 252. Vaulting, heighl ..1 
 spring of, 307. Si I'm rick's Cathi m vi . 
 Aisle, east, in, 172. Piers, "Southern," 
 255. Plan, inspired by Salisbury, 118, 172. 
 
 DUNDRY. Tower, design of top, 606: 
 oversailing parapet, '105. 
 
 DUNFERMLINE. Abaan strings, intei 
 
 nal, 404. '•Durham" school, 100. Shaft 
 ornament, 250. Triforium, windows of, 
 53-- 
 DUNHAM MACNA. Choir, 22,. 
 
 DUNSTABLE. Priory. Altar, Jesus, 180. 
 Arches, elliptical, 2(12 : round, not pointed, 
 retrogressive, 101. Screen, 180. Towers, 
 two western, fall of, 594, note. Triforium, 
 arcade, 527 : absorbed into pier arcade 
 535. Vaulting, semi-elliptical diagonals, 
 310. 
 
 DURHAM. Cathedral. Abacus, octagonal, 
 37, 439. Abacus strings, clerestory, 404. 
 Abutment to walls, 30. Apses, three 
 parallel eastern ; two lateral square out- 
 side ; foundations of, 163. Area, 50, note; 
 Arches — Galilee, 103 ; semicircular, 266 : 
 choir, in triforium chamber, 30 : nave, 
 transverse arches in triforium chamber, 
 370, 374: pointed arch in vaulting, 321, 
 transverse arches omitted between each 
 pair of bays, 316. Bases, nave, 37 : east 
 transept, 447. Capitals, molded, east 
 transept, 443: foliage of capitals, 431. 
 Chapels, Neville, 211: Saint's (St Cuth- 
 bert), 183. Choir — 184; length of, 16. 
 Clerestory, abacus strings, 404. Construc- 
 tion, 60 ; primacy of, 98, 99 and note. 
 Elevation, internal, 26. Flying buttresses, 
 choir, 30: nave, no external, in twelfth cen- 
 tury, 372 : triforium, transverse arches in, 
 369, 370. Foundations, 25 : of apses, 
 163 Fronts, east, 68: west, 28, 71, 73. 
 Galilee, 103, 202. Nave, length of, 16. 
 Piers, alternation of, 243, 317, note: com- 
 pound and cylindrical, 33, 37 : separate 
 shafts, 37 : detached shafts, 237 and note : 
 "Southern," 250, 253: thickness of, 24. 
 Torch, Norman, 582. Reredos, 87. Roman- 
 esque, 10, 45, 46. Roof, stone principals, 
 235, note. Shafts, detached, 237 and 
 note: marble, 250: ornaments of, 236: 
 separate, 37. Spires, 619. Styles, melange 
 of, 45, 46. Tomb, curvilinear, 501. Towers, 
 central and former single western, 600 : 
 western towers, 73. Transepts, east aisle to, 
 16, 197: east transept, 68, 189. Triforium, 
 
74 8 
 
 INDEX OF PLACES. 
 
 arcade, single containing arch, 533: arches, 
 transverse, 369, 370, 374 : internal eleva- 
 tion, 26. Vaulting, 314: pointed arch in, 
 321: Romanesque, 10: ribbed, 13, 298. 
 Aisles (choir), diagonal ribs, 300 ; short 
 voussoirs of ribs, 301 : (nave), 300, oblong 
 bays, 312; stilted transverse arch, 311; 
 segmental diagonals, 311; keystone vault, 
 305 : (transept), 16, 197. Choir, ribbed, 
 298. Nave, pointed arch, 321. Tran- 
 septs, sexpartite, 320. Walls, abutment 
 to, 30 : thickness of, 24. Windows, 
 double, east transept, 513. Prior's 
 Kitchen. Vault ribs, 338. Refectory, 
 Undercroft. Groined vaulting, 292. 
 
 EARL'S BARTON. Anglo-Saxon pilaster 
 
 strips, 358. Tower, battered, 605. 
 EASBY. Abbey. One -aisled nave, 2or. 
 
 Barn, Roof, 558. 
 EASINGTON. Norman capitals; plantain 
 
 leaves, 418, 428. 
 EASTBOURNE. Aisles, 226. 
 EAST DEREHAM. Towers, central lantern, 
 
 and detached campanile, 593, 598. 
 EASTWOOD. Aisled nave, 224. 
 EBRACH. Simplified Cisierian plan, 171. 
 
 EDINBURGH. St Giles. Spire lantern 
 
 or crown, 625. 
 EDINGTON. East window, 500. Dagger 
 
 in tracery, 503. 
 
 ELGIN. Cathedral. Aisles, double, 156. 
 Doonvays, recessed, 579: west doorway, 
 recessed orders, 276. 
 
 ELM. Clerestory windows, 68. 
 
 ELY. Cathedral. Abacus, cruciform, 
 439 : logically subdivided, 37. Abutments 
 — have, 29 ; Lady chapel, 351. Apse, cen- 
 tral, foundation of, 1 64. Aqueduct on upper 
 flying buttress, 400. Arches — foiled, 268 : 
 ogee, 270: three-centred, 267. Aixh 
 moldings, 279. Boss, 305. Buttresses — 
 column buttresses, Gothic, 351 ; Norman, 
 351. Capitals — Galilee, stalk capitals, 43 1 : 
 volute and leaf scroll, 430 : infirmary, 
 capitals with incurved cones, 414, note: 
 presbytery, foliated marble capitals, 252 : 
 Norman transept, painted cubical capitals, 
 413 : transepts, capital with painted detail, 
 445 : interlacing capitals, 414 : Norman 
 Corinthianesque, 428. Chapels — 211: 
 Lady chapel — n, 132 : foliage, 438: tran- 
 septal chapels, walls between, 200. Choir 
 
 — 128, 179, 184; plan, 179. Clerestory — 
 
 26, 56; choir, passage, 545 : Norman, 56, 
 465. Cloister, doorway and foliage, 429. 
 Construction, 56. Corbel-tables, Gothic 
 and Norman, 392. Cusps — Norman tre- 
 foiled doorway, 509 : shrine of St Ethel- 
 dreda, 511; Bishop West's chapel, 511. 
 Doonvays, trefoiled, 429, 509, 580. Drip- 
 stone, Norman, 407. Elevation, external, 
 
 27, 67 ; internal, 26, 67, 80, 530. Eronts 
 — east, 68 ; lancets, 463 ; string courses, 
 405 : west, 73. Flying buttresses, 55, 374, 
 note, 378, 400. Foliage, 429, 430, 431, 
 438. Foundations, continuous, 25. Gali- 
 lee, 116, 202, 204; Marble, 250, 252; 
 Masonry, 20; Nave, 184, 185; length of, 
 16. Octagon — central, 128, 603; on west 
 tower, 600. Parapets, 394 and note. 
 Fiers, 242 ; alternating, 3 1 7, note ; de- 
 tached shafts, 237 and note; "Southern," 
 246, 255. Binnacles, choir and Lady 
 chapel, 364. Presbytery — eastern, 208. 
 Pulpitum, Norman, 182. Roofs — Aisle, flat, 
 540. Nave, span of, collar beam, 560: drain- 
 age, 37. Presbytery, inclination of trussed 
 rafter roof, 560. Transepts, colours of, 
 569. Shafts, detached, 237 note; marble, 
 choir, 250; vaulting, 240. Shrine of St 
 Etheldreda, cusps, 511. Spire, 61 9. Stalls 
 in nave, 185. String courses and stages, 
 404, 405. Styles, melange of, 45, 46. 
 Towers — central, fall of, 594, note: western, 
 and turrets, 599 : octagon on, 600. Tran- 
 septs — aisles, 200; eastern aisles, 197: 
 return or end aisles, 198, 200, 542 : 
 western aisles, 16, 198; galleries in, 16; 
 sacristies, 198. Western transepts, 16. 
 Triforium — arcade, single containing arch, 
 533 : choir, 80, 533 : nave, 26, 27 : pres- 
 bytery, 67, 533, 5 38 : interior and exterior 
 elevation of, 26, 27, 67, 80, 530: glazed, 
 538: too large, 533: use of, 542: win- 
 dows, 67, 532. Vaulting — Chapels, 
 Bp. Alcock's, fan vaulting and pendant, 
 345 : Lady chapel, nearly flat, 351, 541 : 
 Bp. West's, boss, 305 ; conoids, 341 ; 
 knife-edged ribs, 301 ; mitred vault, 305. 
 Choir, 341 ; spring, 307 ; thickness, 304. 
 Presbytery, shafts, 240 ; spring, 307 ; 
 tiercerons, 337. Vestibule, western, 204. 
 
 Walls, upper courses, 37 : between tran- 
 septal chapels, 200. Windows — clere- 
 story, 56: curvilinear, 501 : east windows, 
 177, 463: lancets, 463: Lady chapel, 
 west window, 500 : nave, clerestory, 56 : 
 Norman and glazing, 459 and note: 
 triforium, 67, 532. 
 
INDEX OF PLACES. 
 
 749 
 
 ENGLESQUEVILLE. Norman antral 
 tower, no transepts, 51)0. 
 
 ENSHAM. No buttresses, 362. 
 
 EPINAL. St Maurice. Round tower, 
 601. 
 
 ETCHINGHAM. Arch, compound pointed, 
 272. Brass, 13S. Curvilinear church, 
 501. Vaulted sanctuary, 221. 
 
 ETON. College Chapel. Aiskless, 22S. 
 Buttress chapels, 197. 
 
 ETTON. Spire, 620. 
 
 EU. Aisles, superposed, 529: I'autl web 
 filled in in English manner. 335. 
 
 EVESHAM. Transepts, apsidal chapels, 
 197. Central tower, fall of, 594. 
 
 EVREUX. 1 68, note. Clerestory, skeleton 
 construction, 57. Flying buttresses, 376, 
 used as aqueducts, 400. Lantern, vaulted, 
 593. Transparent triforium, 538, 539. 
 Vault, thrust, 376, of lantern, 593. 
 
 EWERBY. 132. No chancel arch, 227, 
 note. Broach spire, 620. 
 
 EXETER. Cathedral. Aides and arches, 
 eastern, 172. Flying buttresses, 55, 369, 
 377. Foliage, naturalistic, 435. Front, 
 east, 68. Lady chapel, 68, 172. Xave, 
 128. Parapet, internal pierced, 397. 
 Piers, 242: illogical, 241 : "Western," 253. 
 Pinnacles — flying buttresses, 377 : choir, 
 flanking, 363: sedilia, 314. Plan, Nor- 
 man, 187, note Proportion, 53, 54. 
 Sedilia, 314, 364. Towers, twin tran- 
 septal, 593 : on seal, 598. Transepts, 
 eastern, 187, 189. Vaulting, 337, 338: 
 shafts, presbytery, 240. Windows, alter- 
 nating, 490 ; late geometrical tracery, 475, 
 478, j.79, 485, 487, 488, 490. 
 
 EXTON. Spire, on octagonal drum, 629. 
 Tower and turrets, 623. 
 
 FAIRFOR1). Choir, 221. Stained Glass, 
 141, 495, note. Tower, Norman central, 
 59o. 
 
 FAMAGUSTA. Flying buttresses as aque- 
 ducts, 400. 
 
 FARINGDON. Roof, horizontal struts, 
 569- 
 
 FAVKRSHAM. Transept, with double 
 aisle, 198. 
 
 FECAMP. 168, 317, note. 
 
 FEE R INC,. Window mullions, 516. 
 
 FENLAND. Detached tmvers, 598, note. 
 FILBY. Roof, 561. 
 FINCHALL. Abbey. Spire, 61 9. 
 FLEET. T'j,;r t position of, 596. Spire, 
 pel forations of, 626. 
 
 FLORENCE. Duomo. Polychromy, 252. 
 Vaulting, high vault squares, aisles 
 oblongs, 321. San Miniato. Choir, 
 raised, 191. Confessio or crypt, 191: 
 
 capitals and abaci, 410. Piers, alternation 
 
 ol, 317: piers and stone principals, 235, 
 236. Windows, 459, note. 
 
 FONTEFROIDE. Abbey. Crossing and 
 
 cloister, English method of filling web. 
 of vaulting, 335. 
 
 FONTENOY Cistercian plan, id. 
 
 FOSSANOVA. Crossing, vault, ridge ribs, 
 
 335- 
 FOTHERINGHAY. Bases, 455. Lantern 
 
 church, 138. Flying buttresses, 368. 
 
 Octagon on square tower, 600. Parvise 
 
 with piscina, 584. 
 
 FOUNTAINS. Abacus, 441. Arches, 
 joggled, 260 ; pointed, n, 266. Bases, 
 with flattened roll, 451. Basement courses, 
 38, 402, 403. Blindstory, 27. Capitals, 
 foliage, plantain or laurel leaf, 42S, note ; 
 molded, 444. Chapter house, three 
 
 avenues, 383. Cistercian church, one of 
 oldest, 44. Clerestory, no passage, 545. 
 Elevation, internal, 27. Front, east, 68. 
 Narthex, 202. Parapets, 304, 395. Pier, 
 "Southern," 249, 255. Refectory, two 
 avenues, 383. Roof drainage, 37. Sanc- 
 tuary, unaisled formerly, 176. Strings, 
 406. Styles, melange of, 46. Transept, 
 eastern, 6S, 189. Triforium, walled, 27, 
 534. Undercroft, two avenues, 383. 
 Vaults, barrel, 283. 
 
 FRAMLINGHAM. Hybrid roof, 571. 
 
 FRAMPTON. Basement course, round but- 
 tress, 360. Broach spire, 620. 
 
 ]■" RANGE. Transepts, eastern aisle only to, 
 
 rare, 198. 
 F R ESI N GFIELD. Perpendicular windows, 
 
 501. 
 FROSTERLY. Marble, 250. 
 
 FURNESS. Abacus, transitional, 441. 
 Capitals, hollow necked, 442. Cistercian 
 chinch, one of the oldest, 44. Piers, 
 alternation of, 243. Sanctuary, unaisled, 
 176. Strings, 405: shape of, 406. 
 Vaulting ribs, 301. 
 
75o 
 
 INDEX OF PLACES. 
 
 GANNAT. Naff hex, 202. 
 
 GEDDINGTON. Eleanor cross, 11. 
 
 GEDLING. Entasis of spire, 632. 
 
 GEDNEY. Roof, 568. Windows, low side, 
 517 ; tracery, 477. 
 
 GENOA. Cathedral. Basilica, 232. 
 S. Annunciata. Basilican type, 232. 
 
 GERMIGNY-DES-PRES. Apses, three 
 parallel eastern, 163. Totver, central 
 lantern, 592. Piers, 236. 
 
 GERONA. Hall plan, 321. 
 
 GILLINGHAM. No buttresses, 362. 
 
 GIMINGHAM. 7W/i, 582. 
 
 GLASGOW. Cathedral. Crypt, 339. 
 Cross Steeple. Spire lantern or crown, 
 625. 
 
 GLASTONBURY. Abbey. .Jm/m, two 
 eastern, 172. Ambulatory, rectangular, 
 106, 171. Arch, four-centred, Lady chapel, 
 267 : pointed in vaulting, St Joseph's 
 chapel, 321, note: window, 462. Astragal, 
 omission of, 434. Buttresses — transitional, 
 358. Capitals, West of England (choir), 
 434. Crossing, St Andrew's cross at, 382. 
 Doorways, hood molds, finish of, Lady 
 chapel, 407. Flying buttresses, earliest, 
 106: nave, 369, 530, note. Foundations, 
 25. Galilee, 202 and note. Gothic 
 growth, 49. Original osier church, 219, 
 note. Binnacles, Lady chapel, 362. Tran- 
 sept, double eastern aisle to, 198. Tri- 
 forium, absorbed into pier arcade, 535. 
 Vaulting, tas-de-charge, 303 : Lady chapel, 
 ribs, diagonal and wall, 300 : St Joseph's, 
 domical, 310: pointed arch in, 321, note. 
 Windows, arch, nave aisle, 462 : broad 
 lancets, 462 : Lady chapel, 465. St 
 John's- Tower, 604 : buttresses, 605 : 
 disposition of ornament, 608. 
 
 GLINTON. Spire, entasis of, "sugar loaf," 
 632- 
 
 GLOUCESTER. Cathedral. Supreme 
 importance of, in architectural history, 134. 
 Abacus, circular, 37, 439. Abutment, 351. 
 Aisles, high nave, 530 : upper, 542. 
 Altitude, 61. Apses — radiating chapels, 
 28, 167; pentagonal, 16S: three storied, 
 170: transeptal, 197. Arches, south-eastern 
 apse, 266 : molded pier arch, north aisle, 
 33, 277. Bases, 477 : South transept, 454. 
 Ball flower, 83. Choir — 184, 185 : con- 
 struction, skeleton, 58. Clerestory, 497, 
 543 : no passage in choir, 545. Crypt, 191. 
 Cylinders, 33, 234, note, 242. Elevation,\x\- 
 ternal, 26. Flying buttresses, absent, 372. 
 
 Fronts, east and west, 91. Gothic, to, 
 altitude of, 61 ; melange of styles, 45, 46 : 
 veneer over Norman work, 46, 49. Mortar, 
 24. Nave, rebuilding begun, 46 : propor- 
 tion of, S3 : span of, 15. Panelling, 94, 608. 
 Periapsidal plan, radiating chapels, 28, 167, 
 168, 170. Piers (see Cylinders), compound, 
 242 : vault piers, 58, 62, 540. Pinnacles, Ed. 
 II. tomb, 363 : open work, 363. Responds, 
 131. Romanesque and Gothic, 45, 46, 49. 
 Stained glass, 491, 493, 497-9. Stalls, in 
 nave, 185. Towers — Central — design of 
 top, 606 ; effect of, 594 ; panelled all over, 
 608 (see Pinnacles). Former western towers, 
 598 : foundation of north-west tower, 25. 
 Triforium, choir, 88, 91 ; chamber, 527 : 
 nave, two containing arches, 534 and note : 
 internal elevation of, 26 : use of, 542. Vault 
 — Chapels, apsidal, groined, 292 ; Lady 
 chapel, subway under, 53. Choir, aisles, 
 300, 527, note: upper aisle, half barrel, 284, 
 527, note : ambulatory, groined, 292 : Nor- 
 man, probable, 312, note: ribs, lierne, 
 340, 342 : triple longitudinal ridge ribs, 
 336 ; wall ribs, 300 : spring of vault, 306, 
 307 : thrust of vault, 374. Cloister and 
 lavatory, fan vaulting, 343, 344, 345. 
 Crypt, keystone vault, 305. Nave, seg- 
 mental diagonal ribs, 311: north aisle, 
 ribbed vault, 299: south aisle, ornamented 
 ribs, rib moldings, 301 ; wall ribs, 300 : 
 west nave, triple longitudinal ridge ribs, 
 336 ; spring of vaults and abutment, 306, 
 307, 351. South transept, lierne ribs, 
 340, 341, 342 : mitred vaults, 305. Vault 
 under central tower, 593. Verticality, 62. 
 Windoivs — clerestory, 497 : east, 505 : 
 geometrical, 477, 478: perpendicular, 
 49 6 > 497. 5°°. 5°5 : triforium, 532. 
 
 GORING-ON-THAMES. Norman ribbed 
 dome, 281. 
 
 GOTLAND, Island of. WISBY. Pointed 
 arch, 263. 
 
 GRADO. Basilica. 232. 
 
 GRANTHAM. No clerestory, 547. Spire, 
 623: arrangement of lights, 628: orna- 
 mentation of, 632 : shortened and line 
 altered, 632. Toiver, position of, 596. 
 West window, north aisle, 510. 
 
 GRAVILLE. Alternating piers, 317, note. 
 
 GREAT GRIMSBY. Arch moldings, 279. 
 Clerestory, windows, 68. 
 
 GREAT HALE. Acanthus in Anglo-Saxon 
 work, 427. Navel staircase, Anglo-Saxon, 
 601. 
 
INDEX OF PLACES. 
 
 75' 
 
 GREAT HASELEY. Pitch of roof lowered, 
 
 cutting off head of east window, 391. 
 GREENSTEAD. Pre-Conquest church, 219. 
 GRESFORD. Arches— no chancel arch, 
 
 227 : pier arches, no hood-mold, 407. 
 Battlements, 398. Roofs, flat, 541 : arched 
 braces, 562 : tie-beam used as principal 
 rafter, 563. String courses, none, 404. 
 
 GUERON. Apse, elliptical arch, 262: semi- 
 elliptical diagonals, 310. 
 
 GUIBRAY. Apses, three parallel eastern, 
 163. 
 
 GUILDFORD. St Mary's. Alteration of 
 church, 46. Apses, 65, 164. Choir, 221. 
 Melange of styles, 46. Tower, central, 
 Anglo-Saxon, 590. 
 
 GUISBOROUGH. Buttresses, gabled, 367. 
 Canopies, ogee, 270. Choir, 179. East 
 front, 68. Pinnacles, 362. Triforium 
 arcade, no containing arch, trefoiled arches, 
 534. Vaulting, quadripartite, 82 : shaft, 
 
 240. Windows, 475, 476, 4S5, 508 : 
 mullions of south window, 514. 
 
 HADDINGTON. Spire lantern or crown, 
 625. 
 
 HADLEIOH. Plan, 220. 
 
 HALES. Apse, 212. 
 
 HAMPTON POYLE. Capitals with grot- 
 esque figures, 416. 
 
 HARLESTON. Capitals, molded and 
 double molded, 443. 
 
 HARMSTON. Abacus, 441. 
 
 HARTLEPOOL. "Northern school" of 
 Gothic: compared with New Shoreham, 
 106, note. 
 
 HAVERFORDWEST. Capitals: foliage, 
 429, 430, 432. Geometrical windows, 487. 
 
 HAWTON. Buttress, basement course of, 
 360 : gabled buttress, 367. Foliage, 438. 
 Sedilia and Easter sepulchre, 133, 438. 
 
 HECK.INGTON. Advanced in type, 132, 
 133. Capitals, molded, nave, 444 ; tower, 
 443. Ogees, everywhere, 131, note, 133. 
 Perforated parapet, 396. Pinnacles, 364. 
 Roofs, chancel, boarded, nave, open, 560. 
 Sedilia and Easter sepulchre, 133. Win- 
 dows, curvilinear, 479. 
 
 HEDON. Aisles, 67. Buttress, simple de- 
 sign, 360 : gabled, 367. Clerestory, 68, 
 456 : windows, 68. Lantern, 593. Para- 
 pets, 31)5. Piers, arcade, tall, 122: illogical, 
 
 241, note. Transept, fronts, 68. Windows, 
 clerestory, 68 : curvilinear, 484 : flamboy- 
 ant, 485, 48S : geometrical, 477. 
 
 1 1 ELM 1 N( ;T( )N. hi lie, nth-century door, 585. 
 HELPRINGHAM. Arch moldings, 280. 
 
 Steeple, point ol view, I133. 
 
 1 1 EM EL 1 1 EMPSTEA 1 ). Chancel, vaulting, 
 3 r 4i 55° : aisle and sacristy, 225. Plan, 
 
 19. 
 
 HEMINGBOROUGH. Broach spire, 620. 
 
 Proportion, of tower and spin-, 631. 
 
 HEREFORD. Cathedral. Abutment, 351. 
 Aisle, eastern, 172: eastern aisle to north 
 transept, 1 98. . Ipses, three eastern formei l\ , 
 foundations of, 1(14 and note. Arch, 
 straight sided, 259: Norman, much carved, 
 little molded, ^^. Astragal, omitted, 434. 
 Ball J/oiccr, 83. Base, Norman, 450. 
 Basement course, 37. Buttresses, gabled 
 upright, 367 : Lady chapel, 351. Buttress 
 chapels, 197, note. Capitals, with incurved 
 cones, 414 : " West of England,'' no 
 astragal or necking, 434. Clerestory, 
 north transept, peculiar shape, no passage, 
 545: circular windows, 517. Cloister. 
 south wall, Norman baluster window, 458. 
 Crypt, ossuary or Golgotha, 194. Foliage, 
 naturalistic, Cantilupe shrine, 435. Fronts, 
 east, 68 : west, original, 73. Foundations 
 of eastern apses, 164 and note : con- 
 tinuous, 25. Gothic, melange of styles, 
 45,46: " Western school" of, 105. Lady 
 chapel, 68, 172, 351. Plan, 1 64, 170. 
 Porch, north or Bishop Booth's, double, 
 583: vaulting, knife-edge ribs, 301 : win- 
 dows, =504. Romanesque, 45, 46. Shaft, 
 533. Shrine, Cantilupe, 187, 435- Spire, 
 503, 594, 619. Stalls, 185. Towers — 
 central, covered with ball flower, 83, 
 608 : construction, inner shell, 594, 602 : 
 design of top, 606 : piers of, 24, 593, 
 note : former spire, 593, 594, 619 : streng- 
 thened, 594, 602 : western tower, fall of, 
 594, note. Transepts — north, 121, has 
 eastern aisle, 198, but south has square 
 sacristry, 197. Triforium — choir, blind- 
 story ; shaft and tympanum, 533: north 
 transept, two containing arches, 534 : 
 circular windows, 517. Vaulting, thick- 
 ness of, 304. Choir, groined, 294: Nor- 
 man, 312, note: spring of, 307. Porch, 
 knife-edged ribs, 301. Lady chapel, low 
 spring, thrust, 351. Transepts — east tran- 
 sept, ribs, 300, note : north transept, 
 spring of, 307: south transept, Hemes, 343. 
 Windows, Cloister, south wall, Norman 
 baluster, 458. Porch, 504. Transept, 
 north, clerestory and triforium, circular, 
 5 1 7. Bishop's Palace. Norman roof, 572. 
 
752 
 
 INDEX OF PLACES. 
 
 HERNE. Chcuuel aisles, 226. Tower, 
 position of, 596. 
 
 HERSFELD. Basilica. 232. Transept, 
 196. 
 
 HEXHAM. Basement course, 402, 403. 
 Ceiling, retrograde, ir3- Choir, aisled 
 parallelogram, 176, T79. Crypt, xy\. 
 Nave, one aisled, 201. Saxon church, 155, 
 218, 234, note. Styles, melange of, 46. 
 Transept, 46. Triforium too large, single 
 containing arch, 533. Vaulting, Anglo- 
 Saxon barrel vault, 284: filleted ribs, 301 : 
 vaulting shafts, 240. 
 
 HIGHAM FERRERS. St Mary. Chancel 
 aisle and sacristy, 225. Doonnay, rear 
 arch, 585. Spire, flying buttresses to, 624, 
 note. College. 136, note. 
 
 HINGHAM. 132. Capitals, molded, 443- 
 
 Clerestory, 546. Plan, 80. Roof (aisle) 
 hammerbeam, 564, note. Sacristy, 225. 
 Windows, curvilinear, 484, 490. 
 HISTON. Capitals, molded, transept, 444. 
 
 HISPALI. Triple eastern apses. Episcopal 
 chair or throne, 159. 
 
 HITTERDAL. Roof, pitch of, 559. 
 
 HOLBEACH. 132. Clerestory, 82, 546. 
 Door, fourteenth century, 489, 585. Plan, 
 80. Hammerbeam roof, 566, note. Spire, 
 623 : arrangement of lights, 628. 
 
 HOLME. Tower, "Norfolk sound holes," 
 517, 604. 
 
 HOLYROOD. Capitals, 432. 
 
 HORBLING. Abacus, 441. 
 
 HOUGH-ON-THE-HILL. Tower, Anglo- 
 Saxon newel staircase, 60 r. 
 
 HOUGHTON - LE - DALE. Transitional 
 window, 499. 
 
 HOWDEN. Aisles, lofty and wide, 67. 
 Buttresses, gabled, 367. Canopies, ogee 
 arches, 87, 270. Ceiling, 82. Chapter 
 house, cusping, 511. Choir, 122, 179. 
 Clerestory, 68, 546. Fronts, east, 82, 87, 
 270: west, 73, 362. Lantern, 593. Nave, 
 ceiling, no high vault, 82; planning, 67. 
 Niches., ogee canopied, 87, 270. Parapet, 
 choir, 396. Pinnacles, 362. Strings, 
 throat, 405. Tower, 593, 604. T?-ansept, 
 east aisle to, \<f&. Vaulting, choir, quadri- 
 partite, 82; keel molded rib, 312. Win- 
 dows, 475 : clerestory, 68, 82 : west win- 
 dow, transom inserted, 503, 506, 509. 
 
 H O WE LL. Chantry chapel, 225. 
 
 HUISH EPISCOPI. Tower, "Somerset," 
 belfry windows, pierced stonework, 517, 
 605. 
 
 HULL. Capitals, foliage, nave, ^437. 
 Chancel, advanced in type, first ~stone 
 lantern, 133. Buttresses, importance of, 
 367. Front, east, 82; west, 91. Plan, 
 cruciform, 80. Roof, 563 : flat, 54r ; 
 invisible, 391. Tower, central, 591 : pin- 
 nacles, 608. Windows, clerestory, chancel, 
 82: curvilinear, 479, 483, 490: east, 
 transitional, 499 : west, 503, 504 : dagger 
 in tracery, 503. 
 
 HUNSTANTON. "Roue tournante%win- 
 
 dow, 485. 
 
 HURSTMONCEAUX. Capitals, plantain 
 or laurel leaf, 428, note. 
 
 HYTHE. Chancel, spring of vault, 307. 
 Crypt, ossuary or Golgotha, 194. Vaulted 
 subway, 53. 
 
 ICKHAM. Aisled nave, 223. 
 
 ICKLESHAM. Capital, 414. 
 
 ICKLETON. Doorway, roll moldings, 
 277. 
 
 IFFLEY. Enlarged chancel, 225. Choir, 
 221. Front, we>t, 29.. Door7cay, recessed, 
 with thickened wall, 579: Norman drip- 
 stone, 407. Tower, Norman, central, 
 590. Window, circular, 468. 
 
 IRCHESTER. Spire, 6rg. 
 
 IRELAND. Arches, pointed, pre-Conquest, 
 264. Stone roofs, 281. Round lowers, 
 586, 601. 
 
 IRTHINGTON. Capitals, plantain or 
 laurel leaf, 428, note. 
 
 ISLE ABBOTS. Tower, "Somerset/' bel- 
 fry windows, pierced stonework in, 517, 
 605. 
 
 ISLE OF PURBECK. Marble, 250. 
 
 ISLE OF THANET. Arches, 274. 
 
 1SSOIRE. Vaults, barrel and half-barrel, 
 r3. 284. 
 
 J ARROW. Balusters, 445, note. 
 JATAGHAN. C7-uciform church, 195, note. 
 J EDBURGH. Capitals, direction of foliage, 
 
 432. Trifo}ium, absorption of, into pier 
 
 arcade, 535. 
 
NDEX OF PLACES. 
 
 I 
 
 JERVAULX. Bases with flattened roll, 
 451. Chapter house, foliage, intersecting 
 
 stalks, 432 : marble piers and shafts, 250. 
 Choir, aisled parallelogram, 176, 179. 
 Strings, 406. Vault ribs, 301. 
 
 JUMIEGES. Aisles, superposed, 528, 531. 
 Arches, Norman, 276. Bases, 450. Capi- 
 tals, painted cubical, 411. Clerestory, no 
 
 passage, 545, note. Columns, engaged, 
 236. Piers, alternation of, 243, 317, note. 
 String-courses, 404. Triforium chamber, 
 
 527. Van/ting, aisles, groined, 292: no 
 high vault, 290. 
 
 KELLS. St Columba. Pointed arch, 264. 
 KELSO. Porch, two stories high, 205. 
 
 KEMPSFORD. Choir, 221. Tower, Nor- 
 man, central, 590. 
 
 KENILWORTH. Doorway, Norman, re- 
 built, 579. Spire, 617. 
 
 KETTERING. Basement course, 402, 403. 
 Doorway, label to, 5S0. Downpipes, 401. 
 Dripstone, label, 407, 5S0. Masonry, 20. 
 Spire, construction of, thickness of stone, 
 636 : lights, arrangement of, 62S : straight 
 sides, 632 : turrets, 632. Tower, battle- 
 ments, splayed loopholes, 398, note : 
 strings and set-offs, 606 : set back each 
 stage, 605 : tower and spire, proportion of, 
 631 : unity of, 632. 
 
 KETTON. Arches, pointed and semicir- 
 cular, 265, 579. Capitals, molded with 
 nail head, 443. Doorway, semicircular 
 head, 579. Masonry, 20. Spire, 617, 
 note : broach, 620 : lights, arrangement of, 
 628 : proportion of tower and spire, 631. 
 
 KIDDINGTON. Roof, 55S : tiebeam used 
 as principal rafter, 563. 
 
 KIDLINGTON. Chancel aisles, 226. 
 
 KIDWELLY. Broach spire, 620. 
 
 KILK HAMPTON. Bases and benches, 
 447. Barrel roof, 569. 
 
 KILLALOE. Pointed arch, 264. 
 
 KILPECK. Capitals and shafts, intertwin- 
 ing monsters upon, 416, note. Semidome, 
 sculptured keystone, 305. 
 
 K I LW I N N I N G . Horseshoe arch, 262. 
 
 KING'S LYNN. See LYNN. 
 
 KING'S SUTTON. Spire, band round, 
 626 : pinnacles, 625. 
 
 KIPPAX. Norman church, no buttresses, 
 
 362. 
 
 KIRKBURN. Tower, svestern Norman, 19. 
 KIRKHAM. One-aisled nave, 201. 
 
 KIRK HAMMERTON. Arch, Anglo 
 Saxon, with orders, 274. 
 
 KIRKDALE. Arch, Anglo Saxon, with 
 orders, 274. Doonvay, Anglo-Saxon, 
 masonry of, 2^. 
 
 KIRKSTALL. Abacus, 141. Arches, 
 joggled. 200: pointed, 33, 206. Base, 
 waterholding, 451. Basement course, 38. 
 Buttresses, chapter house, 358. Cistercian 
 church, one of oldest, 44 : plan, 19. Pres- 
 bytery oud sanctuary, unaisled, 104, 17(1, 
 321. Planning, 19, 104. Transept, 
 eastern aisle, divided by walls into chapels, 
 103, 19S and note, 200. Triforium, 
 walled, 534. Vaulting, aisles, 311 : barrel, 
 283: choir, Norman, 314, as at Angers, 
 in squares, 321 : nave, moldings of trans- 
 verse ribs, 301. Windows, plate tracery, 
 468 : semicircular headed, 462. 
 
 KNAPTON. Rood beam, 180, note. Roof, 
 angels, 56S : braces, longitudinal, 508 : 
 cornice, 567 : colours of, 569 : hammer- 
 beam, 565, 566, 567. 
 
 LAACH. Arch, elliptical, 2(12. Atrium, 
 201. 
 
 LA CHARITE SUR LOIRE. Arches, 
 pointed, 265. Nave, west end of, 202. 
 Periapsidal plan, 169. Transept, western, 
 202. 
 
 LANCHESTER. Roman altar with chev- 
 ron, 40. 
 
 LANERCOST. Arches, pointed, 103. Nave, 
 one aisled, 201. Sanctuary, unaisled, 176. 
 
 LANGPORT. Marble, 250. 
 
 LAON. Aisles, superposed, 528. Piers, 
 230; alternation of, 317, note. Vaulting, 
 sexpartite, 320 ; shafts, 240 ; wall ribs, 23S. 
 Windows, aisles, superposed of, 531 ; cir- 
 cular, 517 ; plate tracery, 46S, 469, 517. 
 
 LASTINGHAM. Chancel, vaulted, 550. 
 Crypt, 191 ; Norman capitals, 427: vault- 
 ing, centering to, 295. 
 
 LAUGHTON. Capitals, classical reminis 
 cences, 428, note. 
 
 LAVENHAM. Battlements, flint panelled, 
 399. Doorway, label, 580. Pews, screened, 
 1S2. Porch, 582. Roof— arched ibraces, 
 562 ; tie beam, cambered, 562 ; tie beam 
 used as principal rafter, 563. Tower, 
 position of, 596 ; proportions of, 603 ; sky- 
 line, horizontal, 606. 
 
 3 B 
 
754 
 
 INDEX OF PLACES. 
 
 LEADEN HAM. Astragal, 444. Arch 
 moldings, 280. Capitals, molded, 444. 
 
 LECKHAMPFON. Choir, 221. Plan, 
 80. Spire, broach, 620. 
 
 LEDBURY. Ball flower, St Catherine's 
 chapel, 83. Chancel, circular clerestory 
 windows, 517. Doorway, west, foliage, 
 arum, 41 8. Spire, parapetted, 622. 
 Tower, detached, 598. Windows, 476, 
 477, 47S: circular, 517. 
 
 LEEDS. St John. Naves, parallel, 224. 
 Screens and fittings, Jacobean, 203, note. 
 
 LEICESTER. St Martin. Roof, 558; 
 low pitch of, 559 : tie beam used as prin- 
 cipal rafter, 563. St Mary. Buttress, 
 Norman column, 351. Roof, arched 
 braces of, 562 ; hybrid, 569. St Nicholas. 
 Arches, simple, 272; pier arch, 275. 
 Cruciform extension, 222. Nave, aisled, 
 223. 
 
 LEIGHTON BUZZARD. Clerestory, 547. 
 
 LE MANS. Cathedral. Crocket capitals, 
 choir, 429, note. Periapsidal plan, 170. 
 Transeptal towers, 593. Notre Dame 
 de la Couture. Choir, 184. Flan, 
 hall, 321, note: periapsidal, 169. Win- 
 dows, 468, note. Notre Dame du Pre. 
 Piers, alternating, 317, note: piers and 
 stone principals, 235. Periapsidal plan, 
 169. Transept, apsidal chapel, 197. 
 
 LEOMINSTER. Arches, Norman, un- 
 molded square-edged, 33, 276. Doonvay, 
 thirteenth century, removed and reset, 
 579. Periapsidal plan, 167. Windows, 
 geometrical, 478 : west, perpendicular, 
 5°5- 
 
 LE PUY. Cathedral. Capital, 425. 
 Domes, 6, 13. Foliage, fern, 418, note. 
 Folychromy, 250. 
 
 LESSAY. Apses, three parallel eastern, 
 163. Arches, Norman, 276. Bases, 447. 
 Capitals, scalloped, 413. Piers and de- 
 tached shafts, nave, 237. Romanesque, 
 10. Triforium, 532, note. Vaulting — 
 Aisles, groined, 292. Nave — ribbed, 13; 
 Norman high vault, date of, 314, 316: 
 short voussoirs to ribs, 301. 
 
 LEVERINGTON. Porch, 288; ogee 
 crocketted dripstone and finial, 407. 
 Spire pinnacles, 623. 
 
 LEWES. Chapels, radiating eastern, 167 : 
 tangential transeptal, 189. Copy of Cluny, 
 189. Periapsidal plan, 169. Towers, 
 central and western, 599. 
 
 LICHFIELD. Ambulatory, rectangular, 
 171. Arch, pier, 540. Ball floiver, 625. 
 Blindstory, 535. Buttresses, gabled, 367 ; 
 upright, 367, 381. Capitals — Choir, char- 
 acter of foliage, 432 ; incurved cones, 414. 
 Nave, foliage, direction of, 432 ; natura- 
 listic and conventional, 435 ; volutes, 429, 
 note. Choir, original, of " Western school " 
 of Gothic, 1 05. Cleresto?y — choir, 82 ; nave, 
 no passage, peculiar shape, 545 ; spherical 
 triangular windows, 517: presbytery, 82, 
 131, 536. Doorways, double, transepts, 
 577, 584. Dripstones, Lady chapel, ogee 
 crocketted, 270, 407. Flying buttresses, 
 368. Foundations, continuous, 25. Front, 
 west, 71, 72, 73. Gothic, early and late, 
 juxtaposed, 46 ; Gothic growth, 49 ; 
 "Western school" of, 105. Lady chapel, 
 buttresses, 367, 381 ; ogee crocketted 
 dripstone, 270, 407. Niche, ogee, cano- 
 pied, 87. Ogees, 87, 270, 407. Parapet, 
 choir and presbytery, clerestory, interior 
 pierced, 82, 397 : nave, 396. Piers, choir, 
 24, 245. Pinnacles, choir, 364. Propor- 
 tion, 54. Sanctuary, unaisled, 176. Spires, 
 western, ball flower, 625 ; lights, 627 ; 
 strings round, 626. Styles, melange of, 
 46. Towers, west, 71, 72, 73. Transepts, 
 aisle, west, 198 ; not originally vaulted, 
 118. Triforium, absorption of into clere- 
 story, 82, 131, 397, 536: nave, two con- 
 taining arches, 534: blindstory, 535. 
 Vaulting — Nave, ribs, 338 : vault shaft, 
 240 : web of rubble, 299. Transepts, 
 longitudinal ridge ribs and tiercerons, 337 ; . 
 transverse ridge ribs, 339, note. Windows, 
 clerestory, nave, spherical triangle, 517 ; 
 geometrical, 477. 
 
 LILLESHALL. Gothic, " Western school" 
 of, 105. Aave, aisleless, 201. 
 
 LIMBURG. Basilica. 232. 
 
 LIMOGES. Roof of aisle, flat : transparent 
 triforium, 538. 
 
 LINCOLN. Cathedral. Abacus, Nor- 
 man, chamfered, 439 : early Gothic, 441. 
 Abutment, choir, 54. Arcading, aisles, 
 535, note ; foliage of capitals of, 430, 432. 
 Arches, obtuse, nave, 266 : ogee, western 
 towers, 270: transverse, triforium, 369: 
 moldings, nave and presbytery, 279. 
 Bases, western doorways, 452. Buttresses, 
 chamfered, sharp, 358: clerestory, 377: 
 gabled, choir and nave, 363 ; presbytery, 
 367: Norman, angle shaft, choir and tran- 
 septs, 358 : pinnacled, chapterhouse, 363 ; 
 vertical, angel tower, 605. Capitals and 
 
INDEX OF PLACES. 
 
 foliage, 426, 429, 430, 432, 435: Gothic 
 molded capitals, 442 : proportion of 
 capitals, 409 : volute capitals, 430. ( '/lapels, 
 "buttress," 197, note: chantry, 211, Bishop 
 Hugh of Wells', 20S ; semicircular, 65 : 
 walls, low, between, 200. Choir, 184: pro- 
 portion of, 409: length of choir, 16: Angel 
 choir, 179. St Hugh's choir, 105, 109- 
 113. Construction, skeleton, 56. Columns, 
 marble, 252; Roman, 233. Corbel-table, 
 presbytery, 392. Diaper work, 84 ; un- 
 dulating foliage, 436. Doorways, chapter 
 bouse, double, 577: presbytery, south 
 doorway, sculpture, 573. Flying buttresses, 
 54, 374, 375, 377- Foliage, see Capitals, 
 Diaper, and Doorways. Foundations, con- 
 tinuous, 25. Fronts, east, 68 : transepts, 
 central, 71 ; north, 6S : west front, 71, 73. 
 Galilee, 202,584: tooth ornament, 77, note. 
 Gothic, early and late, juxtaposed melange 
 of styles, 46 ; growth of, 49 : St Hugh's 
 choir, not the first piece of, 105. Lantern, 
 vaulted, 593. Marble columns, 252. Nave, 
 Norman, size of, 15. Parapets, nave, 
 394, 398 ; presbytery, 394 ; south transept 
 gable parapet, 398. Piers, "Southern," 
 246, 255. Pinnacles, chapter house, 363 : 
 choir, 367 ; clerestory, 364 ; nave, 398 ; 
 presbytery, and east and central transepts, 
 362. Planning — semicircular chapels, 65. 
 Presbytery, eastern, 20S. Proportion, 53, 
 54. Pulpilum, 179. Roofs, steep, 391, 559. 
 Spires, 594, 6rg. Styles, melange of, 46. 
 Towers, Angel or central, collapse of, 594; 
 interior effect, 603; shells, built in two, 
 602 ; spire, former wooden, 593 ; western 
 towers, 73. Transepts — east, 189: north- 
 eastern transept, return aisle, 200 : south- 
 eastern transept, diaper work, 84. Triforium, 
 chamber not floored, use of, 541 ; choir, 
 transverse arches, 369 : flying buttresses, 
 375: nave, windows of, 533, note; pres- 
 bytery, two containing arches, 534, windows 
 of, 533, note. Vaulting, spring of, 307 : 
 rib moldings, 301. Aisles, quinquepartite, 
 339. Chapels, nave, north-west, vaulting 
 with central pillar, 338 : south-west, vault- 
 ing without central pillar, 338, 339. 
 Chapter house, vaulting with central pillar, 
 339 ; heme and ridge ribs, 340. Choir, 
 335, 33 8 - St Hugh's choir, 336 ; vault- 
 ing shafts, 240. Galilee, sexpartite, 320: 
 rib moldings, 301. Lantern, 593. Nave, 
 span of vault, 291 : Norman high vault, 
 314: ridge rib and tierrerons, 337, 339, 
 note. Presbytery, vaulting shaft, 240 ; 
 tiercerons, 337 : web, French method of 
 
 filling, 535. Tower, central holi . ; 19 
 Transepts, longitudinal ridge rib, ; |6 
 wall ribs, 238 : sexpartite vaulting, , i 
 vaulting shafts, 240. Vestibule, . 
 204. Walls, between chapels, 200. Win 
 
 afcaw, curvilinear windows, 479, 490; I; 1 
 
 465. Ai^lrs and chapter house, lancets 
 in pairs, 463. Clerestory, lancets 111 trip 
 lets, 463. Choir, clerestory, lancets, 543. 
 East window, 177; mullions, 514. Pres 
 bytery, bar tracery, 469 ; clerestory win 
 dows, 56 ; cusps, 510, 511, double win 
 dows, 513; early geometrical tracery, 
 475. Transepts, circular windows, 517. 
 Bishop's Hall. Plate tracery,^. Tran- 
 soms, 503. St Mary-le-Wigford. Bt 
 fry windows Acanthus in Anglo-Saxon 
 work, 427. Crocket capitals, 429. Si 
 Peter-at-Gowts. Belfry windows— Acan- 
 thus in Anglo-Saxon work, 427. 
 
 LINCOLNSHIRE. Churches rebuilt, 132. 
 
 LINDISFARNE. Apse, 164. Apsidal 
 chapel, transept, 197. Flying buttresses, 
 none of twelfth-century date, 372. Roman- 
 esque, "Durham school " of, 100. Tiers, 
 alternating, 3 1 7, note. Vaults, Romanesque 
 high vault, 20, 314: segmental diagonals, 
 stilted transverse arches, 311. St Cuth- 
 bert's Chapel, 2 19. 
 
 LINLITHGOW. Spire lantern or crown, 
 625. 
 
 LISIEUX. Transepts, east aisle only, 198. 
 
 LITTLE ADDINGTON. Arch moldings, 
 
 279. Tracery, 508: soffit cusps, 509. 
 LLANBADARN FAWR. Transepts, 223. 
 LLANBELIG. Holy-water stoup, 207. 
 
 LLANDAFF. Arch, eastern choir, Norman 
 hood-mold, 407 ; ogee arches, 268, 407. 
 Capitals, direction of foliage, 432 : stalked, 
 431 ; great height of stalks, 434. Cham- 
 fers, 277. Dripstone, hood-mold, Nor- 
 man, 407 : ogee, 407. Gothic, " Western 
 school" of, 105. Tiers, "Western," 245. 
 Plan, T70: no central tower, 590; no 
 transept, 593. Ribs, 300, note. Window, 
 ogee head and dripstone, 407. 
 
 LLANIDLOES. Capitals, "West of Eng- 
 land," 433. Pier, "Western," 245. 
 Ifammerbeam roof, 565. See ABBEY 
 CWM-HIR. 
 
 LLANTHONV. Capitals, with incurved 
 cones, 414, note. Triforium, absorption 
 of, into clerestory, 536. 
 
756 
 
 INDEX OF PLACES. 
 
 LOCHES. Nave roof, 637, note. 
 
 LODDON. Presbytery, 226. 
 
 LONDON. Old St Paul's. Area, 15, note. 
 Choir, length of, 53, 179. Crypt, St 
 Faith's church, 191. Gothic growth, 49. 
 Nave, piers of, 24. Spire, 619. Transept, 
 length of, 53 : double aisled, 198. Win- 
 dows, bar tracery, 470: rose, 517. 
 Charing Cross, ii. Ely Palace 
 Chapel. East front, 68. Windows, 
 475. St Andrew Undershaft. No 
 chancel arch, 227. St Bartholomew, 
 Smithfield. Arches, interior arcade, 27 : 
 pointed transept arches, 265 : stilted 
 arches, 261. Choir, retrograde, 43, 99. 
 Triforium, 27 : single containing arch, 
 533. Vaults, groined, 292. St Dunstan- 
 in-the-East. Spire lantern or crown, 
 625. Southwark Cathedral. Aisles, 
 three eastern, 172. Capitals, molded, 446. 
 Internal elevation, 530. Fronts, east, 68; 
 west, 73. Piers, 24: "Northern," 253. 
 Triforium, arcade, blank, 535 ; no contain- 
 ing arch, 534. Reredos, 181. Retrochoir, 
 three parallel naves, 383. Retrograde, 
 113. Vaulting, 335 : "ploughshare," 311 : 
 shaft, 240: spring of, 307 : springers, 303. 
 St Stephen, Walbrook. Bases and 
 benches, 447. Temple Church. Aisles, 
 high, 382. Arch moldings, 279. Bases, 
 449, 452. Capitals, molded, 434, 446. 
 Choir, 11. Roof, three span, 370. Vault- 
 ing, aisles, 382 : semi-elliptical diagonals, 
 310: shafts, 240. Tower. Down pipes, 
 401. (St John's Chapel.) Apse, semi- 
 domed, 283. Console in Capital, 427. 
 Triforium, 527. Vault, barrel, 350. 
 
 LONG MELFORD. 141. Capitals, Lady 
 chapel, 409, note. Chancel arch, none, 
 227. Roof, arched braces, 562 : colours, 
 569: tie beam, cambered, 562; used as 
 principal rafter, 563. Windows, 499, note. 
 
 LOSTWITHIEL Spire, 620. 
 
 LOUTH. Aisles, 226. Bases, 450; chan- 
 cel, 454 ; steeple, 454. Chancel arch, 
 227. Buttresses, importance of, 367. 
 Chantry chapel of Thos. de Luda, 208, 
 note. Clerestory, 547. Doorway, thirteenth 
 century, rebuilt, 579. Flying buttresses, 
 double, of spire, 377, 400, 625. Front, 
 east, 91, gable parapet, 398. Plinth, octa- 
 gonal, 450. Roof, 562 ; arched braces of, 
 562 ; wall, or pendant, post, 563. Spire, 
 angle of, 631 : construction, 636 : crockets, 
 632: entasis, 632: lights, 627; arrange- 
 
 ment of lights, 628: squinches, 636: 
 thickness of stone, 636. Steeple (tower 
 and spire), ornament, 633 : pinnacles, 
 362 : point of view, 636 : proportion of 
 tower and spire, 629, 631 ; unity of, 632. 
 Tower, base of piers, 454 : position of, 
 596 : walls, thickness of, 602 : windows, 
 deeply recessed, 514, 604. 
 
 LOWESTOFT. Doorway, recess for pro- 
 cessional cross, near western or ceremonial, 
 581, note. Presbytery, lighting of, 226. 
 
 LOWICK. Molded capitals, 443. Octagon 
 on square tower, 600. 
 
 LUCCA. Basilica of S. Frediano. 232. 
 
 LUCHUEUX. Vaulting, ridge ribs, 335. 
 
 LUDLOW. Chantry chapels, 224. 
 
 LYDD. Aisled nave, 223. Piers, 232. 
 
 LYMPHENHOE. Roof, 561. 
 
 LYNN. St Margaret. Foliage, both con- 
 ventional and naturalistic, 435. Parisli 
 church (see St Nicholas), 227. Towers, 
 effect of, 594. Window, east, circular, 
 517. St Nicholas. Chapel-ofease to 
 St Margaret's, therefore, aisled chapel on 
 plan, 227. Clerestory, 543. Roof, 562; 
 arched braces of, 562. Tower, position 
 of, 228. 
 
 MADLEY. Apse, polygonal, 158. But- 
 tress, 360. Crypt, 194. 
 
 MAGDEBURG. Cathedral. Arches, 
 transverse, omission of, 316. 
 
 MAIDSTONE. All Saints. Choir, aisles 
 designed for fan vault, 346. Tower, posi- 
 tion of, 596. College. 136, note. 
 
 MALDON. Triangular tower, 600. East 
 windoiv, 516. 
 
 MALMESBURY. Arches, pointed, 266, in 
 vaulting, 321, note: stilted, 262. But- 
 tresses, flanking, 363. Doorway, recessed, 
 578. Flying buttresses, 374. Gothic and 
 Romanesque, melange of styles, 46. 
 Parapets, perforated, 396. Pinnacles, 363, 
 367: "false bearing," 377. Porch, Nor- 
 man, 582. Roof, lowering of, 367, note. 
 Towers, 599. Triforium, single containing 
 arch, 533. Vaulting shafts, 240. 
 
 M ALTON. Triforium arcade, 527. 
 
 MALVERN. Abacus, circular, 439. Clere- 
 story, 497. Internal elevation, 88. Fronts, 
 east and west, 91. Glass, stained, 494, note. 
 Gothic and Romanesque, melange of styles, 
 
INDEX ()l 
 
 'LACKS. 
 
 . ; 
 
 46. Norman ivork, arches, $3, 276: 
 early, no string-courses, 404 : no drip 
 stones or hood-molds, 407. Triforium, 
 blank, 88. Vaulting, no high, 94. 
 
 MANCHESTER. C liantry chapels, 211. 
 
 Tower, 602, 604, 60^, 
 
 MANORBIER. 
 608. 
 
 MARCH. Hammerbeatn roof, 566 ; angels 
 on, 568. 
 
 MARKET DEEPING. Capitals, 442. 
 
 MARSTON. Chantry and sacristy, 225. 
 
 MARTHAM. Roof, lead-covered eaves to, 
 
 39 2 - 
 MATHON. Doorway, Norman, trefoil 
 
 headed, 580. 
 
 MAULBRONN. Abbey. Cistercian plan, 
 16. Sanctuary unaisled, 176, note. South 
 transept, fir roof and ceiling, 571. 
 
 MAYENCE. Vaulting, 317. 
 
 MEAUX. Superposed aisles, 529. 
 
 MELBOURNE. Abacus, cruciform, 439. 
 Apse, 164: made rectangular, 157: tran- 
 septal chapel, 197. Chancel enlarged, 
 225. Roof, arched braces, 562. Towers, 
 effect of, 597 : groined gallery between, j 
 203. 
 
 MELTON MOWBRAY. Toiver, 617, 
 note. 
 
 MERSTHAM. Chancel chapels, 225. 
 
 METHLEY. Aisles, 226, note. 
 
 METZ. Clerestory, windows, skeleton con- 
 struction, 57. 
 
 MILAN. S. Ambrogio. New type of church, 
 155. Apse, 351: three eastern, 163. 
 Aisle, upper, 531. Atrium, 201. But- \ 
 tresses, concealed, Roman, 351. Flying I 
 buttresses, 374, note. Griffes, 455. Twer, 
 586, note : position of choir towers, 599. 
 Vault, repaired, 300 : ribs, 300, 317: thick- 
 ness of, 304, note. Windows in upper 
 aisle, 531. S. Eustorgio. Stone prin- 
 cipals, 235. S. SaTIRO. Tower, 586 : 
 bases, 452. S. Vincent in Prato. 
 Three eastern apses, 163. Groined crypt, 
 292. 
 
 MICHAELCHURCH. 180. 
 
 MILDENHALL. Curvilinear window, 490. 
 
 MI ETON. Door, original fourteenth cen- 
 tury, 585. 
 
 MILTON ABBAS. Eastern > a d 
 
 arches, 172. Choir, quadripartite vaulting, 
 .S2. Lady chapel, 1 -j. Planning, 80. 
 Windows, 477. (This chur< li contain a 
 rarity in the shape of .1 pre Reformation 
 tabernacle fur tin- blessed Sacrament.) 
 
 MINCHINHAMPTON. Stone roofs, 2.S8. 
 Vault, buttress and thrust, 350. Curvi- 
 linear window, 490. 
 
 MOCCAS. Tufa, building material, 303, 
 note. 
 
 MODENA. Cathedral. Piers and stone 
 
 principals, 235. 
 
 MONKWEARMOUTH. Anglo Saxon 
 
 church, 155. Porch, Anglo-Saxon barrel 
 vault, 284 : western, 202. Tower, develop- 
 ment of Anglo-Saxon, 202, 596. 
 
 MONREALE Cathedral, 232. 
 
 MONTARGIS. Cylinders, 243. L/erue 
 vaulting, 342. 
 
 MONTATAIRE. Stone roofs, 288, note. 
 
 MONTMAJOUR. St Croix. Pointed 
 dome, 264. 
 
 MONT ST MICHEL. Triforium, 532, 
 note. 
 
 MONTIVILLIERS. Transepts, elliptical 
 arches, 262 : semi-elliptical diagonal ribs, 
 3 IQ - 
 
 MONZA. 163. 
 
 MORIENVAL. Diagonal ribs, 300. 
 
 MORTON. Roof, arched braces, 562. 
 
 MORTON PINKNEY. Diagonal but- 
 tresses, 361. 
 MORWENSTOW. Norman porch, 582. 
 
 MOULTON. Capitals, direction of foliage, 
 432. Spire, angle of, 631 : construction, 
 thickness of stone, 636 : lights, 627, 62S : 
 arrangement of, 628. Tower versus spire, 
 632 : ornamentation, 633, note. 
 
 MYCEN.E. Treasury. Pointed arches, 
 264. 
 
 NANTES. Tower, early, 590 ; central lan- 
 tern tower, 592. 
 
 NANTVVICH. 132. Clerestory ; none, 547. 
 Plan, cruciform, 80. Tower, central, 591. 
 Windows, curvilinear, 490 : perpendicular, 
 
 500. 
 
753 
 
 INDEX OF PLACES. 
 
 NARBONNE. Roof of aisle, flat, covered 
 with stone slabs, 538, 540, 541. Tower, 
 early, 590 ; central lantern tower, 592. 
 Triforium, transparent, 538 : transverse 
 arches, 369, note. 
 
 NASSINGTON. Steeple, point of view, 
 633, note. 
 
 NATELY. Doorway, trefoil headed, 5S0. 
 
 NAVENBY. Sedilia and Easter sepulchre, 
 133- 
 
 NETLEY. Buttresses, slopes of set-offs, 
 361. Choir, 179: aisle, parapets, 394; 
 porte-a-faux, 394, note. Piers, "Northern," 
 253. Strings, 405, 406. Windows, bar 
 tracery, 470, 472 ; dripstone, 465 ; east 
 window, 5 12 
 
 NEVERS. Cathedral. Basilica bays, 232. 
 Crypt, capitals, 410. St Etienne. 
 Ambulatory, 169. Hollow-necked capitals, 
 442. Trabeated doorway, 573. Monolithic 
 shafts, choir, 411. 
 
 NEWARK. Chancel arch, 227. Screen, 
 227. Spire, 623 : lights, 627 ; arrange- 
 ment of, 628 : ornamentation, 632 : pin- 
 nacles, 623. Steeple, tower versus spire, 
 632 : position of, 596 : proportion of tower 
 and spire, 631. Tower, central, pulled 
 down, 590. 
 
 NEWCASTLE. Cathedral. Spire lantern 
 or crown, 625. Castle. Vaulting, Guard 
 room, tas-de-charge, 303. 
 
 NEWHAVEN. Choir, 221. Spire, 619. 
 
 NEWPORT (Mon.). St Woolos. Abacus, 
 molded, 439, note. Clerestory windows, 
 546. Doorway, foliage in capital, 427. 
 Fore-church, 202. " Southern" piers, 246. 
 
 NEW ROMNEY. Spire, 617. 
 
 NEW SHOREHAM. Arch, moldings, 279 : 
 ornament, 277: leaf-scroll in, 430: semi- 
 circular, retention of, 265. Capitals and 
 foliage, 429, 430, 432, 433 : crocket 
 capitals, 42. Choir, 10S : aisled parallelo- 
 gram, 176, 179: Norman rectangular east 
 end, 28. Clerestory, none, 545. Flying 
 buttresses, 55, 372, 374, 375, 377, 380. 
 Gothic, "Northern school "of, 106. Pier, 
 " Northern," 253. Pinnacles of false bear- 
 ing* 377- Plan, 108. Tower, central, 
 593; horizontal sky-line, 606. Triforium, 
 no containing arch, 534. Vaulting, ashlar 
 vaults, 299, note: domed, 310: ribs, 
 moldings, 301 : wall ribs, 300 : shafts, 240 : 
 spring of, height of, 307 : web, French 
 method of filling, 335. Windows, semi- 
 circular headed, 462. 
 
 NEW WALSINGHAM. Aw/r— aisle, 570: 
 foliated bosses, 94. 
 
 NILOMETER. RODA. Pointed arches, 
 263. 
 
 NIMES. Roman stone ceilings, 285. 
 NICOMEDIA. Early Christian Church, 
 
 145, note. 
 NORFOLK. Churches rebuilt de novo, 46. 
 
 NORTHAMPTON. Eleanor Cross, ii. 
 Ogee arches, 270. St Peter. Abacus 
 strings, 404. Aisles, chancel, 226; nave, 
 223; stone principals over, 235. Basilica 
 form, 232. Buttresses, no Norman, except 
 to tower, 351. Capitals, interlacing, 414. 
 Clerestory windows, 546. Griffes, 455. 
 Nave, aisled, plan, 19, 223, 232. Piers 
 and columns, 37, 236, 255. Strings, 
 abacus, 404. St Sepulchre. Piers, 
 "Western," 245, note. Spire, 622: strings 
 round, 626. Tower and spire, diagonal 
 buttresses, 605 : unity of, 632, note. 
 
 NORTHAMPTONSHIRE. Ogee arch, oc- 
 currence of, 268. 
 
 NORTHBOROUGH. Arch moldings, 280. 
 
 NORTH CREAKE. Aisled nave, 224. 
 
 Windows, 476. 
 NORTH ELMHAM. Newel stair turret, 
 
 601. 
 
 NORTH LUFFENHAM. Spire, broach, 
 620 : lights, 627. 
 
 NORTH MIMMS. Chantry chapel and 
 sacristy, 225. Doors, fourteenth century, 
 S85- 
 
 NORTH NEWBALD. Cruciform plan, 19. 
 Transepts, 222. 
 
 NORTHORPE. Capitals, classical reminis- 
 cences, 428, note. 
 
 NORTH PETHERTON. Door, fifteenth 
 century, 588. 
 
 NORTH WALSHAM. Chancel arch, none, 
 227, note. Clerestory, none, 547. Porch, 
 582. Roof 561; form of, 547. Windoivs, 
 501. 
 
 NORTON. Tower, central, Anglo-Saxon, 
 590. Transepts, 222. 
 
 NORWICH. Cathedral. Abacus, sub- 
 divided, 439. Aisle of nave, vaulting, 
 shafts, 237: ridge ribs, 336; wall ribs, 
 300. Apse, cornice table, 392 : roll mold- 
 ings, 277. Apsidal chapels, 16, radiating, 
 28,167,168: transeptal, 197. Bases, 358, 
 447. Battlements, panelled, 399. But- 
 
INDEX OF PLACES. 
 
 
 tresses, Norman, 358 : spire, 620. ( 'apitals, 
 Norman, on buttresses, 358: Corinthian- 
 esque, 428: scalloped, early, 316, 413. 
 Chapels, see Apsidal chapels : rectangular 
 Lady chapel, 46. Choir, 184: Dripstone, 
 Norman, 407 : continuation of string-course, 
 405. Elevation, external, 27, 532: in- 
 ternal, 27. Flying buttresses, 370, 372, 
 377. Golgotha or ossuary, 194. Gothic, 
 altitude of, 61: growth, 46: melange of 
 styles, 46. Masonry, 20. Moldings, 227. 
 Nave, 1S4, 1S5 : length of, 16; vaulting, 
 lierne, 341; sexpartite intended, 317. 
 Piers, 237, 239; alternating, 317, note: 
 compound, 33. Pinnacles, choir, statues 
 as, 360 : of false-bearing, 377. Spire, 593 ; 
 buttressed, 620 ; on octagonal drum, 629. 
 Stalls in nave, 1S5. String-course, 404 : 
 carved and decorated, 404, 405. Towers, 
 central, 593; diameter of piers of, 594: 
 western, formerly, 59S. Transept, north, 
 front of, 29. Triforium, 370, 519. Vault- 
 ing, see Aisle, Nave. Windows, alternat- 
 ing, 490 : clerestory, 501 : cloister, 45S, 
 497, 5°5 : Norman, 456, 45S, 459: tri- 
 forium, 531, 532. Bishop's Palace. 
 Barrel vault, 284. St Andrew. Chancel 
 arch, none, 227. Clerestory, 547. Pier 
 arches, no hood-mold, 407. Vault piers, 
 540. St Peter Mancroft. Chancel arch, 
 none, 227. Clerestory, 547. Roof, hybrid, 
 571. St Stephen. Arches, four-centred, 
 267. Chancel arch, none, 277. Roof, 
 hammerbeam, 566 : lead-covered eaves, 
 392 : longitudinal braces, 569. Tower, 
 228. Vault piers, 540. 
 
 NOVA R A. Atrium, 201. 
 
 NOYON. Alternating piers, 317, note. Sex- 
 partite van/ting, nave, 320. 
 
 NUN MONKTON. "Northern school'' of 
 Gothic, 108. Windotv passage, 547. 
 
 NUREMBURG. S. Lorenz. Piers, 245, 
 note. 
 
 OAKHAM. Abacus, fourteenth century, 
 scjuare-edged, 442, note. Capitals, with 
 grotesque figures, 416. Castle. Transi- 
 tional foliage, 419. 
 
 OLD BASING. Roofs, 570. 
 
 OLD SARUM. Rectangular east end, 171. 
 
 OLD SHOREHAM. Roof, thirteenth cen- 
 tury, 572. Windows in Norman but- 
 tresses, 358. 
 
 ORBAIS. Bar tracery, 469. 
 
 ORFORD, 126. Columns, 256, note. 
 
 ( )l\ LEANS. Two rypti roined vaults, 
 292. 
 
 ORLEANSVILLE. Early Christian last 
 
 lira, 1 46, 23 1 . 
 
 OTTERY ST MARY. Eastern aisles and 
 arches, 172. East front, 82. Lady chapel, 
 172: screen of, 181. Plan, 80. Tran 
 
 septal towers, 503: effect of, 594. Vault 
 
 ing, 82. Win, lows, 477: late lancets 
 (fourteenth century), 464, note. 
 
 ( >l'ISTkKI I.\ M. in lor 1 urn. absorbed into 
 clerestory, 532, note. 536. Quasi-sexpai 
 lite vault, 321. 
 
 OUNDLE. Spire, construction, thickness 
 ol stone, 636. Tower, battlements, splayed 
 loopholes, 398, note. 
 
 OUTWELL. Chancel chapels, 226. 
 
 OXFORD. Cathedral. Arches, Norman 
 
 square-edged, 276: retrograde semicircu- 
 lar, 101 : pointed arches, 265 : four- 
 centred, 267. Base, "pudding," 450. 
 Capitals and foliage, 414, 41 6, 41N, 428, 
 430, 432, 435. Fixing buttresses, mine, 
 372. Gothic and Romanesque, melange of 
 styles, 46. Internal elevation, 27. Lady 
 chapel, position of, 53. Sanctuary, un- 
 aisled, 176. Shrine of St Frideswide, 183, 
 note; naturalistic foliage, 433 : pedestal of, 
 187. Spire, 617, 621 : lights, 627 : propor- 
 tion, short for tower, 631. Triforium, 
 absorption into pier arcade, 27, 535. Vault- 
 ing, choir, lierne, 341, 347 : pendants, 305 : 
 web, thickness of, 304. Windows, alter- 
 nating, Latin chapel, 490 ; belfry, double, 
 513: plate tracery, 468. All Souls' 
 College. Stained glass, 493, note. JLan 
 vault, 343. Chrisi Church. Staircase, 
 fan vaulting, 344. Divinity School. 
 Arch, straight sided, 260. Basement course, 
 quatrefoil panelling, 403. Buttresses, 361. 
 Vaulting, lierne, 341, 347 : pendants, 305: 
 tiercerons, 337. Windows, 485, 496. 
 Holywell Church. Horseshoe arch, 262. 
 Jesus College. Curvilinear windows, 
 479. Magdalen College. Tower, 610: 
 basement, 604 : buttresses vertical, 605 : 
 pinnacles, 606, 60S : summit, ornament at, 
 610. Merton College. Buttresses, 
 gabled, 363, 3(17. East front, 68. Hood- 
 molds, stops of, 408. Tower, bases of 
 piers, 452. Vestry, 132 : ogee arches, 271. 
 Windows, 476, 479, 484, 4S5, 4S9. Old 
 Convocation House. Vaulting, knife- 
 edge ribs, 301. Proscholium. Vault, 
 lierne ribs, 340. St Giles. Lancet win- 
 
"6o 
 
 INDEX OF PLACES. 
 
 dows, 465. St John's College. Fan 
 vaulting, 344. St Mary. Arch moldings, 
 280. Bases, 449, 454. Capitals, 445, 
 449. Piers, 242. Spire, construction, 
 thickness of stone, 636: lights, 627 : pin- 
 nacles, 624: straight sides, 632. Steeple 
 (tower and spire), height, 631 : ornament 
 of, 633 : unity of tower and spire, 632. St 
 Mary Magdalen. Buttresses, 358. St 
 Michael. Anglo-Saxon tower, 590. 
 Windows, 477. St Peter-in-the-East. 
 Crypt, 191. Griffes, 454. Pinnacles, 
 circular, 362. Priest's chamber, 221. 
 Vault rib, 301. Schools. Tower, vault 
 tiercerons, 337. University. Continu- 
 ance of Gothic in seventeenth century, 142. 
 
 PALATITZA. Compound pier, 236. 
 
 PALENCIA. St John de Bonos. Tran- 
 septs, 196. 
 
 PALMYRA. Temple of the Sun. Com- 
 pound pier, 236. 
 
 PARAY LE MONIAL. Arch, pointed, 264. 
 
 Narthex, 202. Periapsidal plan, 169. 
 
 Vaults — high, barrel, with aisles groined, 
 
 13- 
 PARENZO. Atrium, 201. Cubical capitals, 
 
 411. 
 
 PARIS. Notre Dame. Aisles, double 
 aisles, 378: superposed, 528, 529. But- 
 tress chapels, 197, note. Capitals, plantain 
 leaf, 418. Fleche, 594 ; construction of, 
 620. Flying buttresses, 378. Griffes, water 
 leaf, 418. Piers, 250. Tower, central 
 lantern, originally, 592 ; early tower, 590. 
 Vaulting, ridge ribs, 336 : sexpartite, 
 320 : shafts, 240 : tas-de-charge, none, 
 303 : thickness of vault, 304. Windows, 
 478, 531, and note. Ste Chapelle. 
 Upright buttresses, 367. St Severin. 
 English style, 132. Piers, 242. 
 
 PASTON. Barn. Poof, 560, 566. 
 
 PATRINGTON. Buttresses, 360. Capitals 
 and foliage, 437. Clerestory, none, 547. 
 Elevation, interior, 82. Flying buttresses, 
 377, 624. Plan, cruciform, 80. Porch, 
 registry over, 584. Rebuilt, 132. Retro- 
 gressive, 133, note. Roof, arched braces, 
 563. Spire, on octagonal drum, 629 : 
 pinnacles, 625. Steeple (tower and spire), 
 central, 591 : ornament midway, 633: pro- 
 portion of, 629 : unity of, 632. Styles, 
 melange of, 45. Transepts, 223; double 
 
 aisled, 198; hammerbeam roof, 564. 
 
 Vault, 82. Walls, thickness of, 24, note. 
 
 Windows, alternating curvilinear, 490 : 
 east, 500. 
 
 PATRIXBOURNE. Chancel arch, horse- 
 shoe, 262. East end, Norman, rect- 
 angular, 28. Front, east, 29. Windows, 
 plate tracery, 465, note, 468: wheel, 468, 516. 
 
 PAVIA. Abaci, of brick, 410. St Michele. 
 Aisle, upper, and windows, 531. Flying 
 buttresses, concealed, 374, note. St Theo- 
 dore. Vault, 321. 
 
 PEMBRIDGE. 132. 
 
 PENMON. Spire, 617. 
 
 PENTON MEWSEY. Plan, So. 
 
 PERIGUEUX. Domes, imperfect, 13: 
 pointed, 264. Roofs, over vaults, 550. 
 
 PERSHORE. Apse, polygonal, 15S. Clere- 
 story, absorption of triforium into, 536. 
 Gothic and Romanesque, melange of 
 styles, 46. Pier, " Western," 245. Plan, 
 periapsidal, 167. Roof, weather mold, 
 389. Vault, half-barrel, 285 : spring of, 
 3°7- 
 
 PETERBOROUGH. Cathedral. 99. 
 Abacus — string-course, 404: cruciform, sub- 
 divided, 439 and note. Apses, central, and 
 foundations of lateral, 163. Arches, horse- 
 shoe, 262: pier, 238: pointed, 321, note. 
 Bases, Norman, 449, 450. Buttresses, 
 column, Norman, 351. Capitals, Nor- 
 man, 449 ; subdivided, 439 : molded, 444. 
 Ceilings, 13, 572. Choir, 184. Corbel- 
 table, nebule, 392. Doors, west, original, 
 582. Dripstones, Norman, 407. Founda- 
 tions, bad, 25 ; of apses, 163. Front, west, 
 73, 74, 116. Masonry, 20. Mortar, 24. 
 Nave, 184, 185: span of, 15: stalls in, 
 1S5. Piers, cylinders and octagons, 33: 
 "Southern," 255. Pinnacles, choir, 362: 
 retrochoir, statues, 366 : spire, 624. 
 Porches, three western, 204, 205. Retro- 
 grade, 99. Sacristies, 198. Spires, para- 
 petted, 622: south-west, 128. Styles, 
 melange of, 45. Towers, group of, 600: 
 lantern, rebuilt, 594 : west, 73, 74, 204. 
 Transepts — central, east aisles, 197 : west 
 aisle to south transept, 198; western, 204. 
 Triforium, plate tracery in spandrels, 465. 
 Vaulting — Aisles, ribs, 343: segmental 
 diagonals, 311: spring of, 343: stilted 
 transverse arches, 311. Choir, 439, note. 
 Retrochoir, fan vaulting, 344, 345. Tran- 
 
NDEX OF I'l.ACKS. 
 
 
 septs, aisle, west, pointed arches in, 321, 
 note: shafts, 240. Windows Apse, 175: 
 circular, 516: Norman, unglazed, 459. 
 Triforium, 532. 
 
 PETWORTH. Marble, 250. 
 PIDDINGHOE. Spire, plan of, 629. 
 PISA. Cathedral. Basilica, 232. Lean- 
 ing Tower. Material, excess of, 601. 
 PITTINGTON. Shaft ornaments, 256. 
 
 POITIERS. Cathedral. Vault— high 
 aisles or parallel naves, 382 : ridge ribs 
 of high vaults, 335. Notre Dame. 
 Parapets, 395. Vault, barrel, 13. St 
 Hilaire. Domes, 6, 13. Plan, peri- 
 apsidal, 169. 
 
 POLEBROOK. Buttresses, diagonal. 361. 
 Capitals, volute, 429, note. Roof, 558 : 
 low pitch of, 559. 
 
 POLESWORTH. Eastern tower, 598. 
 
 POMPEII. Compound pier, 236. 
 
 PONT AUDEMER. Triforium, arcade, 
 527 : single containing arch, 533. 
 
 PONTIGNY. Cistercian church, no tower, 
 598. Narthex, 203. Shrine of St 
 Edmund of Canterbury, 1S7, 193. Tran- 
 septs, eastern aisles and chapels, 198, note. 
 Vaults, groined, of aisles, 292. 
 
 PORCHESTER. Church. Re-carved 
 capitals, 413. 
 
 PORTSMOUTH. St Thomas. "South- 
 ern ' piers, 246. Plan, choir aisled 
 parallelogram, 179. 
 
 PRIESTHOLME, PUFFIN ISLAND. 
 Spire, 602, 617, 629. 
 
 PROVENCE. Pointed arch, 264. 
 
 PL' LI LAM. Roof, arched braces of, 562: 
 
 hybrid, 569. 
 PURBECK, ISLE OF. Marble, 250. 
 
 QUALB LOUZEH. Two low western 
 
 towers, 586. 
 QUERQUEYILLE. Central lantern tow, r, 
 
 592- 
 
 RAMSEY. Albey. Capitals, 413. Masonry, 
 20. Towers and transepts, Anglo-Saxon, 
 former, 196, 590, 600. 
 
 RAMSGATE. See ST LAWRENCE. 
 
 RAVENNA. Baptisti r\ 01 S'i ( )u ... 
 ' orbel-table, 392, note, \i.\ ilii \ Group 
 of, 155. \,u ihc\ {< losed), Z02. ( 'n vpei. 
 of Empress I'i vcida. i rue/form, 195. 
 
 M 11 50LEUM OF ( i \l II V l'i.\< [DA. I 
 
 tables, Roman, 392 and note. Windows, 
 45'). Mausoleum 01 Theodork . Win 
 dows, 459. S. Apoi,linare in Cla i. 
 Arches, orders in, 274. Buttresses, pilaster 
 Strips, 351. Confessio or crypt, [91. S. 
 Giovanni. Apses, [59 Round tower, 
 C>o 1. S. Vitale. Apses, 15c). Buttresses, 
 Roman, 351. Capitals, cubical, 411. 
 Dome, 3S9, note. Dosseret, 109. In l,i 
 
 lacings, 414. Vaults, Byzantine groined, 
 2 9 2 < -95, note, 309 : hollow pots in dome, 
 
 ;86. 
 
 ound 
 
 RAUNDS. Cusps, 510. 
 
 389, note. Towers, 
 towers, 601. 
 
 READING. Abbey. Apse and ambulatory, 
 
 167. Vault, barrel, 2S4. 
 
 RECULVERS. Apse, Anglo Saxon, 156. 
 Columns, Anglo-Saxon, 234. Nave, aisled, 
 223. 
 
 REED HAM. Roof 560. 
 
 REIMS. Cathedral. Doorway, with 
 glazed tympanum, 577. Porches, 205. 
 Prototype of Westminster, 121, note. 
 Vaulting, pointed diagonals in, 322. 
 Windows, bar tracery in, 469. St Remi. 
 Aisles, double, 156: superposed, choir, 528. 
 Ceilings, 13. Flying buttresses, 37 4. Piers, 
 illogical, 241. Plan, periapsidal, 169. 
 Vault, 292. 
 
 REPTON. Columns, Anglo-Saxon, 234. 
 Crypt, 191: "Southern"' piers, 246. 
 Transepts, 222. Window of six lancets, 
 463- 
 
 RICKENHALL. Windows, 502. 
 
 RIEVAULX. Basement course, 402, 403. 
 Capitals, molded, 446. Clerestory, passage, 
 choir, 545. Corbels, choir, foliage, 432. 
 Flying buttresses, 372 and note. Fronts, east 
 and transepts, 68. Narthex, 202. Sanctu- 
 ary, formerly unaisled, 176, 179. Styles, 
 melange of, 46. Transept, front, 68: 
 style, 46. Triforium arcade, 465 : two 
 containing arches, 534. Vault, ribs, 301 ; 
 shafts, 240. 
 
 RIPON. Cathedral. Ambulatory, rectan- 
 gular, 172. Anglo-Saxon work, 155: crypt, 
 191. 284. Arches, foiled, transept, 268: 
 pointed, 103 : semicircular, retention of, 
 265. Bases, 447. Battlements, loopholes, 
 39S. Capitals, molded, 441, 442. Choir, 
 arches, semicircular. 265 : plan, Salisbury, 
 
762 
 
 INDEX OF PLACES. 
 
 175, note. Clerestory, passage, nave, 545, 
 note. Crypt, 191,284. Doorways, trefoil 
 headed, north transept, 580 : transitional, 
 509. Fronts, east, 68; west, 71, 73 (see 
 Towers). Narthex, 204. Nave, formerly 
 unaisled, 16, 201. Pulpitum, 179, 381. 
 Roof, flat, of aisle, 538, 540, 541. Spires, 
 former, 593, 594, 619. Stalls, 179. 
 Transepts, aisle, eastern, 198; western, 
 204. Triforium, choir, transparent, 538 : 
 transept, single containing arch, 533. 
 Towers, western flanking, 73, 204, 599. 
 Vaulting, Anglo-Saxon barrel, 284. Choir, 
 wall ribs, 238. Transept aisle, ridge ribs, 
 335, 339 ; web, French method of filling, 
 335. Windows, east, 475 : semicircular 
 headed, 462: splays, 512. Chapel of 
 St Mary Magdalen. 2 1 9. 
 
 ROCHE. Advanced work, " Northern 
 school," vaulted throughout, 103, 106. 
 Sanctuary, unaisled, 176. Strings, pointed 
 bowtell, 406. 
 
 ROCHESTER. Abacus strings, 404. Aisles, 
 unvaulted, 29. Arch, early pointed, 266. 
 Bases, chapter house doorway, 444, 453 : 
 crypt, 451, 452. Capitals, molded, 446: 
 pilaster strips, 358. Choir, raised, with 
 high flight of steps, 191. Crypt, 189, 191. 
 Diaper work, 84. Flying buttress, only 
 one, 113, 372. Fronts, north transept, 
 68: west, 28. Marble shafts, 252. Nave, 
 527: unvaulted, 20. Pier, "Southern," 
 crypt, 246. Pilaster ships, Norman, 358. 
 Pinnacles, octagonal, west front, 362. Pulpi- 
 tum, 179. Sanctuary, unaisled, 176. 
 Shrine, 183, note, 189. Spire, 619. Styles, 
 melange of, 46. Triforium, 84, 266, 532, 
 note, 527 : single containing arch, 533. 
 Tower, central, rebuilt, 594. Transepts, 
 east, 189: north, front of, 68. Tympanum, 
 sculptured, 573. Vaulting, fan, abandoned, 
 Lady chapel, 346: groined, 292, 294: 
 sexpartite, 320 : spring, 307 : vaulting 
 shaft, 240. 
 
 RODA, Island of. Nilometer. Pointed 
 arches, 263. 
 
 RODMELL. Spire, plan of, 629. 
 
 ROMAINMOTIER. Narthex, 202. 
 
 ROME. Arch of Constantine, 230. 
 Arch of Septimus Severus. Acanthus 
 leaf, 426. Basilica of Junius Bassus. 
 Rojnan corbeltable, 392. Basilica of 
 Maxentius or Constantine, 230. But- 
 
 tresses, internal, pierced with aisles, 362. 
 Poof over vault, 550. Vaults and but- 
 tresses, 201. Basilicas, Five Patri- 
 archal, 146. Basilicas, Secular. 
 Atrium, 201. Baths of Caracalla. 
 Vaults, external covering of, 384: material 
 of, 303. Baths of Diocletian (S. 
 Maria degli Angeli). Composite capitals, 
 420, 425. Vault, exterior of, 384 : groined, 
 291. Baths of Titus. Vault material, 
 303. Campaniles, 586. Christian 
 Churches, 145. Cloaca Maxima. Com- 
 pound arch, 272. Colosseum. Capitals, 
 427. Vault material, 303. Monte Tes- 
 taccio. Formed from old broken wine 
 jars, 303. Pantheon. Corinthian capitals, 
 420. Dome, 281. Poof over vault, 550. 
 S. Agnese. 233. S. Balbina. Pilaster 
 strip buttresses, 351. S. Clemente. Altar, 
 592. Atrium, 201. Screen walls, 183. 
 S. Lorenzo. Capitals, 427. Columns, 
 231, 233. S. Maria degli Angeli 
 (Baths of Diocletian). S. Maria in Ara 
 Cceli. Columns, 231, 233. S. Maria in 
 Cosmedin. Apses, three parallel eastern, 
 159. S. Maria in Dominica. Apse, 
 early eastern triple, 159. S. Maria 
 Maggiore. Columns, 231, 233. Ritual- 
 istic divisions, 183. Round tower, in 
 mosaic, 586. St Paul's extra Muros. 
 Columns, 231, 233. St Peter's. Bases, 
 449. Confessio, 191. St Peter's (Old), 
 5. Atrium, 201. Sacristy and library, 
 159. S. Prassede. Columns, 234: alter- 
 nation of, 317. S. Pudenziana. Arches, 
 orders in, 274. Pilaster strip buttresses, 
 351. S. Sabina. Doors, towers, church 
 represented having two western, 586. SS. 
 Sixtus and Cecilia, 145. S. Stefano 
 Rotondo. Dosseret, 411. Temple of 
 Minerva Medica. Apses and buttresses, 
 351. Vault material, 303. Vatican. 
 Tower, 590. Vault Material. 303. 
 ROMSEY. Abacus string-course, 404. Aisle, 
 eastern, 171, 172. Apses, lateral eastern, 
 164: Apsidal chapels, transepts, 197. 
 Arches, pier, square-edged with chevron, 
 33. Capitals, incurved cones, 414, note : 
 leaf scroll, 446 : re-carved, 413. Corbel- 
 tables, 392. Doonvay, western, absent, 
 581, note. Front, west, 73. Gothic, slow 
 growth, melange of styles, 45, 46. Griffes, 
 446, 455. Lady chapel, eastern, 106. 
 Nave, east, Romanesque, not Gothic, 45 : 
 internal elevation, 26, 530. Transept, ap- 
 sidal chapels, 197. Triforium, 26, 530, 
 532, note : absorbed into pier arcade, 535 : 
 
INDEX OF PLACES. 
 
 763 
 
 blindstory, 533 : tympanum pien ed, 533. 
 Vault, stilted transverse arches, 511. 
 Window, 465. 
 
 ROSLYN CHAPEL. Flying buttresses, in 
 two flights, 37S. Vault, barrel, 284. 
 
 ROTHERHAM. Tower, central. 591. 
 
 ROUEN. Cathedral. Cylinders, 242. 
 
 Internal elevation, 529. Periapsidal plan, 
 
 163, note. Transepts, double aisles, [98. 
 
 St Ouen. Windows, 478: triforium, 
 
 double, 513. 
 ROXWELL. Windows, 502. 
 
 RUCQUEVI LLE. C hpilals, Norman, acan- 
 thus in, 427. Dome, 28 1. 
 
 RUDSTON. Windows, mullions, 508 : 
 soffit cusps, 509 : tracery, 472. 
 
 RUE CHAPELLE. Lierne vaulting, 342. 
 
 RUSH DEN. Molded capitals, 443, note. 
 
 Spire crockets, 626. Strainer arc//, 381. 
 RYE. Flying buttresses, 36S. Parallel naves, 
 
 224. Passages in clerestory and aisle walls, 
 
 547- 
 
 ST ALBANS. Cathkdral. Aisles, three 
 eastern, 172. Altars, in nave, with paint- 
 ings, 207. Apses, three parallel eastern, 
 formerly, 164. Arches, eastern, three, 172: 
 Norman, square edged, 276: pier arches, 
 238: rear arch, windows, aisles, 512. 
 Balusters, 445, note. Ceilings, 113. Chapels, 
 apsidal, 16, 197 : chantry, 211 : Lady, 68, 
 172: Saint's, 1S1, 183. Choir, 1S4. 
 Corbel-table, molded, 392. Divisions, ritual- 
 istic, 183. Door, original, 5S5. Founda- 
 tions, 25. Front, east, 68. Length, 16. 
 Masonry, 23, note, 25. Nave, 184, 185: 
 bays, western, retrograde, 113: length of 
 nave, 16, 185. Piers, 25: "Northern," 
 253. Pilasters, 239. Porches, western, 205, 
 Reredos, 181. Screens, eastern, iSi : Rood, 
 180. Shafts, 249. Shrine, pedestal of, 
 187. Stalls in nave, 185. Styles, me- 
 lange of, 45, 46. Powers, central, 593 : 
 strengthened, 594 : western, intended, 204, 
 599. Triforiutn, choir, blank arcade, 535: 
 nave, 26 ; single containing arch, 532. 
 Vaults, groined, 294: presbytery, wood, 
 570: retrochoir, no high vault in four- 
 teenth century, 82. Windows, aisle, 
 south, rear arch, 512: clerestory, choir, 
 477 ; nave, late lancets, fourteenth cen- 
 tury, 464, note ; triforium, 532. St 
 Michael. Aisled nave, 223. 
 
 ST ANDREW'S. Choir, 221. 
 
 ST BENOIT-SUR LOIRE. Ba , crypt, 
 452. Xarthe.x, 202. Transept, ea tern, 
 189. / 'cult, groined, 292. 
 ST Ckoss. (Sic WINCHESTER.) 
 ST CUTHBERT'S. Broach y\ , ,62 
 ST DAVID'S. Cathedral A> . n tro 
 grade, semicircular, 33, toi. Capitals, 
 414: West 11I" England, astragal omitted, 
 434. Elevation, internal, 27. Front, east, 
 Lady chapel, 08. A'ovr, date of, retro 
 grade, coeval with that of Wells, 42: 
 ceiling of, 570. Sanctuary screen, [80. 
 Shrine, and pedestal of, 183, note, 187. 
 Styles, melange of, 46. Tower, central, 
 51)3 : walls, cores of, 24; strengthened, 594. 
 Triforium, absorbed in clerestory, 536: 
 arch, no containing. 534. Vaulting, sex- 
 partite intended, 320. Bishop's I'm \< i. 
 Ogee curb, 268. 
 
 ST DENIS. Acanthus leaf, 426. But- 
 tresses, former flying. 372. Crypt, 194. 
 ( lerestoiy construction, skeleton, 57. Tlan, 
 originally a crux commissa, 105. Trans- 
 parent triforium, 537. Vault, domical, 
 310: ribs, diagonal, 300. Windows, 
 clerestory, 57. 
 
 ST FLORENT. Ridge rib, 339. 
 
 ST GABRIEL. Apses, three eastern, 163. 
 Buttresses, column, 351. Triforium, 532, 
 note : interior Romanesque flying but- 
 tresses, 371. Vault, quasi-sexpartite, 321. 
 
 ST GALL, 6. Apses, three eastern, 163. 
 Chapels, 207. Choir, 1S4. Plan, peri- 
 apsidal, 169. Towers, round, planned, 
 60 1 . Transept, 1 96. 
 
 ST GENEROUX. Apses, three parallel 
 eastern, 163. Transept, 196. Windows, 
 large Romanesque, 458. 
 
 ST GEORGES DE BOSCHERVILLE. 
 Apses, three parallel eastern, 163, 164. 
 Capitals, cubical, painted, 411 : scalloped, 
 413. East end and front, 28. Piers and 
 detached shafts, 237: and stone principals, 
 235. Transepts, end or return aisles, use 
 of, 198, and note, 542. Triforium, 532, 
 note: use of, 542. Vaults, 290: high 
 groined, 292: ridge ribs, 335. note. 
 
 ST <;ERMAIN t -DKS-PRES. Transept of, 
 former church, 196. 
 
 ST GERMER. Arches, pointed, 265. But- 
 tresses, 367. Flying buttresses, internal, 
 Romanesque, 371. Vaults, groined, upper 
 aisles, choir, 292. 
 
764 
 
 INDEX OF PLACES. 
 
 ST GERVAIS. Window, 508. 
 
 ST GENOU. Basilica, 232. 
 
 ST HILAIRE. Ridge rib, 339. 
 
 ST LAWRENCE. Chancel aisles, 226. 
 
 Parallel naves, 224, 382, note. 
 ST LEU D'ESSERENT. Transeptal 
 
 towers, 593. 
 ST LO. Capitals, omitted, 303. Cylinders, 
 
 243- 
 ST MARGARET- AT -CLIFFE. Arches, 
 
 pier, Norman hood-mold, 407. Plan, 
 
 aisled nave, unaisled chancel, 19, 223. 
 
 Windows, clerestory, 546 : Norman, 459. 
 
 ST MARGARET'S. Chapel. Door, origi- 
 nal, thirteenth century, 5S5. 
 
 ST MARTIN-DES-CHAMPS. Ambulatory 
 diagonal ribs, 300. 
 
 ST NEOTS. Tower, ornament, disposition 
 of, 608 : pinnacles, spear-head, 364 : strings 
 and set-offs, 606. 
 
 ST NICHOLAS-AT-WADE. Chancel aisles, 
 226. 
 
 ST PHILBERT-DE-GRANDLIEU. Ar- 
 ches, orders in, 274. 
 
 ST POL-DE-LEON. English style, 132, note. 
 
 ST QUENTIN. Transept, eastern, 189. 
 Triforium transverse arches, 369, note. 
 Vaulting, twisted, 299. 
 
 ST RIQUIER. Tower, central, 590 ; open 
 louvre, 592, 617. Triforium, absent; 
 two stories only, 539. Vaulting, Heme, 
 342 : longitudinal ridge ribs, 336. 
 
 ST SATURNIN. Tower, central lantern, 
 
 59 2 - 
 ST SAVIN. Periapsidal plan, 169. 
 
 ST SIMEON STYLITES. Apses, eastern, 
 159. Corbel-table, arched, 392, note. Plan, 
 cruciform, 195. 
 
 ST THOMAS. Roof on vault, 550. 
 
 ST URBAIN-DE-TROYES. (See 
 TROYES.) 
 
 SAINTES. Spire, entasis of, 632. 
 
 SALAMANCA. Lie me vaulting, 342. 
 
 SALERNO. Atrium, 201. 
 
 SALFORD. Windows, curvilinear, 484 : 
 Flamboyant, 485, 487, 48S. 
 
 SALHOUSE. Large leaf foliage, 437. 
 
 SALISBURY. Cathedral. Aisles, east- 
 ern, 172. Altitude, 61. Arches, eastern, 
 172: strainer, 381. Area, 50, note. Ball 
 
 flower, 625. Bases, cloister, 450. Base- 
 ment course, 405. Boundary line, 50. But- 
 tresses, in three stages, 358, note : saddle- 
 back gables, 363. Capitals, molded, 434, 
 446 : volute, 430. Chapter house, con- 
 struction skeleton, 58 : Gothic, 10. Clere- 
 story, 26, 533: wall, thickness of, 374. 
 Plying buttresses, 30; concealed, 374; 
 choir formerly had none, 372. Profits, 
 east, 68 : transepts, 71 : west, 73. Griffcs, 
 458. Horizontality versus verticalitv, 397, 
 405. Lady chapel, 172 : three parallel 
 naves, 383. Parapets, 394, 397, 405, 
 Piers, "Southern," Lady chapel, 255: 
 nave, 246. Pinnacles, 362 : of spire, 624. 
 Plan and grouping, 175. Proportion, 54, 
 61. Shafts band, 249. Spire, angle, 
 sharp, 631 : angles, with ball flower, 625 : 
 bands round, 626 : construction, 636 ; 
 pinnacles, 624 : squinches, 636 : stone, 
 thickness of, 636 : straight sided. Steeple, 
 tower and spire, central, 593 : proportion 
 of, 629 : tower, construction of, built in 
 two shells, 602 ; effect of, 594; strength- 
 ened, 594. Tower, detached, destroyed, 
 598. Transepts, eastern, 189. Triforium, 
 single containing arch, 533. Vault, 
 material, 304: ploughshare, 311: quin- 
 quepartite, 339 : rib moldings, 301 : 
 shaft, 240 : spring, 307 : chapter house, 
 with central pillar, 339. Walls, thickness 
 of clerestory, 374. Windows, early geo- 
 metrical tracery, 468, note, 472, 490 : plate 
 tracery, 465 : lancets, plain outside, glorious 
 within, 464 : in triplets, 463, 543 : size of 
 chapter house windows, 490. 
 
 SALL. Door, original fourteenth century, 
 585. Doorway, spandrels, 580. Drip- 
 stone label, 407. Parvise with piscina, 
 584. Pinnacles, 366. Porches, vaulted, 
 558, note, 582, 584. Rood beam, 180, note. 
 Roofs, arched braced, 564 : foliated bosses, 
 94: traceried spandrels to trusses, 570. 
 Windows, 503, 504. 
 
 SANDRIDGE. Hybrid rooj r , 569. 
 
 SAUMUR. Hotel de Ville. Four- 
 centred arch, 267. 
 
 SCARBOROUGH. St Mary. Corbel- 
 table, 392. Stone roofs, 2 88. 
 
 SCARTHO. Acanthus, in Anglo-Saxon 
 work, 427. 
 
 SCHAFFHAUSEN. Basilica. 232. 
 
 SEAFORD. Roof, 224, 570. 
 
 SECQUEVILLE. Norman windows, 459. 
 
[NDEX OF PLACES. 
 
 .'■ 
 
 SEDGEFIELD. Cruciform extension, 222. 
 
 SEGOVIA. Lierne vaulting, 342. 
 
 SELBV. Apses, probably three parallel, 
 163. Arches, north porch, semicircular 
 and pointed, 265. Buttresses, choir, 560 : 
 Capitals and foliage, 436. Choi?; [79: 
 clerestory, passage, 82, 397, 545, note. 
 Flying buttresses, absence of, 3(10, 372. 
 Front, east, 82. Gargoyles, 399. Gothic 
 growth, slow : melange of styles, 46 : 
 "Durham school'' of, 100. Parapets, ex- 
 terior, perforated, figures on, 366, 396 : 
 interior pierced, 82, 397. Pinnacles, 362, 
 363, 364, 366, 372. Piers, alternating, 
 317, note. Roofs, 541. Screen, Flam- 
 boyant tracery, 485. Scdilia, 87, note. 
 Shaft ornaments, 256. Tower, central, fall 
 of, 594, note. Transepts, eastern aisle, 198. 
 Triforium, choir, absorbed into clerestory, 
 536: nave, one arched arcade, 532. 
 I 'anlt, diagonals, segmental, 311: shaft, 
 240: wooden vault, 570. Windows and 
 tracery, choir aisle, 508 : curvilinear, 471), 
 490 : Flamboyant, screen, 485 : geometri- 
 cal, early, 472, 4S8. 
 
 SEMPRINGHAM. Door, original Norman 
 deal, 585. Doorway, gabled, 579. 
 
 SEMUR-EN-AUXOIS. Crocket capitals, 
 429, note. Nave, triforium absorbed into 
 clerestory, 537. 
 
 SENS. Copied at Canterbury, 108. Tri- 
 forium, choir, two containing arches, 534. 
 Vault, ribbed, 298: sexpartite, 320: wall 
 ribs, 23S. 
 
 SEVERUS, ROMAN BRIDGE OF. 
 Pointed arch, 263. 
 
 SHERBORNE. Arches, stilted. 262. Choir, 
 east end rectangular, aisles, apsidal, 164. 
 Doorway, Anglo-Saxon masonry, 23. 
 Flying buttresses, choir, none in nave, 345, 
 372, 374 : spandrel pierced, 377. Porch, 
 Norman, 5S2 : two stories high, 205 
 Styles, melange of, Gothic veneer on Nor- 
 man work, 46, 49. Tower, central, piers 
 of, 24, note : strengthened, 594. Tri- 
 forium, 88. Vault, fan, thrust of, 343, 
 345: material, 304: spring of, 307: web, 
 thickness of, 304. Window, east, 505. 
 
 SHIPTON. Spire pinnacles, 623. 
 
 SHOBDON. Capitals and shafts, inter- 
 twining monsters and grotesques, 416 and 
 note. 
 
 SHOTTESBROOKE. 132. Corbel-table, 
 molded, ^,^. To,-,,, ;,,;. Spin; con 
 struction, squim hi s, 636 . parapi tti d, 622. 
 Plan, cruciform, 80. Transepts, 
 
 Windows, east, 503 : Flamboyant, 1 1 
 SHREWSBURY. Abbey. Abacus, cin ular, 
 439- Aisle roof, flat, 537. Arch, square 
 edged, unniolded, Norman, ^\, 27c,. 
 Styles, melange of Gothic and' Roman 
 esque, 45. 46. Towers, western and 
 central, 599. Triforium, transparent, 527, 
 537. Window, west, 504. Battlefield. 
 Windows, 501, note. Si Mary. Arch, 
 chamfered and molded, 277. Capitals, 
 interlacing foliage, 432 : nave, north arcade, 
 Llandaff type, 434 : south arcade, West 
 of England type, 433. Hood-molds, mitring, 
 407. Nave, "Western school " of Gothic, 
 105 : see also Capitals. Porch, plate 
 tracery, 468. 
 
 SIENA. Polychromy, 250. 
 
 SILCHESTER. Romano-British Basilica. 
 Cruciform, 195. Nave, aisled, 223. Tran- 
 septs, 222 : western, 203. 
 
 SIMONBURN. Priest's doorway, 582. 
 SINAI. St Catherine. Apse and side 
 chambers, 151), note. 
 
 SK ELTON. Doorway, 579. 
 
 SKIRLAUGH. Un aisled chapel type, 22S. 
 
 SLAUGHAM. Spire, plan of, 629. 
 
 SLEAFORD. Windows, glazing of, 514. 
 
 SNETTISHAM. Clerestory, 547. Galilee, 
 202. Plan, cruciform, 80. Size and pro- 
 portion, 132, 133. Spire, 12S. Window, 
 west, Flamboyant, 489. 
 
 SOISSONS. Foliage, fern leaf, 418, note. 
 Vault, tas-de-charge, 303. 
 
 SOLIHULL. Chantry chapel, and sacristy, 
 226. Poof hybrid, 569. Spire, perfora- 
 tions, 626 : strings round, 626. Windows, 
 tracery, 477 ; dagger in, 503, note. 
 
 SOMERSET. Churches, rebuilt de novo, 46. 
 
 SOMPTING. Anglo-Saxon -work, acanthus, 
 427 : piers, compound, 237 : pilaster 
 strips, 35S. 
 
 SOPELEV. Tower, Norman, central, 590. 
 
 SOUILLAC. Capitals and shafts, inter- 
 twining monsters on, 416, note. 
 
 SOUTH AM. Clerestory, 547. Spire, 621 : 
 strings round, 626. 
 
 SOUTHEASE. Spire, plan of, 628. Win- 
 dows, shuttered, 459. 
 
■]66 
 
 INDEX OF PLACES. 
 
 SOUTH ELMHAM. Tower, Anglo-Saxon, 
 development of, 596. Porch, western, 
 202. 
 
 Town 
 
 no buttresses, 
 
 SOUTHFLEET. 
 362, 605, note. 
 
 SOUTH HAYLING. Tower, Norman 
 central, 590. 
 
 SOUTHWARK. Cathedral (St Mary 
 Overie or St Saviour). See LONDON. 
 
 SOUTHWELL. Cathedral. Abacus 
 strings, 404. Aisles, choir, apsidal ended, 
 164: bays, oblong, 312. Arches, horse- 
 shoe, 262: pointed, 263: segmental, 
 doorway, 262, 267, 580. Bases, 450, 452. 
 Buttresses, chapter house, pinnacled, 363 : 
 choir, sharp chamfered, 358. Capitals, 
 chapter house, foliage of, naturalistic, 435, 
 436 : choir, 432 : molded, 446. Choir, 
 plan, rectangular east end, 164. Clere- 
 story, 29. Corbel-tables, nebule, and para- 
 pets, 392, 394. Diaper-work, screen, 84. 
 Doorway, segmental headed, south tran- 
 sept, 262, 267, 580. Flying buttresses, 
 choir, formerly none, 372 : top surface 
 channelled for aqueduct, 377, 400. Fronts, 
 east, 68: west. 28, 71, 73. Gargoyles, 
 400. Parapets, 394. Piers, choir, 
 "Northern," 253; nave, cylinders, 33. 
 Pinnacles, chapter house, 363 : choir, 363, 
 367 : porch, 362. Porch, Norman, north, 
 207, 582 : two stories high, with parvise 
 and chimney, 205, 584. Pulpitum, 84, 
 179. Roof drainage, transepts, 37. Sanc- 
 tuary, unaisled, 176. Spires, 617. String- 
 courses, 404, 405, 606. Styles, melange 
 of, 46. Towers, western, 73 : arches, 
 horseshoe and pointed, 262, 263 : string- 
 courses, 606, note. Transepts, eastern, 189. 
 Triforium, choir, absorbed into clerestory, 
 536: nave, 519, 533. Vault — Aisles, 
 quinquepartite, 339; rib moldings, 301; 
 segmental diagonals, 311; stilted trans- 
 verse arches, 311. Chapter house, no 
 central pillar, 339. Choir, longitudinal 
 ridge rib, 336. Windows, east, lancets in 
 groups of four, 463 : nave clerestory, cir- 
 cular, 517. 
 
 SOUTHWOLD. Presbytery, 226. 
 
 SOUVIGNY. Abbey. Narthex, 202. 
 Eastern transept, 189. Vaulting, longitu- 
 dinal ridge rib, 336. 
 
 SPALATO. Diocletian's Palace. Arch 
 071 columns, 233, note. Coursed lintel, 260, 
 note. Roman grippes, 454. 
 
 SPALDING. Windows, North porch, 504. 
 
 SPARSHOLT. Roof, 558: tie beam used 
 as principal, 563. 
 
 SPEYER. Groined vaults, 6, 13, 294. 
 
 SPOLETO. Basilica of the Crucifix. 
 
 Central lantern tower, 592. St Saviour. 
 
 Front, 155, 156. 
 
 SPORLE. Plate tracery, 468. 
 
 STAMFORD. Churches. Masonry, 20. 
 All Saints. Capitals, leaf scroll and 
 volutes, 429, note, 430. Piers, "Southern," 
 249 and note. Tower and spire, turrets, 
 623. St Leonard's Priory. Capitals, 
 leaf scroll and volutes, 430 : direction of 
 foliage, 432. Molding, bead and roll, 41. 
 St Martin. Tower, built for spire, 
 squinches, 632. St Mary. Capitals, 
 tower arch, 431. Spire, 617, note. 
 
 STANION. Broach spire, 620. 
 
 STANTON LACY. Transepts, 222. 
 
 STANWICK. Arches, pier, four-centred, 
 
 267. Foliage, stop to hood-mold, 408. 
 
 Nave, aisled, 224. Spire, plan of, 629. 
 
 Window splays, 512. 
 STEWKLEY. Choir, 221. Vault and 
 
 chamber above, with windows, 221. 
 
 STEYNING. Arch, carving on, 276: pier 
 arches, Norman hood-mold, 407. But- 
 tresses, Norman pilaster strips, 358. 
 Capitals, coniferous, 413; cones omitted, 
 414: foliage, fern leaf, 418. Clerestory, 
 546. Corbel-table, 392. Plait, 19. Win- 
 dows, Norman, 456, 459. 
 
 STOCKTON-ON-TEME. Material, tufa, 
 303, note. 
 
 STOKE. Door, original fourteenth century, 
 585- 
 
 STONE. Aisles, lofty, 67. Doorway, 
 chevron molding, 40. Piers, " Southern," 
 249. Plan, 67. Window s, double, 513; 
 lancets, 464 : plate tracery, 469. 
 
 STOW. Chancel, Norman high vault, 314. 
 Tower, Anglo-Saxon, central, 590. Tran- 
 septs, Anglo-Saxon, 222. 
 
 STRATA FLORIDA. Crockets, 77. 
 
 STRATFORD-ON-AVON. Clerestory, 547. 
 Door, original fifteenth century, 585. 
 
 STRATFORD - ST - MARY. Battlements, 
 flint panelled, 399. 
 
 STRUMPSHAW. Piscina, stop to hood- 
 mold, 408. 
 
[NDEX OF PLACES. 
 
 7<-7 
 
 STUDLAND. Tomer, Norman central, 
 
 59°- 
 STURRY. Aisled nave, 223. 
 STUSTON. Porch, roof, 559, 567. 
 
 SUDBURY. St Gregory. Chancel, ceil- 
 ing, 570. 
 
 SUFFOLK. Churches rebuilt de novo, 46. 
 
 SUTTON ST MARY. Arches, late retro- 
 grade semicircular, 33. Capitals, scalloped 
 and coniferous, 413. Plan, 19. Retro- 
 grade, n, 43, 100. Spire, 617. 
 
 SYRIA. Roman stone ceilings, 2S5. See 
 QUALE LOUZEH, ST SIMEON STY- 
 LITES, TOURMANIN. 
 
 TANGMERE. Tower, no buttresses, 362. 
 TARRAGONA. Cathedral. Double door 
 way, 38, note. 
 
 TAUNTON. St James. Tower, insignifi- 
 cant pinnacles, 608. St Mary Magdalen. 
 Capitals, angels as, 95, 416. Parapets, 
 pierced, 397, 399. Tower, lofty, 604 : 
 parapets and pinnacles, pierced, 366, 399, 
 604, 606 : pinnacles important, 608 ; with 
 hanging buttresses, 606. 
 
 TEMPLE BALSAL, 220. No chancel arch, 
 227, note. 
 
 TEMPSFORD. Door, original fifteenth 
 century, 585. 
 
 TENBY. Doorway, with ogee arch, 268, 
 580. Pier arcade, no capitals, 409. Roof, 
 563- 
 
 TENTERDEN. Roofs, foliated bosses, 94. 
 
 TERRINGTON ST CLEMENT. But- 
 tresses, diagonal, 361. Flying buttress, 
 one only, 368. Front, west, 91. Plan, 
 cruciform, 88. Screen, western, 203. 
 Tower, ornament, disposition of, 608 : 
 position of, 228, 596 : pinnacles, 608. 
 Windows, Perpendicular, 501 : west, 228. 
 
 TERRINGTON ST JOHN. Tower, posi- 
 tion of, and west window, 228. 
 
 TEWKESBURY. Abbey. Abacus, cir- 
 cular, 439. Aisles, nave, high, 530. Am- 
 bulatory, 65. Apse, polygonal, 65, 158. 
 Apsidal chapels, 65, 197. Arches, flat, in 
 porch, 260 : molded slightly in nave, 33 : 
 pointed, 265. Rases, Attic, porch, 451. 
 Clerestory, 80 ; passage, 545. Corbel-table, 
 arched, 392. Flying buttresses, choir, 368 ; 
 nave, absent, 372. Front, west, 28. Mold- 
 
 ings, chevron and roll, 40, 277; ;ligh hi 
 
 ings in arches, 33. Parapets, pierced, 196 
 internal, need of m choir, 397. /' 
 cylinders, short in choir, 23 | ; tall in 
 33. 2 4 2 - Pinnacles, open work, 363. 
 Plan ; periapsidal with tangential apsidal 
 'Impels, 65, 158, 1 (, 7 . Porch, ll,.i an h in. 
 260: Attic bases, .151. I'm/;,, lion, 53. 
 Roof, pitch of, weather molds, 3X9, 559. 
 Stained glass, 493. Styles, melangi of, 
 Gothic veneer over Norman work, 40, 49 
 Tower, central, 593. Triforium, choir, So, 
 3117. Vaulting, aisles, upper, half barrel, 
 285 : choir, lienie ribs, 340 : nave, ridge- 
 rib, triple longitudinal, 336; low spun of, 
 306, 374: fan vault with painted ribs, 
 tomb of Sir Hugh Despenser, 342: tran- 
 sept, south, eastern apse, keystone vault, 
 305 : Trinity chapel, fan vault, 342. Win- 
 dows, choir, upper aisles, 532 : curvilinear, 
 477, 490: alternating, 490: glass in. 495. 
 
 THANET, ISLE OF. Arches, 274. 
 
 THAON. Buttresses, none, 362. Prin- 
 cipals, stone, intended, 235. Spire, 602, 
 617 : plan of, 629. 
 
 THORNEY. Abbey. Masonry, 20. 
 
 THORNTON. Abbey. Choir, 179. 
 
 THRECKINGHAM. Capitals, molded, 
 442, note. Steeple, point of view, 633. 
 
 TICKENCOTE. Vaulting, Norman, sex- 
 partite, 10S, 221, 320, 550. 
 
 TIDESWELL, 132. Basement course, 402. 
 Plan, cruciform, So. Shaft moldings, 256. 
 Windows, curvilinear, 483: Flamboyant, 
 485,487,488. 
 
 TILLINGTON. Capitals, plantain or laurel 
 leaf, 42S, note. 
 
 TILNEV ALL SAINTS. Capitals, leaf 
 scroll, 430 : scalloped and coniferous, 413. 
 Screen, Jacobean, 203, note. Spire, too 
 short for tower, 631. 
 
 TILTEY. Capitals, molded, 443. Window, 
 moldings of mullions, 516. 
 
 TINTERN. Abbey. Bases, 447, 449. 
 Buttresses, gabled, 367. Choir, 1 79. 
 Clerestory, no passage, 545. Flying but- 
 tresses, 30, 372. Fronts, east, 68: tran- 
 septs, 68, 71 : west, 74. Moldings, arch, 
 279. Piers, "Northern,'' 253. Shaft 
 bands, 249. Strings, 406. Triforium, 
 walled, 543. Windows, aisle, rear arch, 
 512 : east, early geometrical tracery, 475 : 
 west, late geometrical, 475, 479. 
 
768 
 
 INDEX OF PLACES. 
 
 TIVERTON. Battlements, 398 : face of, 
 carved, 399. Foliage, capitals, passion 
 flower, 95, 43S : in hollow of string-course, 
 405. Pinnacles, constructed decoration, 
 366. 
 
 TOLEDO. S. Juan de los Reyes. Piers, 
 242. Li erne vaulting, 342. 
 
 TONG. Choir, 221. Tower, Norman cen- 
 tral, 590. 
 
 TOOT BALDON. Aisles, low, 224 : aisles 
 and nave under one roof, 224. 
 
 TORCELLO. Cathedral. Apses, three 
 eastern, 163. 
 
 TOSCANELLA. Cathedral. Basilica, 
 232. 
 
 TOULOUSE. Church of the Corde- 
 liers. Hall plan, 321 : no exterior aisles, 
 228, but aisles pierced through internal 
 buttresses, 362. Church of the Jacobins. 
 Vaults, counter-thrusts of, over parallel 
 naves, 382. St Sernin. Aisles, double, 
 156. Periapsidal plan, 169. 
 
 TOURMANIN. Syria. Eastern limb, 159. 
 
 Towers, two low western, 586. 
 TOURNAY. Superposed aisles, 528. 
 
 TOURNUS. Aisles, high, 531. Dome, 
 central, with tower over, 592. Narthex, 
 202. Piers, alternation of, 243, note. 
 Plan, periapsidal, 169. 
 
 TOURS. St Martin. Aisles, double, 156. 
 Plan, periapsidal, 169. Tin roofs, over 
 stone vaults, 389. Tower, 5S6 : former 
 central lantern, 592. Transept, 196. 
 
 TOWYN. Clerestory, 546. Nave, aisled, 
 223. Retrograde, 43. Windows, Norman, 
 459, 546- 
 
 TREVES. Basilica. Aiskless nave, 201. 
 Roman Palace. Orders in arch, 274. 
 
 TROYES. Cathedral. Battlements, 399. 
 Doorway, western ceremonial, 581. St 
 Urbain. Flying buttresses, 368, 375, 
 note. Piers, 242: "Southern," 249. 
 
 TRUNCH. Arches, pier, no hood-mold, 407. 
 Capitals, molded, 444. Doorway, priest's, 
 
 1 582. Roof, hammerbeam, 365, 566. 
 Tower, buttresses set diagonally, 605, and 
 not carried up to top, 606 : set back of 
 
 . stages, top stage battered, 605 : skyl-ine, 
 horizontal, 606. 
 
 TUDDENHAM. Bases and benches, 447. 
 
 TULUN (CAIRO), Mosque of. Pointed 
 arches, 263. 
 
 TUTBURY. Triforium arcade, 527. 
 
 TWYNEHAM. The Ten Churches. 
 
 207. 
 
 TYNEMOUTH. Choir, ceiling, retrograde, 
 113. East front, 68. Sanctuary unaisled, 
 176. 
 
 TYRE. Windows, with pierced wooden 
 panels, 459, note. 
 
 UFFINGTON. Octagonal tower, 600. 
 UPTON. Keystone vault, 305. 
 UP WELL. Aisle roofs, si o. 
 
 VALLE CRUCIS. Capitals, foliage, char- 
 acter of, 432 : intersecting stalks, 432. 
 West front, 74. Windows, bar tracery, 
 470. 
 
 VAUX-LE-CERNAY. Chapels, elongated 
 transeptal, 167. 
 
 VENDOME. La Trinite. Spire, 617. 
 
 VENICE. St Mark. Cubical capitals, 
 411. Domes, 28 1. 
 
 VERONA. Cathedral. Polychromy, 250. 
 Vaults, high in squares, aisles in oblongs, 
 321. S. Stefano. Ambulatory, semi- 
 circular, groined vaults, 292, 295, note. 
 S. Zeno. Capitals, with intertwining 
 monsters, 416. Piers and stone principals, 
 2 35- 
 
 VEZELAY. Arch, elliptical, nave, 262. 
 Doorway, double, 38, note. Hood-molds, 
 277. Narthex, 202. Vaults groined, upper 
 aisles, choir, 292 : nave, 294. 
 
 VICENZA. Basilica. Roof, 564, note. 
 SS. Felix and Fortunatus. Piers and 
 columns, 236. 
 
 VIENNA. St Stephen. One transeptal 
 tower, 593, note. 
 
 VIGNORY. Periapsidal plan, 169. 
 
 VITERBO. Tower, 586, note. 
 
 WADENHOE. Tower, saddle-back roof, 
 602. 
 
 WAKEFIELD. Aisles, 224. 
 
 WALPOLE. St Mary. Western screen, 203. 
 St Peter. Vault, chancel subway, 53. 
 Windows, perpendicular in curvilinear 
 church, 501. 
 
 WALSINGHAM, OLD. Curvilinear -win- 
 dows, 484, 487, 489. 
 
[NDEX OF PLACES. 
 
 7 r "J 
 
 WALSOKEN. Aisles, chancel, 226. Arch 
 moldings, 279. Capitals, conifer and 
 scalloped, 413, note : waterleaf, 410. 
 
 WALTHAM. Abbey. Clerestory, Norman, 
 465. Hood-mold, pier arcade, Norman, 
 407. Piers, alternating, 317, note. Shaft 
 ornament, 256. Styles, " Durham school," 
 100: melange of, 46. Toivers, western and 
 central, 599. Triforhim, 527 : single con- 
 taining arch, 533, note. Windoivs, circular, 
 4S9, 5 J 7 : glazing, 459 : triforium, 533. 
 
 WALTHAM CROSS. Eleanor Cross. 
 1 1. 
 
 WAREHAM. Piers, compound, Anglo- 
 Saxon, 237. 
 
 WARKWORTH. Vault, high, Norman, 
 
 chancel, 314 : segmental diagonals, 311. 
 
 VVARMINGTON. Arch moldings, 279. 
 Buttresses, diagonal, 361. Capitals, foliage, 
 naturalistic, 435 : molded, double, 443 
 Clerestory, 546. Doorway, enriched, 579 : 
 trefoil headed, 580. Plan, broad aisles, 
 67. Roof, low pitch, 559. Spire, corbel- 
 table, 392. Vault, wooden, 571. Win- 
 dows, 465, 500, note : cusps, 511. 
 
 WARWICK. Panelling, exterior of walls, 
 Beauchamp chapel, 345. Porch, former, 
 library over, 5S4. Vaulting, skeleton, 
 
 chancel, 340. Window, east, 505. 
 
 WATER NEWTON. Arches, Gothic semi- 
 circular, 261. 
 
 WEDMORE. Porch, three-storied, 584. 
 
 WELBOURNE. Basement course, 402 : 
 shadows in, 403. Spire, sugar loaf, entasis, 
 632. 
 
 WELLINGBOROUGH. Spire, 621. 
 
 WELLS. Cathedral. Aisle, eastern, am- 
 bulatory, rectangular, 171, 172. Arches, 
 cinquefoil, 268: eastern, three, 172 : ogee, 
 87 : pier, 266 : pointed, 105. Buttresses, 
 flat, choir, 30 : chapter house, pinnacled, 
 363: clerestory, false bearing, 374. Capitals, 
 foliage, 414, 416, 432, 433, 436, 437: 
 volutes, 432 : " West of England," 433 and 
 note. Chapels, chantry, 211 : Lady, 172, 
 341. Doors, original fourteenth century, 
 585. Flying buttresses, 30: coping of, 377 : 
 internal, 371, 374 : presbytery, 377 : none 
 of twelfth century, 372. Foliage, triforium, 
 spandrels, 418: see also Capitals Fronts, 
 east, 82, 253 : west, commenced, 1 16, and 
 note : towers, 73, 80, 59S : sculpture, 573. 
 Gothic, first complete English, 105; growth, 
 
 49 : melange of styles, early and late juxta 
 posed, 46: nave coeval with thai ol Si 
 David's, 42: "Western school," 105. 
 Niches, ogee canopied, 87. Parapets, 
 choir, 394, 395 : presbyter)', pierced, 396: 
 thickness of walls, 394. Piers, choir, 24 : 
 nave, 266 : western, 245, 253 : wist front, 
 "Southern," 253. Pinnacles, chapter 
 house, 363 ; central tower, 606. I'lan, 
 rectangular east end and ambulatory, 171, 
 172: "Salisbury," 80: Presbytery, [28. 
 Proportion, 54. Retrochoir, 128. Roof, 
 over choir vault, 550. Sacristies, 198. 
 Sculpture, west front, 573. Strings, 
 
 407. Transepts, double aisles, 198 : 
 eastern transept, 189 : western transept, 
 204. Triforium, 82 : nave, continuous 
 arcade, 534 : presbytery, absorbed into 
 clerestory, 131, 536. Towers, central, 
 construction of inner shell, 602 : covered 
 with ornament, 608 : pinnacles, 606 : 
 settlement of, St Andrew crosses under, 
 381, 594 : western towers, position of, 204, 
 59S : tops not finished, 606, note. See also 
 Prun t. 1 "an It, conoid in choir, 34 r : 
 
 diagonals, 300, 
 
 domical, 
 
 early 
 
 ( iothic, ro8: fan vaulting, under tower, 348, 
 note, 593: Heme, chapterhouse, 339, 341 : 
 ploughshare, 311 : pointed acutely, 265, 
 32 r, note, 351 : ribs, 300, 322 : moldings 
 of, 301 : cloister. 301 : wall ribs, 300 : 
 roof over choir, 550 : spring, height of, 
 307: thickness of web, 304: thrust, 351. 
 Walls, thickness of, 394. Windows, geo- 
 metrical, -177 : chapterhouse, 476 : lancets, 
 broad, 462 : tower, shallow fenestration, 
 eliei t of, 603. Bishop's Hall. Windows, 
 plate tracery, 468: transom, 503. StCuth- 
 bert. Piers, "Western," 245. Tower, 
 604. 
 
 WENLOCK. Abbey. Front, west, 73. 
 Gothic, "Western school," 105. Pier, 
 " Northern," 253. 
 
 WESTMINSTER. Abbey. First Norman 
 periapsidal building in England, with 
 ambulatory, 97, 167, 169. Apse, poly- 
 gonal, with ambulatory and tangential 
 1 h.ipels, 65. 68, 157. Arches, pointed. 265, 
 2<>i>. Battlements, St Erasmus' chapel, 399. 
 Capitals, cushion, in all stages of carving, 
 crypt, 413: molded, 434, note, 446. Chapels, 
 65, 68. Chevet, 157. Clerestory and 
 construction, skeleton, 56, 58. Cusps, de- 
 tached, 510. Dripstone, dies into buttress, 
 cloisters, 407. Plying buttresses, 55 ; in 
 double flights, 371, 378; pinnacles on, 
 
 3C 
 
77Q 
 
 INDEX OF PLACES. 
 
 363. Foliage, naturalistic, arcade, chapter 
 house, 434, note. Fronts, east, 68 : north 
 transept, 71. Gothic, French, 121. Piers, 
 choir and nave, "Southern," 246: vault 
 piers, 56, 58. Planning, 65. Polychromy, 
 250, note. Proportion, 54. Shrine of St 
 Edward the Confessor, 187. Stalls in nave, 
 185. Toiver, lantern, 5 93. Transepts, north, 
 double aisled, 198 ; doorways, 584 ; front, 
 71 : south, interior, four-storied, 527. 
 Triforium, two containing arches, 534; 
 blindstory, nave, 535 : use of triforium 
 chamber, 542. Vaults — Chapter house, 
 339. Choir, piers, 56, 58 ; longitudinal 
 ridge ribs, 336, note, 339. Cloisters, east 
 side, domical, 310; south and west sides, 
 semi-elliptical diagonals, 310. Nave, 337; 
 spring of, 307 ; web, thickness of, 304. 
 Transept, short voussoirs to ribs, 301 ; tas- 
 de-charge, 303. Windows, circular, in tri- 
 forium, 517: cusps, detached, 510: size 
 of, chapter house, 490: tracery, bar, 469; 
 early geometrical, 472, 490; curvilinear, 
 484 ; reticulated, 484, 487, 488. Henry 
 the Seventh's Chapel, 208. Flying 
 buttresses, 377. Piers, 94, 540. Pinnacles 
 and ogee cupolas, 364, 54 r. Triforium, 88. 
 Vaulting, fan, or panel vault, 344, 345, 
 348 : pendants, 306 : thickness of web, 
 304. Wall cresting, parapet, battlement, 
 and pinnacle, 541. Windoivs, 504. Hall. 
 Flying buttresses, 368. Roof, 562, 564, 
 569. St Stephen's Chapel. Corbels, 
 with undulatory foliage, 436. Cathedral. 
 Domes, 281. 
 
 WESTON. Transepts, 222. 
 
 WEST WALTON. Aisles, chancel, 226. 
 Arch moldings, 279. Bases, 451. Capitals, 
 foliage, direction of, 432 : volutes, 430, 
 432. Clerestory windows, 68. Cornice, 
 392. Doorways, enriched, 579. Porch, 
 582, note. Tower, detached, 598: as lych 
 gate, 596 : ornament of, 60S. 
 
 WHAPLODE. Abacus, 441. Capitals and 
 foliage, 429, 430, 432. Roof, pitch lowered, 
 head of east window cut off, 391. Tower, 
 position of, 596. 
 
 WHISSENDINE. Doorway, west, 582. 
 Tower, angle buttresses, 605 : disposition 
 of ornament, 608. 
 
 WHITBY. Abbey. Arches, recessed orders, 
 275. Bases, 451. Basement course, 
 feather-edged, 402 : hollows in, 403, But- 
 tresses, saddle-back, or gabled, 363. Ceil- 
 ing, choir, retrograde, 113. Choir, aisled 
 
 parellelogram, 176, 179. Corbels and 
 foliage, 432, 433. Fronts, east, 68 : tran- 
 septs, 68, 71 : west, 73. Parapets, 394. 
 Strings, 406. Transepts, east aisle, 198. 
 Triforium, arcade, two large, single con- 
 taining arch, 533 : plate tracery in, 465. 
 
 WHITCHURCH. Capitals and foliage, 
 43 2 - 434, note. 
 
 WHITCHURCH CANONICORUM. 
 •' Western school" of Gothic, 105. 
 
 WHITE COLNE. Window mulllons, 516. 
 
 WHITHERNE, 218, note. 
 
 WICKHAM MARKET. Spire, plan, 629. 
 
 WIGGENHALL. St Mary the Virgin. 
 Bases and benches, 447. Capitals, molded, 
 double, 443. Roof, hybrid, 561, 562, 569. 
 Style, curvilinear with perpendicular win- 
 dows, 501. St Germans. Bases and 
 benches, 447. 
 
 WIGGENHALL ST MARY MAGDALEN. 
 Screen, destroyed, 203, note. Style, cur- 
 vilinear with perpendicular windows, 501. 
 
 WILBY. Arches, ogee, 268. Spire, on octa- 
 gonal drum, with flying buttresses, 625, 
 629. 
 
 WILEY. Lancets, 465. 
 
 WILLINGHAM. Stone roofs, 288, note. 
 
 WIMBOURNE. Crypt, 194. Sanctuary, 
 unaisled, 176, and note. Styles, melange 
 of, 45. Towers, western and central, 599. 
 Windoivs, 477. 
 
 WIMBOTSHAM. Boarded roof 560. 
 
 WIMMINGTON. Aisles, 226. Curvilinear 
 windows, 501. 
 
 WINCHELSEA, 126. Arch moldings, 280. 
 Sedilia, bases, 452. 
 
 WINCHESTER. Cathedral. Aisles, 
 eastern, three parallel, crypt, 167, 172. 
 Altars, three eastern, 172. Altitude, 61. 
 Ambulatory and apse, crypt, 167. Arches, 
 square-edged, Norman, t,^, 276: three- 
 centred, 267. Area, 15, and note. Bosses, 
 305. Capitals, re-carved cushion, Lady 
 chapel, 413, note. Chantry chapels, 211 : 
 Saint's, 183. Choir, 1S4. Clerestory, 543 : 
 passage in presbytery, 545, note, but not in 
 nave, 545. Corbel-table, 392. Crypt, 167, 
 172, 191, 193. Divisions, ritualistic, 183, 
 184. Dripstone hood-molds, none in early 
 Norman work, 407. Flying buttresses, 30, 
 368: concealed, 374; crocketted, 377. 
 Font, 246. Fronts, east, 68 : west, 91. 
 Galleries, in transepts, 16. Gothic, melange 
 
INDEX OK PLACES. 
 
 // 
 
 of styles, 45, 46, 49 ; veneer over Norman 
 work, 46, 49. Masonry, wide and narrow 
 joints, 23. Nave, 184: area and size, 15 
 and note. Piers, "Southern," font, 246: 
 piers in south transept, 237 and note. 
 Pinnacles, 364. Proportion, 54, 61. 
 Reredos, 181. Retrochoir, 383. Roof, Nor- 
 man, 572 : tie beams, 561 : over- vault, 
 550, note. Sacristies, 198. Stained glass, 
 497. String-course, absence of, in tran- 
 septs, 404. Towers, central lantern, 593 ; 
 collapse of, 594; vault under, 593: western 
 formerly, 598. Transepts, aisles, east and 
 west, 16, 197, 198; return or end, with gal- 
 leries over, 16, 198,542. Triforiiim, 88 : 
 absorbed into clerestory, 536 : use of, 541, 
 note, 542. Vault, aisles, segmental dia- 
 gonals, 311: choir, wooden, 571: crypt, 
 groined, 292 : lantern tower, 593 : nave, 
 bosses, and keystones, 305: lierne, 341 : 
 roof over, 550, note: shafts, 240, 540: 
 spring, height of, 304, 307 : web, thickness 
 of, 304 : presbytery, ornamentation, 95 : 
 transepts, keystone, 305. Windows, Per- 
 pendicular, stained glass, 497: triforium, 
 532. Bishop's Palace. Mortar, 24. 
 Castle Hall. Windows, 465: plate 
 tracery, 468: transom, 503. College 
 Chapel, Round-headed trefoil arch, 26S. 
 St Cross. Choir, aisled parallelogram, 
 176, 179. Flying buttresses, none of 
 twelfth century, 372. Griffes or spurs, 
 455. Roof and eastern gable, 104. Vaults, 
 advanced work, Romanesque high vaults, 
 20, 103: arch, pointed, 321, note: dia- 
 gonals, ornamented, 300, 301, 322 : 
 domical vaulting, aisles, 310. 
 
 WINDSOR CASTLE. St George's 
 Chapel. Buttresses, 351. Clerestory, 
 543 : no passage, 545. Doonvay, western, 
 205. Fixing buttresses, with pierced 
 spandrels, 377. Front, west, 91, 205. 
 Parapet, battlement, and pinnacle, 541. 
 Piers, vault, 540. /'lunacies, 366, 541 : 
 western, 362. Roofs, nearly flat, 541. 
 Triforium, 88 : wall panelled, 535. 
 Vaulting, fiat, 351, 541: lierne, 341: 
 ridge rib, triple longitudinal, 336, 340: 
 panel vault in central part, without ribs, 
 348, note: web, thickness of, 304. Win- 
 dows, perpendicular, 494, 504 : with geo- 
 metrical tracery, 397 : west window, 205. 
 
 WING. Apse, polygonal, 220: with Anglo- 
 Saxon buttresses, 351. Crypt, 191: 
 groined vault, 292. Nave, aisled, 223, 
 with Anglo-Saxon pier arcade, 234. 
 Windows, Anglo-Saxon, 459. 
 
 WINGHAM. Parallel naves, 382, note. 
 Parapetted spire, 622. 
 
 WINNAL HAMPDEN. Soffit cusps, 509. 
 
 WINNAL MAGDALEN. Cttsptd lancet, 
 
 5°9- 
 
 WISBECH. Aisles, 224. Tower, strings 
 
 and dripstones of windows, 606. 
 
 WISBY. ISLE OF GOTLAND. Pointed 
 
 arch, 263. 
 
 WITHAM. " Western school' 1 of Gothic, 
 105. 
 
 WITNEY. Norman porch, 582. Spire, 
 62 t. Tower and spire, position of pin- 
 nacles, 623. 
 
 WITTERING. Spire, entasis, 632. 
 
 WOLLASTON. Spire, 621. 
 
 WOODBASTWICK. Hybrid roof with 
 cambered tie beam, 561, 562, 569. 
 
 \Y< )i )DI ( )kl). Irch, Gothic mi,, i, ,,, ular, 
 261. Gargoyle, lead spout, 399. 
 
 WORCESTER. Cathedral. Aisles, walls 
 raised for light, 530. Arch, cusped, 513. 
 Pases, crypt, Norman, 451, 452: slype, 
 " pudding-base," Anglo-Saxon, 451. Capi- 
 tals, nave, western bays, with incurved 
 cones, 414 and note; earliest Gothic 
 foliage, 106: "West of England" capital, 
 433: slype, "pudding-capitals," Anglo- 
 Saxon, 451. Chapels, Prince Arthur's, 
 208 : transeptal, elongated eastern chapels, 
 167. Crypt, 191 ; apse and ambulatory, 
 no radiating chapels, 167 : Golgotha or 
 ossuary, 194. Elevation, interior, 530. 
 Flying buttresses, 372 and note. Fronts, 
 east transepts, 68. Gothic, melange of 
 styles, 46 : nave, western bays of, earliest 
 Gothic, and of the "Western school," 105, 
 106,433. Mortar, 24. Piers, " Southern," 
 choir and retrochoir, 249 and note, 255 : 
 " Western," 245 : tower, central, 594. Poly- 
 chromy, 250, note. Sanctuary, unaisled, 1 76. 
 Spire, formerly, 619. Tower, detached, 
 destroyed, 598: two small, blown down, 
 and fall of another, 594, note. Transepts, 
 eastern, 189. Triforium, internal elevation, 
 530: blank double arcade, 535. Vaulting, 
 chapter house, central pillar, 339 : choir, 
 height of spring of, 307 : Lady chapel, 
 material, 303, note : nave, earliest Gothic 
 vaulting, 105, 108: aisles, 300; ribs, 
 diagonal, 300 : shafts, 23S, 240, wall ribs, 
 238. Windows, aisles, 530 ; rear arch 
 
77- 
 
 INDEX OF PLACES. 
 
 cusped, 513; double, 513: lancets, plain 
 outside, glorious within, 464. Holy 
 Trinity. Roof, hammerbeam, from 
 Guest Hall, 564 : longitudinal braces, 569. 
 WORMS. Piers, alternation, 243. Vaulting, 
 317 : diagonal ribs of high vault, 7. 
 
 WORSTEAD. Flying buttresses, 368. Pier 
 arcade, no hood-mold, 407. Roof, and 
 clerestory, 501 : eaves, lead-covered, 292 : 
 hammerbeam, 565. Sacristy, 225. Tower, 
 pinnacles, altered in restoration, 608, note : 
 "Norfolk sound holes," 517. 
 
 WORTH. Apse, Anglo-Saxon, 156; roof, 
 219. Transepts, 222. Windows, Anglo- 
 Saxon, 459. 
 
 WREXHAM. Arch, chancel, 227 : pier 
 arcade no hood-mold, 407. Roof, arched 
 braces, 562. Tower, disposition of orna- 
 ment, 608 ; panelled all over, 345, 608 : 
 pinnacles, ogee, cupolas, 362, 364. 
 
 WRINGTON. Shaft bands, 249. 
 
 WYCOMBE. Hospital of St John. In- 
 terlacing foliage, 433. 
 
 WYMINGTON. Date, brass of founder, 
 138. Roof, low pitch of, 559. 
 
 WYMONDHAM. Roof, hammerbeam, 
 566. Tmvers, western and central, 599 : 
 west tower, strings and stages, 405 : with- 
 out ornament, 608 : sky-line, horizontal, 
 606 : turreted angle buttresses, 605. Tri- 
 forium, single arch arcade, 532. 
 
 YARMOUTH. Arches, chamfered, 277; 
 no hood-mold, 407. Crocket capitals, 429. 
 Interior elevation, 68, no clerestory, 547. 
 Transept windows, shuttered, 459, note. 
 
 YATTON. Shaft bauds, 249. West front, 
 91. 
 
 YAXLEY. Double molded capitals, 443. 
 
 YORK. Minster. 11. Battlements, per- 
 forated, 399. Buttresses, pinnacled, 
 chapter house, 363 : eastern transept, 364 : 
 nave, 366. 372. Capitals, with large leaf 
 foliage, 437. Chapterhouse, 126. Choir, 
 internal elevation, 88: plan, "Salisbury" 
 formerly, with rectangular ambulatory, 171, 
 175, note, now an aisled parallelogram, 
 former choir towers, now eastern transepts, 
 599. Clerestory, choir, 543 ; passage, 545, 
 note: presbytery, passage, 545: south 
 transept, walls, 25. Crypt, 191. Door, 
 
 original deal, chapter house, 585. Door- 
 ways, south transept, 584 : west, straight 
 sided gable, 579. Flying buttresses, nave, 
 366, 372 : transepts, concealed, 374. 
 Foliage, crypt, fern leaf, 418 : nave, large 
 leaved, 437. Fronts, east, 91 : transepts, 
 68, 71: west, 579. Gothic, growth, 49: 
 melange of styles, 46. Griffes as spurs, 
 crypt, 455. Masonry, bad, 25. Nave, 67 : 
 internal elevation, 88. Parapets, chapter 
 house, figures on, 396 : choir, perforated, 
 399: " oversailing '•' western towers, 605. 
 Pinnacles, chapter house, 363 : choir, 364 : 
 nave, 363, 364, 366, 372 : transepts, 
 eastern, "spearhead," 364. Plan, see 
 Choir. Pulpit um, 179. Roof, original 
 lead, 389. Screens and divisions, 179, 
 i8t, 179. Stained glass, east window, 494, 
 note: "Five Sisters," 463. Toivers, central, 
 interior effect, 603; lantern, 593; Nor- 
 man, 593, note ; pinnacles absent, top un- 
 finished, 606, note ; settlement, 594 ; 
 western towers, 605. Transepts, aisles, 
 double, 198: eastern, 189, 599. Triforiitm, 
 choir, absorbed into clerestory, 536 : nave, 
 internal elevation, 67, 88 : transept, single 
 containing arch, 533. Vault, chapter 
 house, no central pillar, 339 : choir and 
 nave, shaft, 240: wooden vault, 570: 
 tower, central hole, 339 : transept vault, 
 wood, springers, stone, 118. Windows, 
 clerestory, 67 : curvilinear, 479, 488, 496 
 east, 496 : nave, 499 : presbytery, double 
 windows, 513: west, 132, 306: glazing, 
 Anglo-Saxon windows, 459 ; east, 496 ; 
 " Five Sisters," 463. Holy Trinity, 
 Micklegate. Gothic, "Northern school" 
 of, 108. St Martin. Stained glass, 493, 
 note. St Mary's Abbey. Apses, three 
 parallel eastern, formerly, 164. Arch mold- 
 ings, chapter house, "Southern" pier, 246. 
 Foundations, 26 and note. Strings, fra- 
 gile undercut, 405 : hollow, 406. St 
 Mary Bishop Hill Junior. Belfry 
 window, 456. St Maurice. Plate 
 tracery, 468. St Michael. Chancel arch, 
 absent, 227. Pier, 255. 
 
 YOULGRAVE. Chamfered abacus, 37. 
 
 YPRES. Cloth Hall. Front, 71. 
 
 YTHANCHESTER. Buttresses, Anglo- 
 Saxon, 351. Western porch, 202. 
 
 ZAMORA. Hall churches, 321, note. 
 ZARAGOSSA. Lierne vaulting, 342. 
 
INDEX OF SUBJECT MATTER & GLOSSARY. 
 
 (DEFINITION'S OF ARCHITECTURAL TERMS WILL HE FOUND ON THE PAGES WHICH 
 ARE REFERRED TO IN CLARENDON TYPE.) 
 
 Abacus, 439, 440, 441, 37 
 
 ABUTMENT— See " Buttress " ; Romanesque 
 systems, 29 ; Gothic systems, 54 ; Roman 
 system, 291 ; and 103 
 
 Acanthus, 426, 427 
 
 Advanced Design — See " Progressive " 
 
 Aisles, single, 224; double, 156; several, 
 224 ; early parochial, 224 ; of choir, 
 162 ; of transepts, 16, 197 ; omission of, 
 201, 220 ; how obtained, 211, 223 ; aisled 
 chapel plan, 226 ; why tall, 530 ; super- 
 posed, 527, 52S-530 ; methods of roofing, 
 224, 235, 371, 53S, 570 ; vaulting of and 
 results, 292, 29, 381 
 
 ALTARS, 180, 197, 198, 207, 20S ; high altar, 
 position of, 592 
 
 Altitude, Gothic, 61 
 
 Ambo, 179 
 
 AMBULATORV, semicircular, 164, 172 ; rect- 
 angular, 171 
 
 Angels, 95, 568 
 
 Angevin Gothic, 382 
 
 Anglo-Norman — See " Romanesque " 
 
 Anglo-Saxon, rebuilding, 14; Wilfrid's 
 work, 155 ; presbytery, 156, 157 ; crypt, 
 191 ; transepts, 196 ; west transept, 203 ; 
 complex plan, 218 ; timber churches, 
 218 ; plans of parish churches, 220-223 ; 
 compound piers, 237 ; vaulting, 284 ; 
 buttresses, 351, 358, 362; capitals, 
 cubical, 411 ; Corinthian, 427 ; balusters 
 and lathes, 445 ; windows, 456, 465 ; 
 clerestories, 545 ; towers, 590, 595 ; 
 newel staircase, 601 ; spires, 617 ; west 
 porch, 202 ; survivals in Norman work, 
 42, 596, 599, 600 
 
 Animal Symbolism, 42 
 
 Annulets, 249 
 
 Anthemion, 42, 425 
 
 Apse, 104, 145, 156, 157; parochial, 212; 
 of transept, 197 ; three parallel eastern 
 apses, 162, 163, 164 ; arches, 261 ; West- 
 minster, 121 ; vaulting, 283 
 
 Arcade, 213 
 
 Arch, forms of, 259 ; Romanesque, 36 ; 
 orders, 237, 238 ; how strengthened, 
 272 ; square-edged, 36, 276 ; chamfered, 
 277, 27S, 279 ; ornamentation, 276 ; arches 
 of apse, 261 ; relation to jamb, 277 
 
 Arch, Discharging, 573 ; relieving, 573 
 
 Arch Moldings, 276, 278 
 
 Archaic not early, 43 
 
 Arch-braced Roofs, 563, 564 
 
 Arched Brace, 562, 563 
 
 Architecture, definition of, 1 
 
 Arcisolium, 190 
 
 Arcuated Construction, 257 
 
 Area of Churches, 15, 50 
 
 Arris, 292, 625 
 
 Ashlar, 20, 24, 204 
 
 Assimilation, 49, 80, 12S, 464 
 
 Astragal, 434, 442 
 
 Atrium, 201 
 
 Attic Base, 451 
 
 AUGUSTINIAN (or Austin) CANONS' 
 
 Churches, 15, 44 
 AuVERGNE barrel vaults, 550 
 
 B 
 
 Backward Design — See "Retrogressive" 
 
 Bahut, 395 
 
 Ball-flower, 83, 478, 625 
 
 Baluster Shafts, 445 
 
 Bands, 240, 249 
 
 Bar Tracery, early examples, 469-471 
 
 Barrel Roofs, 569 
 
 Barrel Vaults, 281, 283, 550, 13 ; an- 
 tiquity of, 281 ; corbelled, 263 ; pointed, 
 264 
 
774 
 
 INDEX OF SUBJECT MATTER 
 
 Base, 447-454 ; Attic, 451 ; Romanesque, 
 37 ; Transitional, 451 ; waterholding, 449, 
 
 451 ; double, 451 ; double or triple roll, 
 
 452 ; oversailing, 450 ; how executed, 
 
 449 
 Basement Course, 402-404 — See "Ground 
 
 Course" 
 Basilica, Secular, 201 
 Basilican Architecture, 5, 6, 145, 146, 
 
 156, 159, 183, 222, 230, 231, 232 
 Battlement, 391, 398, 541 
 Bay Proportion, 54, 529 — See "Elevation 
 
 of Interiors" 
 Bead, 441 
 
 Bead and Roll, 41 
 Beak-head, 41, 428 
 Belated Curvilinear Tracery, 501 
 Belfry Windows, 517 
 Bells, early large, 586 
 Benches, 447 
 Benedic Biscop, 155 
 Benedictine Churches, 15, 44 
 Bestiaires, 42 
 Billet, 40 
 
 Bishop's Throne, 155, 162 
 Black Death, 128, 134, 499 
 Blindstory, 532 
 Bone House, 194 
 Boss, 257, 305 
 
 Brace, arched, 562 ; longitudinal, 569 
 Bracket, 256 
 Bretture, 23 
 Broach Spire, 620 
 
 Building Construction, medieval, 275 
 Building, Speed of, 46, 49 
 Bulbous Foliage, 436 
 Buttress, 350-362, 55, 103, 10S ; columnar, 
 
 351; clerestory, 374, ^77; diagonal, 
 
 361 ; pilaster, 351, 352; Romanesque, 
 
 30; set-offs, 361 ; towers, 605, 606; 
 
 vertical, 367 
 Buttressing, the essential element in 
 
 Gothic construction, 7- 11 
 Burgundian, Romanesque, 6, 16 
 Byzantine architecture, 5, 309, 409, 411 
 
 Cable, 41 
 
 Cambered Tie-beam, 562 
 
 Campaniles in Rome and Ravenna, 586 
 
 Canons' Churches, 201 
 
 Canons Regular, 44 
 
 CANOPIES, fourteenth century, 87 
 
 Capitals, function, 409 ; Romanesque, 37 ; 
 cushion or cubical, 409 ; scalloped, 413 ; 
 coniferous, 413 ; incurved cones, 414 ; 
 interlacing, 414 ; figure, 416 ; storied, 
 416 ; Transitional volute, 416 ; water 
 leaf, 416 ; stalked foliage, 431 ; pipy 
 stalks, 431 ; double, 443 ; naturalistic 
 foliage, 434, 435, 436, 438 ; molded caps, 
 442-446, 434, 438, 441 ; octagonal, 444 ; 
 painted, 411 ; recarved, 413; omitted, 
 409 
 
 Capstone, 636 
 
 Carnarvon or shouldered arch, 260 
 
 Carthusian churches, 15 
 
 Casement, 255 
 
 Catacombs, 190 
 
 Cat-heads, 41 
 
 Cathedrals, growth and modifications, 46 
 
 Ceilings, 570 ; Romanesque, 290, 312 ; 
 stone, 285 
 
 CELTIC influence, 212 ; ornament, 42 
 
 Centering, economy of, 275 
 
 Central Tower, 590-594, 195 ; objection- 
 able, 593 ; removal, 590 ; survival, 599 ; 
 omitted, 594 ; constructional value, 381, 
 590 ; collapse, 594 ; diameter of piers, 
 594 ; lantern closed, 593 ; Norman, 590, 
 
 CERCE, 296, 29S 
 
 Chain Ornament, 41 
 
 Chamfer, 278 
 
 Chamfer Cusps, 509 
 
 Chancel, 183, 185, 220 ; rectangular, 212 ; 
 lengthened, 225 ; half aisled, 225, 226 ; 
 aisled, 226 
 
 Chancel Arch, survival, 46 ; omitted, 226, 
 227 ; windows above, 547 
 
 Chantry Chapels, 136, 156, 197, 208, 211, 
 223, 224 
 
 Chapel, 208, 219, 591 ; eastern, 157; tan- 
 gential or radiating, 164 ; of central 
 transept, 16, 197, 222 ; of western tran- 
 sept, 224 ; between buttresses, 197, 
 21 1 
 
 Chapter House, construction, 59 ; vaulting, 
 
 339 
 
 Charnel House, 194 
 
 Chestnut, 571 
 
 Chevron or zigzag, 40 
 
 Choir, 183, 184, 185; definition, 185; apsidal, 
 145 ; rectangular, 145, 157 ; unaisled, 
 16; lengthened, 46, 49 ; number of bays, 
 16; great length, 49 ; rebuilt, 46, 49 : 
 raised, 191 ; beneath central tower, 221 ; 
 flanking towers, 599. 
 
 Choir Screen, constructional value, 381 
 
AND GLOSSARY 
 
 // 5 
 
 Christian Architecture, 5— See " Ba- 
 silica " 
 
 ClNQUEFOILED arch, 268 ; lights, 503 
 
 Circular Windows, 108, 516 
 
 Cistercian Architecture, 15, 16,30,37, 
 37, 44,67,73, 101, 103, 104, 105, 157, 167, 
 171, 176, 179, 185, 197, 202, 266,290,395, 
 418, 441, 445, 59S 
 
 Classical Mythology, 42 
 
 Claveau, 258 
 
 Claw, 454 
 
 Clerestory, 543-549, 371, 497 ; etymology, 
 543; windows, ioo, 517: development 
 of window, diminution of wall, 56, 57, 58, 
 543, 548, 123 ; tallness, 123 ; clerestories 
 added, 547 ; spire needed, 649 : clere- 
 story wall, stability, how secured, 60 ; 
 clerestory buttress, 374, 377 ; clere- 
 story objectionable, 383 : omitted, 547 ; 
 Romanesque, 29 ; wall passage, internal. 
 545 ; external, 545 ; omitted, 30, 121, 
 545 ; parochial clerestories, 19, 67, 68, 
 82, 545, 546, 547 ; wall posts, 562 
 
 Clergy, seats of, 592 
 
 Cloister, position of, 29 
 
 Cluniac Architecture— See "Burgun- 
 dian"; 15, 167, 169, 189, 193, 202, 205,550 
 
 Clustered Columns, 255 
 
 Coats of Arms, 94 
 
 Collar or collar beam, 560, 563 
 
 Collar-braced Roofs, 563 
 
 Colleges, 136 
 
 colonnadks, 230 
 
 Columns, 233, 242, 230, 231, 232 ; export of, 
 232 ; clustered, 255 
 
 Common Rafter, 563 
 Compass Roof, 570 
 Composition, architectural, 127 
 Compound Pier, 233, 234, 242 
 Compression, beam in, 561 
 Cone, double, 41 
 Concessionary, 155, 190, 191 
 Coniferous Capital, 413 
 Console, 427 
 Constantinl's Arch, 230 
 Constructed Decoration, 616 
 Coping, 393, 394 
 Corbel Table, 103, 391, 622 
 Corbelling, 263 
 
 Corinthian Capital, 420-429, 425 
 Cornice, 392, 394, 566, 622 
 Coupe-larmk, 403 
 Crenel, 398 
 
 CROCKET, 77, 83, 95 ; capitals, 429 ; spires, 
 625, 626 
 
 Cri hsillon, 195 
 
 Crossing, 195; piers, 581; arches, 262, 
 
 265 ; (Ionic, 592 ; roofing, 196, 590 
 Cruciform plan, 19, 88. 195, 222 
 Crux commissa and immissa, 195, 196, 592, 
 
 593 
 Crypt, 155, 186, 190; vaulting, 292 
 Cubical Capital, 409-413 
 Cubk ulum, 190 
 Curvilinear tracery, 479-490, 474 ; early, 
 
 484 ; belated, 501 
 Cushion Capital, 409-413 
 CUSPING, 508-511 ; compound, 511 ; arches, 
 
 268 ; doorways, 509 
 
 Cylindrical Piers, 233, 234, 242, 243, 36 
 
 D 
 
 Debased Goi hic, 491 
 
 DECORATED period of English Architecture, 
 S0-S7, 126-134, 472-490 
 
 Demi-beri lac. 284, 374 
 
 Detached Shafts, 233, 244, 246, 249 
 
 DIAGONAL RIBS, 296 ; early examples, 300, 
 440; segmental, 311; pointed, 310; semi- 
 elliptical, 310 ; omitted, 10 
 
 Diaper, 84. 94 
 
 diaphragmatic roofing, 234, 235, 317 
 
 Discharging Arch, 573 
 
 Dissolution of Monasteries, 142 
 
 DOME, 550; antiquity, 281; central, 592; 
 Southern France, 264 ; Romanesque, 
 6, 13 ; pointed, 264 ; corbelled, 264 
 
 Domical Ribbed Vaults, 283, 321, 335, 
 336 
 
 Doors, 585 
 
 Doorways, 573-585 ; Romanesque, 38, 578 ; 
 recessed orders, 276 ; trabeated, 573 : 
 form of arch, 580; trefoiled, 580 ; Gothic, 
 579 ; rear-arch, 5S5 ; west, 581 ; lateral 
 of nave, 583 ; monastic, 583 
 
 Doric Capital. 420 
 
 Dosseret, 409, 41 I 
 
 Drainage of Roofs, 3S4-401, 537 
 
 Drawings, Mediaeval, 2 
 
 Dripping Eaves, 392, 393, 394, 37, 559, 5 6 ° 
 
 Dripstone, 406-408, 405, 277, 465 
 
 Drop Arch, 259 
 
 Durham School of Romanesque, 100 
 
 Early Christian Architecture— See 
 "Basilica"; 145, 146, 159, 183, 222, 230, 
 
 Early Enolish period of English archi- 
 tecture, 65-79, 105, 126 
 
77$ 
 
 INDEX OF SUBJECT MATTER 
 
 East Anglia, isolation of, 501 
 East FRONT, design of, Romanesque, 28 ; 
 thirteenth century, 68 ; fourteenth cen- 
 tury, 82 ; fifteenth century, 91 
 
 Eastern Chapel, 63, 46 
 
 Eastern Transept, 65 
 
 Economy of Stone, Gothic, 58 
 
 Elasticity of ribbed vaults, 298, 299 
 
 Elevation of Interiors, Romanesque, 
 26, 27 ; thirteenth century, 67 ; four- 
 teenth century, So ; fifteenth century, 88 ; 
 proportions, 26; Gothic, 61, 54, 529 
 
 Elliptical — Semi, Arches, 262 
 
 Embrasure, 398 
 
 End Windows, completed later, 500 
 
 Engaged Columns, and shafts, 233, 244 
 
 English Influence on French Gothic, 
 128, 129, 131, 132, 361 
 
 English "Filling-in," 335 
 
 Entablature omitted, 233 
 
 Entasis, 632 
 
 Equilateral Arch, 259 
 
 Extrados, 259 
 
 False Bearing, 374, 277 
 
 Fan Vaults, 342-349 ; transverse arches, 
 346 
 
 Feather-edged, 402 
 
 Feretory, 183 
 
 Fern Capital, 418 
 
 Fifteenth Century work, planning, 88 ; 
 triforium, 88 ; internal elevation, 89 ; 
 east and west fronts, 91 ; vaulting, 93 ; 
 piers, 94 ; ornament, 94 
 
 Figure Capitals, 416 
 
 Fillets, 256 ; triple, 279, 280 
 
 Filling-in, methods of, 323 
 
 Finial, 364 
 
 Fir, 571 
 
 Fireproofing churches, 289 
 
 Flamboyant, 128, 129, 131, 132, 474, 484 
 485 
 
 Flemish art, 12S, 131, 493 
 
 Flint work, 399, 582, 601 
 
 Flowing and Rectilinear Tracery 
 mixed, 499 ; in juxtaposition, 500 
 
 Flowing Tracery, 474, 479-490 ; early, 
 484 ; belated, 501, 503 
 
 Flying Buttress, 368-380, 54-56, 109, 113 ; 
 internal, 30; external, 1 21 ; not essentia 
 to Gothic construction, 9 ; omitted, 372 ; 
 origin, 369 ; function, 380 ; in super- 
 position, 378 ; in two flights, 378 ; aque- 
 duct, 400 ; spires, 377, 625 
 
 Foiled Arch, 260 
 
 Foils, 508 
 
 Fore-court, 201 
 
 Formeret, 299, 300 
 
 Foundations, 25 
 
 Four-centred Arch, 259, 266, 267, 268 
 
 Four-leaved Flower, 83 
 
 Four-story Interiors, 527 
 
 Fourteenth Century work, 80-87, 126- 
 134 ; planning, 80 ; internal elevation, 
 80, 82 ; east and west fronts, 82 ; piers, 
 83 ; bases, 452 ; molded capitals, 443 ; 
 foliage and foliated capitals, 436, 83 ; 
 canopies, 87 ; vaulting, 82 ; buttresses, 
 360; doorwavs, 579 ; doors, 5S5 ; spires, 
 617 
 
 French Flamboyant, 270 
 
 French Gothic, 43, 54-62, 55, 57, 196, 198, 
 205, 300, 372, 374, 375, 376, 378, 382, 389, 
 396, 399, 408, 462, 463, 473, 485, 527, 5 2 8, 
 529, 53i, 535, 537, 53S, 55°, 594 
 
 French Influence on English Gothic, 
 108, 109, 112,(113, I2I i 617, 621 
 
 French Method of " Filling-in," 335 
 
 Fret, 41 
 
 G 
 
 Gable, lowering of, 541 ; gabled aisles, 570 
 
 Galilee, 202 
 
 Gallery, basilican, 233 
 
 Gargoyle, 399 
 
 Gaul, influence of, 218, 590 
 
 Geometrical Tracery, 472 - 47S ; why 
 disused, 485 
 
 German Romanesque, 6, 13, 196, 204, 317 
 
 Gilbertine churches, 15 
 
 Glass — See " Stained Glass " 
 
 Glazing, 456, 459, 464, 514 
 
 Golgotha, 194 
 
 GOTHIC, definition, 7-1 1, 509; relation to 
 Romanesque, 7, 12, 9, 10, 11, 43, 55, 56, 
 57, 58 ; earliest, 43 ; large admixture of 
 Romanesque, 45 ; systems of abutment, 
 54 ; masonry, 24, 25, 60 ; preponderance 
 of voids, 56 ; aerial, 60 ; loftiness, 61 ; 
 vertical line, 61, 62 
 
 Griffe, 418, 446, 455 
 
 Groined Vault, 290-295, 281 ; definition, 
 76, 292 ; construction, 294 ; high, 292 
 
 Ground course, 402-404, 103 
 
 Gutter, 393, 394, 395, 103 
 
 H 
 
 Half Barrel, 284, 374, 13 
 Hall Churches, 228, 321 
 
AND GLOSSARY. 
 
 777 
 
 Hammerbeaw Roofs, 564 seq. ; double, 566 ; 
 alternating with tie-beam, 566 
 
 Hatching, 23 
 
 Height of Churches, 50 ; how deter- 
 mined, 53 ; ratio to span, 53 
 
 Heraldic emblems, 94 
 
 Heterogeneous Design, 45, 46, 49 
 
 High Ribbed Vaults, 312 ; omitted, 290, 
 
 312 ; pier of, 539 
 
 Hinges, 585 
 Honeysuckle, 425, 42 
 Hood-mold, 406-40S, 108, 277 ; straight- 
 sided, 86 
 Horizontal Line in Gothic, 6: 
 
 Horse-shoe arches, 259. 262 
 
 397 
 
 Imbrications, 41 
 
 Interdict, 116 
 
 Interlacings, 41 : capital, 414 
 
 Intermediate rib, 336 
 
 Intermediate Transverse Arch 20 
 
 Incurved-cone capital, 415 
 
 Intrados, 259 
 
 Ionic Capital, 420 
 
 Irish architecture, 212, 213 ; round towers, 
 
 586 
 Italian Gothic, system of abutment, 374 
 
 Jacobean work, 203 
 Jesus Altar, 180 
 JOINTS, Romanesque, 23 
 
 K 
 
 Keel Molding, 256 
 Kentish Tracery, 476 
 KEY ornament, 41 
 Keystone, 257, 304 
 King-post Roof, 561 
 
 Label, 406, 580 
 
 Labour, Cheapness of, 59 
 
 Lady Chapel, 46, 63, 172, 1S1, 202 ; lofty, 
 
 176 
 Laic Gothic, 182 
 Laxc et Arch, 259 
 
 1, 445. 
 and 
 
 532, 
 
 260 ; 
 317, 321,410 
 
 LANI I.I WlNDI W.S, 460 
 
 Lanfranc, influence of, 317 
 
 Lantern type of church, 133, 134, 141, 228 
 
 49' 
 Larmier, 394 
 
 Lathe, use of, 246, 249, 250, 252, 41 
 
 44 r, 
 Latin Style, 5 —See "Basilica" 
 
 " Early Christian " 
 Laurel Capital, 428 
 Le Puy, Romanesque of, 13 
 Leak roofs, 3S9, 541, 559 
 Leaf Scrolls, 42 ; capitals, 429-432 
 Lean-to roofs, 570 
 Length of churches, 15, 49 
 Library, i 59 
 
 Lierne Vaulting, 82, 93, 138,340 
 Lighting, systems of, 26, 53, 60, 61 
 
 222, 228, 371, 3S2, 529i 53o, 53' 
 
 53 r3 > 537, 538- 544, 547, 593 
 Lime-kiln, 116 
 Lintels, 259 ; coursed lintels, 577, 
 
 doorway, 573 
 Lombardic Romanesque, 7, 1 
 Louvre Boards, 517, 605 
 Louvre over crossing, 592 
 Low Side Window, 517 
 Lozenge, 41 
 
 M 
 
 Magister Operis, 2 
 
 Marble, 109, 230, 249, 252 
 
 MASONCRAFT as chronological evidence, 23 
 
 Masonry, 103: Romanesque, 20-25 ; Gothic, 
 60 
 
 Master Masons, 2 
 
 Medallion, 41 
 
 Merlon or Merlan, 398 
 
 Miracles of healing, 1S7, 134 
 
 Mitred Ribs, 305 
 
 Molded Capital, 442-446, 434, 43 s . 44' ; 
 origin, 445 
 
 Moldings, influenced by Purbeck marble, 
 278 
 
 Moldings, arch, 276, 278 ; ribs of vault, 300 ; 
 buttress, 358 ; ground-course, 402 ; string- 
 course, 405 : abacus, 440 ; capital, 442 ; 
 base, 450 ; mullions, 516 
 
 Monasteries, dissolution of, 138 
 
 Monastic Architecture, 44 
 
 Monastic Churches, distribution of, 15 
 
 Monials, 514, 505, 508 
 
 Monolithic Columns and shafts, 230, 246, 
 4" 
 
773 
 
 INDEX OF SUBJECT MATTER 
 
 Mortar, 24 
 
 Mosaic Pavements, Roman, 414 
 
 Mullions, 505, 50S, 514 
 
 N 
 
 Nail-head, 40 
 
 Narthex, 202 
 
 Naturalistic foliage, 434-436, 438 
 
 Nave, 201 ; number of bays, 16, 184 ; why 
 long", 202, 207 ; rebuilding of, 46, 49 ; 
 parallel naves, 224, 382 : without aisles, 
 220 
 
 Nebule, 41, 392 
 
 Necking, 434, 442 
 
 Newel staircase and turret, 601 
 
 Niches, 86 
 
 Nine Altars, Chapel of, 189 
 
 Normandy, Romanesque of — See "Ro- 
 manesque" ; 6, 7, 317, 596 
 
 Normans in Sicily, 218 
 
 Northern School of Gothic, 100-108, 
 253 
 
 O 
 
 Oak, 571 
 
 Obits, 20S 
 
 Obtuse Arch, 259 
 
 Oculus, 516 
 
 Ogee Arch, 260, 268, 482, 271 ; foliage, 436 ; 
 
 in France, 270 
 One-story Interior, 91, 407, 539 
 Opposing Thrusts, 381-383 
 Orders of the Arch, 237, 238, 272, 274, 
 
 37 
 Orders of Window Tracery, 505, 506, 
 
 507 
 Organs, 198 
 Orientation, 155, 221 
 Ornament, Romanesque, 39 ; thirteenth 
 
 century, JJ ; fourteenth century, 83 ; 
 
 fifteenth century, 94 ; on towers, 608 ; 
 
 on spires, 632 
 Osier Churches, 219 
 Ossuary, 194 
 
 PAGAN emblems, 42 
 Painted Capitals, 207, 411, 425 
 Palm Sunday, 207, 581 
 Palmette, 42, 425 
 Panel Vaulting, 337 
 Parapet, 393-398, 391, 541 ; Norman, 394 ; 
 internal, 397 ; spire, 621 
 
 Parclose Screens, 181 
 
 Parish Church, a growth, 46 ; planning, 
 212-229; Romanesque planning, 19; 
 apse, 212 ; pier arcade, 277 ; vaulting, 
 550 ; thirteenth century, 67 ; spacious- 
 ness of late churches, 67, 122, 132, 133, 
 1 38 ; design, 499 
 
 PARVISE, 205, 582 
 
 Passion, symbols of, 95 
 
 Patera, 41 
 
 Pavillon, 537 
 
 Peacock, 183 
 
 Pearl Ornament, 41 
 
 Pellet, 41 
 
 Pendant, 257, 305, 345 
 
 Pendant Post, 563 
 
 Pendentive, 265 
 
 Pentecostals, 207 
 
 Periapsidal Plan, 164 
 
 Perigueux, domes, 13, 550 
 
 Perpendicular period of English archi- 
 tecture, 88-96, 133-142 ; beginning, 134 ; 
 duration, SS 
 
 Perpendicular Window Tracery, 491 
 
 Pews, 447 
 
 PlER, 230-256 ; definition, 233 ; compound, 
 2 33i 2 34 > ' ate Gothic compound, 242, 
 454 ; Romanesque, 36 ; thirteenth cen- 
 tury, 76 ; fourteenth century, 83 ; fifteenth 
 century, 94 ; cylindrical, 233, 234, 242, 
 243 ; logical, 236, 237, 238, 239, 242, 255 ; 
 illogical, 241 ; intermediate, 121 ; alter- 
 nation with column, 235, 243, 317 ; dia- 
 meter and circumference, 24, 594 ; 
 masonry, 24 ; thrust inward, 29 ; Gothic, 
 244-256, 242 ; southern, with detached 
 shafts, 246 ; western, triplets of shafts, 
 245 ; northern, cluster of columns, 253 ; 
 pier of high vault, 57 ; pier of central 
 tower, 594 
 
 Pier Arcade, definition, 233 ; upward 
 growth, 539 ; tall, 530 ; low, 530; longi- 
 tudinal thrusts, 381 
 
 Pilaster, 235, 239 
 
 Pilgrims, 170, 193, 205, 499 
 
 Pillar — See " Pier " 
 
 Pinnacle, 362-367, 377, 391, 541 ; towers, 
 606, 608 ; spires, 623 ; omitted, 367 
 
 Planning, basilican, 156 ; Norman greater 
 churches, 159, 163-170; parish churches, 
 212 ; thirteenth centurj', 65 ; fourteenth 
 century, 80 ; fifteenth century, 88 ; com- 
 plexity, 207 
 
 Plantagenet Gothic, 382 
 
 Plantain Capital, 418, 426, 428 
 
AND GLOSSARY 
 
 77'J 
 
 Plate Tracery, 464-469 ; origin, 465 
 
 Plinth, 450— See " Base " 
 
 Ploughshare vaulting, 299, 311 
 
 Pointed Arch, 262-266, 259, 260 ; antiquity, 
 263 ; early examples, 101, 105, 266 ; 
 origin, 262 ; not essential to Gothic con- 
 struction, 7, 9 ; value, 264, 265 ; rein- 
 vented, 263 ; segmental, 267, 268 ; 
 vaulting, 321 ; diagonals, 310 
 
 Pointed Barrel Vaults, 263 
 
 Polychromy, 250, 252 
 
 PORCH, 5S2-5S4, 205, 206, 596 ; lateral, 205 ; 
 western, 202, 204, 205 : uses, 205 
 
 Porte-a-faux, 374, 377 
 
 Portcullis, 94 
 
 Pre-Conquest— See " Anglo-Saxon " 
 
 Premature Rectilinear Tracery, ;oo 
 
 Premonstratensian, 15, 1S9 
 
 Presbytery, definition, 185 ; apsidal, 104 ; 
 rectangular, 104 ; aisled, 176 : un- 
 aisled, 175, 176 ; parochial, 226 : vaulted, 
 221 ; lofty, 104 ; upper, 226 
 
 Priest's Doorway, 582 
 
 Priest's Room, 221 
 
 Principal rafter or principal, 563 
 
 Principal, stone, 234, 235 
 
 Processional Doorway and cross, 581 
 
 Processions, 170, 207 
 
 Progressive Work, 126, 133, 198,440, 500 
 
 Proportions, 50, 53, 54, 61, 529, 530 
 
 Provence, Romanesque of, 6, 13 
 
 Pulpitum, 179 
 
 Purbeck Marble, 250, 252, 278, 279 
 
 Purlin, 563, 371 
 
 Quarries, 20, 23 
 
 Quasi-sexpartite Vaulting, 321 
 Queen-post Roofs, 561 
 quinquepartite vaulting, 339 
 Quire, 183 
 Quirk, 441 
 Quoin, 601 
 
 R 
 
 Rafters, Common, 559, 563 
 Rafters, Principal, 563 
 Rain, how disposed of, 388-401 
 Rain, protection of walls from, 402 
 Rear-arch of Window, 456, 512 
 RECASING instead of rebuilding, 46, 49, 134, 
 
 136, 138, 141 
 Rectangular Ambulatory, 171 
 
 Kl'.C I ANi .1 ILAK ( II \\( EL, 21 2 
 
 Rectilinear Tracery, 491 ; mixed with 
 flowing, 499 
 
 Relics, 198 
 
 Relieving Arch, 573 
 
 Religious Motive in design, 182, 207, 488 
 
 Remodelling instead of rebuilding, 46, 49, 
 134, 136, 133, Mi 
 
 Renaissance mixed with Gothic work, 142 
 
 Reredos, i Si 
 
 Retable, i Si 
 
 Reticence, 360 
 
 Reticulated Tracery, 484, 489 
 
 Reticulated Vaulting, 342 
 
 Retrogressive Work, 42, 43, 99, 101, 
 103, 113, 122, 133, 13S, 142, 200, 261, 
 265, 279, 299, 464, 497, 501, 570. See 
 "Survivals" and "Assimilation" 
 
 Revelation, Book of, 155 
 
 Ribbed Vaults, 296-307 ; rubble, 299, 304 
 
 Ribs of Vaults, early diagonal, 300 ; stilted 
 transverse, 311 ; wall ribs, 300; breadth 
 of, 299 ; ornament, 301 ; mitring, 305 ; 
 molds, 300 ; advantages, 296-299 
 
 Ridge-piece, 561 
 
 Ridge-ribs, 335-339 
 
 RlNCEAU, 42 
 
 Roman Construction, Basilica of Maxen- 
 tius anticipates Durham, 230 ; groined 
 vaults, 291 ; roofing of vaults, 388 ; abut- 
 ment, 230; buttress, 351 ; internal but- 
 tresses, 362 ; masonry, 20 ; compound 
 pier, 236 ; recessed orders of arch, 272- 
 274 ; pointed arch, 263 ; discharging 
 arch, 576 ; gutter and parapet, 395 ; use 
 of lathe, 411 ; towers, 5S6 ; Corinthian 
 and Composite capitals, 420, 426 ; mosaic 
 pavements, 414 ; interlacings, 41 ; chev- 
 ron, 40 
 
 Romanesque Architecture, definition 
 and date of commencement, 6 ; com- 
 mencement in England, 14 : duration, 
 42 ; schools of Romanesque, 6, 12, 13, 
 16S ; school of West of England, 100 ; 
 school of North of England, 100 ; 
 list of greater churches surviving, 1 5 ; 
 number, scale and dimensions of greater 
 churches, 14, 15, 49, 50, 53 ; length and 
 number of bays in nave, 49 ; span, 1 5, 
 50, 53 ; area, 15 ; relation to Gothic, 7, 
 12, 43; Romanesque ?'. Gothic con- 
 struction, 9, 10, 11, 55, 56, 57, 58; 
 masonryand walling, 20-24, 593 ; founda- 
 tions, 25 ; monumental and gloomy, 26, 
 60 ; internal and external elevations, 26, 
 
780 
 
 INDEX OF SUBJECT MATTER 
 
 27 ; east and west fronts, 28 ; piers, 36 ; 
 
 arches, 36, 260, 261, 262, 273 ; vaulting, 
 
 19. 312, 3H; domes, 13, 282, 590; 
 
 abutment and buttresses, 29, 30, 351 ; 
 
 ceilings, 312 ; dripping eaves, gutter, 
 
 parapet, 37, 392, 394 ; ground-course, 
 
 37, 402 ; strings, 38, 404 ; piers, 233, 
 
 242 ; base, 37, 447 ; abacus, 37, 440 ; 
 
 capitals, 37, 409, 420 ; windows, 38, 456 ; 
 
 clerestory, 30, 545 ; doorways, 38, 577 ; 
 
 towers, 38, 586 ; spires, 617 ; ornament, 
 
 39 ; roofs, 572 
 Rome, Basilicas in, 146 
 ROOD, 180; rood beam, 180; rood screen, 
 
 180 
 Roof Drainage, 384-401, 37, 103 
 Roofs, Open Timber, 558 seg. 
 Roofs, trusses, 559 ; trussed rafter, 559 ; 
 
 tie-beam, sharp pitch, 561 ; depressed, 
 
 562 ; hammerbeam, 564 ; arch-braced, 
 
 563 ; boarded, panelled, 560, 570; barrel, 
 569; pitch of, 389, 391, 558 ; saddle or 
 span, 225 ; aisle roofs, 224, 235, 570 ; 
 triforium, 541 ; over vaults, 388, 389, 
 550 ; over crossing, 196, 590 ; of stone, 
 284; vast spans, 564, 569; early ex- 
 amples, 572 ; colour, 569 ; material, 571 ; 
 omission of, 265 
 
 ROSES, 42, 94 ; rose window, 517 
 Round Towers, 586, 600 
 Royal Chapels, 228, 229 
 Rubele, 20, 24, 295, 303, 593, 594 
 
 SACRARIUM— See " Presbytery'' 
 
 Sacristan, 584 ; sacristy, 159, 198, 225, 226 
 
 Saddle Roof, 570 
 
 Saint's Chapel, 181, 183, 185 
 
 Sanctuary, 185— See "Presbytery" 
 
 Saracenic work, 263 
 
 Saw-tooth, 40 
 
 Scale Work, 41 
 
 Scalloped Capital, 413 
 
 Scandinavian Romanesque, 410, 411 
 
 Schola, 145 
 
 Scoinson Arch, 512 
 
 Screens, 179, 183, 200, 226, 227 ; construc- 
 tional value, 381 
 
 Scrolls of Leafage, 42, 429-432 
 
 Sea Transport, 20 
 
 Seats, fixed, 447 
 
 Secular Canons, churches of, 15 ; archi- 
 tecture, 44 
 
 Segmental Arch, 259, 262, 267 ; pointed, 
 261 
 
 Semicircular Arch, 259, 260 ; retention in 
 Gothic, 261, 265, 322 ; intersecting, 263 
 
 Semidome, 163, 283 
 
 Semi-Norman period, 43 
 
 Set-off, 358 
 
 Severy, definition, 317 ; square and oblong 
 severies, 309 ; various, 295 
 
 Sexpartite Vaulting, 320 
 
 SHAFTS, definition, 233 ; plan, 256 ; pear- 
 shaped, 256; pointed, 258 ; ornamented, 
 256 
 
 Sharpe's Seven Periods, 65 
 
 Shields, 94 
 
 Shingles, 41, 389 
 
 Shouldered Arch, 260 
 
 Shrine, 183, 187, 1S9, 193 
 
 Skeleton Construction, 55, 121, 134, 
 47 1 , 49° 
 
 Skeleton Vaulting, 340 
 
 Skewback, 259 
 
 Sky-line, loss of, 541 
 
 Slates, 389 
 
 Small-stone construction, 258 
 
 Snakes, interlacing, 41 
 
 Snow, 389 
 
 Soffit, 259 ; soffit cusp, 509 
 
 Sole-piece, 559, 564 
 
 Sound Holes, 517, 604 
 
 South Side, sacred, 584 
 
 South-east School of English Gothic, 
 108, 246 
 
 Southern France, domes and barrel 
 vaults, 13, 264 
 
 Spain, Romanesque Domes, 591 : roofing 
 of vaults, 388 
 
 Span, 53 ; span roof, 570; span roof of stone, 
 287 
 
 Spandrel, 259 
 
 Speed of building, 49 
 
 Spire, three types, 625 ; classification, 619- 
 625 ; early examples, 617 ; without 
 parapets, 619, 620 ; with parapets, 621- 
 625 ; broach, 620 ; timber, 619 ; pin- 
 nacles, 623, 624 ; flying buttresses, 624, 
 
 625, 377 ; windows, 627 ; crockets, 625, 
 
 626, 628, 632 ; bands, 626 ; strings, 626 ; 
 rolls, 625 ; arris, 625 ; plate tracery, 626 ; 
 bar tracery, 628 ; plan, 628 ; construc- 
 tion, 636 ; thickness, 636 ; squinches, 
 636 ; jointing, 637 ; entasis, 632 ; cap- 
 stone, 636 ; artistic value, 616 ; spire 
 design, 625, 629, 631, 632, 633 
 
 Splay, 512 
 
AND GLOSSARY 
 
 -s, 
 
 Spout, 401, 503 
 
 Springer, 259, 302, 303 
 
 Spur, 418, 441., 455 
 
 Squinch, 636 
 
 Si Andrew's Cross, 381 
 
 St William < n I )ijon, 317 
 
 Stability, Gothic, how secured, 54-58 
 
 Stained Glass, 61, 134, 141, 487,488,491, 
 
 495, 496, 497, 5°4 
 Stalked Capitals, 431 
 Stalls, position of, 184, 185 
 Star, 40 
 
 Stiff-leaved Capitals, 430 
 Stilted Arch, 259, 261 
 STONE, improper use, 258 
 Stone Construction, in wood, 020 
 Stone Roofs, 283, 285, 287 
 Stop, 407 
 
 Storied Capi i als, 416 
 Stoup, 207 
 Straight-sided Hood-mold, 579 
 
 Straining Arches, 
 
 381 
 
 Strawberry leaf, 94 
 
 String-course, 404-406. 38, 606 
 
 Strut, 559, 566 ; horizontal. 569 
 
 Si ud, 41 
 
 Sunday Procession, 170 
 
 Supermullions, 494, 499 1 
 
 Survivals, 24, 42, 43. 502, 596— See " Re- 
 trogressive " 
 
 Symbolism, 616, 42 
 
 SYRIA, stone ceilings, 285 ; buttresses, 351 ; 
 towers, 586 
 
 Tablinum, 201 
 
 Tapestry, 541 
 
 Tas-de-Charge, 259, 302 
 
 Temple, Jewish, 221 
 
 Temples converted into churches, 145 
 
 Tension, king post in. 561 
 
 Thai, h, y>2, 599 
 
 Thirteenth-Century work, 65-79, 105-126 
 
 Thki.l-i i.ntrkii Arch, 259 
 
 Thri iat, 403 
 
 Thrusts of Vaulting, 9, 10, 29, 30, 54-57, 
 345- 35°, 35', 3°8, 376; opposition of 
 thrusts, 381-383 ; transmission of thrusts, 
 379-380 : thrusts of pier arcade, 381 
 
 Tie-beam, 561 
 
 i ii i' i, rons, 336 
 
 4 11 1 K' >ni 5, 589, 559 
 
 Timber Churi hes, 218, 219 
 
 Tooth Oknamen i, 77 
 
 Tow 1 r . 586-610 : Romanesque, 
 
 plan, 600; wooden, 598; central lan- 
 tern, '103, 24; flanking choir, 204, 599; 
 central and one western, 599; groups, 
 600; detached, 598; construction, 593, 
 601 ; roofs, '102 ; saddleback, 602 ; 
 Somerset, 604 ; batter, '105; buttresses, 
 605, 606; strings, 606; pinnacles, 606 ; 
 fenestration and voids, 603, 004 ; pro- 
 portions, 602 : distribution of orna- 
 ment, 608 
 
 Trabe \iiii Construction, 258 
 
 Tracery, Window, 405 seg.\ functions, 
 505 ; planes, ^,0--, ; bars, 516 
 
 Transept, antiquity and origin, 195, 196, 
 16 ; length, 196 : omitted, 593 ; external 
 elevations, 29, 68 ; towers, 593 ; piers 
 thrust inwardly, 381 ; parochial transept, 
 222; eastern transept, 187, 189: western 
 transept, 203; central transept, apses, 
 197 ; aisles, 16, 197; north and south 
 aisles, 16, 198 
 
 Transitional period of English archi- 
 tecture, 43, 100-105 : volute capital, 
 410 : base. 451 
 
 Translation hi Relics, 193 
 
 Transom, 494, 499, 503, 504 
 
 Transport, 20, 23, 250, 520, 561, 619 
 
 Transverse Rir.s. Sin 11 d, 511 ; omitted, 
 292 
 
 Transverse Arch, Intermediate, 320 
 
 Trefoilkh Arch, 260, 268 
 
 Tribune, 527, 528 
 
 Trifoliated Arch, 260, 580 
 
 Triforium, 519-542: etymology, 519: 
 Romanesque, 26, 27 ; constructional 
 value, 57, 58 : triforium without windows, 
 552 ; triforium with windows, 531, 517 ; 
 transparent, 537; with front wall, 534; 
 with panelled front wall, 88 ; with un- 
 divided arch, 532 ; with subdivided arch, 
 
 533 ; various treatments of arcade, 533, 
 
 534 : absorption into pier arcade. 535 ; 
 
 absorption into clerestory, 
 
 sham 
 
 triforium, 519. 527; height, 529: why 
 tall, 530; roofing of, 540; use of, 541 ; 
 transverse arches in. 369 
 
 Tripartite plan of church, 220, 221, 19 
 
 Tripartite Vaulting, 339 
 
 Triple! s in Shafts, 245 
 
 Trumeau, 577, 578 
 
 TRUSS, 559 ; trussed rafter roof. 559 
 
 TUDOR architecture, 136, 141, 142, 474; 
 
 arch, 259; flower, 95 ; rose. 94 
 I'll a, 303, 304 
 
782 
 
 INDEX OF SUBJECT MATTER. 
 
 Tushes, 366 
 
 Two-story Interior, 538 
 
 Tympanum, 576 ; carved, 576; glazed, 577 
 
 U 
 
 Undercutting, 279, 435 
 Undulatory Foliage, 436 
 Uniformity of design, 45, 49 
 Unity in late Gothic design, 91 
 
 V 
 
 Vaults, primary element in Romanesque 
 and Gothic, 9; Romanesque systems, 
 13 ; Romanesque, 19 ; Viollet-le-Duc on 
 Peterborough vaults, 440 ; fall of high 
 vaults, 30 ; earliest Gothic vaults, 105, 
 106, 10S; thirteenth-century vaults, 74, 
 121 ; fourteenth-century, 82 ; fifteenth- 
 century, 93 ; pointed, 103 ; in parish 
 churches, 82; wooden, 570; omitted, 
 550; material, 303 ; rubble, 304 ; filling- 
 in, methods of, 323 ; thickness, 304; 
 lightened, 60 ; contorted, 299 ; span, 291 ; 
 height of spring, 306 ; aisle vault, thrust, 
 381; value, 29; high vault, value, 30; 
 curve of ribs at spring, 343 ; roofs over 
 vaults, 3S8, 550 
 
 Vault-pier, 56, 57 
 
 Vaulting Shaft, 54, 239, 240, 241 
 
 Verticality of Gothic, 61, 62 
 
 Vesica Piscis, 517 
 
 Vestibule, 203 
 
 Vestry — See " Sacristy " 
 
 Village Churches, excellence of work, 118 
 
 Vine Leaf, 94 
 
 Voids v. Solids, 26, 56-61, 604 
 
 Volute, Transitional, 416 ; volute capitals, 
 425-429 
 
 VOUSSOIR, 257, 301 
 VOUTAIN, 323 
 
 W 
 
 WALLS, thickness of, 24, 374 ; protection 
 
 from rain, 402 
 Wall Arch, converted into pier arch, 56, 
 
 57,58 
 
 Wall Passage, 547 
 
 Wall Plate, 561 
 
 Wall Post, 563, 564 
 
 Wall Rib, 300 
 
 Watching Loft, 187 
 
 Water Table, 406 
 
 Water Transport, 20, 250, 619 
 
 Waterholding Base, 449 
 
 Waterleaf, 426, 416 
 
 Weather Mold, 277, 406 
 
 Weathercock, 636 
 
 Weatherings, 389, 559 
 
 West Doorway, 581, 91, 205, 582 
 
 West Front, Romanesque, 28 ; in thirteenth 
 century, 71 ; in fourteenth century, 82 ; 
 in fifteenth century, 91 ; various methods 
 of design, 71, 72, 73, 91 
 
 West of England School of Roman- 
 esque, 36, 100, 164, 242, 276 
 
 West of England School of Gothic, 
 89, 105, 116, 134, 245, 254, 433, 477, 478, 
 547 
 
 West Porch, 202, 204 
 
 West Towers, 595-598 ; position, 596 ; 
 omission, 598 ; flanking, 599 ; origin, 
 596; artistic value, 71; constructional 
 value, 72, 381, 598 
 
 Wheel Window, 517 
 
 Wilfrid, Bishop, 155 
 
 Wind-beam, 560 
 
 Windows, 456 seg. ; Romanesque, 38, 456- 
 462 ; balustered, 456 ; midshaft, 456 ; 
 diminutive, 459 ; set high, 459 ; unglazed, 
 459 ; window-frames, 459 ; lancet, 462 ; 
 circular, 516 ; construction, 505 seg. ; 
 rear-arch, 456, 512, 513; inner or outer 
 arcade, 513 ; in towers, 603, 604 ; in 
 ringing chamber, 517; in spires, 627, 
 628 ; above chancel arch, 547 ; of odd 
 lights, 475, 479 ; window tracery, 465 ; 
 jointing, 506 ; chamfered, 508 ; setting 
 out of, 496, 508 ; tracery ineffective, 
 497; alternation of, 490 
 Wood Construction in stone, 622 
 Wooden Bell-tower, 598 
 
 Zigzag or chevron, 40 
 
 Printed at The Darien Press, Edinburgh. 
 
«to#2j. 
 
 t 
 
 
 V__ 
 
 GETTY CENTER LIBRARY 
 
 3 3125 00952 7082 
 
■ > -- ,; ■ . ■ . ■ 
 
 ''■••■'•■•;■.>• •'•;.■<:•• 
 
 ■'■''''•■:■ ' ' .... 
 
 ■ ;'■ V •■ , .', ■ i ' '■, 
 
 HBRL 
 
 '•'■'■',' 
 
 ■ Bl 
 
 HI 
 
 1 
 
 Hi 
 
 WmHm 
 
 HE 
 
 Hi 
 
 HHBL